Heart Development and Regeneration (2nd Edition)

Heart Development and Regeneration Volume I

Heart Development and Regeneration Volume I

Edited by

Nadia Rosenthal and

Richard P. Harvey

AMSTERDAM • BOSTON • HEIDELBERG • LONDON • NEW YORK • OXFORD • PARIS

SAN DIEGO • SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO

Academic Press is an imprint of Elsevier

Heart Plate 1 was inspired by the artistry of Franco Mari, a living master of Majolica ceramic decoration working from Deruta, Italy. Each design band repre­ sents a stage in heart development and a theme in the volume. Motifs from center outward: loop, tube, surface, pacer, cushion, crest. Artwork by Nadia Rosenthal.

Heart Plate 2 was inspired by the artistry of Franco Mari, a living master of Majolica ceramic decoration working from Deruta, Italy. Each design band repre­ sents a theme in the volume. Motifs from center outward: circuit, discovery, access, dynamic, progenitor, renewal. Artwork by Nadia Rosenthal.

Academic Press is an imprint of Elsevier 32 Jamestown Road, London NW1 7BY, UK 30 Corporate Drive, Suite 400, Burlington, MA 01803, USA 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA First edition 2010 Copyright © 2010 Elsevier Inc. All rights reserved. Illustrations for Chapter 3.4 (volume I) copyright 2010 Robert H. Anderson 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: ­[email protected]. Alternatively, visit the Science and Technology Books website at www.elsevierdirect.com/rights for further information 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. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress ISBN: 978-0-12-381332-9 (Set) ISBN: 978-0-12-372490-8 (Volume I) ISBN: 978-0-12-381334-3 (Volume II)

For information on all Academic Press publications visit our website at elsevierdirect.com

Typeset by MPS Limited, a Macmillan Company, Chennai, India www.macmillansolutions.com Printed and bound in China 10  11  12  13  14  15  10  9  8  7  6  5  4  3  2  1

 “Come potreste descrivere a parole questo cuore,   senza riempire un libro intero?” (How could you describe this heart in words,   without filling a whole book?) Note written by Leonardo da Vinci beside an anatomical drawing of the heart, 1513.

List of Contributors

Number in parentheses indicates the chapter to which the author contributed. Jeffrey D. Amack (4.1), Department of Cell and Developmental Biology, SUNY Upstate Medical University, Syracuse, NY 13210 Brad A. Amendt (4.3), Institute of Biosciences and Technology, Texas A&M System Health Science Center, 2121 Holcombe Blvd, Houston, TX 77030 Robert H. Anderson (3.2, 3.4), Cardiac Unit, Institute of Child Health, University College, London, UK

Nigel A. Brown (3.4, 4.3), Division of Basic Medical Sciences, St George’s University of London, Cranmer Terrace, London SW17 0RE, UK Benoit G. Bruneau (10.1), Gladstone Institute of Cardiovascular Disease and Department of Pediatrics, University of California, San Francisco, CA 94158 Margaret E. Buckingham (3.1), Department of Developmental Biology, CNRS URA2578, Pasteur Institute, 25 rue du Dr Roux, Paris, France

Piero Anversa (14.3), Cardiovascular Research Institute, Departments of Medicine, New York Medical College, Valhalla, NY 10595

Todd D. Camenisch (6.1), Departments of Pharmacology and Toxicology, Molecular and Cellular Biology, Steele Children’s Research Center, Bio5 Institute, University of Arizona, Tucson, AZ

Ana Paula Azambuja (1.1), Laboratório de Genética e Cardiologia Molecular, Instituto do Coração, Faculdade de Medicina, Universidade de São Paulo, São PauloSP, Brazil; and Departamento de Biologia Celular e do Desenvolvimento, Instituto de Ciências Biomédicas, Universidade de São Paulo, São Paulo-SP, Brazil

Hozana Andrade Castillo (1.1), Laboratório de Genética e Cardiologia Molecular, Instituto do Coração, Faculdade de Medicina, Universidade de São Paulo, São PauloSP, Brazil; and Departamento de Biologia Celular e do Desenvolvimento, Instituto de Ciências Biomédicas, Universidade de São Paulo, São Paulo-SP, Brazil

David Bader (1.5), Stahlman Cardiovascular Research Laboratories, Program for Developmental Biology, and Department of Medicine, Vanderbilt University Medical Center, Nashville, TN 37232-2561

Rodrigo Abe Castro (1.1), Laboratório de Genética e Cardiologia Molecular, Instituto do Coração, Faculdade de Medicina, Universidade de São Paulo, São PauloSP, Brazil; and Departamento de Biologia Celular e do Desenvolvimento, Instituto de Ciências Biomédicas, Universidade de São Paulo, São Paulo-SP, Brazil

Shoumo Bhattacharya (11.3), Wellcome Trust Centre for Human Genetics, University of Oxford, UK Brian L. Black (9.5), Cardiovascular Research Institute and Department of Biochemistry and Biophysics, University of California, San Francisco, CA 94158-2517 Daniel G. Blackmore (13.1), Queensland Brain Institute, The University of Queensland, Brisbane, Queensland, 4072, Australia Rolf Bodmer (1.2), Development and Aging Program, NASCR Center, Burnham Institute for Medical Research, 10901 North Torrey Pines Rd, La Jolla, CA 92037 Thomas Brand (5.1), Heart Science Centre, National Heart and Lung Institute, Imperial College London, UK Marianne Bronner-Fraser (7.1), California Institute of Technology, Pasadena, CA 91125

Bishwanath Chatterjee (11.2), Laboratory of Developmental Biology, National Heart Lung Blood Institute, National Institutes of Health, Bethesda, MD 20892 Vincent M. Christoffels (2.3, 3.2), Heart Failure Research Center, Academic Medical Center, Amsterdam, The Netherlands Ondine Cleaver (8.2), Dept. of Molecular Biology NA8.300, University of Texas Southwestern Medical Center, 5323 Harry Hines Blvd, Dallas, TX 75390 Frank L. Conlon (9.4), Department of Genetics, Fordham Hall, University of North Carolina-Chapel Hill, Chapel Hill, NC 27599 Kimberly R. Cordes (10.3), Gladstone Institute of Cardiovascular Disease and Departments of Pediatrics, Biochemistry and Biophysics, University of California, xxv

xxvi

List of Contributors

San Francisco, 1650 Owens Street, San Francisco, CA 94158 David M. Cox (11.5), MDS Sciex 71 Four Valley Drive, Concord, Ontario, Canada L4K 4V8 Richard M. Cripps (9.5), Department of Biology, University of New Mexico, Albuquerque, NM 87131-0001 Brad Davidson (1.1), Department of Molecular and Cell Biology, Division of Genetics and Development, Center for Integrative Genomics, University of California, Berkeley, CA 94720 José Luis de la Pompa (6.2), Departamento de Biología del Desarrollo Cardiovascular, Centro Nacional de Investigaciones Cardiovasculares (CNIC), Melchor Fernández Almagro 3, E-28029 Madrid, Spain Stefanie Dimmeler (14.2) Professor of Medicine, Molecular Cardiology, Department of Internal Medicine III, University of Frankfurt, Theodor-SternKai 7, 60590 Frankfurt, Germany Pascal Dollé (3.3), Institut de Génétique et de Biologie Moléculaire et Cellulaire (IGBMC), BP 10142, Illkirch, F-67400 France; Inserm, U 964, Illkirch, F67400 France; CNRS, UMR 7104, Illkirch, F-67400 France; Université de Strasbourg, Faculté de Médecine, Strasbourg, F-67000 France Min Du (11.5), Department of Biology, York University, Toronto, Ontario, Canada, M3J 1P3 Victor J. Dzau (14.1), Molecular and Genomic Vascular Biology, Duke University Medical Center, DUMC 3701, Durham, NC 27710 David A. Elliott (9.1), Monash Immunology and Stem Cell Laboratories, Monash University, Clayton, Victoria, 3800, Australia Sylvia M. Evans (2.2), Skaggs School of Pharmacy and Pharmaceutical Sciences, and Department of Medicine, University of San Diego, La Jolla, CA Max Ezin (7.1), California Institute for Technology, Pasadena, CA 91125 Ann C. Foley (1.3), Divison of Cardiology, Weill Cornell Medical College, New York, NY 10065 Richard Francis (11.2), Laboratory of Developmental Biology, National Heart Lung Blood Institute, National Institutes of Health, Bethesda, MD 20892 Manfred Frasch (1.2), Department of Biology, Developmental Biology Unit, University of ErlangenNürnberg, Erlangen, Staudtstr. 5, 91058, Germany Daniel J. Garry (11.1), Lillehei Heart Institute, and Department of Internal Medicine, University of Minnesota, Minneapolis, MN 55455

Lior Gepstein (13.2), The Sohnis Family Research Laboratory for the Regeneration of Functional Myocardium, Department of Biophysics and Physiology, the Bruce Rappaport Faculty of Medicine, Technion-Israel Institute of Technology, Haifa, Israel Massimiliano Gnecchi (14.1), Department of Cardiology, Fondazione I.R.C.C.S. Policlinico San Matteo, University of Pavia, Viale Golgi 19, 27100, Pavia, Italy Joseph Gold (13.2), Geron Corporation, Menlo Park, CA 94025 Joaquim Grego-Bessa (6.2), Sloan-Kettering Institute, 1275 York Ave., New York, NY 10065 Thomas Gridley (8.3), The Jackson Laboratory, Bar Harbor, ME Rosa M. Guzzo (1.3), Department of Orthopedic Surgery, University of Connecticut Health Center, Farmington, CT 06034 Hiroshi Hamada (4.2), Developmental Genetics Group, Graduate School for Frontier Biosciences, Osaka University; and CREST, Japan Science and Technology Corporation (JST), 1-3 Yamada-oka, Suita, Osaka 5650871, Japan Natasha L. Harvey (8.4), Division of Haematology, The Centre for Cancer Biology, SA Pathology, Adelaide, South Australia, Australia Richard P. Harvey (9.1), Victor Chang Cardiac Research Institute, Sydney, NSW, 2010, Australia and Faculties of Science and Medicine, University of New South Wales, Kensington, 2053, Australia Steve Hauschka (11.4), Department of Biochemistry, University of Washington, Seattle, Washington, 98195 Charis Himeda (11.4), Department of Biochemistry, University of Washington, Seattle, Washington, 98195 Willem M. H. Hoogaars (2.3), Heart Failure Research Center, Academic Medical Center, Amsterdam, The Netherlands Guo-Ying Huang (11.2), Pediatric Heart Center, Children’s Hospital of Fudan University, Shanghai, China Mary R. Hutson (7.2), Department of Pediatrics (Neonatology), Neonatal-Perinatal Research Institute, Jones Building, Room 401, Box 103105, Duke University Medical Center, Durham, NC 27710 Kimberly E. Inman (7.1), Stowers Institute for Medical Research, 1000 E. 50th Street, Kansas City, MO 64110 Silviu Itescu (14.2), Director of Transplantation Immunology, Columbia University Medical Center, 630 West 168th Street, New York, NY 10032; Columbia University College of Physicians and Surgeons, and Professor of Medicine, University of Melbourne, Melbourne, Australia

List of Contributors

xxvii

Jan Kajstura (14.3), Cardiovascular Research Institute, Departments of Medicine, New York Medical College, Valhalla, NY 10595

Sigolène M. Meilhac (3.1), Department of Developmental Biology, CNRS URA2578, Pasteur Institute, 25 rue du Dr Roux, Paris, France

Robert G. Kelly (2.2), Developmental Biology Institute of Marseilles-Luminy, UMR 6216 CNRS-Université de la Méditerranée, Campus de Luminy, Marseilles, France

Mark Mercola (1.3), Burnham Institute for Medical Research, 10901 N.Torrey Pines Road, La Jolla, CA 92037

Margaret L. Kirby (7.2), Department of Pediatrics (Neonatology), Neonatal-Perinatal Research Institute, Jones Building, Room 401, Box 103105, Duke University Medical Center, Durham, NC 27710 Edwin P. Kirk (9.1), Victor Chang Cardiac Research Institute, Sydney, NSW, 2010, Australia; Department of Medical Genetics, Sydney Children’s Hospital and School of Women’s and Children’s Health, Faculty of Medicine, University of New South Wales, Australia Paul A. Krieg (8.2), Dept. of Cell Biology and Anatomy, University of Arizona College of Medicine, Molecular Cardiovascular Research Program, Medical Research Building MRB 311, 1656 E. Mabel St, Tucson, AZ 85724 Enrique Lara-Pezzi (14.4), Heart Science Centre, Imperial College London, Harefield, UK; Centro Nacional de Investigaciones Cardiovasculares (CNIC), Madrid, Spain Shuaib Latif (11.1), Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, TX 75390 Kory J. Lavine (5.2), Department of Developmental Biology, Washington University School of Medicine, St. Louis, MO Linda Leatherbury (11.2), Laboratory of Developmental Biology, National Heart Lung Blood Institute, National Institutes of Health, Bethesda, MD 20892 Annarosa Leri (14.3), Cardiovascular Research Institute, Departments of Medicine, New York Medical College, Valhalla, NY 10595 Michael Levine (2.1), Department Molecular and Cell Biology, Division of Genetics, Genomics and Development, Center for Integrative Genomics, University of California, Berkeley, CA 94720 Cecilia W. Lo (11.2), Laboratory of Developmental Biology, National Heart Lung Blood Institute, National Institutes of Health, Bethesda, MD 20892 Roger R. Markwald (6.1), Department of Cell Biology and Anatomy & Developmental Cardiovascular Biology Center, Medical University of South Carolina, SC James F. Martin (4.3), Institute of Biosciences and Technology, Texas A&M System Health Science Center, 2121 Holcombe Blvd, Houston, TX 77030 John C. McDermott (11.5), Department of Biology, York University, Toronto, Ontario, Canada, M3J 1P3

Takashi Mikawa (5.1), Cardiovascular Research Institute, University of California, San Francisco, CA Timothy Mohun (11.3), Division of Developmental Biology, MRC National Institute for Medical Research, London, UK Antoon F.M. Moorman (2.3, 3.2, 3.4), Heart Failure Research Center, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands Katherine Moynihan (1.5), Stahlman Cardiovascular Research Laboratories, Program for Developmental Biology, and Department of Medicine, Vanderbilt University Medical Center, Nashville, TN 37232-2561 Ramón Muñoz-Chápuli (8.1), Departamento de Biología Animal, Facultad de Ciencias, Universidad de Málaga, Campus de Teatinos s/n, E-29071 Málaga, Spain Charles E. Murry (13.2), Department of Pathology, Center for Cardiovascular Biology, Institute for Stem Cell and Regenerative Medicine, University of Washington, Seattle, WA 98109 Georges Nemer (9.2), Research Unit in Cardiac Growth and Differentiation, Institut de recherches cliniques de Montréal (IRCM), Montréal, QC H2W 1R7; Canada and Department of Biochemistry, American University of Beirut, Lebanon Mona Nemer (9.2), University of Ottawa, Laboratory of Cardiac Growth and Differentiation, Department of Biochemistry, Microbiology and Immunology, Ottawa, Ontario, K1N 6N5, Canada Karen Niederreither (3.3), Departments of Medicine and Molecular and Cellular Biology, Center for Cardiovascular Development, Baylor College of Medicine, One Baylor Plaza, Houston, TX 77030 Eric N. Olson (10.2), Department of Molecular Biology, University of Texas Southwestern Medical Center at Dallas, Dallas, TX 75390 David M. Ornitz (5.2), Department of Developmental Biology, Washington University School of Medicine, St. Louis, MO Lil Pabon (13.2), Department of Pathology, Center for Cardiovascular Biology, Institute for Stem Cell and Regenerative Medicine, University of Washington, Seattle, WA 98109

xxviii

José M. Pérez-Pomares (6.2, 8.1), Departamento de Biología Animal, Facultad de Ciencias, Universidad de Málaga, Campus de Teatinos s/n, E-29071 Málaga, Spain Kenneth Poss (12.2), Department of Cell Biology, Howard Hughes Medical Institute, Duke University Medical Center, Durham, NC 27710 Rodney L. Rietze (13.1), Pfizer Regenerative Medicine, Granta Park, Great Abington, Cambridge, UK; Queensland Brain Institute, The University of Queensland, Brisbane, Queensland, 4072, Australia Nadia Rosenthal (14.4), Heart Science Centre, Imperial College London, Harefield, UK, Mouse Biology Unit, European Molecular Biology Laboratory (EMBL), Monterotondo, Italy; and Australian Regenerative Medicine Institute, Monash University, Melbourne, Australia Raymond B. Runyan (6.1), Department of Cell Biology and Anatomy, University of Arizona, Tucson, AZ Allysson Coelho Sampaio (1.1), Laboratório de Genética e Cardiologia Molecular, Instituto do Coração, Faculdade de Medicina, Universidade de São Paulo, São PauloSP, Brazil; and Departamento de Biologia Celular e do Desenvolvimento, Instituto de Ciências Biomédicas, Universidade de São Paulo, São Paulo-SP, Brazil Daniel Schaft (9.1), Victor Chang Cardiac Research Institute, Sydney, NSW 2010, Australia Robert J. Schwartz (9.3), The Institute of Biosciences and Technology, The Texas A&M University System Health Science Center, Houston, TX 77030 Ian C. Scott (1.4), The Hospital for Sick Children, Program in Developmental and Stem Cell Biology, Department of Molecular Genetics, University of Toronto, 101 College Street, East Tower Rm 11-307, Toronto, ON M5G 1L7, Canada Yuan Shen (11.2), Laboratory of Developmental Biology, National Heart Lung Blood Institute, National Institutes of Health, Bethesda, MD 20816 Marcos Sawada Simoes-Costa (1.1), Laboratório de Genética e Cardiologia Molecular, Instituto do Coração, Faculdade de Medicina, Universidade de São Paulo, São Paulo-SP, Brazil; and Departamento de Biologia Celular e do Desenvolvimento, Instituto de Ciências Biomédicas, Universidade de São Paulo, São Paulo-SP, Brazil Jonathan M. W. Slack (12.1), Stem Cell Institute, University of Minnesota, McGuire Translational Research Facility, 2001 6th Street SE, Minneapolis, MN 55455 Deepak Srivastava (10.3), Gladstone Institute of Cardiovascular Disease and Departments of Pediatrics

List of Contributors

and Biochemistry & Biophysics, University of California, San Francisco, 1650 Owens Street, San Francisco, CA 94158 Frank Stockdale (1.5), Department of Medicine, Stanford University, School of Medicine, Stanford, CA 94305-5151 Paul A. Trainor (7.1), Stowers Institute for Medical Research, 1000 E. 50th Street, Kansas City, MO 64110; and Kansas University School of Medicine, 3901 Rainbow Boulevard, Kansas City, KS 66160 Gert van den Berg (3.2), Heart Failure Research Center, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands Sandra Webb (3.4), Department of Basic Sciences, Anatomy and Developmental Biology, St George’s University of London, Cranmer Terrace, London, UK Wolfgang Weninger (11.3), Integrative Morphology Group, Center for Anatomy & Cell Biology, Medical University Vienna, Austria Andy Wessels (11.2), Department of Anatomy and Cell Biology, Medical University of South Carolina, Charleston, SC 29425 José Xavier-Neto (1.1), Laboratório de Genética e Cardiologia Molecular, Instituto do Coração, Faculdade de Medicina, Universidade de São Paulo, São PauloSP, Brazil Deborah Yelon (1.4), Skirball Institute of Biomolecular Medicine, Developmental Genetics Program, New York University School of Medicine, Department of Cell Biology, 540 First Avenue, 4th Floor, Lab 15, New York, NY 10016 H. Joseph Yost (4.1), Department of Neurobiology and Anatomy, University of Utah School of Medicine, Salt Lake City, UT 84112 Bryan D. Young (10.2), Department of Molecular Biology, University of Texas Southwestern Medical Center at Dallas, Dallas, TX 75390 Qing Yu (11.2), Laboratory of Developmental Biology, National Heart Lung Blood Institute, National Institutes of Health, Bethesda, MD 20892 Katherine E. Yutzey (9.4), Division of Molecular Cardiovascular Biology, Cincinnati Children’s Hospital Medical Center, 240 Albert Sabin Way, ML7020, Cincinnati, OH 45229 Zhen Zhang (11.2), Laboratory of Developmental Biology, National Heart Lung Blood Institute, National Institutes of Health, Bethesda, MD 20892 Xiao-Qing Zhao (11.2), Laboratory of Developmental Biology, National Heart Lung Blood Institute, National Institutes of Health, Bethesda, MD 20892

Foreword

Developmental biology is slowly but surely moving to center stage in cardiovascular science and practice as a result of the increasing realization of its potential for understanding the basic mechanisms of a wide variety of congenital and acquired heart diseases. This stems from the fact that many of the factors which are operational during morphogenesis of the heart during fetal life continue to influence growth and adaptation during postnatal life. In addition, the developmental origins and lineages of different cells influence their behavior in adult life. Defining these factors offers exciting possibilities for developing novel therapies. The recent sustained progress in the field has been driven by a handful of enthusiastic individuals around the world who have had the foresight to see the potential of the subspecialty, coupled with the availability of sophisticated experimental approaches involving genetic, molecular, imaging and engineering tools, which are being used alone or in combination.

Heart Development and Regeneration represents a unique achievement in bringing together leaders in the field presenting current knowledge in a lucid form, making it accessible to a wide range of readers. The inclusion of several chapters on regenerative therapy, arguably the hottest topic in medicine today, represents a natural extension to the topics of developmental biology. Surgery is often likened to plumbing. There is no selfrespecting plumber who would tackle a complex plumbing system in a building without demanding to see the blueprint of how this was put together in the first place! I am confident that this book will be of great value and offer unique opportunities to both clinicians and researchers in almost every realm of cardiovascular science, medicine and surgery. Professor Sir Magdi Yacoub FRS Harefield Heart Science Centre Imperial College London

xxix

Preface

It has now been a decade since our book Heart Development was published, and the field has again exploded. A more integrated picture of development, physiology and disease is emerging from our deeper investigations of the genes, proteins and cells, and the functional interactions between them, which build the cardiovascular system. This picture has been immeasurably enhanced by our explorations across the evolutionary landscape. Ten years ago, only a handful of cardiovascular malformations, the most common type of birth defect, could be associated with causative factors, but recent studies have uncovered the genetic basis of many more, and with the advent of forward mutagenic screens in fish and mice, genome-wide association studies in humans, and next generation sequencing, these investigations are now set on an entirely new course, with implications for genetic testing, counseling, long-term patient follow up and choice of intervention. Much of the progress has been technical: advances in genomics, proteomics, transgenesis and imaging have supplied more exacting tools to dissect and manipulate the embryo and the adult. Furthermore, the amount of biological data generated over the past decade has overwhelmed our ability to understand it without powerful computational approaches. It is logical, then, that these approaches have made it possible to explore cardiac functionality with quantitative models of cells, tissues and organs. By far the most important development over the last decade has been in stem cell biology, yielding access to early processes in the embryo determining cell lineage, as mimicked in the differentiation of embryonic stem cells into cardiomyocytes, as well as providing a new understanding of adult lineages heretofore unimagined, ushering in exciting prospects for their use in regenerative medicine. Additional important advances have been more conceptual, reflecting the shifting emphasis of the field towards integrative systems biology and the important role that noncoding RNAs and epigenetics play in the cardiac gene regulatory network and cardiovascular disease. Ten years ago, issues of lineage potency, stemness, robustness and meta-stable chromatin states were the purview of a small group of purists. Now they are fast becoming our bread and butter. In the world of systems biology, individual regulatory genes are seen as part of an information

machine encoded in the genome, rather than individual enzymes working in an isolated pathway. The search for logic in regulatory circuitry is now well and truly focused at this level. The dominance of genetic determination and reductionism has given way to an appreciation of whole cell integration of activity, informing our approach to the study of cardiac growth and morphogenesis, and causality in disease. Epigenetic regulatory mechanisms assume a more prominent position in our models and the relationship between genome and phenotype grows more intimate. In the developing heart, as in other biological systems, it is becoming increasingly difficult to separate genetic instructions from the process of carrying them out, to distinguish plan from execution. Heart Development and Regeneration, published in two volumes, attempts to capture a collective snapshot of this fast moving field. Volume I covers the early stages of cardiovascular determination, growth and morphogenesis across the phylogenetic tree, whereas Volume II reviews recent advances in transcriptional and post-transcriptional regulation, epigenetic circuits, systems analysis, the theory and evolution of stem cells, and the molecular and cellular basis of cardiac repair. The chapters are written by the world’s experts and provide up-to-date reports from their laboratories, while treating the newcomer to a rich grounding in classical developmental biology of the cardiovascular system. These books together represent a comprehensive catalog of the individual parts of the whole as we currently know them. They relay current progress on compiling the instructions for putting the parts together in the context of how form and function are generated in the cardiovascular system to sustain life, and how these processes go awry in disease. The stories told in these chapters are full of models; how we study the biology of the developing and regenerating heart is model-dependent, and therefore our models must remain flexible and responsive to the influx of new data. Integrative models are nowhere more relevant than in the cardiovascular system, where function substantially instructs form. Among all the complex organs, the developing heart is exceptionally susceptible to malformation; human congenital heart defects affect an astonishing 0.7% of live births. Although many defects have genetic underpinnings, very few are attributable to a single gene xxxi

xxxii

or protein malfunction. Their origins will in many cases be subtle, context-dependent and grounded in the architecture of the regulatory network, with the environment playing a major role. The results from modeling the feedback loops between physiological function and gene expression may in the beginning seem counterintuitive, yet as we understand these at a systems level, we will gain vital clues to the development of preventative therapeutic approaches or clinical interventions. As researchers and editors, we have prepared these volumes as a celebration of the beauty and complexity of heart development, its evolution and recapitulation in the context of regeneration. Working closely with our colleagues has been a rewarding experience and we thank them all for their generous donations of time and expertise, for the care and pride in their work, and for their patience in helping us keep these books current through several updates and refinements over the past two years. We thank

Preface

our editor­ial assistants Lars Bochmann, Catarina Catela, Sally Dunwoodie, Marianne Hede, Danielle de Jong, Jason Kovacic, Paschalis Kratsios, Enrique Lara-Pezzi, Tommaso Nastasi, Ekaterina Salimova, Maria Paola Santini, Katia Semenova, Duncan Sparrow, Lieve Temmerman, Pascal te Welscher and David Winlaw for their careful comments on the chapter drafts. We are immensely grateful to our artist Kathy Stern, who has once again surpassed herself in the preparation of beautiful and informative illustrations. This book would not be in your hands without the attention and diligence of our publishers at Elsevier, April Graham, Caroline Johnson and Janice Audet, who have worked with us closely to complete this daunting project. Nadia Rosenthal Richard Harvey Sydney, August 2009

   Chapter 1.1

Evolutionary Origins of Hearts José Xavier-Neto1, Brad Davidson2, Marcos Sawada Simoes-Costa1,3, Rodrigo Abe Castro1,3, Hozana Andrade Castillo1,3, Allysson Coelho Sampaio1,3 and Ana Paula Azambuja1,3 1

Laboratório de Genética e Cardiologia Molecular, Instituto do Coração, Faculdade de Medicina, Universidade de São Paulo, São Paulo-SP, Brazil  2 Department of Molecular and Cell Biology, Division of Genetics and Development, Center for Integrative Genomics, University of California, Berkeley, CA, USA  3 Departamento de Biologia Celular e do Desenvolvimento, Instituto de Ciências Biomédicas, Universidade de São Paulo, São Paulo-SP, Brazil

I.  Introduction The influence of evolutionary concepts in the study of cardiac development is growing. One can argue that few areas in developmental biology are better prepared to profit from an evolutionary perspective than cardiac development. In the last 15 years the study of cardiac development has experienced an accelerated growth that can be attributed, in great measure, to the use of a variety of animal models such as flies, fish, frogs, chicks and mice. Using the intrinsic advantages of each of these models, it was possible to outline the major biological processes that are required to build a pumping organ from a few precursor cells. This success raised interest in cardiac development of nonmodel species that occupy key positions in the phylogenetic tree, which can be assessed with help from recently decoded genomes. Now, cardiac developmental biology can be studied not only from a molecular point of view, but also from a historic­al, evolutionary, perspective. This evolutionary approach utilizes different species as tools to scrutinize the commonalities of development. These commonalities are crucial to understand development at the level of a single organism, and, simultaneously, to provide a backdrop against which we can identify specific developmental mechanisms that operated during evolution to give rise to the amazing diversity of pumping organs that we encounter in animals. To take full advantage of the opportunities provided by the study of nonmodel species at key phylogenetic positions (e.g., cephalochordates, tunicates, hemichordates, echinoderms, arthropods, onychophorans and molluscs), we will have to incorporate language and concepts of both evolution and animal physiology. Through such a ­synthesis Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

we will understand how form evolved, and how it was connected to function. To accomplish this, the following conceptual obstacles need to be addressed. The first obstacle is anthropocentrism. A considerable share of our knowledge was obtained in an effort to address questions that are important for human cardiac development and congenital heart disease. In effect, data generated in species such as fishes and amphibians are often interpreted as an extension of the developmental paradigms created by the study of humans, or of other recent animals such as mice and chicks. However, this approach may not be appropriate from an evolutionary standpoint. Mice and chicks and, above all, humans, appeared on earth hundreds of millions of years after fish and amphibians. Therefore, it is likely that humans, mice and chicks have evolved novel developmental mechanisms, or modified ancestral ones to such an extent that it may no longer be possible to assume that these mechanisms must always be represented in older vertebrates. In sum, it is convenient that developmental mechanisms first characterized in humans, mice and chicks be evaluated in the most inclusive fashion, often with the help of outgroups such as invertebrate chordates, to determine if they are really ancestral in vertebrates, rather than necessarily projected back on the biology of basal vertebrates. The second obstacle is the view that evolution has been a concerted drive to greater complexity through selection. This is a particularly prevalent misconception in the study of cardiac evolution, one that fails to account for the frequent loss of pumping organs that characterize evolution in numerous animal lineages. The third obstacle is the unrestricted use of molecular conservation as evidence for the





homology of morphological characters. The remarkable success in the identification of similar molecular circuits underlying muscle function in animals as distantly related as vertebrates, insects and molluscs has been a powerful advocate for the idea that animal pumping organs share a common evolutionary origin and, therefore, that their designs are homologous. Today, however, there are reasons to believe that molecular conservation may not imply homology of design (reviewed in Xavier-Neto et al., 2007). New ideas about the origins and conservation of gene regulatory networks are stimulating a more disciplined and hierarchical use of the concept of homology (Abouheif et al., 1997; Wray and Abouheif, 1998; Erwin and Davidson, 2002). These alternative interpretations suggest richer evolutionary scenarios that not only incorporate the role of homologous developmental genes, but also allow significant roles for physical, hemodynamic factors as powerful influences that independently shape animal pumping organs (LaBarbera and Vogel, 1982; LaBarbera, 1990; Liebling et al., 2005; Forouhar et al., 2006; Auman et al., 2007). The study of the evolutionary origins of hearts and pumping organs is, therefore, in need of a synthesis that produces a set of common concepts and a common language. Such a synthesis should aim to create a productive environment in which different evolutionary hypotheses can be discussed and critical experiments planned. Our objective is to analyze the advances that have accumulated in our understanding of the anatomy, physiology, developmental biology and evolution of metazoan pumping organs. In this chapter we will discuss first the various kinds of animal circulatory pumps and then utilize a hierarchical set of concepts to analyze their evolution. We will define the position occupied by our vertebrate chambered hearts in relation to the other classes of animal pumping organs, such as arthropod dorsal vessels and mollusc chambered hearts, and trace the origins of these diverse pumping organs. Finally, we will focus on the evolution of deuterostome pumping organs and discuss evolutionary scenarios regarding the origins of our own vertebrate hearts.

II.  Animal circulatory pumps: micropumps and macropumps Pumps are integral components of strategies developed by animals to guarantee a more or less uniform access to ­ oxygen and nutrients for their cells, as well as an ­unimpeded route for the elimination of carbon dioxide and other excreta. Animals have developed an impressive array of mechanisms that facilitate both processes, which have as their final common pathway the physical mechanism of diffusion. Circulation is one such mechanism. Through the use of various kinds of pumping devices, such as ­macropumps and micropumps (Fig. 1), circulation establishes bulk flow of solutes and solvents that effectively homogenize

PART  |  1  Heart Evolution

Circulatory pumps Micropumps Cilia Flagella

Macropumps Ampullar accessory

Pulsating vessels

Tubular vessels

Chambered hearts

Figure 1  A hierarchical sequence of concepts linking animal circulatory pumps. Micropumps are diminutive circulatory devices, such as cilia and flagella, which operate at the cell level to propel fluid within small conduits, cavities or body surfaces. Macropumps are specialized pumping organs that are powered by muscle or myoepithelial cells. Macropumps are traditionally classified as ampullar accessory pumps, pulsating vessels, tubular vessels and chambered hearts.

concentrations across exchanging surfaces, reducing diffusion distances throughout animals. Macropumps are defined as the specialized, muscle-powered pumping organs consisting of various cell types working together to propel fluid inside the largest conduits of a circulatory system (LaBarbera and Vogel, 1982). In contrast, micropumps, such as cilia or flagella, operate at the level of individual cells to propel fluid inside conduits of low crosssectional area (LaBarbera and Vogel, 1982), in cavities, or along body surfaces. Both kinds of pumps can generate and maintain large outputs and are associated with animals that evolved different types and sizes of body plans, including large-bodied colonial animals such as sponges, or solitary animals such as vertebrate blue whales and mollusc giant squids. Here we will concentrate on the evolution of animal macropumps, or specialized pumping organs.

II.A.  Macropumps or Specialized Pumping Organs Macropumps have been traditionally classified into discrete categories including chambered hearts, tubular hearts, pulsating vessels and ampullar accessory hearts (Romer, 1962; Farrell, 1997; McMahon et al., 1997b) (Fig. 1). Chambered hearts are multi-compartment pumps with single or double circuits that include one to four, or more, reservoirs (the atria), and one or two, or more, main contractile compartments (the ventricles) (Romer, 1962; Farrell, 1997; McMahon et al., 1997b). These are the hearts of vertebrates and molluscs (placophoran, bivalve, gastropod and cephalopod) (Lemche and Wingstrand, 1959; Romer, 1962; Jones, 1983; Wells, 1983). Tubular hearts are contractile tubes, with or without perforations (ostia), whose beating is controlled either myogenically, or neurogenically. They are found in annelids, onychophorans and arthropods



Chapter  |  1.1  Evolutionary Origins of Hearts

(McMahon et al., 1997b), as well as in invertebrate chordates such as tunicates (Farrell, 1997). Pulsating vessels are said to be those vascular structures that propel fluid by slow peristaltic contractions, such as those found in annelids (McMahon et al., 1997a), cephalochordates and bat wing veins (Farrell, 1997). Ampullar hearts are defined as accessory pumps that boost the circulation at critical sites of high resistance, of difficult access, or where fluid energy is very low (e.g., in vertebrate lymphatic vessels, which are downstream from arteriolar and capillary beds and collect a plasma ultrafiltrate). These are the accessory pumps of crustaceans and insects, the branchial hearts of cephalopods and the lymph hearts of fish, amphibians and reptiles (Maynard, 1960; Romer, 1962; Satoh and Nitatori, 1980; Farrell, 1997; Pass, 2000).

III.  What is a heart? telling homology from analogy Our ability to expose and objectively discuss our ideas about the evolutionary origin of hearts depends largely on the meaning that we attach to the word. Hearts have been traditionally defined in two distinct ways: as the chambered pumping organs of vertebrates, or, more broadly, as any organ that propels fluid through a circulatory system (e.g., “The American Heritage Dictionary of the English Language,” 2000; “Encyclopaedia Britannica,” 2006). The first concept implies that the anatomy of all vertebrate chambered hearts is similar, because they share a common origin (homology). The second concept of hearts implies that these organs are similar because they perform an analogous function (analogy/homoplasy). Tradition supports the use of both homologous and analogous concepts, so it is reasonable to refer to pumping organs of distantly related animals such as the dorsal vessel of Drosophila melanogaster (see Chapter 1.2, Volume I), or the dorsal, ventral and commissural (transverse) vessels of annelids as hearts (Johansen and Martin, 1965). However, the use of such analogous definitions in evolutionary arguments requires constant attention, to avoid confusing analogy with homology. More importantly, ignoring the duality built into the concept of hearts may lead to the automatic acceptance of homology between the designs of pumping organs of animals whose evolutionary lines diverged hundreds of millions of years ago. This unconscious reasoning biases evolutionary arguments towards homology, because it leaves little room for the equally likely possibility that the designs of animal pumping organs, including chambered hearts, were independently created by convergence, as happened with the wings of insects, pterodactyls, primitive flying mammals, birds and bats (Simpson, 1965; Meng et al., 2006). Since any discussion on the evolutionary origins of hearts requires the tracing of lineage relationships between the various animal pumping organs, we argue that

it is advisable to stick to concepts that convey the notion of homology rather than analogy. To allow for the fact that chambered hearts arose independently in vertebrates and in molluscs, and to acknowledge the fact that developing cardiac segments maintain a dynamic relationship with themselves and with the tissues that surround them (i.e., they regress, merge into others, or divide into more compartments), Simoes-Costa et al. (2005) proposed that hearts can be defined as chambered pumps which, at some point in an animal’s lifetime, contain inflow and outflow segments that are invested with myocytes. In Fig. 1 we depict a hierarchical sequence of concepts that link animal circulatory pumps. Macropumps can be organized into different classes (e.g., tubular pumps, pulsating vessels and ampullar accessory pumps), of which chambered pumps (hearts) are a special case. Here we will utilize the term “hearts” in connection with the chambered pumps of vertebrates and molluscs. Therefore, all other macropumps will be referred to, in general, as pumping organs. We acknowledge that our view of specialized animal pumping organs is consistent with previous classification schemes that set chambered hearts apart from tubular hearts, pulsating hearts and ampullar accessory hearts (Romer, 1962; Farrell, 1997; McMahon et al., 1997b) and note that an equivalent clarifying effect may be achieved by a consistent use of chambered hearts as opposed to hearts in general. We believe these definitions are useful because they illuminate the evident differences between chambered pumps (hearts) and all the other pumping organs.

IV.  What phylogenies tell us about the origins of pumping organs IV.A.  Morphological and Molecular Phylogenies To understand how pumping organs evolved we need to be familiar with the evolutionary relationships between animals. Animals have been classified on the basis of the complexity displayed by their bodies according to the presence or absence of fundamental morphological characters, such as: (1) the occurrence of tissues as opposed to an aggregation of individual cells and extracellular matrix; (2) the number of embryonic layers (diploblastic or triploblastic); (3) the types of spatial symmetry displayed by their bodies (e.g., bilateral versus radial); and (4) the state of coelomic complexity, ranging from the absence (acoelomate) to the presence of a fluid-filled space derived from the blastocoel (blastocoelomate or pseudocoelomate), or from the mesoderm (coelomate) (Hyman, 1951). Morphological criteria further divide coelomate triploblastic bilaterians into two groups according to another series of parameters, of which the most important is the association between the blastopore and the origin of the mouth and anus. Animals that



develop a mouth near the blastopore belong to the group of protostomes (primary mouth). The animals that first form the anus in association with the blastopore and then secondarily develop the mouth belong to the group of deuterostomes (secondary mouth)1 (Fig. 2A). The availability of nuclear and mitochondrial DNA sequences opened up a new world of characters that have revolutionized our views on animal evolution (Adoutte et al., 1999; Adoutte et al., 2000; Blair et al., 2005; Philippe et al., 2005). According to these new molecular phylogenies (Fig. 2B), deuterostomes constitute a natural group (Wada and Satoh, 1994; Halanych, 1995; Cameron et al., 2000; Swalla et al., 2000). Protostomes are divided into two clades, one of animals that periodically shed their cuticles (ecdysozoans) and another of animals that share a ciliated trocophore larva (Hyman, 1951) or sport adult feeding appendices, the lophophores (lophotrochozoans) (Aguinaldo et al., 1997). Ecdysozoans include arthropods such as the insect Drosophila melanogaster, tardigrades, onycophorans and acoelomate nematodes such as the model species Caenorhabditis elegans. Lophotrochozoans include moll­ uscs and annelids, platyhelminthes and others (Fig. 2B).

IV.B.  Phylogenies and the Origin of   Pumping Organs Key observations can be extracted from the molecular phylogenetic tree when we plot the presence or absence of specialized pumping organs onto animal phyla (Fig. 2B). It is evident that all the three major bilaterian groups (deuterostomes, ecdysozoans and lophotrochozoans) contain animals that display pumping organs, such as the chambered hearts of vertebrates, the ostiated dorsal vessels of arthropods and onychophorans, the multiple, imperforated, pulsating vessels of annelids and the chambered hearts of molluscs. The presence of pumping organs in representatives of the three bilaterian groups suggests that the ancestral bilaterian, or at least the protostome–deuterostome ancestor (PDA), was already endowed with some sort of specialized pumping organ(s), an idea also supported by the repeated findings of roughly the same gene regulatory networks in the muscles that drive the very dissimilar pumping organs of deuterostomes, ecdysozoans and lophotrochozoans (discussed in Scott, 1994; Erwin and Davidson, 2002; Gilbert, 2006; Olson, 2006; Xavier-Neto et al., 2007). If the bilaterian ancestor, or the PDA, possessed specialized pumping organs, then Fig. 2B tells us that a great number of animal phyla among protostomes, ecdysozoans and lophotrochozoans must have secondarily lost their pumping organs. This is consistent with the hypothesis that many animal phyla experienced simplification 1. For other criteria and further details see Hyman (1951).

PART  |  1  Heart Evolution

or ­ regression of their pumping organs (Box 1; Table 1). Another key observation can be extracted from Fig. 2B. Apart from vertebrates and molluscs, which display a more sophisticated pumping strategy based on the sequential operation of two or more synchronously contracting compartments (chambers), the majority of animals from other phyla adhere to a more primitive approach to fluid propulsion, whereby fluid is moved by successive waves of contraction that pass along the walls of a hollow muscular structure to force its contents forwards and backwards. This approach is classically known as peristalsis. As depicted in Fig. 2B, peristaltic vessels represent the majority of pumping organs in the animal kingdom, and this dominance makes it likely that all other kinds of pumping organs evolved from a primordial peristaltic vessel in the bilaterian ancestor (see below).

IV.C.  Peristaltic Pumps Arose in Connection With Blood Vascular Systems of Bilaterians IV.C.i  Specialized Pumping Organs in Cnidarians and Ctenophores Cnidarians and ctenophores (corals, anemones, sea pens, medusae, hydroids, jellyfish and comb jellies) are viewed as diploblastic and radial animals because their larvae and polyps lack a third germ layer, and there are multiple planes that can bisect their bodies around the longitud­inal axis to produce roughly symmetrical parts (Hyman, 1940). They are morphologically simpler than most bilaterians, and thus occupy a basal position, a topology that is supported by both molecular and morphological data in phylogenies (Aguinaldo, 1997; Brusca and Brusca, 2003; Phillipe et al., 2005; Delsuc 2006) (Fig. 2). The bodies of cnidarians and ctenophores largely obviate the need for specialized pumping organs. They have remained thin, with two germ layers (endoderm and ectoderm) that are separated by a connective tissue, the mesoglea. The mesoglea is approximately 85% water, and thus does not offer much resistance to diffusion (cnidarian hydrozoans are exceptions, as they display a third layer, the entocodon, at their medusan stages) (Hyman, 1940; Gosline, 1971; Koehl, 1977; LaBarbera and Vogel, 1982; Ruppert and Carle, 1983; Seipel and Schmid, 2005). The circulation in cnidarians and ctenophores is aided by micropumps such as the cilia that line the pharynx in cnidarians, or by the ciliary action of comb plates in ctenophores (Martin and Johansen, 1965; Brusca and Brusca, 2003). In addition, some cnidarians display contractions of the circular and long­ itudinal layers of myoepithelial cells that line their gastrovascular cavities (coelentheron). This draws sea water and food in and out of their bodies, which is thought to be useful for circulatory purposes, as well as for feeding, digestive and excretory behaviors (Shimizu and Fujisawa, 2003).

Chapter  |  1.1  Evolutionary Origins of Hearts



Figure 2  Animal phylogenies and the origins of pumping organs. (A) Morphological metazoan phylogenies classify animals according to the presence of characters such as: true presence of tissues; embryonic layers (diploblastic and triploblastic); types of body symmetry (bilateral versus radial); and absence (acoelomate) or presence of a fluid-filled space derived from the blastocoel (pseudocoelomates), or from the mesoderm (coelomates). Triploblastic bilaterians are further divided into protostomes and deuterostomes. In these phylogenies, acoelomates and pseudocoelomates (mostly pumpless, heartless animals) are located at the base of bilaterians, suggesting that they are intermediates between diploblastic animals and coelomate triploblastic bilaterians (protostomes and deuterostomes), which display specialized pumping organs. (B) Molecular phylogenies placed acoelomate and pseudocoelomates together with protostomes, displacing them from the basal positions they occupied in morphological phylogenies. Protostomes were divided between ecdysozoans and lophotrochozoans. Deuterostomes were reaffirmed in a natural group, but lost a few phyla to protostomes. Molecular phylogenies suggest that all bilaterian pumping organs descend from an ancestral bilaterian peristaltic pump, and that acoelomate and pseudocoelomate animals without pumping organs must have lost them secondarily. See text for details. *Despite the absence of circulatory pumps in most echiurans, a specialized pumping organ is present in the family Ikedae.

Cnidarians and ctenophores display, or have been suggested to possess, myoepithelial, smooth muscle and striated muscle cells (reviewed in Burton, 2008). Since these are the cell types that power the specialized pumps of ­bilaterians, it is pertinent to ask whether bilaterian pumps have anything to do with the muscular apparatus that moves sea water in and out of the cnidarian gastrovascular cavity. In fact, bilaterian pumps have been speculated to originate from muscles similar to those of cnidarians (see Olson, 2006). Recent studies are providing an extra degree of complexity to this hypothesis. It is possible that cnidarians and ­ctenophores are neither completely radial, nor diploblastic

throughout their entire lifecycles (Hernandez-Nicaise and Franc, 1993; Martindale and Henry, 1999; reviewed in Seipel and Schmid, 2005). This in turn raises the challenging possibility that cnidarian, ctenophores as well as bi­laterians, all evolved from a bilateral, triploblastic, ancestor (the early metazoan ancestor), and that radial symmetry and larval diploblasty evolved secondarily among ctenophores and cnidarians, perhaps analogous to the evolution of the echinoderm adult pentaradial symmetry from a bi­laterian ancestor (Ball et al., 2004). This is an important idea, because it suggests that peristaltic pumps first appeared in the early metazoan ancestor, flourished in bilaterians, but



PART  |  1  Heart Evolution

Box 1  Pumpless, heartless animals The reasons why multiple animal phyla are associated with bodies without specialized pumping organs can be surmised from an analysis that includes parameters such as animal sizes, shapes, lifestyles, and the environments in which they live. Such an analysis indicates that the creatures that lack specialized pumping organs display at least one of the following characteristics: small size (micron to millimeter range, as pelagic planktonic animals, or meiofaunal animals), lifestyles including sessile, parasitic, filter feeding and sedentary. An additional condition is the absence of septal divisions between coelomic segments. Pumpless, heartless animals guarantee an adequate transport system for their cells by adopting strategies to minimize diffusion distances to the sources of respiratory gases and nutrients, and to sinks for their excreta. There are two basic strategies: living with very thin (flat) or with very small bodies. Animals thus can be in the ­millimeter to micron size range, or they can even reach considerable dimensions provided that they maintain a large enough ratio between exchange area and body volume (flatness index), so that their cells are never beyond the physical limits for efficient oxygen diffusion (LaBarbera and Vogel, 1982). Flatworms, unsophisticated animals from a morphological point of view, entirely lack vessels or pumping organs. They rely on a combination of generally small size, a thin body, highly-branched digestive surfaces (that increase exchange area relative to volume) and on bodily movements (that circulate their internal fluids) to maintain short diffusion distances. Other animals such as nematodes, rotifers, gastrotrichs, kinorrhynchs, nematodes, nematomorphs,

ectoprocta, loriciferans, cycliophorans and tardigrades are much more complex than flatworms, boasting a thorough gut, excretory structures and well-developed reproductive organs, but nonetheless, they lack specialized pumping organs or vessels. They are generally of very small size (hundreds of micrometers to slightly more than a couple of millimeters), and use their external surface area for gas exchange and nutrient absorption. They circulate bodily fluids via contraction of their guts and/or outer muscular layers, which compress the hydrostatic skeleton formed by their fluid-filled cavities. Absence of septal divisions between coelomic segments is most often observed as a secondary simplification (annelid hirudinoidea, or leeches). It remains possible, however, that this design reflects the basic tripoblastic bilaterian groundplan as in Vernanimalcula guizhouena, a fossil micrometer-size animal which displays coeloms along its  anterior–posterior axis, but no evidence of vessels or pumps (Chen et al., 2004). Recent phylogenies based on molecular characters (Aguinaldo et al., 1997; Regier et al., 2005; Mallatt and Giribet, 2006; Passamaneck and Halanych, 2006) suggest that bilaterian animals without pumping organs may have lost them secondarily and therefore should not be considered representatives of ancestral body plans. In summary, loss of specialized pumping organs probably occurred multiple times, in all the three bilaterian clades, deuterostomes, ecdysozoans and lophotrochozoans (Fig. b). Table 1 displays examples of pumpless, heartless, animals and the conditions associated with their circulatory phenotypes.

Table 1     Condition

Group

Small size

Rotifera,a Gastrotricha,a Kinorhyncha,a Entoprocta,a Gnathostomulida,a Loricifera,a Cycliophoraa Ectoprocta,a Tardigrade,a Tunicates (Kowalevskiidae) Arthropoda, Copepoda (Lernaeopodoida)c Mollusca (Sacoglossa, Alderia)b

Flat body

Platyhelminthesa (Acoela, Haplopharyngida and Cestoda), Nemertea (Hoplonemertea and Heteronemertes)a

Lifestyle

Parasitic: Platyhelminthes (Cestoda and Trematoda),a Nematamorpha,a Nemata (Phasmida),a Acanthocephala,a Annelida (Hirudinoidead) Sessile: Arthropoda (Cirripedia)a Sedentary and/or suspension feeder: Sipunculaa

Loss of coelomic septation

Annelida (Hirudinoidea)

Conditions associated with the absence of specialized pumping organs in bilaterian animals: the selection is illustrative rather than exhaustive. More than one condition may be associated with an animal. a: Brusca and Brusca (2003); b: Fahrner and Haszprunar (2001); c: Maynard (1960); d: Sawyer (1986).

regressed in cnidarians and ctenophores. In this view, gastro­ vascular muscles would be, at best, a vestige of a more sophisticated ancestral pumping organ. One approach to test the proposed homology between the contractile cells that operate the cnidarian gastrovascular cavity and those cells that power bilaterian pumps is to establish whether these cells express orthologous transcription factors (Abouheif et al., 1997). Hydra, a hydrozoan cnidarian, expresses CnNK2, an NK2 transcription factor that occupies

a basal position to bilaterian Nkx2.5/2.3, Nkx2.2 and Nkx2-1 (TTF-1) gene families (Holland et al., 2003). Hydra CnNK2 is preferentially expressed in the peduncle endoderm (Grens et al., 1996) which contains circular myoepithelial cells that contract to expel liquid from the gastro­vascular cavity (Shimizu and Fujisawa, 2003). Moreover, cnidarian and ctenophore contractile proteins are similar to their bilaterian counterparts (Burton, 2008 and references therein). These are all valid arguments for homology.2 However, the case for



Chapter  |  1.1  Evolutionary Origins of Hearts

homology is controversial. Cnidarians and ctenophores constitute groups of highly diversified, but insufficiently studied animals. Moreover, any attempt at character generalization in these animals is quickly confronted with the need to explain the absence of the character in multiple related cnidarian or ctenophore taxa, leading to highly unparsimonious scenarios that postulate multiple, independent losses (Burton, 2008). Pumping from the gastrovascular cavity is no exception, since not all cnidarians necessarily utilize muscular contractions to admit or expel sea water into and out of their cavities (Ruppert and Carle, 1983). Besides, there is no evidence yet that gastro­vascular pumping in cnidarians is associated with any specific set of muscles or myoepithelial cells, since they are also involved in feeding, digestion, excretion, locomotion and defense (Brusca and Brusca, 2003; Shimizu and Fujisawa, 2003; Shimizu et al., 2004). This contrasts with the muscles which power bilaterian peristaltic vessels, which function primarily to propel fluid inside the vessel. The proposed homology between bilaterian pumping organs and the myoepithelial apparatus that operates the gastrovascular cavity of cnidarians cannot yet be settled with the available data. Resolution of this question depends on clarification of the relationships between the origins of diplo­ blasty and triploblasty and the adoption of radial or bilateral symmetries. It may also require a consensual view on the types of contractile cells displayed by cnidarians and ctenophores. In summary, the evidence linking the origin of specialized pumps to cnidarians and ctenophores is weaker than the support for the origins of these structures in bilaterian animals, which will be discussed below.

V.  The rise of blood vascular systems Specialized pumping organs probably originated in connection with the rise of bilaterian blood vascular systems. According to Ruppert and Carle (1983), the first internal fluid transport system of bilaterians was the coelom (Fig. 3A). In the coelom, fluid is moved by micropumps, such as the cilia that equip mesodermal cells lining the coelomic cavity. Muscular action from the body wall also aided the circulation by compressing the hydrostatic 2. Firm evidence of gene homology is needed before organ homologies can be proposed based on similar expression patterns across animals. In this sense, the case for homology between cnidarian gastrovascular muscles and bilaterian macropumps is stronger than that of the C. elegans pharynx. Circulation in C. elegans is driven by contractions of pharyngeal muscles that express ceh-22, a NK2 gene. This prompted the thoughtful suggestion that an evolutionarily-conserved mechanism underlies heart development in vertebrates and insects, and pharyngeal development in nematodes (Haun et al., 1998). However, in contrast to the cnidarian CnNK2 gene, which occupies a basal position to most NK2 gene families, ceh-22 clusters in the Nk2.2-vnd gene family (Okkema and Fire, 1994; Holland et al., 2003), suggesting it is not an ortholog to the Nkx2.5/2.3 family. Thus, the nematode pharynx is most likely an analog of the arthropod dorsal vessel or vertebrate heart.

skeleton formed by liquid-filled coelomic cavities. Exquisite support for this view of the ancestral bilaterian was found by Chen and colleagues in the 600-millionyear-old microfossil Vernanimalcula guizhouena recovered from the Doushantuo formation in China (Box 2). The critical event in the origin of blood vascular systems was the separation of the body into two or three coelomic compartments that were insulated from each other by “bulkheads” of septal tissue (Fig. 3B). Since total separation of the body into compartments would effectively isolate most animal tissues from their respiratory and gut absorptive surfaces, such impervious septal bulkheads may have selected animals for an alternative way to circulate their fluids along their total spans. In this interesting scenario, blood vascular systems emerged and connected body tissues throughout the animal (Fig. 3B). Primitive blood vessels were thus generated at extracellular spaces occupied by basal lamina secreted by epithelia derived from the three germ layers. These spaces appeared at ventral and dorsal positions between the left and right coeloms (i.e., the mesenteries), giving rise to ventral and dorsal vessels. They also appeared between the coelom and gut, giving rise to gut sinuses, and between the coelom and superficial ectoderm, giving rise to lateral vessels (Fig. 3B). These primitive vessels can be pictured as nothing more than empty spaces in the extracellular matrix running through the longitudinal axis and bypassing the blockade of septal tissues. They lacked internal lining by endothelial cells, but were covered on their external aspect by mesodermal coelomic cells (Ruppert and Carle, 1983).

V.A.  Myoepithelial Cells, Myocytes and the Origin of the Ancestral Peristaltic Pump With development of more rigid exo- and endoskeletons, the auxiliary role played by compression of the hydrostatic skeleton gradually lost effectiveness, and animals may have been selected for the development of a circulatory system based on the propulsion of intravascular fluid by specialized pumping organs. In other words, this was a critical transition from a fluid transport system that relied on coelomic micropumps and bodily movements, to a system driven by macropumps (Fig. 3C). In summary, this hypothesis proposes that the basic foundation which all pumping organs were built over was the organization of a layer of contractile myoepithelial or myocyte cells derived from the coelomic epithelium that lined primitive vessels. Therefore, the birth of the first specialized pumping organ coincided with the creation of a myoepithelial or muscular cell type, whose sheets enveloped the external walls of blood vascular channels, thus forming the ancestral peristaltic pump (Ruppert, 1997; see also Chapter 8.1). This view is consistent with the ventral origins of vertebrate pumping organs, the dorsal origins of arthropod vessels and the ventral, dorsal and commissural origins of annelid pumps (Martin, 1980; Milne-Edwards, 1982) (Fig. 3C).

10

PART  |  1  Heart Evolution

Figure 3  Hypothetical origins of blood vascular systems and specialized pumping organs. (A) Scheme of a basal bilaterian. According to Ruppert and Carle (1983) the coelom represented the first fluid transport system of bilaterians in which fluid was propelled either by micropumps lining the coelomic cavity, or by compression of the hydrostatic skeleton by muscular action. (B) Creation of septal bulkheads may have isolated animal tissues from respiratory and gut absorptive surfaces, and selected animals for an alternative way to circulate fluids throughout the body. Primitive blood vessels were generated at extracellular spaces occupied by basal lamina, giving rise to the ventral and dorsal vessels between left and right coeloms, to gut sinuses between coelom and gut, and to peripheral vessels between coelom and superficial ectoderm. These vessels bypassed the blockade of septal tissues, connecting animal segments along the anterior-posterior axis (Ruppert and Carle, 1983). (C) Myoepithelial cells or myocytes of coelomic origin enveloped these channels, creating an external layer of contractile cells. This may have represented the foundation of the ancestral pumping organ, a rudimentary peristaltic pump that gave rise to all specialized pumping organs of bilaterians.

VI.  A gut origin for the ancestral peristaltic pump: an alternative view Another origin for the ancestral peristaltic circulatory pump is suggested by parallels between the vessels of the blood vascular system and the gut. Both structures display a similar anatomical organization, with a delicate internal lining formed by respective endothelial or epithelial layers surrounded by external layers of visceral muscle (Fawcett, 1994). In invertebrates such as Drosophila and decapod crustaceans, the visceral muscles lining the gut and the major arteries (Burnett, 1984) are striated, and the visceral muscles of Drosophila express transcription factors such as Nkx2-5 and its NK2 paralogs (reviewed in Evans, 1999), which hints at a common origin for these muscles. When coupled with the knowledge that guts preceded blood vascular systems, and that vessels likely gave rise to circulatory pumps, the similarities between vessels and

gut suggest a plausible scenario for the origin of pumping organs. In this view, the striated musculature that operates bilaterian pumping organs originated from striated muscles that enveloped the gut (see Harvey, 1996) (Fig. 4A). In vertebrates the external and media layers of gut and vessels are formed by smooth muscle (Fawcett, 1994). Vascular and gut smooth muscle cells express a characteristic set of contractile proteins that include the smooth muscle myosin heavy chain SM MYHC (MYHC 11) (Babu et al., 2000; Desjardins et al., 2002) and transcription factors such as those from the NK2 clade (reviewed in Harvey, 1996; Evans, 1999). These similarities suggest a variation of the gut origin hypothesis. This view entails a transition from visceral smooth muscle to vascular muscle, and to cardiac muscles. The idea is intuitive because it places the seemingly disorganized layout of the smooth muscle contractile apparatus as a primitive character. The implication is that there was a progressive sophistication towards the highlystructured layout of contractile proteins in the sarcomeres

11

Chapter  |  1.1  Evolutionary Origins of Hearts

Box 2  Vernanimalcula: Ground zero for the evolution of vascular systems and their pumps Vernanimalcula guizhouena, a micrometer-size fossil (124–78 m)  of 600 million years, represents the oldest identified triploblastic animal yet (Chen et al., 2004a) (Fig. a–b, reproduced with permission). The biological affinities of Vernanimalcula were questioned (Bengtson and Budd, 2004), but the arguments for an artifactual origin were counteracted in a robust reply that called into question the exquisite preservation characteristics of the Doushantuo formation (Chen et al., 2004b). Vernanimalcula boasts three easily identified layers that may correspond to ectoderm, endoderm and mesoderm (Fig. c). The fossils suggest that the animal had an inner layer organized into a complete gut flanked by a mouth and anus, an external layer with small, paired, sensory-like appendages, and a middle layer forming an uninterrupted coelom that spans the short A–P axis (Chen et al., 2004a). Except for its diminutive size, Vernanimalcula had all the attributes predicted by Ruppert and Carle (1983) for a bilaterian animal before the advent of vascular systems. The continuous 

(A)

(B)

Gut origin

Dorsal vessel Gut Ventral vessel

coelom in Vernanimalcula suggests that fluids could travel the length of the whole animal without the need for specialized conducting systems or a circulatory pump, using only putative ciliated micropumps in its coelom, as well as compression of the coelomic hydrostatic skeleton by body movements. As a primitively designed triploblastic and bilaterian animal, Vernanimalcula likely occupies a place at the bilaterian stem. Its position in relation to the last common ancestor of protostomes and deuterostomes (PDA) and the enigmatic acoels (Hejnol and Martindale, 2007) is a matter for dispute. The lack of vascular systems and circulatory pumps suggest Vernanimalcula is not the PDA, since this animal almost certainly had these characteristics, which are present in the three major bilaterian clades. Whether Vernanimalcula preceded or followed acoels is unknown. The interpretation depends on whether one believes the first bilaterian was coelomated or acoelomated (Hejnol and Martindale, 2008). Figures reproduced from Chen et al., 2004 with permission.

Blood vascular origin

Ectoderm Extracellular matrix

Endoderm

Myoepithelial cell

(C)

Coelomic origin

Peripheral vessel

Coelom

Gut sinus

Mesoderm

Figure 4  A gut origin for the muscles that drive bilaterian pumping organs. (A) Similarities in gene regulatory pathways between visceral muscles and circulatory pump muscles suggest that the striated musculature that operates bilaterian pumping organs originated from striated muscles that enveloped the gut (Harvey, 1996; Evans, 1999). (B) The gut origin scenario in (A) contrasts with the idea that the muscles from bilaterial pumping organs originated from the coelomic mesoderm that overlaid primitive hemal channels in the extracellular matrix (Ruppert and Carle, 1983). (C) In this alternative hypothesis the similarities between visceral and circulatory pump muscle are traced to a common, coelomic origin.

of cardiac muscles, and this notion is further supported by the fact that vertebrate cardiac precursors transiently express smooth muscle alpha actin before turning on typical myocardial contractile proteins (Ruzicka and Schwartz, 1988; Sugi and Lough, 1992; Colas et al., 2000). A smooth muscle origin for the cardiac striated musculature is difficult to uphold, however. Contrary to common

knowledge, SM MYHC is neither ancestral to striated myosins, nor strongly related to them. Multiple phylo­ genetic studies indicate that SM MYHC is closely related to nonmuscle type II myosins (Goodson and Spudich, 1993; Oota and Saitou, 1999; Desjardins et al., 2002; Chiba et al., 2003). Moreover, branch lengths in these phylogenies suggest that SM MYHC diverged from nonmuscle myosin

12

types only after the split of bilaterian animals into deuterostome and protostome clades. This suggests that vertebrate cardiac and skeletal muscles are related to invertebrate muscles, rather than to vertebrate smooth muscles (Oota and Saitou, 1999). Consistent with this, the MYHC expressed in morphologically nonstriated muscles of molluscs is closely related to vertebrate sarcomeric MYHCs (Nyitray et al., 1994), while the mollusc MYHCs that are similar to SM MYHC are all expressed in nonmuscle cells (Hasegawa, 2000; reviewed in Desjardins et al., 2002). Moreover, there are also alternative interpretations for the transient expression of smooth muscle alpha actin in cardiac precursors. Careful characterization of smooth muscle alpha actin expression shows that the gene is not limited to cardiac precursors. Rather, expression is observed throughout the noncardiac component of the lateral plate mesoderm, including the somatic mesothelium and the splanchnic mesothelium, indicating that smooth muscle alpha actin expression marks coelomic derivatives (Colas et al., 2000). Thus, transient smooth muscle alpha actin expression in cardiac precursors is more likely a vestige of coelomic origin than evidence for a smooth muscle origin of cardiac muscle. In summary, there is support for the hypothesis that bilaterian pumping muscles evolved from an ancestral gut striated muscle, but not from a smooth muscle cell type.

VI.A.  Origins of Pumping Organs:   A Synthesis The hypothesis of a gut origin (Fig. 4A) for the circulatory pump musculature can be reconciled with the blood vascular scenario suggested by Ruppert and Carle (1983) (Fig. 4B). In most bilaterians, the gut musculature and most vascular muscles derive from the mesoderm. Thus, it is almost certain that these muscle types share a common origin in this germinal layer which, in the bilaterian ancestor, probably overlaid the endoderm in the form of a coelomic layer (Fig. 3A). Therefore, it is possible that both gut and vascular smooth muscle derived directly from the coelomic epithelium, or perhaps from a mesodermal stem cell population closely attached to the coelomic wall. In other words, blood vascular and gut scenarios are roughly equivalent and can be actually reduced to a scheme in which gut and vascular muscle derived from coelomic mesoderm (Fig. 4C).

VII.  Reconstructing the circulation in the bilaterian ancestor It is likely that multiple, primitive, peristaltic pumps (dorsal, ventral, lateral and gut sinus) powered the circulatory system of the bilaterian ancestor. This decentralized system of multiple and rudimentary pumps presumably lacked efficient coordination within a single pump, as well as among their multiple pumps, which may have resulted

PART  |  1  Heart Evolution

in fluid being propelled backwards and forwards throughout the animal. In the bilaterian ancestor, intravascular fluid was probably in direct contact with the extracellular compartment. In such a system, intravascular fluid circulates to and from major vessels through unlined sinuses and lacunae, rather than utilizing a stereotyped sequence of increasingly smaller conducting vessels, capillary beds and progressively wider collecting vessels, which is characteristic, for instance, of modern vertebrates (LaBarbera, 1990) (Box 3).

VII.A.  The Impact of the Ancestral Peristaltic Pump Concept The concept that a primitive bilaterian peristaltic vessel was the building block on which all the other bilaterian macropumps were built has profound implications for scenarios on the evolution of pumping organs, and provides a number of opportunities for syntheses between many ideas on pumping organ evolution. The concept offers: (1) an objective foundation for reconciliation between two camps on the evolutionary relationships among specialized bi­laterian pumping organs (the proponents of homology and the supporters of convergence); (2) an explanation for the surprising similarities between the gene regulatory networks that were characterized in the pumps of distantlyrelated bilaterians such as the arthropod dorsal vessel and the vertebrate heart; (3) an alternative classification for specialized pumping organs; and finally (4) paradigms for the parallel evolutionary changes that circumvented the common hemodynamic constraints faced by bilaterian peristaltic pumps (LaBarbera, 1990).

VII.B.  Homology, Analogy and Gene Regulatory Networks The similarities between arthropod dorsal vessels, tunicate tubular pumps and vertebrate chambered hearts have been reviewed (Bodmer, 1995; Frasch, 1999; Cripps and Olson, 2002; Zaffran and Frasch, 2002; Davidson, 2007). It is widely accepted that the obvious similarities lie at the initial developmental stages and at the gene regulatory networks, but that adult morphologies are divergent. These resemblances can be interpreted either as reflection of common origins (homology), or as the result of similar function (analogies produced by parallelism or convergent evolution). The discovery of multiple ortholog genes playing critical roles in pump development and function (e.g., tinman/Nkx2-5; d-mef2/Mef2s, pannier/Gata, dpp/Bmps, hand/Hand1-2; mad/ Smad1, medea/Smad4, wingless/Wnt, seven-up/Couptf-II) (Bodmer et al., 1990; Arceci et al., 1993; Kelley et al., 1993; Komuro and Izumo, 1993; Lints et al., 1993; Martin et al., 1993; McDermott et al., 1993; Lilly et al., 1994;

13

Chapter  |  1.1  Evolutionary Origins of Hearts

Box 3  Open and closed circulatory systems The evolution of macropumps in bilaterian animals cannot be understood in isolation from a circulatory context that includes the vessels that lead intravascular fluid to and from the major pumping organs. However, the comparative study of bilaterian vascular beds is marred by arbitrary dichotomies and semantic subtleties that hinder communication. Two main obstacles are to blame. One is the artificial separation between the so-called open and closed circulatory systems. The open or closed nature of a vascular system is defined in relation to the complexity of the layers that make up a vessel wall. Closed systems, in which intravascular fluid is separated from the vascular basal lamina by an inner layer of cells known as endothelium, contrast with open systems, in which fluid is in direct contact with the basal lamina (Farrell, 1997; McMahon et al., 1997b). However, a second concept often appears in the discussion of open or closed systems. The concept is related to the presence or absence of a stereotyped sequence of branching patterns that characterize the vessels leaving or leading to a central pump. This sequence, typified by the so-called closed higher vertebrate pattern, includes the successive branching of large diameter efferent vessels into successively smaller arterioles and capillaries, as well as the progressive coalescence of venules into small, medium and large afferent veins. This contrasts with animals with so-called open circulation, in which intravascular fluid flows through multiple, ill-defined routes that include large sinuses or small lacunas (Fange, 1972; LaBarbera and Vogel, 1982; Satchell, 1991c). These artificial dichotomies associate open circulatory systems with primitive design, sluggishness and low pressure, while closed systems are linked to advanced design, dynamic circulation and high blood pressures (McMahon, 2001). However, these standards have, time and again, proved to be too strict to be useful, because it is not possible to fit the natural variation into these two extreme conditions (Jones, 1983; Burggren et al., 1997; McMahon,

Nguyen et al., 1994; Cserjesi et al., 1995; Frasch, 1995; Hollenberg et al., 1995; Srivastava et al., 1995; Georgias et al., 1997; Gajewski et al., 1999, 2000; Pereira et al., 1999; Cripps and Olson, 2002; Han et al., 2002; Kolsch and Paululat, 2002; Zaffran and Frasch, 2002) in these distantly-related animals has tilted the balance so heavily on the side of homology that it is not uncommon to find attempts to reconstruct the ancestral bilaterian pumping organ as an ostiated, D. melanogasterlike peristaltic pump suspended by alary muscles/ligaments. There are good reasons, however, to question literal reconstructions of ancestral animal morphologies based on the layouts of highly derived model species such as D. melanogaster, Ciona intestinalis or Mus musculus (Xavier-Neto et al., 2007). The major problem with attempts to homologize vertebrate chambered hearts, arthropod dorsal vessels, or tunicate tubular pumps is that homologies in pump design can not be appreciated solely on the basis of pump phenotypes displayed by vertebrates, flies or tunicates. Any thorough

2001). If we consider the presence or absence of endothelial coverage as a criterion, many organisms display vascular beds that mix closed with open segments (Farrell, 1997). In fact, not even vertebrates display a continuous endothelium lining their vessels, as blood is in direct contact with tissues in hepatic and splenic sinusoids (Fawcett, 1994). Moreover, in the cephalochordate Amphioxus, major vessels display considerable coverage of their basal laminas by cells that are alternatively recognized as endothelial-like (Moller and Philpott, 1973; Moller and Ellis, 1974), or dismissed as adhered amoebocytes (Rahr, 1981). If we consider a stereotyped branching pattern as a character (Farrell, 1997), it is evident that there are animals, such as hagfish and lampreys, that display a large amount of sinuses in addition to the typically well-organized vertebrate branching patterns (Satchell, 1991c; Fange, 1972). The large sinuses of hagfish and lampreys are reminiscent of the socalled open circulatory systems and they may account for the high proportion of blood volume over body mass observed in these animals (Fange, 1972, and references therein). There is also a great deal of heterogeneity among animals from the same phylum. Some arthropods, such as insects, have poorlydeveloped vascular beds. As a result, their organs, especially the central nervous system, are superfused, rather than perfused, with hemolymph (Maynard, 1960; Pirow et al., 1999; McMahon, 2001). However, decapod crustaceans display highly-­organized vascular beds which include small vessels that perfuse the central nervous system structures. Indeed, the complexity of the decapod circulation can be fully appreciated by an analysis of their plastic casts (McMahon 2001; McGaw and Reiber, 2002; Farrelly and Greenaway, 2005; McGaw, 2005). In summary, it seems more appropriate and useful to consider that animals display a range of states between the  idealized open and closed systems, regardless of the criterion utilized to define the extremes.

argument in favor of such deep homologies of design must also include the ostial vessels of onychophorans, the imperforated tubular pumps of annelids and the chambered hearts of molluscs (Fig. 5). Given the highly dissimilar designs of all these pumping organs, it is highly unlikely that one can decide, a priori, which of these pumping configurations were utilized in the bilaterian ancestor, or ­figure out a suitable intermediate for these highly dissimilar organs (Xavier-Neto et al., 2007). The idea that a rudimentary ancestral peristaltic pump was formed at the external walls of primitive blood vascular channels by a layer of myoepithelial cells or myocytes solves these problems. It also reconciles the extreme views suggesting either homologous or convergent origins of pumping organs. When we consider the rudimentary peristaltic pump envisioned in Fig. 3C, it becomes easy to understand how the frankly dissimilar pump designs of bi­laterians could have been independently developed from such a basic and homologous feature. This notion is consistent

14

PART  |  1  Heart Evolution

Vertebrates (Teleost fish)

Cephalochordates (Amphioxus)

Tunicates (Ascidian)

Echinoderms (Holothurian)

Hemichordates (Enteropneust)

Onychophorans (Peripatopsis)

Arthropods (Drosophila)

Molluscs (Bivalve)

Annelids (Lumbricus)

Figure 5  Specialized pumping organs of bilaterians. Note the highly dissimilar designs. Deuterostome circulatory pumps based on SimoesCosta et al. (2005). Protostome pumping organs and molluscan heart (bivalve) based on Brusca and Brusca (2003). Topology based on the molecular phylogeny. See text for details.

with novel ideas proposing that the homologies encountered among gene regulatory networks from distantly related animals lie at the level of the genetic pathways assembled early in evolution to give rise to the basic cell types that underlie organ function, rather than at the level of the sophisticated tri-dimensional organ structures (Scott, 1994; Erwin and Davidson, 2002). This idea also provides a convenient fallback position for those that may have been tempted by more extremist views on the convergent evolution of pumping organs, since it has always been difficult to rationalize how a completely independent origin of pumping organs would be associated with almost exactly the same gene regulatory networks in all different bilaterian pumping organs. In summary, the origin of the ancestral peristaltic pumps may have been intimately associated with the rise of a new muscular cell type (Scott, 1994; Erwin and Davidson, 2002; Xavier-Neto et al., 2007).

VII.C.  Parallel Avenues in the Evolution of Hearts and Pumping Organs The origin of bilaterian pumping organs may be approached as a case of evolutionary parallelism (Simpson, 1965). From a primitive peristaltic organ represented in a mere sheet of contractile cells overlying hemal channel(s) (Fig. 3C), deuterostome, ecdysozoan and lophotrochozoan pumps (Fig. 5) may have evolved in parallel. Thus, their superficially similar adaptations may be best understood as functional analogies selected to cope with common physical constraints that limit the performance of peristaltic pumps (Xavier-Neto et al., 2007). This hypothesis suggests the stimulating possibility that hemodynamic factors, acting as selective pressure, interacted with morphogenetic programs of pump design to build specialized pumping organs that were of adaptive value for bilaterians occupying varied niches.

VII.D.  A Reappraisal of the Different Categories of Pumping Organs The idea that the specialized pumping organs of animals derived from a primitive peristaltic vessel suggests that the classical division of these organs into chambered pumps, tubular pumps, pulsating vessels and ampullar accessory pumps can be molded into an alternative scheme that could be employed in deep evolutionary analyses. Indeed, we have recently argued that the only distinction that adds a useful parameter for deep evolutionary comparisons among animal pumping organs is between those organs that evolved as peristaltic pumps and those that evolved as chambered pumps (Xavier-Neto et al., 2007). This proposition is a reaction to the contradictory standards utilized to classify ampullar accessory pumps, tubular hearts or pulsating vessels. The category of ampullar accessory pumps implies the existence of a connection between pump shape and its accessory nature, which does not exist. Moreover, there is little justification for lumping together boosting devices that are actuated by skeletal muscles (e.g., accessory pumps of insects, the cardinal pump of hagfish and caudal pumps of fish) with accessory pumps such as the portal “heart” of hagfish and the branchial “heart” of cephalopods, which are powered by virtually the same cardiomyocytes that make up their systemic chambered hearts (Wells, 1983; Satchell, 1991a,d; Pass, 2000). The distinction between tubular pumps and pulsating vessels is also artificial. Indeed, the pumping organs of arthropods such as chelicerates, crustaceans and insects (Fig. 6; Box 4), as well as annelids such as polichaeta and oligochaeta, are described either as tubular hearts, or pulsating vessels (McMahon et al., 1997b; Brusca and Brusca, 2003) in spite of the fact that all arthropod and annelid pumps share among themselves a common, phylum-specific engineering plan. In fact, tubular pumps are a special case of pulsating vessels, a

Chapter  |  1.1  Evolutionary Origins of Hearts

15

Figure 6  Evolution of the dorsal vessel in Panarthropoda (Tardigrada (water bears), Onychophora (velvet worms) and Arthropoda). Panarthropods are represented in the context of a molecular phylogeny based on Regier et al. (2005) and Giribet et al. (2005). Dorsal vessels and some major efferent vessels are depicted in blue over a gray silhouette of representative species. Tardigrades and onychophorans are outgroups to arthropods, which are represented by an unresolved trichotomy of Chelicerata, Myriapoda and Pancrustacea (Crustacea plus Hexapoda). The onychophoran dorsal vessel, a tubular organ spanning most of the animal anterior–posterior axis is thought to be similar to the ancestral arthropod condition. The absence of pumping organs in tardigrades is probably secondary to the assumption of small body sizes. One major trend in the evolution of arthropod pumping organs was the concentration of pumping work in anterior or posterior segments. This centralization may be a consequence of the fusion of body segments as seen in Eumalacostracans, in which condensation of anterior body segments created a powerful single compartment pump several muscle layers thick (decapods), or secondary to a loss of body segments as seen in Cladocera. In hexapod insects pumping work was concentrated in posterior segments, while anterior segments display very low or no contractility. Sessile cirripeds lost the dorsal vessel, but the class Thoracica developed a new pumping organ actuated by skeletal muscles. Remipedia, a primitive-looking arthropod, often occupies a basal position in morphological phylogenies, but recent molecular evidence suggests it is a derived arthropod (Regier et al., 2005). Figure inspired by Wilkens and drawn from Wilkens (1999); Maynard (1960); Brusca and Brusca (2003); McMahon et al. (1997a).

clear adaptation of an original peristaltic project (see Romer, 1962). The former just happen to contract rapidly enough to give the impression of synchronicity, while the latter display slow contraction waves that give them their distinct peristaltic character (Clark, 1927; Maynard, 1960; McMahon et al.,

1997b). Most tubular pumps are neurogenic, indicating that it is the integrated neural control of segmental contractility, rather than their morphologies, that endows these pumps with synchronicity or near synchronicity of contraction (Clark, 1927; McMahon et al., 1997a).

16

PART  |  1  Heart Evolution

Box 4  Evolution of arthropod pumping organs The classic idea that arthropods and annelids form a clade of segmented animals (Articulata) (Adoutte et al., 2000) was superseded by recent molecular and morphological phylogenies (see Nielsen, 2003, for a dissonant view). This removed support for the view that peristaltic vessels of arthropods and annelids were homologous, and that annelid vessels represented the primitive state of arthropod dorsal vessels (Clark, 1927; McMahon et al., 1997a). The separation of arthropods from annelids is fully consistent with their divergent pumping organs, which only superficially resembled each other in their tubular nature (Xavier-Neto et al., 2007). It is also in line with other purported arthropod–annelid homologies that have been questioned, such as those between their limbs and appendages (see the Snodgrass tapes at http://www.life. umd.edu/entm/shultzlab/snodgrass/). In contrast to Articulata, the idea of a superclade Panarthropoda formed by arthropods, onychophorans and tardigrades is now well-supported by both molecular and morphological evidence (Aguinaldo  et al., 1997; Giribet et al., 2000; Garey, 2001; Giribet et al., 2001; Peterson and Eernisse, 2001; Mallatt et al., 2004). The existence of onychophoran and tardigrade outgroups to arthropods is invaluable, as it makes possible attempts to understand which characters are ancestral and which are derived among the highly complex arthropod phyla (Fig. 5). The layout and physiology of the onychophoran pump is consistent with consensual reconstructions of primitive arthropod pumps (Martin and Johansen, 1965; Hertel et al.,  2002). The ancestral arthropod pump was probably a tubular organ that spanned most of the animal’s A–P axis, and was located at the dorsum of the animal. This dorsal vessel was suspended by allary ligaments and/or muscles and was immersed in a pool of extracellular liquid (hemolymph) contained in a pericardial space. The pericardial space provided a reservoir for the oxygenated hemolymph that perfused body tissues and acted as a filling chamber for the dorsal vessel (McMahon, 2001). The dorsal vessel had a clear segmental organization and displayed one perforation, or ostium, in line with each segment, as is usual in extant arthropods (Tjonneland et al., 1987). Ostia represented the only access for pericardial hemolymph into the dorsal vessel, regulated by valves that open during diastole and close during systole due to the action of ostial muscles (Maynard, 1960). Although the dorsal vessels of basal arthropods, such as chelicerates and myriapods, are often regarded as neurogenic (McMahon et al., 1997; Hertel and Pass, 2002), the weight of evidence suggests that neural control was added on top of an ancestral myogenic arthropod vessel. This idea is consistent with the

VIII.  Hemodynamic constraints may have shaped modern pumping organs out of a primitive peristaltic pump Although peristaltic pumps are dominant, versatile and adaptable, there is evidence that they are far less prepared

myogenic, peristaltic, vessels of onychophorans (Hertel et al., 2002). It agrees with the experimental observation that the embryonic and early juvenile dorsal vessels of the malacostracan isopodan Ligia exotica (Yamagishi and Hirose, 1997) and Homarus americanus (Burrage and Sherman, 1978) are myogenic at embryonic or juvenile stages, but become neurogenic in adults. It is also consistent with mounting evidence that “neurogenic” vessels display intrinsic, albeit slower and less coordinated, myogenic rhythms after separation from their neural inputs (see Maynard, 1960; Greenberg, 1979; Wilkens, 1999; Hertel et al., 2002 and references therein). Apart from the secondary losses of dorsal vessels in small (e.g., some copepods) or sessile arthropods (e.g., cirripedes), there have been two major evolutionary trends in the evolution of arthropod pumping organs. One was the development of accessory pumps in some crustaceans (Maynard, 1960) and hexapods (Pass, 2000). More often than not, these accessory pumps are not related to the dorsal vessel (see Pass, 2000). The other major trend was concentration of pumping work in anterior or posterior segments of the dorsal vessel. This concentration happened in at least two different contexts. Firstly, eumalacostracans, such as decapod lobsters, seem to have fused segments to create powerful, single compartment, pumps  that are several muscle layers thick and boast one-way valves that  keep flow unidirectional. Wilkens (1999) suggested that this derivation from the long and thin ancestral dorsal vessel allowed development of large-bodied and active crustaceans. Second, branchiopods, such as Cladocera (e.g., Daphnia magna) possess a centralized pumping function in a small, single-compartment globular pump with a single ostium. In Daphnia, centralization seems to be secondary to the loss of several body segments (Wilkens, 1999) and to the assumption of small sizes, suggesting that concentration of circulatory work in this animal reflects an ongoing loss of importance of the dorsal vessel as a circulatory pump. This is consistent with the reduced importance of convective oxygen transport over diffusion from the integument in small individuals (Bäumer et al., 2002). The small globular dorsal vessel of Daphnia, however, is not a vestigial organ, as it is required when individuals grow or when they are challenged with hypoxia (Paul et al., 1997). This trend towards concentration of pumping work is also manifested in hexapods. In insects, such Drosophila melanogaster, the contractile function of the dorsal vessel is focused on posterior, abdominal, segments (ventricle), while anterior thoracic segments (aorta) contract only very weakly (Wigglesworth, 1974; Hertel and Pass, 2002). The functional significance of this evolutionary trend is not immediately obvious in hexapods.

than the chambered hearts of vertebrates and molluscs, and the pumps of malacostracan crustaceans, to sustain the high rates of output demanded by large, highly-active and/or homothermous animals (Wilkens, 1999). To understand the mechanical constraints that are inherent in the peristaltic mode of operation, it is useful to examine a simple peristaltic pump, a positive displacement

17

Chapter  |  1.1  Evolutionary Origins of Hearts

Figure 7  Peristaltic pumps. (A) The operation of a roller pump, a man-made positive displacement device. The duty cycle is initiated when the roller engages a flexible tube (t1). At this “pinching site” roller action constricts the tube producing opposite fluid streams (arrows). After the roller first engages the tubing its swinging movement squeezes the tube generating net forward flow (t2) until the roller disengages the tubing (t3). Backflow (red arrows) is inherent to all peristaltic pumps. (B) Physical, hemodynamic, constraints such as backflow, distension of downstream relaxed segments, faulty coordination between contraction and relaxation of pump segments producing increased resistance to flow or reflections limit the performance of biological peristaltic pumps. A completely retrograde sequence of contractions and relaxations can move fluid backwards in biological peristaltic pumps (see text for details). Forward flow: green arrow.

device such as a roller pump. In these machines, propulsion is initiated when the roller engages a flexible tube. At this “pinching site” roller action constricts the tube and produces fluid streams that travel in opposite directions. This physical limitation generates backflow, which is inherent to all peristaltic pumps. After the roller first engages the tubing, its swinging movement squeezes the downstream tube segments, generating net forward flow until the roller disengages the tubing (duty cycle; Fig. 7A) (COMSOL Multiphysics Modeling Guide). Here, roller pumps have a definitive advantage over biological peristaltic pumps in that their pinching and duty cycles are invariate, and as a result their outputs are directly proportional to pinching frequency (Liebling et al., 2005; Forouhar et al., 2006). However, regular pinching and duty cycles are not something to take for granted in biological peristaltic pumps (Fig. 7B). Efficient pumping in biological peristaltic pumps greatly depends on the coordination between the cycles of contraction and relaxation among muscle cell segments. In one favorable scenario, the pressure wave generated by contraction reaches the following contractile segment before it generates enough tension to reduce the vascular cross-section, but well after enough tension has been raised to prevent vessel distension (Fig. 7B). If coordination is not optimal, loss of fluid energy may occur, in the form of wall strain when downstream segments are relaxed, or when the fluid stream encounters contracted segments in which the lumen is reduced (Fig. 7B). In the

latter case, if the segment is significantly constricted, the fluid stream may smash at obstructions, generating reflections that will further add to backflow. Many more scenarios can be modeled, including the extreme case when a completely retrograde sequence of contractions and relaxations moves fluid backward. In summary, the numerous adaptations that are now present in efficient bilaterian macropumps may have been selected in the course of evolution as solutions to hemodynamic limitations of the peristaltic design such as backflow, loss of fluid energy by distension of downstream segments, reflection and reversion (Fig. 7B) (see also Satchell, 1991b; Xavier-Neto et al., 2007).

VIII.A.  Solutions to the Shortcomings of Peristaltic Pumps: A Mixed Bag of Tricks There were many answers to the problems that plague the peristaltic design. An analysis of pumping organs in deuterostomes, ecdysozoans and lophotrochozoans (Fig. 5) tells us that each of these three lineages utilized different combinations of a limited number of adaptations. These adaptations may be understood as recurrent themes that fall into two main groups: those that improved pump performance but conserved a peristaltic mode of operation; and those that modified pumping so extensively as to create new mechanisms.

18

The simplest possible adaptation was, perhaps, to tolerate the inherent bidirectional pumping of peristaltic organs. Rather than organizing circulation around an efficient, regurgitation-proof, unidirectional pump, tunicates and some hexapod arthropods accepted and further exploited bidirectional pumping. Tunicates display two distinct groups of pacemakers that alternate the direction of pumping so as to perfuse their vessels from each direction, which may be understood as a compensation for inefficient uni­ directional pumping (Anderson, 1968). A similar mechanism is present in pupal and adult stages of endopterygote insects (Wigglesworth, 1974; Hertel and Pass, 2002). In fact, bidirectional pumping in hexapods is so important as to have been consolidated by creation of an additional opening at the posterior end of the dorsal vessel (endopterygotes), or by the evolution of valves that avoid forward flow in posterior dorsal vessel segments in Diplura, a class of nonwinged hexapods (Hertel and Pass, 2002). A different mechanical adaptation, which also seems to represent an astute exploitation of the intrinsic deficiencies of the peristaltic design, was recently described in the embryonic zebrafish (Forouhar et al., 2006) (see Chapter 1.4, Volume I). In the zebrafish heart at 26 hours postfertilization, contraction of a posterior segment creates a bidirectional wave whose retrograde component travels towards the inflow tract. The inflow tract displays different impedance characteristics from the main body of the cardiac tube. This impedance mismatch reflects the retrograde wave, creating a local expansion that sucks blood from the inflow tract. This reflected wave travels in the anterograde direction in such a manner that its front coincides with a major relaxation of the body of the cardiac tube, further accelerating the blood towards the outflow tract. This interesting usage of backflow, impedance mismatch and reflection creates special zones in the frequency–output relationship where the output is much higher than predicted for a simple peristaltic model. This led Forouhar et al. (2006) to propose that the early zebrafish heart transcended peristalsis and switched to a hydro-impedance pumping model (Hickerson et al., 2005). Here we argue that although pumping in the zebrafish heart is reminiscent of the impedance model, it is still covered by the classic definition of peristalsis. Peristalsis stems from the Greek peri (enclosing or surrounding), plus stal (contraction or compression), plus sis (suffix), meaning a contraction that originates in the outside to compress what is inside (“The American Heritage Dictionary of the English Language,” 2000). Thus, the traditional concept of peristalsis is an inclusive one, and encompasses all kinds of propagated contractions that mix and propel, forwards and backwards, the contents of a hollow tube (Cannon, 1911; Weems, 1982, 1987) including the mechanical mode of operation typical of man-made positive displacement pumps such as peristaltic roller pumps (COMSOL Multiphysics Modeling Guide). Another adaptation that fights backflow in peristaltic pumps is found in the early embryonic and peristaltic avian heart, which displays alternation between segments

PART  |  1  Heart Evolution

of high (atrium and ventricle) and low conduction velocity (sinus venosus, atrioventricular canal and outflow tract) (see Chapter 1.5, Volume I). This arrangement establishes segments of fast and slow contraction that work together to maximize pumping efficiency by creating functional sphincters that reduce backflow. Whenever a fast contracting segment such as the atrium or the ventricle is activated, most of the flow is directed forward and very little of it escapes backwards, because of the delayed relaxation of the preceding slow conduction segment, such as the sinus venosus, or the atrioventricular canal (de Jong et al., 1992). Energy loss due to mechanical resistance is also minimized by the delayed activation of the slow-conducting cardiac segments downstream from the atrium and ventricle (atrioventricular canal and outflow tract, respectively). An alternative approach for the problem of backflow was the development of unidirectional valves. Valves play key roles in vertebrates and arthropods (Maynard, 1960; Guyton and Hall, 2000; McGaw and Reiber, 2002). Centipedes and millipedes (Myriapods) are perhaps the most eloquent examples of how valves were fully integrated into the segmental body plan of arthropods. The dorsal vessels of these animals span almost the full animal anterior–posterior (AP) axis and each of their many consecutive segments is separated by one flap valve (Lewis, 1981; McMahon et al., 1997b). Also, contrary to common thinking, valves are pervasive in the circulatory systems of hemichordates, echinoderms, molluscs and annelids (Binyon, 1972; Martin, 1980; Jones, 1983). Their different morphologies and tissue composition suggest that they are convergent solutions to the problems of backflow. There are at least three adaptations that transformed the typical ancestral peristaltic operation of dorsal vessels into a more efficient mode characterized by simultaneous or nearly simultaneous contraction. Two of these adaptations are present in arthropods such as chelicerates and crustaceans (Box 4; Fig. 6). The first one was the establishment of neural control over a presumably ancestral myogenic pattern of contraction. Neural control is represented by direct innervation from the central nervous system, by local innervation from peripheral ganglia attached to the dorsal surface of dorsal vessel, or by both. The activation patterns established by neural control are considerably faster and more coordinated than the myogenic rhythms that spread via low resistance junctions among muscle cells (Greenberg, 1979; Wilkens, 1999; Hertel and Pass, 2002 and references therein). The second arthropod adaptation was the centralization of pump work into a single compartment that often displays a thicker layer of contractile myocytes than the longer but thinner ancestral dorsal vessel. This adaptation is characteristic of decapod crustaceans, and it may have played a significant role in their evolution by allowing the appearance of large and very active animals such as giant lobsters (e.g., Palinurus barbarae) and coconut crabs (Birgus latro). Finally, the third major adaptation that transformed ancestral peristaltic vessels into more efficient pumps was the

19

Chapter  |  1.1  Evolutionary Origins of Hearts

evolution of chambered hearts in vertebrates and in molluscs. Chambered hearts represented a leap from the peristaltic vessels. Cardiac chambers revolutionized pumping strategies because they improved two of the most important problems of peristaltic vessels: the lack of coordination between inflow to the pump and the outflow from it, as well as the inefficient usage of energy by propagated peristaltic contractions (see discussion above, Fig. 7B). Vertebrate hearts broke up pump work into inflow and outflow functions, and assigned them to dedicated compartments separated by one-way valves. These compartments, or chambers, operate as reservoirs, or perform the major propulsive roles (e.g., atria and ventricles, respectively). By doing this, vertebrates greatly reduced the interference between inflow and outflow that is often manifested in peristaltic pumps by mechanical restriction to flow, potential reflections and backflow (Fig. 7B). Moreover, by creating compartments where contraction is synchronistic, vertebrate hearts eliminated the loss of energy in the distension of relaxed, downstream muscular segments (Simoes-Costa et al., 2005) (see Satchell, 1991b for a similar opinion), producing much stronger outputs than possible with the peristaltic design (Wilkens, 1999). In summary, bilaterian animals seem to have utilized a host of ingenious adaptations to circumvent the limitations imposed by the ancestral peristaltic design. Although it is possible to identify some common themes, it is evident that each evolutionary lineage used a different combination of strategies, further suggesting the independent, but parallel, nature of these adaptations across bilaterians. In the preceding sections we proposed an organized hierarchy of concepts to understand the relationship between the multiple circulatory pumps of metazoans and reviewed the various ideas about the origin of their specialized pumping organs. Now we direct our attention to the evolution of deuterostome pumping organs, describing the various types of macropumps displayed by these animals and discussing the hypotheses that underlie the origin of chambered hearts in vertebrates.

IX.  The evolution of chambered hearts among deuterostomes IX.A.  Deuterostome Phylogenies Evolutionary relationships among extant3 deuterostomes are depicted in Fig. 2B. Deuterostomes are formed by two major groups, Chordata and Ambulacraria (Halanych 3. The study of deuterostome evolution will benefit from the inclusion of fossils (Jefferies, 1986). Carpoids, extinct animals that displayed calcite skeletons, are interpreted by most as basal echinoderms (Gee, 2001). However, they have also been interpreted as basal deuterostomes, echinoderms, hemichordates, cephalochordates, tunicates or vertebrates (Jefferies, 1986, 2001). This interpretation requires simultaneous loss of calcite skeletons in the three lineages that supposedly lead to tunicates, cephalochordates and vertebrates, plus hemichordates, and thus it is disputed on the grounds of parsimony (Gee, 2001).

et al., 1995). Chordates are animals with notochord: tunicates (urochordates), such as the sea squirt ascidian C. intestinalis (see Chapter 2.1, Volume I); cephalochordates such as the lancelet Branchiostoma floridae (Amphioxus); and vertebrates like Danio rerio (see Chapter 1.4, Volume I), Xenopus laevis (see Chapter 1.3, Volume I), M. musculus and Homo sapiens. The sister group of Chordata is Ambulacraria, a clade formed by Hemichordates (solitary acorn worms such as Saccoglossus kowalevskii and colonial pterobranchs) and Echinoderms (brittle stars, crinoids, sea cucumbers, sea stars and sea urchins, such as Strongylocentrotus purpuratus). The phylogeny depicted in Fig. 2B suggests that cephalochordates and vertebrates are more closely related to each other than to tunicates (Maisey, 1986; Schaeffer, 1987; Ruppert et al., 2004). This is consistent with the obvious similarities observed among the fish-like Amphioxus and the fossils of basal vertebrate fish such as Haikouichthys and Myllokunmingia (Shu et al., 1999, 2003), as well as with the dissimilar bodies and lifestyles of tunicates. Therefore, there is considerable morphological support for the view that vertebrates and cephalochordates form a sister group, and that tunicates are at the base of chordates (Maisey, 1986; Schaeffer, 1987; Ruppert et al., 2004). However, this traditional wisdom was defied by large-scale molecular phylogenetic studies that suggested that tunicates, rather than cephalochordates, are the sister group of vertebrates (Blair and Hedges, 2005; Philippe et al., 2005; Boulart et al., 2006; Delsuc, 2006). These results were received with understandable caution. Nonetheless, they are relevant because they were obtained in phylogenetic studies that incorporated modern strategies to reduce the systematic errors of earlier analyses. Foremost, they are consistent with an increasing number of instances in which vertebrate and tunicate characters have been demonstrated to be more closely related to each other than to those from cephalochordates, including the morphologies of their pumping organs (Ruppert, 2005; reviewed in Schubert et al., 2006). In summary, chordate phylogeny is still an open subject. In the following pages we will discuss the various pumping organs of deuterostomes and suggest scenarios for the evolution of vertebrate hearts based on the competing phylogenies discussed above.

IX.B.  The Pumping Organs of Deuterostomes IX.B.i.  Echinoderms Echinoderms display four circulatory systems: coelomic; water vascular; perihemal; and hemal (Hyman, 1955; Martin and Johansen, 1965; McMahon et al., 1997a; Brusca and Brusca, 2003). Micropumps (flagella) drive fluid in the coelomic and water vascular systems, an echinoderm innovation that generates suction at tube feet, which in turn are primarily utilized in locomotion (Martin and Johansen, 1965; Brusca and Brusca, 2003). The perihemal system is a coelomic space that encloses hemal

20

vessels and connects coelom, hemal and water vascular systems (Hynman, 1940; Boolootian and Campbell, 1964; Millott, 1966; McMahon et al., 1997a). The hemal system of brittle stars, crinoids, sea stars and sea urchins is of particular interest, because it is associated with some rather specialized pumping organs. In contrast, holo­ thurians (sea cucumbers) display a complex but primitively designed hemal system with dorsal and ventral vessels, as well as multiple commissural peristaltic pumps in intestinal and hindgut respiratory outgrowth plexus (Herreid et al., 1976) (Fig. 5). The most interesting macropump of echinoderms is the two-chambered hemal pump of S. purpuratus (Fig. 8A). This pumping organ lies at the aboral end of the axial organ, a complex structure that connects oral and aboral vascular rings from the hemal system (Hyman, 1955; Rupert et al., 2004). The pumping organ receives fluid from rectal hemal vessels and contains two compartments: a larger, flat, thin-walled, chamber that is connected through a valveless ostium to a round, more

PART  |  1  Heart Evolution

conspicuous, chamber, which ends in a bifurcating pulsating vessel (Boolootian and Campbell, 1964) (Fig. 8B). The tri-dimensional organization of this organ is vaguely reminiscent of the two chambered, linear hearts of gastropods (Ripplinger, 1957). Its contractions were suggested to begin in the reservoir-like chamber and progress to the thicker walled efflux chamber connected to a contractile outflow tract (Boolootian and Campbell, 1964) (Fig. 8A) (but see Farmanfarmaian, 1968; McMahon et al., 1997a). The S. purpuratus chambered pump was described as a hemal heart by Boolootian and Campbell (1964). This was followed by controversy, with contrary views generally based on strict interpretations (especially in animals with redundant circulatory systems) of how a circulatory system should behave (Millott and Vevers, 1964; Farmanfarmaian, 1968). The axial pump status as a chambered heart has also been disputed on the grounds that flow in the hemal system may be bidirectional (McMahon et al., 1997a), or on the fact that the axial organ may play other multiple functions besides circulation, such as excretion, defense, reproduction

Figure 8  Larval and adult pumping organs of echinoderms and hemichordates (Ambulacraria). (A) A schematic view of Strongylocentrotus purpuratus (sea urchin) represents the axial organ connecting to oral and aboral hemal rings. (B) Within the axial organ lies a two-chambered hemal pump that ends in a bifurcating pulsating vessel. (C) Example of a larva from Ambulacraria. A pulsatile vesicle propels the fluid from the blastocoel to the hydropore. This larval organ is thought to persist in adults to give rise to the axial organ of echinoderms and the proboscis organ of hemichordates, structures thought to play excretory functions. (D) A lateral view of the acorn worm Saccoglossus kowalevskii showing the proboscis organ (or heart– kidney). (E) Cross-section at the proboscis (indicated by dotted lines at D) shows a hemal sinus sandwiched between a dorsal pericardial sac and the ventral stomochord. Pericardial contractions are thought to squeeze the contents of the hemal sinus, providing increased pressures for ultrafiltration at proboscis capillaries. S. purpuratus axial organ and chambered pump based on Boolootian and Campbell (1964). Ambulacraria larva and S. kowalevskii schemes modified from Ruppert and Balser (1986), and Balser and Ruppert (1990).

21

Chapter  |  1.1  Evolutionary Origins of Hearts

and pigmentation (Hyman, 1955; Millott, 1966; Grimmer and Holland, 1979; McMahon et al., 1997a). Here we take a favorable view of the opinion that the S. purpuratus axial pump is a chambered heart, and suggest that its physiology may be revisited in light of a broader view of circulatory systems as fluid transport systems, whether they are recirculating, unidirectional or bi­directional (LaBarbera and Vogel, 1982; LaBarbera, 1990). The presence of a potential chambered heart in echinoderms invites questions about its functions and its evolutionary relationships with other deuterostome pumping organs. Clues about the function of the hemal pump of S. purpuratus come from the presence of potentially homologous organs in the sea star Asterias forbesi and in the hemichordate acorn worm S. kowalevskii (Ruppert and Balser, 1986; McMahon et al., 1997a) (see below), as well as from the presence of numerous ostia in the walls of both chambers. The latter may be analogous to the pores on the walls of branchial hearts of cephalopods through which fluid is filtrated into the pericardium, suggesting that they have excretory functions (Wells, 1983; Burggren and Keller, 1997). All these facts suggest the hemal pump of S. purpuratus is derived from larval excretory structures that are conserved in adult animals (Ruppert and Balser, 1986; Balser and Ruppert, 1990) (Fig. 8B–C). In summary, the hemal system of echinoderms not only possesses multiple peristaltic pumps on the walls of vessels, as in holothurians (Tiedemann, 1816; Farmanfarmaian, 1968; Herreid et al., 1976) (Fig. 5), but also has a rather specialized pumping organ in sea stars and sea urchins (Boolootian and Campbell, 1964; Ruppert and Barnes, 1994) (Fig. 8A–B).

IX.B.ii.  Hemichordates In contrast to the highly derived, multiple circulatory systems of echinoderms, hemichordates display the common arrangement of dorsal and ventral vessels bridged by transverse channels (Ruppert and Balser, 1986) (Fig. 5). The dorsal vessel is a peristaltic pump. It drives fluid in the anterior direction, while the ventral vessel transports fluid in the caudal direction, contrasting with the orientation of the circulation in cephalochordates and vertebrates (Nübler-Jung and Arendt, 1996). Enteropneust acorn worms and colonial pterobranchs display another interesting pumping organ, which is loosely referred as a heart, a heart–kidney or a cardiac vesicle (Ruppert and Balser, 1986; McMahon et al., 1997a; Brusca and Brusca, 2003; Lowe et al., 2006). This organ is located in the proboscis and employs an original pumping mechanism. It consists of a central hemal sinus (the heart sinus, or central sinus) sandwiched between a dorsal pericardial sac and ventral stomochord, an evagination of the pharynx that lies over the rigid proboscis skeleton (Ruppert and Balser, 1986) (Fig. 8D). The pericardial sac is formed by a layer of

striated muscular cells and encloses the central sinus. The ventral wall of the pericardial sac is firmly attached to the stomochord and, as such, contractions of the pericardium squeeze the contents of the central sinus to boost the pressures generated by dorsal vessel contractions (Fig. 8D–E). These increased pressures are thought to drive ultrafiltration from a network of capillary vessels in the proboscis coelom, producing an ultrafiltrate that is excreted to the exterior via one or two ciliated tubes and pores in the proboscis (Balser and Ruppert, 1990) (Fig. 8D–E). In summary, the circulation of hemichordates is powered by a decentralized system of macropumps that includes a main dorsal peristaltic vessel and the proboscis organ. Although the latter may contribute to circulatory work, the evidence suggests that it is primarily a boosting device attached to an excretory function.

IX.B.iii.  Echinoderms and Hemichordates:   A Synthesis The sister-group relationship between echinoderms and hemichordates (Ambulacraria) is based on morphological and molecular characters (Metchnikoff, 1881; Adoutte et al., 2000; Halanych, 2004; Bourlat et al., 2006). The classic morphological character uniting both phyla is the bilaterally symmetric larvae from hemichordates, and in echinoderms, which contains a pumping organ connected to an excretory system, the larval heart–kidney. This larval organ is retained in adult stages of both phyla and is thought to give rise to the homologous pumping organs of the axial complex of sea stars and the proboscis organ of enteropneusts and pterobranchs (Fig. 8). In summary, hemichordates and echinoderms possess special pumping organs derived from larval excretory organs that do not belong to the main lineage of dorsal and ventral pumping organs that gave rise to the pumping organs of chordates. This interpretation is consistent with recent experiments indicating that the developing hemichordate proboscis organ does not express Nkx2-5/2-3 orthologs (Lowe et al., 2006).

IX.B.iv.  Cephalochordates The circulatory system of Amphioxus is simple and shares primitive attributes with deuterostomes such as hemichordates and holothuroid echinoderms and protostomes such as annelids, lophophorate phoronids and echiurans (Brusca and Brusca, 2003). The circulatory system of Amphioxus shows: (1) dorsal and ventral vessels; (2) commissural vessels between dorsal and ventral vessels; (3) multiple contractile vessels; and (4) a stereotyped system of large conducting channels that resembles the major vessels of vertebrates (Fig. 9A). Endothelial-like cells have been reported in Amphioxus (Moller and Philpott, 1973; Moller and Ellis, 1974; Silva

22

PART  |  1  Heart Evolution

Figure 9  (A) The specialized pumping organs of chordates. Top: scheme of circulation in the basal vertebrate Myxine glutinosa, the hagfish. The hagfish displays a main chambered pump that is enclosed by an open pericardial sac. The heart is often described as a branchial heart, due to its position upstream from the branchial circuit. Arrows identify flow direction (posterior in ventral vessels, anterior in dorsal vessels). The hagfish circulation is characterized by accessory pumps, which include a portal pump powered by cardiac, atrial-like myocytes, as well as pumps actuated by striated skeletal muscle such as cardinal, caudal and pharyngeal (not shown). Hagfish and lampreys display large sinuses spanning the anterior–posterior axis, which accounts for their very high plasma volume/body weight ratios. Middle: circulation in Branchiostoma floridae (Amphioxus). The Amphioxus does not have a chambered heart. Its circulation is driven by four major peristaltic vessels (red arrows) that are powered by nonstriated, smooth-like muscle and are not enclosed by a pericardium. Flow direction is homologous to the vertebrate condition, but sometimes retrograde waves are recorded. Bottom: circulation in the ascidian Ciona intestinalis is driven by a peristaltic pump that is powered by striated muscle and is encased in pericardium. The tunicate circulation is bidirectional due to the presence of two pacemakers at each pump extremity (doubled headed red arrow). M. glutinosa circulation drawn after Cole (1926); Augustinsson et al. (1956); Hol and Johansen (1960); Chapman et al. (1963); Satchell (1992); Janvier (1996). Schemes of tunicate and cephalochordate circulation modified from Simoes-Costa et al. (2005).

Chapter  |  1.1  Evolutionary Origins of Hearts

23

Figure 9   (Continued) (B) The traditional phylogeny places cephalochordates as the sister group of vertebrates and tunicates at the base of chordates. This suggests that the ancestral chordate pump was tunicate-like with striated myocytes and pericardium. In this view, cephalochordates regressed to a decentralized system of peristaltic pumps powered by smooth muscle, while vertebrates improved the ancestral plan developing cardiac chambers and displaying increased concentration of circulatory work in the branchial heart. The alternative chordate phylogeny places tunicates and a main tunicate-like pump for the chordate ancestral. The alternative phylogeny places tunicates as the sister group of vertebrates and cephalochordates at the base of chordates. This suggests that the seemingly primitive pumps of cephalochordates are indeed representative of the ancestral chordate body plan. In this view the last common ancestor of tunicates and vertebrates evolved a more efficient main pump that was encased in pericardium and powered by striated muscle. Tunicates concentrated all circulatory work into this pump, while vertebrates developed cardiac chambers and progressively concentrated circulatory work into a chambered heart.

et al., 1995). They typically occur in major vessels, in which they occupy a large surface but fail to develop junctional complexes or to form a continuous sheet of cells, which may underlie the suggestion that these cells are just adherent amebocytes (Rahr, 1981). This interpretation carries the danger of confusing endothelial cells (a cell type) with endothelium (a cell layer), which may obscure potential origins of vertebrate endothelial cells from protoendothelial cells that adhere to the basal lamina of invertebrate chordates, a possibility acknowledged in Rahr (1981) and formally proposed by Munoz-Chapuli et al. (2005). Amphioxus does not have a chambered heart (Hyman, 1942; Carter, 1967; Moller and Philpott, 1973; Rahr, 1979; Randall and Davie, 1980; Jefferies, 1986). Instead, the Amphioxus circulation is powered by multiple peristaltic pumps; the subintestinal, portal and hepatic veins, as well as the endostylar artery (Fig. 9A). These peristaltic pumps display a type of contractile cell in their walls that does not show organized sarcomeres, and thus resembles smooth muscle (Hirakow, 1985). This constitutes a ­ fundamental difference between the vessels that move circulation in Amphioxus and the chambered hearts of vertebrates, which are powered by striated muscles. The decentralized system of peristaltic pumps imposes a preferred orientation to the blood, which is homologous to the one observed in vertebrates (i.e., posterior in dorsal vessels and anterior in ventral vessels) (Fig. 9A). Nonetheless, peristaltic waves running in the opposite direction have also been recorded in the vessels of Amphioxus (Skramlik, 1938). It remains

to be established, however, whether bilateral pumping is a normal feature of the Amphioxus circulation, as it is with tunicates (see below). The peristaltic vessels of Amphioxus are derivatives of somitic mesoderm, which evaginates bilaterally to generate the visceral peritoneum. Both limbs of the visceral peritoneum join ventral to the digestive tract to form the subintestinal vessel, which expands in the A–P axis. Holland et al. (2003) suggested that growth of hepatic tissues separates the intestinal vessel to generate the major peristaltic vessels of Amphioxus; the subintestinal, portal and hepatic veins, as well as the endostylar artery. A similar mechanism was proposed for the larval lamprey (Keiser, 1914; Baxter, 1957). The particular pump layout of Amphioxus raises the issue of homology between its decentralized system of smooth muscle-like powered peristaltic vessels and the centralized system of striated pumps displayed by tunicates and vertebrates. However, the common theme of bilateral precursors joining at the midline to form a vessel that expresses AmphiNk2-tin, an Amphioxus NK2 gene (Holland et al., 2003), strongly argues in favor of homo­ logy, suggesting that the peristaltic vessels of Amphioxus belong to the same lineage of pumping organs that descend from ventral and dorsal vessels of the bilaterian ancestor. A more contentious subject refers to the nature of the socalled sinus venosus (SV) in Amphioxus. The Amphioxus SV is located at the confluence of major returning veins (Fig. 9A), and thus superficially resembles the vertebrate SV.

24

Also, contrary to what we suggested in Simoes-Costa et al. (2005), there is evidence that the Amphioxus SV contracts (Skramlik, 1938). The controversial SV of Amphioxus constitutes the only support for the presence of a chambered heart in cephalochordates. However, the evidence favoring a cardiac affinity for the Amphioxus SV has always been feeble. While some authors are cautious about this possibility (Randall and Davie, 1980; Rahr, 1981; Jefferies, 1986), others are against it (Moller and Philpott, 1973). Indeed, the anatomical position of the Amphioxus SV is not homologous to its vertebrate counterpart. In vertebrates, the SV is the posterior-most chamber, while in Amphioxus the SV is not the posterior-most segment, but is located between the endostylar artery and the hepatic vein (Simoes-Costa et al., 2005) (Fig. 9A). Therefore, there is little support for a smooth muscle-like driven, sinus venosus-only chambered heart in Amphioxus. The source of confusion over a possible cardiac nature of the Amphioxus SV may have originated from the convention of naming blood reservoirs upstream from the vertebrate heart as “SV” (Kardong, 2002). The connection between the Amphioxus SV and a chambered heart can not be summarily dismissed, however. It remains formally possible that, rather than being the ancestral cardiac chamber the SV is the vestige of a chambered heart from an ancestor, which regressed in Amphioxus (see discussion below). There are multiple hints that the Amphioxus body is derived and that it may have been secondarily simplified (Conway-Morris, 2000; reviewed in Simoes-Costa et al., 2005). In fact, the idea that the Amphioxus SV may represent a cardiac vestige is a plausible solution to the seemingly unlikely possibility that the weak pumping performance of the vertebrate SV (which is consistent with its reservoir function) and its sparse cardiac muscle content in most vertebrates (Farrell and Jones, 1992), was selected as a first choice for a cardiac chamber. In summary, the primitive-looking circulation of Amphioxus is still an enigma because it is compatible with either a regressed, degenerated condition, or with a basal, ancestral circulatory state.

IX.B.v.  Tunicates The tunicate circulation can be depicted as a centralized system with one single tubular pump that alternates direction, driving fluid through two large vessels that are attached to both its ends. The large vessels that branch out of the tunicate pump do so to form smaller vascular channels throughout the body of the animal (Fig. 9A). Overall, the vascular organization of tunicates constitutes a mixture of well-organized, stereotyped, networks of vessels (e.g., the vessels that surround the ascidian pharyngeal basket) and some poorly-defined sinuses. A continuous lining of endothelial cells has been implied by Robb (1965), but never demonstrated in tunicates (Ruppert et al., 2004). The tunicate pump is a ventral organ located just posterior to the pharynx and anterior to the stomach. It shares

PART  |  1  Heart Evolution

some important characteristics with vertebrate hearts (Passamaneck and Halanyck, 2006), but it is not a chambered pump. It is a valveless, peristaltic tube that may also be envisaged as a single compartment pump (Davidson et al., 2006 and see below). In the ascidian Ciona pump and in many other tunicate pumps, there is no distinctive A–P polarity. However, in C. intestinalis there is a dorsal– ­ventral axis distinct from early stages and the adult myocardial tube is V-shaped (Robb, 1965; Kriebel, 1968). This may constitute an adaptation to circumvent some of the problems of fluid stream reflections or mechanical recoil associated with linear pumping organs, such as the gastropod heart (Kilner et al., 2000; Simoes-Costa et al., 2005). However, in many other tunicates, the peristaltic pump is C-shaped (e.g., Ecteinascidia) or linear (e.g., Molgula, Oikopleura), suggesting that this simpler layout may reflect the ancestral character of tunicate pumps (Jones, 1971b). Appendicularians (or larvaceans) constitute an exception among tunicates, since they frequently display rudimentary peristaltic pumps or, as in the Kowalevskiidae family, completely lack pumps (Alldredge, 1976). However, it is probable that the underdeveloped or missing circulatory pumps of these tunicates result from assumption of small body sizes, which is often understood to be a derived character linked to the adoption of a planktonic lifestyle (Zeng and Swalla, 2005). The tunicate pump is encased within a pericardial coelom and is powered by contractile cells that display organized sarcomeres. Bidirectional pumping is characteristic, but not exclusive to tunicates (Fig. 9A). For some time this ability of the ascidian pump was thought to indicate a fundamental difference between the pumps of tunicates and vertebrates, when in truth it only reflects the presence of two myogenic pacemakers, one at either end of the heart (Anderson, 1968). Histological and ultrastructural details of the tunicate pump and pericardium are best understood in C. intestinalis (reviewed in Davidson, 2007). Pericardial and striated muscle cells derive from a single continuous epithelial tube invaginated along the dorsal side, which may also have raised suspicion about homology between tunicate and vertebrate pumps (Harvey, 1996). In truth, the pericardial origin of the tunicate pump is actually one more clue to the common coelomic origin of all bilaterian pumps (see discussion above). The contractile cells of the tunicate pump are of the myoepithelial type. Their myofilaments do not form fibers but, rather, a compact field that is located towards the luminal surface, while the nucleus is opposite and faces the pericardium (Davidson, 2007). The presence of pumping organs in sessile, planktonic, sometimes diminutive, animals such as tunicates constitutes an apparent contradiction of the general associations between size, activity, lifestyles and the presence of pumping organs observed in bilaterians (Box 1; Table 1). It is not known why most tunicates need such vigorous pumps. However, the answer may relate to the defining quality

25

Chapter  |  1.1  Evolutionary Origins of Hearts

of tunicates, the tunic itself. The tunic is a living tissue, which in some species contains blood cells and blood vessels (Ruppert et al., 2004). It is supported by a scaffold of extracellular fibers of tunicin, a cellulose polymer synthesized from the cellulose synthase gene Ces-A4. The tunic is secreted by the ectoderm and grows continually with the animal, incorporating minerals and other materials that confer them variable textures and mechanical resistances (Ruppert et al., 2004). Therefore, it is possible that the continuous need to mobilize and distribute precursors, as well as to produce and secrete cellulose and other materials for the thaliacean “house” (a probable homolog of the tunic) maintained the demand for an active pump that was already present in an ancestor.

IX.C.  Vertebrates A description of the major types of vertebrate hearts, their developmental patterns and molecular mechanisms of morphogenesis is beyond the scope of this review. Here, we will analyze the fossil evidence related to primitive vertebrate hearts, as well as the morphological and developmental patterns of basal extant vertebrates (hagfish and lampreys) in search of concepts that may be useful to reconstruct the primitive vertebrate pattern.

IX.C.i.  Vertebrate Fossils Unfortunately, soft tissues only fossilize under extremely special circumstances. However, the relatively rigid pericardial membranes can leave impressions of the cardiac chamber outlines, such that round cavities are sometimes found posterior to gill slits, endostyle, or vessels in the fossils of fish or animals of disputed affinity, such as Haikouella lanceolata (Janvier, 1996; Chen et al., 1999; Mallatt and Chen, 2003; Shu et al. 2003). These impressions are, however, of insufficient quality to define a heart, much less to allow the definition of chamber identities. Moreover, the heart chambers of primitive vertebrates, such as hagfish and lampreys, display considerable overlap in the A–P and dorsal–ventral axes (Fange, 1972; Janvier, 1996), which may blur chamber outlines and further compound the already problematic identification of hearts in fossils. In summary, fossils of hearts, or of pumping organs, are extremely scarce (Chen et al., 1999; Shu et al., 2003), difficult to interpret (Mallatt and Chen, 2003), controversial (Fisher et al., 2000; Rowe et al., 2001) and thus, are not very informative. 4. Ces-A seems to have been horizontally-acquired by the last common ancestor of tunicates from a bacterial symbiont, making ascidians, thaliaceans and larvaceans the only animals to produce cellulose (Nakashima et al., 2004 and references therein).

IX.D.  Extant Vertebrates Vertebrate circulatory patterns should be considered in the context of the evolution of a very particular model of circulatory system. The vertebrate circulation is committed to a regurgitation-proof, unidirectional pumping model. In this model intravascular fluid volume is minimized but vigorously circulated inside a well-defined network of more or less leak-proof vessels, which display stereo­ typed branching patterns that optimize global transport and guarantee efficient exchange between blood and tissues (LaBarbera, 1990). Overall, vertebrates followed a common trend in the evolution of phyla that display large and/or active species with complex behaviors; a tendency to concentrate circulatory work into a main pump or in special segments of a major pump (e.g., as in decapod crustaceans and insects). The circulation of hagfish, an early vertebrate, contains multiple ingenious assisting devices that are actuated by skeletal muscles (cardinal and caudal pumps), but that also include a portal pump powered by a type of myocardium similar to atrial tissue (Satchell, 1991c and references therein) (Fig. 9A). In contrast, modern vertebrates have reduced their dependence on accessory pumps (Janvier, 1996) and consolidated the “branchial” heart as the major pump. However, this trend in vertebrates has not been complete, since many fish and amphibian species have retained these structures (Satoh and Nitratori, 1980; Satchell, 1991c). Since vertebrates originated in a marine environment, the original cardiac chambers were defined in the context of fish anatomy (Goodrich, 1930; Romer, 1962; Randall, 1970; Fange, 1972; Bourne, 1980; Kardong, 2002). Therefore, there are four conspicuous cardiac compartments traditionally regarded as chambers: sinus venosus; atrium; ventricle; and conus arteriosus. In fish and most other vertebrates, these cardiac segments are located inside the pericardial ­ cavity, guarded by inflow and outflow valves invested with striated cardiac muscle, and contract nearly simultaneously.5 Although these four criteria seem simple enough to qualify a given cardiac segment as a chamber, the concept of a cardiac chamber has been a matter of intense controversy. The problem arises because cardiac segments maintain a dynamic relationship with themselves and with the tissues that surround them. During evolution they have regressed, merged into others, or divided into left, right or more compartments (Simoes-Costa et al., 2005). As an example of this dynamic relationship, Moorman and Christoffels (2003) performed three-dimensional reconstruction in mouse embryos, which show that the SV is a poorly-developed cardiac segment, and that the myocardium of the outflow tract never forms anatomical chambers. This indicates that only atria and ventricles function as true chambers in mice (Christoffels et al., 2000). 5. The fish sinus venosus is bounded by a large sinoatrial valve and there are valves between the sinus venosus and the hepatic veins, but not between the sinus and the ducts of Cuvier (Farrel and Jones, 1992).

26

PART  |  1  Heart Evolution

An especially problematic cardiac segment is the bulbus arteriosus, an elastic cardiac segment located inside the pericardial cavity. The bulbus arteriosus contrasts with the massively muscular, valved, and obviously chambered conus arteriosus, because it does not operate as a propulsive chamber, but acts to damp pressure variations induced by the cyclic ventricular contractions (Farrell and Jones, 1992). Since the bulbus arteriosus contains smooth muscle, it has been suggested that it is an extension of the aortic trunk (Lawson, 1979), or that its smooth muscle character is secondary to a transdifferentiation from an original cardiac muscle pheno­ type (Hu et al., 2000; Guerrero et al., 2004). Recent data in the zebrafish indicates that the bulbus arteriosus does not express cardiac or muscle genes during embryogenesis, and only activates smooth muscle markers in juveniles (Grimes et al., 2006). It is likely that this subject will remain controversial. Part of the problem is the expectation that the extremely variable range of fish phenotypes can ever be represented by one particular species. Despite this, current evidence supports the view that the bulbus arteriosus is not a chamber (Grimes et al., 2006), and that sinus venosus, atrium, ventricle and conus arteriosus are the original four vertebrate chambers. In summary, the evolution of cardiac chambers can be regarded as one extremely important component of the multiple circulatory changes that took place in vertebrates and enabled them with the versatility required to occupy their current niches (Gans and Northcutt, 1983).

Amphioxus and vertebrates, such that most major vessels in Amphioxus are easily homologized to their vertebrate counterparts (Moller and Philpott, 1973; Rahr, 1979) (Fig. 9A). However, the multiple contractile vessels in Amphioxus contrast with the main pumping organs of tunicates and vertebrates. In addition, ultra structural studies indicate that contractile vessels in Amphioxus lack striated muscle, a continuous endothelium, and are poorly-separated from connective tissues, features that link Amphioxus to basal deuterostomes (Rahr, 1981; Hirakow, 1985). We have recently argued that neither Amphioxus, nor C. intestinalis alone provide all elements to infer the ancestral, vertebrate circulatory system (Simoes-Costa et al., 2005). C. intestinalis with a circulatory system centralized in its main pump will likely be a much better model for the ancestral vertebrate pump than the multiple contractile vessels of Amphioxus. Conversely, Amphioxus is the closest chordate model for the vertebrate vascular plan. However, this does not mean that the study of Amphioxus pumps or of tunicate vessels should be neglected. On the contrary, a thorough understanding of the gene regulatory networks that underlie development of these structures will be necessary to establish whether they reflect secondary simplifications or ancestral designs.

X.  Chordates: a synthesis

XI.A.  The Problem

The circulation of tunicates and of cephalochordates share characteristics that are also present in vertebrates. All chordates place their main pumps upstream from pharyn­ geal/gill vessels and, with the exception of tunicates which display regular bidirectional pumping, their dorsal vessels carry blood posteriorly and ventral vessels carry blood anteriorly. This contrasts with enteropneust hemichordates and protostomes such as annelids, lophophorates or echiurans, in which flow direction is reversed (NüblerJung and Arendt, 1996; Brusca and Brusca, 2003). The centralized circulatory system of tunicates resembles that of vertebrates, which gradually consolidated the heart as their main pump (sometimes described as the “branchial” heart because of its position upstream of the branchial circuit) (Augustinsson et al., 1956; Hardisty, 1979; Satchell, 1991d; Kardong, 2006) (Fig. 9A). Another interesting parallel between tunicates and some vertebrates is the encasing of their pumps in stiff pericardial cavities, an adaptation designed to match outflow with inflow (Jones, 1983; Schmidt-Nielsen, 1990; Percy and Potter, 1991; reviewed in Farrell and Jones, 1992) (Fig. 9B). Anatomical and optic microscopic descriptions indicate evident similarities between vascular architectures of

Invertebrate chordates are sessile, planktonic (tunicates), or sedentary filter feeders such as Amphioxus. On the other hand, vertebrates are bigger, more active, have evolved highly-elaborate predatory or other behaviors, developed homeothermy and successfully occupied an extremely varied range of niches on sea, land and air. This major vertebrate leap from marine invertebrate chordate lifestyles has been attributed primarily to the development of a new head, with a new brain and sensory placodes, coupled with a host of other adaptations that include centralization of circulatory work (Gans and Northcutt, 1983). However, it is difficult to understand how active predation, or other dynamic behaviors attributed to vertebrates, could have developed only with a simple centralization of pumping functions (Gans and Northcutt, 1983). Therefore, the increase in tridimensional sophistication and pumping efficiency represented in the evolution of the vertebrate chambered pump was, likely, a key factor in the evolution of vertebrates. When examining the pumping organs of chordates it is evident that there is a large evolutionary gap between the peristaltic pumps of tunicates and cephalochordates and the chambered hearts of vertebrates (Figs 5, 9A). Even the most primitive extant vertebrates, such as hagfish and

XI.  The evolutionary origin of cardiac chambers

27

Chapter  |  1.1  Evolutionary Origins of Hearts

lampreys, already display three or four cardiac chambers. Such leaps in morphology and design are common in the study of character evolution. They are often explained away by the familiar idea that the critical morphological transitions may have occurred in small and isolated populations that preceded major radiations, and whose traces may not have been preserved in the fossil record (Eldredge and Gould, 1972). However, there may be more heuristic interpretations for the abrupt evolutionary transitions between these organs, as we discuss below.

XII.  Chordate phylogenies and the origin of vertebrate hearts XII.A.  Earlier Views The improvement in circulatory abilities among chordates can only be appreciated in a phylogenetic context. In the traditional chordate phylogeny, tunicates are basal and cephalochordates constitute the sister group of vertebrates (Maisey, 1986; Schaeffer, 1987; Ruppert et al., 2004). Based on this phylogeny, Harvey (1996) suggested that the vertebrate heart originated from pumping organs similar to those that equip the Amphioxus. Fishman and Chien (1997) proposed a similar scenario where an organ akin to the peristaltic endostylar artery of Amphioxus was the immediate ancestor of the vertebrate heart, while a tubular, valveless organ related to those of tunicates was suggested as the ancestor chordate pumping organ (Fig. 9B). These earlier ideas were proposed before the availability of multiple chordate and deuterostome genomes when there was still doubt about the evolutionary relationship between vertebrate hearts and tunicate pumps, which were thought to belong to a different evolutionary lineage by virtue of its characteristic bi­directional pumping and pericardial origins (Harvey, 1996). Therefore, this may have prevented analysis of alternative scenarios (reviewed in Simoes-Costa et al., 2005). However, the unveiling of tunicate genomes and recent developmental studies in C. intestinalis, indicates that tunicates and vertebrates utilize essentially the same kernel of transcription factors in the development of their pumps (Davidson, 2007). The scenario postulated by earlier views on the origin of vertebrate hearts is not easily reconciled with the implied idea of a progressive sophistication of pumping organs from invertebrate chordates to vertebrates (Fig. 9B). This is because it is difficult to imagine how the relatively advanced and vigorous, pericardially-enclosed striated pump of tunicates would be a precursor of the much simpler, smooth muscle-like vessels of Amphioxus (Hirakow, 1985). There are only two options for this conundrum: either the decentralized system of multiple smooth muscle-like driven peristaltic pumps of Amphioxus represents a regression to a more ancestral circulatory layout; or the traditional chordate

phylogeny is wrong (Fig. 9B). Evidence for these two possibilities will be discussed below.

XII.B.  Alternative Phylogenies for the Vertebrate Heart Simoes-Costa et al. (2005) analyzed vertebrate cardiac evolution in a broader evolutionary context that also took into consideration factors such as the muscular cell types that power chordate pumps, their pumping modes (e.g., peristaltic or synchronous), the presence of a pericardial casing, the centralization of pumping work (e.g., whether circulatory work is concentrated in one main pump or dispersed among several roughly equivalently powerful pumps), and the presence of a continuous endothelial lining. With these factors considered, they proposed a more parsimonious scenario where a main, tunicate-like pump equipped the chordate ancestor. This pump was retained in tunicates and in the cephalochordate/vertebrate ancestor. Cephalochordates reverted to a decentralized circulatory system, while vertebrates elaborated on the ancestral plan to generate inflow and outflow chambers. Pump design in chordates could thus be explained by a regression event in cephalochordates and a refinement of the ancestral plan in vertebrates (Fig. 9B). However, the support for a secondary simplification of pumping organs in Amphioxus was challenged by studies that proposed that cephalochordates, rather than tunicates, are at the base of chordates, and that vertebrates and tunicates form a sister group (Blair and Hedges, 2005; Philippe et al., 2005; see also Bourlat et al., 2006 and Delsuc et al., 2006). This prompted a re-evaluation of the concepts proposed by Simoes-Costa et al. (2005). We concluded that the suggestion of a sister group relationship between tunicates and vertebrates actually provides stimulating solutions, not only for many similarities between these taxa, but also for the clearly primitive circulatory design of cephalochordates (Schubert et al., 2006). The new chordate phylogeny suggests that the decentralized, smooth muscle-like-powered system of pumps of cephalochordates is primitive, rather than regressed. As such, the ancestor of tunicates and vertebrates may have focused its circulatory work in a main striated muscle pump encased in pericardium. Tunicates further centralized this system by eliminating all auxiliary pumps. Vertebrates improved the pump design by creating cardiac chambers, but maintained redundancy in accessory pumps that are present in many fish and amphibian species and, in special form, in hagfish (Fig. 9). In summary, while it is not possible to take sides on the basis of such limited evidence, the scenarios depicted in Fig. 9B suggest that the origins of our hearts could be traced back to an ancestral, tunicate-like, peristaltic pump project that was modified to give rise to a chambered heart. This important concept offers a number of exciting possibilities for the evolution of cardiac chambers.

28

PART  |  1  Heart Evolution

XIII.  Scrutinizing cardiac chamber evolution

XIII.C.  The Evolution of Chamber Myocardium

XIII.A.  Using the Typical Attributes of Cardiac Chambers as Handles

One of the essential features of cardiac chambers is the coordinated activation of myocytes to generate a nearly synchronic contraction at regularly-spaced intervals. This ability circumvents the limitations associated with peristaltic pumps (Fig. 7) and is a fundamental feature of chambered pumps. In vertebrates, this capacity is developmentally-regulated and manifested only after an initial peristaltic stage symbolized by the primordial cardiac tube. The primordial cardiac tube is formed by myocytes with pronounced automaticity, low contractility and low conduction velocity (de Jong, 1992), probably due to the sparse expression of the low conductance connexin 45 (Alcoléa et al., 1999; van Veen et al., 2001; Coppen and Severs, 2002). After the primordial tube stage, a distinct type of muscle appears at the ventral surface of the common ventricle and at the dorsolateral surface of the common atrium (Delorme et al., 1997; Christoffels et al., 2000). This chamber myocardium expresses a host of transcription factors, sarcomeric and cytoskeletal proteins, as well as sarcoplasmatic reticulum Ca ATPases (Christoffels et al., 2000). The concerted activation of these genes produces a more vigorous pump than the primitive tube. However, the crucial change from a mechanical standpoint is the advent of well-coordinated, nearly synchronistic, contractions, which presumably result from an increased density of high conductance connexins 40 and 43 (Delorme et al., 1997; van Venn et al., 2001). The transition from the primordial muscle type to the chamber myocardium is tightly regulated by T-box pathways that include Tbx5 (Bruneau et al., 2001), Tbx2, Tbx3 and Tbx20 (Habets et al., 2002; Hoogaars et al., 2003; Harrelson et al., 2004; Christoffels et al., 2004; Stennard et al., 2005). These T-box transcription factors are organized in competing genetic hierarchies that promote chamber myocardium (Tbx5), or keep myocytes at the primitive, peristaltic stage (Tbx2, Tbx3). Coordination between these opposing influences is provided by Tbx20, which represses Tbx2 and Tbx3 in the myocardial domains that give rise to chamber myocardium to (reviewed in Stennard and Harvey, 2005; see also Chapter 9). The detailed knowledge obtained about the molecular regulation of the transitions between the peristaltic and the chamber phenotypes suggests that this feature can be utilized as a handle for investigation of cardiac chamber origins.

The evolution of vertebrate cardiac chambers brought a number of advantages over the ancestral peristaltic way of propulsion. These advantages depend on crucial elements such as: a type of muscle that maintains regular operation and provides reserve for top performances; a pattern of muscular interconnection that guarantees the mechanical advantages of simultaneous or near-­simultaneous contraction; a division of circulatory work among units dedicated to inflow (reservoir) and outflow (pumping) functions (Simoes-Costa et al., 2005); and a set of unidirectional valves that avoid reflux and increase circulatory efficiency. The developmental bases for each of these features are reasonably understood to allow their use as entry points to explore the evolutionary origins of cardiac chambers.

XIII.B.  Gene Regulatory Networks and the Emergence of Cardiac Muscle The study of gene regulatory networks from circulatory pumps revealed that the morphological complexity that sets vertebrate hearts and cardiac muscle apart from invertebrate peristaltic pumps, such as the Drosophila dorsal vessel and their muscle cells, is matched by an increased complexity of vertebrate networks (Cripps and Olson, 2002). This difference in complexity between vertebrates and Drosophila stems from sequential gene duplications that occurred in the lineages leading to vertebrates and also from the substantial gene loss observed in ecdysozoans (Putnam et al., 2007). The increased number of players in vertebrate heart development is compounded by an increased complexity of regulation achieved by evolution of new cis-regulatory modules that altered the relationships between genes in these networks, and by co-option of new genes (reviewed in Fishman and Olson, 1997; Cripps and Olson, 2002; Olson, 2006). This fruitful approach established the fundamental notion that most similarities between vertebrate hearts and invertebrate peristaltic pumps concentrate at stages that correspond to the initial peristaltic stages of vertebrate cardiac development, setting the stage for more specific and detailed predictions about the organization of the anatomical and genetic features that characterize the mature vertebrate heart. Now, it is expected that the decoding of invertebrate chordate and ambulacrarian genomes will provide the opportunity to study how vertebrate cardiac gene regulatory networks evolved in their own deuterostome evolutionary context.

XIII.D.  The Inflow/Outflow Organization One of the major advantages of chambered hearts over peristaltic vessels is the division of circulatory work into two types of compartments dedicated to the functions of inflow (atria) and outflow (ventricles). The division of pump progenitors into these two broad domains

29

Chapter  |  1.1  Evolutionary Origins of Hearts

was likely the result of patterning mechanisms that operated early in development to generate diversity inside the pump field (see discussion below). It is probable that these hypothetical cardiac patterning mechanisms were adapted from signaling mechanisms that were already involved in embryonic early axial patterning (see Fishman and Chien, 1998). Therefore, it is useful to concentrate on embryonic signals that pattern the vertebrate cardiac field as targets for an evolutionary analysis. Patterning mechanisms are primary developmental events which could, conceivably, be conserved and thus are potentially useful to scrutinize the origin of cardiac chambers.

XIII.D.i.  One-Way Valves One-way valves are critical for optimal performance in chambered hearts, and they are also pervasive among other bilaterian pumping organs. Despite the paramount importance of valves for cardiac chamber function, we are still lacking a comprehensive evolutionary analysis of valve evolution. It is possible that the extreme diversity in valvular composition and morphology across bilaterians is inhibiting such analyses (Roger Markwald, Medical University of South Carolina, personal communication).

XIV.  The path to a chambered heart The sequence of events leading to the formation of the ancestral vertebrate chambered heart is unknown. However, recent work suggests that subtle changes in the developmental processes that commit cells to a pump muscle fate may generate enough diversity within the pump progenitor field, allowing the opportunity for evolution of multiple cardiac chambers (Davidson et al., 2006; see below and Chapter 2.1, Volume I). By artificially increasing the strength of FGF signaling with a constitutively active form of ETS1/2, Davidson and colleagues increased recruitment to a pump fate at the expense of the skeletal muscle fate in Ciona intestinalis. The manipulation induced an extra pump compartment in a minority of transgenic embryos. In some cases, this second pump compartment worked in unison with the original Ciona peristaltic pump in a two­compartment pump configuration. The use of the label “compartment” is appropriate, because the two pumping organs cannot be considered chambers. They operated peristaltically, were not endowed with one-way valves, and could not be assigned specific roles as inflow/outflow units, as both could reverse peristalsis (Davidson et al., 2006). The important concept derived from these experiments is that two, or perhaps more, anatomical compartments can be generated rather easily if adoption of a pump fate is dissociated in time to produce two or more “bursts” of pump muscle differentiation. In summary, the experiments of Davidson and colleagues suggest a plausible sequence of

events for cardiac chamber evolution, beginning with the formation of at least two compartments. From this initial two-compartment stage, major changes such as polarization (to establish a preferential direction of flow), valve formation (to guarantee efficient unidirectional flow), as well as integration of all myocytes (to produce an efficient contraction) would have to ensue to endow this twocompartment pump with all the attributes of cardiac chambers. However, it is important to realize that at this point many scenarios can be envisioned in which the trigger for cardiac chamber evolution was the evolution of any of the critical attributes of cardiac chambers discussed above. Some of the proposed scenarios for cardiac chamber evolution will be discussed below.

XV.  Hypotheses for the evolution of cardiac chambers In the discussion that follows, we brought together some ideas proposed by many authors during periods that spanned many years under the label of organized hypotheses. This was done exclusively for didactic purposes. Therefore, the responsibility for the organization of these ideas, as such, is ours. However, whenever possible, we discuss the original results and interpretations in the context of what is currently known, to evaluate the pros and cons of each idea.

XV.A.  The Sequential Hypothesis The sequential hypothesis of cardiac chamber evolution suggests that vertebrate chambers arose in succession, on top of an ancestral chordate single-chambered heart (Fishman and Chien, 1997; McRae and Fishman, 2002; Pough et al., 2002) (Fig. 10A). This intuitive concept finds support in the evolution of single compartments in decapod crustaceans (Box 4) which evolved powerful pumps, presumably by “fusion” of at least three consecutive arthropod body segments, generating a thick pumping compartment that establishes unidirectional flow through the action of potent unidirectional valves (Wilkens, 1999). Thus, contrary to our initial views (Simoes-Costa et al., 2005), single compartment pumps with unidirectional valves are not only viable, but also provide efficient circulatory work for large and active arthropods such as lobsters and crabs (Wilkens, 1999). There are problems with the sequential hypothesis, however. There is no indication that chordates followed an evolutionary path similar to that of decapod crustaceans. Moreover, the identity of the putative ancestral cardiac chamber was never clearly proposed, and an analysis of the problematic deuterostome fossil record fails to provide convincing evidence for single-chambered pumping organs (e.g., Chen, 1999; Mallat and Chen, 2003; and others).

30

PART  |  1  Heart Evolution

Figure 10  Three hypotheses for the evolution of vertebrate cardiac chambers. (A) The sequential hypothesis: cardiac chambers arose successively on top of a single-chambered heart from the ancestral chordate. (B) The recruitment hypothesis: cardiac chambers originated when an additional progenitor domain was recruited to a pump fate, creating a two-compartment peristaltic pump that was later molded by morphogenetic and other mechanisms into inflow and outflow chambers. (C) The patterning hypothesis: two or more cardiac chambers appeared simultaneously when patterning mechanism(s) divided the progenitor area of a peristaltic pump into two or more domains that were subsequently fashioned into inflow and outflow cardiac chambers by morphogenetic and other mechanisms.

Thus, the sequential hypothesis leaves us at a loss regarding the nature of the putative original cardiac chamber. One possible solution for this problem is to suggest that the Amphioxus SV is the ancestral chamber and the ventricle the vertebrate-specific chamber. However, there are reasons why this hypothesis is unlikely: it does not

explain the origins of atrium and conus arteriosus, which are already present in basal extant vertebrates; the cardiac affinity of the Amphioxus SV is questionable; and the thin smooth muscle walls and sparse myocytes of the SV make it a poor first chamber candidate from a physiological standpoint.

Chapter  |  1.1  Evolutionary Origins of Hearts

What has been specifically suggested is that the ventricle was the second cardiac chamber, the one designed to provide the high systemic blood pressures that are deemed to be vertebrate attributes (Fishman and Chien, 1997). Here again, there are some problems. The mechanical output of the hagfish ventricle is 13 to 60 times lower than in other fish; moreover, the pressures generated by hagfish and lamprey hearts range from 6.6 mmHg to 11 mmHg (Satchell, 1991d). This is similar to the pressures generated by the single contractile compartment of invertebrates such as lobsters (9–22 mmHg) and by the ventricle of Helix pomatia, a gastropod mollusc (17 mmHg) (Burger and Smythe, 1953; Jones, 1971a). The low pressures generated by hagfish and lamprey ventricles are markedly inferior to the high pressures observed in invertebrates, such as the 50–66 mmHg generated in the ventral vessels of the Glossoscolex giganteus, an annelid (Johansen and Martin, 1965). This questions the idea that ventricles were vertebrate novelties designed to develop the high systemic pressures that are not observed in invertebrates. Moreover, developmental data also fails to support the idea of the ventricle as the second vertebrate chamber. Results in amniotes suggest that the cardiac default may be a ventricular-like cell, rather than a sinus-like cell, since signaling is required in the cardiac field to generate atrial phenotypes in precursors that otherwise would differentiate into ventricular cells (Hochgreb et al., 2003). Finally, the precedence of outflow over inflow in ontogenesis is not consistent with the SV being the ancestral chamber, since outflow segments form before inflow ones in ontogenesis (De la Cruz and Markwald, 1998). In summary, the sequential hypothesis conflicts with multiple lines of evidence.

XV.B.  The Recruitment Hypothesis A new hypothesis has recently emerged from studies of tunicate heart development. According to this view, a second peristaltic compartment or functional chamber arose through recruitment of additional pump precursor cells just posterior to an ancestral anterior precursor field (Fig. 10B). Experimental support for this theory involves manipulation of FGF-mediated pump precursor cell recruitment in embryos of the tunicate, C. intestinalis (Davidson et al., 2006). The Ciona pump lineage can be traced back to a small group of four founder cells. In wild-type embryos, these founder cells divide asymmetrically and FGF signaling drives pump specification only in the smaller rostral daughter cells. Targeted manipulations of Ets1/2, a transcriptional effector of FGF, can drive ectopic pump specification in the caudal daughters, effectively doubling the number of pump precursor cells. These newly-recruited precursors are capable of forming additional pump tissue and, in some cases, underlie the formation of a new independent beating compartment, leading to a juvenile

31

with a two-compartment pump. These results indicate that a quantitative change in pump precursor cell numbers can drive a qualitative change in pump complexity. Similar allometric mechanisms of character evolution have also been invoked to explain the emergence of other complex structures, including feathers and teeth (Harris et al., 2005; Kassai et al., 2005). These results lead to the following scenario for chordate pump evolution: (1) pre-vertebrate chordate embryos contained a broad pool of potential pump precursors; (2) initially a subset of these cells was recruited through FGF signaling to form a single compartment pump similar to that of the tunicates; (3) in the vertebrate ancestor, shifts in FGF signaling led to the recruitment of posterior pump precursors and the spontaneous formation of a second functional compartment. This may either have occurred after the prior evolution of a single chamber, or chamber morphogenesis may have evolved in the context of a two-compartment heart. The latter possibility may seem unlikely, given that a dual compartment heart without an intervening valve would not provide an adaptive advantage. However, one can envision that a partially-obstructed connection between two peristaltic compartments may be sufficient to generate some independence of inflow and outflow, and thus boost the circulatory efficiency. Further study of blood flow in the experimentally-derived tunicate dual compartment heart, as well as computer modeling of such structures, should resolve this question. The speculative recruitment hypothesis gains some nominal support through the following observations. 1. The expression domain of the early cardiac specification factor Mesp indicates that vertebrate embryos contain a broad pool of pre-cardiac mesoderm (Saga et al., 2000; Sawada et al., 2000). In Ciona, Mesp expression defines the pump founder cell lineage and loss of Mesp function abrogates pump specification. In vertebrate embryos, there are two Mesp orthologs. Mesp1/a shows early, broad expression in the nonchordal mesoderm (during pre-gastrula stages) and functional studies in mice indicate that Mesp1 is specifically required for pump specification. Thus, Mesp may define a broad pool of potential cardiac mesoderm, facilitating recruitment of additional precursor cells during vertebrate heart evolution. 2. Recruitment may have occurred during the acquisition of additional heart chambers during vertebrate evolution. As discussed in a following section, a second heart field appears to have emerged within the amniotes, or even earlier in amphibians, contributing primarily to the emergent outflow chambers. The specification network of this secondary field is distinct from the primary field, suggesting that it was not derived through expansion of the primary field, but instead recruited from the neighboring pharyngeal mesoderm (see below).

32

3. Another line of support involves the expression of raldh2 in Ciona embryos. In vertebrates, raldh2 is expressed in the posterior, atrial-precursors of the linear heart tube (Hochgreb et al., 2003). RALDH2 serves to generate the retinoic acid (RA) gradient critical for establishing atrial versus ventricular identity (see below). In late stage Ciona embryos raldh2 is specifically expressed in the posterior sister lineage to the heart cells (SimoesCosta et al., 2005). Thus, when these cells are recruited to form a posterior chamber, the resulting organ field already contains the potential to undergo RA-mediated patterning. This raises the possibility that a recruited population of heart precursor cells may carry a genetic signature facilitating the subsequent axial patterning of the emergent organ (Fig. 10B).

XV.C.  The Inflow/Outflow Patterning Hypothesis We have recently proposed an alternative to the sequential hypothesis. In this new view, two or more cardiac chambers may have originated simultaneously in evolution when the progenitor field of an ancestral peristaltic pump was patterned by embryonic signals into two or more transcriptional domains (Fig. 10C). In principle, multiple ­signaling molecules may be involved in cardiac patterning. Retinoic acid (RA) is one such candidate molecule for cardiac patterning into inflow and outflow units. RA signaling effectively divides the embryonic cardiac field of amniotes such as mice and chicks into A–P domains that will give rise to inflow (atria and sinus venosa) and outflow segments (ventricles plus outflow tract) (Hochgreb et al., 2003). Thus, in mice and chicks, the basic cardiac inflow/ outflow organization is laid down early in development as an A–P code. The evidence in support of a major role of RA in amniote cardiac A–P patterning has been reviewed before (Rosenthal and Xavier-Neto, 2000; Xavier-Neto et al., 2001). Briefly, the results support a two-step model in which posterior cardiac precursors are specified to a sinoatrial fate by low concentrations of RA reaching the posterior cardiac field through diffusion from lateral and paraxial mesoderm. Subsequently, posterior cardiac precursors commit irreversibly to a sinoatrial fate in response to increased concentrations of RA produced by a caudo­ rostral wave of RALDH2 (ALDH1A2), a key enzyme in RA synthesis. This wave develops in the lateral mesoderm and endows posterior precursors with the capacity to produce their own RA at high concentrations, which may seal their sinoatrial fates (Hochgreb et al., 2003). RA signaling is currently the best-characterized mechanism for cardiac A–P patterning because it is expressed at the appropriate developmental times when the A–P identities are established and because gain- and loss-of-function experiments produce the exaggerated posterior and anterior phenotypes

PART  |  1  Heart Evolution

predicted by the association between RA and sinoatrial fates. The idea that cardiac chambers originated from a progenitor field of an ancestral chordate peristaltic pump that was patterned by RA signaling opens a series of interesting opportunities to understand how chordate pumps evolved. In fact, the evolutionary span of this idea has expanded. Recent data indicates that RA signaling is not a chordate innovation as was previously thought (Fujiwara and Kawamura, 2003); RA signaling is present in echinoderms and hemichordates which suggests that it may have already been present in the ancestral bilaterian, or Urbilateria, but lost in protostome lineages (Canestro et al., 2006; Marletaz et al., 2006; Simoes-Costa et al., 2008). Importantly, the patterning model suggests that the abrupt transition from the peristaltic vessels of tunicates and cephalochordates to the three- or four-chambered hearts of vertebrates may not necessarily be a misrepresentation resulting from the absence of intermediate forms in extant animals and/or incomplete fossil record, but rather the outcome of the developmental mechanisms that created the diversity needed for the simultaneous evolution of two or more cardiac chambers from an ancestral progenitor field. Also, if chambered hearts originated from the partition of a field of peristaltic pump progenitors, then it may be useful to regard the present ventricular cell type as a ground state similar to the myocyte that equipped the ancestral peristaltic pump. This could explain why it was never possible to identify “the” ventricular determinants, since these agents may actually be the same factors associated with expression of the basic cardiac phenotype (Simoes-Costa et al., 2005). In summary, the patterning hypothesis is parsimonious because it suggests that cardiac chambers might have originated from further differentiation and morphogenetic reorganization of ventricular-like cells, rather than from the sequential creation of entirely different cardiac cell types (Fig. 10C).

XVI.  Challenges to the inflow/ outflow patterning hypothesis Experiments in mice and chicks have revealed a previously unsuspected degree of complexity in the organization of cardiac chamber progenitors (e.g., the concept of a second cardiac field) (Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001). Moreover, data from the zebrafish and Xenopus respectively suggest that the typical topological relationship between RA signaling and cardiac chamber precursors of amniotes may not apply to all vertebrates, and that the evidence linking RA signaling to cardiac A–P patterning in amphibians is not strong. Thus, it is important to assess to what extent the patterning hypothesis is consistent with experiments in nonamniote vertebrates, chordates or deuterostomes.

33

Chapter  |  1.1  Evolutionary Origins of Hearts

XVI.A.  The Second Heart Field XVI.A.i.  The Early Days In 2001 three studies indicated that three different progenitor populations in the pharyngeal mesoderm contained outflow tract and right ventricle precursors (Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001) (see Chapter 2.2, Volume I). These were named as anterior heart field (Kelly et al., 2001; Mjaatvedt et al., 2001) or secondary heart field (Waldo et al., 2001), and were shown to contribute to the outflow tract and right ventricle in mice, and to the outflow tract in chicks (Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001; reviewed in Abu-Issa et al., 2004). Furthermore, data from retrospective clonal analysis (Meilhac et al., 2004) indicated that the embryonic mouse heart is formed by two major lineages that do not respect the boundaries between cardiac chambers (atria and ventricles) or cardiac segments (sinus venosus, AV canal or outflow tract). Altogether, the data supported the revolutionary notion that amniote cardiac development requires a second, subpharyngeal progenitor population, in addition to the paired, lateral mesoderm progenitor domains classically known as the cardiac crescent (the first heart field or FHF). In summary, in this new model, the first heart field was the origin of the left ventricle, AV canal, atria and sinus venosa, while the anterior/ secondary field accounted for the right ventricle and outflow tract in mice and outflow tract in chicks (Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001). The advent of the pharyngeal cardiac field and the retrospective analyses of Meilhac et al. (2004) dealt a mortal blow to a model of myocardial cell regionalization dubbed by Buckingham and colleagues as the segmental model. In this segmental model, each building block of cardiac progenitor tissue is a clonally-restricted population distributed along the A–P axis of the primitive streak and lateral mesoderm. As such, the blocks are independent units that do not mix with others, sport a unique gene regulatory network, and give rise to a particular cardiac segment according to their rostro-caudal position (Buckingham et al., 2005). The segmental model envisioned by Buckingham et al. (2005) provided a useful contrast for the new view of cardiac development heralded by the anterior/secondary cardiac field (subsequently unified as the second heart field or SHF). However, this particular segmental model represented only an extreme view of the significance of cardiac A–P organization to heart development. In reality, the cardiac fate maps established by Rosenquist, DeHaan and colleagues (DeHaan, 1963; Rosenquist and DeHaan, 1966; Stalsberg and DeHaan, 1969) and by others (GarciaMartinez and Schoenwolff, 1993; Redkar et al., 2001; Hochgreb et al., 2003) were dominantly interpreted as statistical statements about the localization of different cardiac precursors, rather than evidence for clonal restriction in the cardiac progenitor population (Garcia-Martinez and Schoenwolff, 1993; Redkar et al., 2001; Hochgreb

et al., 2003). Indeed, we have proposed that the RA signal that polarizes the heart in the A–P axis is sufficient only for the very broad distinction between inflow (sinoatrial) and outflow (ventricular and conotruncal) domains, rather than for the detailed distinction between each specific cardiac chamber or segment (Rosenthal and Xavier-Neto et al., 2000; Xavier-Neto et al., 2001; Hochgreb et al., 2003; Simoes-Costa et al., 2005). Unfortunately, the shortcomings of the segmental model, as framed by Buckingham et al. (2005) for didactic purposes, have been interpreted as a dismissal of the importance of cardiac A–P organization to heart development in subsequent interpretations. This is a surprising conclusion, because it neglects the overwhelming evidence for cardiac A–P patterning, or, for that matter, for all cardiac axial patterning, as if cardiac development had traded off all its exquisitely intricate layers of regulation for a simpler model of mosaic development, in which fate is directly determined by clonal origins. However, any mosaic model of cardiac development will be severely taxed to explain the outcomes of classic experiments of transplantation and rotation of the cardiac field that demonstrate the extraordinary plasticity inherent to the cardiac field before it is patterned in the A–P axis (Orts-Llorca and Collado, 1967; Pathwardhan et al., 2000; reviewed in Xavier-Neto et al., 2001). In summary, the current emphasis on the contribution of the second heart field has promoted unnecessary controversy and confusion that can be easily resolved if all the different views on cardiac development are considered in a balanced analysis. Indeed, we have been arguing (Simoes-Costa et al., 2005; see below) that there is nothing incompatible with cardiac A–P patterning, and the modern views of cardiac development advanced after recognition of crucial roles for the mechanisms that generate chamber myocardium, the ballooning hypothesis (Christoffels et al., 2000), and after acknowledgement of the heterogeneous nature of the cardiac progenitor population, the second heart field hypothesis (e.g., Buckingham et al., 2005).

XVI.B.  Conflicting Evidence The substantial body of evidence gathered since 2001 has been demanding constant re-evaluations of the concept of a second independent population of cardiac progenitors. In 2003, Cai and colleagues showed that the anterior/secondary heart field is actually part of a continuous, cardiac progenitor domain that expresses the lim-1 homeodomain transcription factor Isl-1 and spans the whole A–P extent of the splanchnic mesoderm dorsal to the looped heart (Cai et al., 2003; Buckingham et al., 2005), as already implied in Kelly et al. (2001). Double in situ hybridization with Isl-1 (a marker of the anterior/secondary field) and MLC2-A (a marker of myocytes from the first heart field) showed that

34

XVI.C.  The Second Heart Field Goes Evolutionary Important clues to the resolution of the controversy between the opposing views of two independent heart fields versus one complex, heterogeneous, heart field were already present in the original study of Cai et al. (2003). Surprisingly, genetic fate-mapping studies using an Isl-1-Cre driver and floxed LacZ reporter mice indicated that a considerable number of Isl-1-expressing cells are found in the left ventricle (Cai et al., 2003; Sun et al., 2007), despite the fact that the left ventricle is considered first heart field territory (Meilhac et al., 2004). These studies raised some doubts about the specificity of Isl-1 as truthful second heart field markers. Nonetheless, the findings were not wholly inconsistent with the idea of a separate second heart field, since Cre–Lox fate mapping is prone to some degree of leakiness. A more fundamental objection to the concept that the second heart field is a separate entity from the first heart field was raised by the demonstration by Brade et al. (2007) that Isl-1 is expressed throughout the cardiac crescent of Xenopus laevis and extensively overlaps the Nkx25 domain, which is considered a marker of the first heart field. The results of Brade and colleagues indicate that the early amphibian cardiac progenitor domain is structured as a single field that contains a heterogeneous, but overlapping, population of Isl-1 and Nkx2-5 precursors. These two populations gradually break away to give rise to two separate

St 17/18

np

nf

nf

cg

St 25 St 30 St 33

at early stages in mice (e.g., 7.0–8.0 dpc) the anterior/secondary heart field and the first heart field are in fact continuous, with the anterior/secondary heart field occupying a position medial and dorsal to the first heart field (Cai et al., 2003). The study of Cai et al. (2003) made it clear that, at later stages, the more anterior section of the Isl-1-expressing population moves from its initial location dorsal and medial to the first heart field and migrates anteriorly to occupy the pharyngeal mesoderm. As revealed by Isl-1 expression, the anterior/secondary heart field keeps its spatial integrity throughout cardiac development, which explains the surprising contribution of this population to the atria and to a limited subset of the left ventricle (Cai et al., 2003). The work by Cai et al. (2003) prompted a revision of the initial concept of a separate cardiac progenitor population at the pharyngeal mesoderm, and indicated that the Isl-1 population could be better understood as a second cardiac field (SHF) closely attached to the first heart field (Buckingham et al., 2005). The notion of a second heart field was an appropriate conclusion at the time, since it highlighted the seemingly diverse transcription factor network employed by these cells (Kelly, 2005; Black, 2007). Nonetheless, an equally reasonable conclusion was that the second heart field was an integral part of a single, but complicatedly-structured cardiac field, as proposed by Abu-Issa et al. (2004).

PART  |  1  Heart Evolution

cg

cg

cg

Nkx-2.5

Nkx-2.5 + Isl-1

Isl-1

Cement Gland

Figure 11  One heterogeneous cardiac field in the Amphibian Xenopus laevis. Double in situ hybridization with Isl-1 and Nkx2-5 antisense riboprobes indicates that the early Amphibian cardiac field is a complex domain that includes overlapping progenitor populations. With development, Isl-1 and Nkx2.5-expressing progenitor populations break away to gradually establish separated domains. Reproduced with permission and adapted from Brade et al. (2007).

progenitor domains at looping stages (Fig. 11). In summary, these important results suggest that the ancestral vertebrate cardiac progenitor domain was a single field that contained a mix of Isl-1 and Nkx2-5-expressing cells. The differences between Xenopus laevis and mice are probably the result of heterochrony, in that mice display Isl-1 and Nkx2-5 populations that are already segregated at cardiac crescent stages, while in Xenopus this is only observed at looping stages. More sampling from reptile, amphibian and fish species will be necessary to confirm this hypothesis. The recent study of Mann et al. (2009) is consistent with the views described above, and lends support to the idea that the ancestral bilaterian pump progenitor field co-expressed Isl and NK2 type transcription factors. The authors showed that tailup (tup), the fly Isl homolog, is co-expressed with tin in dorsal vessel precursors, and that it is required for dorsal vessel development. Therefore, Isl homologs were probably part of the core transcription factor network required to build an ancestral pumping organ from a single ancestral progenitor field (Mann et al., 2009).

Chapter  |  1.1  Evolutionary Origins of Hearts

XVI.D.  One Cardiac Field After All? The results reported by Prall and colleagues in 2007 constitute another fundamental objection to the concept of two different, independent, cardiac fields. Briefly, Prall et al. (2007) demonstrated that Isl-1, previously thought to be a specific marker of the second heart field is, in fact, expressed throughout the coelomic mesoderm that gives rise to the somatic and splanchnic mesoderm, the origin of both the first heart field and second heart field. These results are fully consistent with the previous findings in the early (HH5) chick embryo (Yuan and Schoenwolf, 2000), and provide an explanation for the extensive labeling of the left ventricle in genetic fate maps (Cai et al., 2003; Sun et al., 2007). More importantly, Prall and colleagues uncovered a feedback loop between Nkx2-5 (thought to be a marker of the first heart field) and the Bmp2-Smad1 signaling pathway. Nkx2-5 represses Bmp2-Smad1 signaling, which normally induces cardiac differentiation and inhibits proliferation in the Isl-1 population. Targeted recombination of Nkx2-5 released an impressive surge of Bmp/Smad signaling throughout the cardiac progenitor area which not only reached the Isl-1-expressing population, but also crossed embryonic layers to induce Nkx2-5 in the ectoderm. This particular result questions the whole wisdom of the current paradigm that envisions Nkx2-5 and Isl-1-expressing cardiac progenitor populations as independent first heart field and second heart field, respectively. In summary, the pancardiac nature of early Isl-1 ­expression in the coelomic mesoderm (Yuan and Schoenwolff, 2000), the continuous distribution of Isl-1 and Nkx2-5 in the mouse cardiac field, the contribution of Isl-1-expressing cells to the left ventricle, the extensive overlap between Nkx2-5 and Isl-1 domains in the amphibian heart field, the cross-regulatory interactions between Isl1 and Nkx2-5 populations and the diffusible nature of the mediators of such interactions argue against the existence of clonally restricted Nkx2-5 and Isl-1-expressing populations in the first heart field and second heart field, respectively.

XVI.E.  Reconciling Cardiac Anterior– Posterior Patterning with the Heterogeneous Nature of the Cardiac Field The current research emphasis on the second heart field has established two recent trends: a dismissal of the influence of anterior–posterior (A–P) patterning in cardiac development in favor of a mosaic-like model based on the first heart field and second heart field; and a tendency to overestimate the contributions of the second heart field to the detriment of those of the first heart field. The first trend is a remarkable one, in view of the fact that hearts with clear A–P (inflow/outflow) organization develop in Isl-1 knockout embryos (Cai et al., 2003). The second trend, represented in the suggestion that the primary site of action of RA may

35

not be in the first heart field, but in the second heart field (Ryckebusch et al., 2008), is also surprising in light of the fact that the majority of raldh2 knockout embryos and of embryos treated with RA antagonists or Raldh inhibitors do not reach looping stages and die without forming an atrial chamber (Chazaud et al., 1999; Niederreither et al., 1999, 2001; Xavier-Neto et al., 1999). The results from studies that focus on the second heart field contribution (Cai et al., 2003; Prall et al., 2007; Galli et al., 2008; Ryckebusch et al., 2008) can be easily reconciled with earlier studies on the first heart field and with cardiac A–P patterning by RA to produce an equilibrated synthesis (reviewed in Xavier-Neto, 2001; Hochgreb et al., 2003; see also Sirbu et al., 2008). While the subject is beyond the scope of this chapter, it is important to note that this synthesis can provide explanations for the origins of inflow and outflow precursors from the Isl-1-expressing progenitor population and for the role of the first heart field in a framework of A–P patterning by RA signaling (Fig. 12) (Sirbu et al., 2008; Xavier-Neto, in preparation). In this view, Isl-1-expressing (SHF) inflow precursors represent the most posterior cells in the second cardiac field, the ones that, shortly after receiving the RA signal, stay in the lateral mesoderm and begin to contribute to sinus venosa and atria, while the RA-free anterior progenitors may send their progeny to the outflow cardiac segments only after their migration to the sub-pharyngeal position (Fig. 12). This is consistent with the fact that sinoatrial clones segregate well before outflow tract clones in the retrospective analysis of Meilhac et al. (2004). Alternatively, the contribution of the primary heart field may be understood as a scaffold over which progenitors from the second heart field migrate and find specific cues for their integration, survival, proliferation and differentiation. This idea is supported by a rare, but significant, third population of clones that contribute to all cardiac A–P segments in retrospective analyses (Meilhac et al., 2004) and by the phenotype of Isl-1 knockout embryos (Cai et al., 2003). The behavior of these large, unrestricted, clones is consistent with a scheme where myocardial “pioneers” from the first heart field would undergo axial patterning and then establish a primitive structure represented by primitive inflow (atrium) and outflow (ventricle) chambers (e.g., Cai et al., 2003) that will later receive the delayed second heart field contribution that generates most of the right ventricle and outflow tract and substantial regions of the atria (Fig. 12).

XVI.F.  Retinoic Acid Signaling and Cardiac Anterior–Posterior Patterning in Amphibians and Fish Although cardiac inflow/outflow patterning by RA could have been independently developed in birds and mammals, it is far more likely that it is an ancestral feature of

36

PART  |  1  Heart Evolution

Figure 12  Reconciling cardiac A–P patterning with the heterogeneous nature of the cardiac field. In this model, posterior first heart field progenitors, along with posterior second heart field precursors, receive a posteriorizing signal from retinoic acid, stay in situ, and commit to inflow (sinoatrial) fates. The posterior section of the first heart field quickly forms the outlines of an inflow compartment. In contrast, the caudal second heart field population proliferates, delays cardiac differentiation, and only much later, is incorporated into the nascent, inflow, first heart field-derived compartment. The anterior sections of both first heart field and second heart field escape the initial retinoic acid patterning phase, and thus are free to commit to outflow (ventricular and outflow tract) fates. The anterior section of the first heart field stays in situ and gives rise to the left ventricle, while the anterior section of the second heart field migrates anteriorly to the sub-pharyngeal site, from where it will form outflow tract and right ventricle. In summary, the late differentiating, proliferative, caudal and rostral second heart field sections are incorporated into an inflow/outflow scaffold derived from the early differentiating first heart field. This model is fully-consistent with the phenotypes of Isl-1-null-embryos, which display a truncated heart that is, nonetheless, obviously patterned in the A–P axis to form inflow and outflow chambers.

amniotes, or even of amphibians and fish. However, the role of RA in this process is still unclear in amphibians, perhaps because suitable, early markers of cardiac A–P phenotypes are lacking in these species. Nonetheless, RA is clearly important for cardiac development and differentiation, which is inhibited by the retinoid (Drysdale et al., 1997; Collop et al., 2006). In the zebrafish embryo at the 15-somite stage, atrial and ventricular precursors are distributed in a lateral-to-medial fashion in the anterior lateral mesoderm (Yelon et al., 1999), rather than in the typical A–P pattern of amniotes (Hochgreb et al., 2003). However, fate-map studies by Keegan et al. (2004) indicate that from the tail bud to the 5-somite stage, cardiac precursors are oriented neither in medial to lateral, nor in an anterior to posterior fashion, but in an oblique manner in which ventricular precursors occupy a domain that is anterior to that occupied by atrial precursors, although these domains overlap in the A–P axis. Moreover, these stages coincide with the expression of the zebrafish RALDH2 caudorostral wave, which develops in the lateral mesoderm similarly as in amniotes showing, nonetheless, an interesting feature that may explain how RA can still pattern the zebrafish cardiac progenitors into inflow/outflow domains, even if the latter assume a lateral to medial organization at about the 15-somite stage (Keegan et al., 2004). This is because in the zebrafish, the RALDH2 caudorostral wave obligingly develops in a bracket-like fashion, first expanding laterally and then moving anteriorly, which may endow it with the ability to selectively pattern lateral precursors into inflow, sinoatrial myocytes (Begemann et al., 2001; Simoes-Costa et al., unpublished observations) (Fig. 13). More importantly, reciprocal manipulation of the zebrafish RA signaling pathway during the embryonic period when the RALDH2 caudorostral wave develops produces essentially the same phenotypes observed in amniotes; extension

of posterior, inflow, domains by exogenous RA and reduction of inflow domains after RA synthesis inhibition (Fig. 13). In truth, the spatial segregation of ventricular and atrial progenitors (Hochgreb et al., 2003; Keegan et al., 2004), the development of the caudorostral wave of RALDH2, and the response to exogenous RA treatment are all remarkably similar in amniotes and zebrafish at the critical periods in which atrial and ventricular identities are established in these species (Stainier and Fishman, 1992; Hochgreb et al., 2003).6 The inflow/outflow hypothesis for cardiac chamber evolution has been recently challenged by Waxman et al. (2008). In their view, RA signaling in the zebrafish acts primarily to restrict the number of atrial and ventricular cardiac precursors, rather than to provide instructions for cardiac progenitors to commit to an inflow (sinoatrial) fate, as suggested by the inflow/outflow patterning hypothesis. It is perfectly possible that differential effects of RA on atrial or ventricular specification/proliferation rates may be responsible for the cardiac phenotypes that are observed after manipulation of RA signaling in the zebrafish. Indeed Waxman et al. (2008) provided evidence that RA inhibition increases the number of atrial cells more than it increases the number of ventricular cells. Therefore, if RA is a more potent inhibitor of specification (or proliferation, or both) of atrial cells than of ventricular cells, the outcome of an experimental manipulation that increases the strength of RA signaling will be a predominance of ­ ventricular over 6. The role of RA in vertebrate inflow/outflow patterning should not be confused with its ability to block commitment and/or differentiation to a cardiac fate (Osmond et al., 1991; Stainier and Fishman, 1992; Drysdale et al., 1997; Keegan et al., 2005; Collop et al., 2006). This may be an ancestral chordate mechanism for delimiting morphogenetic fields of pumping organs, from which the ability to pattern the pump progenitor field in the A–P axis evolved secondarily.

Chapter  |  1.1  Evolutionary Origins of Hearts

37

Figure 13  Inflow/outflow patterning by retinoic acid signaling in the zebrafish. Top row: In situ hybridization with an antisense riboprobe for RALDH2, the major retinoic acid synthetic enzyme. A RALDH2 caudorostral wave, homologous to the amniote caudorostral wave (Hochgreb et al., 2003), develops in the zebrafish lateral mesoderm, posterior to the cardiac field. Middle row: Schematic interpretation of RALDH2 expression patterns in relation to the migratory movements displayed by zebrafish atrial and ventricular precursors (based on Simoes-Costa et al., unpublished, and Keegan et al., 2004, respectively). Bottom row: At the tailbud stage, treatment with the RALDH inhibitor DEAB, or with exogenous retinoic acid, induces hearts with outflow (ventricular), or inflow (atrial) dominance, respectively. Therefore, reciprocal manipulation of retinoic acid signaling at embryonic stages that precede and overlap with the RALDH2 caudorostral wave produces essentially the same experimental outcomes observed in amniotes (Xavier-Neto et al., 1999; Hochgreb et al., 2003).

atrial cells. Conversely, an inhibition of RA signaling would result in a proportionally higher number of atrial cells in relation to ventricular cells. The problem with the views put forward by Waxman et al. (2008) is that the actual outcomes of the experiments in which RA signaling is augmented or inhibited are just the opposite to those predicted by the specification and/or proliferation effects that they describe. Stainier and Fishman showed, in 1992, that treatment with increasing doses of RA produced a progressive increase in the relative amount of atrial over nonatrial tissues. Furthermore, here we show in Fig. 13 that treatment with very short pulses of appropriate doses of the Raldh inhibitor DEAB produce hearts with ventricular, rather than atrial, dominance. There are many reasons why we believe the results of Waxman et al. (2008) do not constitute a challenge to the inflow/outflow hypothesis. We believe RA signaling is

dynamic and fast, that its effects are not single, but multiple, and that they are exerted within short and specific windows of opportunity, which open and close in sucession throughout development. Therefore, in our view, the multiple effects of RA signaling on cardiac development cannot be reduced to one or two major effects, and cannot be adequately addressed by the use of protracted treatments that will influence several different developmental processes. Rather, we believe that the effects of RA signaling should be studied with short pulses of effective doses that coincide with periods in which Raldhs are expressed close to or within the tissue under observation. In summary, there are no reasons to believe that vertebrate hearts develop without extensive inflow/outflow patterning. On the contrary, the available evidence suggests that inflow and outflow patterning of the heart field, whether achieved in connection with signaling ­mechanisms

38

spatially aligned to A–P or other embryonic axes, may set the stage for morphogenetic mechanisms to form, in a species-specific fashion, chambers from the four original vertebrate cardiac compartments, SV, atrium, ventricle and conus arteriosus.

PART  |  1  Heart Evolution

by analyses of extant invertebrate chordates or by developmental studies employing gain-of-function strategies. There is, of course, a limit to the extent that all evolutionary studies may contribute to unravel the history of animals. Even with this caveat, we argue that the study of the evolution of animal pumping organs will produce useful concepts for cardiac development.

XVII.  Conclusions Why study the origins of our hearts among such a diverse and highly heterogeneous group as metazoans? In our view, studying the evolution of hearts is just another way to illuminate cardiac development with general principles deduced from the study of species that anteceded mammals and men, and that include in their morphologies adaptations which we may have inherited or emulated. Thus, understanding these adaptations and the constraints that may have selected them can be a powerful aid when we approach development in model systems. In addition, by analyzing the developmental mechanisms characterized in model species in the broader evolutionary context, we may be able to define whether a particular mechanism has also a more general, ancestral, significance, or if it is an evolutionary derivation. If anything, thinking on an evolutionary scale helps us appreciate that the staggering diversity of animal forms and developmental mechanisms cannot be encapsulated into any given species. The comparative study of pumping organs in bilaterian animals is rife with provoking and stimulating similarities that in some cases reflect true homologies. In other cases, however, the distinction is not yet possible, but it is helpful to note that the requisites for homology are very well-known today, and that it is necessary to check the available evidence against these criteria (Abouheif et al., 1997; Wray and Abouheif, 1998; Baguna and Garcia-Fernandez, 2003; Hall, 2003). However, similarities between animal pumping organs need not be homologies to be interesting. Indeed, we have reviewed evidence that, in many cases, these similarities can be conveniently attributed to evolutionary parallelisms or to convergence (Hall, 2003; Xavier-Neto et al., 2007). These parallel or convergent solutions to the problems of animal pumping are, in a sense, actually more challenging to our comprehension than the cases of homology. This is because we are still a long way from grasping the interplay between hemodynamic forces and developmental mechanisms, whose study will be an important topic for further research. In summary, the evolution of chambered hearts from the simple peristaltic organs of invertebrate chordates remains a fascinating problem whose evolutionary, developmental and genetic bases have been left largely untouched. However, data coming from several different areas are beginning to provide material for some hypotheses that have the potential to rise from mere speculative scenarios to offer predictions that can be objectively tested

Acknowledgments We are indebted to Nadia Rosenthal and Richard Harvey for support and encouragement. We would also like to thank Billie Swalla, David Bottjer, Deborah Schechtman, José Maria Perez-Pomares, Junyuan Chen, Linda Holland, Nicholas Holland, Peter Currie, Petra Pandur, Ramon MuñozChapuli and Richard Harvey for advice, criticisms, figures, insights and reviews. We would also like to apologize to the authors of many important contributions that were not cited here due to space constraints. This work was supported by grants from FAPESP (06/50843-0), CNPq (305260/2007-3) and CAPES.

Glossary Amniotes vertebrates defined by the embryonic development of several extensive membranes, the amnion, chorion and allantois. Clade: a taxonomic group containing a common ancestor and all its descendants. Coelom: a body cavity enclosed by mesoderm. Diploblasts: animals with only two germ layers, ectoderm and endoderm. Diplura: a class of nonwinged hexapods that hold the base of their mouthparts inside the head capsule. Endopterygote: insects with distinctive larval, pupal and adult stages. From the Greek endo (within) and pterygos (wing), it refers to the fact that in later immature stages the wing buds are not evident externally, but instead the future wing tissues are entirely internalized. Heterochrony: changes in the timing of developmental processes that occur in the course of evolution. Metazoans: multicellular animals, including sponges. Natural group: a group that includes a common ancestor and all of its descendants. Triploblast: metazoans whose development goes through a three germ-layer stage, ectoderm, mesoderm and endoderm. Ortholog: a gene related to another through a speciation event, as opposed to a gene-duplication event.

References Abouheif, E., Akam, M., Dickinson, W.J., Holland, P.W., Meyer, A., Patel, N.H., Raff, R.A., Roth, V.L., Wray, G.A., 1997. Homology and developmental genes. Trends Genet. 13, 432–433. Abu-Issa, R., Waldo, K., Kirby, M.L., 2004. Heart fields, one, two or more? Dev. Biol. 272, 281–285.

Chapter  |  1.1  Evolutionary Origins of Hearts

Adoutte, A., Balavoine, G., Lartillot, N., de Rosa, R., 1999. Animal evolution. The end of the intermediate taxa? Trends Genet. 15, 104–108. Adoutte, A., Balavoine, G., Lartillot, N., Lespinet, O., Prud’homme, B., de Rosa, R., 2000. The new animal phylogeny, reliability and implications. Proc. Natl. Acad. Sci. USA 97, 4453–4456. Aguinaldo, A.M., Turbeville, J.M., Linford, L.S., Rivera, M.C., Garey, J.R., Raff, R.A., Lake, J.A., 1997. Evidence for a clade of nematodes, arthropods and other moulting animals. Nature 387, 489–493. Alcoléa, S., Théveniau-Ruissy, M., Jarry-Guichard, T., Marics, I., Tzouanacou, E., Chauvin, J.P., Briand, J.P., Moorman, A.F., Lamers, W.H., Gros, D.B., 1999. Downregulation of connexin 45 gene products during mouse heart development. Circ. Res. 84, 1365–1379. Alldredge, A., 1976. Appendicularians. Sci. Am. 235, 94–102. Anderson, M., 1968. Electrophysiological studies on initiation and reversal of the heart beat in Ciona intestinalis. J. Exp. Biol. 49, 363–385. Arceci, R.J., King, A.A., Simon, M.C., Orkin, S.H., Wilson, D.B., 1993. Mouse GATA-4, a retinoic acid-inducible GATA-binding transcription factor expressed in endodermally derived tissues and heart. Mol. Cell Biol. 13, 2235–2246. Augustinsson, K.B., Fange, R., Johnels, A., Ostlund, E., 1956. Histological, physiological and biochemical studies on the heart of two cyclostomes, hagfish (Myxine) and lamprey (Lampetra). J. Physiol. 131, 257–276. Auman, H.J., Coleman, H., Riley, H.E., Olale, F., Tsai, H.J., Yelon, D., 2007. Functional modulation of cardiac form through regionally confined cell shape changes. PLoS Biol. 5, e53. Babu, G.J., Warshaw, D.M., Periasamy, M., 2000. Smooth muscle myosin heavy chain isoforms and their role in muscle physiology. Microsc. Res. Tech. 50, 532–540. Baguna, J., Garcia-Fernandez, J., 2003. Evo-Devo, the long and winding road. Int. J. Dev. Biol. 47, 705–713. Ball, E.E., Hayward, D.C., Saint, R., Miller, D.J., 2004. A simple plan – cnidarians and the origins of developmental mechanisms. Nat. Rev. Genet. 5, 567–577. Balser, E.J., Ruppert, E.E., 1990. Structure, ultrastructure and function of the heart-kidney complex of Saccoglossus kowalevskii (Hemichordata, Enteropneusta). Including new data on the stomochord. Acta Zool. 71, 235–249. Bäumer, C., Pirow, R., Paul, R.J., 2002. Circulatory oxygen transport in the water flea Daphnia magna. J. Comp. Physiol. B. 172, 275–285. Baxter, E.W., 1957. The development of the intestinal blood-system of lampreys (Petromyzonidae), with an account of the origin of vascular couples. Proc. Zool. Soc. Lond. 129, 371–396. Begemann, G., Schilling, T.F., Rauch, G.J., Geisler, R., Ingham, P.W., 2001. The zebrafish neckless mutation reveals a requirement for raldh2 in mesodermal signals that pattern the hindbrain. Development 128, 3081–3094. Bengtson, S., Budd, G., 2004. Comment on “Small bilaterian fossils from 40 to 55 million years before the Cambrian.”. Science 306, 1291. Binyon, J., 1972. Physiology of Echinoderms. Pergamon Press, Oxford. Black, B.L., 2007. Transcriptional pathways in second heart field development. Semin. Cell Dev. Biol. 18, 67–76. Blair, J.E., Hedges, S.B., 2005. Molecular phylogeny and divergence times of deuterostome animals. Mol. Biol. Evol. 22, 2275–2284. Bodmer, R., 1995. Heart development in Drosophila and its relationship to vertebrates. Trends Cardiovasc. Med. 5, 21–28. Bodmer, R., Jan, L.Y., Jan, Y.N., 1990. A new homeobox-containing gene, msh-2, is transiently expressed early during mesoderm formation of Drosophila. Development 110, 661–669.

39

Boolootian, R.A., Campbell, J.L., 1964. A primitive heart in the echinoid Strongylocentrotus purpuratus. Science 145, 173–175. Bourlat, S.J., Juliusdottir, T., Lowe, C.J., Freeman, R., Aronowicz, J., Kirschner, M., Lander, E.S., Thorndyke, M., Nakano, H., Kohn, A.B., Heyland, A., Moroz, L.L., Copley, R.R., Telford, M.J., 2006. Deuterostome phylogeny reveals monophyletic chordates and the new phylum Xenoturbellida. Nature 444, 85–88. Bourne, G.H., 1980. Hearts and Heart-Like Organs. Academic Press, New York. Brade, T., Gessert, S., Kühl, M., Pandur, P., 2007. The amphibian second heart field, Xenopus islet-1 is required for cardiovascular development. Dev. Biol. 311, 297–310. Bruneau, B.G., Nemer, G., Schmitt, J.P., Charron, F., Robitaille, L., Caron, S., Conner, D.A., Gessler, M., Nemer, M., Seidman, C.E., Seidman, J.G., 2001. A murine model of Holt-Oram syndrome defines roles of the T-box transcription factor Tbx5 in cardiogenesis and disease. Cell 106, 709–721. Brusca, R.C., Brusca, G.J., 2003. Invertebrates, second ed. Sinauer Associates, Sunderland, MA. Buckingham, M., Meilhac, S., Zaffran, S., 2005. Building the mammalian heart from two sources of myocardial cells. Nat. Rev. Genet. 6, 826–835. Burger, J.W., Smythe, C.M, 1953. The general form of circulation in the lobster, Homarus. J. Cell Physiol. 42, 369–383. Burggren, W.W., Keller, B.B., 1997. Development of Cardiovascular Systems: Molecules to Organisms. Cambridge University Press, Cambridge. Burggren, W.W., Farrell, A.P., Lillywhite, H., 1997. Vertebrate cardiovascular physiology. In: Dantzler, W.H. (Ed.), The Handbook of Physiology. American Physiological Society, New York, pp. 215–308. Burnett, B.R., 1984. Striated muscle in the wall of the dorsal abdominal aorta of the California Spiny Lobster Panulirus interruptus. J. Crustacean. Biol. 4, 560–566. Burrage, T.G., Sherman, R.G., 1978. Cellular organization of the embryonic lobster heart. Cell Tissue Res. 188, 171–187. Burton, P.M., 2008. Insights from diploblasts; the evolution of mesoderm and muscle. J. Exp. Zoolog. B. Mol. Dev. Evol. 310 (1), 5–14. Cai, C.L., Liang, X., Shi, Y., Chu, P.H., Pfaff, S.L., Chen, J., Evans, S., 2003. Isl1 identifies a cardiac progenitor population that proliferates prior to differentiation and contributes a majority of cells to the heart. Dev. Cell 5, 877–889. Cameron, C.B., Garey, J.R., Swalla, B.J., 2000. Evolution of the chordate body plan, new insights from phylogenetic analyses of deuterostome phyla. Proc. Natl. Acad. Sci. USA 97, 4469–4474. Canestro, C., Postlethwait, J.H., Gonzalez-Duarte, R., Albalat, R., 2006. Is retinoic acid genetic machinery a chordate innovation? Evol. Dev. 8, 394–406. Cannon, W.B., 1911. The nature of gastric peristalsis. Am. J. Physiol. 29, 250–266. Carter, G.S., 1967. Structure and Habit in Vertebrate Evolution. Sidgwick & Jackson, London. Chapman, C.B., Jensen, D., Wildenthal, K., 1963. On circulatory control mechanisms in the Pacific hagfish. Circ. Res. 12, 427–440. Chen, J.Y., Huang, D.Y., Li, C., 1999. An early Cambrian craniate-like chordate. Nature 403, 518–522. Chen, J.Y., Bottjer, D.J., Oliveri, P., Dornbos, S.Q., Gao, F., Ruffins, S., Chi, H., Li, C.W., Davidson, E.H., 2004a. Small bilaterian fossils from 40 to 55 million years before the Cambrian. Science 305, 218–222.

40

Chen, J.Y., Oliveri, P., Davidson, E.H., Bottjer, D.J., 2004b. Response to comment on “Small bilaterian fossils from 40 to 55 million years before the Cambrian. Science 306, 1291b. Chiba, S., Awazu, S., Itoh, M., Chin-Bow, S.T., Satoh, N., Satou, Y., Hastings, K.E., 2003. A genomewide survey of developmentally relevant genes in Ciona intestinalis. IX. Genes for muscle structural proteins. Dev. Genes Evol. 213, 291–302. Christoffels, V.M., Habets, P.E., Franco, D., Campione, M., de Jong, F., Lamers, W.H., Bao, Z.Z., Palmer, S., Biben, C., Harvey, R.P., Moorman, A.F., 2000. Chamber formation and morphogenesis in the developing mammalian heart. Dev. Biol. 223, 266–278. Christoffels, V.M., Hoogaars, W.M., Tessari, A., Clout, D.E., Moorman, A.F., Campione, M., 2004. T-box transcription factor Tbx2 represses differentiation and formation of the cardiac chambers. Dev. Dyn. 229, 763–770. Clark, A.J., 1927. Comparative Physiology of the Heart. The Macmillan Company, New York, NY. Colas, J.F., Lawson, A., Schoenwolf, G.C., 2000. Evidence that translation of smooth muscle alpha-actin mRNA is delayed in the chick promyocardium until fusion of the bilateral heart-forming regions. Dev. Dyn. 218, 316–330. Cole, F.J., 1926. A monograph on the general morphology of the Myxinoid fishes, based on a study of Myxine. Part IV. The morphology of the vascular system. Trans. Roy. Soc. Edin. 54, 309–342. Collop, A.H., Broomfield, J.A., Chandraratna, R.A., Yong, Z., Deimling, S.J., Kolker, S.J., Weeks, D.L., Drysdale, T.A., 2006. Retinoic acid signaling is essential for formation of the heart tube in Xenopus. Dev. Biol. 291, 96–109. COMSOL Multiphysics Modeling Guide. http://www.chem.ucla. edu/~frax/exch/Courses/Advanced/peristaltic_pump/  . Conway-Morris, S., 2000. The Cambrian “explosion,” slow-fuse or megatonnage? Proc. Natl. Acad. Sci. USA 97, 4426–4429. Coppen, S.R., Severs, N.J., 2002. Diversity of connexin expression patterns in the atrioventricular node, vestigial consequence or functional specialization? J. Cardiovasc. Electrophysiol. 13, 625–626. Cripps, R.M., Olson, E.N., 2002. Control of cardiac development by an evolutionarily conserved transcriptional network. Dev. Biol. 246, 14–28. Cserjesi, P., Brown, D., Lyons, G.E., Olson, E.N., 1995. Expression of the novel basic helix-loop-helix gene eHAND in neural crest derivatives and extraembryonic membranes during mouse development. Dev. Biol. 170, 664–678. Davidson, B., 2007. Ciona intestinalis as a model for cardiac development. Semin. Cell Dev. Biol. 18, 16–26. Davidson, B., Shi, W., Beh, J., Christiaen, L., Levine, M., 2006. FGF signaling delineates the cardiac progenitor field in the simple chordate, Ciona intestinalis. Genes. Dev. 20, 2728–2738. DeHaan, R.L., 1963. Migration patterns of the precardiac mesoderm in the early chick embryo. Exp. Cell Res. 29, 544–560. de Jong, F., Opthof, T., Wilde, A.A., Janse, M.J., Charles, R., Lamers, W.H., Moorman, A.F., 1992. Persisting zones of slow impulse conduction in developing chicken hearts. Circ. Res. 71, 240–250. De la Cruz, M.V., Markwald, R.R., 1998. Living Morphogenesis of the Heart. Birkhäuser, Boston, MA. Delorme, B., Dahl, E., Jarry-Guichard, T., Briand, J.P., Willecke, K., Gros, D., Théveniau-Ruissy, M., 1997. Expression pattern of connexin gene products at the early developmental stages of the mouse cardiovascular system. Circ. Res. 81, 423–437. Delsuc, F., Brinkmann, H., Chourrout, D., Philippe, H., 2006. Tunicates and not cephalochordates are the closest living relatives of vertebrates. Nature 439, 965–968.

PART  |  1  Heart Evolution

Desjardins, P.R., Burkman, J.M., Shrager, J.B., Allmond, L.A., Stedman, H.H., 2002. Evolutionary implications of three novel members of the human sarcomeric myosin heavy chain gene family. Mol. Biol. Evol. 19, 375–393. Drysdale, T.A., Patterson, K.D., Saha, M., Krieg, P.A., 1997. Retinoic acid can block differentiation of the myocardium after heart specification. Dev. Biol. 188, 205–215. Eldredge, N., Gould, S.J., 1972. Punctuated equilibria, an alternative to phyletic gradualism. In: Schoff, T.J.M. (Ed.), Models in Paleobiology. Freeman Cooper, San Francisco, pp. 82–115. Encyclopaedia Britannica (2006). Encyclopaedia Britannica Premium Service. http://www.britannica.com/eb/article-9039718/   Erwin, D.H., Davidson, E.H., 2002. The last common bilaterian ancestor. Development 129, 3021–3032. Evans, S.M., 1999. Vertebrate tinman homologues and cardiac differentiation. Semin. Cell Dev. Biol. 10, 73–83. Fahrner, A., Haszprunar, G., 2001. Anatomy and ultrastructure of the excretory system of a heart-bearing and a heart-less sacoglossan gastropod (Opisthobranchia, Sacoglossa). Zoomorphol. 121, 85–93. Fange, R., 1972. The circulatory system. In: Hardisty, M.W., Potter, I.C. (Eds.) Lampreys, vol. 2. Academic Press, New York, pp. 287–306. Farmanfarmaian, A., 1968. The controversial echinoid heart and hemal system function effectiveness in respiratory exchange. Comp. Biochem. Physiol. 24, 855–863. Farrell, A.P., Jones, D.R., 1992. The heart. In: Hoar, W.S., Randall, D.J., Farrell, A.P. (Eds.) Fish Physiology, Part A, The cardiovascular system, vol. XII. Academic Press, San Diego, pp. 1–24. Farrell, A.P., 1997. Evolution of cardiovascular systems, insights into ontogeny. In: Burggren, W.W., Keller, B.B. (Eds.) Development of Cardiovascular Systems: Molecules to Organisms. Cambridge University Press, Cambridge, pp. 101–113. Farrelly, C.A., Greenaway, P., 2005. The morphology and vasculature of the respiratory organs of terrestrial hermit crabs (Coenobita and Birgus), gills, branchiostegal lungs and abdominal lungs. Arthropod. Struct. Dev. 34, 63–87. Fawcett, D.W., 1994. A Textbook of Histology, twelfth ed. New York, Chapman & Hall. Fisher, P.E., Russell, D.A., Stoskopf, M.K., Barrick, R.E., Hammer, M., Kuzmitz, A.A., 2000. Cardiovascular evidence for an intermediate or higher metabolic rate in an ornithischian dinosaur. Science 288, 503–505. Fishman, M.C., Chien, K.R., 1997. Fashioning the vertebrate heart, earliest embryonic decisions. Development 124, 2099–2117. Fishman, M.C., Olson, E.N., 1997. Parsing the heart, genetic modules for organ assembly. Cell 91, 153–156. Forouhar, A.S., Liebling, M., Hickerson, A., Nasiraei-Moghaddam, A., Tsai, H.J., Hove, J.R., Fraser, S.E., Dickinson, M.E., Gharib, M., 2006. The embryonic vertebrate heart tube is a dynamic suction pump. Science 312, 751–753. Frasch, M., 1995. Induction of visceral and cardiac mesoderm by ectodermal Dpp in the early Drosophila embryo. Nature 374, 464–467. Frasch, M., 1999. Intersecting signaling and transcriptional pathways in Drosophila heart specification. Semin. Cell Dev. Biol. 10, 61–71. Fujiwara, S., Kawamura, K., 2003. Acquisition of retinoic acid signaling pathway and innovation of the chordate body plan. Zoolog. Sci. 20, 809–818. Gajewski, K., Choi, C.Y., Kim, Y., Schulz, R.A., 2000. Genetically distinct cardial cells within the Drosophila heart. Genesis 28, 36–43.

Chapter  |  1.1  Evolutionary Origins of Hearts

Gajewski, K., Fossett, N., Molkentin, J.D., Schulz, R.A., 1999. The zinc finger proteins Pannier and GATA4 function as cardiogenic factors in Drosophila. Development 126, 5679–5688. Galli, D., Domínguez, J.N., Zaffran, S., Munk, A., Brown, N.A., Buckingham, M.E., 2008. Atrial myocardium derives from the posterior region of the second heart field, which acquires left-right identity as Pitx2c is expressed. Development 135, 1157–1167. Gans, C., Northcutt, R.G., 1983. Neural crest and the origin of vertebrates: A new head. Science 220, 268–274. Garcia-Martinez, V., Schoenwolf, G.C., 1993. Primitive-streak origin of the cardiovascular system in avian embryos. Dev. Biol. 159, 706–719. Garey, J.R., 2001. Ecdysozoa: The relationship between Cycloneuralia and Panarthropoda. Zool. Anz. 240, 321–330. Gee, H., 2001. Deuterostome phylogeny: The context for the origin and evolution of the vertebrates. In: Ahlberg, P. E. (Ed.), Major Events in Early Vertebrate Evolution, Systematics Association Special Volume, Serie 61. Taylor & Francis, London and New York. Georgias, C., Wasser, M., Hinz, U., 1997. A basic-helix-loop-helix protein expressed in precursors of Drosophila longitudinal visceral muscles. Mech. Dev. 69, 115–124. Gilbert, S.F., 2006. Developmental Biology, eighth ed. Sinauer Associates, Inc., Sunderland, MA. Giribet, G., Distel, D.L., Polz, M., Sterrer, W., Wheeler, W.C., 2000. Triploblastic relationships with emphasis on the acoelomates and the position of Gnathostomulida, Cycliophora, Platyhelminthes, and Chaetognatha, a combined approach of 18S rDNA sequences and morphology. Syst. Biol. 49, 539–562. Giribet, G., Edgecombe, G.D., Wheeler, W.C., 2001. Arthropod phylogeny based on eight molecular loci and morphology. Nature 413, 157–161. Goodrich, E.S., 1930. Studies on the Structure and Development of Vertebrates. Macmillan and Company, London. Goodson, H.V., Spudich, J.A., 1993. Molecular evolution of the myosin family: Relationships derived from comparisons of amino acid sequences. Proc. Natl. Acad. Sci. USA 90, 659–663. Gosline, J.M., 1971. Connective tissue mechanics of Metridium senile. I. Structural and compositional aspects. J. Exp. Biol. 55, 763–774. Greenberg, M.J., 1979. Discussion: Common themes in cardiac control. Am. Zool. 19, 175–190. Grens, A., Gee, L., Fisher, D.A., Bode, H.R., 1996. CnNK2, an NK2 homeobox gene, has a role in patterning the basal end of the axis in hydra. Dev. Biol. 180, 473–488. Grimes, A.C., Stadt, H.A., Shepherd, I.T., Kirby, M.L., 2006. Solving an enigma: Arterial pole development in the zebrafish heart. Dev. Biol. 290, 265–276. Grimmer, J.C., Holland, N.D., 1979. Haemal and coelomic circulatory systems in the arms and pinnules of Florometra serratissima (Echinodermata, Crinoidea). Zoomorphology 94, 93–109. Guerrero, A., Icardo, J.M., Duran, A.C., Gallego, A., Domezain, A., Colvee, E., Sans-Coma, V., 2004. Differentiation of the cardiac outflow tract components in alevins of the sturgeon Acipenser naccarii (Osteichthyes, Acipenseriformes): Implications for heart evolution. J. Morphol. 260, 172–183. Guyton, A.C., Hall, J.E., 2000. Textbook of Medical Physiology, tenth ed. W. B. Saunders, Philadelphia, PA. Habets, P.E., Moorman, A.F., Clout, D.E., van Roon, M.A., Lingbeek, M., van Lohuizen, M., Campione, M., Christoffels, V.M., 2002. Cooperative action of Tbx2 and Nkx2.5 inhibits ANF expression in the atrioventricular canal: Implications for cardiac chamber formation. Genes Dev. 16, 1234–1246.

41

Halanych, K.M., 1995. The phylogenetic position of the pterobranch hemichordates based on 18S rDNA sequence data. Mol. Phylogenet. Evol. 4, 72–76. Halanych, K.M., 2004. The new view of animal phylogeny. Annu. Rev. Ecol. Evol. Syst. 35, 229–256. Halanych, K.M., Bacheller, J.D., Aguinaldo, A.M., Liva, S.M., Hillis, D.M., Lake, J.A., 1995. Evidence from 18S ribosomal DNA that the lophophorates are protostome animals. Science 267, 1641–1643. Hall, B.K., 2003. Descent with modification: The unity underlying homology and homoplasy as seen through an analysis of development and evolution. Biol. Rev. Camb. Philos. Soc. 78, 409–433. Han, Z., Fujioka, M., Su, M., Liu, M., Jaynes, J.B., Bodmer, R., 2002. Transcriptional integration of competence modulated by mutual repression generates cell-type specificity within the cardiogenic mesoderm. Dev. Biol. 252, 225–240. Hardisty, M.W., 1979. Biology of the Cyclostomes. Chapman and Hall, London. Harrelson, Z., Kelly, R.G., Goldin, S.N., Gibson-Brown, J.J., Bollag, R.J., Silver, L.M., Papaioannou, V.E., 2004. Tbx2 is essential for patterning the atrioventricular canal and for morphogenesis of the outflow tract during heart development. Development 131, 5041–5052. Harris, M.P., Williamson, S., Fallon, J.F., Meinhardt, H., Prum, R.O., 2005. Molecular evidence for an activator-inhibitor mechanism in development of embryonic feather branching. Proc. Natl. Acad. Sci. USA 102, 11734–11739. Harvey, R.P., 1996. NK2 homeobox genes and heart development. Dev. Biol. 178, 203–216. Hasegawa, Y., 2000. Isolation of a cDNA encoding the motor domain of nonmuscle myosin which is specifically expressed in the mantle pallial cell layer of scallop (Patinopecten yessoensis). J. Biochem. 128, 983–988. Haun, C., Alexander, J., Stainier, D.Y., Okkema, P.G., 1998. Rescue of Caenorhabditis elegans pharyngeal development by a vertebrate heart specification gene. Proc Natl Acad Sci USA 95, 5072–5075. Hejnol, A., Martindale, M.Q., 2008. Acoel development supports a simple planula-like urbilaterian. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 363 (1496), 1493–1501. Hernandez-Nicaise, M.L., Franc, J.M., 1993. Embranchement des Cténaires. In: Grassé, P.-P. (Ed.), Traitéde Zoologie. Anatomie, Systématique, Biologie. Cnidaires, hydrozoaires, scyphozoaires, cubozoaires, cténaires, Tome III, fascicule 2, Masson, Paris, pp. 943–1055. Herreid, C.F., LaRussa, V.F., DeFesi, C.R., 1976. Blood vascular system of the sea cucumber, Stichopus moebii. J. Morphol. 150, 423–452. Hertel, W., Pass, G., 2002. An evolutionary treatment of the morphology and physiology of circulatory organs in insects. Comp. Biochem. Physiol. A. Mol. Integr. Physiol. 133, 555–575. Hertel, W., Wirkner, C.S., Pass, G., 2002. Studies on the cardiac physiology of Onychophora and Chilopoda. Comp. Biochem. Physiol. A. Mol. Integr. Physiol. 133, 605–609. Hickerson, A.I., Rinderknecht, D., Gharib, M., 2005. Experimental study of the behavior of a valveless impedance pump. Exps. Fluids 38, 534–540. Hirakow, R., 1985. The vertebrate heart in phylogenetic relation to the prochordates. Forschr. Zool. 30, 367–369. Hochgreb, T., Linhares, V.L., Menezes, D.C., Sampaio, A.C., Yan, C.Y., Cardoso, W.V., Rosenthal, N., Xavier-Neto, J., 2003. A caudorostral wave of RALDH2 conveys anteroposterior information to the cardiac field. Development 130, 5363–5374. Hol, R., Johansen, K., 1960. A cinematographic study of the central circulation in the hagfish, Myxine glutinosa L. J. Exp. Biol. 37, 469–473.

42

Holland, N.D., Venkatesh, T.V., Holland, L.Z., Jacobs, D.K., Bodmer, R., 2003. AmphiNk2-tin, an Amphioxus homeobox gene expressed in myocardial progenitors: Insights into evolution of the vertebrate heart. Dev. Biol. 255, 128–137. Hollenberg, S.M., Sternglanz, R., Cheng, P.F., Weintraub, H., 1995. Identification of a new family of tissue-specific basic helix-loop-helix proteins with a two-hybrid system. Mol. Cell Biol. 15, 3813–3822. Hoogaars, W.M., Tessari, A., Moorman, A.F., de Boer, P.A., Hagoort, J., Soufan, A.T., Campione, M., Christoffels, V.M., 2004. The transcriptional repressor Tbx3 delineates the developing central conduction system of the heart. Cardiovasc. Res. 62, 489–499. Hu, N., Sedmera, D., Yost, H.J., Clark, E.B., 2000. Structure and function of the developing zebrafish heart. Anat. Rec. 260, 148–157. Hyman, L.H., 1940. The Invertebrates, Protozoa through Ctenophora. McGraw-Hill Book Company, New York. Hyman, L.H., 1942. Comparative Vertebrate Anatomy. The University of Chicago Press, Chicago. Hyman, L.H., 1951. Introduction to the bilateria. In: “The Invertebrates, Platyhelminthes and Rhynchocoela” The Acoelomate Bilateria, vol. 2. McGraw-Hill Book Company, New York, pp. 1–51. Hyman, L.H., 1955. The Invertebrates, Echinodermata, the Coelomate Bilateria. McGraw-Hill Book Company, New York. Janvier, P., 1996. “Early Vertebrates” Oxford Monographs in Geology and Geophysics, 33. Oxford University Press, Oxford. Jefferies, R.S., 1986. The Ancestry of the Vertebrates. British Museum (Natural History), London & Cambridge. Jefferies, R.S., 2001. The origin and early fossil history of the chordate acustico-lateralis system, with remarks on the echinoderm-hemichordate clade. In: Ahlberg, P. E. (Ed.), Major Events in Early Vertebrate Evolution, Systematics Association Special Volume, Serie 61, Taylor & Francis, London and New York. Johansen, K., Martin, A.W., 1965. Circulation in a giant earthworm, Glossoscolex giganteus. I. Contractile processes and pressure gradients in the large blood vessels. J. Exp. Biol. 43, 333–347. Jones, H.D., 1971a. Circulatory pressures in Helix pomatia L. Comp. Biochem. Physiol. 39, 289–295. Jones, H.D., 1983. The circulatory systems of gastropods and bivalves. In: Saleuddin, A.S.M., Wilbur, K.M. (Eds.) The Mollusca, Physiology, Part 2, vol. 5. Academic Press Inc, New York, pp. 189–238. Jones, J.C., 1971b. On the heart of the orange tunicate Ecteinascidia turbinata. Herdman Biol. Bull. 141, 130–145. Kardong, K.V., 2002. Vertebrates: Comparative Anatomy, Function, Evolution. McGraw-Hill, Boston, MA. Kardong, K.V., 2006. Vertebrates: Comparative Anatomy, Function, Evolution, third ed. McGraw-Hill, Boston, MA. Kassai, Y., Munne, P., Hotta, Y., Penttila, E., Kavanagh, K., Ohbayashi, N., Takada, S., Thesleff, I., Jernvall, J., Itoh, N., 2005. Regulation of mammalian tooth cusp patterning by ectodin. Science 309, 2067–2070. Keegan, B.R., Meyer, D., Yelon, D., 2004. Organization of cardiac chamber progenitors in the zebrafish blastula. Development 131, 3081–3091. Keegan, B.R., Feldman, J.L., Begemann, G., Ingham, P.W., Yelon, D., 2005. Retinoic acid signaling restricts the cardiac progenitor pool. Science 307, 247–249. Keiser, F., 1914. Untersuchungen uber die erste anlagen des Herzens, der beiden Langsgefassstamme und die Blutes bei Embryonen von Petromyzon planeri. Jena. Z. Naturw. 51, 579. Kelley, C., Blumberg, H., Zon, L.I., Evans, T., 1993. GATA-4 is a novel transcription factor expressed in endocardium of the developing heart. Development 118, 817–827.

PART  |  1  Heart Evolution

Kelly, R.G., 2005. Molecular inroads into the anterior heart field. Trends Cardiovasc. Med. 15, 51–56. Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell. 1, 435–440. Kilner, P.J., Yang, G.Z., Wilkes, A.J., Mohiaddin, R.H., Firmin, D.N., Yacoub, M.H., 2000. Asymmetric redirection of flow through the heart. Nature 404, 759–761. Koehl, M.A., 1977. Mechanical diversity of the connective tissue of the body wall of sea anemones. J. Exp. Biol. 69, 107–125. Kolsch, V., Paululat, A., 2002. The highly conserved cardiogenic bHLH factor Hand is specifically expressed in circular visceral muscle progenitor cells and in all cell types of the dorsal vessel during Drosophila embryogenesis. Dev. Genes. Evol. 212, 473–485. Komuro, I., Izumo, S., 1993. Csx, a murine homeobox-containing gene specifically expressed in the developing heart. Proc. Natl. Acad. Sci. USA 90, 8145–8149. Kriebel, M.E., 1968. Electrical coupling between tunicate heart cells. Life Sci. II. 7, 181–186. LaBarbera, M., 1990. Principles of design of fluid transport systems in zoology. Science 249, 992–1000. LaBarbera, M., Vogel, S., 1982. The design of fluid transport systems in organisms. Am. Sci. 70, 54–60. Lawson, R., 1979. The comparative anatomy of the circulatory system. In: Wake, M.H. (Ed.), Hyman’s Comparative Anatomy, third ed. University of Chicago Press, Chicago, pp. 448–454. Lemche, H., Wingstrand, K.G., 1959. The anatomy of Neopilina galatheae Lemche 1957 (Mollusca, Tryblidiacea). Galathea Rep. 3, 9–71. Lewis, J.G., 1981. The Biology of Centipedes. Cambridge University Press, Cambridge, UK. Liebling, M., Forouhar, A.S., Gharib, M., Fraser, S.E., Dickinson, M.E., 2005. Four-dimensional cardiac imaging in living embryos via postacquisition synchronization of nongated slice sequences. J. Biomed. Opt. 10 054001. Lilly, B., Galewsky, S., Firulli, A.B., Schulz, R.A., Olson, E.N., 1994. D-MEF2, a MADS box transcription factor expressed in differentiating mesoderm and muscle cell lineages during Drosophila embryogenesis. Proc. Natl. Acad. Sci. USA 91, 5662–5666. Lints, T.J., Parsons, L.M., Hartley, L., Lyons, I., Harvey, R.P., 1993. Nkx-2.5, a novel murine homeobox gene expressed in early heart progenitor cells and their myogenic descendants. Development 119, 419–431. Lowe Jr, C.J., Terasaki, M., Wu, M., Freeman, R.M., Runft, L., Kwan, K., Haigo, S., Aronowicz, J., Lander, E., Gruber, C., Smith, M., Kirschner, M., Gerhart, J., 2006. Dorsoventral patterning in hemichordates, insights into early chordate evolution. PLoS Biol. 4, 1603–1617. Maisey, J.G., 1986. Heads and tails: A chordate phylogeny. Cladistics 2, 201–256. Mallatt, J., Chen, J.Y., 2003. Fossil sister group of craniates, predicted and found. J. Morphol. 258, 1–31. Mallatt, J., Giribet, G., 2006. Further use of nearly complete 28S and 18S rRNA genes to classify Ecdysozoa, 37 more arthropods and a kino­ rhynch. Mol. Phylogenet. Evol. 40, 772–794. Mallatt, J.M., Garey, J.R., Shultz, J.W., 2004. Ecdysozoan phylogeny and Bayesian inference, first use of nearly complete 28S and 18S rRNA gene sequences to classify the arthropods and their kin. Mol. Phylogenet. Evol. 31, 178–191. Mann, T., Bodmer, R., Pandur, P., 2009. The Drosophila homolog of vertebrate Islet1 is a key component in early cardiogenesis. Development 136, 317–326.

Chapter  |  1.1  Evolutionary Origins of Hearts

Marletaz, F., Holland, L.Z., Laudet, V., Schubert, M., 2006. Retinoic acid signaling and the evolution of chordates. Int. J. Biol. Sci. 2, 38–47. Martin, A.W., 1980. Some invertebrate myogenic hearts, the hearts of worms and molluscs. In: Bourne, G.H. (Ed.), Hearts and Heart-Like Organs: Comparative Anatomy and Development, vol. 1. Academic Press Inc, New York, pp. 1–39. Martin, A.W., Johansen, K., 1965. Adaptations of the circulation in invertebrate animals. In: Hamilton, W.F., Dow, P. (Eds.), Handbook of Physiology, Sect. 2, vol. 3. American Physiological Society, Washington, pp. 2545–2581. Martin, J.F., Schwarz, J.J., Olson, E.N., 1993. Myocyte enhancer factor (MEF). 2C, a tissue-restricted member of the MEF-2 family of transcription factors. Proc. Natl. Acad. Sci. USA 90, 5282–5286. Martindale, M.Q., Henry, J.Q., 1999. Intracellular fate mapping in a basal metazoan, the ctenophore Mnemiopsis leidyi, reveals the origins of mesoderm and the existence of indeterminate cell lineages. Dev. Biol. 214, 243–257. Maynard, D.M., 1960. Circulation and heart function. In: Waterman, T.H. (Ed.), The Physiology of Crustacea. Metabolism and Growth, vol. 1. Academic Press, New York, pp. 161–226. McDermott, J.C., Cardoso, M.C., Yu, Y.T., Andres, V., Leifer, D., Krainc, D., Lipton, S.A., Nadal-Ginard, B., 1993. hMEF2C gene encodes skeletal muscle- and brain-specific transcription factors. Mol. Cell Biol. 13, 2564–2577. McGaw, I.J., 2005. The decapod crustacean circulatory system, a case that is neither open nor closed. Microsc. Microanal. 11, 18–36. McGaw, I.J., Reiber, C.L., 2002. Cardiovascular system of the blue crab Callinectes sapidus. J. Morphol. 251, 1–21. McMahon, B.R., 2001. Control of cardiovascular function and its evolution in Crustacea. J. Exp. Biol. 204, 923–932. McMahon, B.R., Bourne, G.B., Chu, K.H., 1997a. Invertebrate cardiovascular development. In: Burggren, W.W., Keller, B.B. (Eds.), Development of Cardiovascular Systems: Molecules to Organisms. Cambridge University Press, Cambridge, pp. 127–144. McMahon, B.R., Wilkens, J.L., Smith, P.J.S., 1997b. Invertebrate circulatory systems. In: Dantzler, W.H. (Ed.), Handbook of Physiology, Sect. 13, vol. 2. American Physiological Society, Oxford University Press, Inc, New Delhi, pp. 931–1008. McRae, C.A., Fishman, M.C., 2002. Zebrafish: the complete cardiovascular compendium. Cold Spring Harb. Symp. Quant. Biol., Symp. 67, 301–308. Meilhac, S.M., Esner, M., Kerszberg, M., Moss, J.E., Buckingham, M.E., 2004. Oriented clonal cell growth in the developing mouse myocardium underlies cardiac morphogenesis. J. Cell Biol. 164, 97–109. Meng, J., Hu, Y., Wang, Y., Wang, X., Li, C., 2006. A Mesozoic gliding mammal from northeastern China. Nature 444, 889–893. Metchnikoff, V.E., 1881. Uber die systematische stellung von Balanoglossus. Zoll. Anz. 4, 139–157. Millott, N., 1966. A possible function for the axial organ of Echinoids. Nature 209, 594–596. Millott, N., Vevers, H.G., 1964. Axial organ and fluid circulation in Echinoids. Nature 204, 1216–1217. Milne-Edwards, H., 1982. Recherches pour server a l’historie de la circulation du sang chez les annelids. Ann. Sci. Nat. Zool. Biol. Anim. Ser. 10, 193–221. Mjaatvedt, C.H., Nakaoka, T., Moreno-Rodriguez, R., Norris, R.A., Kern, M.J., Eisenberg, C.A., Turner, D., Markwald, R.R., 2001. The ­outflow tract of the heart is recruited from a novel heart-forming field. Dev. Biol. 238, 97–109.

43

Moller, P.C., Ellis, R.A., 1974. Fine structure of the excretory system of Amphioxus (Branchiostoma floridae) and its response to osmotic stress. Cell Tissue Res. 148, 1–9. Moller, P.C., Philpott, C.W., 1973. The circulatory system of Amphioxus (Branchiostoma floridae). I. Morphology of the major vessels of the pharyngeal area. J. Morphol. 139, 389–406. Moorman, A.F., Christoffels, V.M., 2003. Development of the cardiac conduction system, a matter of chamber development. In: Foundation, N., Chadwick, D.J., Goode, J. (Eds.), Development of the Cardiac Conduction System, vol. 250. CIBA Found Symp Ser. John Wiley & Sons, Chichester, UK. Munoz-Chapuli, R., Carmona, R., Guadix, J.A., Macias, D., PerezPomares, J.M., 2005. The origin of the endothelial cells, an evo-devo approach for the invertebrate/vertebrate transition of the circulatory system. Evol. Dev. 7, 351–358. Nakashima, K., Yamada, L., Satou, Y., Azuma, J., Satoh, N., 2004. The evolutionary origin of animal cellulose synthase. Dev. Genes Evol. 214, 81–88. Nguyen, H.T., Bodmer, R., Abmayr, S.M., McDermott, J.C., Spoerel, N.A., 1994. D-mef2, a Drosophila mesoderm-specific MADS box­containing gene with a biphasic expression profile during embryogenesis. Proc. Natl. Acad. Sci. USA 91, 7520–7524. Niederreither, K., Subbarayan, V., Dollé, P., Chambon, P., 1999. Embryonic retinoic acid synthesis is essential for early mouse post-implantation development. Nat. Genet. 21, 444–448. Niederreither, K., Vermot, J., Messaddeq, N., Schuhbaur, B., Chambon, P., Dollé, P., 2001. Embryonic retinoic acid synthesis is essential for heart morphogenesis in the mouse. Development 128, 1019–1031. Nielsen, C., 2003. Proposing a solution to the Articulata−Ecdysozoa controversy. Zool. Scr. 32, 475–482. Nübler-Jung, K., Arendt, D., 1996. Enteropneusts and chordate evolution. Curr. Biol. 6, 352–353. Nyitray, L., Jancsó, A., Ochiai, Y., Gráf, L., Szent-Györgyi, A.G., 1994. Scallop striated and smooth muscle myosin heavy-chain isoforms are produced by alternative RNA splicing from a single gene. Proc. Natl. Acad. Sci. USA 91, 12686–12690. Okkema, P.G., Fire, A., 1994. The Caenorhabditis elegans NK2 class homeoprotein CEH-22 is involved in combinatorial activation of gene expression in pharyngeal muscle. Development 120, 2175–2186. Olson, E.N., 2006. Gene regulatory networks in the evolution and development of the heart. Science 313, 1922–1927. Oota, H., Saitou, N., 1999. Phylogenetic relationship of muscle tissues deduced from superimposition of gene trees. Mol. Biol. Evol. 16, 856–867. Orts-Llorca, F., Collado, J.J., 1967. Determination of heart polarity (arterio venous axis) in the chicken embryo. Roux. Arch. Entwicklungsmechanik 158, 146–163. Osmond, M.K., Butler, A.J., Voon, F.C., Bellairs, R., 1991. The effects of retinoic acid on heart formation in the early chick embryo. Development 113, 1405–1417. Pass, G., 2000. Accessory pulsatile organs, evolutionary innovations in insects. Annu. Rev. Entomol. 45, 495–518. Passamaneck, Y., Halanych, K.M., 2006. Lophotrochozoan phylogeny assessed with LSU and SSU data, evidence of lophophorate polyphyly. Mol. Phylogenet. Evol. 40, 20–28. Patwardhan, V., Fernandez, S., Montgomery, M., Litvin, J., 2000. The rostro-caudal position of cardiac myocytes affect their fate. Dev. Dyn. 218, 123–135. Paul, R.J., Colmorgen, M., Hüller, S., Tyroller, F., Zinkler, D., 1997. Circulation and respiratory control in millimetre-sized animals

44

(Daphnia magna, Folsomia candida) studied by optical methods. J. Comp. Physiol. B. 167, 399–408. Percy, L.R., Potter, I.C., 1991. Aspects of the development and functional morphology of the pericardia heart and associated blood vessels of lampreys. J. Zool. 223, 49–66. Pereira, F.A., Qiu, Y., Zhou, G., Tsai, M.J., Tsai, S.Y., 1999. The orphan nuclear receptor COUP-TFII is required for angiogenesis and heart development. Genes Dev. 13, 1037–1049. Peterson, K.J., Eernisse, D.J., 2001. Animal phylogeny and the ancestry of bilaterians: Inferences from morphology and 18S rDNA gene sequences. Evol. Dev. 3, 170–205. Philippe, H., Lartillot, N., Brinkmann, H., 2005. Multigene analyses of bilaterian animals corroborate the monophyly of Ecdysozoa, Lophotrochozoa, and Protostomia. Mol. Biol. Evol. 22, 1246–1253. Pirow, R., Wollinger, F., Paul, R.J., 1999. The sites of respiratory gas exchange in the planktonic crustacean Daphnia magna, an in vivo study employing blood haemoglobin as an internal oxygen probe. J. Exp. Biol. 202, 3089–3099. Pough, F.H., Janis, C.M., Heiser, J.B., 2002. Vertebrate Life. Prentice Hall, Upper Saddle River, NJ. Prall, O.W., Menon, M.K., Solloway, M.J., Watanabe, Y., Zaffran, S., Bajolle, F., Biben, C., McBride, J.J., Robertson, B.R., Chaulet, H., Stennard, F.A., Wise, N., Schaft, D., Wolstein, O., Furtado, M.B., Shiratori, H., Chien, K.R., Hamada, H., Black, B.L., Saga, Y., Robertson, E.J., Buckingham, M.E., Harvey, R.P., 2007. An Nkx2-5/ Bmp2/Smad1 negative feedback loop controls heart progenitor specification and proliferation. Cell 128, 947–959. Putnam, N.H., Srivastava, M., Hellsten, U., Dirks, B., Chapman, J., Salamov, A., Terry, A., Shapiro, H., Lindquist, E., Kapitonov, V.V., Jurka, J., Genikhovich, G., Grigoriev, I.V., Lucas, S.M., Steele, R.E., Finnerty, J.R., Technau, U., Martindale, M.Q., Rokhsar, D.S., 2007. Sea anemone genome reveals ancestral eumetazoan gene repertoire and genomic organization. Science 317, 86–94. Rahr, H., 1979. Circulatory system of Amphioxus Branchiostoma lanceolatum (Pallas). – light microscopic investigation based on intra­vascular injection technique. Acta Zool. 60, 1–18. Rahr, H., 1981. The ultrastructure of the blood-vessels of branchiostomalanceolatum (Pallas). (Cephalochordata), 1. Relations between bloodvessels, epithelia, basal laminae, and connective-tissue. Zoomorphology 97, 53–74. Randall, D.J., 1970. The circulatory system. In: Hoar, W.S., Randall, D.J. (Eds.), Fish Physiology: The Nervous System, Circulation and Respiration. Academic Press, New York. Randall, D.J., Davie, P.S., 1980. The hearts of urochordates and cephalochordates. In: Bourne, G.H. (Ed.), Hearts and Heart-Like Organs: Comparative Anatomy and Development, vol. 1. Academic Press Inc, New York, pp. 41–59. Redkar, A., Montgomery, M., Litvin, J., 2001. Fate map of early avian cardiac progenitor cells. Development 128, 2269–2279. Regier, J.C., Shultz, J.W., Kambic, R.E., 2005. Pancrustacean phylogeny, hexapods are terrestrial crustaceans and maxillopods are not monophyletic. Proc. R. Soc. B. 272, 395–401. Ripplinger, J., 1957. Contribution à l’étude de la physiologie du coeur et de son innervation extrinsèque chez l’Escargot (Helix pomatia). Ann. Sci. Univ. Besançon. 2, 3–173. Robb, J.S., 1965. Comparative Basic Cardiology. Grune & Stratton, New York. Romer, A.S., 1962. Circulation of body fluids. In: The Vertebrate Body, second ed. W. B. Saunders & Company, Philadelphia, pp. 386–434.

PART  |  1  Heart Evolution

Rosenquist, G.C., deHaan, R.L., 1966. Migration of precardiac cells in the chick embryo, a radiographic study. Contrib. Embryol. 38, 111–121. Rosenthal, N., Xavier-Neto, J., 2000. From the bottom of the heart, antero­ posterior decisions in cardiac muscle differentiation. Curr. Opin. Cell. Biol. 12, 742–746. Rowe, T., McBride, E.F., Sereno, P.C., 2001. Dinosaur with a heart of stone. Science 291, 783. Ruppert, E.E., 1997. Aschelminthes. In: Harrison, F.W., Woollacott, R.M. (Eds.), Microscopic Anatomy of Invertebrates. Wiley-Liss, New York. Ruppert, E.E., 2005. Key characters uniting hemichordates and chordates: Homologies or homoplasies? Can. J. Zool. 83, 8–23. Ruppert, E.E., Carle, K.J., 1983. Morphology of metazoan circulatory systems. Zoomorphology 103, 193–208. Ruppert, E.E., Balser, E.J., 1986. Nephridia in the larvae of hemichordates and echinoderms. Biol. Bull. 171, 188–196. Ruppert, E.E., Barnes, R.D., 1994. Invertebrate Zoology, sixth ed. Saunders, Philadelphia. Ruppert, E.E., Fox, R.S., Barnes, R.D., 2004. Invertebrate Zoology: A Functional Evolutionary Approach, seventh ed. Brooks Cole/ Thomson, Belmont, CA. Ruzicka, D.L., Schwartz, R.J., 1988. Sequential activation of alpha-actin genes during avian cardiogenesis, vascular smooth muscle alphaactin gene transcripts mark the onset of cardiomyocyte differentiation. J. Cell Biol. 107, 2575–2586. Ryckebusch, L., Wang, Z., Bertrand, N., Lin, S.C., Chi, X., Schwartz, R., Zaffran, S., Niederreither, K., 2008. Retinoic acid deficiency alters second heart field formation. Proc. Natl. Acad. Sci. USA 105, 2913–2918. Saga, Y., Kitajima, S., Miyagawa-Tomita, S., 2000. Mesp1 expression is the earliest sign of cardiovascular development. Trends Cardiovasc. Med. 10, 345–352. Satchell, G.H., 1991a. Physiology and Form of Fish Circulation. Cambridge University Press, Cambridge. Satchell, G.H., 1991b. The heart. In: Physiology and Form of Fish Circulation. Cambridge University Press, Cambridge, pp. 10–40. Satchell, G.H., 1991c. Venous return and venous pumps. In: Physiology and Form of Fish Circulation. Cambridge University Press, Cambridge, pp. 126–140. Satchell, G.H., 1991d. Myxine, a speculative conclusion. In: Physiology and Form of Fish Circulation. Cambridge University Press, Cambridge, pp. 183–197. Satchell, G.H., 1992. The venous system. In: Hoar, W.S., Randall, D.J., Farrell, A.P. (Eds.), Fish Physiology. Academic Press, San Diego, CA. Satoh, Y.I., Nitratori, T., 1980. On the fine structure of lymph hearts in amphibia and reptiles. In: Bourne, G.H. (Ed.), Hearts and Heart-Like Organs: Comparative Anatomy and Development, vol. 1. Academic Press Inc, New York, pp. 149–169. Sawada, A., Fritz, A., Jiang, Y.J., Yamamoto, A., Yamasu, K., Kuroiwa, A., Saga, Y., Takeda, H., 2000. Zebrafish Mesp family genes, mesp-a and mesp-b are segmentally expressed in the presomitic mesoderm, and Mesp-b confers the anterior identity to the developing somites. Development 127, 1691–1702. Sawyer, R.T., 1986. Leech Biology and Behaviour. Oxford University Press, Oxford. Schaeffer, B., 1987. Deuterostome monophyly and phylogeny. Evol. Biol. 21, 179–235. Schmidt-Nielsen, K., 1990. Animal Physiology, Adaptation and Environment. Cambridge University Press, Cambridge. Schubert, M., Escriva, H., Xavier-Neto, J., Laudet, V., 2006. Amphioxus and tunicates as evolutionary model systems. Trends Ecol. Evol. 21, 269–277.

45

Chapter  |  1.1  Evolutionary Origins of Hearts

Scott, M.P., 1994. Intimations of a creature. Cell 79, 1121–1124. Seipel, K., Schmid, V., 2005. Evolution of striated muscle: Jellyfish and the origin of triploblasty. Dev. Biol. 282, 14–26. Shimizu, H., Fujisawa, T., 2003. Peduncle of Hydra and the heart of higher organisms share a common ancestral origin. Genesis 36, 182–186. Shimizu, H., Koizumi, O., Fujisawa, T., 2004. Three digestive movements in Hydra regulated by the diffuse nerve net in the body column. J. Comp. Physiol. A. 190, 623–630. Shu, D.G., Luo, H.L., Conway-Morris, S., Zhang, X.L., Hu, S.X., Chen, L., Han, J., Zhu, M., Li, Y., Chen, L.Z., 1999. Lower Cambrian vertebrates from South China. Nature 402, 42–46. Shu, D.G., Morris, S.C., Han, J., Zhang, Z.F., Yasui, K., Janvier, P., Chen, L., Zhang, X.L., Liu, J.N., Li, Y., Liu, H.Q., 2003. Head and backbone of the Early Cambrian vertebrate Haikouichthys. Nature 421, 526–529. Silva, J.R., Mendes, E.G., Mariano, M., 1995. Wound repair in the Amphioxus (Branchiostoma platae), an animal deprived of inflammatory phagocytes. J. Invertebr. Pathol. 65, 147–151. Simoes-Costa, M.S., Azambuja, A.P., Xavier-Neto, J., 2008. The search for non-chordate retinoic acid signaling: Lessons from chordates. J. Exp. Zoolog. B. Mol. Dev. Evol. 310 (1), 54–72. Simoes-Costa, M.S., Vasconcelos, M., Sampaio, A.C., Cravo, R.M., Linhares, V.L., Hochgreb, T., Yan, C.Y., Davidson, B., Xavier-Neto, J., 2005. The evolutionary origin of cardiac chambers. Dev. Biol. 277, 1–15. Simpson, G.G., 1965. The Meaning of Evolution. A Study of the History of Life and of Its Significance to Man. Yale University Press, New Haven and London. Sirbu, I.O., Zhao, X., Duester, G., 2008. Retinoic acid controls heart anteroposterior patterning by downregulating Isl1 through the Fgf8 pathway. Dev. Dyn. 237, 1627–1635. Skramlik, E.V., 1938. Uber den Blutumlaufbei Amphioxus lanceolatus Y. Pubbl Staz zool. Napoli 17, 130–157. Srivastava, D., Cserjesi, P., Olson, E.N., 1995. A subclass of bHLH proteins required for cardiac morphogenesis. Science 270, 1995–1999. Stainier, D.Y., Fishman, M.C., 1992. Patterning the zebrafish heart tube: Acquisition of anteroposterior polarity. Dev. Biol. 153, 91–101. Stalsberg, H., DeHaan, R.L., 1969. The precardiac areas and formation of the tubular heart in the chick embryo. Dev. Biol. 19, 128–159. Stennard, F.A., Harvey, R.P., 2005. T-box transcription factors and their roles in regulatory hierarchies in the developing heart. Development 132, 4897–4910. Stennard, F.A., Costa, M.W., Lai, D., Biben, C., Furtado, M.B., Solloway, M.J., McCulley, D.J., Leimena, C., Preis, J.I., Dunwoodie, S.L., Elliott, D.E., Prall, O.W., Black, B.L., Fatkin, D., Harvey, R.P., 2005. Murine T-box transcription factor Tbx20 acts as a repressor during heart development, and is essential for adult heart integrity, function and adaptation. Development 132, 2451–2462. Sugi, Y., Lough, J., 1992. Onset of expression and regional deposition of alpha-smooth and sarcomeric actin during avian heart development. Dev. Dyn. 193, 116–124. Sun, Y., Liang, X., Najafi, N., Cass, M., Lin, L., Cai, C.L., Chen, J., Evans, S. M., 2007. Islet 1 is expressed in distinct cardiovascular lineages, including pacemaker and coronary vascular cells. Dev. Biol. 304, 286–296. Swalla, B.J., Cameron, C.B., Corley, L.S., Garey, J.R., 2000. Urochordates are monophyletic within the deuterostomes. Syst. Biol. 49, 52–64. The American Heritage Dictionary of the English Language. (2000). fourth ed. Houghton Mifflin Company, Boston, MA.

The

Snodgrass Tapes. http://www.life.umd.edu/entm/shultzlab/ snodgrass/  . Tiedemann, F., 1816. Anatomie der Röhren-Holothurie, des Pomeranzfarbigen Seesterns, und Stein-Seeigels [Anatomy of the tube holothurians, the orange seastars, and the stone sea urchins.] Joseph Thomannschen Buchdruckerei, Landshut, Germany. (Translated by A. Böttger and edited by J. M. Lawrence). Herizos Press, Tampa, FL. Tjonneland, A., Bkland, S., Nylund, A., 1987. Evolutionary aspects of the arthropod heart. Zool. Scr. 16, 167–175. van Veen, A.A., van Rijen, H.V., Opthof, T., 2001. Cardiac gap junction channels: Modulation of expression and channel properties. Cardiovasc. Res. 51, 217–229. Wada, H., Satoh, N., 1994. Details of the evolutionary history from invertebrates to vertebrates, as deduced from the sequences of 18S rDNA. Proc. Natl. Acad. Sci. USA 91, 1801–1804. Waldo, K.L., Kumiski, D.H., Wallis, K.T., Stadt, H.A., Hutson, M.R., Platt, D.H., Kirby, M.L., 2001. Conotruncal myocardium arises from a secondary heart field. Development 128, 3179–3188. Waxman, J.S., Keegan, B.R., Roberts, R.W., Poss, K.D., Yelon, D., 2008. Hoxb5b acts downstream of retinoic acid signaling in the forelimb field to restrict heart field potential in zebrafish. Dev. Cell. 15, 923–934. Weems, W.A., 1982. Intestinal wall motion, propulsion, and fluid movement: Trends toward a unified theory. Am. J. Physiol. 243, G177–G188. Weems, W.A., 1987. Intestinal fluid flow, its production and control. In: Johnson, L.R. (Ed.), Physiology of the Gastrointestinal Tract, second ed. Raven Press, New York, pp. 571–593. Wells, M.J., 1983. Circulation in cephalopods. In: Saleuddin, A.S.M., Wilbur, K.M. (Eds.), The Mollusca, Physiology, Part 2, vol. 5. Academic Press Inc, New York, pp. 239–290. Wigglesworth, V.B., 1974. The Principles of Insect Physiology, seventh ed. Chapman and Hall, London. Wilkens, J.L., 1999. Evolution of the cardiovascular system in Crustacea. Am. Zool. 39, 199–214. Wray, G.A., Abouheif, E., 1998. When is homology not homology? Curr. Opin. Genet. Dev. 8, 675–680. Xavier-Neto, J., Rosenthal, N., Silva, F.A., Matos, T.G., Hochgreb, T., Linhares, V.L., 2001. Retinoid signaling and cardiac anteroposterior segmentation. Genesis 31, 97–104. Xavier-Neto, J., Castro, R.A., Sampaio, A.C., Azambuja, A.P., Castillo, H.A., Cravo, R.M., Simoes-Costa, M.S., 2007. Parallel avenues in the evolution of hearts and pumping organs. Cell Mol. Life Sci. 64, 719–734. Yamagishi, H., Hirose, E., 1997. Transfer of the heart pacemaker during juvenile development in the isopod crustacean Ligia exotica. J. Exp. Biol. 200, 2393–2404. Yelon, D., Horne, S.A., Stainier, D.Y., 1999. Restricted expression of cardiac myosin genes reveals regulated aspects of heart tube assembly in zebrafish. Dev. Biol. 214, 23–37. Yuan, S., Schoenwolf, G.C., 2000. Islet-1 marks the early heart rudiments and is asymmetrically expressed during early rotation of the foregut in the chick embryo. Anat. Rec. 260, 204–207. Zaffran, S., Frasch, M., 2002. Early signals in cardiac development. Circ. Res. 91, 457–469. Zeng, L., Swalla, B.J., 2005. Molecular phylogeny of the protochordates, chordate evolution. Can. J. Zool. 83, 24–33.

Chapter 1.2

Development and Aging of the Drosophila Heart Rolf Bodmer1 and Manfred Frasch2 1

Development and Aging Program, NASCR Center, Burnham Institute for Medical Research, La Jolla, CA, USA Department of Biology, Developmental Biology Unit, University of Erlangen-Nürnberg, Erlangen, Germany

2

I.  Introduction The Drosophila heart is a linear tube that is reminiscent of the primitive heart tube in vertebrate embryos. In both invertebrates and vertebrates, the heart originates from embryologically equivalent, bilaterally symmetrical groups of meso­dermal cells. In the past one and a half decades, many of the genetic mechanisms of heart specification have been elucidated in Drosophila. Here, we summarize the functions and interactions of genes that subdivide the early mesoderm, orchestrate the specification and initial differentiation of the Drosophila heart, and lead to the structural and functional diversification of cell fates within this organ. The emerging picture of interplay between mesoderm-intrinsic transcription factors and inductive signals from the ectoderm has provided a framework that has been widely used as a prototype to explore the genetic basis of cardiogenesis in vertebrates. Inroads have also been made into understanding the processes of morphogenesis, the extensive remodeling of the larval heart and cardiac physio­ logy, and aging in adult Drosophila. Notably, many of the genes involved in the specification, differentiation and aging of the Drosophila heart have vertebrate counterparts with analogous cardiogenic functions, suggesting that basic molecular control mechanisms of heart development and function are conserved.

II.  Morphology and morphogenesis of the drosophila heart II.A.  Heart Structure The heart of Drosophila has a simple tubular structure and, apart from the accessory pulsatile organs for the wings, is probably the only structural component of the fly’s

Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

circulatory system (Poulson, 1950; Rizki, 1978; Bodmer, 1995; Tögel et al., 2008). Its major function is to pump the hemolymph, including the various types of blood cells and nutrients, through the body cavity in an open circulatory system. Because insects have elaborate tracheal systems, the heart is unlikely to be critical as a provider of oxygen. The heart of Drosophila is located at the dorsal midline (Fig. 1), which differs from vertebrates where the heart lies ventral to the gut. Because the Drosophila heart looks more like a pulsating blood vessel, it is also called the dorsal vessel, which is the term we will use from here on when describing the entire pulsatile organ. In the larva, the dorsal vessel extends along most of the body axis and is suspended at each segment border beneath the dorsal epidermis by the skeletal alary muscles (Fig. 1). The dorsal vessel consists of two major cell types; the inner, contractile muscle cells (the “cardial” or “myocardial” cells) are aligned in two rows flanked on each side by an outer row of pericardial cells (Fig. 1B). The two rows of cardial cells form a central cavity, generating the lumen of the heart (Rizki, 1978; Rugendorff et al., 1994; Haag et al., 1999). As for many other muscles, the cardial cells contain numerous muscle-specific structural proteins, such as actins, tropomyosins, myosin heavy chain and 3-tubulin. The cardial cells also express muscle-specific transcription factors, such as Mef2. The subcellular arrangement of myofilaments and adherence junctions has been described by Rugendorff et al. (1994) and Tepass and Hartenstein (1994) (see also Figs 5; 13) (Monier et al., 2005; Zaffran et al., 2006; Mery et al., 2008; Taghli-Lamallem et al., 2008). The pericardial cells are loosely associated with the cardial cells (Fig. 1B). They do not contain muscle-specific proteins as do the cardial cells, and are not contractile. The pericardial cells within the abdominal region have a function as nephrocytes (Das et al., 2008; Weavers et al., 2009),

47

48

PART  |  1  Heart Evolution

Figure 1  Morphology and cellular composition of the larval dorsal vessel. (A) Dorsal vessel from a third instar larva stained with phalloidin-FITC. The wider heart portion attached to fan-shaped alary muscles is seen in the right 1/3 of the image and the thinner aorta flanked by thin alary muscle fibers in the left 2/3. Dorsal somatic muscle fibers are positioned underneath. The myofibrils of the dorsal vessel are arranged in a spirally-intersecting fashion around the dorsal vessel. (B) Schematic drawing of the dorsal vessel at late embryonic stages. There are 104 cardioblasts, which are divided into Tinman (Tin)-expressing “working myocardial” cells and Seven-up (Svp)-expressing myocardial cells, which form ostia (inflow valves) within the heart portion. The anterior aorta (left) lacks Svp-positive cardioblasts. Pericardial cells flanking the myocardial tube are also diversified (not shown). The anterior aorta is flanked by the neuroendocrine ring gland and the hemocyte-forming lymph gland (TMC: Tinman myocardial cell; SMC: Seven-up (Svp) myocardial cell).

as well as a nonautonomous role in modulating differentiation and functional properties of the myocardium in the adult (Fujioka et al., 2005; H.-Y. Lim, T. Buechling, and R. B., unpublished). There are two broadly different populations of pericardial cells, one of which expresses the gene odd-skipped whereas the other is more heterogeneous and expresses several different markers such as even-skipped, tinman or ladybird (Fig. 1A,B; see also Fig. 9) (see for example, Ward and Skeath, 2000; Alvarez et al., 2003; Han and Bodmer, 2003). In the thoracic region, the oddskipped pericardial cells are arranged in two bilaterally symmetrical clusters, termed lymph glands, which serve as blood forming organs during larval stages (Fig. 1B) (Rizki, 1978). In the embryo, the blood cells derive from the head mesoderm (Tepass et al., 1994), which is distinct from the trunk mesodermal origin of the cardial and pericardial cells (see also Section II.B). Anterior to the lymph glands, surrounding the anterior-most part of the dorsal vessel, is the ring gland (Fig. 1B), an endocrine organ of which part is derived from the anterior lip of the ventral furrow (De Velasco et al., 2004). Some of the even-skipped pericardial cells within the thoracic region give rise to the adult wing hearts (Tögel et al., 2008).

During mid-embryonic stages of heart development, before the tube of the dorsal vessel has formed, no morphological differences are detected between heart precursors along the anterior–posterior axis. After tube formation, the anterior portion of the dorsal vessel (also called “aorta”) is narrower than the posterior portion (the heart proper) (Fig. 1). In later larval stages, a valve separates the aorta and the wider portion of the heart. Posterior to this aortic valve in the heart wall there is a set of small inflow valves, the ostia (Rizki, 1978). The openings and closings of the ostia control the flow of hemolymph and blood cells in coordination with the aortic valve. Thus, during the rhythmic contractions of the heart the ostia close and the aortic valve opens, causing the hemolymph to be pumped out through the aorta. When the heart expands again, the aortic valve closes and the ostia open to allow entry of the body fluid (see Ponzielli et al., 2002).

II.B.  Embryology of Heart Development In the trunk region of the Drosophila embryo, the mesoderm gives rise to four major derivatives: the skeletal (or somatic)

49

Chapter  |  1.2  Development and Aging of the Drosophila Heart

(A)

(B)

(C)

(D)

(E)

(F)

Figure 2  Mesoderm formation and differentiation of mesodermal tissues. Shown are cross-sections through embryos at progressively advanced developmental stages (ventral sides are down; dorsal (d) and ventral (v) borders of the germ bands in elongation stages (B–D) are as indicated in C). Embryos in (A–C) were stained for Twist protein and those in (D–F) for both Mef2 (light brown) and Tinman (dark brown). (A) Stage 7 (3 hours after fertilization). The mesoderm invaginates at the ventral side of the gastrulating embryo. (B) Stage 8 (3.5 hours). The mesoderm spreads dorsally (see arrows). (C) Stage 9 (4 hours). The mesoderm has reached the dorsal border of the ectoderm. (D) Late stage 11 (7 hours). Heart precursors (hp; circled) and visceral mesoderm (vm) have segregated from the somatic mesoderm (sm). Fat body primordia (between vm and sm) are unstained. (E) Stage 13 (10 hours). Both mesoderm and ectoderm extend towards the dorsal side. (F) Stage 15 (11.5 hours). Mesodermal primordia differentiate into somatic and visceral muscles (sm and vm), and into the cardioblasts (cb) and pericardial cells (pc) of the future dorsal vessel.

muscles of the body wall are arranged in a segmentally repeated pattern of syncytial myofibers; the visceral muscles surround the gut; the fat body is located between the somatic and visceral muscles; and the heart is located most dorsally. At the beginning of gastrulation the ventral third of the circumference of the blastoderm embryo, the presumptive mesoderm, forms a furrow along the ventral midline and invaginates into the interior of the embryo (Fig. 2A). Shortly thereafter, the mesodermal cell mass flattens and spreads dorsally to form a monolayer of cells in close apposition to the ectoderm (Fig. 2B,C). After the mesoderm has reached the dorsal ectoderm, the first morphologically visible subdivision of mesodermal lineages occurs as the dorsal half of the mesoderm segregates into an inner and outer mesodermal layer. The inner layer contributes to the gut musculature (dorsally) and to the fat body (ventrally); the outer layer gives rise to the heart and dorsal body wall muscles (Fig. 2D) (Dunin Borkowski, 1995). A portion of the mesodermal cells which have migrated most dorsally on either side of the embryo (in the trunk region) become specified as heart progenitor cells (Fig. 2D,E). The cardiac mesoderm then subdivides further into two rows of bilaterally symmetrical cells, the future cardial and pericardial cells (Figs 1B; 2F). In the context of the dorsal extension of the entire germ band,

the heart precursors from either side of the embryo move toward each other and form the linear heart at the dorsal midline. The assembly of the contractile heart tube occurs in a highly ordered fashion, in that the left- and righthand cardial precursors align perfectly with each other (Fig. 1B) and form a lumen between them. Thus, the structure and formation of the Drosophila heart appears to be relatively simple and is easily amenable to histological and genetic studies. A prerequisite for the usefulness of the Drosophila heart as an experimental system for studying heart development in general is the existence of sufficient similarities between the Drosophila and vertebrate heart. Superficially, however, the vertebrate heart looks very different from the Drosophila heart in that it consists of multiple chambers with numerous specialized cell types; it is looped and connected to an elaborate circulatory system. In contrast, the Drosophila heart is a linear structure composed of only a few cell types. Nevertheless, when comparing the very early developmental events between these species, several interesting similarities become apparent (for reviews see Bodmer, 1995; Harvey, 1996; Olson, 2006 and Chapter 1.1). Similar to the Drosophila heart, the vertebrate heart is also formed from bilaterally symmetrical rows of mesodermal cells, and these appear to be of equivalent embryological origin

50

in that they have migrated most distally from the point of invagination during gastrulation (Fig. 3) (Bodmer, 1995). Since the dorsal–ventral axis is reversed between vertebrates and invertebrates (Francois et al., 1994; DeRobertis and Sasai, 1996) the cardiac precursors end up dorsally in Drosophila, and first laterally, then ventrally, in vertebrates (anterior lateral plate mesoderm). As in Drosophila, the bilateral heart primordia of vertebrates also fuse together at the midline and initially form a linear heart tube. In addition, there is now a significant body of evidence that heart (and perhaps vascular) development initiates through molecular and cellular mechanisms that have been largely conserved during the evolution of invertebrates and vertebrates (Harvey, 1996; Hartenstein and Mandal, 2006; Olson, 2006). The most compelling evidence stems from studies comparing the gene tinman, which is required for heart formation in Drosophila, with the vertebrate gene family of Nkx2-5 and related genes (Fig. 3) (see Section IV.A and Chapter 9.1). These genes are similar not only in amino acid sequence and cardiac pattern of expression (Fig. 3), but they also appear to be functionally equivalent to tinman. When the vertebrate tinman-like genes are expressed in transgenic Drosophila, they are capable of rescuing some of the abnormalities caused by the lack of tinman function (Park et al., 1998; Ranganayakulu et al., 1998) and can mimic some effects of ectopic tinman expression within the dorsal vessel (Zaffran et al., 2006). Taken together, there seem to be significant molecular and embryological similarities between Drosophila and vertebrates, and thus Drosophila can serve as an excellent model for studying basic molecular-genetic principles during the early

Figure 3  Comparison of NK homeodomain gene expression (tinman and Nkx2-5) and heart tube formation in Drosophila and vertebrates. Left panels: Expression of tinman RNA in the heart-forming cells of Drosophila embryos before (top) and after (bottom) assembly of the cardiac tube at the dorsal midline. Right panels: Nkx2-5 expression in chick embryos before (top, cardiac crescent) and after (bottom, primitive heart tube) assembly of the primitive heart tube in the anterior ventral region of the embryo (micrograph provided by T. Schultheiss). Note that tinman/ Nkx2-5 marks the cardiac primordia in regions of equivalent embryological origin in both species.

PART  |  1  Heart Evolution

steps of cardiogenesis, and perhaps also at later stages (see Section X).

III.  Genetic control of the formation and dorsal expansion of the mesoderm III.A.  Twist and Snail In Drosophila, the formation of the mesoderm in ventral regions of blastoderm embryos is largely controlled autonomously, i.e., by mechanisms which act within the nuclei or cells that acquire mesodermal fates. This is in contrast to vertebrate embryos where mesoderm formation depends largely on inductive processes. Drosophila mesoderm formation is initiated by a nuclear gradient of the maternally provided, NFB-related morphogen Dorsal1 (reviewed in Rusch and Levine, 1996). Peak levels of nuclear Dorsal protein are present along the ventral midline of blastoderm embryos and are required to activate two zygotic genes, twist and snail, that are essential for mesoderm formation and encode transcription factors of the basic helix-loop-helix and zinc-finger protein families, respectively (Boulay et al., 1987; Thisse et al., 1987). Mutations of either of the two genes cause virtually identical phenotypes, which consist of the complete lack of invagination and mesoderm differentiation (Simpson, 1983; Grau et al., 1984). However, the two genes play different roles in mesoderm development, as snail appears to have a permissive role in mesoderm formation by repressing nonmesodermal genes in the prospective mesoderm, whereas twist has a key role in activating genes that are required for the pro­ cesses of invagination, patterning and differentiation of the mesoderm (Kosman et al., 1991; Leptin, 1991). Candidates or known target genes of twist at blastoderm stage include folded gastrulation, which codes for a secreted molecule controlling aspects of mesoderm invagination (Costa et al., 1994); heartless, which encodes an FGF-receptor homolog that is involved in mesoderm migration (see below); the homeobox gene tinman that is crucial for mesoderm patterning and heart formation (see Section IV.A); mef2, which encodes a MADS-domain transcription factor that functions in later aspects of muscle and heart differentiation (see VII.C.ii); zfh-1, a zinc-finger and homeobox-containing gene, which is required for the differentiation of a subset of heart cells (see Section IV.E), and mir-1, a micro-RNA encoding gene that is involved in late steps of heart and somatic muscle differentiation (Biemar et al., 2005; Kwon et al., 2005; Sokol and Ambros, 2005). A large number of 1. According to general practice in the Drosophila literature, we write names of genes in italics and lowercase (e.g., dorsal), and names of the corresponding gene products in plain text with the first letter in uppercase (e.g., Dorsal).

Chapter  |  1.2  Development and Aging of the Drosophila Heart

additional candidates of twist targets have recently been identified through genome-wide “ChIP-on-Chip” approaches (Sandmann et al., 2007; Zeitlinger et al., 2007).

51

IV.  Transcription factors controlling cardiac specification IV.A.  The NK Homeobox Gene tinman

III.B.  Heartless Heartless and its downstream effector gene dof are expressed uniformly in the early mesoderm in a twistdependent manner (Shishido et al., 1993; Vincent et al., 1998). In both heartless and dof mutants the formation of the dorsal vessel and midgut visceral mesoderm is severely reduced. However, the primary phenotype of these mutants appears to be the inability of the invaginated mesoderm mass to spread towards the dorsal ectoderm. Instead, the mesoderm flattens only partially, and only portions of it are fortuitously reaching more dorsal positions (Beiman et al., 1996; Gisselbrecht et al., 1996; Shishido et al., 1997; Vincent et al., 1998). These observations suggest that the extension of the mesoderm towards dorsal areas of the germ band is triggered by FGF-signaling. The FGF8-like ligands of Heartless, Pyramus and Thisbe exhibit a dynamic but rather broad pattern of expression in the overlying ectoderm during this migratory phase (Gryzik and Müller, 2004; Stathopoulos et al., 2004). Nevertheless, the activated Heartless receptor, as monitored with an anti-diphospho MAPK antibody, is restricted to the dorsolateral edges of the spreading mesoderm of wild-type embryos (Gabay et al., 1997). Hence, it has been argued that FGF signaling may be permissive for migration and that additional ectodermal cues are required for the local activation of MAPK via Heartless (Wilson et al., 2005). Why does a reduction of dorsal spreading of the meso­ derm cause defects in heart and visceral mesoderm formation? Independent data have shown that signals from dorsal ectodermal cells are required to induce heart and visceral mesoderm formation in the underlying cells of the dorsal mesoderm (see Section V.A). As discussed below, the major inducing signal in this process has been identified as Dpp, a secreted protein homologous to vertebrate bone morphogenetic proteins (BMPs). In mutants for Heartless pathway components, disruption of the dorsal extension of the mesoderm appears to prevent most cells from receiving the dorsally restricted Dpp signal. Thus, the loss of heart and visceral mesoderm in the absence of Heartless signaling can be explained largely by the inability of Dpp to induce a sufficient number of mesodermal cells to form these tissues. Consistent with this explanation, ectopic expression of Dpp in ventral areas of heartless mutants is able to restore the formation of both tissues (Beiman et al., 1996; Gisselbrecht et al., 1996); however, after completion of mesoderm spreading Heartless signaling also has a more specific function during the induction of certain pericardial cell progenitors (Michelson et al., 1998) (see also Chapter 2.1 for FGF in Ciona cardiogenesis).

IV.A.i.  The Function of tinman in Cardiac Specification The gene tinman is expressed initially in the entire mesoderm of the trunk (Bodmer et al., 1990) (Fig. 4A,B), as are heartless and twist (Fig. 2A–C), but it is neither required for gastrulation nor for dorsal migration of the mesoderm. Rather, the major role of tinman is in the initial specification of tissues in dorsal portions of the mesoderm (Azpiazu and Frasch, 1993; Bodmer, 1993). Indeed, in tinman mutants, all derivatives from the dorsal mesoderm appear to be absent. The heart and midgut musculature are not formed (Fig. 5A,B), and the absence of the dorsal muscle markers such as eve, msh, and of dorsal muscle specific enhancer/lacZ expression in tinman mutants indicate that dorsal body wall muscles (including muscles 1, 2, 9 and 10) also fail to be specified (Azpiazu and Frasch, 1993; Bodmer, 1993; Yin and Frasch, 1998). The phenotypes with these and additional markers demonstrate that tinman is essential for the specification of the progenitors of the dorsal vessel, visceral musculature and dorsal body wall muscles in the early mesoderm. By contrast, most tissues and cell types derived from ventral and ventrolateral portions of the mesoderm develop in the absence of tinman function, and thus the majority of the somatic muscles and the fat body are present. However, tinman does have functions in the development of some ventrally derived cell types, such as a distinct subset of ventrolateral muscle founders and their corresponding muscles, as well as a set of mesodermallyderived glia-like cells along the ventral midline of the mesoderm that are involved in guidance of peripheral nerves (Azpiazu and Frasch, 1993; Gorczyca et al., 1994).

IV.A.ii.  The Expression of tinman As suggested by its genetic function, tinman is a crucial link between the determination of mesoderm by twist and later events that lead to the differentiation of mesodermal sublineages. Indeed, tinman is expressed only minutes after Twist protein is present in the mesodermal anlagen at blastoderm stage and this expression depends on twist function (Fig. 4A,B) (also see below). On completion of the dorsal spreading of the mesoderm, tinman expression becomes restricted to the dorsal half of the mesodermal monolayer, the “dorsal mesoderm” (Fig. 4C) (Bodmer et al., 1990). The tinman expression maintained in the dorsal mesoderm is likely to be critical, since this is where tinman exerts its major functions during the specification of cardiac and visceral mesoderm. Accordingly, with a synthetic tinman allele that provides broad tinman expression only

52

PART  |  1  Heart Evolution

(A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

Figure 4  Expression of Tinman protein during embryogenesis. The developmental stages shown in (A–F) are similar to those shown in Fig. 2A–F. During gastrulation (A) and dorsal spreading of the mesoderm (B), Tinman is expressed ubiquitously in the trunk mesoderm. On completion of the dorsal spreading, Tinman expression becomes restricted to the dorsal mesoderm (dms). At later stages (D–F), Tinman is observed in heart precursors (hp), which segregate into a dorsal row of cardioblasts (cb) and ventrally adjacent rows of pericardial cells (pc). In addition, tinman mRNA is transiently expressed in the visceral mesoderm (vm), where detectable protein (but not mRNA) levels are maintained until late stages (see (F)). (G) Shows Tinman protein expression in the heart precursors (hp) and visceral mesoderm (vm) of an early stage 12 embryo (side view); and (H) shows Tinman in cardioblasts (cb) and pericardial cells (pc) of a stage 14 embryo (dorsal view).

in the early mesoderm (being driven by its native twistdependent enhancer but lacking the dorsal and cardiac enhancers; see below) the formation of cardiac and visceral muscle progenitors is not as complete as with an analogous allele containing both the early mesodermal and the dorsal mesodermal enhancers (Zaffran et al., 2006). However, it is notable that the embryos lacking specifically the dorsallyrestricted tinman expression phase do show a significant amount of residual cardiac and visceral muscle tissues. This observation points to the crucial role of additional dorsallyrestricted inputs that are needed together with tinman for the induction of these tissues (see below). After the mesoderm has morphologically subdivided, and just before the retraction of the germband, tinman RNA expression ceases in the forming visceral mesoderm and becomes exclusive to cardiac precursors, where its expression persists to adulthood (Figs 3; 4D–H)(Bodmer et al., 1990). As the heart tube forms, Tinman protein is present in a segmentally-repeated pattern in which four out of six myocardial and pericardial cells in each hemisegment

express Tinman protein, while the remaining two cells lack expression (Fig. 1B; Fig. 9A; Fig. 10A).

IV.A.iii.  The Function of tinman Within the Dorsal Vessel The dorsal vessel-specific expression of tinman suggested a role of this gene in the correct differentiation and morphogenesis of the heart, in addition to its requirement for the initial specification of cardiac progenitors. This expectation was confirmed by the findings with the synthetic tinman allele that retained the native early and dorsal mesodermal enhancers, but lacked the dorsal vessel-specific enhancer. Embryos with this genetic background do form a largely normal dorsal vessel and can grow into adult flies, even though there is no cardiac tinman expression (Zaffran et al., 2006). However, the analysis of subtype-specific markers for cardioblasts shows that there are cell fate switches (see Section VII.A). Moreover, during larval stages there are significant structural changes in the tinman-deficient

Chapter  |  1.2  Development and Aging of the Drosophila Heart

53

Figure 5  Loss-of-function phenotypes of tinman mutants. (A) Dorsal view of late stage wild-type control embryo stained with an enhancer trap for cardioblasts and with anti-Even-skipped for pericardial cells. (B) A tin null mutant embryo stained as the embryo in (A) is lacking the dorsal vessel. (C) Portion of a third instar larval dorsal vessel stained with phalloidin-FITC. The myofibrils show an intersecting spiral arrangement. (D) Portion of the dorsal vessel of a third instar larva, in which tinman expression is specifically abrogated in the myocardial cells of the dorsal vessel. The myofibrils are arranged in parallel with the long axis of the dorsal vessel. (E) Dorsal vessel of an adult wild-type fly stained with phalloidin-FITC. The larval posterior aorta has been remodeled into the adult heart (arrows). (F) Dorsal vessel in an adult fly, in which tinman expression is specifically abrogated in the myocardial cells as in (D). The adult heart is extremely hypotrophic (arrows).

dorsal vessel, which are characterized by a highly aberrant arrangement of the myofibrils (Fig. 5C,D). In adults the heart defects are even more dramatic and include severe hypotrophy combined with grave morphological, ultrastructural and functional defects (Fig. 5E,F) (Zaffran et al., 2006). The adults with this genetic background have much reduced lifespans.

IV.A.iv.  Conserved Molecular Aspects of the Function and Regulation of tinman Tinman and its vertebrate relatives encode a homeodomain of the NK type (Kim and Nirenberg, 1989) that has been shown to bind preferentially to DNA sites containing sequences with a CAAG core (with the canonical high affinity binding sequence being TCAAGTG) (Chen and Schwartz, 1995; Damante et al., 1996; Gajewski et al., 1997). In addition, the tinman homeodomain can dimerize with itself and with the NK homeodomain of Bag­ pipe (Zaffran and Frasch, 2005). The products of the NK homeo­box gene family also contain a conserved stretch of 10–12 amino acids (TN-domain) close to the amino terminus (Bodmer, 1995), which has homology to the engrailed repressor element (Smith and Jaynes, 1996). The homeodomain of tinman can interact with the Groucho corepressor, whereas the N-terminal 110 residues of tinman may contain a transcriptional activator domain. Based on these, as well as on genetic data, it has been proposed that tinman can act both as transcriptional activator and as a repressor in a context-specific manner (Choi et al., 1999; Zaffran et al., 2006). Unlike the tinman-related

gene products in vertebrates, tinman itself does not contain a conserved hydrophobic domain carboxy-terminal to the homeodomain (the NK2-specific domain, NK2-SD) (Harvey, 1996). Expression of full-length tinman-related genes, Nkx2-5, Nkx2-3 and Nkx2-7, from several vertebrate species in tinman mutant fly embryos is able to rescue the visceral mesoderm, but is less effective in restoring heart development (Park et al., 1998; Ranganayakulu et al., 1998; see Chapter 9.1). By contrast, expression of a chimeric tinman gene product, in which the 3 portion including the homeobox was replaced by the mouse Nkx2-5 homeobox and NK2-SD, restores heart and visceral mesoderm development, suggesting that the tinman and Nkx2-5 homeodomains are functionally interchangeable despite their limited similarity (65% within the homeodomain). In addition, this result suggests that the N-terminal sequence from the Drosophila tinman protein has an important and specific role in heart development. Within the dorsal vessel, tinman appears to have repressive functions towards some of its targets, such as the Dorsocross T-box genes, which can be exerted in a similar fashion by mouse Nkx2-5 (Zaffran et al., 2006). Together, these and other findings clearly demonstrate that tinman and its vertebrate relatives are similar in their structure and expression patterns (Fig. 3), and suggest that they are also equivalent in many functional aspects. At the level of the tinman locus, the dynamic expression of the tinman gene is controlled by a modular array of enhancer elements. Each of these enhancers drives the tinman gene in a specific temporal pattern and spatial domain (Yin et al., 1997). Among these, an intronic enhancer is broadly active in the early mesoderm, which is due to three functional

54

binding sites for Twist. Downstream of the tinman gene there is an enhancer driving cardioblast-specific expression in the four tinman-positive cardioblasts within each hemisegment (Yin et al., 1997). A third enhancer, also located downstream, is driving expression in the dorsal mesoderm prior to the specification of cardiac progenitors. This enhancer is targeted by Dpp, a Drosophila BMP (see below). A discrete enhancer recapitulating endogenous pericardial cell expression has not been identified; rather, regulatory sequences contributing to pericardial expression seem to be present both within the Dpp-responsive enhancer element and within the second intron of tinman (Xu et al., 1998; Venkatesh et al., 2000). As in Drosophila, a modular arrangement of enhancers acting in temporally and spatially distinct patterns, including several that are targeted by BMP signals, have also been described for the tinman-related gene Nkx2-5 in the mouse (Schwartz and Olson, 1999; see Chapter 9.1). Hence, in addition to the functional parallels between the tinman-related genes from Drosophila and vertebrates, several aspects of the regulation of these genes are shared among these organisms.

IV.B.  The Dorsocross T-box Genes The proteins encoded by the Dorsocross genes belong to the family of T-box factors, whose amino acid sequences within their T-box are equally close to vertebrate Tbx5 and Tbx6 (Reim et al., 2003). Similar to the Dorsocross genes in Drosophila, Tbx5 has a key role in cardiac development in vertebrate species. In Drosophila, there are three clustered Dorsocross genes, Doc1, Doc2 and Doc3, which are a result of relatively recent gene duplications and which appear to be genetically-redundant. Hence, their developmental functions were studied mainly by using deletions for all three of these genes. The Dorsocross genes feature a dynamic expression pattern in both ectodermal and mesodermal tissues (Reim et al., 2003). The first mesodermal expression occurs within segmentally-distributed quadrants of cells in the dorsal mesoderm at the stage when the spreading of the mesoderm underneath the ectoderm has been completed. During this period, Dorsocross expression overlaps with dorsally-restricted tinman expression (Fig. 6G). Dorsocross gene induction occurs in parallel with tinman, as it is unaffected in tinman mutant embryos (Reim and Frasch, 2005). Unlike tinman, Dorsocross expression is interrupted segmentally. During the ingression of the presumptive visceral mesodermal cells from the segmental dorsal areas that express tinman but not Dorsocross, the cells expressing both Dorsocross and tinman rearrange to form the continuous band of cardiogenic mesoderm along the dorsal margin of the mesoderm (Fig. 6H,I). Genetic data showed that the Dorsocross genes are critical cardiogenic factors. As in tinman mutants, no dorsal vessel is formed in their absence and both the myocardial and pericardial cells are missing. At earlier stages,

PART  |  1  Heart Evolution

none of the known markers for myocardial or pericardial progenitors, such as H15/neuromancer1 and midline/ neuromancer2 (nmr1&2, Drosophila Tbx20 genes), sevenup (COUP-TF2), Hand, or even-skipped are observed along the dorsal margin of the mesoderm of Dorsocross-deficient embryos (Reim and Frasch, 2005). Hence, the Dorsocross genes are required, apparently in combination with tinman, to promote early specification events of cardiac progenitors. This cardiogenic function is underscored by the data from ectopic expression experiments, in which combined forced expression of Dorsocross and tinman yielded stronger expansion of myocardial cell identities, as compared to the expression of either gene alone (Reim and Frasch, 2005). Notably, Dorsocross function is dosage-dependent, as in embryos with a reduced number of Dorsocross gene copies fewer myocardial cells are formed. This situation is reminiscent of mutations in Tbx5, which lead to haploinsufficiency and cause the heart defects of Holt-Oram syndrome (Mori and Bruneau, 2004; see Chapter 9.4). It appears that one of the earliest and critical functions of Dorsocross during cardiogenesis is the activation of the GATA gene pannier, which additionally requires tinman (Fig. 6I) (Gajewski et al., 1999; Reim and Frasch, 2005). As described in the following section, Pannier is also a critical cardiogenic factor during early myocardial specification, and is required for differentiation of most cardiac lineages (Alvarez et al., 2003; Klinedinst and Bodmer, 2003).

IV.C.  The GATA-Encoding Gene pannier Even though tinman and Dorsocross are absolutely required for cardiac specification, either one is insufficient to expand the cardiogenic region significantly at the dorsal margin of the mesoderm when overexpressed in the meso­ derm (Lockwood and Bodmer, 2002; Klinedinst and Bodmer, 2003; Reim and Frasch, 2005). This suggests that Tinman and Dorsocross may require additional factors within the context of the mesoderm to provide heartspecifying competence. Of the three GATA factor-type zincfinger genes found in Drosophila (pannier, serpent and grain) pannier is co-expressed with tinman and Dorsocross in a narrow dorsal domain during cardiac mesoderm induction (Fig. 6F,I) (Gajewski et al., 1999; Alvarez et al., 2003; Klinedinst and Bodmer, 2003). The expression of pannier in these cells requires both tinman and Dorsocross, and Tinman is known to act as a direct upstream activator that binds to a heart-specific enhancer of pannier (Gajewski et al., 2001; Reim et al., 2005). In pannier mutants or by mesodermal expression of a dominant-negative form of pannier (pnr-EnR), early cardiac tinman and Dorsocross expression as well as all cardiac lineages (myocardial and pericardial cells) (Figs 1B; 9A) are reduced but not abolished. When pannier is co-expressed with tinman and/or Dorsocross in the mesoderm, a synergistic expansion

Chapter  |  1.2  Development and Aging of the Drosophila Heart

55

Figure 6  Wild-type expression patterns of ectodermal wg (A–C, turquoise; G, H, blue) and dpp (A–C, purple) in comparison with mesodermal tinman (D–F, purple; G, H, red), Dorsocross (G–I, green) and Pannier (I, red). (A) Stage 9 wg and dpp patterns. dpp mRNA is expressed in a broad band in the dorsal ectoderm. wg (detected via wg-LacZ) is expressed in 15 continuous transverse stripes. (B) Stage 10 wg and dpp patterns. The broad band of dpp expression decreases during late stage 9 and is further reduced during stage 10. During stage 10 the continuous wg stripes become interrupted laterally and persist only in the dorsal and ventral regions. This interrupted pattern continues throughout embryogenesis. (C) Late stage 11 wg and dpp patterns. The dorsal line of dpp expression has reappeared strongly (arrowhead) and a second stripe of dpp appears at a lateral position (arrow), just dorsal to the CNS. The lateral interruption of the wg stripes is complete. (D) At stage 10 tin has become restricted to the dorsal mesoderm (by dpp, see text). (E) At mid-stage 11 tin is further restricted to segmental regions corresponding to the wg stripes, in two clusters. The dorsal cluster corresponds to heart precursors, and is dependent on wg signaling. (F) At late stage 11 tin expression is confined to the cardial and pericardial cells of the developing heart. (G) Early stage 10 embryo stained with antibodies against Wingless, Dorsocross, and Tinman. Dorsocross is induced in the Tin cells (which receive Dpp signals) underneath the ectodermal Wingless stripes. (H) Stage 11 embryo with laterally-interrupted Wingless stripes. The dorsal mesodermal cells underneath the Wingless domains that express Tin  Doc (yellow signals; white arrowheads) become specified as cardioblasts, whereas the dorsal cells in the cardiogenic mesoderm that contain only Tin (red signals, red arrowhead) will become pericardial cells. Dorsocross expression is also seen in a lateral cluster of somatic mesodermal cells (arrows). (I) At stage 12 Dorsocross and Pannier are co-expressed within developing cardioblasts (yellow signals) along the dorsal margin of the mesoderm. Micrographs courtesy of W. Lockwood and I. Reim.

and ectopic expression of heart markers or target genes is observed (e.g., Hand, SUR, mid/nmr2), suggesting that combinations of these genes are more effective in cardiac induction than either one alone (Klinedinst and Bodmer, 2003; Reim and Frasch, 2005). Therefore, it is likely that these transcription factors act as an interrelated group to induce cardiogenesis. Similarly, in mammals including humans, GATA-4 together with Nkx2-5 and Tbx5 are expressed in the early cardiac crescent and are critical for heart development and function (Kuo et al., 1997; Molkentin and Olson 1997; Schott et al., 1998; Bruneau et al., 2001; Garg et al., 2003; Oka et al., 2006; see Chapters 9.1–9.4). In vitro, GATA-4 interacts with a number of other transcription factors, including the multi-zinc-finger protein FOG-2, which primarily acts as a repressor (Durocher et al., 1997; Svensson et al., 1999; Tevosian et al., 1999), but which can also be an

activator (Lu et al., 1999; see Chapter 9.2). As for GATA-4, FOG-2 is present in developing and adult cardiomyocytes, and FOG-2-deficient mice exhibit severe developmental heart defects (Svensson et al., 2000; Tevosian et al., 2000). Pannier is also known for its requirement during closure of the embryonic and adult epidermis. The Drosophila counterpart of FOG-2, U-shaped, can physically interact with Pannier, and as with GATA-4 and FOG-2 this interaction is mediated by the N-terminal zinc-finger of Pannier, which antagonizes transcriptional activation (Cubadda et al., 1997; Haenlin et al., 1997). In contrast to pannier, u-shaped is not required for the initial specification of the cardiac mesoderm, but plays a role in maintaining cardiogenic gene expression and in correct differentiation of both myocardial and pericardial lineages (Fossett et al., 2001; Klinedinst and Bodmer, 2003). Interestingly, overexpression of u-shaped (and likewise, FOG-1 and FOG-2)

56

also diminishes cardiac differentiation and antagonizes the cardiogenic effect of pannier when co-overexpressed in the mesoderm, which is consistent with U-shaped acting in a complex that is disrupted when too much or too little U-shaped protein is present (Fossett et al., 2001).

PART  |  1  Heart Evolution

Mann et al., 2009). They appear to cooperate with each other in an interdependent genetic network, as well as with certain signaling activities (discussed in Section V), to activate downstream genes that promote the specification of different types of cardiac progenitors within the cardiac mesoderm.

IV.D.  The islet-1 Ortholog tailup The LIM homeodomain gene tailup, the Drosophila homolog of islet-1, is reported to be expressed in the dorsal mesoderm from stage 10, and by late embryonic stages it is expressed in all cardioblasts, alary muscles, as well as in the Odd pericardial and lymph gland cells (Tao et al., 2007; Mann et al., 2009). Based on the phenotype of mutations and deletions of tailup, this gene is not absolutely required for the formation of the cardiac mesoderm, but early cardiac expression of tinman, pannier and Dorsocross is to some extent reduced. In addition, genetic interaction studies suggest that tailup, tinman, pannier and Dorsocross strongly cooperate in early cardiogenesis (Mann et al., 2009). These observations suggest that tailup acts at a similar step in cardiac specification as pannier. Although its activity is not as critical as that of pannier, tailup appears to participate in the early cardiac regulatory network. In addition to its contribution to the specification of myocardial and pericardial cells, tailup activity in late stage embryos is also required for normal cardiac cell differentiation (Tao et al., 2007; Mann et al., 2009). At least in the pericardial cells, downstream genes of tailup include odd-skipped (odd) and the bHLH gene Hand (see Section VII.C.i). The expression of odd and a Hand-GFP reporter in pericardial cells (but not of Hand-GFP in most myocardial cells) is abolished in tailup mutants, and the results with a mutated Tailup binding site in the corresponding Hand enhancer indicate that Hand is a direct target of Tailup in these cells (Tao et al., 2007). Hence, like its ortholog islet-1 in vertebrate heart development, tailup is involved in early and late phases of cardiogenesis in Drosophila. However, because tailup is expressed and apparently active in all myocardial cells, there is currently no indication that it is functionally connected to the development of structures homologous to the “second heart field”, as is its vertebrate ortholog islet-1 (Buckingham et al., 2005; Chapter 2.2). Indeed, there is currently no indication that a homologous equivalent of such a field exists in Drosophila (note, however, that a few anterior myocardial cells are derived from the head mesoderm; see Section VII.B.ii). Based on the available loss-of-function and genetic interaction data with tinman, pannier, Dorsocross and tailup, as well as the data from single and combined ectopic expression, it was proposed that these genes encode the critical cardiogenic transcription factors that, in essence, define the heart-forming mesoderm (Klinedinst and Bodmer, 2003; Qian et al., 2005a; Reim and Frasch, 2005;

IV.E.  The Zinc-Finger Encoding Gene zfh1 and the Homeobox Gene eve in Pericardial Cell Specification In addition to its divergent homeodomain (Fortini et al., 1991) zfh-1 contains nine zinc-fingers and is expressed initially in all mesoderm, similar to tinman, heartless and mef 2, which depend on the mesoderm determinant twist. Later, zfh-1 is present in a variety of tissues, including the cardiac mesoderm (Lai et al., 1991). After formation of the heart tube, cardiac expression of zfh-1 becomes restricted primarily to the pericardial cells. Embryos which are zfh-1-mutant show moderate abnormalities, with variable penetrance, in the body wall muscle pattern and in heart muscle morphology (Lai et al., 1991; Su et al., 1999; Liu et al., 2006). The formation and cell type specification of most myocardial and pericardial cell lineages is not noticeably affected by loss of zfh-1 function, except for the even-skipped pericardial cells (EPCs, see below), which are completely absent (Su et al., 1999). This suggests that zfh-1 is not essential for early cardiac mesoderm specification, but rather for the formation of a specific pericardial sublineage (see Fig. 9A). Interestingly, mesodermal rescue with the mouse version of zfh-1 (SIP1, a ZFHX1B family member) (Verschueren et al., 1999) restores normal heart morphology and specification of the even-skipped pericardial cell lineage (Liu et al., 2006). Although expressed earlier as a pair-rule gene, the homeobox gene even-skipped (eve) is expressed during mid-embryonic development in segmentally-arranged, mesodermal clusters of 3–4 cells at the dorsal mesodermal edge among the other cardiac progenitor cells (Fig. 8A) (Frasch et al., 1987). Expression of eve in these cells is apparently essential for their specification and further differentiation into an epicardial-like envelope of the embryonic heart (Su et al., 1999; Fujioka et al., 2005). A discrete, well-defined enhancer 3 to the eve coding region is responsible for eve expression in this small subset of highly patterned mesodermal cells. The regulation of this well-studied enhancer will be discussed in Section V.C. After germband retraction, one of the mesodermal eve lineages will give rise to two EPCs in each hemisegment (Fig. 9). In zfh-1 mutants, the initial Eve clusters seem to be formed normally, whereas EPC differentiation is abolished (Su et al., 1999; Liu et al., 2006). The correct development of the EPCs also requires eve itself, since inactivation of eve function (using a temperature-sensitive

Chapter  |  1.2  Development and Aging of the Drosophila Heart

allele of eve or a mesodermal eve enhancer-deficient rescue construct) causes most EPCs to be missing or misplaced, and the overall number of pericardial cells in larvae or adults reduced (Su et al., 1999; Fujioka et al., 2005). The physiological consequences of mesodermal eve deficiency will be discussed in Section X.A.

V.  Combinatorial signals during cardiac induction V.A.  The Signaling Factors Encoded by decapentaplegic and wingless Immediately after gastrulation, the fate of mesodermal cells appears to be largely uncommitted (Beer et al., 1987). Consistent with this, expression of tinman is ubiquitous in the trunk mesoderm at this time, and only later does its expression become confined to the dorsal portion of the mesoderm where it is known to have its major function. Experiments in other insects had suggested that a likely source of patterning information for mesodermal subdivisions is the ectoderm, which the invaginated mesoderm is in close contact with (Seidel et al., 1940). Results from more recent experiments that used drug treatment to block gastrulation and the apposition of ectoderm and mesoderm, and others that blocked the spreading of the mesoderm towards the dorsal ectoderm genetically, suggest that the ectoderm has an instructive role in mesoderm patterning and cardiac induction (Baker and Schubiger, 1995; Maggert et al., 1995). Two secreted factors that are primarily expressed in the ectoderm have been shown to serve as inductive signals for patterning the mesoderm, and in particular, for the formation of the cardiac and visceral mesoderm. One of these molecules is the product of the gene decapentaplegic (dpp), a member of the bone morphogenetic protein subgroup of the TGF- superfamily proteins (for review, see Chen et al., 2004). The other secreted molecule is encoded by wingless (wg), a homolog of the vertebrate wnt-1 gene (for review, see Clevers, 2006; Chapter 1.3).

V.A.i.  Dorsolateral Signaling Inputs by dpp The Drosophila bone morphogenetic protein (BMP; see also Chapters 1.3 and 2.2) Dpp is expressed throughout the dorsal half of the embryo beginning at blastoderm stage, where it is required for the development of dorsal and lateral derivatives of the epidermis as part of the zygotic pathway that specifies the embryo­nic dorsal–ventral axis of ectodermal cuticular structures (Spencer et al., 1982; Ferguson and Anderson, 1992; Francois et al., 1994). After gastrulation, Dpp expression becomes restricted to the dorsal ectoderm where it continues to function in this developmental pathway of epidermal development. This maintained expression of Dpp also requires the ectodermal

57

activity of pannier (Klinedinst and Bodmer, 2003). Importantly, ectodermally-expressed dpp has a direct role in patterning the mesoderm during this same period. After gastrulation, when tinman expression becomes restricted to the dorsal mesoderm, its ventral limits coincide with the ventral limits of dpp in the dorsal ectoderm directly above (Fig. 6A,D,G). Mesodermal cells that do not “contact” dpp expressing ectodermal cells lose tinman expression shortly after the mesoderm has reached its dorsal margin. This led to the hypothesis that dpp may encode an inductive ectodermal signal necessary for patterning the mesoderm (Staehling-Hampton et al., 1994; Frasch, 1995). Indeed, in dpp-mutant embryos tinman ex­p­r­ession fades shortly after gastrulation, indicating that dpp signaling from the dorsal ectoderm is needed to maintain tinman expression in the dorsal mesoderm. A second essential function of Dpp in heart development during this period is the induction of Dorsocross expression in the dorsal mesoderm (Fig. 6G) (Reim and Frasch, 2005). Moreover, it is likely that the induction of pannier in the dorsal mesoderm, which occurs downstream of tinman and Dorsocross, also requires continued inputs from Dpp (Klinedinst and Bodmer, 2003). The absence of Dpp signaling activity in the dorsal mesoderm results in a failure to develop cardiac mesoderm, which can be explained by the failure to induce the expression of the cardiogenic factors (Tinman, Dorsocross, Pannier, Tailup, Midline/Neuromancer2 and others). In addition, loss of Dpp signaling results in the absence of visceral mesoderm and dorsal somatic muscles. Further evidence that dpp encodes a direct signal for mesodermal patterning stems from experiments where dpp is expressed ectopically. Expression of dpp in the entire ectoderm maintains tinman in the ventral mesoderm at a stage when its expression is normally confined to the dorsal mesoderm (Frasch, 1995). However, dpp-dependent patterning cannot exclusively be mediated by maintaining tinman expression in the dorsal mesoderm, since ubiquitous tinman expression alone at the time of its dorsal restriction is insufficient to confer dorsal mesodermal fates to ventral mesoderm. Furthermore, the phenotype with the synthetic allele of tinman that lacks the Dpp-responsive enhancer is much weaker with respect to the loss of dorsal mesodermal derivatives than that of dpp mutants (Zaffran et al., 2006). This is explained by the fact that additional patterning genes and cardiogenic factors are induced in response to dpp signaling within the dorsal mesoderm. Some of these factors, such as Dorsocross, are induced independently of tinman, but more typically Tinman and Dpp signals are required together for their induction. In summary, dpp is producing an ectodermal signal to specify dorsal mesoderm, which among other important effects involves the maintenance of tinman and the induction of Dorsocross expression. This appears to generate one of the first subdivisions of the mesoderm, leading to a distinction

58

between “dorsal” and “ventral” mesoderm that adopt separate developmental pathways. After the cardiac progenitors have been specified and are aligned along the dorsal margin of the mesoderm, Dpp expression becomes further restricted to the dorsal margin of the ectoderm, from where it is able to exert continued inputs into later events of cardiogenesis (Fig. 6B–I) (Lockwood and Bodmer, 2002; Johnson et al., 2003; Johnson et al., 2007). The Dpp receptors Punt and Thickveins, which encode serine/threonine transmembrane kinases, are known to med­ iate dpp-dependent signaling in the mesoderm. Muta­tion of thickveins and ectopic mesodermal expression of cons­titutively active Thickveins receptors have similar effects on tinman expression as mutation and overexpression of dpp, respectively (Yin and Frasch, 1998). Likewise, the phenotype of embryos lacking the maternal and zygotic activity of the Dpp effector Medea shows that Smad4 (and presumably Smad1/Mad) is also required (Xu et al., 1998). By contrast, mutation of schnurri, which codes for a zincfinger-containing transcription factor that confers a double-negative effect downstream of Dpp via repressing the expression of the transcriptional repressor Brinker, causes only a mild reduction of the dorsal tinman domains and heart formation (Yin and Frasch, 1998; Jazwinska et al., 1999; Marty et al., 2000). Hence, Dpp acts largely through the canonical pathway via activating the positively-acting Smad proteins Mad and Medea during cardiac induction and not, or at least not prominently, via blocking Brinker expression. The functional organization of the Dpp-responsive enhancer of tinman sheds further light on the mechanisms of Dpp-induced gene expression in the dorsal mesoderm. This enhancer includes several binding sites for the Smad proteins Mad and Medea, which feature either CG-rich or AGAC core sequences and are essential for enhancer activity. In addition, there are two Tinman binding sites juxtaposed with Smad sites, which are also critical for the response of this enhancer to Dpp (Xu et al., 1998). Hence, it appears that the tinman products derived from the early activation of the tinman gene by twist serve to make the dorsal mesodermal enhancer of tinman competent to res­ pond to Dpp. Once the dorsal mesodermal cells receive Dpp signals, it is the combination of bound Tinman and Dpp-dependent Smad proteins that is needed to activate this enhancer, and thus, induce tinman in the dorsal mesoderm. In addition, Tinman/Smad protein interactions and binding of a high mobility group (HMG) protein to both the enhancer and to Tinman are involved, and interestingly, an HMG protein is similarly involved in Nkx2-5 enhancer activation and Nkx2-5/Smad-mediated cardiogenesis in vertebrates (Zaffran et al., 2002; Monzen et al., 2008). Similar molecular mechanisms that utilize synergistic Tinman/Smad enhancer interactions are known or expected to be active during the induction of other Dpp and tinman targets, such as pannier, even-skipped, and other genes.

PART  |  1  Heart Evolution

V.A.ii.  Anterior–Posterior Signaling Inputs by wingless: Direct Effects on Cardiogenesis While dpp is required to induce all dorsal mesodermal derivatives, additional spatial cues play a role in distinguishing cardiac versus visceral versus somatic muscle cell fates within the dorsal mesoderm. A second important signaling molecule that functions in mesodermal patterning is encoded by wingless (wg), a Drosophila homolog of the mouse oncoprotein Wnt-1, which is a secreted glycoprotein (Baker, 1987). Like members of the TGF- family, Wnt family proteins are highly conserved in different species and play key roles in many different developmental processes (for review, see Clevers, 2006; Chapters 1.3 and 2.2). In Drosophila, wg is required for a variety of inductive signaling events during both embryonic and imaginal development. In the early embryo, wg is expressed in 15 transverse stripes in the trunk region (Fig. 6B,C,G,H) (van den Heuvel et al., 1989) and plays a central role in determining the anterior–posterior segmental polarity of the ectoderm. Notably, wg is also directly involved in heart formation (Wu et al., 1995; Park et al., 1996). In contrast to a loss of dpp function, which causes a failure of both cardiac and visceral mesoderm to form, wg mutants lack only the heart and its precursors, but do form visceral and some somatic mesoderm (albeit abnormally patterned, see Baylies et al., 1995; Azpiazu et al., 1996; Ranganayakulu et al., 1996). Thus, it appears that wg is involved in further subdividing the dorsal mesoderm and is necessary for specifying cardiac cell fates. Given the requirement for wg in ectodermal segmentation, a temperature-sensitive wg allele was instrumental in addressing the question of timing and specificity for wg in cardiac mesoderm formation. Elimination of wg function shortly after gastrulation, at a time when tinman becomes restricted to the dorsal mesoderm, results in the selective loss of heart progenitor cells with little effect on segmental patterning of the cuticle or other mesodermal derivatives (Wu et al., 1995). As expected, tinman expression is only lost in the presumptive cardiac precursors, and is normal in the early mesoderm and during its subsequent dorsal restriction. Although wg is expressed at its highest level in the ectoderm, transplantation experiments with syncytial blastoderm nuclei between wild-type and wg-mutant embryos have shown that the germlayer-specific origin of the wg activity is not of crucial importance, and cardiogenic wg activity can be provided either by the ectoderm or by the mesoderm (Lawrence et al., 1995). In light of the specific effects of wg on heart formation, it was important to determine whether the Wg pathway provides a direct cardiogenic signal or whether it exerts its effects by regulating the expression or activity of other segmentation genes. Indeed, another secreted factor encoded by hedgehog (hh) is also required for heart precursor formation during the same period as wg (Park et al., 1996). The interpretation of this effect is made complicated by the fact that

Chapter  |  1.2  Development and Aging of the Drosophila Heart

maintenance of wg expression after gastrulation requires hh function, and vice versa (Klingensmith and Nusse, 1994). This complex issue has been resolved by genetic epistasis experiments, which showed that wg activity is absolutely necessary for cardiogenesis, and cannot be substituted by hh or other segmentation genes that regulate wg function (Park et al., 1996). These genetic experiments have implicated the intracellular Wingless signal transdu­cers Dishevelled (Dsh; a conserved adapter protein), Zeste-white3/Shaggy (Zw3/Sgg; a GSK-3b-related serine/threonine kinase), Armadillo (Arm; a -catenin homolog), and Pangolin (Pan; a TCF/LEF-1-related HMG box-containing DNA binding protein) in the mesodermal response to Wingless immediately downstream of the Wg receptor, Dfz2, during early cardiogenesis (for Wg pathway see review by Dierick and Bejsovec, 1999). Taken together, these data clearly indicate that Wg is a direct signal for specifying cardiac cell fates. It is likely that the Dorsocross genes are among the key targets of the Wg signals in the dorsal mesoderm during early cardiac induction. In addition to being regulated by Dpp, these genes criti­ cally require the activity of wg, and thus become induced in the dorsal mesodermal quadrants of cells that receive combinatorial inputs of Dpp and Wg signals (Fig. 6G,H) (Reim and Frasch, 2005). Hence, the Dorsocross genes are important mediators of the combined Dpp and Wg signals during early cardiogenesis. It remains to be shown whether the Wingless signaling pathway, through the activity of Pangolin/dTCF, acts directly and in combination with Smads to control Dorsocross expression in these areas of the precardiac mesoderm.

V.A.iii.  Wg-Mediated Mesoderm Segmentation: Indirect Effects on Cardiac Development The striped expression of wg in the anterior (A) compartments of the ectoderm extends from the dorsal margin to the ventral midline. Histological examination and the analysis of early markers for heart precursors, including mid/nmr2, eve and ladybird (lb), confirmed that induction of heart precursors is restricted to areas below the wingless stripes, i.e., below the ectodermal A compartments. The segmentally-arranged clusters of cardiac progenitors subsequently merge to form continuous structures along the anterior–posterior axis (Fig. 7; top) (Dunin-Borkowski et al., 1995; Jagla et al., 1997a; Miskolczi-McCallum et al., 2005; Qian et al., 2005a; Reim et al., 2005). Within these segmental areas, heart induction occurs near the dorsal margins of the wingless stripes. Importantly, wingless functions also in more ventral areas of the mesoderm, where it is required for the development of somatic muscles and their progenitor cells (Baylies et al., 1995; D’Alessio and Frasch, 1996; Ranganayakulu et al., 1996). Thus, it appears that the development of the entire mesodermal areas below each of the wg stripes is under the influence

59

of wg, although the response to the Wg signal differs along the dorsal–ventral axis, which is at least in part due to the differential presence of Dpp signals. Although Wg signals presumably induce some cardiogenic and myogenic genes directly within these striped domains of the mesoderm (see previous paragraph and Section V.C), they also have indirect effects on cardiogenesis and myogenesis. These effects are due to the induction of the expression of the segmentation gene pair sloppy paired (slp), which encodes forkhead domain repressor factors related to vertebrate FoxG. The gene slp is a direct target of the Wg signaling cascade, and as a result is expressed in mesodermal (as well as ectodermal) stripes that are in register with the wingless stripes (Lee and Frasch, 2000). Genetically, slp is required for both heart and somatic muscle formation similar to wg (Park et al., 1996; Riechmann et al., 1997; Lee and Frasch, 2000). The known feed-back regulation from slp to wg (Cadigan et al., 1994), which only occurs in the ectoderm and not in the mesoderm, provides only part of the explanation for the essential role of slp in cardiogenesis and myogenesis. Within the mesoderm, the function of slp is strictly downstream of wg and is known to involve at least two distinct regulatory events. First, in the dorsal mesoderm slp blocks the induction of visceral mesodermal regulators, such as the NK homeobox gene bagpipe and the FoxFrelated forkhead domain gene biniou (Zaffran et al., 2001) by Dpp and Tinman within the cardiogenic quadrants of cells (Riechmann et al., 1997; Lee and Frasch, 2000). In the case of bagpipe, this effect is mediated by the direct binding of Slp repressor proteins to a Dpp- and Tinmanresponsive bagpipe enhancer (Lee and Frasch, 2000). Apparently, preventing the expression of specification genes of the visceral mesoderm (presumptive midgut musculature) within these areas is an essential prerequisite for cardiogenesis, as indicated by the observed loss of cardiogenic markers on forced pan-mesodermal expression of biniou (Zaffran et al., 2001). Hence, Wg signals are only able to induce cardiogenic regulators in cells in which the expression of visceral mesoderm regulators is blocked by Slp (Lee and Frasch, 2000). A second slp-dependent event occurs in lateral and ventral areas of the mesoderm, where slp is required for the upregulation of twist, which therefore assumes a striped pattern along the anteroposterior axis of the mesoderm (Fig. 7) (Riechmann et al., 1997). In turn, high levels of Twist appear to promote somatic myogenesis, whereas low levels or absence of Twist during this stage is thought to permit visceral mesoderm differentiation (Baylies and Bate, 1996).

V.B.  A Combinatorial Model for Specifying the Precardiac Mesoderm Beyond the genes involved in the initial specification of the mesoderm, we have discussed in some detail the

60

PART  |  1  Heart Evolution

Cardiac mesoderm/heart progenitors Visceral mesoderm (bagpipe domains) Dorsal mesoderm (tinman domains) Other mesoderm Dorsal ectoderm (dpp domains) Ectodermal (wingless domains) Ventrolateral ectoderm

Stage 8 (3.5h)

Stage 9 (45h)

Dp

p

Dpp

tin

slp wg

hh

Stage 10 (5h)

slp wg

Stage 11 (7.5h)

slp wg

Dp

hh

slp wg

Dorsal

p

Doc

bap

Doc

CM

tin Doc pnr

( )

VM

DSM

CM

tin Doc pnr

( )

DSM

sog slp

twi

slp wg Stage 9

slp hh Stage 10

twi

slp wg

slp twi slp wg

hh Stage 11

slp twi - Mesoderm slp wg - Ectoderm Ventral

Figure 7  Schematic drawings of early mesoderm development in wild-type embryos (top) and summary of genetic interactions during early mesoderm patterning (bottom). At the bottom, the left halves of the germband at stage 9 (to the left), stage 10 (middle) and stage 11 (to the right) are shown. At stage 9, Dpp induces the maintenance of tinman (tin) expression in the dorsal ectoderm, thus triggering a dorsoventral subdivision of the mesodermal layer. The activity of Dpp itself is limited to the dorsal ectoderm by short gastrulation (sog), which encodes an inhibitor of Dpp. During stage 10, Dorsocross is induced by Dpp  Wingless in segmentally-repeated areas within the dorsal mesoderm, whereas bagpipe is induced by Dpp in the segmental areas that lack active Wingless signals. In addition, there is cross-regulation between wg and slp in the ectoderm and induction of slp expression by Wg signals in the mesoderm, which upregulates Twist in striped domains. Altogether, these events result in anteroposterior subdivisions of the dorsal mesoderm, and thus, in differential determination of the primordia of the cardiac mesoderm (CM), visceral mesoderm (VM) and dorsal somatic mesoderm (DSM) (stage 11; see text).

function of a number of genes that play a role in specifying individual cell types within the mesoderm. Based on their functions and expression patterns, these genes can be divided into three basic groups. The first group contains genes that pattern the mesoderm along the dorsoventral axis. The primary members of this group, dpp and tinman, are needed for specifying the dorsal mesoderm, which leads to the formation of the heart, the visceral muscles and the dorsal somatic muscles (Azpiazu and Frasch, 1993; Bodmer, 1993; Staehling-Hampton et al., 1994; Frasch, 1995). The second group comprises genes that act in the anteroposterior patterning of the mesoderm and subdivide it into segmental and subsegmental units. These genes include wg and slp, which are required for the formation of the heart (and most skeletal muscles), but block the specification of visceral muscles (Wu et al., 1995; Park et al., 1996; Riechmann et al., 1997; Lee and Frasch, 2000; Lockwood and Bodmer, 2002). Another member of this group is hh, which is required for the efficient formation of visceral muscles, but not (at least not

directly) for the formation of the heart and the majority of somatic muscles (Azpiazu et al., 1996) (note, however, that Hh signals provide negative inputs to further subdivide the cardiac mesoderm; see Section V.C). Finally, genes of the third group are expressed in defined dorsoventral and anteroposterior areas in the mesoderm and control the tissue specification of the cells they are expressed in. The best-studied examples for genes in this third group are Dorsocross, pannier and tailup, which promote cardiogenesis, and bagpipe as well as biniou, which specify visceral mesoderm identities in their respective, mutually exclusive areas of expression (Azpiazu et al., 1993; Gajewski et al., 1999; Klinedinst et al., 2003; Reim et al., 2005; Mann et al., 2009). The combined action of these genes then leads to the activation of additional regulatory genes, including mid/nmr2, Hand, ladybird, eve and seven-up, in all or subsets of cardiac progenitors. Some of them contribute to maintaining tinman or Dorsocross expression, but their cross-regulatory interactions are not yet clearly established.

Chapter  |  1.2  Development and Aging of the Drosophila Heart

The expression patterns of genes from the first and second group intersect in every segment, and it is at these intersections where the genes of the third group are activated and serve to specify individual tissues. Specifically, Dorsocross induction requires combinatorial Dpp and Wg signals; therefore Dorsocross becomes expressed in the segmental quadrants of cells underneath the intersecting Dpp and Wg domains, where it promotes cardiac development (Reim and Frasch, 2005). A similar argument can be made for the subsequent restriction of tinman expression to the cardiac mesoderm at the intersection of Dpp and Wg in the dorsal mesodermal edge (Lockwood and Bodmer, 2002).2 Conversely, bagpipe and biniou activation requires intersections of the dorsal domains of dpp and tinman with the transverse stripes of Hh and the absence of Wg and Slp. The heart-forming third group of genes as a set may be selecting the cardiogenic fate, which is reminiscent of the subdivision of the embryonic body plan by the homeotic selector genes (McGinnis and Krumlauf, 1992). This would mean that once these cardiogenic selectors are activated they would no longer require the inductive patterning information provided by Dpp and Wg. This notion is supported by the results with forced ectopic expression of combinations of these genes, which provokes ectopic cardiogenesis (Klinedinst and Bodmer, 2003; Reim and Frasch, 2005; Akasaka et al., 2006; S. Klinedinst and R. B., unpubl.). It is not certain whether ectopic expression of these “cardiac selector” genes can override the requirement of potentially continued cardiogenic influences of wg and dpp. As a test, tinman and pannier have been overexpressed in the mesoderm, along with a dominant-negative form of dTCF/pangolin to reduce cardiogenic Wg signaling directly. Forced expression of tinman and pannier either alone or in combination was unable to rescue the loss of cardiogenesis (S. Klinedinst and R. B., unpubl.). Thus, Wg signaling is essential in addition to tinman and pannier, presumably being required for the activation of Dorsocross, which is initially tinman- and pannierindependent. It would be interesting to see whether forced mesodermal expression of Dorsocross with tinman or pannier was able to promote cardiogenesis with compromised Wg signaling. In summary, the convergence of wg, dpp, tinman, Dorsocross (see Figs 6; 7; 11) and probably additional yet unidentified patterning cues appear to be a pivotal and obligatory event in the specification of the heart. Since the intersection of wg and dpp also plays a role in the development of 2. It should be noted that, in these quadrants, an additional dorsoventral subdivision appears to occur, which restricts precardiac mesoderm formation to their dorsal-most areas. This indicates the existence of an additional yet unidentified cue, which is active in narrow domains along the dorsal margins of the germ band, and acts in combination with dpp, tin, wg and slp to determine precardiac mesoderm.

61

imaginal discs and other tissues of the embryo, wg and dpp are not instructive in terms of the specific tissue type that is formed. Rather, it is the mesoderm-specific expression of tinman (which then acts in conjunction with Dorsocross, pannier and others) that appears to “prime” the dorsal mesodermal cells to respond properly to the converging Wg and Dpp signals, which results in the formation of cardiac mesoderm at the dorsal edge of the mesoderm (see also Lockwood and Bodmer, 2002; Zaffran and Frasch, 2002). Taken together, the patterning events described above lead to an alternating arrangement of cardiac progenitors and visceral muscle precursors along the anterior–posterior axis within the dorsal mesoderm. Subsequently, morphogenetic rearrangements initiate, during which the visceral mesoderm precursors segregate towards the inside and the segmentally-derived clusters connect to form the internal layer of trunk visceral mesoderm. During the same process of cell rearrangements, the segmental clusters of cardiac progenitors remain in contact with the ectoderm and also connect along the anterior–posterior axis to form a continuous string of developing cardiac tissue along the dorsal margin of the mesoderm (Fig. 7).

V.C.  Early Diversification within the Cardiac Mesoderm V.C.i.  Hedgehog Organizes Anterior–Posterior Cell-Type Specification: Eve and Lbe The early set of cardiogenic transcription factors discussed above is expressed almost uniformly within the cardiac mesoderm (stage 11). However, these anlagen are subdivided rapidly in a metamerically (segmentally) repeated pattern of cellular identities that arise from specific lineages (Figs 8A; 9). This becomes first apparent during stage 11 as distinct groups of cardiac progenitors emerge within the tinman/Dorsocross/pannier-positive cardiogenic region of each segment (Figs 6F,I; 8A). Among them are those expressing the homeobox genes eve, ladybird early (lbe), the zinc-finger gene odd and the Coup-TFII transcription factor gene seven-up (svp), each marking a subpopulation of cardiac progenitors and progeny (Frasch et al., 1987; Jagla et al., 1997a; Su et al., 1999; Gajewski et al., 2000; Ward and Skeath, 2000; Lo and Frasch, 2001; Han et al., 2002; Liu et al., 2006). These genes are likely to contribute to the specification of the identity of the cardiac progenitors they are expressed in (Jagla et al., 2002). Although the segment polarity gene hh initially has an indirect role in cardiogenesis by maintaining wg expression (Park et al., 1996), it seems to be involved in generating the repeated pattern of identity gene expression independently of its requirement for maintaining wg expression (Azpiazu et al., 1996; Ponzielli et al., 2002; Liu et al., 2006). hh is required for the formation of two clusters of cardiac progenitors that are positioned in a

62

PART  |  1  Heart Evolution

(A)

Lbe Eve HhLacZ

(B)

Rho

sSpi

? EGF-R

Ht (FGF-R) Ras

Hh

MAPK

Pnt

Lbc

X

Cire

Eve

(C) wt

Meso>hh

hh Lbe

Mesoderm Ectoderm

Eve

Lbe Rho Hh

Lbe

Lbe

LbeRbe

Mesoderm

Mesoderm

Ectoderm

Ectoderm

Eve Rho

Rho

Hh

Figure 8  Hedgehog and RTK/Ras signaling in the diversification of cardiac progenitors. (A) Stage 12 embryos stained for Lbe (blue), Eve (pink) and hh-LacZ (green). Confocal scan of ectodermal Hh-lacZ staining was merged with the scan of mesodermal Lbe and Eve in a deeper focal plane to visualize the segmental register of the stainings. (B) Model of RTK/Ras and Hh signaling in the specification of Eve versus Lbe cells. Hh signaling promotes expression of Rho, an EGF receptor ligand activator, and/or blocks production of a repressor form of Ci, Cire. Hh may also play a Ras-independent role in repressing cardiac Lbe expression, again via Cire involving an unidentified negative regulator of Lbe (X). (C) The spatial relationship of Hh and RTK/Ras signaling and the responding cells along the anterior–posterior axis in the cardiac mesoderm. Hh, emanating from the posterior ectodermal compartment, promotes Rho expression and formation of Eve cells, which inhibits Lbe (B).

stereotyped and precise location relative to the segmentally-repeated ectodermal hh stripes; the eve-expressing clusters are immediately anterior (Fig. 8A) (Liu et al., 2006), and the svp-expressing cells are immediately posterior to the Hh stripes (Ponzielli et al., 2002). A third set of clusters expressing lbe are further removed anteriorly from the hh stripes (Fig. 8A,C). In a genetic combination where hh activity is abolished but wg expression is artificially maintained by forced transgene expression in “Wg stripes” or throughout the mesoderm, both the Eve- and the Svppositive cells are diminished or absent, but lbe expression is expanded to encompass most, if not all, of the cardiogenic region (Ponzielli et al., 2002; Liu et al., 2006). Eve progenitor formation also requires local activation of Ras signaling in the dorsal mesoderm (Carmena et al., 1998; Halfon et al., 2000). In hh-only mutants (hh-off, wg-on, see above), Ras signaling was diminished,

as evidenced by reduced expression of rhomboid (rho) (Liu et al., 2006), which codes for a protease that is required for proteolytically cleaving (and thereby activating) the EGF receptor ligand Spitz (Spi). Thus, hh signaling seems to determine the level of Ras pathway activation by transcriptionally-regulating rho (Fig. 8). Consistent with this idea is the observation that mesodermal overexpression of ras can partially restore eve expression in hh mutants (hh-off, wg-on). hh also functions independently of Ras in repressing lbe expression, perhaps through an indirect mechanism (Fig. 8B). Thus, anterior–posterior positioning of cardiac cell fates is set by ectodermal Hh signaling, which places the eve expressing cells near the Hh stripes and those expressing lbe more distantly (Fig. 8A,C) (Liu et al., 2006). This relative position is maintained and further reinforced by mutual transcriptional repression between eve and lbe (Fig. 8B) (Han et al., 2002; Jagla et al., 2002).

Chapter  |  1.2  Development and Aging of the Drosophila Heart

63

Figure 9  Summary diagrams of embryonic cardiac lineages and the effect of cell cycle arrest as well as Numb-Notch function in determining cardiac cell fates. Cardiac lineages of posterior aorta and heart region are shown. (A, top) Cell types formed in a wild-type (wt) hemisegment (black lines indicate lineage relationships). The effects of cell cycle arrest, numb mutants, and double mutants of numb and cycA. Note the changes in cell types formed in cell-cycle and numb mutants. (A, bottom). Summary diagram of the effect of numb/Notch manipulation. Block of cell divisions and asymmetric segregation of Numb into one daughter cell or blocking precursor division promotes myocardial cell fate by inhibition of Notch signaling. In contrast, activation of Notch signaling or the absence of Numb causes the daughter cell or undivided precursor to adopt a non myogenic pericardial cell fate (SMC: Svp myocardial cell; TMC: Tinman myocardial cell; SOPC: Svp-Odd pericardial cell; EPC: Eve pericardial cell; TMLC: Tinman-Lbe myocardial cell; LPC: Lbe pericardial cell; OPC: odd pericardial cell; DA1: dorsal acute muscle; DO2: dorsal oblique muscle 2; DA1sib; SSP: Svp-positive super progenitor; TSP: Tin-positive super progenitor). FEPC, FDO2, FDA1 and FDA1sib are founders of EPCs, DO2, DA1 and DA1sib. P2 and P15 are progenitors of the above founder cells. (B) Tinman and Eve double-labeling of wt; cycA and cycA;numb double mutants illustrate the changes in cell type specification. Note the differences in symmetrical versus asymmetrical lineages. Compared to wt, cycA mutants show half the number TMC and TLMC, DA1 muscle, but no EPCs. The cycA;numb double mutants also have half the number of Tin-positive cardioblasts, but not DA1 muscles. Instead there is one EPC per hemisegment.

V.C.ii.  msh Reinforces Restriction of Cardiac Cell Fates to the Dorsal Mesodermal Edge In addition to regulatory interactions between lbe and eve, which cross-repress each other to maintain distinct domains of expression (Fig. 8A), a third player, the homeobox transcription factor Msh, participates similarly in this system (Jagla et al., 1997a, 2002). These three identity genes specify a subset of heart and dorsal muscle progenitors: lbe a subset of myocardial and pericardial cells; eve a subset of pericardial and dorsal-most muscle founders; and msh a set of dorsal muscle founders. Loss- or gain-of-function of each of these genes changes the number of cells expressing the two

other genes, suggesting that they act in a cross-repressive network to ensure that correct ratios of myocardial, pericardial and dorsal muscle founders are formed (Jagla et al., 2002). For example, msh mutants or lbe overexpression causes cardiac hyperplasia, and msh overexpression reduces the number of tinman-expressing myocardial and pericardial cells, which includes Lbe and Eve cells dorsal to the Msh muscle founders. The cross-repressive interactions of these identity genes are likely to reinforce the demarcation of their localized expression domains over time. Based on these observations it has been proposed that eve, lbe and msh are “identity genes” that contribute to

64

proper cell fate specification of specific myocardial, pericardial and dorsal muscle progenitors downstream of cardiac mesoderm determination by tinman, pannier, Dorsocross, and tailup, as well as the inductive ectodermal signals Wg, Dpp and Hh, which desig­nate the localized expression of eve, lbe and msh expression (see Fig. 11).

VI.  Lineage decisions and notch signaling during the specification of cardiac cells The eve-expressing lineages give rise to two Eve pericardial cells (EPCs) and two dorsal muscles within each hemisegment. In contrast, the lbe expressing progenitors develop into two Lbe-positive myocardial cells and two Lbe-positive pericardial cells. The svp-expressing progenitors give rise to two Svp myocardial cells (the future inlet valves or ostia that no longer express tinman) and two Svp pericardial cells. Unlike the Svp progenitors, the Eve and Lbe progenitors also express Tinman. An additional set of Tinman progenitors not expressing eve or lbe generate the remaining two myocardial cells. Finally, a set of pericardial progenitors produces odd-skipped-expressing pericardial cells. In addition to the main part of the dorsal vessel described above, the anterior aorta is generated by a different set of lineages (for detailed lineage descriptions see Ward and Skeath, 2000; Alvarez et al., 2003; Han and Bodmer, 2003; reviewed in Bodmer et al., 2005). As the Drosophila heart develops, the myocardial cells are placed dorsally to the pericardial cells to form a continuous row of cells at the dorsal margin of the mesoderm, except for the eve-expressing EPCs whose nuclei are located dorsally and seem to wrap the developing myocardium in an epicardial-like fashion (Fujioka et al., 2005). The cardiac progenitors undergo a stereotyped set of cell divisions that require the activity of cell cycle genes (cyclinA and B) and determinants of asymmetric division (numb and Notch) (Ruiz Gomez and Bate, 1997; Park et al., 1998; Ward and Skeath, 2000; Han et al., 2002; Jagla et al., 2002; Alvarez et al., 2003; Han and Bodmer, 2003; for detailed review see Bodmer et al., 2005). For the main part of the dorsal vessel there are two sets of asymmetric lineages that are related to the eve and svp progenitors (Fig. 9A). As in the nervous system, membrane-associated protein Numb in the eve and svp lineages antagonizes the activity of transmembrane receptor Notch in the daughter cell inheriting asymmetrically localized Numb, which in turn leads to the distinction between progeny cell fates. In numb mutants there are supernumerary Eve and Svp pericardial cells, whereas on numb overexpression these pericardial cells are transformed towards a myogenic fate. In cell-cycle mutants and in embryos in which cellcycle inhibitors are overexpressed, cardiac progenitors con­ tinue to differentiate in the absence of cell division, but they

PART  |  1  Heart Evolution

preferentially adopt a myogenic cell fate in the case of asymmetric lineages (Fig. 9). This may be because Notch activation is prevented in the progenitor due to the presence of Numb that normally segregates to the myogenic sibling during division. To test this, cell division was arrested in numb mutants (or on overexpression of an activated form of Notch, which is insensitive to inhibition by Numb). Indeed, cell division arrested Eve progenitors adopt a nonmyogenic pericardial cell fate in the absence of Notch inhibition (Fig. 9). In contrast, the symmetric lineages are insensitive to the presence or absence of Notch (or Numb) activity. Thus, cell type diversity is increased by adopting the Numb/Notch system to implement asymmetry during cell division.

VII.  Axial patterning, diversification and differentiation of the myocardium VII.A.  Intrasegmental Myocardial Diversification Within the Developing   Dorsal Vessel VII.A.i.  Myocardial Diversification Via Regulatory Interactions Among seven-up, tinman and Dorsocross As described in Sections V. and VI., various events that include inductive signaling, mutual repression and asymmetric lineage decisions lead to a diverse array of cardiac progenitors that express distinct regulatory factors and are fated to generate different cardiac cell types. Within the myocardium, a major subdivision occurs between cardioblasts that express the COUP-TFII-related nuclear orphan receptor Seven-up (Svp; hence termed Svp cells) versus those that express Tinman (termed Tin cells) (see also Chapter 8.3.II.F). This mutually-exclusive pattern of expression is established during stage 13, after the expression of Svp has been established in a segmental subset of cardioblasts (downstream of Hh and early Tin) (Ponzielli et al., 2002; Ryan et al., 2007) (for other cardiogenic Hh functions see Section V.C and Liu et al., 2006) and in turn causes repression of tin in these cells (Lo and Frasch, 2001). The Svp cells in the myocardium also continue to express Dorsocross, whereas the Tin cells no longer express Doc after this stage (Fig. 10A,C). Consequently, among the six cardioblasts in each hemisegment within the dorsal vessel, two are positive for Svp and Doc, whereas the remaining four are positive for Tin (Figs 9; 10). The major exception to this pattern is in the anterior aorta, where no Svp cells are specified, and thus each hemisegment consists of four Tin cells. The distinction between Svp and Tin myocardial cells has very important consequences for the morphological differentiation and physiology of the dorsal vessel.

65

Chapter  |  1.2  Development and Aging of the Drosophila Heart

Figure 10  Intrasegmental diversification of myocardial cells and the role of Hox genes in the axial organization of the dorsal vessel. (A) The mutuallyexclusive expression of Tinman (blue) and Dorsocross (red)/Seven-up (green) in cardioblasts within each segment. The Tin cells are termed “working myocardial” cells, whereas the Svp/Doc cells will give rise to inflow valves (ostia) (arrow: ring gland). The three bilateral pairs of Svp/Doc cells form ostia during late embryonic stages, whereas the five anterior bilateral pairs form the adult ostia during the remodeling of the dorsal vessel at metamorphosis. (B) Late embryonic dorsal vessel stained for Pericardin (red) and Abdominal-A (green). Within the dorsal vessel, Abd-A is only expressed in the wider heart portion at the posterior. (C) Schematic summary of the expression patterns of segmental regulators and Hox genes within the dorsal vessel.

Orthologs or homologs Drosophila Dpp Doc Hand Hh Ibe pnr Mef2 mid/H15 (=nmr2/nmr1) Sur tin tup Wg

Vertebrates BMP Tbx5/6 hand 1/2 Shh Lbx1/2/3 GATA4/5/6 Mef2 Tbx20 sur1/2 Nkx2-5 Isl1 Wnt

Figure 11  Summary of known regulatory interactions during cardioblast development. Ectodermal signals are shown in yellow and mesodermal transcription factors in red. Regulatory interactions within ostial myocardial cells are shown in the red area and those occurring in working myocardial cells (“generic cardioblasts”) within the green area. The ocher area indicates events that are activated in both types of myocardial cells (albeit via different inputs) (cb: cardioblast).

66

Specifically, during late embryogenesis each of the three pairs of Svp/Doc myocardial cells within each hemisegment in the posterior portion of the dorsal vessel, which forms the larval heart, differentiate into an ostium (inflow valve). Although the other four pairs of Svp/Doc cells within the middle portion of the dorsal vessel (termed the posterior aorta) are not known to fulfill any specialized functions during larval stages, they do develop into ostia during metamorphosis, when this portion of the dorsal vessel is being remodeled into the adult heart (see Section IX). By contrast, the nonostial Tin cells have been termed the “working myocardium”. Apart from their different cell morphologies, the ostial versus working myocardial cells also express distinct sets of differentiation markers, which likely contribute to their different physiological functions in the dorsal vessel. For example, the Svp cells in the heart portion express wingless, whereas the Tin cells in the heart specifically express the Na-dependent bicarbonate anion exchanger 1 (NDAE1) (Perrin et al., 2004) and the two-pore domain potassium channel ORK1 (Perrin et al., 2004; Lalevee et al., 2006). Likewise, only the Tin cells of the working myocardium throughout the dorsal vessel express the KATP channel subunit Sulfonylurea receptor (SUR; Seino, 1999) as a result of the direct activation of the SUR gene by tinman via a Tin-dependent enhancer element (Nguyen and Xu, 1998; Nasonkin et al., 1999; Lo and Frasch, 2001; Akasaka et al., 2006; Zaffran et al., 2006; Hendren et al., 2007). Genetic and pharmacological approaches have confirmed that the expression and function of these channels is essential for normal heart function and performance (see Section X.A) (Akasaka et al., 2006). Perhaps unexpectedly, even differentiation genes that are expressed equally in the Svp/Doc and Tin cells are regulated via different enhancer elements within the two subtypes of cells. For example, a cardioblast enhancer element of the gene for the transmembrane receptor Toll includes a binding site for Tin that is essential for its expression in Tin cells, whereas at least one of three Doc-binding sequences that are also present is required for its expression in the Spv/Doc cells (Wang et al., 2005). A similar situation has been demonstrated for the myocyte enhancer 2 (mef2) gene, which is controlled by regulatory sequences in the Tin cells that are directly targeted by Tin, as well as by enhancer sequences that are needed for Svp cells and may be targeted by Svp or Doc (Gajewski et al., 1997; Nguyen and Xu, 1998) (see Section VII.C). The convergence of differentially-expressed cardiogenic factors on the same downstream genes in distinct myocardial populations may be a general phenomenon in cardiac differentiation. Due to the major functional consequences of the mutually-exclusive pattern of Svp/Doc and Tinman expression in myocardial cells, an important question is how these patterns are established and maintained during

PART  |  1  Heart Evolution

cardiac development. Recent studies have partially clarified the genetic circuits involved in this diversification event. Specifically, it has been shown that, in principle, all myocardial cells express tinman early on; however, in the Svp cardioblasts tinman is repressed by Svp, whereas it persists in svp mutants. Likewise, Doc is repressed in the Tin cardioblasts by Tinman, which restricts Doc expression to the Svp cells. This has been shown with a genetic combination, in which tinman function was knocked-out specifically in the forming dorsal vessel without affecting its earlier dorsal mesodermal expression, which caused expansion of Doc expression into all the Svp myocardial cells (Zaffran et al., 2006). In turn, the expanded expression of Doc causes at least a partial cell fate switch from working myocardium to ostial cells as indicated, for example, by the expanded expression of the ostial marker Wg. Conversely, Doc can also repress tinman, which is presumably an additional mechanism to stabilize the mutually-exclusive expression of tinman and Doc in myocardial cells, analogous to the mutual repression earlier between eve- and lbe-expressing clusters (see Section V.C).

VII.A.ii.  Tbx20 Genes (mid/nmr2 and   H15/nmr1) in Myocardial Diversification and Differentiation Based on the observations of the regulation of tinman and Doc within cardioblasts via mutual repression and Svp-mediated repression of tinman, the question arises: which are the positive regulators that initiate and maintain the expression of these genes in cardioblasts during development? For Doc, the answer is not known, although it is likely that its expression in cardioblasts is initially established by the activity of tinman, pannier and perhaps Doc itself. However, for tinman it was determined that its expression in cardioblasts is (re)established and maintained by the tbx20 orthologs mid/nmr2 and H15/ nmr1 (see also Chapter 9.4.II.F). Expression of mid/nmr2 depends on tinman, Doc and pannier, and initiates during cardiac progenitor specification. Later, the expression of its paralog, H15/nmr1 is also initiated in differentiating cardioblasts, and both genes continue to be expressed in all myocardial cells of the dorsal vessel (Figs 12; 13D) (Miskolczi-McCallum et al., 2005; Qian et al., 2005a; Reim et al., 2005). In mutants for mid/nmr2 or both paralogs, the cardioblasts are formed but fail to express tinman. Hence, mid/nmr2 mutants have a similar phenotype as the mutants in which tinman is knocked out specifically within the dorsal vessel (Zaffran et al., 2006), which consists of the ectopic expression of Doc in Svp cardioblasts and the concomitant expansion of ostial identities (Reim et al., 2005). Interestingly, the nmr genes are also expressed in the adult heart and interact strongly with tinman in establishing and maintaining adult heart function (Qian et al., 2008).

Chapter  |  1.2  Development and Aging of the Drosophila Heart

67

Figure 12  Epithelial polarity features of the myocardium. Myocardial cell nuclei are in red and cell polarity markers in green. Diagrams on the left: Before closure, Discs large protein (green) is distributed at basal and lateral sides of myocardial cells. When the bilateral myocardial rows come into contact, Discs large is redistributed to the apical(dorsal)–lateral sides of the myocardial cells. Top right: Low magnification picture showing the whole heart stained with nuclear Dmef2 (red) and secreted Slit along the heart lumen (green). Middle and bottom right: Higher magnification picture of two segments of the heart with nuclear stain in red and apical-basal Dystroglycan and apical–lateral Discs large in green, respectively.

VII.A.iii.  ladybird Genes and Myocardial Cell Diversification The expression of the ladybird genes (lbe and lbl) represents another interesting example for the segmental organization of the dorsal vessel. In late stage embryos, lbe expression is observed in two of the four tinman-expressing cardioblasts in each hemisegment, and in addition in two pericardial cells (Fig. 9A) (Jagla et al., 1997a). Very low levels of expression are also detected in the adjacent posterior pair of Svp/Doc cardioblasts. The Lb pericardial cells are distinct from the EPCs. The expression of lbe is established at early stage 12, when it is activated in segmental clusters of about four cells each in the segregating heart anlagen. These cells are likely to correspond to the two myocardial and two pericardial cells per hemisement observed later. At the same stage, evenskipped expression is observed in clusters of three adjacent cells each, corresponding to two pericardial cells and one somatic muscle founder. This cluster is located posteriorly adjacent to the lbe clusters (Fig. 8A). During the subsequent processes of heart segregation, cell rearrangements which involve a 90° clockwise rotation of heart precursors within each segment move the lbe expressing cells to the dorsal side and the eve-expressing cells ventrally to them. Some of the regulatory processes involved in the establishment of these expression patterns are discussed in Section V.C. The analysis of the role of lbe in cardiac patterning and differentiation has been hampered by the lack of specific

loss-of-function mutants, and the fact that ectodermal lbe is upstream of late wg expression, which could cause indirect effects on cardiac development (Jagla et al., 1997b). However, the results from forced ectopic expression of lbe and of lbe RNAi constructs indicate that lbe does have roles in specifying distinct cell fates within the cardiac mesoderm (Jagla et al., 1997a; Junion et al., 2007). Possibly, these include positive effects on the regulation of tinman and Dorsocross in the respective cardioblast subpopulations, although this supposition needs to be evaluated in more detail (Junion et al., 2007). In addition, it appears that lbe can also affect, directly or indirectly, the expression of numerous terminal differentiation genes that are needed in all myocardial cells. Orthologs of lbe have been identified in a number of vertebrate species, but none have been reported to be expressed in embryonic heart tissues (Jagla et al., 1995).

VII.B.  Axial Patterning and Subdivision of   the Dorsal Vessel VII.B.i.  The Role of Hox Genes in the Axial Patterning of the Dorsal Vessel Morphologically, the most obvious subdivision of the cardiac tube along the anterior–posterior axis is the subdivision into an anterior portion, termed aorta, and a posterior portion, termed heart. The aorta is characterized by a narrower bore,

68

a lower contractility and, in embryonic and larval stages, an absence of ostia. Conversely, the heart displays a wider chamber diameter, higher contractility and three bilateral pairs of ostia that serve as inflow valves. The cardioblasts of the heart are also distinguished by their specific expression of various differentiation markers, including the Na-driven anion exchanger Ndae I, the two-pore domain potassium channel ORK1, the hyperpolarization-activated cyclic nucleotide-gated (HCN) channel Ih and the troponic-C-like protein Tina-1 in the working myocardial cells and Wingless in the ostial cells (Lovato et al., 2002; Lo and Frasch, 2003; Perrin et al., 2004; Monier et al., 2005; Lalevee et al., 2006). The aorta lies underneath the region from the end of the second thoracic to the fourth abdominal segments of the ectoderm (T2/T3 to A4) and the heart extends from segments within A5 to A8 (Figs 1; 10; 13). Within the aorta, the anterior portion can be clearly distinguished from the posterior portion by using molecular markers. Specifically, each of the three segments of the anterior aorta contains only four bilateral pairs of cardioblasts (two cells per hemisegment being Tin and the other two Tin Lb) (Fig. 10C), which additionally express the Meis family homeodomain factor Homothorax (Hth) (Lo and Frasch, 2003; Perrin et al., 2004). By contrast, the three segments of the posterior aorta contain the standard pattern of four pairs of cardioblasts as in the heart (two pairs Tin, two pairs Tin Lb, two pairs Svp Doc) and lack Hth (Fig. 10C). Likewise, the anterior aorta is flanked by the cells of the lymph gland, which express the trans­cription factor Collier (Crozatier et al., 2004), whereas the remaining cardiac tube (in A2 to A8) is flanked by pericardial cells that express the type IV collagen-like protein Pericardin (Chartier et al., 2002; Perrin et al., 2004). Several studies have described the role of homeotic genes (Hox genes) as key regulators of these anterior– posterior (AP) subdivisions of the dorsal vessel. These studies found that Hox genes have important functions during both early and late stages of cardiogenesis in determining the regional anterior–posterior polarity of the cardiac tube. An early role of Antennapedia (Antp), Ultrabithorax (Ubx), abdominal-A (abd-A), which presumably is exerted during cardioblast specification, is to provide the cardioblasts with posterior aorta/heart identities, and likewise, the flanking pericardial cells with their particular identity that includes pericardin expression (Perrin et al., 2004; Ryan et al., 2005). Each of these Hox genes is individually sufficient to promote this developmental pathway. Hence, the development of the cardioblasts and pericardial cells that are located anteriorly to the anterior borders of expression of Antp, abd-A, and Abd-B can be considered as the default pathway in the absence of Hox gene activity (in Ubx;abd-A double mutants more posterior cardioblasts also express hth). The default pathway gives rise to the particular types of cardioblasts and the lymph gland cells of the anterior aorta. Apparently, a major function of these Hox genes during cardioblast specification is to generate, together

PART  |  1  Heart Evolution

with segmental regulators such as Hedgehog, the progenitors of the Svp cardioblasts and their sibling pericardial cells in the region of the future posterior aorta and heart within the cardiogenic mesoderm. In addition, the presence of these Hox genes affects lineage decisions of the Tin cells, as the progenitors of the Tin cardioblasts within the region of the future posterior aorta/heart divide symmetrically, whereas those in the region of the future anterior aorta divide asymmetrically into one cardioblast and one pericardial cell (Alvarez et al., 2003; Ryan et al., 2005). In summary, the early function of Hox genes serves to provide a fundamental subdivision of the dorsal vessel into an anterior aorta portion and a posterior aorta/heart portion (Perrin et al., 2004). Of note, during metamorphosis the anterior aorta develops into the adult aorta, whereas the posterior aorta is remodeled into the adult heart (see Section IX). By contrast, Abdominal-B (Abd-B), which during early stages is expressed posteriorly to the heartforming region, appears to have a suppressive function in the formation of cardiac progenitors during early cardiogenesis (Lo et al., 2002; Lovato et al., 2002). The late functions of the Hox genes are reflected in their spatial patterns of expression along the anterior– posterior axis within the cardiac tube. In particular, Antp is expressed in four pairs of cardioblasts at the border between the anterior and posterior aorta, Ubx expression marks the remainder of the posterior aorta, and abd-A is expressed in the heart proper, except for the two posterior-most cardioblast pairs, which express Abd-B (Fig. 10) (Lo et al., 2002; Lovato et al., 2002; Ponzielli et al., 2002; Lo and Frasch, 2003; Perrin et al., 2004). The sharp border of abd-A expression that coincides with the border between the aorta and the heart suggested a major role of abd-A in specifying heart identities. This notion was confirmed genetically, as in abd-A mutants the heart is transformed into aorta, whereas ectopic expression of abd-A in the entire cardiac tube leads to an opposite transformation, in which the aorta acquires heart-like properties (Lo et al., 2002; Lovato et al., 2002; Ponzielli et al., 2002). Although in ectopic expression experiments Antp and Ubx are also able to activate certain heart-specific markers in the posterior aorta (Perrin et al., 2004), the specific expression of abd-A and the observed effects with loss-of-function of abd-A provide strong evidence that in normal development abd-A provides the key activity in promoting heart development in the posterior dorsal vessel. The specific functions of abd-A include the activation of heart-specific differentiation markers such as ndae1, which is only observed in the Tin Abd-A cardioblasts, and wg, which is only observed in Svp Doc AbdA ostial cardioblasts. Presumably, abd-A acts together with tinman and Dorsocross, respectively, to promote the expression of specific differentiation markers in these cells and thus to provide them with their heart-specific morphological and functional features (Perrin et al., 2004; Reim and Frasch, 2005).

Chapter  |  1.2  Development and Aging of the Drosophila Heart

VII.B.ii.  A Separate Origin of the Most Anterior Cells of the Cardiac Tube from Head Mesoderm: Neural Crest-Like Contributions to the Dorsal Vessel? In addition to the trunk mesoderm, tinman is also expressed in a broad domain within the presumptive head region of the early embryo, which is regulated by a distinct enhancer ­ element with yet uncharacterized inputs (Yin et al., 1997). During stage 10, i.e., significantly later than the time of invagination of the trunk mesoderm, cells from small bilateral areas within this domain close to the stomodeum, delaminate, migrate dorsally, and form two rows of tinman-expressing cells at the dorsal roof of the esophagus (de Velasco et al., 2006). Afterwards the number of these tinman-positive cells is reduced, presumably by apoptosis. The remaining 15 cells, termed cephalic vascular cells, join the anterior aorta derived from the trunk mesoderm at its anterior tip. This structure, which is dependent on tinman activity and contains some Mef cells and a larger proportion of Mef2 cells, is also found in other arthropods, where it is often more prominent and is called the head aorta (de Velasco et al., 2006; Janssen and Damen, 2008). The Tin -Mef2 cephalic vascular cells that attach to the anterior aorta are connected to specific somatic muscles derived from the procephalic mesoderm, called cephalic outflow muscles (COM) and to so-called heart anchoring cells (HANC) derived from the dorsal epidermis. In addition, the HANC cells contact the most anterior Lb cardioblasts of the aorta. Together, these cells serve to anchor the anterior end of the dorsal vessel and form a funnel-shaped outflow tract. The assembly of this structure requires the Slit/Robo pathway (Zikova et al., 2003; Zmojdzian et al., 2008). Due to its mixed origin from the cephalic mesoderm and ectoderm, it has been speculated that this structure may be equivalent to the vertebrate outflow tract, which is formed from head mesoderm and migrating neural crest cells (Zmojdzian et al., 2008; Chapters 1.1 and 7.2).

VII.C.  Differentiation of the Dorsal Vessel VII.C.i.  The bHLH Domain Encoding   Gene Hand Drosophila has a single ortholog of the vertebrate Hand genes (Srivastava, 1999; see Chapter 10.3), which is expressed in the precursors of both the myocardial and pericardial cells starting from mid-cardiogenesis, as well as in the lymph glands and the midgut visceral mesoderm (Kölsch and Paululat, 2002). The expression continues in these tissues until the adult stages. The activity of tinman, Dorsocross and pannier is essential for the initiation of Hand expression in all dorsal vessel precursors, whereas tailup appears to be needed for its expression in pericardial cells and the GATA gene serpent (srp) for expression in the prohemocytes of the lymph

69

gland (Han and Olson, 2005; Reim and Frasch, 2005; Sellin et al., 2006; Tao et al., 2007). An intronic enhancer of Hand, which recapitulates endogenous dorsal vessel and lymph gland expression, features four conserved Tinman binding sites and five conserved GATA sites. Whereas the Tinman sites are needed for full enhancer activity in the Tin cardioblasts, the GATA sites are required specifically for enhancer activity in the Svp Doc cardioblasts and the lymph glands (Han et al., 2006). As simultaneous mutation of all Tinman and GATA sites abolishes enhancer activity in the dorsal vessel and lymph gland altogether, it appears that Tinman and Pannier have partially redundant roles in the establishment of cardiac Hand expression, except for the Svp Doc cells, in which Tinman expression is repressed and Pannier is essential on its own (see Section VII.A.i). The analysis of Hand-null mutants indicates that Hand is required neither for the normal specification of myocardial and pericardial cells, nor for normal morphogenesis and differentiation of the dorsal vessel during embryogenesis (Lo et al., 2007). However, the reduced lymph glands in Hand mutants reveal a role in the specification or maintenance of prohemocytes (Han et al., 2006; Lo et al., 2007). Unlike the embryonic and larval dorsal vessels, the dorsal vessels of adult mutant flies lacking Hand feature clear structural abnormalities (Lo et al., 2007). Normally the myofibrils are arranged in a highly-ordered semicircular fashion, extending from the dorsal to the ventral midline of the tube along the periphery of each myocardial cell and thus forming a spiral pattern around the dorsal vessel (Figs 1; 13A). By contrast, in Hand mutant flies the myofibrils are severely disarranged and often appear to surround the entire cortex of each myocardial cell, which reflects a major disruption of the cellular ultrastructure. In addition, and probably as a consequence, the diastolic and systolic widths of the heart are significantly reduced (Lo et al., 2007). The strongly reduced lifespan of Hand mutant flies may be a combined result of these heart defects, and of the observed severe gut defects that are related to the expression of Hand in the midgut visceral mesoderm (Lo et al., 2007).

VII.C.ii.  The MADS-Box Gene Mef2 Myocyte-specific enhancer-binding factor 2 in Drosophila (Mef2) is a member of the MADS-box family of transcription factor genes. MEF2 proteins are distinguished from other MADS-box-containing genes by an adjacent 29-amino-acid sequence called the MEF2 domain. The expression of the vertebrate mef2 genes correlates with the development of skeletal, cardiac and smooth muscle cells (Edmondson et al., 1994; Chapter 9.5) and suggests a role for this gene in all muscle lineages (for reviews see Potthoff and Olson, 2007). In contrast to Mef2, the myogenic factors of the MyoD class are not expressed in the cardiac muscle progenitors and thus do not play a role in heart muscle development (for review see Rudnicki and Jaenisch, 1995).

70

Like tinman, the single Drosophila ortholog of mef2 is expressed first ubiquitously in the early mesoderm, under the control of twist, and then undergoes dynamic changes in its expression (Lilly et al., 1994; Nguyen et al., 1994; Taylor et al., 1995). Drosophila mef2 expression finally persists in all progenitor cells of contractile muscle cells, and thus also in the myocardial cells of the heart (Fig. 12). In mef2 mutants, the formation of cardiac, skeletal and visceral muscle progenitors per se does not seem to be affected, suggesting that mef2 is not required for the initial specification of the major mesodermal subtypes. Rather, the major requirement for mef2 appears to be in muscle differentiation and muscle-specific gene expression (Bour et al., 1995; Lilly et al., 1995; Ranganayakulu et al., 1995; Lin et al., 1996; Lin and Storti, 1997; Stronach et al., 1999). Most strikingly, myocytes of the somatic musculature do not fuse into poly-nucleated muscle fiber syncytia, although the expression of identity genes in the founder cells of individual muscles is not affected. In contrast, some of the muscle structural genes required for muscle function, including Drosophila muscle myosin heavy chain (DMM) and Tropomyosin I (TmI), are not expressed in mef2 mutants. The requirement of mef2 for muscle-specific gene expression in the heart suggests that mef2 is a crucial differentiation factor not only for the body wall muscles, but also for the heart of Drosophila. For some known mef2 downstream genes such as actin57B and 3-tubulin (Tub60D) it has been demonstrated via enhancer analysis that they are direct target genes of Mef2 in the dorsal vessel (Damm et al., 1998; Kelly et al., 2002). A large number of DNA regions that are bound by Mef2 in vivo at different stages of development have been identified by ChIP-on-Chip approaches. These include both known mef2 downstream genes and yet unknown candidate targets, some of which are likely to require mef2 activity for their expression in myocardial cells (Junion et al., 2005; Sandmann et al., 2006). The analysis of regulatory regions at the mef2 locus revealed a highly modular arrangement of enhancer elements within 12 kb upstream of the start site of the major transcript (which corresponds to the first intron if an alternative upstream mef2 transcription start site is used). These include an enhancer that is active in the early mesoderm and is a direct target of Twist, another that is active in the dorsal mesoderm after the spreading of the mesoderm that is a direct target of Dpp signaling, several enhancers that are active in the developing somatic musculature at different stages, and two separate enhancers that are active in the dorsal vessel. Expression in the circular or longitudinal gut muscles is also driven by several of these elements (Cripps et al., 1998, 1999; Gajewski et al., 1998; Nguyen and Xu, 1998). These results demonstrate that the ubiquitous and continuous expression of mef2 in the mesoderm is actually the result of the additive and consecutive function of a complex set of enhancer elements that receive diverse inputs from various upstream regulators. Of note, one of the two myocardial enhancers is

PART  |  1  Heart Evolution

known to contain two Tinman binding sites, each of which is required for driving expression in the dorsal vessel (Gajewski et al., 1997; Cripps et al., 1999). Accordingly, this enhancer is only active in the Tin “working myocardial” cells. In addition, this enhancer includes a GATA site, presumably being targeted by Pannier, which is required for enhancer activation in myocardial cells and for repression in pericardial cells (Gajewski et al., 1998). In summary, tinman and pannier appear to be direct upstream regulators during the establishment of myocardial mef2 expression, and tinman is additionally required as a direct regulator of the maintained myocardial expression specifically within the Tin cardioblasts. Whether Dorsocross or any of the other cardiogenic transcription factors are also required for the initiation of myocardial mef2 expression and for its maintenance in the Svp Doc set of cardioblasts remains to be shown.

VII.C.iii.  The Chromatin-Remodeling Factor Myocardin Drosophila myocardin-related transcription factor (MRTF) is a member of the SAP family of chromatin-remodeling proteins (Han et al., 2004). Vertebrate myocardin acts as a transcriptional co-factor for serum response factor (SRF), and regulates smooth and cardiac muscle genes as well as its own expression (Wang et al., 2001; Chapter 9.3.XI). Notably, for its own expression myocardin appears to require Mef2 as a co-factor (Creemers et al., 2006). Loss-of-function mutants for Mrtf generated by homologous recombination in flies do not exhibit obvious developmental cardiac defects, which is possibly due to maternally-provided compensation. In contrast, overexpression of a dominant-negative form of MRTF in the mesoderm impairs spreading and dorsal migration of the early mesoderm, which is reminiscent of heartless mutants. This causes abnormal cardiac morphogenesis with gaps and misaligned myocardial cells (see also Section III.B). Similar to vertebrates, MRTF is likely a transcriptional co-activator of Drosophila SRF (Han et al., 2004). These data suggest that myocardin and SRF are another example of functionally-conserved genes between flies and vertebrates (see Chapter 9.3).

VII.C.iv.  The Role of MicroRNAs in Cardiac Differentiation MicroRNAs (miRs) are a recently discovered class of genomically-encoded, small nonprotein-coding RNAs (20-nucleotide) that post-transcriptionally target other mRNA transcripts for degradation or translational inhibition by hybridizing to their UTRs (Ambros, 2001, 2003; Bartel, 2004). Many hundred miRNAs affecting diverse biological processes have been identified in both vertebrates and invertebrates, including Drosophila, and elucidation of their functions in cardiac development, function and response to stress are under intense investigation (Latronico et al., 2007;

Chapter  |  1.2  Development and Aging of the Drosophila Heart

van Rooij et al., 2007; Chapter 10.3). Studies of miRNAs in Drosophila heart development began recently with work on miR-1 (Kwon et al., 2005; Sokol and Ambros, 2005), the ortholog of vertebrate miR-1-1 and miR-1-2, which are critical for mammalian cardiogenesis downstream of SRF, for example by targeting Hand2 mRNA (Zhao et al., 2005). Drosophila miR-1 is expressed in the early mesoderm, as well as later in most if not all muscles, including the heart, under the control of Twist and Mef2 transcription factors (Sokol and Ambros, 2005). As in mice, Drosophila miR-1 expression also depends on SRF consensus binding sites for expression in the heart-forming region. Genetic evidence indicates that miR-1 contributes to the maintained expression of muscle genes. During cardiogenesis, miR-1 has been reported to target the Notch ligand Delta mRNA for translational inhibition, which may normally restrict the number and differentiation of cardiac progenitors (Kwon et al., 2005). Stereotyped asymmetrical cell division via differential Notch signaling is a well-established mechanism for specifying alternative cardiac cell fates in Drosophila (see Section VI). In a small portion of the miR-1 mutant embryos, cardiac tinman expression is expanded, which is accompanied by supernumerary myocardial and Eve pericardial cells (Kwon et al., 2005). Whether this reflects reduced Notch signaling and participation in cardiac lineage decisions will need to be substantiated in future studies.

VIII.  Assembly of the cardiac tube VIII.A.  Myocardial Polarity As in vertebrates, the bilateral primordia of cardiac mesoderm of Drosophila assemble at the midline in a concerted and well-orchestrated fashion (Fig. 3; Chapter 3.1). This process begins after dorsal migration of the mesoderm (Fig. 2F), at the time when the cardiac lineages and specification of individual cell types have been essentially completed. The myocardial cells on either side of the embryo then line up in a perfectly aligned double row (Figs 1; 10; 12), flanked by more irregular rows of pericardial cells. The myocardial cells on either side of the midline establish close contact with their contralateral counterparts, and seem to undergo a mesenchyme-to-epithelium transition, resulting in an apical-basallike polarization of the myocardial cells. The bilateral aligned rows of myocardial cells fuse at the dorsal midline to form the heart tube enclosing a lumen (Zaffran et al., 1995; Fremion et al., 1999; Haag et al., 1999; Chartier et al., 2002; Qian et al., 2005a,b). The epithelial polarity of these lumen-forming myocardial cells is unusual in that both basal and apical markers (Dystroglycan and Discs large, respectively) are localized at the site of apposition (Fig. 12) (Qian et al., 2005b). Recent confocal sectioning and reconstruction during the process of lumen formation revealed that basal markers

71

become localized at the interior, luminal surface, whereas apical markers are restricted to the apposition or contact points between two contralateral myocardial cells that resemble two half-pipes forming a tube (Medioni et al., 2008). When the bilateral myocardial rows are near each other, they adopt epithelial cell shapes and display characteristic cell polarity features (Fig. 12). How this process is controlled and coordinated, and how correct cell–cell contacts in the emerging heart tube are formed is largely unexplored. It is also not known what the causal relationship is between cardiac cell polarity, patterning or cellular alignments of the forming heart tube. For future elucidation of cardiac morphogenesis more cardiac cell polarity markers will have to be characterized. Some recent studies have begun to shed light on this process.

VIII.A.i.  Neuromancer (H15/mid) As discussed in Section VII.A.ii, mid/nmr2 is expressed in all cardioblasts, where it seems to be required for continued tinman expression. In addition, it was reported that the nmr locus is critical for the assembly and morphogenesis of the heart tube (Miskolczi-McCallum et al., 2005; Qian et al., 2005a), although another study did not discern any significant heart tube morphogenesis defects in nmr loss-offunction mutants (Reim et al., 2005). Normally, the myocardial cells align in a highly-ordered fashion that includes stereotyped cellular localization of cell polarity markers (Fig. 12) (Qian et al., 2005a). H15/nmr1 deletion mutants combined with pan-mesodermal knockdown of mid/nmr2 (or when both genes are deleted in a larger deficiency; Miskolczi-McCallum et al., 2005), appeared to cause defects in myocardial alignment and normal cell polarity, as detected with antibodies against Discs Large, Armadillo, Spectrin and Dystroglycan (Qian et al., 2005a). Likewise, when mid/nmr2 was knocked down by RNAi specifically in the dorsal vessel, polarity markers such as Discs large and Dystroglycan were mislocalized (Qian et al., 2005a). Based on these data, it seems that nmr is also required for establishing and maintaining myocardial cell polarization, in addition to maintaining tinman expression. However, in view of the possibility of RNAi off-target effects (Kulkarni et al., 2006) and the discrepancies in observed heart morphogenesis defects by different groups, a definitive elucidation of the role of the Drosophila tbx20 genes in heart morphogenesis must await further experimental evidence.

VIII.A.ii.  Slit/Robo A large number of extracellular, transmembrane and intracellular membrane-associated proteins have been found to contribute to cardiac morphogenesis. These include cell adhesion and extracellular matrix molecules (Faint Sausage, Laminin A, E-cadherin, Pericardin, Tincar, Dystroglycan) (Yarnitzky and Volk, 1995; Haag et al., 1999; Chartier et al.,

72

2002; Hirota et al., 2002; Deng et al., 2003); guidance molecules (Slit, Robo, Robo2) (Qian et al., 2005b; MacMullin and Jacobs, 2006) and cell polarity genes (Go47A protein a.k.a. Brokenheart, PDZ pro­tein Discs large) (Fig. 12), Spectrin, Armadillo/-catenin, Toll, and others (Fremion, 1999; Chartier et al., 2002; Qian et al., 2005b; Wang et al., 2005). Of them, slit and the robo genes have been studied in some detail (Qian et al., 2005b; MacMullin and Jacobs, 2006; Santiago-Martínez et al., 2006, 2008; Medioni et al., 2008). The secreted leucine-rich repeat protein Slit is expressed before the myocardioblasts meet each other at the dorsal midline, but subsequently accumulates between each bilateral pair where the lumen of the heart is forming (Fig. 12). Similarly, the transmembrane Slit receptor Robo also localizes to the lumen-facing side of the forming myocardium, where it presumably interacts with Slit. The Robo2 receptor is prominently expressed in pericardial cells, and perhaps also in the myocardium (Santiago-Martínez et al., 2006). Interestingly, in robo mutants, robo2 is upregulated in the myocardium, and presumably compensates for the lack of robo function (see below) (Qian et al., 2005b). In slit mutants, or in robo;robo2 double mutants, dorsal migration and heart cell identity specification proceeds normally. Before the bilateral primordia meet at the midline, expression of polarity markers normally occurs as well. However, as the bilateral primordia meet at the midline, cardiac morphogenesis is severely perturbed, leaving gaps and misaligned myocardial cells that have largely lost their stereotyped polarity features (Qian et al., 2005b). Single mutants of robo exhibit a much milder heart phenotype because of partial redundancy provided by robo2, which is corroborated by a variety of rescue experiments. In addition, slit interacts genetically with many of the above-mentioned genes affecting cardiac morphogenesis, but notably not with crumbs, which is not surprising, because Crumbs does not seem to localize in a polarized fashion in the myocardium (Qian et al., 2005b; MacMullin and Jacobs, 2006). As shown in Fig. 12, the myocardial cells are polarized differently before and after heart tube assembly at the midline, which is also reflected in the localization of Slit and Robo protein. Interestingly, robo;robo2 or slit mutant hearts are correctly polarized initially but fail to maintain their proper polarization after the bilateral rows of heart cells reach the dorsal midline (Qian et al., 2005b). In this context, Medioni et al. (2008) and Santiago-Martínez et al. (2008) have proposed similar models, in which Slit/Robo signaling allows lumen formation by preventing the formation of adherence junctions within the prospective luminal surface of cardioblasts. Recently, the Rho GTPase Cdc42 was found to be required for the stereotyped alignment of the myocardioblasts, suggesting that it may be potentially involved in transducing cell polarity information to the cytoskeleton and other effectors, in order to implement morphogenetic changes during dorsal vessel formation (J. Liu, G. Vogler, L. Qian and R. B., unpubl.).

PART  |  1  Heart Evolution

VIII.B.  The Mevalonate Pathway The rows of pericardial cells on either side of the embryo are normally closely associated with the heart tube (Fig. 10). Studies of the genetic basis of this association have recently begun. A P-element-based screen using Hand-GFP expression in the myocardial, pericardial and lymph gland cells (Han and Olson, 2005) revealed mutants that cause malformations of the heart tube late in embryogenesis. One of the phenotypic classes of mutants featured the dissociation of the pericardial cell rows from the myocardium (Yi et al., 2006). Five genetic loci, which gave this “broken hearted” phenotype, turned out to be members of the metabolic mevalonate pathway, including HMG-CoA reductase and downstream enzymes, which is necessary for the isoprene derivatives and the farnesylation and geranylgeranylation of proteins. One of the proteins which is geranylgeranylated is G-protein G1, a protein required autonomously within the myocardium for proper association between myocardial and pericardial cells during cardiac morphogenesis (Yi et al., 2006). Since the mevalonate pathway seems to act within the heart tube to keep intimate pericardial cell apposition, it will be interesting to find out the basis for this cell–cell interaction. Additional insights into this pathway have been provided by Yi et al. (2008), who showed that various septate junction proteins act in conjunction with heteromeric G-proteins to ensure proper adhesion of pericardial cells to the cardiac tube. Of note, the absence of septate junctions in the embryonic cardiac tube points to a novel, noncanonical function of these proteins. Given the evolutionary conservation of heart development, it seems likely that components of this pathway may be involved in congenital heart disease. The emerging studies in cardiac morphogenesis suggest the involvement of a complex network of interactions between extracellular matrix, signal transduction, epithelial polarization, cytoskeletal components and adhesion molecules. Understanding the orchestration and regulation of these factors and their roles in cell–cell communication will eventually clarify how the highly-specialized structure (and function) of this organ is accomplished.

IX.  Remodeling of the larval to the adult dorsal vessel At the end of embryogenesis the heart begins to beat in slow irregular contractions. After hatching, the cardiac tube increases in size during the three larval instar stages and develops a rhythmical myogenic heart beat of 2–4 Hz. In the fly’s open circulatory system, the hemolymph enters the dorsal vessel through the ostia, the external valves in the heart region (Fig. 13). On contraction, the ostia close and the internal valve between heart and aorta portions of the dorsal vessel opens to allow posterior-to-anterior directed flow of hemolymph, which then recirculates back

73

Chapter  |  1.2  Development and Aging of the Drosophila Heart

(A)

Heart

Posterior aorta

(B)

Larva

Adult

Terminal chamber

Heart Ubx

abd-A

(D)

Adult

Embryo & Larva

(C)

Figure 13  Remodeling of the dorsal vessel. (A, B) Larval (A) and adult (B) cardiac tubes, stained with phalloidin for polymerized actin (F-actin), are shown at the same magnification. Scale bars: 50 m. (C) Diagram of late embryo/larval and adult cardiac tube. The svp, tin, Ubx and abdA expression patterns are indicated. Modified from Monier et al. (2005); reproduced with permission of the Company of Biologists; segmental register has been renumbered according to Lo and Frasch (2003). (D) Part of an adult segment of the heart stained with myofibrillar Z-line marker -actinin (red) and myocardial nuclei with H15/nmr1-lacZ (green).

through the body cavity along the gut and other internal organs. Each myocardial cell of the larval dorsal vessel contains a number of myofibrils that are arranged in an unusual spiral fashion (Figs 1A; 13A). This arrangement of the contractile units permits contraction of the heart in the transverse, as well as in the longitudinal, direction to generate maximal output. In the free crawling larva, the heart beat may stop or change its rhythm intermittently, depending on activity or various environmental stimuli or stresses. Interestingly, even a prolonged cardiac arrest due to genetic manipulations is not immediately lethal, and the larvae can survive for hours or longer (R. J. Wessells and R. B., unpubl.). It will be interesting to find out what vital functions robust heart performance is required for. One way the larval heart beat can be stopped is by overexpression of the two-pore potassium channel ORK1, which is normally involved in setting the basal heart rate (Lalevee et al., 2006). A detailed review of larval heart physiology is given in Bodmer et al. (2005). During metamorphosis, the larval heart undergoes extensive morphological and functional changes in a pro­ cess of myocyte reprograming or remodeling (Curtis et al., 1999; Molina and Cripps, 2001; Zikova et al., 2003; Monier et al., 2005; Zeitouni et al., 2007). The contractile myocardium of the adult heart originates from the larval posterior aorta in the abdominal segments A2–5, whereas the terminal adult chamber is derived from A6 segment of the anterior larval “heart” tube (Fig. 13A–C) (the literature presently does not agree on the exact segmental register of

the heart; here we follow Lo and Frasch, 2003). The posterior two segments of the larval heart undergo histolysis during metamorphosis. How the anterior aorta is restructured is not known, but it does extend into the adult head to allow exchange of nutrients and hormones between head and body (Fig. 14A) (Rizki, 1978). During this remodeling process the cardiomyocyte morphology is dramatically changed. The adult heart is significantly expanded, exhibiting a wider diameter than its larval counterpart, and it contains four main chambers separated by three internal valves in A3–5 (Figs 13C; 14A,B) (Monier et al., 2005; Ocorr et al., 2007a; Zeitouni et al., 2007). In addition, the larval svp-expressing myocardial cells of the posterior aorta (A2–5) develop into functional ostia in the adult for the inflow of hemolymph during the diastolic relaxation phase after contraction (Fig. 13C). Recently, two ostia not described previously in the posterior thorax (presumably derived from the anterior-most Svp cells) and a posterior opening in the termi­ nal segment have been discovered, which explains some of the hemodynamics of the adult heart, such as the periodic reversal of pumping from anterior to posterior (Dulcis and Levine, 2005; Monier et al., 2005; Wasserthal, 2007). The terminal chamber derives from the larval A6 heart region which, in contrast to the more anterior adult heart, becomes thinner (Fig. 13A–C). A striking new feature of the adult heart is the addition of a layer of longitudinal muscles, which associate with the cardiac tube along its ventral side (Molina and Cripps,

74

2001). These longitudinal muscles are syncytial and contain several nuclei, which express mef2 but not tinman. Their origin and exact functional contribution to the cardiac contractions are not yet known, but their formation apparently requires FGF signaling, since interference with the FGF receptor (Htl) function prevents the formation of these imaginal heart-associated muscles (Zeitouni et al., 2007). A further remodeling change of the heart concerns the myofibrillar content and their orientation within the cardiomyocytes. In the adult terminal chamber, for example, the myofibers are reoriented from spiral/transverse in the larvae to longitudinal in the adult (Monier et al., 2005). In contrast to larvae, the adult heart is innervated anteriorly along the conical chamber (segments A1–3) by glutaminergic nerve endings and posteriorly (A5 and 6) by peptidergic nerve endings (Dulcis and Levine, 2003, 2005). To visualize the remodeling process of the heart from larva to adult, Monier and colleagues (2005) conducted time-lapse video analysis of the heart during metamorphosis, which substantiated conclusions of an earlier study (Molina and Cripps, 2001), namely that the adult heart is primarily generated by remodeling of the larval aorta and histolysis of the posterior larval heart during pupal development. This happens apparently without cell proliferation of the tinman- or svp-expressing larval cardiomyocytes (Monier et al., 2005). The remodeled adult heart tube functions in an open circulatory system as in the larva and contracts rhythmically and efficiently (reviewed in Bodmer et al., 2005; Ocorr et al., 2007b), but with age the heart progressively declines in its performance (see Section X.C). The remodeling of the heart also includes transcriptional reprograming. The larval svp-expressing cells of the aorta begin expressing wg as they differentiate into ostia of the adult heart. Wg signaling seems to be required for cardiac metamorphosis, since inhibition of Wg signal transduction by expressing a dominant-negative form of dTCF inhibits ostia formation and remodeling of the larval aorta into adult heart (Zeitouni et al., 2007). Ih and Ndae also expand their expression domains into the remodeled adult Tin cardiomyocytes of the anterior abdomen (Monier et al., 2005). Ih is a hyperpolarization-activated cyclic nucleotide-gated (HCN) ion channel, whose vertebrate HCN homologs are thought to be involved in cardiac pacemaking (e.g., Robinson et al., 2006; DiFrancesco and Borer, 2007). Ndae is an Na-driven anion exchanger involved in ionic homeostasis. Since cardiac pacemaking initiates in the heart region of the larval as well as the adult dorsal vessel, it is conceivable that these ionic regulators have a critical role in the cardiac cycle, and especially in the case of Ih participate in setting or modulating the pace of the heart beat. Cardiac reprograming also involves participation of the Hox genes (Monier et al., 2005). In the embryo, abdA specifies the heart region of the dorsal vessel in segments A6–8 (Section VII.B.i). In the adult, AbdA activity is required autonomously in the A6 portion of the heart for its

PART  |  1  Heart Evolution

remodeling to form the terminal chamber with its unique morphological features (Fig. 13A–C). As expected, abdA does not play a role in the remodeling of the larval aorta to the adult heart. Rather, it is the repression of Ubx in the A2–5 Tin cardiomyocytes and its restriction to the Svp cardiomyocytes that is critical for remodeling of the posterior aorta into the adult heart, since maintained expression of Ubx (or abdA) in all cells during metamorphosis prevents acquisition of adult heart characteristics (Fig. 13) (Monier et al., 2005). Additional heart-autonomous regulators required for normal cardiac remodeling include the transcription factors Tinman and Hand (see Sections IV.A and VII.C) (Fig. 5F) (Zaffran et al., 2006; Lo et al., 2007). Cardiac remodeling is temporally coincident with the peak release of ecdysone. Thus, it was suspected that ecdysone signaling was driving cardiac remodeling during metamorphosis. Indeed, interference with ecdysone receptor activity within the cardiomyocytes inhibited metamorphosis of the heart (Monier et al., 2005). This suggests that the ecdysonemediated metamorphosis program controls cardiac remodeling in part by downregulating Ubx expression in the heart and by modifying AbdA activity in the terminal segment. In a recent genome-wide approach, changes in transcriptional patterns were identified within the heart during metamorphosis (Zeitouni et al., 2007). This was achieved by isolating RNA from dissected heart tubes at progressively later time points of cardiac remodeling. Microarrays then revealed the modulation of genes in a number of signaling pathways. Among them were the FGF, the Wg and the PDGF–VEGF signaling pathways. Subsequent analysis by Zeitouni et al. (2007) showed that FGF signaling is required for formation of the longitudinal ventral muscles that are associated with the adult heart (Molina and Cripps, 2001), and Wg signaling takes part in the cardiac remodeling and in adult ostia formation. pvr, which codes for a receptor tyrosine kinase related to mammalian PDGF and VEGF receptors, is expressed in the recently described internal valves of the adult heart that apparently partition the heart into four chambers along the anterior–posterior axis (Fig. 13C) (Monier et al., 2005). Although little is still known about the distinguishing characteristics of these valves, modulation of the activity of Pvr can alter their abundance or differentiation (Zeitouni et al., 2007). Thus, the PDGF–VEGF pathway in flies seems to participate in valve formation, as does the VEGF pathway in mammals. However, the vertebrate valves derive from the endocardium, which appears to be an invention of chordates and vertebrates and does not exist in flies (Fishman and Chien, 1997; Chapters 6.1 and 6.2).

X.  Controls of the physiology and aging of the adult heart Recent studies suggest that the Drosophila heart is not only an excellent model for a basic understanding of cardiovascular

Chapter  |  1.2  Development and Aging of the Drosophila Heart

development, but also for investigating the molecular controls of cardiac physiology and of the decline in heart function as the organism ages. Studying the genetics of heart function and aging in flies became possible on the development of assays that allow detailed and precise measurements of various aspects of heart function, such as heart rate, contractility, arrhythmias, ejection volume, electrical activity and response to pacing stress (Wessells and Bodmer, 2004; Ocorr et al., 2007a; Fink et al., 2009). The combination of these assays with the genetic tools available resulted in the development of a number of Drosophila models of heart disease, including dilated and restrictive cardiomyopathy and arrhythmias (see below) (Akasaka et al., 2006; Wolf et al., 2006; Allikian et al., 2007; Ocorr et al., 2007a; Cammarato et al., 2008; Frolov et al., 2008; Mery et al., 2008; Qian et al., 2008; Taghli-Lamallem et al., 2008). In these cardiac disease models, the progressive deterioration of the heart with age is usually aggravated, as is the case for their human counterparts. This makes the fly with its short lifespan an attractive model for studying the age-dependent changes in models of cardiac disease (Wessells et al., 2004, 2007; Ocorr et al., 2007c). Moreover, a recent study shows that rare alleles causing “heart disease,” such as arrhythmias and dilated cardiomyopathy, exist in wild populations of flies (Ocorr et al., 2007d). Examining the nature of these naturally occurring genetic variations may provide valuable insights into the population genetics of heart disease.

X.A.  Control of the Heart Rhythm Recent studies showed that despite its simplicity the fly heart functions remarkably similarly to a vertebrate heart, and is composed of cardiomyocytes with contractile and electrical properties that are also conserved, thus providing a powerful model for elucidating basic mechanisms of contractility and the generation and maintenance of a regular heart rhythm. Ion channel and transporter genes, first studied in the larval heart, have been found to alter heart rate, for example SERCA, the Sarco-endoplasmic reticulum Ca2-ATPase, a Ca2-channel encoded by cacophony and a Ca2-gated K channel encoded by slowpoke (Johnson et al., 1997; Ray and Dowse, 2005; Sanyal et al., 2006). “Cardiac arrhythmia” genes encoding potassium channels responsible for repolarizing the cardiac action potential have also been studied in larval, and more recently in adult, hearts (Fig. 14C): ether-a-gogo (eag) and seizure are related to HERG and generate the repolarizing IKr current; and the single Drosophila KCNQ gene also produces IKs-like currents in flies (Wen et al., 2005). Drosophila KNCQ is related to the well-known KCNQ1 arrhythmia gene in humans, which causes type 1 long QT syndrome (LQT1) and is associated with an increased risk for torsades des pointes (TdP) arrhythmias and sudden death (for review see Jentsch, 2000; Towbin, 2004; Sanguinetti and Tristani-Firouzi, 2006).

75

Mutants or RNAi knockdown in each of these genes have dramatic effects on heart function in flies, supporting the idea of functional parallels in basic cardiac physiology between flies and humans (Johnson et al., 1998; Bodmer et al., 2005; Ocorr et al., 2007a; K. Ocorr and R. B., unpubl.). An obvious parameter to monitor is the heart rate (Dowse et al., 1995; Johnson et al., 1998, 2002; Mispelon et al., 2003; Wessells et al., 2004; Ocorr et al., 2007a). Intracellular and extracellular electrical recordings have been achieved using suction or floating electrodes and multielectrode arrays, respectively (Papaefthmiou and Theophilidis, 2001; Lalevée et al., 2006; Ocorr et al., 2007a). Several methods for monitoring heart wall movements have been developed, for example, an edge tracing system (IonOptix) and a noncontact ultrasound-like method called optical coherence tomography (OCT), which led to the identification of cardiomyopathy phenotypes in adult flies (Wessells and Bodmer, 2004; Wessells et al., 2004; Wolf et al., 2006; Allikian et al., 2007; Frolow et al., 2008; Fink et al., 2009). An automated analysis of high-speed movie recordings from partially-dissected adult fly heart preparations made it possible to process a large body of image-based data, and led to the establishment of a number of cardiac disease models (Ocorr et al., 2007a,b; Cammarato et al., 2008; Mery et al., 2008; Qian et al., 2008; TaghliLamallem et al., 2008). The power of this method is illustrated in M-mode traces to gauge, for example, the severity of arrhythmias or dilated cardiomyopathy (Fig. 14D) (Ocorr et al., 2007a,d). In another assay aimed at examining cardiac performance under stress, the heart is paced to a higher rate by external electrical stimulation and subsequently assessed for heart failure, defined as cardiac arrest or uncoordinated fibrillation (Fig. 15) (Paternostro et al., 2001; Wessells and Bodmer, 2004; Wessells et al., 2004; Luong et al., 2006; Ocorr et al., 2007a) (see also Section X.C). Using these heart function assays, it was shown that, as in humans, Drosophila mutants for the KCNQ potassium channel (see above) cause arrhythmias and prolonged contractions (Fig. 14D), which are likely due to compromised repolarization of the cardiac action potential, as revealed by delayed relaxation or fibrillation of the myocardium, increases in the duration of phasic contractions, extracellular field potentials and systolic tension, as well as by elevated susceptibility to pacing stress in adult flies (Ocorr et al., 2007a). Because KCNQ1 is a major arrhythmia susceptibility locus in humans, these findings suggest that the Drosophila heart could be well-suited for studying the genetics of arrhythmias, which may be caused by mutations, as well as by aging (see Section X.C). Indeed, mutants or heart-specific knockdown of HERG-related fly genes also exhibit a significantly elevated frequency of arrhythmias (K. Ocorr and R. B., unpubl.), again reminiscent of HERG mutations, which also affect human heart function and lead to the LQT2. Manipulation of another potassium channel, ORK1 – a two-pore K2P type channel – dramatically alters the

76

PART  |  1  Heart Evolution

(A)

A1 CC

A6.7

h

Pc Os4

Os L

(B)

Leak K

Pacemaker

V-D Cation Ca2+/Na+

K Ca2+

Kr Ks Sei/HERG Eag KCNQ

Threshold

(C) 1-Week wild-type

5-Week wild-type

(D) 1-Week KCNQ

5-Week KCNQ

Figure 14  (A) Diagram and image of a semi-intact heart preparation where the abdominal heart is exposed by removing the head, ventral thorax, gut and fat bodies. The exposed heart can beat for hours in oxygenated, artificial hemolymph. The red rectangle indicates a one pixel-wide portion of the movie frame used to generate the four M-mode traces shown at the bottom. (B) Electrical recording of two action potentials with arrows illustrating the currents during the cardiac cycle. (C, D) M-mode traces prepared from high speed movies of dissected flies. A one pixel-wide region with both edges of the heart tube is defined in a single movie frame. The same regions are electronically cut from all of the frames in the movie and aligned horizontally to produce the trace. Arrhythmic heart beats are evident in old wild-type (C) and KCNQ mutant (D) flies. The incidence and severity of arrhythmias increases in wild-type and dramatically more in KCNQ mutants with age. For quantitative analysis, see Ocorr et al. (2007a).

frequency of the larval heartbeat. Whereas cardiac overexpression of an ORK1 transgene stops the heart from beating, downregulation of ORK1 speeds it up by shortening the slow ramp depolarization phase of the action

potential (Lalevée et al., 2006). This suggests that ORK1 is an important novel regulator of heart rate, perhaps also in humans. SUR, which is a class of genes involved in ATP-dependent potassium channel function and protection

77

Chapter  |  1.2  Development and Aging of the Drosophila Heart

(A)

(B)

Electrical pacing

(D)

6 Hz for 30s

0.7

Failure rate

0.6

Dilps

Amino acids

WT female

0.5

WT male

0.4

InR

PIP3

0.3 PTEN

IRS

0.2

TSC1 & TSC2

1

2

3

P110-P13K P60-P13K

6

7

(E)

PKB

TOR

5 4 Age (weeks)

0.8

Heart-specific overexpression of:

0.7

Growth

S6K

FOXO

(C)

Aging Metabolism

Failure rate

4E-BP

InR

0.6 0.5 0.4

foxo 0.3

PTEN

0.2 1

2

3

5 4 Age (weeks)

6

7

Figure 15  InR–TOR pathway is implicated in modulating aging, including that of the heart. (A) Diagram of possible interconnections between InR and TOR pathways. (B) Electrical pacing set-up. Flies are placed between tips of electrode jelly connected to electrodes. (C) GMH5 Gal4 driver specifically expresses GFP in the myocardium (adult myocardial segments A2–3 are shown). (D) Electrical pacing-induced risk of heart failure (heartbeat arrest or fibrillation) of wild-type flies shows a progressive age-dependent increase in both males and females. (E) Activating the insulin signaling pathway by expression of InR only in the heart (with the GMH5 driver) causes a premature elevation of pacing-induced risk of heart failure in young (1–2-week-old) flies. Cardiac-specific expression of antagonists of insulin signaling, such as Foxo and PTEN (see (A)), abolishes the age-dependent increased risk in heart failure, which results in old flies with well-performing hearts. For details see Wessells et al. (2004).

from ischemia reperfusion and other stresses (Hanley and Daut, 2005), has also been studied in the Drosophila heart. Cardiac knockdown of SUR makes the heart more susceptible to hypoxia and pacing-induced heart failure. Thus, SUR protects the heart from various stresses in flies as in humans (Akasaka et al., 2006; see also Section X.C). Both of these potassium channel functions are thus thought to protect the heart from excessive activity and stressinduced damage, probably by reducing the heart’s baseline excitability.

X.B.  Pericardial Influences on Heart Function: even-skipped In vertebrates, the epicardium is the outermost layer of the heart, an epithelial cell layer that starts migrating during the process of heart looping to cover the surface of the myocardium. The epicardium and derivatives are critical for proper heart development by contributing to the maturation of the myocardium (reviewed in Männer and RuizLozano, 2008; Chapters 5.1 and 5.2). Within the forming heart of Drosophila, even-skipped is specifically expressed in a subset of pericardial cells. These pericardial cells in the fly have recently been implicated in contributing to the functional properties of the heart (Fujioka et al., 2005),

reminiscent of the nonautonomous influence of the epicardium to the vertebrate heart (Merki et al., 2005; Zamora et al., 2007; Männer and Ruiz-Lozano, 2008). All regulatory influences necessary for this highlyrestricted pattern of mesodermal eve expression seem to be contained within a small and well-defined enhancer element 3 to the gene “eme” (Fujioka et al., 1999), and many binding sites in eme for pertinent cardiogenic transcription factors and signaling pathway effectors have been identified (Halfon et al., 2000; Knirr and Frasch, 2001; Han et al., 2002; Liu et al., 2008). In order to investigate the role of these cells in cardiac differentiation and their potential influence on cardiac function two approaches have been taken, one that is based on rescue of the evenull phenotype and the other that uses a repressor of eve expressed in the mesodermal Eve lineages. Transgenic flies were generated with a large genomic fragment containing the entire eve locus, except for a deletion of eme, and combined with eve-null mutants (evemeso) (Fujioka et al., 2005). evemeso flies are viable but lack the mesodermal Eve lineages, which seem to be transformed into lbe expressing cells due to the lack of eve function (Fig. 8B). They also exhibit a slower heart rate compared to controls and increased sensitivity to pacing stress, as well as a reduced flying ability and lifespan (Fujioka et al., 2005).

78

Similarly, when the eve repressor encoded by lbe is ectopically expressed in the mesodermal Eve lineages using a construct that drives lbe with eme (eme-lbe), embryonic heart development (with reduced Eve lineages) and adult physiology (with increased pacing-induced heart failure and arrhythmias) are disturbed (Buechling et al., 2009). In addition, the hearts are significantly hypoplastic, but not as severely as cardiac-only tinman mutants (Fig. 5F). Remarkably, all structural and functional abormalities of eme-lbe animals are restored by concomitant overexpression of eve in these same lineages (eme-Gal4UASeve; eme-lbe). These findings suggest that the eve-positive epi/pericardium is likely to play a critical, nonautonomous role in regulating myocardial structure and contractility.

X.C.  Insulin Receptor Signaling, Potassium Channels and Cardiac Aging As is the case in humans, and presumably in all organisms with a heart, cardiac performance declines with age (reviewed in Lakatta, 2001). In Drosophila, the heart also deteriorates as the fly ages, which is manifest in the progressive increase in heart period length, arrhythmias (Fig. 14D), susceptibility to pacing-induced heart failure (Fig. 15B) and disorganization of the regular spiral arrangements of myofibrils within cardiomyocytes (Fig. 13B,D) (Paternostro et al., 2001; Wessells and Bodmer, 2004; Wessells et al., 2004; Ocorr et al., 2007a,b; Wessells and Bodmer, 2007; Cammarato et al., 2008; Mery et al., 2008; Taghli-Lamallem et al., 2008; reviewed in Bodmer et al., 2005). Since heart development and basic aspects of heart function are also conserved, it seems likely that mechanisms controlling or accompanying the process of “cardiac aging” are also conserved. In addition, it is possible in flies to manipulate “cardiac aging” without necessarily affecting overall survival of the organism, and to study how aging of an organ is coordinated with that of the whole organism. Since Drosophila is the simplest genetic model with a heart (Bier and Bodmer, 2004), it may be uniquely suited to gain genetic insights into cardiac aging, also because the survival of the fly, which has a tracheal system for oxygen delivery, is not as critically dependent on heart function as that of vertebrates.

X.C.i.  SUR and KCNQ A number of genes and associated pathways have been identified that seem to play a role in the cardiac aging process in Drosophila. An example is the ATP-sensitive potassium channel-associated gene product of the SUR gene, which functions in coupling cell metabolism with electrical activity, and in protecting the heart from the effects of hypoxia (Akasaka et al., 2006). In the aging heart, SUR expression is reduced to a third, and young fly hearts with RNAi-mediated knockdown of SUR, like aged wild-type

PART  |  1  Heart Evolution

hearts, are more sensitive to pacing-induced stress, suggesting that SUR function is protecting against declining performance during cardiac aging (Akasaka et al., 2006). This result is reminiscent of mutations in human SUR2, which lead to cardiomyopathy (Bienengraeber et al., 2004). As with SUR, KCNQ encoded potassium channels (Section X.A) are also dramatically downregulated in the aging heart (Ocorr et al., 2007a). Reduced expression of these potassium channels may therefore contribute to the observed increase in arrhythmias of aging hearts. In agreement with this hypothesis, deterioration of heart function is further aggravated in old KCNQ mutants, which exhibit sustained periods of fibrillation alternating with asystoles (Fig. 14D), consistent with a severe deficit in repolarization capacity. In an attempt to replenish the age-dependent decline in cardiac potassium channel activity, KCNQ was overexpressed with a heart-specific Gal4 driver, in the hope of reversing the age-dependent deterioration in heart function. Indeed, old cardiac KCNQ transgenic flies exhibited fewer arrhythmias than wild-type with a heart rhythm pattern typical of younger flies (Ocorr et al., 2007a). Young flies with a cardiac KCNQ transgene, however, exhibited a much higher level of arrhythmias than their wild-type counterparts (K. Ocorr and R. B. unpubl.). This implies that the maintenance of a regular, robust heart beat depends on a delicate balance of ion channel activities, which apparently is not preserved at old age. In addition to ion channels, other structural gene functions are also necessary for normal cardiac function, and the pathology in mutants of these genes is enhanced with age. For example, myosin heavy chain mutants exhibit restrictive or dilated cardiomyopathies – depending on the nature of the mutation – in flies, as do their counterparts in humans, and the pathology is often more severe in old age (Cammarato et al., 2008). Moreover, deficiency in dystrophin function in flies causes a dilated cardiomypathy phenotype, which is associated with an age-dependent increase in myofibrillar disorganization (Taghli-Lamallem et al., 2008), consistent with the observed aggravation of muscle deterioration in old mdx mice (Chamberlain et al., 2007).

X.C.ii.  Insulin–TOR Signaling Mutations in insulin–insulin-like growth factor (IGF) signaling pathway are well-known to affect longevity in a variety of organisms (Fig. 15A) (reviewed in Kenyon, 2001; Barbieri et al., 2003; Helfland and Rogina, 2003; Tatar et al., 2003). For example, reduced function of the insulinlike receptor (InR) or its substrate encoded by chico extend the fly’s lifespan (Clancy et al., 2001; Tatar et al., 2001), and reverse the decline in age-related cardiac performance; old flies no longer exhibit slower heart rates or a higher incidence of pacing-induced heart failure, as observed in wild-type flies (Fig. 15B,D) (Wessells et al., 2004).

Chapter  |  1.2  Development and Aging of the Drosophila Heart

A similar amelioration of cardiac senescence was observed when reducing the function of the nutrient-sensing protein kinase TOR, a key component of a pathway that is tightly interlinked with InR signal transduction (Fig. 15A) (Oldham and Hafen, 2003; Luong et al., 2006). Interestingly, ablation of the cell cluster in the brain that secretes the fly’s insulin-like peptides also prolongs demographic life expectancy and improves heart function in old age. Therefore, systemic or nonautonomous manipulation of the InR-TOR signaling modulates cardiac senescence along with overall aging. This raises the question of how the decline in different body parts is coordinated, and whether aging characteristics can be manipulated organ-specifically. To approach the question of cardiac aging, Wessells et al. (2004) used a heart-specific Gal4 driver to manipulate the InR pathways specifically in this organ, and then examined how the age-dependent changes were affected. Indeed, it was found that heart-specific expression of InR caused accelerated aging of heart function, in that young flies already exhibited high susceptibility to pacing stress. Conversely, cardiac overexpression of the PI3Kinase anta­gonist dPTEN or the Akt/PKB responsive transcription factor encoded by dFOXO, both negative effectors of InR signaling, maintain robust cardiac performance in old age (Fig. 15C,E) (Wessells et al., 2004). These “young-at-heart” flies, however, do not live longer, consistent with the idea that altered cardiac senescence does not cause a systemic alteration in insulin signaling, which in turn would alter organism aging. In contrast, overexpressing dFOXO in the adult fat body that functions as an endocrine organ results in long-lived flies (Giannakou et al., 2004; Hwangbo et al., 2004). Cardiac-specific overexpression of TOR, as with InR overexpression, also causes premature decline of the heart’s performance. Thus, InR and TOR signaling may interact in a complex network to influence cardiac aging. Analysis of potential downstream effectors of InR–TOR signaling in modulating the cardiac age-related decline suggested the involvement of protein translation via alteration of the levels or activity of specific translation factors. Cardiac overexpression of the TOR target 4E-BP (Fig. 15A), which inhibits the eukaryotic initiation factor eiF4E, improves cardiac performance at old ages, whereas eif4E overexpression has the opposite effect (Wessells et al., 2009). In conclusion, cardiac-specific aging is significantly influenced by heart-autonomous InR and TOR signal transduction, suggesting that these signaling pathways can directly modulate the aging process of individual organs (Wessells et al., 2004; Luong et al., 2006). Thus, it seems that cardiac aging can be investigated in an organ-autonomous fashion in this genetic model system. Future studies will reveal whether mechanisms underlying cardiac functional aging in flies are conserved in higher organisms with potentially significant implications for mammalian cardiac aging.

79

References Akasaka, T., Klinedinst, S., Ocorr, K., Bustamante, E.L., Kim, S.K., Bodmer, R., 2006. The ATP-sensitive potassium (KATP) channelencoded dSUR gene is required for Drosophila heart function and is regulated by tinman. Proc. Natl. Acad. Sci. USA 103, 11999–12004. Allikian, M.J., Bhabha, G., Dospoy, P., Heydemann, A., Ryder, P., Earley, J.U., Wolf, M.J., Rockman, H.A., McNally, E.M., 2007. Reduced lifespan with heart and muscle dysfunction in Drosophila sarcoglycan mutants. Hum. Mol. Genet. 16, 2933–2943. Alvarez, A.D., Shi, W., Wilson, B.A., Skeath, J.B., 2003. Pannier and pointedP2 act sequentially to regulate Drosophila heart development. Development 130, 3015–3026. Ambros, V., 2001. MicroRNAs: tiny regulators with great potential. Cell 107, 823–826. Ambros, V., 2003. MicroRNA pathways in flies and worms: growth, death, fat, stress, and timing. Cell 114, 673–676. Azpiazu, N., Frasch, M., 1993. Tinman and bagpipe: two homeo box genes that determine cell fates in the dorsal mesoderm of Drosophila. Genes Dev. 7, 1325–1340. Azpiazu, N., Lawrence, P., Vincent, J.-P., Frasch, M., 1996. Segmentation and specification of the Drosophila mesoderm. Genes Dev. 10, 3183–3194. Baker, N.E., 1987. Molecular cloning of sequences from wingless, a segment polarity gene in Drosophila: the spatial distribution of a transcript in embryos. EMBO J. 6, 1765–1773. Baker, R., Schubiger, G., 1995. Ectoderm induces muscle-specific gene expression in Drosophila embryos. Development 121, 1387–1398. Barbieri, M., Bonafe, M., Franceschi, C., Paolisso, G., 2003. Insulin/ IGF-I-signaling pathway: an evolutionarily conserved mechanism of longevity from yeast to humans. Am. J. Physiol. Endocrinol. Metab. 285, 1064–1071. Bartel, D.P., 2004. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell 116, 281–297. Baylies, M., Martinez Arias, A., Bate, M., 1995. Wingless is required for the formation of a subset of muscle founder cells during Drosophila embryogenesis. Development 121, 3829–3837. Baylies, M.K., Bate, M., 1996. Twist: A myogenic switch in Drosophila. Science 272, 1481–1484. Beer, J., Technau, G.M., Campos-Ortega, J.A., 1987. Lineage analysis of transplanted individual cells in embryos of Drosophila melanogaster. Commitment and proliferative capabilities of mesodermal cells. Roux’s Arch. Dev. Biol. 196, 220–230. Beiman, M., Shilo, B., Volk, T., 1996. Heartless, a Drosophila FGF receptor homolog, is essential for cell migration and establishment of several mesodermal lineages. Genes Dev. 10, 2993–3002. Biemar, F., Zinzen, R., Ronshaugen, M., Sementchenko, V., Manak, J.R., Levine, M.S., 2005. Spatial regulation of microRNA gene expression in the Drosophila embryo. Proc. Natl. Acad. Sci. USA 102, 15907–15911. Bienengraeber, M., Olson, T.M., Selivanov, V.A., Kathmann, E.C., O’Cochlain, F., Gao, F., Karger, A.B., Ballew, J.D., Hodgson, D.M., Zingman, L.V., Pang, Y.P., Alekseev, A.E., Terzic, A., 2004. ABCC9 mutations identified in human dilated cardiomyopathy disrupt catalytic KATP channel gating. Nat. Genet. 36, 382–387. Bier, E., Bodmer, R., 2004. Drosophila, an emerging model for cardiac disease. Gene 342, 1–11. Bodmer, R., 1993. The gene tinman is required for specification of the heart and visceral muscles in Drosophila. Development 118, 719–729.

80

Bodmer, R., 1995. Heart development in Drosophila and its relationship to vertebrates. Trends Cardiovasc. Med. 5, 21–28. Bodmer, R., Jan, L.Y., Jan, Y.N., 1990. A new homeobox-containing gene, msh-2, is transiently expressed early during mesoderm formation of Drosophila. Development 110, 661–669. Bodmer, R., Wessells, R., Johnson, E.C., Dowse, H., 2005. Heart development and function. In: Gilbert, L., Latrau, K., Gill, S. (Eds.), Comprehensive Insect Science, vol. 2. Elsevier, Amsterdam, pp. 199–250. Boulay, J.L., Dennefeld, C., Alberga, A., 1987. The Drosophila developmental gene snail encodes a protein with nucleic acid binding fingers. Nature 330, 395–398. Bour, B., O’Brien, M., Lockwood, W., Goldstein, E., Bodmer, R., Taghert, P., Abmayr, S., Nguyen, H., 1995. Drosophila MEF2, a transcription factor that is essential for myogenesis. Genes Dev. 9, 730–741. Bruneau, B.G., Nemer, G., Schmitt, J.P., Charron, F., Robitaille, L., Caron, S., Conner, D.A., Gessler, M., Nemer, M., Seidman, C.E., Seidman, J.G., 2001. A murine model of Holt-Oram syndrome defines roles of the T-box transcription factor Tbx5 in cardiogenesis and disease. Cell 106, 709–721. Buechling, T., Akasaka, T., Vogler, G., Ruiz-Lozano, P., Ocorr, K., Bodmer, R., 2009. Non-autonomous modulation of heart rhythm, contractility and morphology in adult fruit flies. Dev. Biol. 328, 483–492. Cadigan, K., Grossniklaus, U., Gehring, W., 1994. Localized expression of sloppy paired protein maintains the polarity of Drosophila para­ segments. Development 8, 899–913. Cammarato, A., Dambacher, C.M., Knowles, A.F., Kronert, W.A., Bodmer, R., Ocorr, K., Bernstein, S.I., 2008. Myosin transducer mutations differentially affect motor function, myofibril structure, and the performance of skeletal and cardiac muscles. Mol. Biol. Cell 19, 553–562. Carmena, A., Gisselbrecht, S., Harrison, J., Jimenez, F., Michelson, A.M., 1998. Combinatorial signaling codes for the progressive determination of cell fates in the Drosophila embryonic mesoderm. Genes Dev. 12, 3910–3922. Chamberlain, J.S., Metzger, J., Reyes, M., Townsend, D., Faulkner, J.A., 2007. Dystrophin-deficient mdx mice display a reduced life span and are susceptible to spontaneous rhabdomyosarcoma. Faseb. J. 21, 2195–2204. Chartier, A., Zaffran, S., Astier, M., Semeriva, M., Gratecos, D., 2002. Pericardin, a Drosophila type IV collagen-like protein is involved in the morphogenesis and maintenance of the heart epithelium during dorsal ectoderm closure. Development 129, 3241–3253. Chen, C., Schwartz, R., 1995. Identification of novel DNA binding targets and regulatory domains of a murine tinman homeodomain factor, nkx-2.5. J. Biol. Chem. 270, 15628–15633. Chen, D., Zhao, M., Mundy, G.R., 2004. Bone morphogenetic proteins. Growth Factors 22, 233–241. Choi, C.Y., Lee, Y.M., Kim, Y.H., Park, T., Jeon, B.H., Schulz, R.A., Kim, Y., 1999. The homeodomain transcription factor NK-4 acts as either a transcriptional activator or repressor and interacts with the p300 coactivator and the Groucho corepressor. J. Biol. Chem. 274, 31543–31552. Clancy, D.J., Gems, D., Harshman, L.G., Oldham, S., Stocker, H., Hafen, E., Leevers, S.J., Partridge, L., 2001. Extension of life-span by loss of CHICO, a Drosophila insulin receptor substrate protein. Science 292, 104–106. Clevers, H., 2006. Wnt/beta-catenin signaling in development and disease. Cell 127, 469–480.

PART  |  1  Heart Evolution

Costa, M., Wilson, E., Wieschaus, E., 1994. A putative cell signal encoded by the folded gastrulation gene coordinates cell shape changes during Drosophila gastrulation. Cell 76, 1075–1089. Creemers, E.E., Sutherland, L.B., Oh, J., Barbosa, A.C., Olson, E.N., 2006. Coactivation of MEF2 by the SAP domain proteins myocardin and MASTR. Mol. Cell 23, 83–96. Cripps, R.M., Zhao, B., Olson, E.N., 1999. Transcription of the myogenic regulatory gene Mef2 in cardiac, somatic, and visceral muscle cell lineages is regulated by a Tinman-dependent core enhancer. Dev. Biol. 215, 420–430. Cripps, R.M., Black, B.L., Zhao, B., Lien, C.L., Schulz, R.A., Olson, E.N., 1998. The myogenic regulatory gene Mef2 is a direct target for transcriptional activation by Twist during Drosophila myogenesis. Genes Dev. 12, 422–434. Crozatier, M., Ubeda, J.M., Vincent, A., Meister, M., 2004. Cellular immune response to parasitization in Drosophila requires the EBF orthologue collier. PLoS Biol. 2, E196. Cubadda, Y., Heitzler, P., Ray, R.P., Bourouis, M., Ramain, P., Gelbart, W., Simpson, P., Haenlin, M., 1997. U-shaped encodes a zinc finger protein that regulates the proneural genes achaete and scute during the formation of bristles in Drosophila. Genes Dev. 22, 3083–3095. Curtis, N.J., Ringo, J.M., Dowse, H.B., 1999. Morphology of the pupal heart, adult heart, and associated tissues in the fruit fly, Drosophila melanogaster. J. Morphol. 240, 225–235. D’Alessio, M., Frasch, M., 1996. Msh may play a conserved role in dorsoventral patterning of the neuroectoderm and mesoderm. Mech. Devel. 58, 217–231. Damante, G., Pellizzari, L., Esposito, G., Fogolari, F., Viglino, P., Fabbro, D., Tell, G., Formisano, S., Di Lauro, R., 1996. A molecular code dictates sequence-specific DNA recognition by homeodomains. EMBO J. 15, 4992–5000. Damm, C., Wolk, A., Buttgereit, D., Loher, K., Wagner, E., Lilly, B., Olson, E.N., Hasenpusch-Theil, K., Renkawitz-Pohl, R., 1998. Independent regulatory elements in the upstream region of the Drosophila beta 3 tubulin gene (beta Tub60D) guide expression in the dorsal vessel and the somatic muscles. Dev. Biol. 199, 138–249. Das, D., Aradhya, R., Ashoka, D., Inamdar, M., 2008. Post-embryonic pericardial cells of Drosophila are required for overcoming toxic stress but not for cardiac function or adult development. Cell Tissue Res. 331, 565–570. De Velasco, B., Mandal, L., Mkrtchyan, M., Hartenstein, V., 2006. Subdivision and developmental fate of the head mesoderm in Drosophila melanogaster. Dev. Genes Evol. 216, 39–51. De Velasco, B., Shen, J., Go, S., Hartenstein, V., 2004. Embryonic development of the Drosophila corpus cardiacum, a neuroendocrine gland with similarity to the vertebrate pituitary, is controlled by sine oculis and glass. Dev. Biol. 274, 280–294. Deng, W.M., Schneider, M., Frock, R., Castillejo-Lopez, C., Gaman, E.A., Baumgartner, S., Ruohola-Baker, H., 2003. Dystroglycan is required for polarizing the epithelial cells and the oocyte in Drosophila. Development 130, 173–184. DeRobertis, E., Sasai, Y., 1996. A common plan for dorsoventral patterning in Bilateria. Nature 380, 37–40. Dierick, H., Bejsovec, A., 1999. Cellular mechanisms of Wingless/Wnt signal transduction. Curr. Top. Dev. Biol. 43, 153–190. DiFrancesco, D., Borer, J.S., 2007. The funny current: cellular basis for the control of heart rate. Drugs 67 (Suppl. 2), 15–24. Dowse, H., Ringo, J., Power, J., Johnson, E., Kinney, K., White, L., 1995. A congenital heart defect in Drosophila caused by an action-potential mutation. J. Neurogenet. 10, 153–168.

Chapter  |  1.2  Development and Aging of the Drosophila Heart

Dulcis, D., Levine, R.B., 2003. Innervation of the heart of the adult fruit fly, Drosophila melanogaster. J. Comp. Neurol. 465, 560–578. Dulcis, D., Levine, R.B., 2005. Glutamatergic innervation of the heart initiates retrograde contractions in adult Drosophila melanogaster. J. Neurosci. 25, 271–280. Dulcis, D., Levine, R.B., Ewer, J., 2005. Role of the neuropeptide CCAP in Drosophila cardiac function. J. Neurobiol. 64, 259–274. Dunin-Borkowski, O., Brown, N., Bate, M., 1995. Anterior–posterior subdivision and the diversification of the mesoderm in Drosophila. Development 121, 4183–4193. Durocher, D., Charron, F., Warren, R., Schwartz, R.J., Nemer, M., 1997. The cardiac transcription factors Nkx2-5 and GATA-4 are mutual cofactors. EMBO J. 16, 5687–5696. Edmondson, D., Lyons, G., Martin, J., Olson, E., 1994. Mef2 gene expression marks the cardiac and skeletal muscle lineages during mouse embryogenesis. Development 120, 1251–1263. Ferguson, E., Anderson, K., 1992. Decapentaplegic acts as a morphogen to organize dorsal-ventral pattern in the Drosophila embryo. Cell 71, 451–461. Fink, M., Callol-Massot, C., Chu, A., Ruiz-Lozano, P., Izpisua-Belmonte, J.C., Giles, W., Bodmer, R., Ocorr, K., 2008. Semi-automatic detection and quantification of heart beat parameters in Drosophila, zebrafish and embryonic mouse hearts. BioTechniques 45, 101–113. Fortini, M.E., Lai, Z.C., Rubin, G.M., 1991. The Drosophila zfh-1 and zfh-2 genes encode novel proteins containing both zinc-finger and homeodomain motifs. Mech. Dev. 34, 113–122. Fossett, N., Tevosian, S.G., Gajewski, K., Zhang, Q., Orkin, S.H., Schulz, R.A., 2001. The Friend of GATA proteins U-shaped, FOG-1, and FOG-2 function as negative regulators of blood, heart, and eye development in Drosophila. Proc. Natl. Acad. Sci. USA 98, 7342–7347. Francois, V., Solloway, M., O’Neill, J., Emery, J., Bier, E., 1994. Dorsal– ventral patterning of the Drosophila embryo depends on a putative negative growth factor encoded by the short gastrulation gene. Genes Dev. 8, 2602–2616. Frasch, M., 1995. Induction of visceral and cardiac mesoderm by ectodermal Dpp in the early Drosophila embryo. Nature 374, 464–467. Frasch, M., Hoey, T., Rushlow, C., Doyle, H., Levine, M., 1987. Characterization and localization of the even-skipped protein of Drosophila. Embo. J. 6, 749–759. Fremion, F., 1999. The heterotrimeric protien Go is required for the formation of heart epithelium in Drosophila. J. Cell Biol. 145, 1063–1076. Frolov, R.V., Berim, I.G., Singh, S., 2008. Inhibition of delayed rectifier potassium channels and induction of arrhythmia: a novel effect of celecoxib and the mechanism underlying it. J. Biol. Chem. 283, 1518–1524. Fujioka, M., Emi-Sarker, Y., Yusibova, G.L., Goto, T., Jaynes, J.B., 1999. Analysis of an even-skipped rescue transgene reveals both composite and discrete neuronal and early blastoderm enhancers, and multistripe positioning by gap gene repressor gradients. Development 126, 2527–2538. Fujioka, M., Wessells, R.J., Han, Z., Liu, J., Fitzgerald, K., Yusibova, G.L., Zamora, M., Ruiz-Lozano, P., Bodmer, R., Jaynes, J.B., 2005. Embryonic even skipped-dependent muscle and heart cell fates are required for normal adult activity, heart function, and lifespan. Circ. Res. 97, 1108–1114. Gabay, L., Seger, R., Shilo, B.-Z., 1997. MAP kinase in situ activation atlas during Drosophila embryogenesis. Development 124, 3535–3541. Gajewski, K., Kim, Y., Lee, Y., Olson, E., Schulz, R., 1997. D-mef2 is a target for Tinman activation during Drosophila heart development. EMBO J. 16, 515–522.

81

Gajewski, K., Kim, Y., Choi, C., Schulz, R., 1998. Combinatorial control of Drosophila mef2 gene expression in cardiac and somatic muscle cell lineages. Dev. Genes Evol. 208, 382–392. Gajewski, K., Fossett, N., Molkentin, J., Schulz, R., 1999. The zinc finger proteins Pannier and GATA4 function as cardiogenic factors in Drosophila. Development 126, 5679–5688. Gajewski, K., Choi, C., Kim, Y., Schulz, R., 2000. Genetically distinct cardial cells within the Drosophila heart. Genesis 28, 36–43. Gajewski, K., Zhang, Q., Choi, C., Fossett, N., Dang, A., Kim, Y., Kim, Y., Schulz, R., 2001. Pannier is a transcriptional target and partner of Tinman during Drosophila cardiogenesis. Dev. Biol. 233, 425–436. Garg, V., Kathiriya, I.S., Barnes, R., Schluterman, M.K., King, I.N., Butler, C.A., Rothrock, C.R., Eapen, R.S., Hirayama-Yamada, K., Joo, K., Matsuoka, R., Cohen, J.C., Srivastava, D., 2003. GATA4 mutations cause human congenital heart defects and reveal an interaction with TBX5. Nature 424, 443–447. Giannakou, M.E., Goss, M., Junger, M.A., Hafen, E., Leevers, S.J., Partridge, L., 2004. Long-lived Drosophila with overexpressed dFOXO in adult fat body. Science 305, 361. Gisselbrecht, S., Skeath, J., Doe, C., Michelson, A., 1996. Heartless encodes a fibroblast growth factor receptor (DFR1/DFGF-R2) involved in the directional migration of early mesodermal cells in the Drosophila embryo. Genes Dev. 10, 3003–3017. Gorczyca, M.G., Phillis, R.W., Budnik, V., 1994. The role of tinman, a mesodermal cell fate gene, in axon pathfinding during the development of the transverse nerve in Drosophila. Development 120, 2143–2152. Grau, Y., Carteret, C., Simpson, P., 1984. Mutations and chromosomal rearrangements affecting the expression of snail, a gene involved in embryonic patterning in Drosophila melanogaster. Genetics 108, 347–360. Gryzik, T., Müller, H.A., 2004. FGF8-like1 and FGF8-like2 encode putative ligands of the FGF receptor Htl and are required for mesoderm migration in the Drosophila gastrula. Curr. Biol. 14, 659–667. Haag, T.A., Haag, N.P., Lekven, A.C., Hartenstein, V., 1999. The role of cell adhesion molecules in Drosophila heart morphogenesis: faint sausage, shotgun/DE-cadherin, and laminin A are required for discrete stages in heart development. Dev. Biol. 208, 56–69. Haenlin, M., Cubadda, Y., Blondeau, F., Heitzler, P., Lutz, Y., Simpson, P., Ramain, P., 1997. Transcriptional activity of pannier is regulated negatively by heterodimerization of the GATA DNA-binding domain with a cofactor encoded by the u-shaped gene of Drosophila. Genes Dev. 11, 3096–3108. Halfon, M.S., Carmena, A., Gisselbrecht, S., Sackerson, C.M., Jimenez, F., Baylies, M.K., Michelson, A.M., 2000. Ras pathway specificity is determined by the integration of multiple signal-activated and tissuerestricted transcription factors. Cell 103, 63–74. Han, Z., Bodmer, R., 2003. Myogenic cells fates are antagonized by Notch only in asymmetric lineages of the Drosophila heart, with or without cell division. Development 130, 3039–3051. Han, Z., Olson, E.N., 2005. Hand is a direct target of Tinman and GATA factors during Drosophila cardiogenesis and hematopoiesis. Development 132, 3525–3536. Han, Z., Li, X., Wu, J., Olson, E.N., 2004. A myocardin-related transcription factor regulates activity of serum response factor in Drosophila. Proc. Natl. Acad. Sci. USA 101, 12567–12572. Han, Z., Yi, P., Li, X., Olson, E.N., 2006. Hand, an evolutionarily conserved bHLH transcription factor required for Drosophila cardiogenesis and hematopoiesis. Development 133, 1175–1182. Han, Z., Fujioka, M., Su, M., Liu, M., Jaynes, J.B., Bodmer, R., 2002. Transcriptional integration of competence modulated by mutual

82

repression generates cell-type specificity within the cardiogenic mesoderm. Dev. Biol. 252, 225–240. Hanley, P.J., Daut, J., 2005. K(ATP) channels and preconditioning: a reexamination of the role of mitochondrial K(ATP) channels and an overview of alternative mechanisms. J. Mol. Cell Cardiol. 39, 17–50. Hartenstein, V., Mandal, L., 2006. The blood/vascular system in a phylogenetic perspective. Bioessays 28, 1203–1210. Harvey, R., 1996. NK-2 homeobox genes and heart development. Dev. Biol. 178, 203–216. Helfland, S.L., Rogina, B., 2003. Molecular genetics of aging in the fly: is this the end of the beginning? Bioessays 25, 134–141. Hendren, J.D., Shah, A.P., Arguelles, A.M., Cripps, R.M., 2007. Cardiac expression of the Drosophila Sulphonylurea receptor gene is regulated by an intron enhancer dependent upon the NK homeodomain factor Tinman. Mech. Dev. 124, 416–426. Hirota, Y., Sawamoto, K., Okano, H., 2002. Tincar encodes a novel transmembrane protein expressed in the Tinman-expressing cardioblasts of Drosophila. Gene Expr. Patterns 2, 323–327. Hwangbo, D.S., Gersham, B., Tu, M.P., Palmer, M., Tatar, M., 2004. Regulation of aging and neuronal insulin transcription by dFOXO in fat body of adult Drosophila. Nature 480, 652–656. Jagla, K., Dolle, P., Mattei, M., Jagla, T., Schuhbaur, B., Dretzen, G., Bellard, F., Bellard, M., 1995. Mouse Lbx1 and human LBX1 define a novel mammalian homeobox gene family related to the Drosophila lady bird genes. Mech. Dev. 53, 345–356. Jagla, K., Frasch, M., Jagla, T., Dretzen, G., Bellard, F., Bellard, M., 1997a. ladybird, a new component of the cardiogenic pathway in Drosophila required for diversification of heart precursors. Development 124, 3471–3479. Jagla, K., Jagla, T., Heitzler, P., Dretzen, G., Bellard, F., Bellard, M., 1997b. ladybird, a tandem of homeobox genes that maintain late wingless expression in terminal and dorsal epidermis of the Drosophila embryo. Development 124, 91–100. Jagla, T., Bidet, Y., Da Ponte, J.P., Dastugue, B., Jagla, K., 2002. Crossrepressive interactions of identity genes are essential for proper specification of cardiac and muscular fates in Drosophila. Development 129, 1037–1047. Janssen, R., Damen, W.G., 2008. Diverged and conserved aspects of heart formation in a spider. Evol. Dev. 10, 155–165. Jazwinska, A., Rushlow, C., Roth, S., 1999. The role of brinker in mediating the graded response to Dpp in early Drosophila embryos. Development 126, 3323–3334. Jentsch, T.J., 2000. Neuronal KCNQ potassium channels: physiology and role in disease. Nat. Rev. Neurosci. 1, 21–30. Johnson, A.N., Bergman, C.M., Kreitman, M., Newfeld, S.J., 2003. Embryonic enhancers in the dpp disk region regulate a second round of Dpp signaling from the dorsal ectoderm to the mesoderm that represses Zfh-1 expression in a subset of pericardial cells. Dev. Biol. 262, 137–151. Johnson, A.N., Burnett, L.A., Sellin, J., Paululat, A., Newfeld, S.J., 2007. Defective decapentaplegic signaling results in heart overgrowth and reduced cardiac output in Drosophila. Genetics 176, 1609–1624. Johnson, E., Ringo, J., Dowse, H., 1997. Modulation of Drosophila heartbeat by neurotransmitters. J. Comp. Physiol. (B) 167, 89–97. Johnson, E., Ringo, J., Bray, N., Dowse, H., 1998. Genetic and pharmacological identification of ion channels central to the Drosophila cardiac pacemaker. J. Neurogenet. 12, 1–24. Johnson, E., Sherry, T., Ringo, J., Dowse, H., 2002. Modulation of the cardiac pacemaker of Drosophila: cellular mechanisms. J. Comp. Physiol. (B) 172, 227–236.

PART  |  1  Heart Evolution

Junion, G., Bataille, L., Jagla, T., Da Ponte, J.P., Tapin, R., Jagla, K., 2007. Genome-wide view of cell fate specification: ladybird acts at multiple levels during diversification of muscle and heart precursors. Genes Dev. 21, 3163–3180. Junion, G., Jagla, T., Duplant, S., Tapin, R., Da Ponte, J.P., Jagla, K., 2005. Mapping Dmef2-binding regulatory modules by using a ChIPenriched in silico targets approach. Proc. Natl. Acad. Sci. USA 102, 18479–18484. Kelly, K.K., Meadows, S.M., Cripps, R.M., 2002. Drosophila MEF2 is a direct regulator of Actin57B transcription in cardiac, skeletal, and visceral muscle lineages. Mech. Dev. 110, 39–50. Kenyon, C., 2001. A conserved regulatory system for aging. Cell 105, 165–168. Kim, Y., Nirenberg, M., 1989. Drosophila NK-homeobox genes. Proc. Natl. Acad. Sci. USA 86, 7716–7720. Klinedinst, S.L., Bodmer, R., 2003. Gata factor Pannier is required to establish competence for heart progenitor formation. Development 130, 3027–3038. Klingensmith, J., Nusse, R., 1994. Signaling by wingless in Drosophila. Dev. Biol. 166, 396–414. Knirr, S., Frasch, M., 2001. Molecular integration of inductive and mesoderm-intrinsic inputs governs even-skipped enhancer activity in a subset of pericardial and dorsal muscle progenitors. Dev. Biol. 238, 13–26. Kölsch, V., Paululat, A., 2002. The highly conserved cardiogenic bHLH factor Hand is specifically expressed in circular visceral muscle progenitor cells and in all cell types of the dorsal vessel during Drosophila embryogenesis. Dev. Genes Evol. 212, 473–485. Kosman, D., Ip, Y.T., Levine, M., Arora, K., 1991. Establishment of the mesoderm-neuroectoderm boundary in the Drosophila embryo. Science 252, 118–122. Kulkarni, M.M., Booker, M., Silver, S.J., Friedman, A., Hong, P., Perrimon, N., Mathey-Prevot, B., 2006. Evidence of off-target effects associated with long dsRNAs in Drosophila melanogaster cell-based assays. Nat. Methods 3, 833–838. Kuo, C.T., Morrisey, E.E., Anandappa, R., Sigrist, K., Lu, M.M., Parmacek, M.S., Soudais, C., Leiden, J.M., 1997. GATA4 transcription factor is required for ventral morphogenesis and heart tube formation. Genes Dev. 11, 1048–1060. Kwon, C., Han, Z., Olson, E.N., Srivastava, D., 2005. MicroRNA1 influences cardiac differentiation in Drosophila and regulates Notch signaling. Proc. Natl. Acad. Sci. USA 102, 18986–18991. Lai, Z.C., Fortini, M.E., Rubin, G.M., 1991. The embryonic expression patterns of zfh-1 and zfh-2, two Drosophila genes encoding novel zinc-finger homeodomain proteins. Mech. Dev. 34, 123–134. Lakatta, E.G., 2001. Heart aging: a fly in the ointment? Circ. Res. 88, 984–986. Lalevee, N., Monier, B., Senatore, S., Perrin, L., Semeriva, M., 2006. Control of cardiac rhythm by ORK1, a Drosophila two-pore domain potassium channel. Curr. Biol. 16, 1502–1508. Latronico, M.V., Catalucci, D., Condorelli, G., 2007. Emerging role of microRNAs in cardiovascular biology. Circ. Res. 101, 1225–1236. Lawrence, P., Bodmer, R., Vincent, J., 1995. Segmental patterning of heart precursors in Drosophila. Development 121, 4303–4308. Lee, H., Frasch, M., 2000. Wingless effects mesoderm patterning and ectoderm segmentation events via induction of its downstream target sloppy paired. Development 127, 5497–5508. Leptin, M., 1991. Twist and snail as positive and negative regulators during Drosophila mesoderm formation. Genes Dev. 5, 1568–1576. Lilly, B., Galewsky, S., Firulli, A., Schulz, R., Olson, E., 1994. D-MEF2: a MADS box transcription factor expressed in differentiating mesoderm

Chapter  |  1.2  Development and Aging of the Drosophila Heart

and muscle cell lineages during Drosophila embryogenesis. Proc. Natl. Acad. Sci. USA 91, 5662–5666. Lilly, B., Zhao, B., Ranganayakulu, G., Paterson, B., Schulz, R., Olson, E., 1995. Requirement of MADS domain transcription factor DMEF2 for muscle formation in Drosophila. Science 267, 688–693. Lin, M., Nguyen, H., Dybala, C., Storti, R., 1996. Myocyte-specific enhancer factor 2 acts cooperatively with a muscle activator region to regulate Drosophila tropomyosin gene muscle expression. Proc. Natl. Acad. Sci. USA 93, 4623–4628. Lin, S., Storti, R., 1997. Developmental regulation of the Drosophila Tropomyosin I (TmI) gene is controlled by a muscle activator enhancer region that contains multiple cis-elements and binding sites for multiple proteins. Dev. Genet. 20, 297–306. Liu, J., Qian, L., Wessells, R.J., Bidet, Y., Jagla, K., Bodmer, R., 2006. Hedgehog and RAS pathways cooperate in the anterior–posterior specification and positioning of cardiac progenitor cells. Dev. Biol. 290, 373–385. Liu, J., Qian, L., Han, Z., Wu, X., Bodmer, R., 2008. Spatial specificity of mesodermal even-skipped expression relies on multiple repressor sites. Dev. Biol. 313, 876–886. Lo, P.C., Frasch, M., 2001. A role for the COUP-TF-related gene sevenup in the diversification of cardioblast identities in the dorsal vessel of Drosophila. Mech. Dev. 104, 49–60. Lo, P.C., Frasch, M., 2003. Establishing A–P polarity in the embryonic heart tube: a conserved function of Hox genes in Drosophila and vertebrates? Trends Cardiovasc. Med. 13, 182–187. Lo, P.C., Zaffran, S., Senatore, S., Frasch, M., 2007. The Drosophila Hand gene is required for remodeling of the developing adult heart and midgut during metamorphosis. Dev. Biol. 311, 287–296. Lo, P.C., Skeath, J.B., Gajewski, K., Schulz, R.A., Frasch, M., 2002. Homeotic genes autonomously specify the anteroposterior subdivision of the Drosophila dorsal vessel into aorta and heart. Dev. Biol. 251, 307–319. Lockwood, W.K., Bodmer, R., 2002. The patterns of wingless, decapentaplegic, and tinman position the Drosophila heart. Mech. Dev. 114, 13–26. Lovato, T.L., Nguyen, T.P., Molina, M.R., Cripps, R.M., 2002. The Hox gene abdominal-A specifies heart cell fate in the Drosophila dorsal vessel. Development 129, 5019–5027. Lu, J.R., McKinsey, T.A., Xu, H., Wang, D.Z., Richardson, J.A., Olson, E.N., 1999. FOG-2, a heart- and brain-enriched cofactor for GATA transcription factors. Mol. Cell Biol. 6, 4495–4502. Luong, N., Davies, C.R., Wessells, R.J., Graham, S.M., King, M.T., Veech, R., Bodmer, R., Oldham, S.M., 2006. Activated FOXOmediated insulin resistance is blocked by reduction of TOR activity. Cell Metab. 4, 133–142. MacMullin, A., Jacobs, J.R., 2006. Slit coordinates cardiac morphogenesis in Drosophila. Dev. Biol. 293, 154–164. McGinnis, W., Krumlauf, R., 1992. Homeobox genes and axial patterning. Cell 68, 283–302. Maggert, K., Levine, M., Frasch, M., 1995. The somatic–visceral subdivision of the embryonic mesoderm is initiated by dorsal gradient thresholds in Drosophila. Development 121, 2107–2116. Mann, T., Bodmer, R., Pandur, P., 2009. The Drosophila homolog of the vertebrate second heart field marker gene Islet1 is a key component in early cardiogenesis. Development 136 (2), 317–326. Männer, J., Ruiz-Lozano, P., 2008. Development and function of the epicardium. In: Bodmer, R. (Ed.), Advances in Developmental Biology: Cardiovascular Development, vol. 18. Elsevier B.V., The Netherlands, pp. 334–357.

83

Marty, T., Müller, B., Basler, K., Affolter, M., 2000. Schnurri mediates Dpp-dependent repression of brinker transcription. Nat. Cell Biol. 2, 745–749. Medioni, C., Astier, M., Zmojdzian, M., Jagla, K., Sémériva, M., 2008. Genetic control of cell morphogenesis during Drosophila melanogaster cardiac tube formation. J. Cell Biol. 182, 249–261. Merki, E., Zamora, M., Raya, A., Kawakami, Y., Wang, J., Zhang, X., Burch, J., Kubalak, S.W., Kaliman, P., Belmonte, J.C., Chien, K. R., Ruiz-Lozano, P., 2005. Epicardial retinoid X receptor alpha is required for myocardial growth and coronary artery formation. Proc. Natl. Acad. Sci. USA 102, 18455–18460. Mery, A., Taghli-Lamallem, O., Clark, K.A., Beckerle, M.C., Wu, X., Ocorr, K., Bodmer, R., 2008. The Drosophila muscle LIM protein, Mlp84B, is essential for cardiac function. J. Exp. Biol. 211, 15–23. Michelson, A.M., Gisselbrecht, S., Zhou, Y., Baek, K.H., Buff, E.M., 1998. Dual functions of the heartless fibroblast growth factor receptor in development of the Drosophila embryonic mesoderm. Dev. Genet. 22, 212–229. Miskolczi-McCallum, C.M., Scavetta, R.J., Svendsen, P.C., Soanes, K.H., Brook, W.J., 2005. The Drosophila melanogaster T-box genes midline and H15 are conserved regulators of heart development. Dev. Biol. 278, 459–472. Mispelon, M., Thakur, K., Chinn, L., Owen, R., Nichols, R., 2003. A nonpeptide provides insight into mechanisms that regulate Drosophila melanogaster heart contractions. Peptides 24, 1599–1605. Molina, M.R., Cripps, R.M., 2001. Ostia, the inflow tracts of the Drosophila heart, develop from a genetically distinct subset of cardial cells. Mech. Dev. 109, 51–59. Molkentin, J.D., Olson, E.N., 1997. GATA4: a novel transcriptional regulator of cardiac hypertrophy? Circulation 11, 3833–3835. Monier, B., Astier, M., Semeriva, M., Perrin, L., 2005. Steroid-dependent modification of Hox function drives myocyte reprogramming in the Drosophila heart. Development 132, 5283–5293. Monzen, K., Ito, Y., Naito, A.T., Kasai, H., Hiroi, Y., Hayashi, D., Shiojima, I., Yamazaki, T., Miyazono, K., Asashima, M., Nagai, R., Komuro, I., 2008. A crucial role of a high mobility group protein HMGA2 in cardiogenesis. Nat. Cell Biol. 10, 567–574. Mori, A.D., Bruneau, B.G., 2004. TBX5 mutations and congenital heart disease: Holt-Oram syndrome revealed. Curr. Opin. Cardiol. 19, 211–215. Nasonkin, I., Alikasifoglu, A., Ambrose, C., Cahill, P., Cheng, M., Sarniak, A., Egan, M., Thomas, P., 1999. A novel sulfonylurea receptor family member expressed in the embryonic Drosophila dorsal vessel and tracheal system. J. Biol. Chem. 274, 29420–29425. Nguyen, H., Bodmer, R., Abmayr, S., McDermott, J., Spoerel, N., 1994. D-mef2: a Drosophila mesoderm-specific MADS box-containing gene with a biphasicexpression profile during embryogenesis. Proc. Natl. Acad. Sci. USA 91, 7520–7524. Nguyen, H.T., Xu, X., 1998. Drosophila mef2 expression during mesoderm development is controlled by a complex array of cis-acting regulatory modules. Dev. Biol. 204, 550–566. Ocorr, K., Reeves, N.L., Wessells, R.J., Fink, M., Chen, H.S., Akasaka, T., Yasuda, S., Metzger, J.M., Giles, W., Posakony, J.W., Bodmer, R., 2007a. KCNQ potassium channel mutations cause cardiac arrhythmias in Drosophila that mimic the effects of aging. Proc. Natl. Acad. Sci. USA 103, 3943–3948. Ocorr, K., Perrin, L., Lim, H.Y., Qian, L., Wu, X., Bodmer, R., 2007b. Genetic control of heart function and aging in Drosophila. Trends Cardiovasc. Med. 17, 177–182. Ocorr, K., Akasaka, T., Bodmer, R., 2007c. Age-related cardiac disease model of Drosophila. Mech. Ageing Dev. 128, 112–116.

84

Ocorr, K.A., Crawley, T., Gibson, G., Bodmer, R., 2007d. Genetic variation for cardiac dysfunction in Drosophila. PLoS ONE 2, e601. Oka, T., Maillet, M., Watt, A.J., Schwartz, R.J., Aronow, B.J., Duncan, S. A., Molkentin, J.D., 2006. Cardiac-specific deletion of Gata4 reveals its requirement for hypertrophy, compensation, and myocyte viability. Circ. Res. 98, 837–845. Oldham, S., Hafen, E., 2003. Insulin/IGF and target of rapamycin signaling: a TOR de force in growth control. Trends Cell Biol. 13, 79–85. Olson, E.N., 2006. Gene regulatory networks in the evolution and development of the heart. Science 313, 1922–1927. Papaefthmiou, C., Theophilidis, G., 2001. An in vitro method for recording the electrical activity of the isolated heart of the adult Drosophila melanogaster. In Vitro Cell. Dev. Biol. Anim. 37, 445–449. Park, M., Wu, X., Golden, K., Axelrod, J.D., Bodmer, R., 1996. The wingless signaling pathway is directly involved in Drosophila heart development. Dev. Biol. 177, 104–116. Park, M., Lewis, C., Turbay, D., Chung, A., Chen, J.N., Evans, S., Breitbart, R.E., Fishman, M.C., Izumo, S., Bodmer, R., 1998. Differential rescue of visceral and cardiac defects in Drosophila by vertebrate tinman-related genes. Proc. Natl. Acad. Sci. USA 95, 9366–9371. Paternostro, G., Vignola, C., Bartsch, D.U., Omens, J.H., McCulloch, A.D., Reed, J.C., 2001. Age-associated cardiac dysfunction in Drosophila melanogaster. Circ. Res. 88, 1053–1058. Perrin, L., Monier, B., Ponzielli, R., Astier, M., Semeriva, M., 2004. Drosophila cardiac tube organogenesis requires multiple phases of Hox activity. Dev. Biol. 272, 419–431. Ponzielli, R., Astier, M., Chartier, A., Gallet, A., Therond, P., Semeriva, M., 2002. Heart tube patterning in Drosophila requires integration of axial and segmental information provided by the Bithorax Complex genes and hedgehog signaling. Development 129, 4509–4521. Potthoff, M.J., Olson, E.N., 2007. MEF2: a central regulator of diverse developmental programs. Development 134, 4131–4140. Poulson, D.F., 1950. Histogenesis, organogenesis and differentiation in the embryo of Drosophila melanogaster. In: Demerec, M. (Ed.), The Biology of Drosophila. John Wiley & Sons, Inc., New York, pp. 168–274. Qian, L., Liu, J., Bodmer, R., 2005a. Neuromancer-Tbx20-related genes (H15/midline) promote cell fate specification and morphogenesis of the Drosophila heart. Dev. Biol. 279, 509–524. Qian, L., Liu, J., Bodmer, R., 2005b. Slit and Robo control cardiac cell polarity and morphogenesis. Curr. Biol. 15, 2271–2278. Qian, L., Mohapatra, B., Akasakaa, T., Liu, J., Ocorr, K., Towbin, T. A., Bodmer, R., 2008. Transcription factor neuromancer/TBX20 is required for cardiac function in Drosophila with implications for human heart disease. Proc. Natl. Acad. Sci. USA 105, 19833–19838. Ranganayakulu, G., Schulz, R.A., Olson, E.N., 1996. wingless signaling induces nautilus expression in the ventral mesoderm of the Drosophila embryo. Dev. Biol. 176, 143–148. Ranganayakulu, G., Elliott, D., Harvey, R., Olson, E., 1998. Divergent roles for NK-2 class homeobox genes in cardiogenesis in flies and mice. Development 125, 3037–3048. Ranganayakulu, G., Zhao, B., Dokidis, A., Molkentin, J., Olson, E., Schulz, R., 1995. A series of mutations in the D-MEF2 transcription factor reveal multiple functions in larval and adult myogenesis in Drosophila. Dev. Biol. 171, 169–181. Ray, V.M., Dowse, H.B., 2005. Mutations in and deletions of the Ca2 channel-encoding gene cacophony, which affect courtship song in Drosophila, have novel effects on heart beating. J. Neurogenet. 19, 39–56.

PART  |  1  Heart Evolution

Reim, I., Frasch, M., 2005. The Dorsocross T-box genes are key components of the regulatory network controlling early cardiogenesis in Drosophila. Development 132, 4911–4925. Reim, I., Lee, H.H., Frasch, M., 2003. The T-box-encoding Dorsocross genes function in amnioserosa development and the patterning of the dorsolateral germ band downstream of Dpp. Development 130, 3187–3204. Reim, I., Mohler, J., Frasch, M., 2005. Tbx20-related genes, mid and H15, are required for tinman expression, proper patterning, and normal differentiation of cardioblasts in Drosophila. Mech. Dev. 132, 4911–4925. Riechmann, V., Irion, U., Wilson, R., Grosskortenhaus, A., Leptin, M., 1997. Control of cell fates and segmentation in the Drosophila mesoderm. Development 124, 2915–2922. Rizki, T.M., 1978. The circulatory system and associated cells and tissues. In: Ashburner, M., Wright, T.R.F. (Eds.), The Genetics and Biology of Drosophila. Academic Press, London and New York, pp. 397–452. Robinson, R.B., Brink, P.R., Cohen, I.S., Rosen, M.R., 2006. I(f) and the biological pacemaker. Pharmacol. Res. 53, 407–415. Rudnicki, M., Jaenisch, R., 1995. The MyoD family of transcription factors and skeletal myogenesis. Bioessays 17, 203–209. Rugendorff, A., Younossi-Hartenstein, A., Hartenstein, V., 1994. Embryonic origin and differentiation of the Drosophila heart. Roux’s Arch. Dev. Biol. 203, 266–280. Ruiz Gomez, M., Bate, M., 1997. Segregation of myogenic lineages in Drosophila requires numb. Development 23, 4857–4866. Rusch, J., Levine, M., 1996. Threshold responses to the dorsal regulatory gradient and the subdivision of primary tissue territories in the Drosophila embryo. Curr. Opin. Genet. Dev. 6, 416–423. Ryan, K.M., Hoshizaki, D.K., Cripps, R.M., 2005. Homeotic selector genes control the patterning of seven-up expressing cells in the Drosophila dorsal vessel. Mech. Dev. 122, 1023–1033. Ryan, K.M., Hendren, J.D., Helander, L.A., Cripps, R.M., 2007. The NK homeodomain transcription factor Tinman is a direct activator of seven-up in the Drosophila dorsal vessel. Dev. Biol. 302, 694–702. Sandmann, T., Jensen, L.J., Jakobsen, J.S., Karzynski, M.M., Eichenlaub, M.P., Bork, P., Furlong, E.E., 2006. A temporal map of transcription factor activity: mef2 directly regulates target genes at all stages of muscle development. Dev. Cell 10, 797–807. Sandmann, T., Girardot, C., Brehme, M., Tongprasit, W., Stolc, V., Furlong, E.E., 2007. A core transcriptional network for early mesoderm development in Drosophila melanogaster. Genes Dev. 21, 436–449. Sanguinetti, M.C., Tristani-Firouzi, M., 2006. hERG potassium channels and cardiac arrhythmia. Nature 440, 463–469. Santiago-Martínez, E., Soplop, N.H., Kramer, S.G., 2006. Lateral positioning at the dorsal midline: Slit and Roundabout receptors guide Drosophila heart cell migration. Proc. Natl. Acad. Sci. USA 103, 12441–12446. Santiago-Martínez, E., Soplop, N.H., Patel, R., Kramer, S.G., 2008. Repulsion by Slit and Roundabout prevents Shotgun/E-cadherinmediated cell adhesion during Drosophila heart tube lumen formation. J. Cell Biol. 182, 241–248. Sanyal, S., Jennings, T., Dowse, H., Ramaswami, M., 2006. Conditional mutations in SERCA, the Sarco-endoplasmic reticulum Ca(2)ATPase, alter heart rate and rhythmicity in Drosophila. J. Comp. Physiol. (B) 176, 253–263. Schott, J.J., Benson, D.W., Basson, C.T., Pease, W., Silberbach, G. M., Moak, J.P., Maron, B.J., Seidman, C.E., Seidman, J.G., 1998.

Chapter  |  1.2  Development and Aging of the Drosophila Heart

Congenital heart disease caused by mutations in the transcription factor NKX2-5. Science 281, 108–111. Schwartz, R.J., Olson, E.N., 1999. Building the heart piece by piece: modularity of cis-elements regulating Nkx2-5 transcription. Development 126, 4187–4192. Seidel, F., Bock, E., Krause, G., 1940. Die Organisation des Insekteneies. Naturwissenschaften 28, 433–446. Seino, S., 1999. ATP-sensitive potassium channels: a model of heteromultimeric potassium channel/receptor assemblies. Annu. Rev. Physiol. 61, 337–362. Sellin, J., Albrecht, S., Kölsch, V., Paululat, A., 2006. Dynamics of heart differentiation, visualized utilizing heart enhancer elements of the Drosophila melanogaster bHLH transcription factor Hand. Gene. Expr. Patterns 6, 360–375. Shishido, E., Higashijima, S., Emori, Y., Saigo, K., 1993. Two FGFreceptor homologues of Drosophila: one is expressed in mesodermal primordium in early embryos. Development 117, 751–761. Shishido, E., Ono, N., Kojima, T., Saigo, K., 1997. Requirements of DFR1/Heartless, a mesoderm-specific Drosophila FGF-receptor, for the formation of heart, visceral and somatic muscles, and ensheathing of longitudinal axon tracts in CNS. Development 124, 2119–2128. Simpson, P., 1983. Maternal-zygotic gene interactions during formation of the dorsoventral pattern in Drosophila embryos. Genetics 105, 615–632. Smith, S.T., Jaynes, J.B., 1996. A conserved region of engrailed, shared among all en-, gsc-, Nk1-, Nk2- and msh-class homeoproteins, mediates active transcriptional repression in vivo. Development 122, 3141–3150. Sokol, N.S., Ambros, V., 2005. Mesodermally expressed Drosophila microRNA-1 is regulated by Twist and is required in muscles during larval growth. Genes Dev. 19, 2343–2354. Spencer, F.A., Hoffmann, F.M., Gelbart, W.M., 1982. Decapentaplegic: a gene complex affecting morphogenesis in Drosophila melanogaster. Cell 28, 451–461. Srivastava, D., 1999. HAND proteins: molecular mediators of cardiac development and congenital heart disease. Trends Cardiovasc. Med. 9, 11–18. Staehling-Hampton, K., Hoffmann, F.M., Baylies, M.K., Rushton, E., Bate, M., 1994. Dpp induces mesodermal gene expression in Drosophila. Nature 372, 783–786. Stathopoulos, A., Tam, B., Ronshaugen, M., Frasch, M., Levine, M., 2004. Pyramus and thisbe: FGF genes that pattern the mesoderm of Drosophila embryos. Genes Dev. 18, 687–699. Stronach, B.E., Renfranz, P.J., Lilly, B., Beckerle, M.C., 1999. Muscle LIM proteins are associated with muscle sarcomeres and require dMEF2 for their expression during Drosophila myogenesis. Mol. Biol. Cell 10, 2329–2342. Su, M.T., Fujioka, M., Goto, T., Bodmer, R., 1999. The Drosophila homeobox genes zfh-1 and even-skipped are required for cardiac-specific differentiation of a numb-dependent lineage decision. Development 126, 3241–3251. Svensson, E.C., Tufts, R.L., Polk, C.E., Leiden, J.M., 1999. Molecular cloning of FOG-2: a modulator of transcription factor GATA-4 in cardiomyocytes. Proc. Natl. Acad. Sci. USA 3, 956–961. Svensson, E.C., Huggins, G.S., Dardik, F.B., Polk, C.E., Leiden, J.M., 2000. A functionally conserved N-terminal domain of the friend of GATA-2 (FOG-2) protein represses GATA4-dependent transcription. J. Biol. Chem. 27, 20762–20769. Taghli-Lamallem, O., Akasaka, T., Hogg, G., Nudel, U., Yaffe, D., Chamberlain, J.S., Ocorr, K., Bodmer, R., 2008. Dystrophin

85

deficiency in Drosophila reduces lifespan and causes a dilated cardiomyopathy phenotype. Aging Cell 7, 237–249. Tao, Y., Wang, J., Tokusumi, T., Gajewski, K., Schulz, R.A., 2007. Requirement of the LIM homeodomain transcription factor Tailup for normal heart and hematopoietic organ formation in Drosophila melanogaster. Mol. Cell Biol. 27, 3962–3969. Tatar, M., Bartke, A., Antebi, A., 2003. The endocrine regulation of aging by insulin-like signals. Science 299, 1346–1351. Tatar, M., Kopelman, A., Epstein, D., Tu, M.P., Yin, C.M., Garofalo, R.S., 2001. A mutant Drosophila insulin receptor homolog that extends life-span and impairs neuroendocrine function. Science 292, 107–110. Taylor, M., Beatty, K., Hunter, H., Baylies, M., 1995. Drosophila MEF2 is regulated by twist and is expressed in both the primordia and differentiated cells of the embryonic somatic, visceral and heart musculature. Mech. Dev. 50, 29–41. Tepass, U., Hartenstein, V., 1994. The development of cellular junctions in the Drosophila embryo. Dev. Biol. 161, 563–596. Tepass, U., Fessler, L., Aziz, A., Hartenstein, V., 1994. Embryonic origin of hemocytes and their relationship to cell death in Drosophila. Development 120, 1829–1837. Tevosian, S.G., Deconinck, A.E., Cantor, A.B., Rieff, H.I., Fujiwara, Y., Corfas, G., Orkin, S.H., 1999. FOG-2: A novel GATA-family cofactor related to multitype zinc-finger proteins Friend of GATA-1 and U-shaped. Proc. Natl. Acad. Sci. USA 96, 950–955. Thisse, B., el Messal, M., Perrin-Schmitt, F., 1987. The twist gene: isolation of a Drosophila zygotic gene necessary for the establishment of dorsoventral pattern. Nucl. Acids Res. 15, 3439–3453. Tögel, M., Pass, G., Paululat, A., 2008. The Drosophila wing hearts originate from pericardial cells and are essential for wing maturation. Dev. Biol. 318, 29–37. Towbin, J.A., 2004. Molecular genetic basis of sudden cardiac death. Pediatr. Clin. North Am. 51, 1229–1255. van den Heuvel, M., Nusse, R., Johnston, P., Lawrence, P., 1989. Distribution of the wingless gene product in Drosophila embryos; a protein involved in cell–cell communication. Cell 59, 739–749. van Rooij, E., Sutherland, L.B., Qi, X., Richardson, J.A., Hill, J., Olson, E.N., 2007. Control of stress-dependent cardiac growth and gene expression by a microRNA. Science 316, 575–579. Venkatesh, T.V., Park, M., Ocorr, K., Nemaceck, J., Golden, K., Wemple, M., Bodmer, R., 2000. Cardiac enhancer activity of the homeobox gene tinman depends on CREB consensus binding sites in Drosophila. Genesis 26, 55–66. Verschueren, K., Remacle, J.E., Collart, C., Kraft, H., Baker, B.S., Tylzanowski, P., Nelles, L., Wuytens, G., Su, M.T., Bodmer, R., Smith, J.C., Huylebroeck, D., 1999. SIP1, a novel zinc finger/ homeodomain repressor, interacts with Smad proteins and binds to 5-CACCT sequences in candidate target gene. J. Biol. Chem. 274, 20489–20498. Vincent, S., Wilson, R., Coelho, C., Affolter, M., Leptin, M., 1998. The Drosophila protein Dof is specifically required for FGF signaling. Mol. Cell 2, 515–525. Wang, D., Chang, P.S., Wang, Z., Sutherland, L., Richardson, J.A., Small, E., Krieg, P.A., Olson, E.N., 2001. Activation of cardiac gene expression by myocardin, a transcriptional cofactor for serum response factor. Cell 105, 851–862. Wang, J., Tao, Y., Reim, I., Gajewski, K., Frasch, M., Schulz, R.A., 2005. Expression, regulation, and requirement of the Toll transmembrane protein during dorsal vessel formation in Drosophila. Mol. Cell Biol. 25, 4200–4210.

86

Ward, E.J., Skeath, J.B., 2000. Characterization of a novel subset of cardiac cells and their progenitors in the Drosophila embryo. Development 127, 4959–4969. Wasserthal, L.T., 2007. Drosophila flies combine periodic heartbeat reversal with a circulation in the anterior body mediated by a newly discovered anterior pair of ostial valves and “venous” channels. J. Exp. Biol. 210, 3707–3719. Weavers, H., Prieto-Sanchez, S., Grawe, F., Garcia-Lopez, A., Artero, R., Wilsch-Brauninger, M., Ruiz-Gomez, M., Skaer, H., Denholm, B., 2009. The insect nephrocyte is a podocyte-like cell with a filtration slit diaphragm. Nature 457 (7227), 322–326. Wen, H., Weiger, T.M., Ferguson, T.S., Shahidullah, M., Scott, S.S., Levitan, I.B., 2005. A Drosophila KCNQ channel essential for early embryonic development. J. Neurosci. 25, 10147–10156. Wessells, R.J., Bodmer, R., 2004. Screening assays for heart function mutants in Drosophila. Biotechniques 37, 58–60, 62, 64. Wessells, R.J., Bodmer, R., 2007. Cardiac aging. Semin. Cell Dev. Biol. 18, 111–116. Wessells, R.J., Fitzgerald, E., Cypser, J.R., Tatar, M., Bodmer, R., 2004. Insulin regulation of heart function in aging fruit flies. Nat. Genet. 36, 1275–1281. Wessells, R.J., Fitzgerald, E., Piazza, N., Ocorr, K., Morley, S., Davies, C., Lim, H.-Y., Elmén, L., Hayes, M., Oldham, S., Bodmer, R., 2009. d4eBP acts downstream of both dTOR and dFOXO to modulate cardiac functional aging in Drosophila. Aging Cell 8, 542–552. Wilson, R., Vogelsang, E., Leptin, M., 2005. FGF signalling and the mechanism of mesoderm spreading in Drosophila embryos. Development 132, 491–501. Wolf, M.J., Amrein, H., Izatt, J.A., Choma, M.A., Reedy, M.C., Rockman, H.A., 2006. Drosophila as a model for the identification of genes causing adult human heart disease. Proc. Natl. Acad. Sci. USA 103, 1394–1399. Wu, X., Golden, K., Bodmer, R., 1995. Heart development in Drosophila requires the segment polarity gene wingless. Dev. Biol. 169, 619–628. Xu, X., Yin, Z., Hudson, J., Ferguson, E., Frasch, M., 1998. Smad proteins act in combination with synergistic and antagonistic regulators to target Dpp responses to the Drosophila mesoderm. Genes Dev. 12, 2354–2370. Yarnitzky, T., Volk, T., 1995. Laminin is required for heart, somatic muscles, and gut development in the Drosophila embryo. Dev. Biol. 169, 609–618. Yi, P., Han, Z., Li, X., Olson, E.N., 2006. The mevalonate pathway controls heart formation in Drosophila by isoprenylation of Ggamma1. Science 313, 1301–1303. Yi, P., Johnson, A.N., Han, Z., Wu, J., Olson, E.N., 2008. Heterotrimeric G proteins regulate a noncanonical function of septate junction

PART  |  1  Heart Evolution

proteins to maintain cardiac integrity in Drosophila. Dev. Cell 15, 704–713. Yin, Z., Frasch, M., 1998. Regulation and function of tinman during dorsal mesoderm induction and heart specification in Drosophila. Dev. Genet. 22, 187–200. Yin, Z., Xu, X.-L., Frasch, M., 1997. Regulation of the Twist target gene tinman by modular cis-regulatory elements during early mesoderm development. Development 124, 4871–4982. Zaffran, S., Astier, M., Gratecos, D., Guillen, A., Semeriva, M., 1995. Cellular interactions during heart morphogenesis in the Drosophila embryo. Biol. Cell 84, 13–24. Zaffran, S., Küchler, A., Lee, H.H., Frasch, M., 2001. biniou (FoxF), a central component in a regulatory network controlling visceral mesoderm development and midgut morphogenesis in Drosophila. Genes Dev. 15, 2900–2915. Zaffran, S., Frasch, M., 2005. The homeodomain of Tinman mediates homo- and heterodimerization of NK proteins. Biochem. Biophys. Res. Commun. 334, 361–369. Zaffran, S., Xu, X., Lo, P.C., Lee, H.H., Frasch, M., 2002. Cardiogenesis in the Drosophila model: control mechanisms during early induction and diversification of cardiac progenitors. Cold Spring Harb. Symp. Quant. Biol. 67, 1–12. Zaffran, S., Reim, I., Qian, L., Lo, P.C., Bodmer, R., Frasch, M., 2006. Cardioblast-intrinsic Tinman activity controls proper diversification and differentiation of myocardial cells in Drosophila. Development 133, 4073–4083. Zamora, M., Männer, J., Ruiz-Lozano, P., 2007. Epicardium-derived progenitor cells require beta-catenin for coronary artery formation. Proc. Natl. Acad. Sci. USA 46, 18109–18114. Zeitlinger, J., Zinzen, R.P., Stark, A., Kellis, M., Zhang, H., Young, R. A., Levine, M., 2007. Whole-genome ChIP-chip analysis of Dorsal, Twist, and Snail suggests integration of diverse patterning processes in the Drosophila embryo. Genes Dev. 21, 385–390. Zeitouni, B., Senatore, S., Severac, D., Aknin, C., Semeriva, M., Perrin, L., 2007. Signalling pathways involved in adult heart formation revealed by gene expression profiling in Drosophila. PLoS Genet. 3, 1907–1921. Zhao, Y., Samal, E., Srivastava, D., 2005. Serum response factor regulates a muscle-specific microRNA that targets Hand2 during cardiogenesis. Nature 436, 214–220. Zikova, M., Da Ponte, J.P., Dastugue, B., Jagla, K., 2003. Patterning of the cardiac outflow region in Drosophila. Proc. Natl. Acad. Sci. USA 100, 12189–12194. Zmojdzian, M., Da Ponte, J.P., Jagla, K., 2008. Cellular components and signals required for the cardiac outflow tract assembly in Drosophila. Proc. Natl. Acad. Sci. USA 105, 2475–2480.

Chapter 1.3

Cardiac Development in the Frog Mark Mercola, Rosa M. Guzzo and Ann C. Foley Burnham Institute for Medical Research, La Jolla, CA, USA

I.  Introduction Amphibians have a long and prominent history as an experimental organism for embryology research because their general availability, the large size of their embryos and the relative ease of explant culture make them ideal for both microsurgical manipulation and gene and protein misexpression studies. As experimental embryologists realize, any given question can often be best addressed using a particular species. The relative ease of manipulating early amphibian embryos makes them particularly well-suited for deciphering the early mechanisms that specify and commit cells to cardiac lineages; thus, amphibians have been, and continue to be, instrumental for elucidating the signaling relationships between tissues that induce heart. Equally important, the rich knowledge of early amphibian development provides a detailed context against which we are able to understand cardiogenesis. Beginning in the 1920s, first with urodeles and later with anurans, experimental embryologists mapped the regions of the embryo fated to form the heart and this pioneering research was the foundation for subsequent studies that identified the tissue sources of the inducing factors and, eventually, the signaling pathways and molecules themselves. Importantly, the components of the molecular signaling pathways that regulate amphibian cardiogenesis have turned out to be evolutionarily conserved in other species, such that they regulate heart formation in animals with divergent developmental strategies, including those of birds and mammals. The knowledge of the mechanisms and signaling pathways that regulate cardiogenesis has contributed to the theoretical framework now being used to devise methods for directing the differentiation of cardiac cell types from embryonic and other stem cells. This chapter, therefore, focuses on the long heritage of classical and molecular genetic studies Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

of amphibian heart development and discusses how the inductive, patterning and morphogenetic processes are part of the broader mechanisms that establish the body plan, including axial pattern, germ layer and organ specification, and morphogenesis.

II.  The heart-forming region in   the early embryo As in all vertebrates, the amphibian heart originates from paired primordia located on either side of the dorsal midline. Amphibian embryos develop as a ball of cells with a blastocoel cavity that forms in the animal hemisphere at blastula stages. Mesodermal cells are located equatorially, and heart primordia in the frog Xenopus laevis have been mapped to deep cells located between 30° and 45° in the equatorial (marginal) zone to either side of dorsal midline at the onset of gastrulation (Keller, 1976; Sater and Jacobson, 1989) (Fig. 1), consistent with earlier fate and specification mapping experiments in urodeles (Jacobson, 1961; Jacobson and Duncan, 1968). Earlier, at the 32-cell stage, the stereotypic Xenopus embryo consists of four tiers (A–D, with A being most animal and D most vegetal) of eight cells each. The four blastomeres on each side are numbered, with 1 being most dorsoanterior and 4 most ventroposterior (Fig. 1A, see Dale and Slack (1987) for a complete fate map of the 32-cell stage Xenopus laevis embryo). Although by convention the meridians that define the embryo’s plane of bilateral symmetry are referred to as the dorsal and ventral midlines at this stage of development, it should be noted that these terms are inaccurate since the anteroposterior and dorsoventral body axes are not yet orthogonal (Lane and Sheets, 2000), and for this reason we refer in this review to the cleavage stage axes as animal–vegetal and dorsoanterior–ventroposterior, respectively. The heart primordia derive primarily 87

88

PART  |  1  Heart Evolution

from the equatorial region spanning approximately 30–45° on either side of the dorsal midline, corresponding to the upper (animal) and lateral portion of the C1 blastomere and extending into the adjoining region of C2 (Fig. 1A). The spatial location of blastomeres and hence the fate map remain constant during early cleavage; thus, microinjection of mRNA or other molecules into this region at the 8–32-cell stage embryo effectively targets delivery to the developing heart.

(A)

As in all vertebrates, the tissue movements of gastrulation establish the shape of the embryo. In amphibians this is initiated by extensive cell intercalation movements that drive convergent-extension tissue rearrangements and cause the equatorial and vegetal regions of the late blastula to involute, and eventually lead to a pronounced elongation along the eventual anteroposterior body axis of the tadpole (Keller and Winklbauer, 1992). Gastrulation begins at stage 10 (Nieuwkoop and Faber, 1994) with involution of

(B)

Ventroposterior

Involution of mesoderm during gastrulation

Dorsoanterior

A

ht

B C1 C

ht

ht

ht

D 32-cell

(C)

(D)

Ventroposterior

Dorsoanterior

Nodal

Nodal antagonism (cerberus)

ht en

FGF BMP antagonism (cerberus)

so ht

Vegetal view at onset of gastrulation, stage 10

Canonical wnt/β–catenin signaling

Mesoderm

Wnt antagonism (e.g. Dkk1, crescent, Frzb)

BMP Non-canonical wnts (e.g. Wnt11) Canonical wnts (e.g. Wnt3)

Hex-dependent endodermal protein Dorsoanterior mesoderm Wnt/β–catenn, BMP

Cardiogenic mesoderm Nkx2.5+

Cardiac mesoderm Nkx2.5+, αMHC+

Figure 1  Heart induction in Xenopus laevis. The heart primordia originate on either side of the dorsal midline. At the 32-cell stage (A), the fated heart region (red, ht) corresponds primarily to the C1 blastomere. The tissue movements of gastrulation (B) cause the mesoderm mantle to involute and move anteriorly. The heart region (ht) is located at the vegetal region of the mesoderm annulus and therefore involutes early. Heart induction occurs in Xenopus laevis beginning at the onset of gastrulation (C, vegetal pole view) and important sources of inducing molecules include the dorsoanterior region of the deep endoderm (en, yellow), fated to give rise to the liver, pancreas and lung, and from Spemann’s Organizer (so, blue) that gives rise to notochord and pre-chordal mesoderm. (D) Diffusible signals that induce the heart. Early Wnt signaling is important in establishing dorsoanterior organizing centers in the embryo, including Spemann’s Organizer. At the onset of gastrulation, Wnt antagonists (e.g. Dkk-1, Crescent and Frz-b) and Nodal-related proteins are produced on the dorsoanterior region of the embryo. These signals act on deep endoderm to initiate a signaling cascade that ultimately provides diffusible factors, including Cerberus, that initiate cardiogenesis in mesoderm. Non-canonical Wnts are also likely to be involved either through inhibition of canonical Wnt signaling or a b-catenin-independent pathways. Other proteins, in particular BMP, operate subsequently to maintain the cardiac field.

89

Chapter  |  1.3  Cardiac Development in the Frog

(A)

(B)

region of the tailbud-stage embryo based on the ability of mesodermal tissue from this region to form hearts when transplanted to ectopic sites in the embryo (Ekman, 1925). Although these studies accurately identified the heart field, the conclusion that its myocardial fate is determined at this time was uncertain since the grafted regions also included pharyngeal endoderm that might have contributed heartinducing signals. More recent data have clarified the signaling mechanisms that specify myocardial fate and apportion myocardial versus nonmyocardial lineages in the heart field. These signals, discussed below and diagrammed in Fig. 1D, act during the late tailbud stages as the Nkx2.5 sheet of cardiac mesoderm buckles and descends into the enlarging pericardial cavity to form the heart tube, dorsal mesocardium and roof of the pericardium.

III.  Sources of heart-inducing signals

(C)

(D)

Figure 2  Early cardiac gene expression in late tailbud-stage embryos. Lateral and ventral views of whole Xenopsu laevis embryos (stage 26–27) after in situ hybridization show the characteristic bilateral expression of Nkx2.5 (A) and Mlc2a (B) mRNA. Nkx2.5 marks the heart field in the mesoderm, of which the ventral region expresses myocardial markers such as Mlc2a. In histological section (C, D), Nkx2.5 expression extends up to the pharyngeal arch mesoderm, whereas Mlc2a expression is more ventrally delimited (see text). The bilateral expression domains are separated initially at the ventral midline of the tailbud-stage embryo presumably by cardiac cells that do not express these genes. Images are reproduced from Mohun et al. (2000) with permission.

the dorsoanterior-most equatorial and vegetal regions, and progresses laterally so that the entire equatorial and vegetal region eventually ingresses at a blastopore lip that by stage 11 circumscribes the embryo (Keller and Winklbauer, 1992). The heart primordia and underlying endoderm involute together shortly after gastrulation begins (Fig. 1B), and they migrate anteriorly with the advancing archenteron cavity. The two heart primordia then migrate ventrally until they join at the ventral midline (marked by Nkx2.5 expression in Fig. 2A,C). At this time in development (stages 26– 27 in Xenopus), the heart field is a thin layer of mesoderm situated between the surface ectoderm and pharyngeal endoderm. It is located posterior to the cement gland and buccopharyngeal plate and extends laterally from the dorsal midline approximately 90° where it abuts pharyngeal arch mesoderm. The region that forms the heart muscle, or myocardium (marked by Mlc2a expression in Fig. 2B,D) develops in the ventral portion of the Nkx2.5 domain. Classical experiments mapped the heart field to a comparable

Heart tissue is not specified until after the onset of gastrulation. Jacobson and colleagues (Jacobson, 1961; Jacobson and Duncan, 1968) using urodele amphibians showed that early gastrula-stage cardiac primordia do not form heart tissue when explanted and placed into dilute saline culture medium and, therefore, are not specified at this point. Sater and Jacobson later confirmed this observation in Xenopus laevis, and went on to show that inclusion of a region of midline mesoderm located between the primordia (15° from the dorsal midline in the equatorial region) stimulated the incidence of heart formation in the explants (Sater and Jacobson, 1989, 1990b). This dorsoanterior marginal zone (DMZ) region gives rise to notochord and head mesoderm, and importantly includes the early gastrula-stage blastopore lip tissue known as Spemann’s Organizer. The blastopore lip in cross-section resembles a standing wave, and is the point where the sheet of mesendodermal cells folds inward and moves into the embryo during gastrulation. Therefore, different tissue progenitor populations are present at different times within the lip. At the onset of gastrulation (stage 10), the tissue present at the lip has the remarkable ability, first shown by Hilde Mangold and Hans Spemann, to induce the formation of an ectopic body axis, including heart tissue, when transplanted to the opposite side of recipient embryos (reviewed in Harland and Gerhart, 1997). It shares signaling properties with analogous dorsal organizing centers in the chick (Hensen’s node) (Waddington, 1932) and mouse (node) (Beddington, 1994). Classical experiments performed with cultured chick and mouse embryos had suggested that hypoblast and definitive endoderm also produce heart-inducing signals (Hommes, 1957; Orts-Llorca, 1963; Orts-Llorca and Gil, 1965), and early experiments in several urodele amphibians suggested an inductive role for the deep endoderm that gives rise to pharyngeal endoderm in these species

90

(Balinsky, 1939; Nieuwkoop, 1947; Chuang and Tseng, 1957; Jacobson, 1960, 1961). In Xenopus laevis, however, the walls and roof of the pharynx derive from a superficial layer of cells surrounding the marginal zone of the embryo known as the suprablastoporal endoderm that early experiments indicated is not required for heart induction (Sater and Jacobson, 1990b). Subsequent experiments demonstrated that the deep dorsoanterior endoderm (which forms the ventral cells of the pharynx), as well as the DMZ region containing Spemann’s Organizer, are both necessary for heart induction, and showed that both together are sufficient to induce heart development after grafting to noncardiac mesoderm (Nascone and Mercola, 1995), suggestive of at least two separate activities (Fig. 1C). The sequential inductions in response to DMZ and deep dorsoanterior endoderm tissue are largely complete in Xenopus laevis by stage 10.5, since removal of these tissues after that point has minimal effect on the incidence of hearts that form in isolated explants. Considerable evidence shows that the inducing properties in the region of the late-blastula/early-gastrula-stage Organizer and adjacent tissue are not uniformly distribu­ ted; rather, distinct trunk, head and heart-inducing activities are localized within different regions (Harland and Gerhart, 1997; Zoltewicz and Gerhart, 1997). Head-inducing activity is strongest within the portion of the blastopore lip region that during gastrulation will become anteriorly located and which spans approximately 15° from the dorsal midline at the onset of gastrulation. Both superficial cells which will give rise to pharyngeal endoderm and deeper prechordal mesoderm cells have inducing activity (Shih and Keller, 1992). Trunk-organizing activity is localized more animally, involutes later during gastrulation and gives rise to more posterior tissue. Experiments extirpating the head and trunk organizer separately in Xenopus laevis showed that heart-inducing activity is localized within the head-inducing region and declines animally in the trunk organizer (Schneider and Mercola, 1999). In terms of signaling properties, the signals produced by the Xenopus head organizer might resemble those of anterior visceral endoderm (AVE) of the mouse with which it shares a profile of gene expression. Mouse embryos lacking Cerberus1, Nodal, Otx2 or Lim1 in the AVE show anterior truncations suggestive that this tissue is needed to generate pattern anterior to the mid-hindbrain (Shawlot and Behringer, 1995; Ang et al., 1996; Varlet et al., 1997; Acampora et al., 1998; Rhinn et al., 1998; Shawlot et al., 1998). A distinct heart-inducing activity is present in the deep dorsoanterior endoderm underlying the head organizer region of stage 10–10.5 Xenopus laevis embryos. This activity resides within a region that extends about 45° to either side of the dorsal midline (Schneider and Mercola, 1999), and coincides with the mRNA expression domain of Cerberus, a member of the DAN (differential screeningselected gene aberrative in neuroblastoma) family of

PART  |  1  Heart Evolution

secreted proteins that have the ability to bind and inhibit signaling from certain TGF family proteins, including bone morphogenetic protein isoforms (Dionne et al., 2001). Xenopus Cerberus inhibits bone morphogenetic proteins and Nodal, as well as Wnt activities (Piccolo et al., 1999). Although strongly heart-inducing, the deep endoderm is clearly not required for head induction (Schneider and Mercola, 1999) despite the potent ability of Cerberus to induce ectopic head structures when misexpressed in ventroposterior marginal zone (VMZ) blastomeres (Bouwmeester et al., 1996). Within the DMZ, the region of heart-inducing activity has not been tightly localized, but overlaps at least with that responsible for head-induction (Schneider and Mercola, 1999). Taken together, therefore, the microsurgical ablation and explant studies demonstrated heart-inducing activity within at least two regions, the superficial DMZ overlapping the region of head-inducing activity in Spemann’s Organizer and the deep dorsoanterior endoderm that underlies the Organizer and heart fields (Fig. 1C). The involvement of two spatially-separated tissues first suggested that multiple signaling pathways mediate heart induction, and this has been borne out by the molecular genetic studies discussed below.

IV.  Inhibitory signals and the concept of a cardiac field As discussed above, Ekman’s classical experiment (Ekman, 1925) using newt embryos showed that tissue located just dorsal to the portion of the heart field fated to form myocardium is capable of forming beating heart tubes if transplanted to ectopic sites in the embryo, where it would be distanced from inhibitory signals. Similar experiments in chick embryos (Rawles, 1936; DeHaan, 1965) generalized the conclusion that a broad region of heart potency exists in early embryos. These and other (see below) studies indicated that mesodermal cells located outside the fated myocardial region are transiently specified to form myocardial structures, but that they become respecified during normal development (although the relative location of this tissue is the same, it is described as lateral to the fated myocardial region in amphibians and medial to it in amniotes, because the terms apply to a spherical frog versus a flat chick embryo). Lineage labeling of the lateral portion of the Xenopus laevis heart field, as classically identified by transplantation and explant studies, demonstrated that this region normally gives rise to the dorsal mesocardium and pericardial roof (Raffin et al., 2000). The myocardial tube, in contrast, arises from the ventral portion of the heart field (compare Fig. 2A,C with Fig. 2B,D). The Nkx2.5 expression domain overlaps both the ventral and lateral regions of the classically-defined heart field. These lineage studies refined the earlier view that the heart field itself becomes restricted during development (as in Sater and Jacobson, 1990a), indicating instead

91

Chapter  |  1.3  Cardiac Development in the Frog

that the sheet of Nkx2.5 cells does not diminish in size, but rather shifts ventrally and folds to form the myocardial tube, dorsal mesocardium and pericardial roof. Inhibitory signals, therefore, function to distinguish myocardial from dorsal mesocardial and pericardial roof fate. What are the tissues and signals that delimit the extent of the myocardial domain? Explant culture and tissue recombination studies in Xenopus laevis showed that the neural plate provides inhibitory signals (Raffin et al., 2000; Garriock and Drysdale, 2003). It is intriguing to speculate that the neural crest cells migrating into the branchial arch region contribute to repression of cardiomyogenesis, and therefore correct patterning of dorsal mesocardial and pericardial roof tissue. As in amniotes, ablation of the Xslug-expressing lateral neural fold tissue (comprising putative cardiac neural crest cells) alters outflow and inflow tract development, and as well reduces the level of the transcription factor Id2, which is a repressor of cardiogenic bHLH factors (Martinsen et al., 2004). The ventral portion of the heart field, which normally forms myocardium, also inhibits myocardial differentiation in the lateral region in Xenopus laevis (Raffin et al., 2000). This mechanism probably accounts for the earlier experimental results of Copenhaver (1926), who ablated the fated myocardial region of salamander (Ambystoma punctatum) embryos and found that beating hearts nonetheless arose from tissue located just dorsal to the site of ablation. Little information exists regarding the factors that suppress cardiomyogenesis in the lateral (medial in chick) regions of the heart field. Chick experiments showing that Wnt secretion by neural plate ectoderm inhibits myocardial differentiation in paraxial mesoderm, while the secreted Wnt antagonist Crescent produced by the endoderm underneath the heart field supports myocardial development (Tzahor and Lassar, 2001), suggesting that Wnt could be such an inhibitory signal. This mechanism was postulated to position the medial border of the Nkx2.5 expression domain in chick, but whether it might also apportion myocardial versus nonmyocardial fate within the heart field is unclear. Within the Xenopus heart field, Notch1 signaling in response to Jagged1 has been shown to suppress myogenesis and promote nonmyocardial differentiation in the lateral heart field region, such that experimental suppression of Notch enlarges the myocardial domain (Rones et al., 2000). Interestingly, Notch-mediated suppression of cardiomyocyte differentiation was first noted in Drosophila, where it also serves to select between myocytic and nonmyocytic lineages (Hartenstein et al., 1992), suggesting evolutionary conservation of action. This model might also apply to mammals since targeted disruption of RBPJ, which mediates responsiveness from all Notch receptors, in embryonic stem cells (ESCs) enhanced cardiogenesis relative to differentiation in wild-type, parental embryonic stem cells (Schroeder et al., 2003). Since extra-myocardial portions of the Xenopus heart field contain progenitors that are transiently specified

as myocardial, it is tempting to draw a connection to the secondary or anterior heart field of higher vertebrates that gives rise to the outflow tract and right ventricle somewhat later than when the primary heart field forms the heart tube (Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001; Cai et al., 2003). The evolutionary origin of the amniote secondary heart field is currently unknown. It is reasonable to speculate that the ability of precursors outside of the primary heart field to contribute new units of cardiac structure and function to the lower vertebrate heart arose by recruiting cells, such as those in lateral or anterior domains of the Xenopus heart field (see Brade et al., 2007 and below) whose potential for myocardial differentiation existed but was held in check by developmental signals. Thus, the secondary heart field might have originated in a population of progenitors by altering patterns of inhibitory gene activity (see below).

V.  Signaling pathways that induce heart development V.A.  Early Wnt Signaling Establishes Dorsoanterior Mesoderm As described above, the amphibian heart forms from dorsoanterior mesoderm, and the inducing tissues are dorsoanterior midline mesoderm and deep dorsoanterior endo­­ derm. Consequently, the first embryological signals that are required for heart formation are those that specify dorsoanterior mesendoderm as distinct from ventroposterior mesendoderm. The unfertilized amphibian egg is considered to be radially symmetrical around the animal–vegetal axis (reviewed in Harland and Gerhart, 1997; De Robertis and Kuroda, 2004). Sperm enters in the animal hemisphere and triggers cortical microtubules to align parallel to the meridian where entry occurred. Prior to the first cell division, molecular motors along the microtubule network drive the egg cortex (about 4 m thick) to rotate approximately 30° relative to cytoplasm. Cortical rotation leads to the stabilization of -catenin in a region of the vegetal hemisphere that is 180° opposite that of sperm entry and the meridian linking the sperm entry site with this domain will become the dorsal midline of the embryo. Cleavage of the fertilized egg ensues, and large-scale transcription is triggered at about the twelfth cell division, a point known as the midblastula transition (MBT). Beginning at this time, the nuclear-localized -catenin in dorsoanterior blastomeres activates transcription of genes that are responsible for the eventual formation of Spemann’s Organizer in mesendoderm at the onset of gastrulation, by which time ongoing division has given rise to approximately 10–50,000 cells. Until recently, experiments have failed to identify a Wnt that triggered the pathway, leading to speculation that it is activated cytoplasmically.

92

Recently, however, Tao et al. (2005) showed that Wnt11, which in many settings activates the -catenin-independent noncanonical signaling, cooperates with the glycosyl transferase X.ETX1 and the EGF-CFC protein FRL1 to stabilize -catenin. A cascade of -catenin-dependent gene activation ensues in the dorsoanterior mesendoderm, involving Siamois and Goosecoid, to control the spatio– temporal expression patterns of signaling proteins and additional transcription factors, including members of the Sox family (Zhang et al., 2005), to initiate cardiogenesis through the activity of the secreted proteins described below and summarized in Fig. 1D.

V.B.  Specification of the Heart Field by   Wnt Antagonists, TGF family members   and Cerberus Although activation of the Wnt/-catenin pathway initiates dorsoanterior development, specific Wnt antagonists are needed at the onset of gastrulation in order to induce cardiogenesis, as marked by expression of the heart field markers Nkx2.5, Tbx5 and GATA4. This apparent paradox can be explained because although -catenin-dependent transcription establishes dorsoanterior tissues at mid- to late-blastula stages, it then operates as a potent ventroposteriorizing signal during pregastrula to gastrula stages. The effect of timing is seen when comparing differences in effects of mRNA versus plasmid misexpression of Wnt agonists (Christian and Moon, 1993). Directed expression of Wnt proteins, -catenin or other activators of the pathway by microinjection of mRNA, which is translated immediately to yield protein, into cleavage stage embryos will dorsoanteriorize the tissue, but injection of DNA plasmid, which is transcribed only after midblastula transition, will ventroposteriorize the embryo. Considerable evidence indicates that canonical Wnt/-catenin signaling in the pregastrula- and gastrula-stage embryo is highest in the ventroposterior portion, while an essential activity of the dorsoanterior organizer region is to produce antagonists that block this activity. Screening known secreted factors from the Organizer region by misexpression in the noncardiogenic VMZ tissue, cultured in isolation, revealed the remarkable ability of the secreted Wnt antagonist Dickkopf-1 (Dkk1) and, to a lesser extent the secreted Frizzled-like protein Crescent, to induce beating heart tubes (Schneider and Mercola, 2001). Analogous experiments in the chick showed the same result (Marvin et al., 2001). Heart induction also requires XDbf4 (a Xenopus homolog of the yeast Dumbell former 4) that, in addition to regulating cell-cycle, inhibits Wnt signaling through its interactions with Frodo, which is itself a component of the Wnt pathway acting at the level of binding to chromatin and the transcription factor TCF (Brott and Sokol, 2005). Cell-cycle interacting domains of the XDbf4 protein are not relevant to heart induction,

PART  |  1  Heart Evolution

indicating that it is the interaction with the Wnt pathway that is important. Dkk1, which binds to the low density lipoproteins 5 and 6 to disrupt Wnt signaling, and other Wnt antagonists such as Crescent that bind and sequester Wnt, are therefore likely to constitute part of the Organizer region’s heart-inducing activity. The classical studies, however, indicated that neither the Organizer region nor underlying deep endoderm alone is sufficient to induce heart, but that both are required. Part of the explanation why antagonists such as Dkk1 alone are sufficient, yet the Organizer region is not, is that Wnt antagonism acts by stimulating the endoderm to produce diffusible signals that in turn initiate cardiogenesis in mesoderm. A key mediator in this molecular relay is the homeodomain protein Hex, which is expressed in deep dorsoanterior endoderm where it is important for the differentiation of anterior endodermal derivatives (Smithers and Jones, 2002) and operates downstream of at least Dkk1 to induce heart (Foley and Mercola, 2005). The cardiogenic role of Hex is as a transcriptional repressor, but the relevant downstream targets are not yet known. In mice, Wnt/-catenin control of endodermally-derived diffusible molecules is supported by the appearance of multiple hearts in embryos having targeted disruption of the -catenin gene in definitive endoderm (Lickert et al., 2002). Hex could be a mediator, since its targeted disruption affects heart morphogenesis (Martinez Barbera et al., 2000), but it has not been possible to distinguish whether the heart defects arise as a consequence of improper heart induction versus vascular anomalies in these animals. Recently, a genetic hierarchy in endodermal cells of murine embryonic stem cells involving Sox17 and Hex has been shown to regulate cardiomyocyte induction (Liu et al., 2007) indicating that the cardiogenic function of Hex is conserved in mammals. mRNA encoding Nodal family members, in particular Xenopus Nodal related-1 (XNr-1), are expressed in the Organizer region overlapping that of Dkk1 and Crescent. XNr-1 is a member of the Nodal family, which is part of the transforming growth factor- (TGF) superfamily. Nodal proteins have been implicated as potential heart inducers, not only as a consequence of their role in the induction of axial mesoderm in Xenopus (Jones et al., 1995; Agius et al., 2000), zebrafish (Toyama et al., 1995; Feldman et al., 1998), mouse (Zhou et al., 1993; Conlon et al., 1994) and chick (Bertocchini and Stern, 2002; Bertocchini et al., 2004), but also because they are essential for the induction and patterning of the endoderm (Henry et al., 1996; Alexander and Stainier, 1999; Agius et al., 2000; Chang et al., 2000; David and Rosa, 2001), which, as discussed above, is important for heart induction. In addition, several members of the Xenopus Nodal-related family, including XNr1 (Reissmann et al., 2001), the closely-related TGF family member activin (Logan and Mohun, 1993; Mangiacapra et al., 1995; Yatskievych et al., 1997; Ladd

Chapter  |  1.3  Cardiac Development in the Frog

et al., 1998), and a constitutively active form of the Nodal receptor, Alk4 (Takahashi et al., 2000) are each capable of inducing cardiac tissue in Xenopus and chick explant cultures. A critical role for Nodal-dependent signaling in vertebrate mesendoderm development has also been confirmed by homozygous deletion of the gene encoding the Nodal co-receptor Cripto in mice, which results in embryonic lethality at day 7.5 due to lack of primitive streak formation, absence of a node, and defects in embryonic mesoderm formation (Ding et al., 1998; Xu et al., 1999; Liguori et al., 2003). These studies also revealed that Cripto/ mice display severe defects in cardiogenesis, while exhibiting precocious differentiation of anterior neuroectoderm (Ding et al., 1998). As for Dkk1, a major challenge for elucidating the role of Nodal proteins has been to trace the downstream signaling paths that pertain to heart induction, perhaps distinguishing them from paths that control aspects of dorsoanterior patterning. We have recently shown through misexpression and morpholino knockdown studies that the heart-inducing activity of XNr-1 functions through Cerberus in the endoderm (Foley et al., 2007). Xenopus Cerberus is a multifunctional antagonist that binds and blocks Wnt, Bmp and Nodal proteins (Bouwmeester et al., 1996; Piccolo et al., 1999) (murine Cerberus-like blocks only bone morphogenetic proteins (BMP) and Wnt), and it is Cerberus expression that has been shown to correspond precisely, spatially and temporally, with heart-inducing activity in deep endoderm (Schneider and Mercola, 1999). A truncated version of Cerberus, Cerberus-short (Cer-S), which only inhibits Nodal proteins, and not BMP and Wnt, was able to induce heart tissue in conjunction with a BMPantagonist. This implicated feedback inhibition of XNr-1 as one essential activity of Cerberus in heart induction (Foley et al., 2007). Interestingly, induction of Cerberus by ectopic XNr-1 shows temporal and spatial restriction, possibly reflecting feedback function (Osada and Wright, 1999; Piccolo et al., 1999; Yamamoto et al., 2003). Thus, Cerberus might block Nodal, BMP and Wnt signaling in the heart field at a time when these signals would otherwise inhibit heart induction. Although the combination of early Wnt/-catenin signaling with subsequent Wnt antagonism (Dkk1, Crescent, or XDbf4) and Nodal signaling plus feedback (XNr-1, Cripto and Cerberus) is sufficient to direct differentiation of uncommitted mesoderm to cardiac mesoderm (at which point it expresses markers such as Nkx2.5 and Tbx5), additional signals are certain to cooperate with Wnt and Nodal pathways in complex ways. For instance, fibroblast growth factor (FGF) isoforms, which are capable of inducing and patterning mesoderm, might positively influence heart field specification. The influence of FGF on amphibian heart induction has not been examined thoroughly, apart from its role in mesoderm induction and subsequent patterning during gastrula and neurula stages. FGF isoforms

93

have been studied extensively during the development of amniotes, in particular chicks, and is well-known for induction and patterning of mesoderm and neurectoderm. Relevant to heart induction, FGF is produced by endodermal cells underneath the heart field and localized application to late-streak stage chick embryos can expand the size of the heart field posteriorly in regions of BMP signaling (Lough et al., 1996; Alsan and Schultheiss, 2002) and it might also function similarly in the mouse and chicken secondary heart fields (Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001; Cai et al., 2003). Additionally, the anti-dorsalizing morphogenetic protein-2 (ADMP2), a divergent TGF- family member expressed in the deeper cells of the marginal zone but excluded from the Spemann Organizer, has been recently shown to influence axial patterning and heart development (Kumano et al., 2006). Targeted reduction of ADMP2 levels caused a striking loss of cardiac and ventral blood island structures, but only a mildly-ventralized phenotype in the embryo. The observation that Nkx2.5 expression was eliminated in these studies, together with the ADMP2 expression pattern, suggests that it could act either directly on the cardiogenic mesoderm or more broadly within the DMZ to apportion fate and secondarily influence development of the heart field. Another pathway that is likely to cooperate with the Nodal/Cerberus and antagonists of the canonical Wnt/catenin pathways is the class of so-called noncanonical Wnts, which include Wnt4, Wnt5a and Wnt11. In Xenopus embryos and many cell culture settings, these Wnts regulate intracellular calcium (see Kuhl et al., 2000; Kohn and Moon, 2005) and planar cell polarity and convergent extension movements (reviewed by Mlodzik, 2002), but characteristically do not induce secondary axes on overexpression by microinjection of mRNA into cleavage-stage Xenopus embryos. Gain- and loss-of-function studies in chick and Xenopus have demonstrated that signaling by Wnt11 acts to establish the early heart field (Eisenberg and Eisenberg, 1999; Pandur et al., 2002). A second Wnt11-related gene has been identified in Xenopus (Wnt11-R) that shares signaling properties, but appears not to be involved in heart induction, instead functioning later during heart tube morphogenesis (Garriock, 2005). Since signaling evoked by Wnt11 isoforms can block canonical Wnt/-catenin signaling, this has been proposed as one way in which the noncanonical pathway might stimulate cardiogenesis (Pandur et al., 2002; Maye et al., 2004). Repression of canonical Wnt signaling by Wnt11 might occur by either: (1) intracellular calciummediated repression of -catenin-dependent transcriptional activity (Kuhl et al., 2001; Li and Iyengar, 2002; Maye et al., 2004); or (2) inhibition at the level of the cell surface, possibly via receptor (Frizzled) competition (Maye et al., 2004). Since there is cross-talk between the canonical and noncanonical Wnt signaling pathways, it has been technically

94

difficult to separate these pathways, as activating one inhibits the other and vice versa. Evidence that inhibition of the canonical Wnt signaling pathway serves a primary role in the establishment of the precardiac mesoderm is that blocking Wnt at any level, including down to the transcriptional response, induces gene markers of the early heart field (Foley and Mercola, 2005). A second intriguing possibility involves phosphorylation and activation the Jun kinase (JNK) pathway (Pandur et al., 2002) to stimulate not only differentiation, but also perhaps some aspect of cardiac maturation or tissue morphogenesis. Dkk1 synergizes with the heparin sulfate proteoglycan Kynpek to activate JNK (Caneparo et al., 2007), and might be a link to JNK in cardiogenesis. Interestingly, while Dkk-1 induces the formation of well-organized beating heart tubes in Xenopus laevis, downstream antagonists of the Wnt/-catenin pathway give poorly-organized cardiac tissue that rarely beats and expresses primarily early markers. In addition to Wnt antagonism, which resides in the carboxyl terminal cysteine-rich domain of Dkk1, a second activity that maps to its amino terminal cysteine-rich domain is also important for heart induction (Korol et al., 2008). Taken together, these studies show that inhibition of canonical Wnt/-catenin signaling is critical for early heart induction, and that activation of the noncanonical phospho-JNK pathway, possibly as a novel function of Dkk1, might contribute to cardiomyocyte maturation and possibly heart tube morphogenesis.

V.C.  Bone Morphogenetic Proteins and Progression to Beating Myocardium The preceding factors appear largely responsible for specifying the heart field by stage 10.5, as quantified classically by the propensity to differentiate into beating tissue when cultured in traditional saline medium. This does not indicate that other signals are not required, for they might operate in situ or might become expressed in the explanted tissue, which generally includes other tissues in addition to the developing myocardium. Indeed, other signals are required to sustain the differentiated state and stimulate progression to form myocardium and heart. Bone morphogenetic proteins (BMPs) were first implicated as heart-inducing molecules by the observations that the chick heart field develops in a region having high BMP and low Wnt signaling (Schultheiss et al., 1997; Marvin et al., 2001) and that zebrafish Swirl/bmp2b mutants show reduced or absent expression of Nkx2.5 (Kishimoto et al., 1997). As for many inducing factors, discerning the precise role(s) played by BMPs in heart development has proven difficult because their cardiogenic function has been hard to distinguish from their overall effects on patterning the early embryo, a problem made even more complicated by the early death of mouse embryos possessing homozygous deletions of either the BMP receptor BMPRII,

PART  |  1  Heart Evolution

ALK3/BMPR1A or ALK2/ActRIA that mask the effects of BMP at later stages (Mishina et al., 1995; Gu et al., 1999; Beppu et al., 2000). In chicks, application of pellets of cells expressing BMP or BMP-soaked beads expanded the heart field laterally, whereas similar treatment with the antagonist Noggin contracted the field (Schultheiss et al., 1997; Andree et al., 1998; Schlange et al., 2000; Yamada et al., 2000; Nakajima et al., 2002). Moreover, again from chick studies, BMP can stimulate ectopic cardiogenesis in isolated explants of normally noncardiac mesendoderm that lie just medial to the normal heart field (Schultheiss et al., 1997; Schlange et al., 2000). More posterior mesoderm is refractory (Ladd et al., 1998; Barron et al., 2000), consistent with a prior need for other signals such as Wnt antagonists or possibly FGF. Critically, BMP, like Wnt, is expressed by ventroposterior mesoderm in Xenopus laevis or corresponding posterior tissue in amniotes, where it provides a ventroposteriorizing signal to all three germ layers. Consequently, it will antagonize heart formation if given too early (for instance, see Ladd et al., 1998; Matsui et al., 2005). Conversely, blocking BMP signaling by overexpression of dominant negative forms of downstream effectors such as dominant negative ALK3 (dnALK3), truncated BMP receptor (tBRII) and the inhibitory SMAD6, does not block the initial induction of cardiac markers in Xenopus laevis, but rather prevents progression to a state of terminal differentiation (Shi et al., 2000; Walters et al., 2001). Together these findings indicate that BMP, like Wnt, must be inhibited to induce cardiogenic mesoderm, but is required after the field is specified to maintain the specified state and possibly stimulate differentiation of beating cardiomyocytes. Genetic studies in mice are consistent with such a role for BMP in sustaining heart specification after initial induction. For example, homozygous deletion mutants of BMP2 (Zhang and Bradley, 1996), BMP4 (Winnier et al., 1995), SMAD5 (Chang et al., 1999) or the double BMP 5;7 mutants (Solloway and Robertson, 1999) all specify the heart field correctly, but go on to develop cardiac structural defects. In addition, the specific deletion of the type I serine-threonine kinase receptor ALK3 within the mouse epiblast reveals that BMP signaling is not required for the initial induction of cardiogenic markers but is necessary to maintain cardiac-specific gene expression (Mishina et al., 2002).

V.D.  Transcription Factor Control of   Cardiac Muscle Gene Activity Cardiac transcription factors have been reviewed extensively and their description is beyond the focus of this chapter, and therefore will be discussed only briefly. As in other vertebrates, normal heart development depends on Nkx2 genes, which encode homeodomain transcription factors. The single Nkx2 gene in Drosophila, tinman, is expressed in the

95

Chapter  |  1.3  Cardiac Development in the Frog

dorsal vessel and is essential for cardioblast differentiation, whereas mutation or loss of the Nkx2.5 gene leads to an abnormally-structured heart and myocardial dysfunction in both mice and humans (Schott et al., 1998; Benson et al., 1999; Tanaka et al., 1999). Whether the multiple Nkx2 homologs that are expressed in overlapping but nonidentical patterns in the developing Xenopus cardiac region (Nkx2.3, Nkx2.5 and Nkx2.10) have overlapping cardiogenic function has been investigated by targeted knockdowns and dominant-negative approaches, revealing some degree of functional redundancy, but also distinct requirements for the individual proteins (Chen and Fishman, 1996; Cleaver et al., 1996; Fu et al., 1998; Grow and Krieg, 1998; Allen et al., 2006). The reason for multiple Nkx2 genes is not entirely clear, and although functional differences between Xenopus Nkx2 proteins have been noted (Newman et al., 2000) such differential activity has not been correlated to developmental function. It is tempting to speculate that the different Nkx2 proteins might be involved in regionalization of cardiac mesoderm, similar to the function of Tbx transcription factors (for review, see Plageman and Yutzey, (2005) and Part 9 of this book). Regardless, it is clear that Nkx2 factors are insufficient to specify cardiogenesis, since overexpression causes a limited and spatially-constrained expansion of the field, and ectopic cardiogenesis does not occur. Moreover, the cardiac Nkx2 proteins function in other regions of the embryo, notably pharyngeal endoderm derivatives and tongue, therefore expression of these factors is not an indicator of cardiac fate. Cardiogenic function of Nkx2 proteins involves wellcharacterized interactions with SRF and GATA proteins to induce genes essential for myocardial differentiation (Durocher et al., 1997; Gove et al., 1997; Sepulveda et al., 1998). SRF with its co-factor myocardin and myocardinrelated factors, MRTF-A and MRTF-B, are important regulators of smooth muscle and cardiac genes (Wang et al., 2001, 2002; Small et al., 2005). MRT-A and B orthologs are not present in Xenopus cardiac mesoderm, consequently myocardin appears to be required for heart differentiation. It is not sufficient to elicit a complete muscle developmental program, however, since ectopic expression induces certain muscle genes but not myofibrillar structures (Small et al., 2005). In addition, Nkx2.5 and possibly other cardiogenic transcription factors might be regulated by an unusually small homeodomain-only protein, HOP, that is incapable of binding DNA but is induced by Nkx2.5 and suppresses SRF-dependent transcription of cardiac promoters (Chen et al., 2002; Shin et al., 2002). Studied primarily in mice, it might be part of an Nkx2-dependent mechanism that regulates myocardial gene expression, and as such might operate to suppress Nkx2 function in the heart field and, possibly, at slightly later stages of ventricular myogenesis to promote cell-cycle withdrawal associated with trabeculation and terminal differentiation (e.g., Kochilas et al., 1999; Pasumarthi and Field, 2002).

VI.  Morphogenetic studies of   heart tube formation VI.A.  Formation and Closure of the   Heart Tube The heart tube and pericardium form from the sheet of Nkx2.5, Nkx2.3 cardiac mesoderm at the anterior end of the tadpole between the pharyngeal endoderm and ventral surface ectoderm (Fig. 2C). Histologically, fissures appear separating the thin splanchnic (inner) and somatic (outer) layers in the heart region. The splanchnic layer, which will develop as the myocardium and pericardial roof, delaminates from the outer, somatic portion that will form the pericardium. With time (stages 29–32 in Xenopus laevis), the detaching splanchnic layer becomes raised on both sides, folds and joins to form a heart tube suspended from the pericardial roof by the dorsal mesocardium, as beautifully documented in three-dimensional reconstructions by Mohun and colleagues (Fig. 3) (Mohun et al., 2000). The heart tube, which up to this time has remained one cell deep, thickens and the inner portion gives rise to the myocardium, while the outer layer forms the thin visceral portion of the pericardium. The heart tube becomes disconnected entirely from the parietal layer of pericardium, so that no ventral mesocardium persists and the pericardium continuously lines the pericardial space. Anteriorly in the tadpole, the truncus ateriosus divides into two branches that continue to the aortic arches and connect to the gill filaments. The endothelial lining of the heart tube, the endocardium, becomes visible as the myocardial tube forms. Lineage relationships between the endocardium, myocardium and pericardium in the amphibian are not well-established, but cell tracing studies (Raffin et al., 2000) have shown that at least the myocardial and pericardial roof cells arise from the original heart field which is marked by mesodermal expression of Nkx2.5. Endocardial cells also appear to arise from this field, but it is not clear that this is their exclusive source. Whether a secondary heart field, such as contributes to the rostral and caudal poles of amniote hearts (Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001; Cai et al., 2003), exists in amphibians remains a matter of speculation. Recently, Brade et al. (2007) described a domain of cells overlapping with and extending anteriorwards to the Nkx2.5 heart field at neurula stages that are marked by expression of Isl-1, which encodes a transcription factor that is associated with the mammalian secondary heart field. Morpholino-mediated knockdown of Isl-1 in Xenopus gave small, abnormally-structured hearts and correspondingly downregulated myocardial gene expression. Although amphibians lack a right ventricle that in mammals is derived from the secondary heart field, it is possible that the mechanisms which control the contribution of precursor cells to the heart might be regulated similarly in the frog heart field

96

PART  |  1  Heart Evolution

(A)

(B)

A

P

A

P

e m p

P Dorsal

(C)

(D)

Oblique A

Ventral

A

P

A

P

aa d p

e

m

da as aa P Dorsal

Oblique

A

Ventral

Figure 3  Formation and closure of the heart tube. Transverse sections (A, C) through the forming heart tube at the level of the future ventricle show myocardial tube formation (A, B, stage 29) and closure (C, D, stage 32). Distinct pericardial (p), myocardial (m) and endocardial (e) layers are visible. The three-dimensional computer reconstructions (B, D, viewed in panels as indicated) show the myocardium (red) and endocardium (yellow). The sheet of myocardial and pericardial roof cells descends into the expanding pericardial cavity and folds to form a trough containing the epicardial tube. Lateral regions of the thickened myocardium join to form a tube connected to the pericardial roof by a dorsal mesocardium (d). Blue dashed lines in (B) and (D) indicate the approximate positions of the histological sections. Images are reproduced from Mohun et al. (2000) with permission.

as in the amniote secondary heart field. Moreover, given the region of nonoverlap anterior to the Nkx2.5 domain, it is also reasonable to speculate that the uniquely Isl-1 cells might contribute to discrete regions from the Nkx2.5 cells, although definitive demonstration of a secondary heart field waits for lineage labeling of these cells. As the heart tube elongates, it undergoes a characteristic leftward bend of the outflow tract and rightward looping of the conus (stages 33–35 in Xenopus laevis). Acquisition of its characteristic spiral shape is seen clearly in the Mohun et al. (2000) three-dimensional reconstructions (Fig. 4), and in two-dimensional projections of a Z-series from confocal fluorescent microscopy of Kolker et al. (2000) that visualized successive stages of tube formation and looping. Atrial septation and anteriorwards movement occurs during stages 35–40, and ventricular myocardial wall thickening and trabeculation occur over the next few stages. The left–right orientation of asymmetric morphogenesis is established earlier during development. Asymmetric morphogenesis and its determining mechanisms are reviewed in Part 4 of this book, but it is relevant to reiterate here that the direction of left–right asymmetric morphogenesis appears to be established considerably earlier in amphibians than in amniotes. As early as the 4-cell stage in Xenopus laevis, left–right differences are apparent in the expression of genes encoding certain proteins, such as H/K-ATPase (Levin et al., 2002) that are thought to promote differences in cell physiology and, consequently, activation of specific left- and right-sided cascades of gene expression. Although some aspects of the signaling cascades that orient left–right asymmetry are shared between

avian and mammalian species, the timing and some of the earliest events differ, and this might reflect important differences in developmental strategies. Determination of left–right asymmetry so soon after fertilization in amphibians might reflect the fixed orientation of the embryo’s dorsoventral and ventroposterior body axes relative to the sperm entry point and subsequent rotation of the egg cortex that orients the embryo’s dorsoventral and anterioposterior axes (Harland and Gerhart, 1997). Interestingly, while the mouse embryonic body axes are also oriented with respect to each other and morphologic landmarks prior to cleavage in at least a proportion of unperturbed embryos (for instance, Ciemerych et al., 2000; Gardner, 2001), they can be reoriented quite readily (consider the generation of chimeric mice created by aggregation of two morula-stage embryos) and this capability might require that mammalian left–right asymmetry is determined relatively later, once the embryo has become highly-multicellular. It is not surprising that cultured heart field explants from tailbud-stage amphibian embryos show poorly-developed myofibrillar organization and do not develop a characteristic looped heart tube, implicating additional signals that control development of mature contractile properties and tissue organization. Although the deficits are probably due in part to the lack of physical influences, such as from hemodynamics and the body cavity, it is also possible that the explants lack signals from cardiac neural crest cells (Farrell et al., 1999) and epicardial cells (Dettman et al., 1998; Gittenberger-de Groot et al., 1998), which contribute to the heart and influence myocardial structure in amniotes, as well as any secondary heart field cells discussed above. Morphogenesis of the amphibian heart tube and acquisition

97

Chapter  |  1.3  Cardiac Development in the Frog

(A)

function without the potential confounding influence of misexpression elsewhere in the embryo. There are a good number of questions that can be addressed, not least of which is definition of the alterations in cell shape and possibly localized proliferation that drive morphogenesis.

(B)

e c P Oblique

(C)

A

(D)

Figure 4  Spiral looping of the heart tube of Xenopus laevis. Transverse section through the conus region (A) shows its rightward displacement. The conus (c) and endocardium (e) are indicated. The three-dimensional oblique model (B) shows the myocardium (red), endocardium (yellow), dorsal mesocardium and pericardial roof (gray), revealing the rotation of the heart tube. Frontal views of stage 46 heart stained with antibodies to cardiac troponin-T (red) and fibrillin (green) (C) or type II collagen (purple) and red delineates the heart (D). The images are presented as a compressed Z-stack of serial confocal images spaced 5–7 m apart. Images in panels (A) and (B) are reproduced from Mohun et al. (2000) with permission, and in panels (C) and (D) were provided by Sandra Kolker and Dan Weeks (Kolker et al., 2000). Ventricular trabeculation and outflow tract looping are clearly visible (C), as is the curvature of the spiral valve (D).

of the characteristic structures of the heart and its chambers remains poorly-understood at the molecular genetic level. This is in large part due to the difficulty in modulating gene expression in late-stage embryos uniformly and selectively within a region of the developing heart, as has been possible in mice because of powerful transgenic technology and in chicks because of viral transduction technology. Progress in transgenesis in Xenopus tropicalis embryos, however, should help alleviate this shortcoming, although the paucity of good promoters remains an impediment and has slowed widespread use of this technology. To date, Xenopus tropicalis transgenesis has contributed to an understanding of the transcriptional regulation of certain genes that are regionally expressed within the developing heart (for instance, see Latinkic et al., 2002; Smith et al., 2005). However, the ability to control the spatio–temporal expression of transgenes should eventually allow the conditional gain and loss of function experiments that are required to probe gene

VI.B.  The Three-Chambered   Amphibian Heart The late stage tadpole develops the three-chambered heart typical of the adult frog with two atria and a single ventricle. Despite the single ventricle, Mayer (1835) showed that distinct streams of oxygen-rich and oxygen-poor blood will course from the cut tip of the ventricle in a pithed Rana. The classical model postulated by Brücke and later Sabatier, based on anatomical considerations and direct observation of beating hearts, explained that systemic and pulmonary blood flow, at least in several Rana species, and might remain mostly distinct owing to the structural and physiological features of the heart (Brücke, 1852; Sabatier, 1873). Other investigators have challenged whether this is always the case, suggesting that mixing occurs in some species or even varies among individuals (discussed in de Graaf, 1957; Haberich, 1965). Reinvestigation of the question of functional separation by a number of laboratories using colored or fluorescent dyes taking care to avoid transient anomalous filling, and potential associated perturbation of flow, clearly showed separation in several species, including Rana and Xenopus. The anatomy of the frog (Rana species) heart was wellknown to early physiologists and that of the Xenopus laevis heart has been described by de Graaf as similar to Rana. Blood from the pulmonary veins enters the left atrium and contraction closes the valveless opening preventing reflux. Blood from the sinus venosus enters the right atrium through thickenings of the sinoatrial aperture that presumably act as valves. Blood from both atria enter into the ventricle through the atrioventricular aperture, which contains two large, thick valves (dorsal and ventral) and two smaller semi-lunar valves (left and right). The ventricle has an open chamber and a series of deep slit-like chambers leading to a fairly thin outer wall. Separating the ventricle and bulbus cordis are three semilunar valves. Within the bulbus is the spiral valve, an unusually long fold of tissue attached to the dorsal (twisting towards right) wall of the bulbus with its ventral side suspended freely (Fig. 4D). Anteriorly, the bulbus is followed by the truncus arteriosus, which is divided by a septum into dorsal pulmo-cutaneous and ventral aortic trunks. This septum is continuous with the spiral valve; thus, the left side leaving the ventricle twists clockwise towards dorsal moving anteriorly. In the classical Brücke and Sabatier model, atrial contractions were envisaged to direct oxygenated left atrial blood to the left side of the ventricle while right atrial

98

blood enters the right and middle regions, and reception by the slit-like chambers would prevent mixing. Ventricular contraction first occurs on the right and would direct presumably oxygen-poor blood to the relaxed bulbus to the left of the spiral valve and towards the pulmonary trunk. The model supposes that the spiral valve directs blood towards the pulmonary trunk at the beginning of contraction because the peripheral resistance of the pulmonocutaneous trunk is less than that of the carotid and aorta. As ventricular contraction continues and left (presumably oxygenated) blood is expelled, bulbus contraction would cause the spiral valve to press against the opening to the pulmocutaneous trunk, thereby deflecting blood towards the carotid arches and systemic circulation. There have been several attempts to test this model directly. Notably, blood flow has been visualized by ultraviolet cinematography in hearts of living Hyla caerulea, an Australian treefrog (Simons and Michaelis, 1953) and Xenopus laevis (de Graaf, 1957). The Hyla caerulea cinematography showed that systemic and pulmonary flow through the ventricle, bulbus and truncus could indeed be maintained distinct. The Xenopus cinematography showed a more complicated situation. Right atrial (oxygen-poor) blood was transferred to the right side of the ventricle and primarily to the pulmonocutaneous arches. A considerable portion of oxygen-rich left atrial blood, however, mixed with that on the right side of the ventricle and entered into the systemic and pulmonary circulation. Thus, flow is not mixed randomly in either species, but the situation appears more complex in Xenopus than predicted by the classical model or as borne out in Hyla caerulea. Perfusion studies using dyed liquids in Rana also supported separation, but there was no consistent difference between pulmonocutaneous and systemic or carotid arterial pressure, either systolic or diastolic, as predicted by the classical model (Haberich, 1965). Thus, the mechanisms involved in maintaining separate laminar flow of the blood are not understood in detail, but it would appear that the spiral shape of the bulbus cordis and the shape of the spiral valve might direct laminar flow patterns of the streaming blood.

VII.  Conclusions and prospects Two centuries of amphibian cardiogenesis research have been instrumental in providing a coherent picture of how signaling pathways specify and form vertebrate heart tissue during early embryonic development. The past decade in particular has been remarkable because of the advance made in uncovering the nature of embryonic signals, including the complex Wnt, Wnt antagonism, Nodal, FGF, Notch and BMP signal transduction pathways, attesting to the power of amphibian embryology in teasing apart early developmental events. Importantly, these signals have turned out to play evolutionarily-conserved roles in heart

PART  |  1  Heart Evolution

induction in species with such divergent anatomies as represented by amphibians, birds and mammals. Improved methods of imaging heart development and the advent of stable transgenesis for conditional, tissue-specific gene modulation should open the door to more detailed morphogenetic studies, as well as research into later stages of heart development.

References Acampora, D., Avantaggiato, V., Tuorto, F., Briata, P., Corte, G., Simeone, A., 1998. Visceral endoderm-restricted translation of Otx1 mediates recovery of Otx2 requirements for specification of anterior neural plate and normal gastrulation. Development 125, 5091–5104. Agius, E., Oelgeschlager, M., Wessely, O., Kemp, C., De Robertis, E.M., 2000. Endodermal Nodal-related signals and mesoderm induction in Xenopus. Development 127, 1173–1183. Alexander, J., Stainier, D.Y., 1999. A molecular pathway leading to endoderm formation in zebrafish. Curr. Biol. 9, 1147–1157. Allen, B.G., Allen-Brady, K., Weeks, D.L., 2006. Reduction of XNkx210 expression leads to anterior defects and malformation of the embryonic heart. Mech. Dev. 123, 719–729. Alsan, B.H., Schultheiss T.M., 2002. Regulation of avian cardiogenesis by Fgf8 signaling. Development 129, 1935–1943. Andree, B., Duprez, D., Vorbusch, B., Arnold, H.-H., Brand, T., 1998. BMP-2 induces ectopic expression of cardiac lineage markers and interferes with somite formation in chicken embryos. Mech. Dev. 70, 119–131. Ang, S.-L., Jin, O., Rhinn, M., Daigle, N., Stevenson, L., Rossant, J., 1996. A targeted mouse Otx2 mutation leads to severe defects in gastrulation and formation of axial mesoderm and to deletion of rostral brain. Development 122, 243–252. Balinsky, B.I., 1939. Experiments on total extirpation of the whole endoderm in Triton embryos. C. R. Acad. Sci. URSS 23, 196–198. Barron, M., Gao, M., Lough, J., 2000. Requirement for BMP and FGF signaling during cardiogenic induction in non-precardiac mesoderm is specific, transient, and cooperative. Dev. Dyn. 218, 383–393. Beddington, R.S., 1994. Induction of a second neural axis by the mouse node. Development 120, 613–620. Benson, D.W., Silberbach, G.M., Kavanaugh-McHugh, A., Cottrill, C., Zhang, Y., Riggs, S., Smalls, O., Johnson, M.C., Watson, M.S., Seidman, J.G., Seidman, C.E., Plowden, J., Kugler, J.D., 1999. Mutations in the cardiac transcription factor NKX2.5 affect diverse cardiac developmental pathways. J. Clin. Invest. 104, 1567–1573. Beppu, H., Kawabata, M., Hamamoto, T., Chytil, A., Minowa, O., Noda, T., Miyazono, K., 2000. BMP type II receptor is required for gastrulation and early development of mouse embryos. Dev. Biol. 221, 249–258. Bertocchini, F., Stern, C.D., 2002. The hypoblast of the chick embryo positions the primitive streak by antagonizing nodal signaling. Dev. Cell 3, 735–744. Bertocchini, F., Skromne, I., Wolpert, L., Stern, C.D., 2004. Determination of embryonic polarity in a regulative system: evidence for endogenous inhibitors acting sequentially during primitive streak formation in the chick embryo. Development 131, 3381–3390. Bouwmeester, T., Kim, S.-H., Sasai, Y., Lu, B., De Robertis, E.M., 1996. Cerberus is a head-inducing secreted factor expressed in the anterior endoderm of Spemann’s Organizer. Nature 382, 595–601.

Chapter  |  1.3  Cardiac Development in the Frog

Brade, T., Gessert, S., Kuhl, M., Pandur, P., 2007. The amphibian second heart field: Xenopus islet-1 is required for cardiovascular development. Dev. Biol. 311, 297–310. Brott, B.K., Sokol, S.Y., 2005. A vertebrate homolog of the cell cycle regulator Dbf4 is an inhibitor of Wnt signaling required for heart development. Dev. Cell 8, 703–715. Brücke, E., 1852. Beiträge zur vergleichenden Anatomie und Physiologie des Gefässsystems, I. Ueber die Mechanik des Kreislaufes bei den Amphibien. Denkschriften: Akademie der Wissenschaften Wien, Mathematisch-Naturwissenschaftliche Classe 3, 335–367. Cai, C.L., Liang, X., Shi, Y., Chu, P.H., Pfaff, S.L., Chen, J., Evans, S., 2003. Isl1 identifies a cardiac progenitor population that proliferates prior to differentiation and contributes a majority of cells to the heart. Dev. Cell 5, 877–889. Caneparo, L., Huang, Y.L., Staudt, N., Tada, M., Ahrendt, R., Kazanskaya, O., Nierhrs, C., Houart, C., 2007. Dickkopf-1 regulates gastrulation movements by coordinated modulation of Wnt/betacatenin and Wnt/PCP activities, through interaction with the Dallylike homolog Knypek. Genes Dev. 21, 465–480. Chang, H., Huylebroeck, D., Verschueren, K., Guo, Q., Matzuk, M.M., Zwijsen, A., 1999. Smad5 knockout mice die at mid-gestation due to multiple embryonic and extraembryonic defects. Development 126, 1631–1642. Chang, H., Zwijsen, A., Vogel, H., Huylebroeck, D., Matzuk, M.M., 2000. Smad5 is essential for left–right asymmetry in mice. Dev. Biol. 219, 71–78. Chen, F., Kook, H., Milewski, R., Gitler, A.D., Lu, M.M., Li, J., Nazarian, R., Schnepp, R., Jen, K., Biben, C., Runke, G., Mackay, J.P., Novotny, J., Schwartz, R.J., Harvey, R.P., Mullins, M.C., Epstein, J.A., 2002. Hop is an unusual homeobox gene that modulates cardiac development. Cell 110, 713–723. Chen, J.N., Fishman, M.C., 1996. Zebrafish tinman homolog demarcates the heart field and initiates myocardial differentiation. Development 122, 3809–3816. Christian, J.L., Moon, R.T., 1993. Interactions between Xwnt-8 and Spemann organizer signaling pathways generate dorsoventral pattern in the embryonic mesoderm of Xenopus. Genes Dev. 7, 13–28. Chuang, H.H., Tseng, M.P., 1957. An experimental analysis of the determination and differentiation of the mesodermal structures of neurula in urodeles. Scientia Sinica 6, 669–708. Ciemerych, M.A., Mesnard, D., Zernicka-Goetz, M., 2000. Animal and vegetal poles of the mouse egg predict the polarity of the embryonic axis, yet are nonessential for development. Development 127, 3467–3474. Cleaver, O.B., Patterson, K.D., Krieg, P.A., 1996. Overexpression of the tinman-related genes XNkx-2.5 and XNkx-2.3 in Xenopus embryos results in myocardial hyperplasia. Development 122, 3549–3556. Conlon, F.L., Lyons, F.M., Takaesu, N., Barth, K.S., Kispert, A., Herrmann, B., Robertson, E.J., 1994. A primary requirement for nodal in the formation and maintenance of the primitive streak in the mouse. Development 120, 1919–1928. Copenhaver, W.M., 1926. Experiments on the development of the heart of Amblystoma punctatum. J. Exper. Zool. 43, 321–371. Dale, L., Slack, J.M., 1987. Fate map for the 32-cell stage of Xenopus laevis. Development 99, 527–551. David, N.B., Rosa, F.M., 2001. Cell autonomous commitment to an endodermal fate and behaviour by activation of Nodal signalling. Development 128, 3937–3947. de Graaf, A.R., 1957. Investigations into the distribution of blood in the heart and aortic arches of Xenopus laevis (Daud.). J. Exp. Biol. 34, 143–172.

99

De Robertis, E.M., Kuroda, H., 2004. Dorsal-ventral patterning and neural induction in Xenopus embryos. Annu. Rev. Cell Dev. Biol. 20, 285–308. DeHaan, R.L., 1965. Morphogenesis of the vertebrate heart. In: DeHaan, R.L., Ursprung, H. (Eds.), Organogenesis, vol. 16. Holt, Rinehart and Winston, New York, pp. 377–419. Dettman Jr., R.W., Denetclaw, W., Ordahl, C.P., Bristow, J., 1998. Common epicardial origin of coronary vascular smooth muscle, perivascular fibroblasts, and intermyocardial fibroblasts in the avian heart. Dev. Biol. 193, 169–181. Ding, J., Yang, L., Yan, Y.T., Chen, A., Desai, N., Wynshaw-Boris, A., Shen, M.M., 1998. Cripto is required for correct orientation of the anterior–posterior axis in the mouse embryo. Nature 395, 702–707. Dionne, M.S., Skarnes, W.C., Harland, R.M., 2001. Mutation and analysis of Dan, the founding member of the Dan family of transforming growth factor beta antagonists. Mol. Cell Biol. 21, 636–643. Durocher, D., Charron, F., Warren, R., Schwartz, R.J., Nemer, M., 1997. The cardiac transcription factors Nkx2-5 and GATA-4 are mutual cofactors. Embo. J. 16, 5687–5696. Eisenberg, C.A., Eisenberg, L.M., 1999. WNT11 promotes cardiac tissue formation of early mesoderm. Dev. Dyn. 216, 45–58. Ekman, G., 1925. Experimentelle Beiträge zur Herzentwicklung der Amphibien. Wilhelm Roux Arch. EntwMech. Org. 106, 320–352. Farrell, M., Waldo, K., Li, Y.X., Kirby, M.L., 1999. A novel role for cardiac neural crest in heart development. Trends Cardiovasc. Med. 9, 214–220. Feldman, B., Gates, M.A., Egan, E.S., Dougan, S.T., Rennebeck, G., Sirotkin, H.I., Schier, A.F., Talbot, W.S., 1998. Zebrafish organizer development and germ-layer formation require nodal-related signals. Nature 395, 181–185. Foley, A., Mercola, M., 2005. Heart induction by Wnt antagonists depends on the homeodomain transcription factor Hex. Genes Dev. 19, 387–396. Foley, A., Korol, O., Timmer, A.M., Mercola, M., 2007. Nodal and Cerberus cooperate in heart induction. Dev. Biol. 303 (1), 57–65. Fu, Y., Yan, W., Mohun, T.J., Evans, S.M., 1998. Vertebrate tinman homologues XNkx2.3 and XNkx2.5 are required for heart formation in a functionally redundant manner. Development 125, 4439–4449. Gardner, R.L., 2001. Specification of embryonic axes begins before cleavage in normal mouse development. Development 128, 839–847. Garriock, R.J., Drysdale, T.A., 2003. Regulation of heart size in Xenopus laevis. Differentiation 71, 506–515. Gittenberger-de Groot, A.C., Vrancken Peeters, M.P., Mentink, M.M., Gourdie, R.G., Poelmann, R.E., 1998. Epicardium-derived cells contribute a novel population to the myocardial wall and the atrioventricular cushions. Circ. Res. 82, 1043–1052. Gove, C., Walmsley, M., Nijjar, S., Bertwhistle, D., Guille, M., Partington, G., Bomford, A., Patient, R., 1997. Over-expression of GATA-6 in Xenopus embryos blocks differentiation of heart precursors. EMBO J. 16, 355–368. Grow, M.W., Krieg, P.A., 1998. Tinman function is essential for vertebrate heart development: elimination of cardiac differentiation by dominant inhibitory mutants of the tinman-related genes, Nkx2.3 and Nkx2.5. Dev. Biol. 204, 187–196. Gu, Z., Reynolds, E.M., Song, J., Lei, H., Feijen, A., Yu, L., He, W., MacLaughlin, D.T., van den Eijnden-van Raaij, J., Donahoe, P.K., Li, E., 1999. The type I serine/threonine kinase receptor ActRIA (ALK2) is required for gastrulation of the mouse embryo. Development 126, 2551–2561.

100

Haberich, F.J., 1965. The functional separation of venous and arterial blood in the univentricular frog heart. Ann. NY Acad. Sci. 127, 459–476. Harland, R., Gerhart, J., 1997. Formation and function of Spemann’s organizer. Annu. Rev. Cell Dev. Biol. 13, 611–667. Hartenstein, A.Y., Rugendorff, A., Tepass, U., Hartenstein, V., 1992. The function of the neurogenic genes during epithelial development in the Drosophila embryo. Development 116, 1203–1220. Henry, G.L., Brivanlou, I.H., Kessler, D.S., Hemmati-Brivanlou, A., Melton, D.A., 1996. TGF-b signals and a prepattern in Xenopus laevis endodermal development. Development 122, 1007–1015. Hommes, O.R., 1957. Primary Entodermal Defects, Development of Body Form and Gential Organs of Acardii in Uni-Vitelline Twins. Jacob van Campen, Amsterdam. Jacobson, A.G., 1960. Influences of ectoderm and endoderm on heart differentiation in the newt. Dev. Biol. 2, 138–154. Jacobson, A.G., 1961. Heart determination in the newt. J. Exper. Zool. 146, 139–152. Jacobson, A.G., Duncan, J.T., 1968. Heart induction in salamanders. J. Exper. Zool. 167, 79–103. Jones, C.M., Kuehn, M.R., Hogan, B.L., Smith, J.C., Wright, C.V., 1995. Nodal-related signals induce axial mesoderm and dorsalize mesoderm during gastrulation. Development 121, 3651–3662. Keller, R., Winklbauer, R., 1992. The cellular basis of amphibian gastrulation. Curr. Top. Dev. Biol. 27, 39–89. Keller, R.E., 1976. Vital dye mapping of the gastrula and neurula of Xenopus laevis II. Prospective areas and morphogenetic movements of the deep layer. Dev. Biol. 51, 118–137. Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1, 435–440. Kishimoto, Y., Lee, K.-H., Zon, L., Hammerschmidt, M., Schulte-Merker, S., 1997. The molecular nature of zebrafish swirl: BMP2 function is essential during early dorsoventral patterning. Development 124, 4457–4466. Kochilas, L.K., Li, J., Jin, F., Buck, C.A., Epstein, J.A., 1999. p57Kip2 expression is enhanced during mid-cardiac murine development and is restricted to trabecular myocardium. Pediatr. Res. 45, 635–642. Kohn, A.D., Moon, R.T., 2005. Wnt and calcium signaling: beta-cateninindependent pathways. Cell Calcium. 38, 439–446. Kolker, S.J., Tajchman, U., Weeks, D.L., 2000. Confocal imaging of early heart development in Xenopus laevis. Dev. Biol. 218, 64–73. Korol, O., Gupta, R.W., Mercola, M., 2008. A novel activity of the Dickkopf-1 amino terminal domain promotes axial and heart development independently of canonical Wnt inhibition. Dev. Biol. 324, 131–138. Kuhl, M., Sheldahl, L.C., Park, M., Miller, J.R., Moon, R.T., 2000. The Wnt/Ca2 pathway: a new vertebrate Wnt signaling pathway takes shape. Trends Genet. 16, 279–283. Kuhl, M., Geis, K., Sheldahl, L.C., Pukrop, T., Moon, R.T., Wedlich, D., 2001. Antagonistic regulation of convergent extension movements in Xenopus by Wnt/beta-catenin and Wnt/Ca2 signaling. Mech. Dev. 106, 61–76. Kumano, G., Ezal, C., Smith, W.C., 2006. ADMP2 is essential for primitive blood and heart development in Xenopus. Dev. Biol. 299, 411–423. Ladd, A., Yatskievych, T.A., Antin, P.B., 1998. Regulation of avian cardiac myogenesis by activin/TGFb and bone morphogenetic proteins. Dev. Biol. 204, 407–419. Lane, M.C., Sheets, M.D., 2000. Designation of the anterior/posterior axis in pregastrula Xenopus laevis. Dev. Biol. 225, 37–58.

PART  |  1  Heart Evolution

Latinkic, B.V., Cooper, B., Towers, N., Sparrow, D., Kotecha, S., Mohun, T.J., 2002. Distinct enhancers regulate skeletal and cardiac musclespecific expression programs of the cardiac alpha-actin gene in Xenopus embryos. Dev. Biol. 245, 57–70. Levin, M., Thorlin, T., Robinson, K., Nogi, T., Mercola, M., 2002. Asymmetries in H()/K()-ATPase and cell membrane potentials comprise a very early step in left-right patterning. Cell 111, 77–89. Li, G., Iyengar, R., 2002. Calpain as an effector of the Gq signaling pathway for inhibition of Wnt/beta-catenin-regulated cell proliferation. Proc. Natl. Acad. Sci. USA 99, 13254–13259. Lickert, H., Kutsch, S., Kanzler, B., Tamai, Y., Taketo, M.M., Kemler, R., 2002. Formation of multiple hearts in mice following deletion of beta-catenin in the embryonic endoderm. Dev. Cell 3, 171–181. Liguori, G.L., Echevarria, D., Improta, R., Signore, M., Adamson, E., Martinez, S., Persico, M.G., 2003. Anterior neural plate regionalization in cripto null mutant mouse embryos in the absence of node and primitive streak. Dev. Biol. 264, 537–549. Liu, Y., Asakura, M., Inoue, H., Nakamura, T., Sano, M., Niu, Z., Chen, M., Schwartz, R.J., Schneider, M.D., 2007. Sox17 is essential for the specification of cardiac mesoderm in embryonic stem cells. Proc. Natl. Acad. Sci. USA 104, 3859–3864. Logan, M., Mohun, T., 1993. Induction of cardiac muscle differentiation in isolated animal pole explants of Xenopus laevis embryos. Development 118, 865–875. Lough, J., Barron, M., Brogley, M., Sugi, Y., Bolender, D.L., Zhu, X., 1996. Combined BMP-2 and FGF-4, but neither factor alone, induces cardiogenesis in non-precardiac embryonic mesoderm. Devl. Biol. 178, 198–202. Mangiacapra, F.J., Fransen, M.E., Lemanski, L.F., 1995. Activin A and transforming growth factor-beta stimulate heart formation in axolotls but not rescue cardiac lethal mutants. Cell Tissue Res. 282, 227–236. Martinez Barbera, J.P., Clements, M., Thomas, P., Rodriguez, T., Meloy, D., Kioussis, D., Beddington, R.S., 2000. The homeobox gene Hex is required in definitive endodermal tissues for normal forebrain, liver and thyroid formation. Development 127, 2433–2445. Martinsen, B.J., Frasier, A.J., Baker, C.V., Lohr, J.L., 2004. Cardiac neural crest ablation alters Id2 gene expression in the developing heart. Dev. Biol. 272, 176–190. Marvin, M.J., Di Rocco, G., Gardiner, A., Bush, S.M., Lassar, A.B., 2001. Inhibition of Wnt activity induces heart formation from posterior mesoderm. Genes Dev. 15, 316–327. Matsui, H., Ikeda, K., Nakatani, K., Sakabe, M., Yamagishi, T., Nakanishi, T., Nakajima, Y., 2005. Induction of initial cardiomyocyte alpha-actin – smooth muscle alpha-actin – in cultured avian pregastrula epiblast: a role for nodal and BMP antagonist. Dev. Dyn. 233, 1419–1429. Maye, P., Zheng, J., Li, L., Wu, D., 2004. Multiple mechanisms for Wnt11-mediated repression of the canonical Wnt signaling pathway. J. Biol. Chem. 279, 24659–24665. Mayer, H.F.J.C., 1835. Analekten für Vergleichende Anatomie, Bonn, Germany. Mishina, Y., Suzuki, A., Ueno, N., Behringer, R.R., 1995. Bmpr encodes a type I bone morphogenetic protein receptor that is essential for gastrulation during mouse embryogenesis. Genes Dev. 9, 3027–3037. Mishina, Y., Hanks, M.C., Miura, S., Tallquist, M.D., Behringer, R.R., 2002. Generation of Bmpr/Alk3 conditional knockout mice. Genesis 32, 69–72. Mjaatvedt, C.H., Nakaoka, T., Moreno-Rodriguez, R., Norris, R.A., Kern, M.J., Eisenberg, C.A., Turner, D., Markwald, R.R., 2001. The

Chapter  |  1.3  Cardiac Development in the Frog

outflow tract of the heart is recruited from a novel heart-forming field. Dev. Biol. 238, 97–109. Mlodzik, M., 2002. Planar cell polarization: do the same mechanisms regulate Drosophila tissue polarity and vertebrate gastrulation? Trends Genet. 18, 564–571. Mohun, T.J., Leong, L.M., Weninger, W., Sparrow, D.B., 2000. The morphology of heart development in Xenopus laevis. Dev. Biol. 218, 74–88. Nakajima, Y., Yamagishi, T., Ando, K., Nakamura, H., 2002. Significance of bone morphogenetic protein-4 function in the initial myofibrillogenesis of chick cardiogenesis. Dev. Biol. 245, 291–303. Nascone, N., Mercola, M., 1995. An inductive role for the endoderm in Xenopus cardiogenesis. Development 121, 515–523. Newman, C.S., Reecy, J., Grow, M.W., Ni, K., Boettger, T., Kessel, M., Schwartz, R.J., Krieg, P.A., 2000. Transient cardiac expression of the tinman-family homeobox gene, XNkx2-10. Mech. Dev. 91, 369–373. Nieuwkoop, P.D., 1947. Experimental investigations on the origin and determination of the germ cells, and on the development of the lateral plates and germ ridges in Urodeles. Archs. Neerl. Zool. 8, 1–205. Nieuwkoop, P.D., Faber, J., 1994. Normal Table of Xenopus laevis (Daudin): A Systematical and Chronological Survey of the Development from the Fertilized Egg Till the End of Metamorphosis. Garland Pub., New York, NY. Orts-Llorca, F., 1963. Influence of the endoderm on heart differentiation during the early stages of development of the chicken embryo. Wilhelm Roux Arch. Entw. Mech. Org. 154, 533–551. Orts-Llorca, F., Gil, D.R., 1965. Influence of the endoderm on heart differentiation. Wilhelm Roux Arch. Entw. Mech. Org. 156, 368–370. Osada, S.I., Wright, C.V., 1999. Xenopus nodal-related signaling is essential for mesendodermal patterning during early embryogenesis. Development 126, 3229–3240. Pandur, P., Lasche, M., Eisenberg, L.M., Kuhl, M., 2002. Wnt-11 activation of a non-canonical Wnt signalling pathway is required for cardiogenesis. Nature 418, 636–641. Pasumarthi, K.B., Field, L.J., 2002. Cardiomyocyte cell cycle regulation. Circ. Res. 90, 1044–1054. Piccolo, S., Agius, E., Leyns, L., Bhattacharyya, S., Grunz, H., Bouwmeester, T., De Robertis, E.M., 1999. The head inducer Cerberus is a multifunctional antagonist of Nodal, BMP and Wnt signals. Nature 397, 707–710. Plageman Jr, T.F., Yutzey, K.E., 2005. T-box genes and heart development: putting the “T” in heart. Dev. Dyn. 232, 11–20. Raffin, M., Leong, L.M., Rones, M.S., Sparrow, D., Mohun, T., Mercola, M., 2000. Subdivision of the cardiac Nkx2.5 expression domain into myogenic and non-myogenic compartments. Dev. Biol. 218, 326–340. Rawles, M.E., 1936. A study in the localization of organ-forming areas in the chick blastoderm of the head process stage. J. Exp. Zool. 138, 505–555. Reissmann, E., Jornvall, H., Blokzijl, A., Andersson, O., Chang, C., Minchiotti, G., Persico, M.G., Ibanez, C.F., Brivanlou, A.H., 2001. The orphan receptor ALK7 and the Activin receptor ALK4 mediate signaling by Nodal proteins during vertebrate development. Genes Dev. 15, 2010–2022. Rhinn, M., Dierich, A., Shawlot, W., Behringer, R.R., Le Meur, M., Ang, S.-L., 1998. Sequential roles for Otx2 in visceral endoderm and neuroectoderm for forebrain and midbrain induction and specification. Development 125, 845–856. Rones, M.S., McLaughlin, K.A., Raffin, M., Mercola, M., 2000. Serrate and Notch specify cell fates in the heart field by suppressing cardiomyogenesis. Development 127, 3865–3876.

101

Sabatier, A., 1873. Études sur le coeur et la circulation centrale dans la série des Vertébrés. Ann. Sci. Nat. 18, 1–89. Sater, A.K., Jacobson, A.G., 1989. The specification of heart mesoderm occurs during gastrulation in Xenopus laevis. Development 105, 821–830. Sater, A.K., Jacobson, A.G., 1990a. The restriction of the heart morphogenetic field in Xenopus laevis. Dev. Biol. 140, 328–336. Sater, A.K., Jacobson, A.G., 1990b. The role of the dorsal lip in the induction of heart mesoderm in Xenopus laevis. Development 108, 461–470. Schlange, T., Andree, B., Arnold, H.H., Brand, T., 2000. BMP2 is required for early heart development during a distinct time period. Mech. Dev. 91, 259–270. Schneider, V.A., Mercola, M., 1999. Spatially distinct head and heart inducers within the Xenopus organizer region. Curr. Biol. 9, 800–809. Schneider, V.A., Mercola, M., 2001. Wnt antagonism initiates cardio­ genesis in Xenopus laevis. Genes Dev. 15, 304–315. Schott, J.J., Benson, D.W., Basson, C.T., Pease, W., Silberbach, G.M., Moak, J.P., Maron, B.J., Seidman, C.E., Seidman, J.G., 1998. Congenital heart disease caused by mutations in the transcription factor NKX2-5. Science 281, 108–111. Schroeder, T., Fraser, S.T., Ogawa, M., Nishikawa, S., Oka, C., Bornkamm, G.W., Honjo, T., Just, U., 2003. Recombination signal sequence-binding protein Jkappa alters mesodermal cell fate decisions by suppressing cardiomyogenesis. Proc. Natl. Acad. Sci. USA 100, 4018–4023. Schultheiss, T.M., Burch, J.B., Lassar, A.B., 1997. A role for bone morphogenetic proteins in the induction of cardiac myogenesis. Genes Dev. 11, 451–462. Sepulveda, J.L., Belaguli, N., Nigam, V., Chen, C.Y., Nemer, M., Schwartz, R.J., 1998. GATA-4 and Nkx-2.5 coactivate Nkx-2 DNA binding targets: role for regulating early cardiac gene expression. Mol. Cell Biol. 18, 3405–3415. Shawlot, W., Behringer, R.R., 1995. Requirement for Lim1 in headorganizer function. Nature 374, 425–430. Shawlot, W., Deng, J.M., Behringer, R.R., 1998. Expression of the mouse cerberus-related gene, Cerr1, suggests a role in anterior neural induction and somitogenesis. PNAS 95, 6198–6203. Shi, Y., Katsev, S., Cai, C., Evans, S., 2000. BMP signaling is required for heart formation in vertebrates. Dev. Biol. 224, 226–237. Shih, J., Keller, R., 1992. The epithelium of the dorsal marginal zone of Xenopus has organizer properties. Development 116, 887–899. Shin, C.H., Liu, Z.P., Passier, R., Zhang, C.L., Wang, D.Z., Harris, T.M., Yamagishi, H., Richardson, J.A., Childs, G., Olson, E.N., 2002. Modulation of cardiac growth and development by HOP, an unusual homeodomain protein. Cell 110, 725–735. Simons, J.R., Michaelis, A.R., 1953. A cinematographic technique, using ultraviolet illumination, for amphibian blood circulation. Nature 171, 801. Small, E.M., Warkman, A.S., Wang, D.Z., Sutherland, L.B., Olson, E.N., Krieg, P.A., 2005. Myocardin is sufficient and necessary for cardiac gene expression in Xenopus. Development 132, 987–997. Smith, S.J., Ataliotis, P., Kotecha, S., Towers, N., Sparrow, D.B., Mohun, T.J., 2005. The MLC1v gene provides a transgenic marker of myocardium formation within developing chambers of the Xenopus heart. Dev. Dyn. 232, 1003–1012. Smithers, L.E., Jones, C.M., 2002. Xhex-expressing endodermal tissues are essential for anterior patterning in Xenopus. Mech. Dev. 119, 191–200.

102

Solloway, M.J., Robertson, E.J., 1999. Early embryonic lethality in Bmp5;Bmp7 double mutant mice suggests functional redundancy within the 60A subgroup. Development 126, 1753–1768. Takahashi, S., Yokota, C., Takano, K., Tanegashima, K., Onuma, Y., Goto, J., Asashima, M., 2000. Two novel nodal-related genes initiate early inductive events in Xenopus Nieuwkoop center. Development 127, 5319–5329. Tanaka, M., Chen, Z., Bartunkova, S., Yamasaki, N., Izumo, S., 1999. The cardiac homeobox gene Csx/Nkx2.5 lies genetically upstream of multiple genes essential for heart development. Development 126, 1269–1280. Tao, Q., Yokota, C., Puck, H., Kofron, M., Birsoy, B., Yan, D., Asashima, M., Wylie, C.C., Lin, X., Heasman, J., 2005. Maternal wnt11 activates the canonical wnt signaling pathway required for axis formation in Xenopus embryos. Cell 120, 857–871. Toyama, R., O’Connell, M.L., Wright, C.V., Kuehn, M.R., Dawid, I.B., 1995. Nodal induces ectopic goosecoid and lim1 expression and axis duplication in zebrafish. Development 121, 383–391. Tzahor, E., Lassar, A.B., 2001. Wnt signals from the neural tube block ectopic cardiogenesis. Genes Dev. 15, 255–260. Varlet, I., Collignon, J., Robertson, E.J., 1997. Nodal expression in the primitive endoderm is required for specification of the anterior axis during mouse gastrulation. Development 124, 1033–1044. Waddington, C.H., 1932. Experiments on the development of chick and duck embryos, cultivated in vitro. Philos. Trans. R. Soc. Lond. B 221, 179–230. Waldo, K.L., Kumiski, D.H., Wallis, K.T., Stadt, H.A., Hutson, M.R., Platt, D.H., Kirby, M.L., 2001. Conotruncal myocardium arises from a secondary heart field. Development 128, 3179–3188. Walters, M.J., Wayman, G.A., Christian, J.L., 2001. Bone morphogenetic protein function is required for terminal differentiation of the heart but not for early expression of cardiac marker genes. Mech. Dev. 100, 263–273. Wang, D., Chang, P.S., Wang, Z., Sutherland, L., Richardson, J.A., Small, E., Krieg, P.A., Olson, E.N., 2001. Activation of cardiac

PART  |  1  Heart Evolution

gene expression by myocardin, a transcriptional cofactor for serum response factor. Cell 105, 851–862. Wang, D.Z., Li, S., Hockemeyer, D., Sutherland, L., Wang, Z., Schratt, G., Richardson, J.A., Nordheim, A., Olson, E.N., 2002. Potentiation of serum response factor activity by a family of myocardin-related transcription factors. Proc. Natl. Acad. Sci. USA 99, 14855–14860. Winnier, G., Blessing, M., Labosky, P.A., Hogan, B.L., 1995. Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 2105–2116. Xu, C., Liguori, G., Persico, M.G., Adamson, E.D., 1999. Abrogation of the Cripto gene in mouse leads to failure of postgastrulation morphogenesis and lack of differentiation of cardiomyocytes. Development 126, 483–494. Yamada, M., Revelli, J.P., Eichele, G., Barron, M., Schwartz, R.J., 2000. Expression of chick Tbx-2, Tbx-3, and Tbx-5 genes during early heart development: evidence for BMP2 induction of Tbx2. Dev. Biol. 228, 95–105. Yamamoto, S., Hikasa, H., Ono, H., Taira, M., 2003. Molecular link in the sequential induction of the Spemann organizer: direct activation of the cerberus gene by Xlim-1, Xotx2, Mix.1, and Siamois, immediately downstream from Nodal and Wnt signaling. Dev. Biol. 257, 190–204. Yatskievych, T., Ladd, A., Antin, P., 1997. Induction of cardiac myogenesis in avian pregastrula epiblast: the role of the hypoblast and activin. Development 124, 2561–2570. Zhang, C., Basta, T., Klymkowsky, M.W., 2005. SOX7 and SOX18 are essential for cardiogenesis in Xenopus. Dev. Dyn. 234, 878–891. Zhang, H., Bradley, A., 1996. Mice deficient for BMP2 are nonviable and have defects in amnion/chorion and cardiac development. Development 122, 2977–2986. Zhou, X., Sasaki, H., Lowe, L., Hogan, B.L., Kuehn, M.R., 1993. Nodal is a novel TGF-beta like gene expressed in mouse node during gastrulation. Nature 361, 543–547. Zoltewicz, J.S., Gerhart, J.C., 1997. The Spemann organizer of Xenopus is patterned along its anteroposterior axis at the earliest gatrula stage. Dev. Biol. 192, 482–491.

Chapter 1.4

Cardiac Development in the Zebrafish Ian C. Scott1 and Deborah Yelon2 1 The Hospital for Sick Children, Program in Developmental and Stem Cell Biology, Department of Molecular Genetics, University of Toronto, Toronto, Canada 2 Skirball Institute of Biomolecular Medicine, Developmental Genetics Program, New York University School of Medicine, New York, USA

I.  Introduction Over the past decade, the zebrafish (Danio rerio) has become a popular model organism for the study of vertebrate development. Since the optically transparent embryo develops rapidly, the entirety of embryogenesis can be visualized in real time (Kimmel et al., 1995). The zebrafish embryo is also readily amenable to embryological approaches such as microinjection, lineage tracing and transplantation. The feasibility of conducting large-scale classical genetic screens in zebrafish has further increased interest in its use for discovering genes essential for vertebrate development (Patton and Zon, 2001). In addition to forward (phenotype-driven) genetic approaches, the zebrafish also offers opportunities for reverse (gene-driven) genetic approaches, via the use of antisense morpholino (MO) oligonucleotides to target specific genes (Ekker, 2000). Together, the combination of optical accessibility, embryonic manipulability, and feasibility of genetic approaches makes the zebrafish a unique and exciting model organism for developmental biologists. The attributes of the zebrafish have special appeal for the analysis of heart development. The embryonic zebrafish heart can be imaged live and develops quickly, with the heartbeat starting by 22 hours post-fertilization (hpf). This has allowed discovery of mutations affecting heart development via straightforward visual examination of mutant embryos under a dissecting microscope. To date, a large number of mutations affecting multiple aspects of heart development have been isolated from a variety of forward genetic screens (Chen et al., 1996; Stainier et al., 1996; Alexander et al., 1998; Warren et al., 2000; Beis et al., 2005). Studies of mutant phenotypes have provided significant new insights into the genetic mechanisms Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

underlying cardiogenesis. The ease with which MOs can be used to examine the roles of specific genes has further extended the number of cardiac researchers turning to the zebrafish model to gauge gene function. Analysis of cardiac defects in mutant or MO-injected embryos is relatively convenient, since zebrafish embryos can survive up to seven days post-fertilization (dpf) in the absence of cardiovascular function via oxygen diffusion. In contrast, mouse mutations that compromise aspects of heart or vascular development typically cause early embryonic lethality, necessitating the use of complex temporal and spatial regulation of gene function to assay roles in later developmental steps. Here, we review the utility of zebrafish for the study of heart development. We begin by summarizing the experimental approaches that have been particularly useful for analysis of heart formation. We then outline the major phases of heart development in zebrafish, emphasizing our current understanding of the molecular and cellular regulation of each step. Finally, we reflect on the application of the zebrafish model to the study of human heart disease. Throughout, we highlight the important impact of the combinatorial use of imaging, embryology and genetics in zebrafish on our understanding of how genes regulate cardiac development.

II.  Experimental approaches for analysis of heart development in zebrafish While most often recognized as an opportune organism for forward genetic screens, zebrafish have numerous other experimental strengths. This is, in large part, due to the

103

104

fact that fertilization and embryonic development occur entirely external to the mother. Embryos can be collected, examined and manipulated at the earliest stages of development. Mated pairs of fish yield several hundred embryos per week, providing an abundance of material. Since adult zebrafish are small in size (typically 1.5–2 inches in length) and relatively easy to maintain, an individual laboratory can keep a variety of wild-type, mutant and transgenic lines inexpensively. Therefore, a growing number of researchers are exploiting the zebrafish model for the study of heart development. In this section, we summarize the experimental approaches routinely used in the zebrafish.

II.A.  Genetics The application of forward genetic screens in the fruit fly Drosophila and the nematode Caenorhabditis elegans has had a profound impact on our understanding of the genetic events that govern animal development (Brenner, 1974; Nusslein-Volhard and Wieschaus, 1980). These phenotypedriven screens are ideally suited for discovering new genes whose function is critical to a particular developmental process. Although the Drosophila heart equivalent (the dorsal vessel) is a relatively simple organ, forward genetic screens in this organism have uncovered several genes that are also crucial for vertebrate heart development. One strong example is the Drosophila gene tinman (see Chapter 1.2); mutation of tinman eliminates the dorsal vessel (Bodmer, 1993) and mutation of the mouse homolog Nkx2-5 results in profound perturbations in heart development (Lyons et al., 1995) (see Chapter 9.1). Furthermore, mutations in NKX2-5 are causative for congenital heart disease (CHD) in humans (Schott et al., 1998). Thus, it appears that many of the genetic pathways that regulate heart development are highly-conserved throughout much of the animal kingdom. Of course, the Drosophila dorsal vessel lacks much of the complexity found in the vertebrate heart. Forward genetic screens in the chick and frog are not feasible, whereas this approach in the mouse embryo is laborintensive, due to its in utero development, and expensive, due to the costs associated with housing large populations of mice. The zebrafish represents an ideal solution. Due to the small size, high fecundity and relatively low costs of maintenance of zebrafish, laboratories can easily establish zebrafish populations sufficient in size to carry out forward genetic screens (Patton and Zon, 2001). As the translucent zebrafish embryo develops externally in a simple salt solution, defects in heart development can be assayed rapidly in live embryos under a dissecting microscope. More detailed phenotypic examination is facilitated by screening with molecular markers, especially transgenes that express fluorescent reporters in specific cell types (Jin et al., 2007).

PART  |  1  Heart Evolution

Traditionally, forward genetic screens in zebrafish have taken advantage of the potent mutagen N-ethylN-nitrosourea (ENU), an alkylating agent that introduces random point mutations into the genome (Patton and Zon, 2001). Male founder (F0) adults are treated with ENU, resulting in the mosaic accumulation of point mutations in their sperm. Founders are repeatedly bred with wild-type female fish to establish multiple F1 progeny per F0 male. These F1 fish are then bred to derive F2 families, in which half of the siblings from a given cross will carry a particular point mutation. F2 siblings are then intercrossed, and embryos are screened for defects in heart development. This standard diploid genetic screen is very effective for recovering recessive mutations, and more than 300 mutations affecting discrete steps in zebrafish heart development have been identified using this approach (Chen et al., 1996; Stainier et al., 1996; Warren et al., 2000; Beis et al., 2005). As will be discussed later in this chapter, the processes disrupted by these mutations include specification of cardiac progenitors, assembly of the linear heart tube, chamber morphogenesis and valve formation. While forward genetic screens are very powerful, they can be time-consuming and laborious, given the gen­ eration time of zebrafish (around three months) and the number of F2 families that must be established for a largescale screen. An interesting feature of zebrafish is that haploid and diploid embryos can be created gynogenetically, such that their genome is derived solely from their mother (Corley-Smith et al., 1996). This technique has been exploited in screens analyzing cardiac defects in the hemizygous haploid progeny of F1 generation females, greatly reducing the time and number of fish necessary for an effective screen (Alexander et al., 1998). Although haploid embryos do not survive beyond embryogenesis, the early steps of heart development proceed relatively normally. This allows rapid identification of mutations affecting heart development that can then be further characterized and mapped in the diploid state. The high efficiency of ENU mutagenesis is unfortunately counterbalanced with the challenge of identifying the affected genes. Positional cloning of ENU-induced mutations typically involves high resolution linkage analysis requiring the isolation of 500–3,000 mutant embryos (Bahary et al., 2004). While the nearly complete sequence of the zebrafish genome has facilitated positional cloning, identification of a mutated gene can still take many months of work. An alternative approach to ENU mutagenesis is insertional mutagenesis using retroviruses. While not as efficient as ENU in introducing mutations, retroviral mutagenesis has the advantage of marking mutated genes with a molecular tag, making their identification possible in a much shorter period of time (Amsterdam et al., 1999). A similar insertional approach employing transposons to disrupt genes is currently being developed (Sivasubbu et al., 2007; Nagayoshi et al., 2008).

Chapter  |  1.4  Cardiac Development in the Zebrafish

The ability to modify genes in mice precisely via homologous recombination in embryonic stem cells has revolutionized the study of gene function in vertebrates (Capecchi, 2005). The power of this technique has been further extended by the use of spatially- and temporallyregulated gene modifications. Unfortunately, homologous recombination approaches are not currently feasible in zebrafish, and it will be challenging to overcome the technical hurdle of the early establishment of germ cell fate through inheritance of maternally provided cytoplasmic determinants (Raz, 2003). While the introduction of MOs has been extremely beneficial (see below), the availability of a loss-of-function mutant is often required to ensure that the MO-induced phenotype is specific. A promising development has been the use of targeting induced local lesions in genomes (TILLING) to isolate mutations in a given gene (Sood et al., 2006). In this reverse-genetic approach, PCR targeted against a known gene is used to screen DNA samples derived from a library of F1 fish following ENU mutagenesis. Mutations are detected by sequencing or by using a restriction enzyme that recognizes DNA base pair mismatches. In the coming years, the large-scale application of TILLING and insertional mutagenesis will create an invaluable resource of mutations in many (if not all) zebrafish genes. Commercial and academic consortia are currently scaling-up efforts toward this goal. While they are not strictly gene-based, chemical genetic approaches have recently become popular in zebrafish (Peterson et al., 2000; Murphey and Zon, 2006). The aqueous habitat of the zebrafish embryo makes it particularly suitable for treatment with small molecules. Inhibitors of known specificity can be used to ascertain the role of select signaling pathways in heart development. An advantage is that chemicals can be added and removed at defined time points, allowing for the temporal resolution of requirements for specific signals. Complementing this targeted approach, the large clutch size of zebrafish makes it possible to screen through large libraries of several thousand small molecules in search of previously uncharacterized chemicals that cause specific developmental defects, or that mitigate or rescue specific mutant phenotypes. As will be discussed later in this chapter, this chemical genetic approach may prove to be a powerful method for analyzing zebrafish models of human heart disease, since the effects of small molecules may illuminate disease mechanisms and suggest possible treatments.

II.B.  Regulation of Gene Activity Besides standard genetic approaches, additional techniques are available to analyze gene function in zebrafish. These strategies take advantage of the ease of introducing exogenous material into the zebrafish embryo at early stages. At the 1-cell stage, the zebrafish embryo is large,

105

roughly 0.7 mm in diameter. It is therefore straightforward to microinject RNA, DNA, or protein into the 1-cell stage embryo using a fine glass needle. Following fertilization, a stream of maternally deposited proteins and RNA flows from the yolk into the embryo proper; these gene products are instrumental in supporting the initial phases of embryo­ genesis (Kimmel and Law, 1985). Thus, material microinjected into the yolk at the 1-cell stage is passed into the cell, inherited by its descendants, and thereby distributed into every cell of the embryo. As embryonic cell cleavages proceed, a finer glass needle can similarly be used to transfer material to individual blastomeres, although this technique becomes impractical beyond the 1,000-cell stage without the use of more sophisticated iontophoresis equipment. Given the large clutch sizes and ease of injection of zebrafish, researchers can routinely inject hundreds of embryos in a single session. Because of the convenience of microinjection, the advent of chemically-modified antisense MO DNA oligonucleotides has allowed the selective knockdown of genes in the zebrafish embryo (Ekker, 2000). Injected at the 1-cell stage, MOs target either the translational start site or splice sites of a specific mRNA. Interaction of an MO and its target sterically blocks translation or splicing, effectively interfering with gene function. MOs are highly stable, so they remain effective for at least the first five days of embryogenesis, providing ample time to analyze the major phases of heart development. While MOs provide powerful tools for analysis of gene function, care must be taken in the interpretation of MO results. The efficacy of individual MOs varies widely and must be determined empirically. Moreover, MOs may have off-target effects and sometimes appear to cause nonspecific toxicity (Robu et al., 2007). The amount of MO injected should be carefully titrated to aid in gauging specific versus nonspecific effects. Importantly, multiple independent MOs, targeting different regions of a gene, should also be used to verify that an observed phenotype is gene-specific. In C. elegans, use of bacterial libraries encoding double-stranded RNA (dsRNA) has proven to be an effective means of knockingdown genes in a high-throughput manner (Simmer et al., 2003). In contrast to MOs, dsRNA does not currently seem to be effective for specific gene knockdown in zebrafish (Oates et al., 2000). As a complement to loss-of-function approaches like MO knockdown, gain-of-function experiments, in which wild-type genes or their modified forms are overexpressed, are often informative. In vitro transcribed RNA is readily translated following injection. When available, RNA encoding dominant-negative or constitutively active forms of a protein can be used for functional and epistasis analysis. Typically, gain-of-function experiments in zebrafish overexpress a gene throughout the entire embryo from the 1-cell stage. Unlike the case in the frog Xenopus laevis, the fates of early blastomeres in the zebrafish embryo are not

106

reproducible (Kimmel et al., 1990; Ho and Kimmel, 1993), so injections cannot specifically target the heart lineage. To achieve this, transgenic approaches are required. DNA transgenes injected at the 1-cell stage result in mosaic expression in F0 embryos. The zebrafish cardiac myosin light chain 2 (cmlc2) promoter drives strong expression specifically in cardiomyocytes (Huang et al., 2003). In F0 injected embryos it is feasible to achieve expression in a large proportion of the myocardium. To attain stable, nonmosaic transgenic lines, F0 embryos are raised to adulthood and screened for germline transmission of the transgene. Use of transposons to derive transgenic animals has been shown to increase the frequency of germline transmission greatly, and has become a favored method for transgenesis in zebrafish (Kawakami, 2005). Gene overexpression can often cause a lethal phenotype. Temporal control of transgene expression is therefore desirable. Additionally, temporally-regulated transgene expression can allow examination of when gene activity is required. The zebrafish hsp70 promoter provides a useful tool to conditionally overexpress gene products throughout the embryo (Halloran et al., 2000). This heat-shock promoter is quiescent at standard temperature (28.5°C), but is activated on short incubations at 37°C. Other methods, including the binary tetracycline-regulated and Gal4/UAS systems, provide additional means of temporally and spatially restricting transgene expression in zebrafish (Koster and Fraser, 2001; Huang et al., 2005). A further option is a recombination-based system, in which transgenic expression of Cre can modify a second transgene (Langenau et al., 2005). The future derivation of multiple regionally restricted cardiac promoters will increase the usefulness of these transgenic approaches for gene misexpression.

II.C.  Embryological Manipulation In order to analyze the cellular consequences of altering gene function in the zebrafish embryo, it is extremely beneficial to employ embryological techniques such as fate mapping and transplantation. The external development of the zebrafish embryo makes such embryonic manipulations possible. Embryonic cells can be individually labeled, tracked in real time, and transferred between embryos. When coupled with the availability of mutants, the use of MOs and the ability to misexpress genes, embryological approaches in zebrafish can reveal a wealth of information about the mechanisms regulating fate assignment and developmental potential. A fate mapping or lineage tracing approach is used to determine the progeny of a labeled cell or group of cells retrospectively. This is especially useful when molecular markers specific to a progenitor population do not exist, as is the case for myocardial progenitors in the early embryo prior to and during gastrulation (Fig. 1A). Putative

PART  |  1  Heart Evolution

progenitor cells can be labeled at early stages, and the fate, migration and growth of these cells can be determined from the locations and numbers of labeled cells found at later stages (Fig. 2A–D). Individual cells can be labeled via direct injection of fluorophore-conjugated dextrans. However, beyond the 1,000-cell stage, when microinjection of single cells is technically challenging, it is particularly useful to label cells with a caged fluorescein dextran conjugate; its fluorescence is inhibited by a chemical moiety until it is activated by UV light (Kozlowski et al., 1997) (Fig. 2A–D). Similarly, the photoconvertible protein Kaede changes its fluorescence to emit red rather than green light following UV exposure (Hatta et al., 2006). By finely focusing UV light via microscopy, individual cells or groups of cells in embryos injected with caged fluorescein or kaede RNA at the 1-cell stage can be labeled at any optically accessible stage and location and followed subsequently. These strategies have been employed to generate fate maps of myocardial and endocardial progenitors in wild-type zebrafish embryos (Stainier et al., 1993; Lee et al., 1994; Keegan et al., 2004; Schoenebeck et al., 2007) (Fig. 1A,B). Comparison of the wild-type fate map to fate maps in embryos with cardiac defects has been instrumental in elucidating whether and how particular genes or pathways impact cardiac fate assignment (Keegan et al., 2004, 2005; Schoenebeck et al., 2007). Transplantation approaches provide a powerful means to assay the potential of a cell in various environments (Carmany-Rampey and Moens, 2006) (Fig. 2E–H). Using a fine glass needle and mild suction, cells are removed from a donor embryo and placed into a host embryo, typically at midblastula stages. The donor embryo is labeled with a lineage tracer, such that the fate of donor-derived cells in the host embryo can be followed over time. This provides an assay to gauge the autonomy, or cellular requirement, of gene function. By placing mutant or wild-type cells into a host of the contrary genotype, the influence of environment on phenotype can be observed.

II.D.  Imaging Aside from providing aesthetic pleasure, the optical clarity and rapid development of the zebrafish embryo make it an ideal model organism for high-resolution imaging of the developing heart (Schoenebeck and Yelon, 2007). Internal organs are readily visualized via whole-mount in situ hybridization and immunohistochemistry. Online databases contain the results of high-throughput in situ analysis of thousands of genes, and numerous myocardial and endothelial markers exist for zebrafish (Thisse et al., 2004). Furthermore, the optical properties of the zebrafish embryo allow detailed cellular and subcellular examination of the cell biological traits of the developing heart. Myocardial cell polarity, cytoskeletal organization and cell

107

Chapter  |  1.4  Cardiac Development in the Zebrafish

(A)

(B)

40% epiboly

(C)

7 somites

(D)

21 somites

(E)

30 hpf

48 hpf

Figure 1  Major phases of heart formation in the zebrafish embryo. (A) Lateral view of the blastula at the 40% epiboly stage (5 hpf), depicting locations of ventricular (red) and atrial (yellow) progenitors as determined by fate mapping (Keegan et al., 2004). (B) Dorsal view, anterior up, at the 7-somite stage (12.5 hpf), depicting locations of ventricular and atrial progenitors in the lateral mesoderm as determined by fate mapping (Schoenebeck et al., 2007). (C) By the 21-somite stage (19.5 hpf), cardiac fusion has created a cardiac cone at the embryonic midline. Atrial (yellow) cardiomyocytes surround the ventricular (red) cardiomyocytes (Berdougo et al., 2003). (D) At 30 hpf, heart tube elongation is complete. (E) Frontal view, anterior up, at 48 hpf; the ventricle and atrium are morphologically distinct and asymmetrically looped. Artwork based on images from Schoenebeck and Yelon (2007).

(A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

Figure 2  Techniques for embryological manipulation in zebrafish. (A–D) Fate mapping using caged fluorescein dextran as a lineage tracer. (A) Caged fluorescein dextran is injected into the embryo at the 1-cell stage. The lineage tracer is inherited by all of the cells of the embryo, but does not fluoresce until it is photoactivated. (B) Photoactivation using UV light creates fluorescence in selected cells. In this example, five blastomeres (green) are selected at the margin of the embryo at 40% epiboly, as in the fate-mapping experiments of Keegan and colleagues (2004). (C) The location of photoactivated cells is recorded relative to anatomical landmarks or transgene expression patterns. In this view of the animal pole of the embryo at 50% epiboly, the longitudinal location (90°) of the photoactivated cells is recorded relative to the expression of Tg(gsc:gfp) at the dorsal midline. (D) The fate of photoactivated cells is determined at later developmental stages. Here, immunohistochemistry at 44 hpf detects fluorescein (blue) in two cardiomyocytes derived from the photoactivated cells. (E–H) Transplantation using fluorescein dextran as a lineage tracer. (E) Fluorescein dextran is injected into the donor embryo at the 1-cell stage. All of the cells of the donor embryo are therefore fluorescent. (F) Using mild suction, cells are removed from the donor embryo at the midblastula stage. (G) The removed cells are transferred to a host embryo. Placement near the margin of the host embryo facilitates contribution to cardiac lineages. (H) The fate of donor-derived cells is determined at later developmental stages. Only the donorderived cells contain the lineage tracer. In this example, two fluorescent cardiomyocytes are detectable at 44 hpf.

108

shape can be analyzed in fixed samples via immunofluorescence and confocal microscopy (Trinh and Stainier, 2004; Trinh et al., 2005; Rohr et al., 2006; Auman et al., 2007). Transgenic zebrafish that express cytoplasmic, nuclear or membrane-localized fluorescent proteins allow dynamic analysis of cell movement, cell number and cell shape in the living embryo (Rottbauer et al., 2002; Auman et al., 2007; Holtzman et al., 2007). It is particularly exciting to use transgenic zebrafish for time-lapse analysis during stages of heart development when considerable morphogenesis is occurring, including the assembly of the heart tube, the expansion of the cardiac chambers and the formation of the atrioventricular valve (Beis et al., 2005; Rohr et al., 2006; Holtzman et al., 2007). High-resolution imaging of the live heart is, of course, complicated by the heartbeat, which by 48 hpf occurs at a rate of roughly 180 beats per minute in the zebrafish embryo (Baker et al., 1997). Heartbeat can be temporarily inhibited by addition and removal of chemicals such as 2,3-butanedione monoxime and the anesthetic tricaine (Bartman et al., 2004). Although it is often desirable to stop the heart in order to collect images, it is also of interest to use imaging techniques to assess cardiac function. By combining high-speed microscopy and custom computational algorithms with transgenic embryos expressing fluorescent proteins in the myocardium, endothelium and circulating red blood cells, it is possible to measure fractional shortening, flow velocity, ejection fraction and additional functional parameters (Hove et al., 2003; Forouhar et al., 2006). Additionally, calcium indicator dyes and transgenic calcium sensor proteins can be used to assess cardiac conduction in the live embryo (Sedmera et al., 2003; Milan et al., 2006; Arnaout et al., 2007). These techniques are fundamental for analysis of the large number of zebrafish mutations that affect cardiac function. The ongoing development of more rapid microscopic techniques will allow more detailed analysis of cell morphology in a functional heart. One promising technique is selective plane illumination microscopy (SPIM), which is well-suited to imaging samples the size of the zebrafish heart (Huisken et al., 2004; Huisken and Stainier, 2007). Altogether, the confluence of genetic, embryological and imaging techniques available for zebrafish facilitates what has been termed an “in vivo cell biology” approach (Beis and Stainier, 2006). A variety of cell behaviors – including fate assignment, differentiation, proliferation, adhesion, polarity, movement, shape change, contractility and conduction – can be examined in the context of the developing embryo. This presents an ideal model for the direct analysis of how genes regulate heart development at the level of the cell. Thus, work in zebrafish can reveal key mechanisms by which mutations are translated into developmental defects and disease.

PART  |  1  Heart Evolution

III.  Mechanisms regulating zebrafish heart development In a relatively short period of time, our understanding of the mechanisms regulating zebrafish heart development has expanded dramatically. Large-scale genetic screens have yielded plentiful collections of mutations, and the characterization of mutant phenotypes and cloning of mutated genes have revealed key genetic pathways controlling cardiogenesis. In concert with these genetic studies, high-resolution imaging of the developing heart has provided insight into the fundamental cellular mechanisms underlying cardiac morphogenesis. In this section, we outline the major phases of the first two days of heart development in zebrafish, emphasizing our current understanding of the molecular and cellular regulation of each step.

III.A.  Overview of Stages of Zebrafish Development Following fertilization, the zebrafish embryo develops in a simple salt solution at 28.5°C (Kimmel et al., 1995). The embryo is found on top of a yolk, and is surrounded by a transparent chorion that provides protection. A series of rapid cell divisions converts the large single cell of the embryo into an unpatterned cap of over 1,000 small cells by 3 hpf. Maternally provided protein, RNA and other biomolecules play key roles in supporting these early cleavages. At the midblastula transition (roughly 3 hpf), zygotic transcription commences and is required for the initiation of gastrulation and further embryonic development. Morphogenesis of the embryo begins at dome stage (just after 4 hpf) as the yolk pushes upwards into the embryo, moving cells away from the inner core and creating a monolayer, termed the epiblast, that covers the upper 30% of the yolk. This is followed by epiboly – the spreading of the epiblast vegetally over the yolk. At 40% epiboly (5 hpf) gastrulation commences, with cells at the margin (the leading edge of the embryo over the yolk) involuting underneath the epiblast to form a new cell layer, termed the hypoblast. This process is critical to embryogenesis, as cells of the hypoblast will form the mesodermal and endodermal germ layers. Gastrulation movements consist of a combination of convergence towards the embryonic midline and extension along the future anterio–posterior (A–P) axis. The combination of these convergent extension (CE) movements converts the blastula into an organized, multi-layered embryo with discernible axes. Gastrulation and epiboly are completed in the zebrafish embryo at tailbud stage (10 hpf), by which time the yolk is completely engulfed. The embryo now has an obvious anteroposterior axis, with a presumptive head and tail, and the notochord is apparent at the embryonic midline. Between 10 and 24 hpf, a number of major developmental

Chapter  |  1.4  Cardiac Development in the Zebrafish

events occur. Rudimentary organs become evident; for example, the eye is visible by 12 hpf, and the otic vesicle forms by 14 hpf. Somite formation begins, with new somites being added initially at a rate of 2–3 somites per hour. At the same time, the embryo lengthens in the antero­ posterior dimension, primarily via lengthening of the tail. Heartbeat in the zebrafish embryo commences at 22 hpf, with blood flow apparent by 24 hpf. Between 24 and 48 hpf, the embryo straightens, with the head moving from a position curled over the yolk to be in line with the trunk axis. Spontaneous contractions of skeletal muscle become more frequent, and pigmentation arises in the eyes and body. The vascular network in the embryo becomes more elaborate over this time period, as angiogenesis results in a fine network of vessels branching from the primary vasculature in a stereotypic fashion. By 48 hpf, much of organogenesis is completed and the embryo is motile and responsive to external stimuli. Over the next 24 hours, the embryo hatches from its chorion and starts to inflate its swim bladder. By 120 hpf, the embryo has depleted the energy stores present in the yolk and begins eating. Sexual maturity is reached at 2.5–3 months of age. The strikingly rapid and external development of zebrafish makes it an excellent model for the study of cardiogenesis.

III.B.  Cardiac Progenitor Specification Fate maps of several vertebrate species indicate that cardiac progenitor cells ingress early during gastrulation, migrating to reach their destination in the anterior lateral plate mesoderm (ALPM) (Schoenwolf and GarciaMartinez, 1995; Tam et al., 1997). In the zebrafish embryo, extensive fate-mapping studies have determined the origins of cardiac progenitors prior to and following gastrulation (Stainier et al., 1993; Lee et al., 1994; Keegan et al., 2004; Schoenebeck et al., 2007). At early blastula stages (256 to 512 cells), progenitors are found at the ventro­ lateral margin, 90° to 270° from the dorsal midline of the embryo (Stainier et al., 1993). The multipotent myocardial progenitor cells are intermingled with cells fated to form other tissues, including muscle and blood. Endocardial progenitor cells are also found near the margin, at more ventral positions closer to 180° (Lee et al., 1994). In the early blastula, individual myocardial progenitor cells can contribute to both the atrium and ventricle, indicating that cardiac chamber lineages are not yet separated at this stage of development (Stainier et al., 1993). After an additional hour of development at midblastula stage (1,000–2,000 cells), individual atrial and ventricular myocardial progenitor cells can be identified via fate mapping, indicating lineage separation by this time. At the onset of gastrulation, the distribution of myocardial progenitors has shifted to encompass bilateral regions 60°–140° from the dorsal midline (Keegan et al., 2004)

109

(Fig. 1A). Roughly 10–20% of blastomeres within the first three rows next to the embryonic margin in this area will give rise to myocardial progeny. Organization of the atrial and ventricular progenitors is evident at this stage of development. Progenitors of ventricular myocardium are found closer to the margin and the shield, whereas those of atrial fate are found further from the margin in more ventral positions (Fig. 1A). In contrast, endocardial progenitors do not demonstrate a chamber-specific organization, and individual cells can contribute to the endocardium of both chambers. The spatial organization of ventricular and atrial myocardial progenitors persists after gastrulation, when they occupy discrete regions of the ALPM (Schoenebeck et al., 2007) (Fig. 1B). Fatemapping experiments indicate that ventricular progenitors occupy a more medial portion of the ALPM than atrial progenitors (Fig. 1B), and this mediolateral organization correlates well with the subsequent complementary mediolateral expression patterns of the chamber-specific genes ventricular myosin heavy chain (vmhc) and atrial myosin heavy chain (amhc) (Yelon et al., 1999; Berdougo et al., 2003). The regions of the early embryo containing myocardial progenitors qualify as heart fields, areas of the embryo possessing cardiac developmental potential (Jacobson and Sater, 1988). Transplantation of donor cells from noncardiogenic regions of the blastula into the heart fields of host embryos can confer myocardial fate, although not with 100% efficiency (Lee et al., 1994; Scott et al., 2007). Thus, the signals available to cells within the heart fields are clearly important for cardiac specification. However, as other types of mesendodermal progenitor cells also arise from these regions, additional signals must be critical for commitment to cardiac lineages. These signals may be delivered at later stages, as the progenitor cells migrate to and reside within the ALPM. Studies of zebrafish mutations have implicated multiple signaling pathways in the regulation of cardiac specification within the heart field. For example, mutations in components of the Bmp and Fgf signaling pathways indicate that these pathways are essential for promoting myocardial fate assignment. In swirl (bmp2b) and acerebellar (fgf8) mutants, the number of cardiomyocytes is severely decreased compared to their wild-type siblings; this myocardial deficiency is preceded by gene expression defects, including reduced expression of the transcription factor gene nkx2.5, thought to be an early marker of myocardial progenitors (Reifers et al., 2000; Reiter et al., 2001). A similar phenotype is exhibited by one-eyed pinhead mutants, which lack an essential coreceptor in the Nodal signaling pathway (Reiter et al., 2001). Furthermore, in acerebellar (fgf8) and one-eyed pinhead mutants, the effects on ventricular cardiomyocytes are more dramatic than the effects on atrial cardiomyocytes, suggesting that Fgf8 and Nodal signaling may be involved in chamber fate assignment. Indeed, fate-map analysis in embryos

110

PART  |  1  Heart Evolution

with reduced levels of Nodal signaling supports a model in which differential exposure to Nodal ligands, secreted from the embryonic margin, promotes the assignment of ventricular identity in the myocardial progenitors closest to the margin (Keegan et al., 2004). Integration of Bmp, Fgf and Nodal signaling is likely to regulate expression of pivotal transcription factor genes that drive cardiomyocyte differentiation. Mutation of the transcription factor gene faust (gata5) results in reduced expression of nkx2.5 and severe cardiomyocyte deficiencies, especially for ventricular cardiomyocytes (Reiter et al., 1999). Epistasis experiments suggest that Gata5 functions downstream of Bmp2b and Nodal signaling (Reiter et al., 2001); although Fgf8 is also necessary for normal expression of nkx2.5, it does not seem to be essential for gata5 expression. Mutants lacking the transcription factor gene hands off (hand2) also have too few cardiomyocytes but do not exhibit defects in nkx2.5 or gata5 expression, suggesting a downstream or parallel role for Hand2 (Yelon et al., 2000). Fate mapping in hands off (hand2) mutants reinforces this notion: the dimensions of the heart fields appear unchanged in hands off (hand2) mutants, yet production of differentiated cardiomyocytes from these fields is inefficient (Schoenebeck et al., 2007). The inductive signals that promote cardiac specification are counterbalanced by repressive signals that restrict the formation of cardiac progenitors. In contrast to zebrafish mutations causing a shortage of cardiomyocytes, mutation of neckless (retinaldehyde dehydrogenase 2) causes

a significant cardiomyocyte surplus (Keegan et al., 2005). Retinaldehyde dehydrogenase 2 controls the rate-limiting step of retinoic acid (RA) synthesis; therefore, the neckless mutant phenotype, together with the phenotype of embryos treated with inhibitors of RA signaling (Fig. 3A–D), suggests that RA signaling is critical for restricting cardiomyocte production. Furthermore, fatemap analysis in embryos with reduced RA signaling has shown that RA signaling during gastrulation stages is essential for setting limits on the number of myocardial progenitors. Canonical Wnt signaling also appears to inhibit cardiac specification, although its influence varies at different developmental stages (Ueno et al., 2007). Use of heat-shock inducible transgenes that activate or inhibit Wnt/-catenin signaling has demonstrated that signaling during gastrulation blocks expression of nkx2.5, whereas signaling at earlier stages promotes nkx2.5 expression. In addition to the RA and canonical Wnt pathways, the pathways that control endothelial and hematopoietic specification contribute to the restriction of cardiac specification (Schoenebeck et al., 2007). Zebrafish mutants deficient in vessel and blood lineages, such as cloche, exhibit ectopic cardiomyocytes and enlarged hearts. Fate mapping indicates that this phenotype is caused by fate transformation of a territory located adjacent to the heart field; normally, this region produces vessel and blood progenitors but not cardiac progenitors. It is not yet known how the multiple repressive influences on cardiac specification work in opposition to the inductive influences on this process. Future studies will illuminate

(A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

Figure 3  Examples of signaling pathways regulating cardiac progenitor specification in zebrafish. (A–D) RA signaling restricts the number of cardiomyocytes in the zebrafish embryo. (A, B) Dorsal views, anterior to the top, comparing cmlc2 expression at the 16-somite stage (17 hpf) in wild-type (wt) embryos and siblings treated with the RA receptor antagonist BMS189453. Inhibition of RA signaling creates a significant surplus of cardiomyocytes residing in bilateral populations in the ALPM. Images adapted from Keegan et al. (2005). (C, D) Frontal views of hearts at 48 hpf; both chambers express the transgene Tg(cmlc2:DsRed2-nuc) (red; Mably et al., 2003) and atria are labeled with the anti-atrial myosin heavy chain (Amhc) antibody S46 (green). Inhibition of RA signaling causes the formation of enlarged hearts containing an excess number of cardiomyocytes. Images courtesy of J. S. Waxman and D. Yelon (unpublished data). (E–H) Agtrl1b signaling promotes cardiomyocyte formation in the zebrafish embryo. (E, F) Dorsal views, anterior to the top, comparing nkx2.5 expression at the 10-somite stage (14 hpf) in wt and grinch (agtrl1b) mutant embryos. Loss of Agtrl1b signaling inhibits formation of nkx2.5-expressing myocardial progenitors. (G, H) Lateral views, anterior to the left, of cmlc2 expression at 48 hpf. Inhibition of Agtrl1b signaling results in a severe deficiency of cardiomyocytes. Images adapted from Scott et al. (2007).

111

Chapter  |  1.4  Cardiac Development in the Zebrafish

how the network of inductive and repressive pathways is integrated before, during and after gastrulation stages to establish an appropriate number of cardiac progenitors. It is unclear at present if myocardial fate assignment is simply a matter of migration to the proper region in the ALPM. The path traveled by mesendodermal progenitor cells during gastrulation may be important to ensure their exposure to the appropriate mixture of inductive and repressive signals over a suitable length of time. The complementary organization of the myocardial fate map in the blastula and the myocardial fate map in the ALPM may reflect organized patterns of cell movements during gastrulation. Several signals are known to be required for efficient gastrulation (Rohde and Heisenberg, 2007). However, the regulation of migration patterns of specific progenitor populations during gastrulation is not yet understood. In the grinch mutant, myocardial progenitors are lost, as evidenced by an absence of nkx2.5 expression (Scott et al., 2007) (Fig. 3E–H). The grinch locus encodes the G proteincoupled receptor (GPCR) Agtrl1b, a relative of the chemokine GPCRs. Transplantation experiments demonstrate that Agtrl1b signaling is required autonomously in myocardial progenitors during the initial phases of gastrulation (Scott et al., 2007). MO knockdown of the Agtrl1b ligand Apelin inhibits directed cell migrations during gastrulation (Zeng et al., 2007). Agtrl1b signaling may therefore represent a key pathway that regulates migration of cardiac progenitors to the correct position in the ALPM for the initiation of cardiogenesis.

cone, through a process termed cardiac fusion (Fig. 1C). Cardiac fusion begins with contacts between posterior subsets of the contralateral cardiomyocyte populations; then, connections between anterior subsets of contralateral cells complete formation of the cone (Fig. 4A–C). Time-lapse analysis has indicated that particular angular movements of the most anterior and posterior cardiomyocytes create the specific circumference of the myocardial cone, engulfing the centrally located endocardial precursors (Bussmann et al., 2007; Holtzman et al., 2007). At 20 hpf (22-somite stage), the cardiac cone is a lumenized structure projecting in a dorsoventral direction (Glickman and Yelon, 2002). Atrial cardiomyocytes reside in its broad base, and ventricular cardiomyocytes are found more dorsally and medially in the apex of the cone (Fig. 1C). The cone then tilts posteriorly, placing the ventricle posterior to the atrium. In this orientation, the cone gradually elongates, generating a primitive myocardial tube

(A)

(B)

(C)

(D)

(E)

(F)

(G)

III.C.  Heart Tube Assembly By 16 hpf (14-somite stage), myocardial differentiation is underway within the bilateral heart fields of the zebrafish ALPM. The differentiating cardiomyocytes, while not yet contractile, begin expressing genes encoding sarcomere components, including cardiac myosin light chain 2 (cmlc2) in all cardiomyocytes, vmhc in ventricular cardiomyocytes, and amhc in atrial cardiomyocytes (Yelon et al., 1999; Berdugo et al., 2003). Between 16 and 20 hpf, these bilateral sheets of cells, sandwiched between the extraembryonic yolk syncytial layer (YSL) on their ventral side and the anterior endoderm on their dorsal side, migrate toward the embryonic midline, where they will begin to assemble the heart tube (Fig. 1C,D). Time-lapse studies utilizing a myocardial reporter transgene have revealed that cardiomyocytes migrate medially in an orderly and predictable manner (Holtzmann et al., 2007) (Fig. 4A–C). The coherent pattern of cell behavior is consistent with the finding that the migrating cardiomyocytes begin to form a polarized epithelium as they approach the embryonic midline (Trinh and Stainier, 2004) (Fig. 4D,E). As the bilateral cardiomyocyte populations approach each other, they create a ring of cells referred to as the cardiac

Figure 4  Medial migration of cardiomyocytes is an orderly, coherent process. (A–C) Images from a time-lapse study of cardiac fusion in a wild-type embryo expressing the myocardial reporter transgene Tg(cmlc2:egfp). Dorsal views, anterior to the top, at the (A) 16-somite; (B) 18-somite; and (C) 20-somite stages. Cardiac fusion begins with contacts between posterior cardiomyocytes (B) and concludes with contacts between anterior cardiomyocytes (C). Images adapted from Holtzman et al. (2007). (D–G) The migrating myocardium is a polarized epithelium, and its integrity requires Fibronectin function. (D, F) Transverse sections, dorsal to the top, of wt and natter (fibronectin) mutant embryos at the 20-somite (19 hpf) stage. (E, G) Magnified views of the right side of each section. Expression of Tg(cmlc2:egfp) (pseudocolored blue) indicates bilateral locations of cardiomyocytes, and immunohistochemistry for -catenin (red) and PRKCi (green) indicates protein localization. (D, E) The myocardium exhibits apicobasal polarity, with -catenin localized basolaterally and PRKCi localized apicolaterally. (F, G) Loss of Fibronectin disrupts myocardial polarity and causes cardia bifida. Images adapted from Trinh and Stainier (2004).

112

(Fig. 1D). At the same time, the endocardial precursors migrate from their central position to generate an endothelial lining for the tube (Bussmann et al., 2007). The geometry of the developing heart is tightly confined by its position above the yolk and beneath the head. With resorption of the yolk, the atrium, attached to the venous inflow, shifts to a more posterior position, leaving the heart in its familiar anteroposterior orientation, with the ventricle positioned more anteriorly (Stainier and Fishman, 1992). Genetic analysis has revealed that the endoderm, YSL and endocardium all contribute to regulation of the dynamic patterns of cell behavior that drive cardiomyocyte migration and cardiac fusion. Inhibition of cardiomyocyte migration results in the formation of two separate hearts in lateral positions, a phenotype known as cardia bifida (Glickman and Yelon, 2002). Remarkably, bifid hearts can each contain a ventricle and an atrium (Yelon et al., 1999); cardiac fusion therefore appears dispensable for myocardial differentiation, even though it is essential for proper alignment with the vascular system. Mutations in casanova (sox32), bonnie and clyde, faust (gata5) and one-eyed pinhead all disrupt endoderm specification and cause cardia bifida, suggesting that the endoderm provides an important signal or substrate utilized by migrating cardiomyocytes (Peyrieras et al., 1998; Reiter et al., 1999; Kikuchi et al., 2000, 2001). The molecular underpinnings of endodermal-myocardial interactions remain mysterious, although these may somehow involve sphingolipid sig­ naling, since transplantation experiments indicate that the sphingosine-1-phosphate receptor gene miles apart (edg5) plays a cell nonautonomous role in promoting myocardial migration (Kupperman et al., 2000). In addition to putative interactions with the endoderm, migrating cardiomyocytes require interactions with components of the extracellular matrix, particularly Fibronectin, which is deposited by the YSL. This requirement is demonstrated by the cardia bifida phenotype of natter (fibronectin) mutants (Trinh and Stainier, 2004). Mutation of natter (fibronectin) also disrupts apicobasal polarity in cardiomyocytes (Fig. 4F,G), suggesting that formation of a polarized epithelium is an important prerequisite for the coordination of myocardial migration (Trinh and Stainier, 2004; Trinh et al., 2005). Once interactions with the endoderm and YSL have recruited cardiomyocytes toward the embryonic midline, interactions between the myocardium and the endocardium regulate the cell behaviors that create the specific shape of the cardiac cone (Holtzman et al., 2007). Mutation of cloche, which eliminates the endocardium, alters the pattern of cardiomyocyte movements during cardiac fusion. Although cloche mutant embryos do not display cardia bifida, the cloche mutant cardiac cone has a significantly distorted morphology. Although the molecular nature of the relevant myocardial-endocardial interactions is not yet clear, it seems that communication between myocardium and endocardium controls the induction, direction and

PART  |  1  Heart Evolution

duration of angular cardiomyocyte movements during the final phase of cardiac fusion. Thus, the endocardium helps to organize the myocardium into a configuration appropriate for the assembly of the heart tube. Much less is known about the mechanisms that drive the tilting and elongation of the cardiac cone to create the heart tube. One emerging theme is that the regulation of apicobasal polarity is critical during heart tube elongation, just as it is during cardiomyocyte migration. Mutations in the genes heart and soul (prcki), snakehead (atp1a1a. 1) and nagie oko (mpp5) inhibit heart tube elongation; mutant hearts appear as arrested cardiac cones or stunted heart tubes (Yelon et al., 1999; Horne-Badovinac et al., 2001; Peterson et al., 2001; Shu et al., 2003; Yuan and Joseph, 2004; Rohr et al., 2006). PRKCi and Mpp5 are components of apically-localized protein complexes, and analysis of prkci and mpp5 mutants demonstrates that both genes are required cell-autonomously for normal apicobasal polarity of the myocardium and heart tube elongation (Rohr et al., 2006) (Fig. 5). The function of Atp1a1a.1 also relates to the maintenance of myocardial apicobasal polarity, since its activity as an ion pump is required for the maintenance of junctional belts that connect the myocardial epithelium (Cibrian-Uhalte et al., 2007). Together, these data point to the importance of organization and coherence of the myocardium during the transformation of the cardiac cone into the heart tube.

III.D.  Morphogenesis of the Cardiac Chambers and Atrioventricular Cushions Between 24 and 48 hpf, the simple heart tube gradually transforms into a two-chambered organ, with an atrium and a ventricle separated by the constriction of the atrioventricular canal (Fig. 1D,E). The zebrafish heart tube is initially positioned with its ventricular end pointing toward the right side of the embryo and its atrial end pointing toward the left side of the embryo (Fig. 1D). Left–right asymmetry of the heart is maintained as the tube loops to create an S-shaped structure (Fig. 1E). By 36 hpf the ventricle is clearly displaced to the right of the atrium and an obvious pinching appears at the atrioventricular (AV) canal. While cardiac looping is underway, chamber curvatures emerge through a process called ballooning, in which localized bulges deform the cylindrical wall of the heart tube (see Chapter 3.2, Vol. I). By 48 hpf each expanded chamber exhibits two characteristic curvatures: a bulging curvature called the outer curvature (OC); and a recessed curvature called the inner curvature (IC). Recent studies have started to uncover the cellular mechanisms responsible for zebrafish chamber morphogenesis. In particular, analysis of cardiomyocyte morphology and organization in the zebrafish heart has demonstrated that regional changes in cell size and shape are associated

113

Chapter  |  1.4  Cardiac Development in the Zebrafish

(A)

(C)

(E)

(B)

(A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

(I)

(D)

(F)

Figure 5  Heart tube elongation requires normal apicobasal polarity in the myocardium. Dorsal views, anterior to the top, at the (A, C, E) 20somite and (B, D, F) 28-somite (23 hpf) stages, comparing myocardial morphology of (A, B) wt embryos with that of embryos injected with (C, D) anti-prkci MO or (E, F) anti-mpp5 MO. Expression of Tg(cmlc2:egfp) (green) indicates arrangement of cardiomyocytes, and immunohistochemistry for PRKCi (red) and ZO-1 (blue) facilitates visualization of the midline. In wt embryos, tilting and elongation of the cardiac cone (A) creates a primitive heart tube (B) by the 28-somite stage. (C, D) Loss of PRKCi blocks both tilting and elongation, and (E, F) loss of Mpp5 delays cardiac fusion and blocks further steps of heart tube assembly. Images adapted from Rohr et al. (2006).

with curvature formation in the ventricle (Auman et al., 2007). Within the linear heart tube, all ventricular cardiomyocytes are similarly small and round (Fig. 6A). As chamber curvatures emerge, both OC and IC cells increase their surface area (Auman et al., 2007). Notably, however, only OC cells significantly change their shape, becoming flattened and elongated (Fig. 6B), whereas IC cells maintain a cuboidal morphology (Fig. 6C). Thus, regionally-confined cell shape changes underlie the acquisition of chamber morphology; the elongation and orientation of OC cells, coupled with the cuboidal shape of IC cells, create the characteristic curvatures of the expanded ventricle. The dynamic cellular remodeling of the heart tube is essential for creating chambers of appropriate functional

Figure 6  Cardiac function influences morphogenesis of the cardiac chambers and atrioventricular cushions. (A–C) Regionally-confined cell shape changes underlie the emergence of chamber curvatures in the ventricle. Live wt hearts expressing Tg(cmlc2:egfp) and exhibiting mosaic expression of Tg(cmlc2:dsredt4). Arrows point to representative cells expressing both dsredt4 and egfp. Ventricular cells in the linear heart tube (LHT) at 28 hpf (A) and in the IC at 52 hpf (C) are relatively cuboidal, whereas cells in the OC at 52 hpf (B) are flattened and elongated. Images adapted from Auman et al. (2007). (D–F) Cardiomyocyte cell shape changes are regulated by cardiac function. Live hearts expressing Tg(cmlc2:egfp) and exhibiting mosaic expression of Tg(cmlc2:dsredt4) at 52 hpf. Arrows point to representative ventricular OC cells expressing both dsredt4 and egfp. In contrast to the enlargement and elongation of OC cells in wt embryos (D), when blood flow is reduced, as in amhc mutant embryos (E), ventricular cardiomyocytes, including the OC cells, remain small and fail to elongate normally. (F) In vmhc mutant embryos, ventricular cardiomyocytes, including the OC cells, are excessively enlarged and elongated. Images adapted from Auman et al. (2007). (G– I) AV cushion formation is regulated by cardiac function. Lateral views of live embryos, anterior to the left, expressing the endothelial reporter transgene Tg(tie2:gfp) at 48 hpf. Red arrows point to the AV canal. (G) In wt embryos, enhanced expression of Tg(tie2:gfp) is visible in the AV endocardial cushions. In silent heart (sih; tnnt2) mutant (H) and cardiofunk (cfk; actc1) mutant embryos, cardiac contractility is disrupted and AV cushions do not form. Images adapted from Bartman et al. (2004).

capacity, yet this entire process must occur while the heart is beating. Interestingly, studies of zebrafish mutants with functional deficiencies have demonstrated that cardiac function has a potent influence on chamber curvature formation. For example, mutation of the weak atrium (amhc) locus, which encodes an atrium-specific myosin heavy chain, disrupts atrial contractility and, consequently, hinders blood flow (Berdougo et al., 2003). Additionally, weak atrium (amhc) mutants have significant ventricular defects; their ventricular cardiomyocytes fail to enlarge or elongate normally (Fig. 6D,E), resulting in an abnormally small and round ventricle (Auman et al., 2007). Since amhc is not expressed in the ventricle, this ventricular

114

phenotype is presumed to reflect a potent secondary effect of diminished blood flow on the cell shape changes that normally create chamber curvatures. The precise mechanism by which blood flow promotes myocardial cell shape change remains unclear. One possibility is that the shear forces produced by blood flow result in mechanosensation by the endocardium which then transmits another set of signals to the myocardium. Several zebrafish mutant phenotypes provide evidence for the importance of endocardial–myocardial signaling during chamber formation. The cloche mutation, which blocks endocardium formation, also causes dysmorphic, dilated chambers (Stainier et al., 1995; Liao et al., 1997). The heart of glass, santa (krit1) and valentine (ccm2) genes are also critical for regulating chamber morphology; mutation of any of these genes causes severe chamber dilation (Mably et al., 2003, 2006). Since all three of these genes are expressed in endothelial cells, these mutant phenotypes again implicate endocardial–myocardial signaling in the control of chamber shape. Whereas blood flow encourages cells to enlarge and elongate, the inherent contractility of a cardiomyocyte seems to limit its degree of shape change. Mutation of the pickwick locus, encoding the giant scaffold protein Titin and mutation of the half-hearted locus, encoding the ventricle-specific myosin heavy chain Vmhc, both result in distortion of cardiomyocyte morphology (Xu et al., 2002; Auman et al., 2007) (Fig. 6F). Transplant experiments have demonstrated that these proteins are required cell-autonomously for maintenance of cardiomyocyte cell shape, indicating that the integrity of the contractile apparatus plays an important role in chamber morphogenesis. Altogether, the phenotypes of zebrafish mutants with both chamber shape defects and functional defects suggest that the acquisition of normal cardiomyocyte morphology requires a balance between external physical forces, such as blood flow, and internal physical forces, such as contractility. Future studies are likely to focus on the mechanisms for the mechanotransduction of hemodynamic forces and on the cytoskeletal attributes that are linked to sarcomere integrity; potential interplay between the relevant pathways may provide the basis for establishing and maintaining cardiomyocyte cell shape during chamber curvature formation. At the same time as chamber curvatures are emerging, AV valve morphogenesis is underway (Beis et al., 2005) (see Chapters 6.1 and 6.2). Valve formation initiates within the AV canal, a constriction of the heart tube found at the boundary between the atrium and the ventricle. Gene expression patterns also distinguish the developing AV canal; expression of bmp4 and versican becomes restricted to the AV myocardium and AV endocardial cells exhibit upregulation of endothelial reporter transgenes (Fig. 6G), restricted expression of notch1b and characteristic lateral localization of the adhesion molecule Dm-grasp (Walsh and Stainier, 2001; Beis

PART  |  1  Heart Evolution

et al., 2005). Additionally, myocardial and endocardial cells in the AV canal begin to exhibit cell morphologies distinct from those in the flanking chambers. Notably, the AV endocardial cells transition from a squamous to a more cub­oidal appearance and aggregate to establish endocardial cushions (Beis et al., 2005). Subsequent remodeling events convert these cushions into the leaflets of the AV valve. A variety of zebrafish mutations provide insight into the mechanisms responsible for patterning the AV canal and executing AV valve morphogenesis. For example, jekyll (udp-glucose dehydrogenase) mutants have defects in AV canal patterning; they fail to restrict expression of bmp4, versican and notch1b, and do not form endocardial cushions (Walsh and Stainier, 2001). The enzymatic activity of UDP-glucose dehydrogenase (Ugdh) is required for the modification of extracellular matrix proteins that facilitate Wnt and Fgf signaling (Lander and Selleck, 2000), suggesting a mechanism by which Ugdh might regulate the signal transduction that is responsible for AV patterning. In particular, the jekyll mutant phenotype might reflect a defect in Wnt signaling; activation of Wnt signaling via mutation of the tumor suppressor gene apc leads to excessive endocardial cushion formation beyond the boundaries of the AV canal, and inhibition of Wnt signaling blocks endocardial cushion formation (Hurlstone et al., 2003). In addition to requiring proper AV canal patterning, AV valve morphogenesis, like chamber morphogenesis, requires input generated by biomechanical forces. Mutation of genes encoding contractile proteins, such as the cardiac troponin T gene silent heart (tnnt2) and the cardiac actin gene cardiofunk (actc1), blocks endocardial cushion formation, suggesting that either shear forces, resulting from blood flow, or a stretch response, resulting from contractility, trigger valve morphogenesis (Bartman et al., 2004; Beis et al., 2005) (Fig. 6G–I). The role of shear forces in promoting valve development has also been demonstrated by experiments implanting large beads into embryos; when the implanted bead obstructs blood flow, endocardial cushions do not form (Hove et al., 2003). However, not all aspects of AV canal development are dependent on cardiac function. Restriction of bmp4 expression to the AV canal proceeds normally, even when contractility is defective (Bartman et al., 2004). Also, formation of specialized slow conductive tissue in the AV myocardium, which is induced by endocardial-myocardial signaling, does not require cardiac contractility (Milan et al., 2006). Together, studies of chamber and valve formation in zebrafish highlight the important interrelationship of form and function, chamber and valve morphogenesis are both required for, and dependent on, normal cardiac function. This suggests numerous etiologies for congenital heart defects, since small alterations in cardiac function could easily influence cardiac morphology, and vice versa. The 24 hpf heart tube has been described as a peristaltic pump, allowing blood flow via unidirectional propagation of

Chapter  |  1.4  Cardiac Development in the Zebrafish

contractions from the atrial pole to the ventricular pole. However, recent examination of contractions and blood flow using high-speed confocal microscopy has suggested that the early heart tube acts instead as a suction pump (Forouhar et al., 2006). In this case, the precise engineering required to allow the heart tube to pump blood properly implies that even mild perturbations in morphogenesis will have dramatic effects on cardiac performance, leading to potentially devastating consequences for cardiac development.

IV.  Use of zebrafish as models of heart disease Despite the great progress made in identifying mutations associated with heart disease, heart dysfunction remains the leading cause of death in the Western world. Congenital heart disease is found in 1–2% of live births, with complex fetal or postnatal surgery in some cases being required for survival (Ransom and Srivastava, 2007). Additionally, cardiomyopathies and arrhythmias in adults often have an underlying genetic origin (Keating and Sanguinetti, 2001; Towbin and Bowles, 2002). Identification of mutations that cause or increase the likelihood of heart disease provides a means to pinpoint individuals at risk via genetic analysis. This offers an opportunity for the initiation of measures to prevent or mitigate the severity of disease. However, curative approaches are not currently available. It is difficult at present to envisage gene therapy as an alternative to repair genetic defects in all cells of the heart. Therefore, a greater mechanistic understanding of how heart disease arises from genetic mutations is the current priority. This approach may indicate pathways to target for pharmacological or dietary intervention. Congenital heart disease primarily involves morphogenetic defects. These include incomplete septation of the chambers, defects in valve development, hypoplastic chambers and anomalies in outflow tract septation/patterning (Gruber and Epstein, 2004). The zebrafish does not require pulmonary circulation, and therefore the zebrafish heart is much simpler in design than the mammalian heart. In zebrafish, the single ventricle and atrium lack septa, and the outflow tract consists of a simple bulbus arteriosus. However, the zebrafish bulbus is invested with smooth muscle (Grimes et al., 2006), as is observed in the mammalian arterial pole. Additionally, valve formation in zebrafish appears to occur via a similar mechanism as in mammals (Chang et al., 2004; Beis et al., 2005), and the zebrafish ventricular myocardium forms a compact layer and trabeculae similar to those seen in mammals (Hu et al., 2000, 2001). Despite the relative simplicity of the zebrafish heart, the genes that regulate heart development appear conserved. There are multiple examples in which mutations

115

of genes associated with human congenital heart disease cause cardiac developmental defects in zebrafish. The zebrafish heartstrings mutation disrupts tbx5, a gene associated with Holt-Oram syndrome (Basson et al., 1997; Li et al., 1997; Garrity et al., 2002). Similarly, mutation or MO knockdown of congenital heart disease genes such as tbx1, gata4, fog1 and raf1 cause early developmental heart defects in zebrafish (Piotrowski et al., 2003; Walton et al., 2006; Peterkin et al., 2007; Razzaque et al., 2007). It therefore seems likely that the same molecular pathways enact common morphogenetic mechanisms in both the human and zebrafish heart. Since zebrafish cardiac morphogenesis can be observed in real time, this model presents an excellent tool to study how congenital heart disease genes regulate heart development and to test the roles of candidate disease genes. Forward genetic screens and the use of MOs in zebrafish have identified a number of phenotypes featuring physiological defects in which cardiac contractility or rhythm is compromised (Rottbauer et al., 2001; Sehnert et al., 2002; Xu et al., 2002; Berdougo et al., 2003; Langheinrich et al., 2003; Bartman et al., 2004; Ebert et al., 2005; Langenbacher et al., 2005; Rottbauer et al., 2006; Arnaout et al., 2007). Interestingly, many of these mutations affect human familial cardiomyopathy or arrhythmia genes, such as those encoding Titin or Kcnh2 (Xu et al., 2002; Langheinrich et al., 2003; Arnaout et al., 2007). While cardiomyopathies and arrhythmias are not strictly defined as developmental disorders, mutation of disease-associated genes causes profound defects in the zebrafish embryo. Given the experimental tractability of the zebrafish, these mutations provide excellent models to study the signals that lead to heart failure in humans. From the viewpoint of considering mechanisms for cardiac repair, it is also important to note that the zebrafish heart has tremendous regenerative capacity (Poss et al., 2002). While the precise cellular and molecular basis for regeneration is not yet known, it appears that novel cardiomyocytes are generated to repair the injured zebrafish heart (Lepilina et al., 2006). A greater understanding of how this process is regulated in zebrafish may have significant implications for the repair of damage caused by heart failure in humans. Aside from providing a means to study the mechanisms leading to disease, animal models are also useful for the evaluation of possible therapeutic interventions. The great success of a folic acid-rich diet in preventing the complex genetic spectrum of neural tube closure defects suggests that other congenital anomalies may also benefit from chemical intervention. Chemical genetics, in which large libraries of compounds are screened for specific biological effects, is one promising approach toward identifying useful small molecules. The zebrafish embryo provides an excellent in vivo model for this technique (Murphey and Zon, 2006). Successful chemical genetic screens in zebrafish have identified compounds that regulate

116

processes including pigmentation, cell-cycle progression, angiogenesis and heart rate (Milan et al., 2003; Peterson et al., 2004; Murphey et al., 2006; Ni-Komatsu and Orlow, 2007). In an example utilizing an automated method to measure heart rate, drugs known to cause long QT syndrome in humans almost uniformly produced bradycardia and AV block in zebrafish embryos (Milan et al., 2003). Therefore, as with genetic pathways, drug effects may be conserved from zebrafish to human. Chemical suppression of a mutant cardiovascular phenotype has been demonstrated in the case of the zebrafish gridlock mutant (Peterson et al., 2004). Mutation of gridlock (hey2) causes aortic coarctation, resulting in the absence of blood flow to the trunk. Screening of several thousand chemicals revealed two modulators of VEGF signaling that restored proper blood flow to gridlock mutants. While it is tempting to speculate that compounds able to rescue zebrafish mutant phenotypes could be clinically applicable, this may be a naïve hope. Zebrafish null mutants are not equivalent to patients heterozygous for a particular mutation. Furthermore, many compounds are likely to have undesirable off-target effects. Even so, future pursuit of this strategy will provide molecular inroads to elaborate promising pathways with curative potential, further strengthening the utility of the zebrafish as a model of heart disease.

PART  |  1  Heart Evolution

of how the complex terrestrial heart evolved from the simpler aquatic organ. Studies in this regard are likely to focus on how the multiple heart fields found in amniotes (Buckingham et al., 2005) relate to the zebrafish heart fields; it is not yet clear whether zebrafish possess a second heart field equivalent (see Chapter 2.2). It will also be fascinating to examine whether advanced morphogenetic processes such as chamber septation are regulated by mechanisms similar to those driving zebrafish heart tube assembly and chamber formation. Overall, it is probable that the commonalities between vertebrate hearts will ultimately outweigh their differences, at least in terms of the fundamental mechanisms regulating many cardiac cell behaviors that are highly relevant to the etiology of human disease.

Acknowledgments We thank S. Abdelilah-Seyfried, T. Bartman, D. Stainier, S. Rohr, L. Trinh and J. Waxman for providing images for the figures. We regret having to limit or omit discussion of many interesting studies due to space constraints. Work in the Scott laboratory is supported by funding from the Canadian Institutes of Health Research, the Natural Sciences and Engineering Research Council of Canada, and the Canada Foundation for Innovation; work in the Yelon laboratory is supported by funding from the National Institutes of Health, the American Heart Association and the March of Dimes.

V.  Conclusions The past decade has been a highly productive period for the study of zebrafish heart development. By combining multiple genetic and embryological experimental appro­ aches, investigators have taken considerable strides toward elucidating the mechanisms that drive heart patterning and morphogenesis, including and in addition to the work encapsulated here. Notably, the ability to study development at the cellular level in the zebrafish embryo has been especially useful for linking essential genes to the specific cell behaviors that they regulate. As the number of investigators using zebrafish increases and the sophistication of zebrafish technology advances, the pace of discovery will undoubtedly continue to accelerate. In particular, emerging methods for targeted gene inactivation, enhanced strategies for spatial and temporal manipulation of gene function, and new techniques for automated image analysis are likely to facilitate high-throughput analysis of phenotypes generated by forward, reverse and chemical genetics. Importantly, future work is also likely to provide a clearer resolution of the parallels and contrasts between cardiogenic mechanisms in zebrafish and other vertebrates. It is certain that some aspects of heart development have changed over the course of vertebrate evolution; there are clear differences in the structure of fish and mammalian genomes and in the complexity of fish and mammalian hearts. It will be particularly exciting to elucidate models

References Alexander, J., Stainier, D.Y., Yelon, D., 1998. Screening mosaic F1 females for mutations affecting zebrafish heart induction and patterning. Dev. Genet. 22, 288–299. Amsterdam, A., Burgess, S., Golling, G., Chen, W., Sun, Z., Townsend, K., Farrington, S., Haldi, M., Hopkins, N., 1999. A large-scale insertional mutagenesis screen in zebrafish. Genes Dev. 13, 2713–2724. Arnaout, R., Ferrer, T., Huisken, J., Spitzer, K., Stainier, D.Y., TristaniFirouzi, M., Chi, N.C., 2007. Zebrafish model for human long QT syndrome. Proc. Natl. Acad. Sci. USA 104, 11316–11321. Auman, H.J., Coleman, H., Riley, H.E., Olale, F., Tsai, H.J., Yelon, D., 2007. Functional modulation of cardiac form through regionally confined cell shape changes. PLoS Biol. 5, e53. Bahary, N., Davidson, A., Ransom, D., Shepard, J., Stern, H., Trede, N., Zhou, Y., Barut, B., Zon, L.I., 2004. The Zon laboratory guide to positional cloning in zebrafish. Methods Cell Biol. 77, 305–329. Baker, K., Warren, K.S., Yellen, G., Fishman, M.C., 1997. Defective “pacemaker” current (Ih) in a zebrafish mutant with a slow heart rate. Proc. Natl. Acad. Sci. USA 94, 4554–4559. Bartman, T., Walsh, E.C., Wen, K.K., McKane, M., Ren, J., Alexander, J., Rubenstein, P.A., Stainier, D.Y., 2004. Early myocardial function affects endocardial cushion development in zebrafish. PLoS Biol. 2, E129. Basson, C.T., Bachinsky, D.R., Lin, R.C., Levi, T., Elkins, J.A., Soults, J., Grayzel, D., Kroumpouzou, E., Traill, T.A., Leblanc-Straceski, J., Renault, B., Kucherlapati, R., Seidman, J.G., Seidman, C.E., 1997. Mutations in human TBX5 cause limb and cardiac malformation in Holt-Oram syndrome. Nat. Genet. 15, 30–35.

Chapter  |  1.4  Cardiac Development in the Zebrafish

Beis, D., Stainier, D.Y., 2006. In vivo cell biology: following the zebrafish trend. Trends Cell Biol. 16, 105–112. Beis, D., Bartman, T., Jin, S.W., Scott, I.C., D’Amico, L.A., Ober, E.A., Verkade, H., Frantsve, J., Field, H.A., Wehman, A., Baier, H., Tallafuss, A., Bally-Cuif, L., Chen, J.N., Stainier, D.Y., Jungblut, B., 2005. Genetic and cellular analyses of zebrafish atrioventricular cushion and valve development. Development 132, 4193–4204. Berdougo, E., Coleman, H., Lee, D.H., Stainier, D.Y., Yelon, D., 2003. Mutation of weak atrium/atrial myosin heavy chain disrupts atrial function and influences ventricular morphogenesis in zebrafish. Deve­lopment 130, 6121–6129. Bodmer, R., 1993. The gene tinman is required for specification of the heart and visceral muscles in Drosophila. Development 118, 719–729. Brenner, S., 1974. The genetics of Caenorhabditis elegans. Genetics 77, 71–94. Buckingham, M., Meilhac, S., Zaffran, S., 2005. Building the mammalian heart from two sources of myocardial cells. Nat. Rev. Genet. 6, 826–835. Bussmann, J., Bakkers, J., Schulte-Merker, S., 2007. Early endocardial morphogenesis requires Scl/Tal1. PLoS Genet. 3, e140. Capecchi, M.R., 2005. Gene targeting in mice: functional analysis of the mammalian genome for the twenty-first century. Nat. Rev. Genet. 6, 507–512. Carmany-Rampey, A., Moens, C.B., 2006. Modern mosaic analysis in the zebrafish. Methods 39, 228–238. Chang, C.P., Neilson, J.R., Bayle, J.H., Gestwicki, J.E., Kuo, A., Stankunas, K., Graef, I.A., Crabtree, G.R., 2004. A field of myocardialendocardial NFAT signaling underlies heart valve morphogenesis. Cell 118, 649–663. Chen, J.N., Haffter, P., Odenthal, J., Vogelsang, E., Brand, M., van Eeden, F.J., Furutani-Seiki, M., Granato, M., Hammerschmidt, M., Heisenberg, C.P., Jiang, Y.J., Kane, D.A., Kelsh, R.N., Mullins, M.C., Nusslein-Volhard, C., 1996. Mutations affecting the cardiovascular system and other internal organs in zebrafish. Development 123, 293–302. Cibrian-Uhalte, E., Langenbacher, A., Shu, X., Chen, J.N., AbdelilahSeyfried, S., 2007. Involvement of zebrafish Na, K ATPase in myocardial cell junction maintenance. J. Cell. Biol. 176, 223–230. Corley-Smith, G.E., Lim, C.J., Brandhorst, B.P., 1996. Production of androgenetic zebrafish (Danio rerio). Genetics 142, 1265–1276. Ebert, A.M., Hume, G.L., Warren, K.S., Cook, N.P., Burns, C.G., Mohideen, M.A., Siegal, G., Yelon, D., Fishman, M.C., Garrity, D.M., 2005. Calcium extrusion is critical for cardiac morphogenesis and rhythm in embryonic zebrafish hearts. Proc. Natl. Acad. Sci. USA 102, 17705–17710. Ekker, S.C., 2000. Morphants: a new systematic vertebrate functional genomics approach. Yeast 17, 302–306. Forouhar, A.S., Liebling, M., Hickerson, A., Nasiraei-Moghaddam, A., Tsai, H.J., Hove, J.R., Fraser, S.E., Dickinson, M.E., Gharib, M., 2006. The embryonic vertebrate heart tube is a dynamic suction pump. Science 312, 751–753. Garrity, D.M., Childs, S., Fishman, M.C., 2002. The heartstrings mutation in zebrafish causes heart/fin Tbx5 deficiency syndrome. Development 129, 4635–4645. Glickman, N.S., Yelon, D., 2002. Cardiac development in zebrafish: coordination of form and function. Semin. Cell. Dev. Biol. 13, 507–513. Grimes, A.C., Stadt, H.A., Shepherd, I.T., Kirby, M.L., 2006. Solving an enigma: arterial pole development in the zebrafish heart. Dev. Biol. 290, 265–276.

117

Gruber, P.J., Epstein, J.A., 2004. Development gone awry: congenital heart disease. Circ. Res. 94, 273–283. Halloran, M.C., Sato-Maeda, M., Warren, J.T., Su, F., Lele, Z., Krone, P.H., Kuwada, J.Y., Shoji, W., 2000. Laser-induced gene expression in specific cells of transgenic zebrafish. Development 127, 1953–1960. Hatta, K., Tsujii, H., Omura, T., 2006. Cell tracking using a photoconvertible fluorescent protein. Nat. Protoc. 1, 960–967. Ho, R.K., Kimmel, C.B., 1993. Commitment of cell fate in the early zebrafish embryo. Science 261, 109–111. Holtzman, N.G., Schoenebeck, J.J., Tsai, H.J., Yelon, D., 2007. Endocardium is necessary for cardiomyocyte movement during heart tube assembly. Development 134, 2379–2386. Horne-Badovinac, S., Lin, D., Waldron, S., Schwarz, M., Mbamalu, G., Pawson, T., Jan, Y., Stainier, D.Y., Abdelilah-Seyfried, S., 2001. Positional cloning of heart and soul reveals multiple roles for PKC lambda in zebrafish organogenesis. Curr. Biol. 11, 1492–1502. Hove, J.R., Koster, R.W., Forouhar, A.S., Acevedo-Bolton, G., Fraser, S.E., Gharib, M., 2003. Intracardiac fluid forces are an essential epigenetic factor for embryonic cardiogenesis. Nature 421, 172–177. Hu, N., Sedmera, D., Yost, H.J., Clark, E.B., 2000. Structure and function of the developing zebrafish heart. Anat. Rec. 260, 148–157. Hu, N., Yost, H.J., Clark, E.B., 2001. Cardiac morphology and blood pressure in the adult zebrafish. Anat. Rec. 264, 1–12. Huang, C.J., Tu, C.T., Hsiao, C.D., Hsieh, F.J., Tsai, H.J., 2003. Germ-line transmission of a myocardium-specific GFP transgene reveals critical regulatory elements in the cardiac myosin light chain 2 promoter of zebrafish. Dev. Dyn. 228, 30–40. Huang, C.J., Jou, T.S., Ho, Y.L., Lee, W.H., Jeng, Y.T., Hsieh, F.J., Tsai, H.J., 2005. Conditional expression of a myocardium-specific transgene in zebrafish transgenic lines. Dev. Dyn. 233, 1294–1303. Huisken, J., Stainier, D.Y., 2007. Even fluorescence excitation by multidirectional selective plane illumination microscopy (mSPIM). Opt. Lett. 32, 2608–2610. Huisken, J., Swoger, J., Del Bene, F., Wittbrodt, J., Stelzer, E.H., 2004. Optical sectioning deep inside live embryos by selective plane illumination microscopy. Science 305, 1007–1009. Hurlstone, A.F., Haramis, A.P., Wienholds, E., Begthel, H., Korving, J., Van Eeden, F., Cuppen, E., Zivkovic, D., Plasterk, R.H., Clevers, H., 2003. The Wnt/beta-catenin pathway regulates cardiac valve formation. Nature 425, 633–637. Jacobson, A.G., Sater, A.K., 1988. Features of embryonic induction. Development 104, 341–359. Jin, S.W., Herzog, W., Santoro, M.M., Mitchell, T.S., Frantsve, J., Jungblut, B., Beis, D., Scott, I.C., D’Amico, L.A., Ober, E.A., Verkade, H., Field, H.A., Chi, N.C., Wehman, A.M., Baier, H., Stainier, D.Y., 2007. A transgene-assisted genetic screen identifies essential regulators of vascular development in vertebrate embryos. Dev. Biol. 307, 29–42. Kawakami, K., 2005. Transposon tools and methods in zebrafish. Dev. Dyn. 234, 244–254. Keating, M.T., Sanguinetti, M.C., 2001. Molecular and cellular mechanisms of cardiac arrhythmias. Cell 104, 569–580. Keegan, B.R., Meyer, D., Yelon, D., 2004. Organization of cardiac chamber progenitors in the zebrafish blastula. Development 131, 3081–3091. Keegan, B.R., Feldman, J.L., Begemann, G., Ingham, P.W., Yelon, D., 2005. Retinoic acid signaling restricts the cardiac progenitor pool. Science 307, 247–249. Kikuchi, Y., Trinh, L.A., Reiter, J.F., Alexander, J., Yelon, D., Stainier, D.Y., 2000. The zebrafish bonnie and clyde gene encodes a Mix family

118

homeodomain protein that regulates the generation of endodermal precursors. Genes Dev. 14, 1279–1289. Kikuchi, Y., Agathon, A., Alexander, J., Thisse, C., Waldron, S., Yelon, D., Thisse, B., Stainier, D.Y., 2001. Casanova encodes a novel Sox-related protein necessary and sufficient for early endoderm formation in zebrafish. Genes Dev. 15, 1493–1505. Kimmel, C.B., Law, R.D., 1985. Cell lineage of zebrafish blastomeres. I. Cleavage pattern and cytoplasmic bridges between cells. Dev. Biol. 108, 78–85. Kimmel, C.B., Warga, R.M., Schilling, T.F., 1990. Origin and organization of the zebrafish fate map. Development 108, 581–594. Kimmel, C.B., Ballard, W.W., Kimmel, S.R., Ullmann, B., Schilling, T.F., 1995. Stages of embryonic development of the zebrafish. Dev. Dyn. 203, 253–310. Koster, R.W., Fraser, S.E., 2001. Tracing transgene expression in living zebrafish embryos. Dev. Biol. 233, 329–346. Kozlowski, D.J., Murakami, T., Ho, R.K., Weinberg, E.S., 1997. Regional cell movement and tissue patterning in the zebrafish embryo revealed by fate mapping with caged fluorescein. Biochem. Cell Biol. 75, 551–562. Kupperman, E., An, S., Osborne, N., Waldron, S., Stainier, D.Y., 2000. A sphingosine-1-phosphate receptor regulates cell migration during vertebrate heart development. Nature 406, 192–195. Lander, A.D., Selleck, S.B., 2000. The elusive functions of proteoglycans: in vivo veritas. J. Cell Biol. 148, 227–232. Langenau, D.M., Feng, H., Berghmans, S., Kanki, J.P., Kutok, J.L., Look, A.T., 2005. Cre/lox-regulated transgenic zebrafish model with conditional myc-induced T cell acute lymphoblastic leukemia. Proc. Natl. Acad. Sci. USA 102, 6068–6073. Langenbacher, A.D., Dong, Y., Shu, X., Choi, J., Nicoll, D.A., Goldhaber, J.I., Philipson, K.D., Chen, J.N., 2005. Mutation in sodium–calcium exchanger 1 (NCX1) causes cardiac fibrillation in zebrafish. Proc. Natl. Acad. Sci. USA 102, 17699–17704. Langheinrich, U., Vacun, G., Wagner, T., 2003. Zebrafish embryos express an orthologue of HERG and are sensitive toward a range of QT-prolonging drugs inducing severe arrhythmia. Toxicol. Appl. Pharmacol. 193, 370–382. Lee, R.K., Stainier, D.Y., Weinstein, B.M., Fishman, M.C., 1994. Cardiovascular development in the zebrafish. II. Endocardial progenitors are sequestered within the heart field. Development 120, 3361–3366. Lepilina, A., Coon, A.N., Kikuchi, K., Holdway, J.E., Roberts, R.W., Burns, C.G., Poss, K.D., 2006. A dynamic epicardial injury response supports progenitor cell activity during zebrafish heart regeneration. Cell 127, 607–619. Li, Q.Y., Newbury-Ecob, R.A., Terrett, J.A., Wilson, D.I., Curtis, A.R., Yi, C.H., Gebuhr, T., Bullen, P.J., Robson, S.C., Strachan, T., Bonnet, D., Lyonnet, S., Young, I.D., Raeburn, J.A., Buckler, A.J., Law, D.J., Brook, J.D., 1997. Holt-Oram syndrome is caused by mutations in TBX5, a member of the Brachyury (T) gene family. Nat. Genet. 15, 21–29. Liao, W., Bisgrove, B.W., Sawyer, H., Hug, B., Bell, B., Peters, K., Grunwald, D.J., Stainier, D.Y., 1997. The zebrafish gene cloche acts upstream of a flk-1 homologue to regulate endothelial cell differentiation. Development 124, 381–389. Lyons, I., Parsons, L.M., Hartley, L., Li, R., Andrews, J.E., Robb, L., Harvey, R.P., 1995. Myogenic and morphogenetic defects in the heart tubes of murine embryos lacking the homeo box gene Nkx2-5. Genes. Dev. 9, 1654–1666.

PART  |  1  Heart Evolution

Mably, J.D., Chuang, L.P., Serluca, F.C., Mohideen, M.A., Chen, J.N., Fishman, M.C., 2006. Santa and valentine pattern concentric growth of cardiac myocardium in the zebrafish. Development 133, 3139–3146. Mably, J.D., Mohideen, M.A., Burns, C.G., Chen, J.N., Fishman, M.C., 2003. Heart of glass regulates the concentric growth of the heart in zebrafish. Curr. Biol. 13, 2138–2147. Milan, D.J., Peterson, T.A., Ruskin, J.N., Peterson, R.T., MacRae, C.A., 2003. Drugs that induce repolarization abnormalities cause bradycardia in zebrafish. Circulation 107, 1355–1358. Milan, D.J., Giokas, A.C., Serluca, F.C., Peterson, R.T., MacRae, C.A., 2006. Notch1b and neuregulin are required for specification of central cardiac conduction tissue. Development 133, 1125–1132. Murphey, R.D., Zon, L.I., 2006. Small molecule screening in the zebrafish. Methods 39, 255–261. Murphey, R.D., Stern, H.M., Straub, C.T., Zon, L.I., 2006. A chemical genetic screen for cell cycle inhibitors in zebrafish embryos. Chem. Biol. Drug Des. 68, 213–219. Nagayoshi, S., Hayashi, E., Abe, G., Osato, N., Asakawa, K., Urasaki, A., Horikawa, K., Ikeo, K., Takeda, H., Kawakami, K., 2008. Insertional mutagenesis by the Tol2 transposon-mediated enhancer trap approach generated mutations in two developmental genes: tcf7 and synembryn-like. Development 135, 159–169. Ni-Komatsu, L., Orlow, S.J., 2007. Identification of novel pigmentation modulators by chemical genetic screening. J. Invest. Dermatol. 127, 1585–1592. Nusslein-Volhard, C., Wieschaus, E., 1980. Mutations affecting segment number and polarity in Drosophila. Nature 287, 795–801. Oates, A.C., Bruce, A.E., Ho, R.K., 2000. Too much interference: injection of double-stranded RNA has nonspecific effects in the zebrafish embryo. Dev. Biol. 224, 20–28. Patton, E.E., Zon, L.I., 2001. The art and design of genetic screens: zebrafish. Nat. Rev. Genet. 2, 956–966. Peterkin, T., Gibson, A., Patient, R., 2007. Redundancy and evolution of GATA factor requirements in development of the myocardium. Dev. Biol. 311 (2), 623–635. Peterson, R.T., Link, B.A., Dowling, J.E., Schreiber, S.L., 2000. Small molecule developmental screens reveal the logic and timing of vertebrate development. Proc. Natl. Acad. Sci. USA 97, 12965–12969. Peterson, R.T., Mably, J.D., Chen, J.N., Fishman, M.C., 2001. Convergence of distinct pathways to heart patterning revealed by the small molecule concentramide and the mutation heart-and-soul. Curr. Biol. 11, 1481–1491. Peterson, R.T., Shaw, S.Y., Peterson, T.A., Milan, D.J., Zhong, T.P., Schreiber, S.L., MacRae, C.A., Fishman, M.C., 2004. Chemical suppression of a genetic mutation in a zebrafish model of aortic coarctation. Nat. Biotechnol. 22, 595–599. Peyrieras, N., Strahle, U., Rosa, F., 1998. Conversion of zebrafish blastomeres to an endodermal fate by TGF-beta-related signaling. Curr. Biol. 8, 783–786. Piotrowski, T., Ahn, D.G., Schilling, T.F., Nair, S., Ruvinsky, I., Geisler, R., Rauch, G.J., Haffter, P., Zon, L.I., Zhou, Y., Foott, H., Dawid, I.B., Ho, R.K., 2003. The zebrafish van gogh mutation disrupts tbx1, which is involved in the DiGeorge deletion syndrome in humans. Development 130, 5043–5052. Poss, K.D., Wilson, L.G., Keating, M.T., 2002. Heart regeneration in zebrafish. Science 298, 2188–2190. Ransom, J., Srivastava, D., 2007. The genetics of cardiac birth defects. Semin. Cell Dev. Biol. 18, 132–139.

Chapter  |  1.4  Cardiac Development in the Zebrafish

Raz, E., 2003. Primordial germ-cell development: the zebrafish perspective. Nat. Rev. Genet. 4, 690–700. Razzaque, M.A., Nishizawa, T., Komoike, Y., Yagi, H., Furutani, M., Amo, R., Kamisago, M., Momma, K., Katayama, H., Nakagawa, M., Fujiwara, Y., Matsushima, M., Mizuno, K., Tokuyama, M., Hirota, H., Muneuchi, J., Higashinakagawa, T., Matsuoka, R., 2007. Germline gain-of-function mutations in RAF1 cause Noonan syndrome. Nat. Genet. 39, 1013–1017. Reifers, F., Walsh, E.C., Leger, S., Stainier, D.Y., Brand, M., 2000. Induction and differentiation of the zebrafish heart requires fibroblast growth factor 8 (fgf8/acerebellar). Development 127, 225–235. Reiter, J.F., Alexander, J., Rodaway, A., Yelon, D., Patient, R., Holder, N., Stainier, D.Y., 1999. Gata5 is required for the development of the heart and endoderm in zebrafish. Genes. Dev. 13, 2983–2995. Reiter, J.F., Verkade, H., Stainier, D.Y., 2001. Bmp2b and Oep promote early myocardial differentiation through their regulation of gata5. Dev Biol. 234, 330–338. Robu, M.E., Larson, J.D., Nasevicius, A., Beiraghi, S., Brenner, C., Farber, S.A., Ekker, S.C., 2007. p53 activation by knockdown technologies. PLoS Genet. 3, e78. Rohde, L.A., Heisenberg, C.P., 2007. Zebrafish gastrulation: cell movements, signals, and mechanisms. Int Rev Cytol. 261, 159–192. Rohr, S., Bit-Avragim, N., Abdelilah-Seyfried, S., 2006. Heart and soul/ PRKCi and nagie oko/Mpp5 regulate myocardial coherence and remodeling during cardiac morphogenesis. Development 133, 107–115. Rottbauer, W., Baker, K., Wo, Z.G., Mohideen, M.A., Cantiello, H.F., Fishman, M.C., 2001. Growth and function of the embryonic heart depend upon the cardiac-specific L-type calcium channel alpha1 subunit. Dev. Cell 1, 265–275. Rottbauer, W., Saurin, A.J., Lickert, H., Shen, X., Burns, C.G., Wo, Z.G., Kemler, R., Kingston, R., Wu, C., Fishman, M.C., 2002. Reptin and pontin antagonistically regulate heart growth in zebrafish embryos. Cell 111, 661–672. Rottbauer, W., Wessels, G., Dahme, T., Just, S., Trano, N., Hassel, D., Burns, C.G., Katus, H.A., Fishman, M.C., 2006. Cardiac myosin light chain-2: a novel essential component of thick-myofilament assembly and contractility of the heart. Circ. Res. 99, 323–331. Schoenebeck, J.J., Yelon, D., 2007. Illuminating cardiac development: Advances in imaging add new dimensions to the utility of zebrafish genetics. Semin. Cell Dev. Biol. 18, 27–35. Schoenebeck, J.J., Keegan, B.R., Yelon, D., 2007. Vessel and blood specification override cardiac potential in anterior mesoderm. Dev. Cell 13, 254–267. Schoenwolf, G.C., Garcia-Martinez, V., 1995. Primitive-streak origin and state of commitment of cells of the cardiovascular system in avian and mammalian embryos. Cell Mol. Biol. Res. 41, 233–240. Schott, J.J., Benson, D.W., Basson, C.T., Pease, W., Silberbach, G.M., Moak, J.P., Maron, B.J., Seidman, C.E., Seidman, J.G., 1998. Congenital heart disease caused by mutations in the transcription factor NKX2-5. Science 281, 108–111. Scott, I.C., Masri, B., D’Amico, L.A., Jin, S.W., Jungblut, B., Wehman, A.M., Baier, H., Audigier, Y., Stainier, D.Y., 2007. The g proteincoupled receptor agtrl1b regulates early development of myocardial progenitors. Dev. Cell 12, 403–413. Sedmera, D., Reckova, M., deAlmeida, A., Sedmerova, M., Biermann, M., Volejnik, J., Sarre, A., Raddatz, E., McCarthy, R.A., Gourdie, R.G., Thompson, R.P., 2003. Functional and morphological evidence for a ventricular conduction system in zebrafish and Xenopus hearts. Am. J. Physiol. Heart Circ. Physiol. 284, H1152–H1160.

119

Sehnert, A.J., Huq, A., Weinstein, B.M., Walker, C., Fishman, M., Stainier, D.Y., 2002. Cardiac troponin T is essential in sarcomere assembly and cardiac contractility. Nat. Genet. 31, 106–110. Shu, X., Cheng, K., Patel, N., Chen, F., Joseph, E., Tsai, H.J., Chen, J.N., 2003. Na,K-ATPase is essential for embryonic heart development in the zebrafish. Development 130, 6165–6173. Simmer, F., Moorman, C., van der Linden, A.M., Kuijk, E., van den Berghe, P.V., Kamath, R.S., Fraser, A.G., Ahringer, J., Plasterk, R. H., 2003. Genome-wide RNAi of C. elegans using the hypersensitive rrf-3 strain reveals novel gene functions. PLoS Biol. 1, E12. Sivasubbu, S., Balciunas, D., Amsterdam, A., Ekker, S.C., 2007. Insertional mutagenesis strategies in zebrafish. Genome Biol. 8 (Suppl 1), S9. Sood, R., English, M.A., Jones, M., Mullikin, J., Wang, D.M., Anderson, M., Wu, D., Chandrasekharappa, S.C., Yu, J., Zhang, J., Paul Liu, P., 2006. Methods for reverse genetic screening in zebrafish by resequencing and TILLING. Methods 39, 220–227. Stainier, D.Y., Fishman, M.C., 1992. Patterning the zebrafish heart tube: acquisition of anteroposterior polarity. Dev. Biol. 153, 91–101. Stainier, D.Y., Lee, R.K., Fishman, M.C., 1993. Cardiovascular development in the zebrafish. I. Myocardial fate map and heart tube formation. Development 119, 31–40. Stainier, D.Y., Weinstein, B.M., Detrich, H.W., Zon, L.I., Fishman, M.C., 1995. Cloche, an early acting zebrafish gene, is required by both the endothelial and hematopoietic lineages. Development 121, 3141–3150. Stainier, D.Y., Fouquet, B., Chen, J.N., Warren, K.S., Weinstein, B.M., Meiler, S.E., Mohideen, M.A., Neuhauss, S.C., Solnica-Krezel, L., Schier, A.F., Zwartkruis, F., Stemple, D.L., Malicki, J., Driever, W., Fishman, M.C., 1996. Mutations affecting the formation and function of the cardiovascular system in the zebrafish embryo. Development 123, 285–292. Tam, P.P., Parameswaran, M., Kinder, S.J., Weinberger, R.P., 1997. The allocation of epiblast cells to the embryonic heart and other mesodermal lineages: the role of ingression and tissue movement during gastrulation. Development 124, 1631–1642. Thisse, B., Heyer, V., Lux, A., Alunni, V., Degrave, A., Seiliez, I., Kirchner, J., Parkhill, J.P., Thisse, C., 2004. Spatial and temporal expression of the zebrafish genome by large-scale in situ hybridization screening. Methods Cell. Biol. 77, 505–519. Towbin, J.A., Bowles, N.E., 2002. The failing heart. Nature 415, 227–233. Trinh, L.A., Stainier, D.Y., 2004. Fibronectin regulates epithelial organization during myocardial migration in zebrafish. Dev. Cell. 6, 371–382. Trinh, L.A., Yelon, D., Stainier, D.Y., 2005. Hand2 regulates epithelial formation during myocardial diferentiation. Curr. Biol. 15, 441–446. Ueno, S., Weidinger, G., Osugi, T., Kohn, A.D., Golob, J.L., Pabon, L., Reinecke, H., Moon, R.T., Murry, C.E., 2007. Biphasic role for Wnt/beta-catenin signaling in cardiac specification in zebrafish and embryonic stem cells. Proc Natl Acad Sci USA 104, 9685–9690. Walsh, E.C., Stainier, D.Y., 2001. UDP-glucose dehydrogenase required for cardiac valve formation in zebrafish. Science 293, 1670–1673. Walton, R.Z., Bruce, A.E., Olivey, H.E., Najib, K., Johnson, V., Earley, J.U., Ho, R.K., Svensson, E.C., 2006. Fog1 is required for cardiac looping in zebrafish. Dev. Biol. 289, 482–493. Warren, K.S., Wu, J.C., Pinet, F., Fishman, M.C., 2000. The genetic basis of cardiac function: dissection by zebrafish (Danio rerio) screens. Philos. Trans. R. Soc. Lond. B Biol. Sci. 355, 939–944.

120

Xu, X., Meiler, S.E., Zhong, T.P., Mohideen, M., Crossley, D.A., Burggren, W.W., Fishman, M.C., 2002. Cardiomyopathy in zebrafish due to mutation in an alternatively spliced exon of titin. Nature Genet. 30, 205–209. Yelon, D., Horne, S.A., Stainier, D.Y., 1999. Restricted expression of cardiac myosin genes reveals regulated aspects of heart tube assembly in zebrafish. Dev. Biol. 214, 23–37. Yelon, D., Ticho, B., Halpern, M.E., Ruvinsky, I., Ho, R.K., Silver, L.M., Stainier, D.Y., 2000. The bHLH transcription factor hand2 plays

PART  |  1  Heart Evolution

parallel roles in zebrafish heart and pectoral fin development. Development 127, 2573–2582. Yuan, S., Joseph, E.M., 2004. The small heart mutation reveals novel roles of Na/K-ATPase in maintaining ventricular cardiomyocyte morphology and viability in zebrafish. Circ. Res. 95, 595–603. Zeng, X.X., Wilm, T.P., Sepich, D.S., Solnica-Krezel, L., 2007. Apelin and its receptor control heart field formation during zebrafish gastrulation. Dev. Cell 12, 391–402.

Chapter 1.5

An Overview of Avian Heart Structure and Development Katherine Moynihan1, Frank Stockdale2 and David Bader1 1 Stahlman Cardiovascular Research Laboratories, Program for Developmental Biology, and Department of Medicine, Vanderbilt University Medical Center, Nashville, TN, USA 2 Department of Medicine, Stanford University, School of Medicine, Stanford, CA, USA

I.  Overview of chapter The avian, and more specifically the chick, heart has provided a key experimental model for the developmental biologist. Advances in the regulation of cardiac myogenesis using genetic models have provided key insights. Still, the avian heart remains as one of the most accessible models for the manipulation of cardiac, epicardial and endothelial differentiation in heart development. The avian heart has so many advantages to the experimental zoologist that it seems hard to think of the analysis of cardiac developmental biology without it. We have been asked to provide a chapter on the avian heart in overview. In writing this chapter, we have made a conscious decision to focus solely on the avian, and in most cases, the chick heart, and have directed our attention to the cardiac myocyte and myocardium with reference to the overall differentiation and morphogenesis of the heart. The discussion of the cardiac myocyte is framed in the overall context of the developing heart. For a more in depth discussion of the development of the endo- and epicardium, readers are asked to refer to Chapters 3.1, 3.2, 5.1 and 5.2. Finally, this chapter relies heavily on work that examines avian heart structure and studies particularly using experimental zoology to understand mechanisms regulating cardiac myogenesis.

II.  Anatomy of the chick heart The anatomy of the avian heart is very similar to that of the mammalian heart, yet offers deviations unique to its specific niche. Both of these vertebrate classes demonstrate Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

a significant demand for oxygen while maintaining homiothermy. The highly-conserved four-chambered heart separates the pulmonary and systemic circulatory pathways to meet these demands (Proctor and Lynch, 1993). The anterior cardiac chambers (the atria) and the posterior chambers (the ventricles) lie in series between the two circulation paths and serve to repressurize the blood. The entire organ is enclosed in fluid within the thin, fibrous perdicardial sac and is located in the thoracoabdominal cavity (Whittow and Sturkie, 2000). However, the physical act of flight, mostly restricted to birds, is more strenuous than any other activity and requires some basic modification of the heart. The avian heart is significantly heavier in proportion to its body weight than all other vertebrates (Proctor and Lynch, 1993). By combining this larger size with its bigger stroke volume, the output of the avian cardiovascular system can carry enough oxygen to muscular tissue to sustain flight.

II.A.  Circulation, Chambers and Valves The four-chambered heart moves the blood through the pulmonary and somatic circuits with each contraction. Deoxygenated blood returns from the body and enters anteriorly into the right atrium. This atrium tends to be much larger than its left counterpart, but both have relatively thin walls and no inflow valves (Whittow and Sturkie, 2000). The blood then moves through the right atrioventricular (AV) valve and into the right ventricle. This valve is a single spiral plane of myocardium, significantly different from the fibrous flap of the mammalian heart (Lu et al., 1993a,b) (Fig. 1).

121

122

Figure 1  Anterior view of adult chicken heart. This diagram shows a longitudinal section from an adult avian heart from Lu et al. (1993a). This diagram illustrates the basic structure of the avian heart with the relative thicknesses of atria (RA and LA) and ventricles (RV and LV). The ventricles are relatively thicker than the atria, and have significant trabeculation. The interventricular septum is also thick. Note that the right ventricle does not extend to the apex of the heart. The left atrioventricular valve leaflets are relatively thin and are linked to the ventricular wall by chordae tendinae and papillary muscles. The right atrioventricular value has only one lateral leaflet. The atria can be very thin.

It even has a ring of conduction tissue around the aperture, suggesting that muscular contraction may play a role during valve closure (Lu et al., 1993a,b). The right ventricle is the less muscular of the two lower chambers, and does not extend to the apex of the heart. It compresses and sends a volume of blood through the pulmonary arteries to the lungs, a route not requiring extensive pressure. The outflow valve here is made up of three semilunar cusps. Once oxygenated, the blood returns to the heart via the pulmonary veins and enters the left atrium. It passes through the left AV valve and into the highly-muscular left ventricle. This AV valve is also encircled by conduction fibers, again suggesting deviation from the passive nature of mammalian valve leaflets. The muscular wall of the left ventricle is two-to-three times thicker than its right counterpart, and its cone shape extends all the way to the apex of the heart (Whittow and Sturkie, 2000). This chamber empties with considerable force through the aortic outflow valve, comprised of three rigid cusps linked to the left AV valve by an arch of cardiac muscle (Lu et al., 1993a). This chamber delivers blood to the rest of the body first as it passes through the aortic arch, which branches to the right in birds unlike mammals.

II.B.  Coronary Circulation and Conduction The coronary vessels originate immediately above the aortic outflow valve as two sinuses leading to the right and left branches. The right coronary artery tends to be the

PART  |  1  Heart Evolution

dominant of the two in most species, but both begin with a split into a superficial and deep branch on either side of the heart. The dominant branch supports the left ventricular myocardium, intraventricular septum, the apex and the dorsal walls of the atria (West, 1981). The arteries form frequent anastomoses en route to capillary beds and venous blood is returned to the heart via five groups of cardiac veins, also prone to anastomoses. Relative to other avian tissues this circulatory system leads to a highly perfused myocardium. Additionally, blood flow increases with reduction in the oxygen supply, allowing highaltitude flight. Systemic circulation emanating from the heart is dependent on the conduction system prompting chamber contraction. This conduction system relies on electrical impulses initiated by the pacemaker cells of the sinoatrial (SA) node on the anterior side of the right atrium. Cardiac myocytes carry this signal through low-resistance pathways; the intercalated disc uses physical (desmosome) and electrical (nexus) connections between cells to couple many individual cells to work as a unit. This signal sweeps through both atria causing depolarization and thus, contraction. Mathur and Shrivastava demonstrated the presence of distinct bundle branches connecting the SA node with the atrioventricular (AV) node across the atria (Mathur and Shrivastava, 1979). The presence of these structures in mammalian hearts has been a point of debate. Once the signal reaches the AV node, it is then delayed, moving at a rate two-to-three orders of magnitude slower at the atrioventricular (AV) node to allow the atria to empty. It should be noted that Szabo et al. (1986) reported that light and electron microscopic examination of the AV junction in avian species did not reveal a distinct AV node when compared to other classes of vertebrates. They determined that myocytes of an AV ring that has contact with working atrial myocytes served as the AV node. Conduction system myocytes represent the only myogenic connection between the adult atria and ventricles. The His bundle and its three branches extend from the AV node, and are invested in connective tissue (Szabo et al., 1986). Two of the bundles branch into subendocardium of the left and right ventricles and follow the pathways of the coronary arteries. The third bundle takes an unusual route up around the aorta and connects to the avian-specific ring around the right AV valve (Lu et al., 1993a,b). These branches disperse as they enter the myocardium and connect with the terminal cells of the conduction system, the Purkinje fibers. These cells are large, elongated cardiac myocytes that conduct electrical impulses to the working myocardium. Their conductivity is much faster than cardiomyocytes of the atrial and ventricular walls. Many groups have studied the variation in gene expression in avian conduction system myocytes. There are many immunochemical reagents available to identify these cells within the developing and adult heart

Chapter  |  1.5  An Overview of Avian Heart Structure and Development

(de Groot et al., 1985; Gonzalez-Sanchez and Bader, 1985; Takebayashi-Suzuki et al., 2000, 2001). Gonzalez-Sanchez and Bader suggested that a common myosin heavy chain was expressed in both the atrial and ventricular muscle components of the avian conduction system (GonzalezSanchez and Bader, 1985). However, all of these studies attest to the diversity among these specialized myocytes in the ventricular wall, even though they are derived from common working myocytes (see below).

II.C.  Histology The histology of the avian heart is very similar to the mammalian heart. The heart tissue is comprised of three layers – the epicardium, myocardium and endocardium. The epicardium and endocardium are similar in structure; loose connective tissue supporting a simple squamous layer of mesothelium or endothelium. The two layers are separated by myocardium, a layer of striated muscle fibers, which differs within the heart only by its significantly thicker presence in the ventricles. This particular layer does differ in three ways from mammalian cardiac muscle. Avian striated muscle bands lack M-bands, the significance of which is unknown. The muscle bands are also much shorter in diameter than those of mammals, leading to a much higher number of cells per unit volume within the organ. Avian myocardial cells also lack transverse tubules (T tubules), and instead use junctional processes as “couplings” between the sarcoplasmic reticulum and plasmalemma (Sommer and Johnson, 1969; Hirakow, 1970; Sommer et al., 1991; Whittow and Sturkie, 2000). Differentiated atrial, ventricular and conduction myocytes express distinct myosin heavy chains, which have been used to trace the embryogenesis of their cell lineages. Interestingly, avian hearts express more than just the alpha and beta-type MyHCs detected in mammalian hearts and antibodies exist for many of these additional myosin isoforms (Gonzalez-Sanchez and Bader, 1984; GonzalezSanchez and Bader, 1985; Evans et al., 1988; Stockdale, 1992; Wang et al., 1996; Stockdale et al., 2002). The regulation of gene expression in developing and adult hearts is explored in depth in other chapters.

III.  Development: an overview The avian embryo has provided a critical model for the analysis of cardiac development. Aristotle is one of the first scientists to publish on the use of avian embryos to study heart development. More recent publications (Malpighi, 1672; Rosenthal and Harvey, this scholarly tome) attest to the value of the avian embryo as a model of heart development. The work of Waddington, Butler, Patton, DeHaan and colleagues in the chick embryo laid the foundation for

123

molecular approaches to cardiogenic commitment and differentiation (Manner, 2000). Anyone who has worked with both avian and mammalian embryos knows that mRNA expression boundaries seem to be sharper, that cardiogenic tissue is more easily obtained, and that experimental manipulations are more easily executed in the embryos of birds. The procurement and availability of the embryo has certainly been a large factor in its use. More importantly, experimental zoologists have long understood that the hardy nature and resilience of this embryo makes it an outstanding model for study of cardiac development, even in the absence of genomic manipulation. McCain and McLaughlin give a wonderful compilation of methodologies pertaining to isolation, morphology and analysis of whole chick hearts of the early embryo at http://www.zoo. utoronto.ca/able/volumes/vol-20/4-mccain.pdf (McCain, 1998). Additionally, advances in lineage marking with viral systems (Mikawa and Fischman, 1992; Mikawa et al., 1992; Mikawa, 1995; Mjaatvedt et al., 2001) and vital dyes (Satin et al., 1988; Gonzalez-Sanchez and Bader, 1990; Mjaatvedt et al., 2001) are particularly amenable to the avian embryo and the heart in particular. In this section of the chapter we will discuss the basic development of the avian heart with reference to the cardiac myocytes and present methods that are particularly useful for experimentation in this system.

IV.  Determination and earliest development The avian epiblast contains cells that will give rise to the embryo proper. The posterior thickening of the epiblast, called Koller’s sickle, can be seen at stage 2 (Hamburger and Hamilton, 1992). With its anterior, midline movement or extension, Koller’s sickle gives rise to the primitive streak. Cardiomyogenic cells of the heart are some of the first to ingress during formation of the primitive streak. The studies of Garcia-Martinez and Schoenwolf (1993) established the location of cardiomyogenic progenitors in the pregastrulated chick embryo and are pictured in Fig. 2. Cells were mapped using vital dyes in the primitive streak (stages 2 and 3), and these studies revealed that progenitors were located in the anterior region of the structure. This domain extends 125–700 microns from the anteriormost tip of the primitive streak. It should be noted that the cells at the very anterior tip of the streak are not cardiogenic, but are fated to the notochord (Garcia-Martinez and Schoenwolf, 1993). In addition, cardiac progenitors were shown to be arranged in an anteroposterior fashion within the streak, and the anterior-most cells were destined to form the conus and ventricles. In turn, the more posteriorly positioned cells were fated to the atria and outflow tract. Experimental manipulation of avian cardiac progenitors in vitro is a highly important and continuously employed

124

Figure 2  Fate maps of cardiogenic progenitors and myocytes in the early avian embryo. The relative position of atrial and ventricular myogenic cells in the primitive streak at stage 3 through 13 (Garcia-Martinez and Schoenwolf, 1993; Yutzey et al., 1994; Yutzey and Bader, 1995; modified from these three publications). Solid black represents the relative position of atrial progenitors and myocytes, while the stippled area depicts the regions occupied by ventriculogenic cells. For precise localization using in situ hybridization in postgastrulated embryos see Yutzey et al. (1994).

procedure used by scientists interested in heart development. Identification and isolation of these progenitors is key for further work in this area to elucidate mechanisms regulating cardiogenic commitment and differentiation. Further, because disagreement exists on the position of these progenitors in the lateral plate, we will emphasize recent mapping studies here. Several mapping and extirpation analyses of gastrulating chick embryos demonstrate the presence of cardiac progenitors in the primitive streak at late stage 3 (Garcia-Martinez and Schoenwolf, 1993), and all agree that progenitors have migrated out of the streak at stage 4. Controversy arises regarding the position of cardiomyogenic cells after they leave the primitive streak. Our goal here is to present a clear picture of the two views concerning the positions of postgastrulated, predifferentiated progenitors in the lateral plate. In that way, the investigator should be able to isolate and culture cells and determine the location of progenitors for their experiment. The first issue to consider is the medial–lateral position of cardiac progenitors after gastrulation. Rosenquist, and later Redkar et al., used radioactive and DiI labeling of stage 5–8 chick embryos with subsequent differentiation to localize progenitors to mesodermal regions directly ventral to the developing neural plate (Rosenquist, 1970b; Redkar et al., 2001). This is a “very medial” position with reference to the zona pellucida. Conversely, DeHaan (1963), Stahlsberg and DeHaan (1989), and Ehrman and Yutzey (1999) localized cells of cardiomyogenic potential to a much more lateral position (Dehaan, 1963; Stalsberg and DeHaan, 1969; Ehrman and Yutzey, 1999). Indeed, these studies suggested

PART  |  1  Heart Evolution

that these progenitors are near to the lateral boundary of the zona pellucida (i.e., at the lateral edge of the embryonic mesoderm). While there is agreement concerning the anterior deposition of progenitors just superior to the anterior intestinal portal (Yutzey et al., 1995), the limits of the posterior (atrial-forming domain) boundary are disputed. Surface labeling and time lapse studies of Ehrman and Yutzey (1999) put the posterior boundary of heart-forming cells at the level of the first condensing somite in the stage 7 embryo. But radioisotope and DiI labeling analyses determined that this boundary extended more posteriorly (Rosenquist, 1970b; Redkar et al., 2001). With isolation of stage 4–8 chick embryos using the methods outlined by Gannon and Bader (1995), the individual investigator can easily resolve this issue in their own laboratory. There is little chance for interaction between anterior and posterior progenitors as they leave the streak in a temporally-determined fashion (anterior, first; posterior, last) and they apparently move to like-positioned stations in the lateral mesoderm. The lateral mesoderm is first to lose mesenchymal structure, quickly forming a definitive epithelium (Osler and Bader, 2004). All cardiomyogenic cells have left the streak by stage 4. Conversely, endocardial cells are still present in the streak at this stage. Studies on the position of cardiogenic mesoderm in the anterior lateral plate mesoderm of the chick embryo began over 60 years ago (Rawles, 1952) and have continued to nearly the present day (Dehaan, 1963; DeHaan and Ursprung, 1965; Stalsberg and DeHaan, 1968, 1969; Rosenquist, 1970a,b; Satin et al., 1988; Montgomery et al., 1994; Yutzey et al., 1994; Gannon and Bader, 1995, 1997; Yutzey and Bader, 1995; Ehrman and Yutzey, 1999; Redkar et al., 2001). While the precise location of myocyte progenitors is still not completely resolved, it is generally agreed that two regions of anterior splanchnic mesoderm positioned lateral to midline and separated by 800 microns comprise much of the embryo’s native cardiogenic potential (Stalsberg and DeHaan, 1969). When embryos are isolated by a “paper ring” method described below and viewed in a dissecting microscope, cardiac progenitors can be visualized by their greater opacity relative to the adjoining tissue (Gannon and Bader, 1997). This bilaterally positioned tissue has been referred to as the cardiogenic crescent. At stage 6, these cells form an elongated horseshoe that approaches or possibly abuts the midline anterior to the forming anterior intestinal portal (Fig. 3 and Chapter 3.2). The actual “mechanism of movement” of myogenic cells from the streak to their positions in lateral mesoderm has not been established. It is assumed that single cells or small groups of cells move laterally prior to stage 4 and take up assigned positions in the lateral mesoderm, but to our knowledge real time analysis of cell movement in the avian embryo has not been reported. Interestingly, lineage marking and grafting studies support the idea that

Chapter  |  1.5  An Overview of Avian Heart Structure and Development

Figure 3  Position of cardiogenic cells at stage 6 in the developing chicken embryo. Redrawn from DeHaan and Ursprung (1965). Fate maps of cells at stage 6 have determined the relative positions of cells that incorporate into the chicken heart tube. Several investigators have used this classic map to isolate cells of ventriculogenic and atriogenic potential, and DeHaan’s early work remains a reliable departure point for studies in this area (CV: conoventricular tissue; V: ventricles; SA: sinoatrium).

anteroposterior mixing of cells does not occur during this or subsequent timeframes (Stalsberg and DeHaan, 1969; Rosenquist, 1970b; Satin et al., 1988; Gonzalez-Sanchez and Bader, 1990). By stage 6 and beginning in the anterior region of the cardiogenic mesoderm, an intact and polarized epithelium is formed (for examples of this situation see Osler and Bader, 2004). Cardiogenic mesoderm and resulting cardiac muscle remains polarized throughout formation of the looping heart tube and subsequent chamber formation (Peng et al., 1990). Until trabeculation begins, cardiomyogenic morphogenesis is driven by epithelial movement, folding and remodeling. The importance of forming a cardiogenic epithelium has not been resolved. Differences in the cardiogenic potential along the anteroposterior axis, with regard to ventricular or atrial differentiation, have been noted (Yutzey et al., 1994; Gannon and Bader, 1995; Yutzey and Bader, 1995; Gannon and Bader, 1997).

IV.A.  Cardiogenic Determination Slack defines “specification” as the ability of a cell to express a certain phenotype in a neutral environment (such as cell culture) whereas “determination” is the ability to differentiate within different cellular environments (such as a different locale in the embryo) (Slack, 1991). Gilbert defines “determination” as the cell’s commitment to eventually differentiate into a specific cell type and into no other tissue (Gilbert, 2006). Applying this principle to cardiac myogenesis is difficult and, to our knowledge, has not often been directly tested. Le Douarin conducted heterotopic transplantation of cardiogenic tissue to noncardiogenic

125

regions of the embryo and reported an early commitment to the lineage (Le Douarin et al., 1965). When epiblast cells of the chick embryo are isolated and plated at high density in vitro cardiac myocytes are observed (Montgomery et al., 1994), suggesting that the physical act of gastrulation is not required for cardiac myocyte differentiation. At low density, this differentiation is not observed. Conversely, stage 4 cardiogenic mesoderm can be isolated, dispersed into single cells, and cultured at clonal density. The cells within these clones readily expressed markers of cardiac myocytes (Gonzalez-Sanchez and Bader, 1990), suggesting that no additional signals are needed to propel these cells into their final state of differentiation. In addition, Antin and colleagues have demonstrated the importance of activin in the early commitment of epiblast cells to the cardiogenic cell lineage (Yatskievych et al., 1997). Taken together, these data suggest that pregastrulated cells are not as “determined” as postgastrulated cells. Still, while genetic regulatory networks have been identified that are essential for cardiogenic determination (Firulli and Thattaliyath, 2002), the exact program governing conversion of mesodermal cells into determined cardiac “myoblast” has not been identified for the chick or any other organism. Cardiomyogenic potential is not completely restricted to the cardiogenic crescent. As we will see below, conversion of noncardiogenic mesoderm to this lineage can be accomplished by culturing explants with specific growth factors that promote cardiac differentiation (Schultheiss et al., 1995; Lough et al., 1996; Barron et al., 2000). These conversions were “induced” by application of exogenous factors. Additionally, Mjaatvedt et al. and Waldo et al. have defined an “anterior heart field” that contributes myocytes to the conus and truncus and originates separately from mesoderm located anterior to the initial cardiac heart tube (Mjaatvedt et al., 2001; Waldo et al., 2005). These results are corroborated by studies in the mouse (Kelly et al., 2001; Verzi et al., 2005). Also, Kruithof et al. have recently demonstrated that the proepicardium contributes cardiac myocytes to the inflow tract of the heart (Kruithof et al., 2006). Thus, while it is clear that the cardiogenic crescent is the major contributor to the differentiated myocardium, other cells (the so-called anterior heart field) and proepicardium normally differentiate into cardiac myocytes. Additionally, other cells in the embryo can be induced to take on this cellular phenotype.

IV.B.  Inducers of Cardiomyogenic Determination A major question in the development of cardiac myocytes is what signals regulate the conversion of mesodermal, or even premesodermal, cells to the cardiomyogenic cell lineage. Obviously, a first step in these analyses is to determine when cells take on cardiogenic potential. Interestingly,

126

few reports exist examining this issue. As stated above, Montgomery et al. demonstrated that pregastrulated epiblast cells can differentiate into cardiac myocytes when grown at high density in serum-containing medium (Montgomery et al., 1994). This suggests that simple events may govern the conversion of embryonic cells to the cardiogenic lineage. Corroborating this theme are the many reports of mouse ES cells robustly differentiating into cardiac myocytes under simple culture conditions (Bettiol et al., 2006; van Laake et al., 2006). While these studies in avian species supported the notion that embryonic cells are receptive to inducing signals, they did not identify the factors. Several groups have used the chick or quail embryo to analyze soluble factors that regulate cardiogenic specification. This specification is driven by inducing signals that might arise from adjacent tissues or from the target cells themselves. The avian heart has been used as a model system to identify factors inducing cardiogenic determination (see chapters in Part 9). BMPs, FGFs and TGF-beta have all been identified as key inducers (Sugi and Lough, 1994; Ladd et al., 1998; Barron et al., 2000; Pabon-Pena et al., 2000; Schlange et al., 2000; Marvin et al., 2001). Ectopic application of these agents can induce cardiogenic differentiation in medial, anterior or posterior mesoderm (Schultheiss et al., 1995; Andree et al., 1998; Marvin et al., 2001). As yet, the exact step in specification/determination that is mediated by these factors is unknown (Gannon and Bader, 1995). In addition, most of these studies do not report a complete elimination of cardiogenic differentiation, but cite a decrease or inhibition of differentiation. It is important to note the relationship of these avian studies of inducing reagents to the literature on other species. While elimination of specific BMPs, FGFs and TGFbeta gene products with knockout technology in mice leads to defects in heart development, no such knockout reported thus far has eliminated cardiac muscle (PubMed search, May 2007). Thus, in strictest terms, it is not clear whether a single “magic bullet” of cardiac induction has been identified.

IV.C.  In vitro Analysis of Cardiogenic Mesoderm It should be noted that in vitro analysis of avian cardiogenic mesoderm can be carried out with relative ease. Several groups have published on these methods (Stalsberg and DeHaan, 1969; Gonzalez-Sanchez and Bader, 1990; Sugi and Lough, 1994; Yutzey et al., 1994; Schultheiss et al., 1995; Yutzey and Bader, 1995; Gannon and Bader, 1995, 1997). One important consideration in the culturing of chick cardiogenic mesoderm is the ease of explanting the entire embryo, isolation of relatively pure samples of tissue and obtaining large numbers of precisely staged embryos. Importantly, avian cardiogenic mesoderm can be

PART  |  1  Heart Evolution

isolated and cultured with or without anterior endoderm as well as with emerging pharmacological and recombinant agents to test effects on cardiogenic determination and differentiation. We refer readers to Gannon and Bader (1997) for detailed protocols for these procedures. Finally, with the dawn of stem cell biology and its potential for the analysis of cell determination and differentiation, it is important to remember that a clear understanding of cardiogenic cell lineage processes is essential. The avian system remains one of the few experimental models where the embryonic precursors of the myocardium can be identified, isolated, cultured, manipulated and analyzed for its developmental potential.

V.  Early morphogenetic changes in the forming heart tube The heart tube is formed from the rostral to caudal fusion of the paired lateral primordia beginning at stage 6. These lateral elements are brought medially and ventrally by the action of the lateral body folds. Accompanying endocardial tubes are formed medial to the myocardial mantle of progenitors on both sides of midline. While left and right endocardial progenitors form bilateral tubes, left and right myocardial mantles do not truly form tube structures. They are epithelial, nonlumenal structures that reside laterally to the endocardial tubes. With progression of the lateral folds, bilateral endocardial tubes fuse (a line can be seen at the interface of the two tubes as this process begins), the two myocardial mantles meet and the foregut is positioned dorsal to the heart tube. Now the myocardium forms a single epithelium of two or three myocytes in thickness. No epicardium is present at this time (Manasek, 1969a,b, 1970; Reese et al., 2002; Nahirney et al., 2003). Fusion or meeting of the two bilaterally positioned wings of the cardiogenic crescent has been recently re-examined in the chick (Moreno-Rodriguez et al., 2006). A summary of their studies is given in Fig. 4. Originally, Stalsberg and DeHaan proposed that the anterior-most region of the cardiogenic crescent fused at the midline and zippered down posteriorly, bringing the two wings together as a single heart tube (Stalsberg and DeHaan, 1969). More recent vital dye labeling and explant studies have shown that the wings of the cardiogenic crescent actually meet and fuse at a position posterior to the cranial-most heart forming cells. Subsequent zippering shut of the crescent into a single heart tube occurs anteriorly and posteriorly from this point (Fig. 3 and Chapter 3.2). The addition of cells to the forming anterior heart tube from the second heart field occurs after the initial fusion event (see Chapters 2.2 and 3.1). Thus, the avian heart represents an outstanding experimental model to examine the early stages of crescent fusion and tube formation.

Chapter  |  1.5  An Overview of Avian Heart Structure and Development

Figure 4  Closure of the bilateral wings of the cardiogenic crescent and formation of a heart tube. A model of fusion and formation of the primitive heart tube is given. Transverse sections through an embryo in the region of the cardiogenic crescent are diagrammed in (A–C) at stages 6–9. In panel (A), the coelom (arrow) is formed when lateral mesoderm cavitates to form somatic and splanchnic layers at approximately stage 6. The two wings of the cardiogenic crescent approach (B) and fuse (C) as the anterior intestinal portal closes. (A–C) These are ventral views showing mesoderm only. Moreno-Rodriguez et al. present their theory on the fusion of myocardial mantles via a “zipper complex” that is formed and fuses in a bidirectional fashion (anterior/cephalic versus posterior/ caudal; arrows). The mesoderm turns (seen as twisted arrows in (A)), and the outer margins (blue) form a ventral fusion line that can be seen through a simple dissecting microscope. This originates in the prospective right ventricle (seen as green horizontal lines in (C)). The forming pericardial cavity is seen at the same time (two arrowheads). Fusion proceeds in both directions and permits the caudal differentiation of the left ventricle (white area). The future outflow tract is seen as the anterior yellow area. To summarize, the outlet tissue (yellow), right ventricle (horizontal hatch), left ventricle and AV canal (white), sino-atrial tissue (perpendicular hatch) and ventral fusion line (blue) are identified during these stages of development. The green represents differentiating myocytes identified by MF20 staining during several developmental stages (B, C, D). Modified by permission from Moreno-Rodriguez et al. (2006).

V.A.  Diversification of Myogenic Cell Lineages Several studies suggest that anterior lateral mesoderm takes on cardiomyogenic potential long before the heart is formed. Moorman and colleagues recognized this in a study of the three-dimensional distribution of atrial and ventricular isomyosins during the formation of the tubular chick heart (HH stage 7 to 12) using antibodies specific for atrial and ventricular myosins. This study demonstrated that both myosins were expressed at stage 8, when the cardiogenic crescent is still two separate plates. The antiventricular myosin antibody bound cardiogenic cells adjacent to the anterior intestinal portal. The atrial myosin was identified in regions caudal and lateral to ventricular myosin expression. Regions medial to atrial myosin expression did not express sarcomeric myosin (De Jong et al., 1990). Moorman has subsequently published many studies on this topic in several different species, and the reader is encouraged to review those publications and Chapter 3.2 for a comprehensive analysis. Later studies have found largely similar, yet specifically variant, results at these early stages of heart development in the chick (Gonzalez-Sanchez and

127

Figure 5  Position of ventricular and atrial MHC-expressing cells at stage 10. The region of VMHC1 staining depicts the position of all differentiated myocytes as this gene product is expressed in all striated muscle cells (left panel). In the right panel, the regions of AMHC1 staining show the regions of the forming heart tube that will later incorporate into the embryonic atria. It should be noted that addition of the secondary heart fields and later morphogenesis are not considered in this representation. Modified from Yutzey et al. (1994).

Bader, 1990; Sugi and Lough, 1994; Schultheiss et al., 1995; Yutzey and Bader, 1995; Gannon and Bader, 1995, 1997; Yatskievych et al., 1997). Interestingly, diversification of ventricular and atrial myocytes appears to occur very early along the anteroposterior axis in the forming heart tube. In the chick, the straight heart tube is never really “straight,” and there is never a stage where anterior is “equal” to posterior. In situ hybridization analyses demonstrate that anterior progenitors express ventricular myosin heavy chain gene products prior to bilateral myocardial mantle fusion (Yutzey et al., 1994). The same progenitors do not express atrial myosin transcripts at detectable levels at these same times (Fig. 5). Obviously, variation in probe preparation and size, as well as length of hybridization time and stringency, can play a role in the interpretation of in situ experiments (Somi et al., 2006), but general patterns of expression do exhibit consistent results. Expression of the atrial myosin heavy chain is largely confined to the posterior regions of the cardiogenic crescent and forming tube, suggesting that cardiomyogenic differentiation varies along the anteroposterior axis. Isolation of anterior and posterior progenitors followed by culture in standard medium demonstrated that only posterior cells activate the atrial-specific myosin gene, further indicating that differentiative potential varies along the anteroposterior axis (Yutzey et al., 1994). These experiments also suggest that expression of atrialspecific genes is not dependent on formation of the definitive atrium, as posterior myogenic cells will activate these genes in vitro. Interestingly, this initial diversification of myocytes can be modulated by treatment with retinoic acid (Yutzey et al., 1995; Yutzey and Bader, 1995), and is controlled in the chick and mouse by the same mechanism

128

PART  |  1  Heart Evolution

ct v sv

36 Hour embryo

a

a

ct

sv

v

48 Hour embryo

ct a sv v

72 Hour embryo

Figure 6  Early morphogenetic moments of the heart tube. Changes in the relative positions of the conotruncus (ct), ventricles (v), atria (a) and sinus venosus (sv) in the developing chicken heart is seen in these illustrations. Modified from McCain and McLaughin (McCain and McLaughin, 1998) by permission.

(Wang et al., 2001). In this situation, application of retinoic acid results in conversion of anterior ventriculogenic cells to an atrial phenotype. Conversely, misexpression of the ventricular region-specific iroquois gene product in developing avian embryos is thought to activate the ventricular phenotype along the anterior–posterior axis (Bao et al., 1999). The atrial myosin heavy chain gene product encoded by AMHC1 has the most highly restricted pattern of gene expression along this axis (Yutzey et al., 1994). Additionally, it should be noted that the chick heart, unlike mammalian hearts, expresses more than two heavy chain genes, and one atrial isoform becomes restricted to the atria only much later in development (Evans et al., 1988). While differences in gene expression have garnered much attention in recent years as a model for myocyte and chamber diversification, simple observation of heart morphology at the time of bilateral fusion demonstrates the initial diversification of the heart in terms of anteroposterior variation. Most anteriorly, the aortic sac, truncus arteriosis and bulbus chordus are restricted in diameter when compared to the ventriculogenic region that bulges to the right. It should be remembered that the aortic outlet (the aortic sac, truncus arteriosis, bulbus chordus) are derived from the secondary heart field (Evans et al., 1988; Mjaatvedt et al., 2001; Waldo et al., 2005). Posteriorly, a restriction is easily seen in the heart tube. This restriction coincides with the line that separates AMHC1-expressing myocytes from anterior myocytes that do not express AMHC1. The right and left posterior cardiac regions of

the stage 10 embryo straddle the anterior intestinal portal and look nothing like their anterior counterparts. It is interesting to note that chick embryos isolated and grown on paper rings develop hearts that undergo these anteroposterior morphogenetic events. Diversification of atrial and ventricular cells along the anteroposterior axis appears to be a component of the initial differentiation process in chick heart development (Yutzey and Bader, 1995). This is not to say that the final fully-differentiated phenotype of the ventricular or atrial phenotype is established at this early stage (for example, see Evans et al., 1988). However, this initially diverse state may play a role in driving differential development of ventricles and atria. As stated above, the avian heart tube is “diverse” along its anteroposterior and left–right axes from the moment it is formed with its specific constrictions and protrusions. It is also obvious to even the most casual observer that the avian heart is diverse along its left–right axis. The ventricular region of the heart loops to the right. Without doubt, early mechanisms in the establishment of right–left asymmetry in the embryo as a whole drive subsequent asymmetry in the heart (Raya and Belmonte, 2006; Tabin, 2006). One of the more fascinating aspects of myogenic diversification in the developing avian heart is the generation of the conduction system. Interestingly, the avian heart, like that of the mammal, displays rhythmic posterior-to-anterior contractions very early in development. This topic is extensively reviewed in Moorman and Christoffels (2003a,b and Chapter 2.3). While this indicates the presence of a defined

Chapter  |  1.5  An Overview of Avian Heart Structure and Development

conduction system, morphological or molecular characterization of these cells is not advanced. Moorman and colleagues have proposed that the embryonic heart may be reflective of the hearts of more primitive organisms where every myocyte might be considered a “conduction system” cell (Anderson et al., 2004). For theories on the generation of the central conduction system, see Gourdie et al. and other contributors to the Novartis Foundation Symposium (2003) (Gourdie et al., 2003). Generation of Purkinje cells in the embryonic ventricle has been more firmly established in the avian heart. Mikawa and co-workers have established that working ventricular myocytes give rise to conduction system cells within the developing heart wall (Gourdie et al., 1995, 1998, 2003; Hyer et al., 1999). Interaction with the developing coronary system is critical in lineage determination of some Purkinje fibers, and endothelin appears to be a major signaling factor in this process. Obviously, morphogenesis and growth of the heart continue long after the initial establishment of the asymmetric tube and primitive chambers. This is demonstrated in Fig. 4. At day three, the heart is seen as a looped structure where atrial and ventricular myocytes are clearly positioned within their developing chambers. While the diversification process is still incomplete (Stockdale, 1992; Wang et al., 1996; Stockdale et al., 2002), atrial and ventricular myocytes are clearly positioned in the heart and express at least one MHC specific to their respective adult chambers (Gonzalez-Sanchez and Bader, 1984). While other chapters focus on the differentiation of the cardiac chambers and their overall structure, it is interesting to consider the atrioventricular canal seen in Fig. 7. This is a major structure in the embryonic heart and provides a muscular connection between ventricular and atrial chambers. It is a critical signaling structure for the generation of AV values. However, this myogenic connection between ventricular and atrial chambers is completely missing in the adult. The mechanisms underlying this morphogenetic process are almost completely unknown. The chick embryo may be an excellent model for this study. Thus, the avian system has provided critical data in the examination of myogenic lineage determination, and continues to provide a model for these analyses by the experimental zoologist.

VI.  Trabeculation and cardiac myocytes Trabeculation is a complex process whereby the epithelial myocardium expands to produce finger-like projections into the cardiac lumen. A diagram of this overall process is given in Fig. 8 (modified from Sedmera et al., 1999). Depending on the species studied, trabeculae may be retained in the adult heart without subsequent compaction. This is generally true for species with circulatory systems

129

Figure 7  Prominence of the atrioventricular canal in the early chick heart. This scanning EM modified from McCain and McLaughin (McCain and McLaughin, 1998) demonstrates the atrioventricular canal that is prominently seen from day 3 through 5. This is an important signaling center for generation of the atrioventricular valves. Interestingly, this structure and its muscular band is absent in the adult heart.

under lower blood pressures. For example, most amphibian species and many fish have noncompacted, trabeculated ventricles. All avian hearts studied thus far have compacted hearts (Hicks, 2002). As with other classes of vertebrates, there are significant differences in the trabeculating potential of the avian atria and ventricles. While atria undergo trabeculation, most of the literature has focused on the process in the ventricles (Sedmera et al., 1999). In the chick beginning around stage 13 or 14, the polarized epithelial myocardium is one-to-three cells thick (Peng et al., 1990). At this time, spaces in the myocardium are seen between myocytes (Nahirney et al., 2003). It should be remembered that cardiac myocytes are the only cell type seen in the myocardium before the arrival of the epicardium (Reese et al., 2002, see references cited therein). As the outer heart wall expands by proliferation, myocytes change their distribution of adhesion molecules including cadherins ZO1 and Bves (Osler and Bader, unpublished results). These myocytes have numerous cell processes that appear to interact with other myocytes (Mikawa and Fischman, 1992;

130

PART  |  1  Heart Evolution

Proliferating cardiomocytes Differentiating cardiomocytes Endocardial cells (A)

Endothelial cells Epicardial cells

(B)

(C)

(D) Figure 8  Diagram of the morphogenetic events regulating trabeculation and compaction. This diagram from an insightful review by Sedmera and colleagues summarizes the proliferation, growth, movement and differentiation of cardiac myocytes in the transmural axis. Along with this cell type, the position of endocardial and epicardial cells is noted. Modified from Sedmera et al. (2000).

Nahirney et al., 2003). The Fischman and Mikawa laboratories have shown that clones of cardiac myocytes extend from the outer wall transmurally to the lumen. Thus, trabeculae are polyclonal and derived from radial expansion of proliferative myocytes of the outer wall. Butcher et al. (2007) have recently generated a new and very sensitive method for analysis of trabeculation in the chick heart. Proliferation is a key to trabeculation, and disruption of many gene products leads to a blunting of growth and extension. For example, Hatcher et al. have determined that Tbx5 is a critical regulator of myocyte proliferation of avian cardiomyogenic cells in vitro and in vivo, potentially regulating trabeculation (Hatcher et al., 2004; Hatcher and McDermott, 2006). Lee and colleagues have made critical contributions to our understanding of trabeculation using the mouse as a model. In these studies, Erb2, a receptor tyrosine kinase, and its ligand neuregulin-1 were shown to play essential roles in the selective trabeculation of the

ventricular wall (Lee et al., 1995; Negro et al., 2004). Loss of Erb2 caused major blunting of ventricular trabeculation and death in embryonic mice (Lee et al., 1995). Other growth factors including FGFs are critical for the progression of myocyte division and expansion of the avian heart wall (Mima et al., 1995a,b). Thompson and colleagues have provided extensive analysis of myocyte proliferation patterns in the developing avian ventricular wall that have been reviewed (Sedmera et al., 1999, 2000). Briefly, myocytes of the outer wall remain proliferative and addition of new myocytes is appositional (Fig. 6). Proliferation of myocytes at the tips of forming trabeculae is modest. It is of interest to question when the last mesodermal cells are added to the ventricular wall, and when addition of new myocytes is strictly from the division of pre-existing myocytes. No stem cells have been identified in avian hearts, as is claimed in mammalian counterparts (Anversa et al., 2007).

Chapter  |  1.5  An Overview of Avian Heart Structure and Development

During this phase of wall growth, spaces in the myocardium are observed and cells of the developing coronary system move to form the vascular system of the heart (Reese et al., 2002). The mechanisms whereby spaces in the proliferating, differentiating myocardium are eliminated and intervening endocardial cells are moved or removed are not well-understood. Species of birds with noncompacted heart walls have not been reported.

VII.  Summary The avian heart remains an important component in the analysis of heart development. While not yet easily manipulated at the genomic level, descriptive embryology, experimental manipulation, lineage tracing and in vitro analysis of cardiomyogenesis, as well as determination of endothelial and epicardial differentiation, are often best approached in this class of vertebrates. With the recent sequencing of the chick genome along with the many advances in nongenomic manipulation of gene product expression, the avian heart will continue to play an essential role in discovering the mechanistic basis of cardiac morphogenesis.

References Anderson, R.H., Christoffels, V.M., Moorman, A.F., 2004. Controversies concerning the anatomical definition of the conduction tissues. Anat. Rec. B. New Anat. 280, 8–14. Andree, B., Duprez, D., Vorbusch, B., Arnold, H.H., Brand, T., 1998. BMP-2 induces ectopic expression of cardiac lineage markers and interferes with somite formation in chicken embryos. Mech. Dev. 70, 119–131. Anversa, P., Leri, A., Rota, M., Hosoda, T., Bearzi, C., Urbanek, K., Kajstura, J., Bolli, R., 2007. Concise review: stem cells, myocardial regeneration, and methodological artifacts. Stem. Cells 25, 589–601. Bao, Z.Z., Bruneau, B.G., Seidman, J.G., Seidman, C.E., Cepko, C.L., 1999. Regulation of chamber-specific gene expression in the developing heart by Irx4. Science 283, 1161–1164. Barron, M., Gao, M., Lough, J., 2000. Requirement for BMP and FGF signaling during cardiogenic induction in non-precardiac mesoderm is specific, transient, and cooperative. Dev. Dyn. 218, 383–393. Bettiol, E., Clement, S., Krause, K.H., Jaconi, M.E., 2006. Embryonic and adult stem cell-derived cardiomyocytes: lessons from in vitro models. Rev. Physiol. Biochem. Pharmacol. 157, 1–30. Butcher, J.T., Norris, R.A., Hoffman, S., Mjaatvedt, C.H., Markwald, R.R., 2007. Periostin promotes atrioventricular mesenchyme matrix invasion and remodeling mediated by integrin signaling through Rho/PI 3-kinase. Dev. Biol. 302, 256–266. de Groot, I.J., Hardy, G.P., Sanders, E., Los, J.A., Moorman, A.F., 1985. The conducting tissue in the adult chicken atria. A histological and immunohistochemical analysis. Anat. Embryol. (Berl). 172, 239–245. De Jong, F., Geerts, W.J., Lamers, W.H., Los, J.A., Moorman, A.F., 1990. Isomyosin expression pattern during formation of the tubular chicken

131

heart: a three-dimensional immunohistochemical analysis. Anat. Rec. 226, 213–227. DeHaan, R.L., 1963. Migration patterns of the precardiac mesoderm in the early chick embrvo. Exp. Cell Res. 29, 544–560. DeHaan, R.L., Ursprung, H., 1965. Organogenesis. Holt, Reinhart & Winston, New York, NY. Ehrman, L.A., Yutzey, K.E., 1999. Lack of regulation in the heart forming region of avian embryos. Dev. Biol. 207, 163–175. Evans, D., Miller, J.B., Stockdale, F.E., 1988. Developmental patterns of expression and coexpression of myosin heavy chains in atria and ventricles of the avian heart. Dev. Biol. 127, 376–383. Firulli, A.B., Thattaliyath, B.D., 2002. Transcription factors in cardiogenesis: the combinations that unlock the mysteries of the heart. Int. Rev. Cytol. 214, 1–62. Gannon, M., Bader, D., 1995. Initiation of cardiac differentiation occurs in the absence of anterior endoderm. Development 121, 2439–2450. Gannon, M., Bader, D., 1997. Avian cardiac progenitors: methods for isolation, culture, and analysis of differentiation. Methods Cell Biol. 52, 117–132. Garcia-Martinez, V., Schoenwolf, G.C., 1993. Primitive-streak origin of the cardiovascular system in avian embryos. Dev. Biol. 159, 706–719. Gilbert, S.F., 2006. Developmental Biology. Sinauer Associates, Inc., Sunderland, MA. Gonzalez-Sanchez, A., Bader, D., 1984. Immunochemical analysis of myosin heavy chains in the developing chicken heart. Dev. Biol. 103, 151–158. Gonzalez-Sanchez, A., Bader, D., 1985. Characterization of a myosin heavy chain in the conductive system of the adult and developing chicken heart. J. Cell Biol. 100, 270–275. Gonzalez-Sanchez, A., Bader, D., 1990. In vitro analysis of cardiac progenitor cell differentiation. Dev. Biol. 139, 197–209. Gourdie, R.G., Mima, T., Thompson, R.P., Mikawa, T., 1995. Terminal diversification of the myocyte lineage generates Purkinje fibers of the cardiac conduction system. Development 121, 1423–1431. Gourdie, R.G., Wei, Y., Kim, D., Klatt, S.C., Mikawa, T., 1998. Endothelin-induced conversion of embryonic heart muscle cells into impulse-conducting Purkinje fibers. Proc. Natl. Acad. Sci. USA 95, 6815–6818. Gourdie, R.G., Harris, B.S., Bond, J., Edmondson, A.M., Cheng, G., Sedmera, D., O’Brien, T.X., Mikawa, T., Thompson, R.P., 2003. His–Purkinje lineages and development. Novartis Found Symp. 250, 110–122; discussion 122–124, 276–279. Hamburger, V., Hamilton, H.L., 1992. A series of normal stages in the development of the chick embryo. Dev. Dyn. 195, 231–272. Hatcher, C.J., McDermott, D.A., 2006. Using the TBX5 transcription factor to grow and sculpt the heart. Am. J. Med. Genet. A. 140, 1414–1418. Hatcher, C.J., Diman, N.Y., Kim, M.S., Pennisi, D., Song, Y., Goldstein, M.M., Mikawa, T., Basson, C.T., 2004. A role for Tbx5 in proepicardial cell migration during cardiogenesis. Physiol. Genomics. 18, 129–140. Hicks, J.W., 2002. The physiological and evolutionary significance of cardiovascular shunting patterns in reptiles. News Physiol. Sci. 17, 241–245. Hirakow, R., 1970. Ultrastructural characteristics of the mammalian and sauropsidan heart. Am. J. Cardiol. 25, 195–203. Hyer, J., Johansen, M., Prasad, A., Wessels, A., Kirby, M.L., Gourdie, R.G., Mikawa, T., 1999. Induction of Purkinje fiber differentiation

132

by coronary arterialization. Proc. Natl. Acad. Sci. USA 96, 13214–13218. Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1, 435–440. Kruithof, B.P., van Wijk, B., Somi, S., Kruithof-de Julio, M., Perez Pomares, J.M., Weesie, F., Wessels, A., Moorman, A.F., van den Hoff, M.J., 2006. BMP and FGF regulate the differentiation of multipotential pericardial mesoderm into the myocardial or epicardial lineage. Dev. Biol. 295, 507–522. Ladd, A.N., Yatskievych, T.A., Antin, P.B., 1998. Regulation of avian cardiac myogenesis by activin/TGFbeta and bone morphogenetic proteins. Dev. Biol. 204, 407–419. Le Douarin, G., Le Douarin, N., Cuminge, D., 1965. Ultrastructural study of the autodifferentiation capabilities of the mesoderm of the cardiac area in the chick embryo. C. R. Acad. Sci. Hebd. Seances Acad. Sci. D. 260, 6998–7001. Lee, K.F., Simon, H., Chen, H., Bates, B., Hung, M.C., Hauser, C., 1995. Requirement for neuregulin receptor erbB2 in neural and cardiac development. Nature 378, 394–398. Lough, J., Barron, M., Brogley, M., Sugi, Y., Bolender, D.L., Zhu, X., 1996. Combined BMP-2 and FGF-4, but neither factor alone, induces cardiogenesis in non-precardiac embryonic mesoderm. Dev. Biol. 178, 198–202. Lu, Y., James, T.N., Bootsma, M., Terasaki, F., 1993a. Histological organization of the right and left atrioventricular valves of the chicken heart and their relationship to the atrioventricular Purkinje ring and the middle bundle branch. Anat. Rec. 235, 74–86. Lu, Y., James, T.N., Yamamoto, S., Terasaki, F., 1993b. Cardiac conduction system in the chicken: gross anatomy plus light and electron microscopy. Anat. Rec. 236, 493–510. Malpighi, M., 1672. De Formatione Pulli in Ovo. Royal Society, London. Manasek, F.J., 1969a. Embryonic development of the heart. II. Formation of the epicardium. J. Embryol. Exp. Morphol. 22, 333–348. Manasek, F.J., 1969b. The appearance of granules in the Golgi complex of embryonic cardiac myocytes. J. Cell Biol. 43, 605–610. Manasek, F.J., 1970. Histogenesis of the embryonic myocardium. Am. J. Cardiol. 25, 149–168. Manner, J., 2000. Cardiac looping in the chick embryo: a morphological review with special reference to terminological and biomechanical aspects of the looping process. Anat. Rec. 259, 248–262. Marvin, M.J., Di Rocco, G., Gardiner, A., Bush, S.M., Lassar, A.B., 2001. Inhibition of Wnt activity induces heart formation from posterior mesoderm. Genes. Dev. 15, 316–327. Mathur, R., Shrivastava, A., 1979. Cardiac conducting system of Gallus domesticus with special reference to the atrial bundles. Acta. Morphol. Neerl. Scand. 17, 181–189. McCain, E.R., McLaughlin, J., 1998. Developmental and physiological aspects of the chicken embryonic heart. Proceedings of the 20th Workshop/Conference of the Association for Biology Laboratory Education 20, 1–15. Mikawa, T., 1995. Retroviral targeting of FGF and FGFR in cardiomyocytes and coronary vascular cells during heart development. Ann. NY Acad. Sci. 752, 506–516. Mikawa, T., Fischman, D.A., 1992. Retroviral analysis of cardiac morphogenesis: discontinuous formation of coronary vessels. Proc. Natl. Acad. Sci. USA 89, 9504–9508. Mikawa, T., Borisov, A., Brown, A.M., Fischman, D.A., 1992. Clonal analysis of cardiac morphogenesis in the chicken embryo using a

PART  |  1  Heart Evolution

replication-defective retrovirus: I. Formation of the ventricular myocardium. Dev. Dyn. 193, 11–23. Mima, T., Ohuchi, H., Noji, S., Mikawa, T., 1995a. FGF can induce outgrowth of somatic mesoderm both inside and outside of limb-forming regions. Dev. Biol. 167, 617–620. Mima, T., Ueno, H., Fischman, D.A., Williams, L.T., Mikawa, T., 1995b. Fibroblast growth factor receptor is required for in vivo cardiac myocyte proliferation at early embryonic stages of heart development. Proc. Natl. Acad. Sci. USA 92, 467–471. Mjaatvedt, C.H., Nakaoka, T., Moreno-Rodriguez, R., Norris, R.A., Kern, M.J., Eisenberg, C.A., Turner, D., Markwald, R.R., 2001. The outflow tract of the heart is recruited from a novel heart-forming field. Dev. Biol. 238, 97–109. Montgomery, M.O., Litvin, J., Gonzalez-Sanchez, A., Bader, D., 1994. Staging of commitment and differentiation of avian cardiac myocytes. Dev. Biol. 164, 63–71. Moorman, A.F., Christoffels, V.M., 2003a. Cardiac chamber formation: development, genes, and evolution. Physiol. Rev. 83, 1223–1267. Moorman, A.F., Christoffels, V.M., 2003b. Development of the cardiac conduction system: a matter of chamber development. Novartis Found Symp. 250, 25–34; discussion 34–43, 276–279. Moreno-Rodriguez, R.A., Krug, E.L., Reyes, L., Villavicencio, L., Mjaatvedt, C.H., Markwald, R.R., 2006. Bidirectional fusion of the heart-forming fields in the developing chick embryo. Dev. Dyn. 235, 191–202. Nahirney, P.C., Mikawa, T., Fischman, D.A., 2003. Evidence for an extracellular matrix bridge guiding proepicardial cell migration to the myocardium of chick embryos. Dev. Dyn. 227, 511–523. Negro, A., Brar, B.K., Lee, K.F., 2004. Essential roles of Her2/erbB2 in cardiac development and function. Recent Prog. Horm. Res. 59, 1–12. Osler, M.E., Bader, D.M., 2004. Bves expression during avian embryogenesis. Dev. Dyn. 229, 658–667. Pabon-Pena, L.M., Goodwin, R.L., Cise, L.J., Bader, D., 2000. Analysis of CMF1 reveals a bone morphogenetic protein-independent component of the cardiomyogenic pathway. J. Biol. Chem. 275, 21453–21459. Peng, I., Dennis, J.E., Rodriguez-Boulan, E., Fischman, D.A., 1990. Polarized release of enveloped viruses in the embryonic chick heart: demonstration of epithelial polarity in the presumptive myocardium. Dev. Biol. 141, 164–172. Proctor, N.S., Lynch, P.J., 1993. Manual of Ornithology: Avian Structure & Function. Yale University Press, New Haven, CT. Rawles, M.E., 1952. Transplantation of normal embryonic tissues. Ann. NY Acad. Sci. 55, 302–312. Raya, A., Belmonte, J.C., 2006. Left–right asymmetry in the vertebrate embryo: from early information to higher-level integration. Nat. Rev. Genet. 7, 283–293. Redkar, A., Montgomery, M., Litvin, J., 2001. Fate map of early avian cardiac progenitor cells. Development 128, 2269–2279. Reese, D.E., Mikawa, T., Bader, D.M., 2002. Development of the coronary vessel system. Circ. Res. 91, 761–768. Rosenquist, G.C., 1970a. Cardia bifida in chick embryos: anterior and posterior defects produced by transplanting tritiated thymidine-labeled grafts medial to the heart-forming regions. Teratology 3, 135–142. Rosenquist, G.C., 1970b. Location and movements of cardiogenic cells in the chick embryo: the heart-forming portion of the primitive streak. Dev. Biol. 22, 461–475. Satin, J., Fujii, S., DeHaan, R.L., 1988. Development of cardiac beat rate in early chick embryos is regulated by regional cues. Dev. Biol. 129, 103–113.

Chapter  |  1.5  An Overview of Avian Heart Structure and Development

Schlange, T., Andree, B., Arnold, H., Brand, T., 2000. Expression analysis of the chicken homologue of CITED2 during early stages of embryonic development. Mech. Dev. 98, 157–160. Schultheiss, T.M., Xydas, S., Lassar, A.B., 1995. Induction of avian cardiac myogenesis by anterior endoderm. Development 121, 4203–4214. Sedmera, D., Pexieder, T., Rychterova, V., Hu, N., Clark, E.B., 1999. Remodeling of chick embryonic ventricular myoarchitecture under experimentally changed loading conditions. Anat. Rec. 254, 238–252. Sedmera, D., Pexieder, T., Vuillemin, M., Thompson, R.P., Anderson, R.H., 2000. Developmental patterning of the myocardium. Anat. Rec. 258, 319–337. Slack, J.M., 1991. From Egg to Embryo: Regional Specification in Early Development. Cambridge University Press, Cambridge, UK, and New York. Somi, S., Klein, A.T., Houweling, A.C., Ruijter, J.M., Buffing, A.A., Moorman, A.F., van den Hoff, M.J., 2006. Atrial and ventricular myosin heavy-chain expression in the developing chicken heart: strengths and limitations of non-radioactive in situ hybridization. J. Histochem. Cytochem. 54, 649–664. Sommer, J.R., Bossen, E., Dalen, H., Dolber, P., High, T., Jewett, P., Johnson, E.A., Junker, J., Leonard, S., Nassar, R., et al., 1991. To excite a heart: a bird’s view. Acta. Physiol. Scand. Suppl. 599, 5–21. Sommer, J.R., Johnson, E.A., 1969. Cardiac muscle. A comparative ultrastructural study with special reference to frog and chicken hearts. Z Zellforsch Mikrosk Anat. 98, 437–468. Stalsberg, H., DeHaan, R.L., 1968. Endodermal movements during foregut formation in the chick embryo. Dev. Biol. 18, 198–215. Stalsberg, H., DeHaan, R.L., 1969. The precardiac areas and formation of the tubular heart in the chick embryo. Dev. Biol. 19, 128–159. Stockdale, F.E., 1992. Myogenic cell lineages. Dev. Biol. 154, 284–298. Stockdale, F.E., Nikovits, W., Jr, Espinoza, N.R., 2002. Slow myosins in muscle development. Results Probl. Cell Differ. 38, 199–214. Sugi, Y., Lough, J., 1994. Anterior endoderm is a specific effector of terminal cardiac myocyte differentiation of cells from the embryonic heart forming region. Dev. Dyn. 200, 155–162. Szabo, E., Viragh, S., Challice, C.E., 1986. The structure of the atrioventricular conducting system in the avian heart. Anat. Rec. 215, 1–9. Tabin, C.J., 2006. The key to left–right asymmetry. Cell 127, 27–32.

133

Takebayashi-Suzuki, K., Pauliks, L.B., Eltsefon, Y., Mikawa, T., 2001. Purkinje fibers of the avian heart express a myogenic transcription factor program distinct from cardiac and skeletal muscle. Dev. Biol. 234, 390–401. Takebayashi-Suzuki, K., Yanagisawa, M., Gourdie, R.G., Kanzawa, N., Mikawa, T., 2000. In vivo induction of cardiac Purkinje fiber differentiation by coexpression of preproendothelin-1 and endothelin converting enzyme-1. Development 127, 3523–3532. van Laake, L.W., Hassink, R., Doevendans, P.A., Mummery, C., 2006. Heart repair and stem cells. J. Physiol. 577, 467–478. Verzi, M.P., McCulley, D.J., De Val, S., Dodou, E., Black, B.L., 2005. The right ventricle, outflow tract, and ventricular septum comprise a restricted expression domain within the secondary/anterior heart field. Dev. Biol. 287, 134–145. Waldo, K.L., Hutson, M.R., Ward, C.C., Zdanowicz, M., Stadt, H.A., Kumiski, D., Abu-Issa, R., Kirby, M.L., 2005. Secondary heart field contributes myocardium and smooth muscle to the arterial pole of the developing heart. Dev. Biol. 281, 78–90. Wang, G.F., Nikovits, W., Schleinitz, M., Stockdale, F.E., 1996. Atrial chamber-specific expression of the slow myosin heavy chain 3 gene in the embryonic heart. J. Biol. Chem. 271, 19836–19845. Wang, G.F., Nikovits, W., Jr, Bao, Z.Z., Stockdale, F.E., 2001. Irx4 forms an inhibitory complex with the vitamin D and retinoic X receptors to regulate cardiac chamber-specific slow MyHC3 expression. J. Biol. Chem. 276, 28835–28841. West, N.H., Langille, B.L., Jones, D.R., 1981. Cadiovascular system. Form and Function in Birds 2, 235–339. Whittow, G.C., Sturkie, P.D., 2000. Sturkie’s Avian Physiology. Academic Press, San Diego, CA. Yatskievych, T.A., Ladd, A.N., Antin, P.B., 1997. Induction of cardiac myogenesis in avian pregastrula epiblast: the role of the hypoblast and activin. Development 124, 2561–2570. Yutzey, K.E., Bader, D., 1995. Diversification of cardiomyogenic cell lineages during early heart development. Circ. Res. 77, 216–219. Yutzey, K.E., Rhee, J.T., Bader, D., 1994. Expression of the atrial-specific myosin heavy chain AMHC1 and the establishment of anteroposterior polarity in the developing chicken heart. Development 120, 871–883. Yutzey, K., Gannon, M., Bader, D., 1995. Diversification of cardiomyogenic cell lineages in vitro. Dev. Biol. 170, 531–541.

Chapter 2.1

The Specification of Myocardial Progenitor Cells in the Ciona Tadpole Michael Levine Department of Molecular and Cell Biology, Division of Genetics, Genomics and Development, Center for Integrative Genomics, University of California, Berkeley, CA, USA

I.  Overview

II.  Specification of the heart field

Ciona intestinalis is a urochordate or tunicate, the simplest members of the Phylum Chordata. The Ciona tadpole offers a number of advantages for studying gene networks and how these networks control cellular behavior (reviewed by Passamaneck and Di Gregorio, 2005; Davidson, 2007). First, Ciona possesses a small and compact genome (Dehal et al., 2002). At about 150 Mb and with approximately 16,000 genes it contains the same complexity as the Drosophila genome, yet it is chordate and thereby presents an opportunity for studying developmental processes directly applicable to vertebrates. Second, the definitive swimming tadpole is composed of only around 2,500 cells, and a complete lineage map is available. It is possible to trace even complex processes, such as heart morphogenesis, to the fertilized egg. And third, transgenic DNAs can be introduced rapidly into large, synchronous populations of developing embryos via electroporation (Corbo et al., 1997). This procedure permits the rapid characterization of cis-regulatory DNAs, as well as the assessment of gene function through the expression and misexpression of dominant-negative and constitutively-active transcription factors and cell signaling components (e.g., Takahashi et al., 1999). The combination of a simple genome, detailed cell lineage information, and the facility of transgenic manipulation places Ciona in a unique position as a model for exploring how gene regulatory networks control cellular morphogenesis. Here, I summarize recent advances in our understanding of the earliest events in the specification of myocardial progenitor cells in the Ciona tadpole.

The adult sea squirt possesses a one-chamber, U-shaped cardiac tube enclosed by a pericardial sheath (e.g., Anderson, 1968). The heart has a muscular, rhythmic beat that is reversible, due to the appearance of pacemaker cells on either side of the heart tube. After beating an average of once every second for minutes or hours, it suddenly changes direction and then beats equally well in the opposite direction, before reversing once again. Compared to the overall simplicity of the adult body plan, the Ciona heart is rather sophisticated and shares obvious affinity with the vertebrate heart (see Chapter 1.1 Evolutionary Origins of Hearts). Fate-map studies in vertebrate embryos suggest that the heart field is established prior to gastrulation. In mouse embryos, it appears that the field is composed of approximately 50 cells on either side of the ventral midline in anterior regions of the epiblast (reviewed by Abu-Issa and Kirby, 2007). These cells undergo invagination into the primitive streak and then emerge in a thoracic position of the lateral mesoderm to form the cardiac mesoderm. A similar process is seen in the Ciona embryo, but it is considerably simpler due to the reduced number of cells. The heart field is initially composed of just two cells, one on either side of the animal–vegetal axis (Fig. 1). These are the so-called B7.5 blastomeres, which form at the intersection of the presumptive tail muscles (possibly analogous to the vertebrate paraxial mesoderm) and primitive gut (Davidson and Levine, 2003). They are the first cells of the mesoderm to enter the blastopore during gastrulation, immediately after the ingression of the gut.

Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

137

138

PART  |  2  Cardiac Precursor Populations and Lineages

An interesting property of the early heart progenitors is that they undergo coherent gastrulaton. That is, they always appear at the rostral-most position of the invaginated mesoderm after the completion of gastrulation. This coherent invagination might depend on transient contacts with the endoderm. It is also conceivable that the heart field helps guide the invagination of the more caudal mesoderm derivatives (the future tail muscles). After gastrulation, the heart field is positioned between the mesoderm and ectoderm of neurula-stage embryos (Fig. 2A). Prior to gastrulation, the B7.5 blastomeres exhibit selective expression of a bHLH regulatory gene, Mesp (Satou et al., 2004; Davidson et al., 2005). There are two homologous genes in vertebrates, Mesp1 and Mesp2, and both have been implicated in early heart specification, as well as somitogenesis (Saga et al., 2000; Morimoto et al., 2006). In Ciona, Mesp expression is observed shortly after the clonal restriction of the heart field. All of the derivatives of the B7.5 blastomeres form either the heart primoridium or anterior tail muscles. At earlier stages (Fig. 1) a common precursor gives rise to both the B7.5 blastomere, as well as the progenitors of the germline (Davidson et al., 2005). It is unclear whether the early heart field of vertebrates shares any such connection to primordial germ cells.

B7.3

B

B7.7 B7.5

B7.6

B7.8

Figure 1  The cardiac mesoderm in 64–110 cell embryos, prior to gastrulation. The entire heart field is derived from the pair of B7.5 cells. The sister lineage, B7.6, produces germ cells. The shaded cells give rise to the tail muscles.

(A)

(B)

Figure 2  The cells labeled in red correspond to the B7.5 lineage (each B7.5 blastomere has undergone one division), and those in green are the future tail muscles. The large cells stained blue correspond to the endoderm. The B7.5 cells (arrowheads) are located between endoderm, muscles, and epidermis (outer cells in (A)). The B7.5 cells undergo an asymmetric division at the onset of neurulation (purple arrows in (B) indicate the future heart cells).

Mesp is essential for the specification of the heart field. Morpholino-based gene disruption assays lead to a loss of the heart primordium in tadpole-stage embryos (Satou et al., 2004). However, in morpholino-treated embryos, the B7.5 blastomeres undergo coherent invagination during gastrulation and appear in the anterior-most region of the mesoderm in the neurula and tailbud stages. Hence, Mesp is required for delineating the early heart field, but not for normal gastrulation movements (Davidson et al., 2006). The regulatory factors and signaling molecules responsible for coherent gastrulation of the heart field are currently unknown.

II.A.  Fibroblast Growth Factor Signaling   and Heart Induction The B7.5 blastomeres undergo one division at the conclusion of gastrulation, and a second division at the conclusion of neurulation (Fig. 2). Thus, each of the original heart field blastomeres produces four grand-daughter cells; each group of four cells is located on either side of the ventral midline at the leading edge of the developing tail muscles. The first division appears to be more or less symmetric, in that it produces daughter cells of equivalent size. However, the second division is asymmetric, producing small rostral cells and large caudal cells (Fig. 2B). The small cells are destined to form the heart primordium, while the large cells give rise to the anterior-most tail muscles (Davidson et al., 2006). This asymmetric division is the first indication of the cardiac mesoderm, although the basis for this asymmetry is unknown. The small rostral B7.5 cells are induced to form myocardial progenitor cells. This induction depends on a localized fibroblast growth factor (FGF) signal. There are only six FGF genes in the Ciona genome (as compared with over 20 such genes in vertebrates) (Satou et al., 2002), and only one of the six, FGF9, is expressed at the right time and place to serve as the inductive signal (Imai et al., 2002, 2006; Bertrand et al., 2003). The B7.5 lineage is sensitized for response to FGF due to the prior action of Mesp, which is only transiently expressed during gastrulation. Mesp either directly or indirectly activates EtsP2, which encodes a transcriptional effector of FGF signaling (Tsang and Dawid, 2004; Davidson et al., 2006). EtsP2 is expressed in the entire B7.5 lineage, and persists during the asymmetric division that produces the small rostral cells and large caudal cells. Normally, the EtsP2 transcription factor is only activated in the rostral cells, due to its proximity to the localized FGF9 signal. In contrast, the caudal B7.5 cells do not receive the signal, or at least not sufficient levels of the signal, and form anterior tail muscles by default. There is evidence that FGF signaling defines the anterior heart field in vertebrates (Zaffran et al., 2003). A variety of evidence is consistent with this simple gene network for the specification of heart precursor cells (Mesp-EtsP2FGF9). For example, a dominant negative form of the

139

Chapter  |  2.1  The Specification of Myocardial Progenitor Cells in the Ciona Tadpole

EtsP2 transcription factor blocks heart induction in the rostral B7.5 cells (Fig. 3), which are transformed into supernumerary tail muscles (Davidson et al., 2006). Conversely, ectopic activation of FGF signaling in the caudal cells causes them to form myocardial progenitor cells, leading to the formation of an enlarged heart rudiment with twice the normal number of cells.

II.B.  Directed Migration of Myocardial Progenitor Cells FoxF is a direct target gene of the Mesp-EtsP2-FGF9 signaling pathway. It is selectively expressed in the myocardial progenitor cells almost immediately after the asymmetric division of the B7.5 lineage in neurula-stage embryos (Fig. 4C) (Beh et al., 2007). FoxF is essential for their directed migration from the anterior tail region of the developing tadpole to the posterior region of the head (Fig. 4A,B). This portion of the Ciona tadpole is thought to correspond to thoracic regions of vertebrate embryos, due to the evolutionary loss of anterior regions of the ascidian tail (Hox genes 7–9 are absent) (Ikuta et al., 2004; Keys et al., 2005). Overall, directed migration of the myocardial progenitor cells is evocative of related processes in vertebrate embryos (reviewed by Abu-Issa and Kirby, 2007). In Ciona, migration is restricted to a total of just four cells, two on either side of the ventral midline. It is possible to visualize this process in detail, including the live imaging of labeled cells using Mesp-GFP fusion genes (Davidson et al., 2006). Disruption of FoxF gene activity inhibits heart cell migration (Beh et al., 2007). Different methods were used to attenuate FoxF function. Fertilized eggs were injected with morpholinos directed against the FoxF mRNA, similar to the methods used in zebrafish and Xenopus (e.g., Heasman, 2002). The identification of cell-specific enhancers, coupled with the ease of electroporation, also permitted the targeted expression of a dominant-negative form of FoxF, FoxF/WRPW (Beh et al., 2007). The WRPW peptide mediates interactions with the general corepressor protein, Groucho/TLE, which functions as a robust (A)

(B)

Figure 3  Inhibition of FGF signaling blocks heart formation in the Ciona tadpole. The red cells correspond to the four descendants of a single B7.5 blastomere. Normally, the anterior cells migrate into posterior head regions to form the heart rudiment (A). However, the expression of a dominant-negative form of EtsP2 causes the anterior cells to remain in the tail to form supernumerary tail muscles (B).

transcriptional repressor (Jennings et al., 2006). Directed expression of the FoxF/WRPW fusion protein in the B7.5 lineage using the Mesp enhancer produced the same phenotype as that obtained by MO injection, loss of directed heart cell migration (Fig. 5B; compare with A) (Beh et al., 2007). Electroporation also permitted genetic epistasis studies. As discussed earlier, constitutive activation of FGF signaling in both the rostral and caudal B7.5 lineages results in the transformation of caudal cells into additional myocardial progenitor cells (Fig. 5C; compare with A) (Davidson et al., 2006). Such activation was achieved with a constitutive EtsP2 fusion protein, EtsP2/VP16, which no longer depends on FGF signaling for post-translational activation. Targeted expression of EtsP2/VP16 using the Mesp enhancer resulted in the specification and directed migration of both the rostral and caudal B7.5 grand-daughter cells, and a doubling in the size of the heart rudiment (A)

(B)

(C)

Figure 4  Migration of myocardial progenitor cells. The small, rostral descendants of the B7.5 blastomeres (purple arrows) are associated with their larger sister cells (white arrows) in neurula-stage embryos (A). They undergo directed migration during tailbud stages, while the sister lineage forms muscles in the anterior tail (B). FoxF is activated shortly after the asymmetric division of the B7.5 lineage to form the rostral myocardial progenitor cells (C).

(A)

(C)

(B)

(D)

Figure 5  FoxF is essential for migration. Embryos were electroporated with a Mesp-GFP fusion gene to label the B7.5 lineage. (A) Normal tadpole. The anterior B7.5 cells have migrated into the head, while the posterior cells remain in the tail and form muscles. (B) The same as (A), except that the embryo was co-electroporated with a fusion gene containing the Mesp enhancer driving the expression of a dominant-negative form of FoxF (FoxF/WRPW). Both the anterior and posterior lineages remain in the tail. (C) The same as (A), except that the embryo was coelectroporated with a Mesp fusion gene containing a constitutively active form of EtsP2 (EtsP2/VP16). Both the anterior and posterior B7.5 cells migrate into the head and form myocardial progenitor cells. (D) The same as (C), except that the embryo was also electroporated with the MespFoxF/WRPW fusion gene. The inhibition of FoxF activity causes all of the cells to remain in the tail.

140

(Davidson et al., 2006). Simultaneous expression of a Mesp-FoxF/WRPW transgene led to a striking block in migration (Fig. 5D) (Beh et al., 2007). Both the normal myocardial progenitors (the rostral B7.5 lineage) and the transformed heart cells (the caudal lineage) remain in the anterior tail and fail to migrate. Thus, FoxF is epistatic over the Mesp-EtsP2-FGF9 signaling pathway with respect to directed migration.

II.C.  Uncoupling Heart Migration and Differentiation After migration, the four myocardial progenitor cells (two on either side of the ventral midline) undergo a series of asymmetric divisions. The myocardial “mother cell” divides to produce one large and one small daughter cell (Davidson and Levine, 2003). The larger cell remains on the outer part of the heart rudiment, and after several rounds of asymmetric divisions the smaller internal cells meet at the ventral midline to form a single rudiment composed of 12–24 cells (depending on the developmental stage) (Fig. 6). It is possible, but unproven, that the two cell types in the heart rudiment form distinct tissues in the beating hearts of juveniles and adults. For example, the larger mother cells might form the pericardial sheath, while the smaller daughter cells might form the cardiomyo­ cyes and heart tube. The Ciona tadpole lacks a beating heart and does not require one. It is only 1 mm in length, a vigorous swimmer and quite thin. As a result, there is no problem with the passive exchange of O2/CO2. After the tadpole swims around for 8–12 hours it attaches to a substrate (e.g., the bottom of a boat in a marina) and undergoes metamorphosis. After 1–2 days a beating heart is clearly evident in the juvenile. It undergoes reversals at more frequent intervals than the adult heart, but nonetheless exhibits steady, rhythmic beats (Davidson et al., 2006).

PART  |  2  Cardiac Precursor Populations and Lineages

The heart is located between the endostyle (primitive thyroid gland) and the stomach in the body cavity of juveniles and adults (Fig. 7A). During metamorphosis the entire tail is resorbed and destroyed via programed cell death. This includes all of the tail muscles, as well as the notochord and associated neural tube. The anterior tail muscles, derived from the caudal B7.5 lineage, are also destroyed in this process during metamorphosis. Interestingly, B7.5 cells expressing EtsP2-VP16 and FoxFWRPW are not destroyed during the programed destruction of the tail. Instead, these cells differentiate into a beating heart located on the wrong side of the body cavity (Fig. 7B). It is located between the stomach and the mass of resorbed tissue from the tail. These observations suggest that FoxF is required specifically for the process of directed migration of the myocardial progentior cells, and not for heart differentiation. However, the hearts that form in the tail region of FoxF mutants are not normal. They have erratic beating rhythms and fail to produce blood vessels (Beh et al., 2007). As a result, there is no blood flow and the mutant juveniles die. It is currently unclear whether the “tail hearts” are defective due to a failure in migration, or reflect a requirement of FoxF gene activity for heart differentiation in addition to migration.

II.D.  Heart Differentiation Network A core network of interacting transcription factors, Nkx2.5/tinman, Mef2 and Gata4,5,6, have been implicated in the differentiation of cardiomyocytes in a broad spectrum of animals, including flies and humans (Harvey, 1996; Bodmer and Venkatesh, 1998; Frasch, 1999; Olson, 2006). Orthologs of Nkx2.5 (NK-4) and Gata4,5,6 (GataA) are expressed in the heart rudiment of Ciona tadpoles (Davidson and Levine, 2003). GataA expression first appears in the myocardial progenitors prior to migration,

(A)

Figure 6  Asymmetric division of the myocardial progenitor cells leads to the formation of a single heart rudiment. The large, outer “mother” cells produce smaller, internal daughter cells.

(B)

Figure 7  Block in migration causes heart to form in the wrong place. (A) Normal juvenile, high magnification view. The arrowhead identifies the beating heart, which is located between the endostyle (e) and stomach (s). (B) Juvenile obtained from mutant tadpole that was co-electroporated with both the Mesp-EtsP2/VP16 and Mesp-FoxF/WRPW fusion genes (see Fig. 5D). The myocardial progenitor cells fail to migrate out of the tail, and as a result the heart forms on the wrong side of the stomach (s). (*) indicates the normal location of the heart, while the arrow identifies the misplaced heart.

Chapter  |  2.1  The Specification of Myocardial Progenitor Cells in the Ciona Tadpole

while NK-4 is activated after migration. The Ciona Mef2 gene does not appear to be expressed in the heart rudiment, but it might be expressed in differentiating heart cells during metamorphosis. After migration, the heart cells come into contact with ventral epidermal cells in the posterior region of the head that express BMP2/4. This signal is critical for heart differentiation in both flies and vertebrates (Olson, 2006), so it is conceivable that it is also important for the activation of heart differentiation genes after the directed migration of the myocardial progenitor cells. Perhaps defects in the “tail hearts” arise from the failure of the misplaced heart rudiment to receive this critical BMP signal. It is currently unclear how the Mesp-EtsP2-FGF9-FoxF heart migration network in Ciona intersects with the established cardiomyocyte gene cassette in vertebrates. One potential complication is the lag between heart migration and differentiation in Ciona. As discussed earlier, the heart rudiment remains quiescent in the tadpole, but there is no comparable lag in the heart differentiation program of vertebrates. Preliminary studies suggest that GataA might provide a link between FoxF-mediated migration and heart differentiation. GataA appears to be activated by the MespEtsP2 network and is essential for early migration.

III.  Future prospects The directed migration of myocardial progenitor cells in the developing Ciona tadpole is evocative of comparable processes in complex vertebrate systems (e.g., Abu-Issa and Kirby, 2007; Schoenebeck and Yelon, 2007). The heart progenitors located on either side of the ventral midline migrate to “thoracic” regions of the tadpole, and subsequently fuse at the ventral midline (Li et al., 2004). In Ciona this process appears to depend on asymmetric divisions of myocardial “mother” cells (Fig. 6), similar to the asymmetric divisions of CNS neurons in the Drosophila embryo (Wodarz, 2005). Future studies will identify FoxF target genes required for the initial phase of directed heart cell migration. This process almost certainly depends on changes in cell polarity and adhesion, as well as the formation of directed membrane protrusions at the leading edge of the migrating cells. Classical biochemical methods, along with more recent proteomics technology, have identified many of the protein complexes and signaling pathways responsible for these cellular processes. It is currently unclear how the Mesp-EtsP2-FGF9-FoxF network impinges on these processes. Perhaps FoxF induces the transcription of just a few key rate-limiting components essential for migration, or weakly influences the expression of many such components. Indeed, the integration of gene networks and cellular effectors is one of the great promises of the Ciona tadpole as an experimental model for organogenesis.

141

References Abu-Issa, R., Kirby, M.L., 2007. Heart field: from mesoderm to heart tube. Annu. Rev. Cell Dev. Biol. 23, 45–68. Anderson, M., 1968. Electrophysiological studies on initiation and reversal of the heart beat in Ciona intestinalis. J. Exp. Biol. 49, 363–385. Beh, J., Shi, W., Levine, M., Davidson, B., Christiaen, L., 2007. FoxF is essential for FGF-induced migration of heart progenitor cells in the ascidian Ciona intestinalis. Development 134, 3297–3305. Bertrand, V., Hudson, C., Caillol, D., Popovici, C., Lemaire, P., 2003. Neural tissue in ascidian embryos is induced by FGF9/16/20, acting via a combination of maternal GATA and Ets transcription factors. Cell 115, 615–627. Bodmer, R., Venkatesh, T.V., 1998. Heart development in Drosophila and vertebrates: conservation of molecular mechanisms. Dev. Genet. 22, 181–186. Corbo, J.C., Levine, M., Zeller, R.W., 1997. Characterization of a notochord-specific enhancer from the Brachyury promoter region of the ascidian, Ciona intestinalis. Development 124, 589–602. Davidson, B., 2007. Ciona intestinalis as a model for cardiac development. Semin. Cell Dev. Biol. 18, 16–26. Davidson, B., Levine, M., 2003. Evolutionary origins of the vertebrate heart: Specification of the cardiac lineage in Ciona intestinalis. Proc. Natl. Acad. Sci. 100, 11469–11473. Davidson, B., Shi, W., Levine, M., 2005. Uncoupling heart cell specification and migration in the simple chordate Ciona intestinalis. Development 132, 4811–4818. Davidson, B., Shi, W., Beh, J., Christiaen, L., Levine, M., 2006. FGF signaling delineates the cardiac progenitor field in the simple chordate, Ciona intestinalis. Genes Dev. 20, 2728–2738. Dehal, P., Satou, Y., Campbell, R.K., Chapman, J., Degnan, B., De Tomaso, A., Davidson, B., Di Gregorio, A., et al., 2002. The draft genome of Ciona intestinalis: insights into chordate and vertebrate origins. Science 298, 2157–2167. Frasch, M., 1999. Intersecting signalling and transcriptional pathways in Drosophila heart specification. Semin. Cell Dev. Biol. 10, 61–71. Harvey, R.P., 1996. NK-2 homeobox genes and heart development. Dev. Biol. 178, 203–216. Heasman, J., 2002. Morpholino oligos: making sense of antisense? Dev. Biol. 243, 209–214. Ikuta, T., Yoshida, N., Satoh, N., Saiga, H., 2004. Ciona intestinalis Hox gene cluster: Its dispersed structure and residual colinear expression in development. Proc. Natl. Acad. Sci. 101, 15118–15123. Imai, K.S., Satoh, N., Satou, Y., 2002. Early embryonic expression of FGF4/6/9 gene and its role in the induction of mesenchyme and notochord in Ciona savignyi embryos. Development 129, 1729–1738. Imai, K.S., Levine, M., Satoh, N., Satou, Y., 2006. Regulatory blueprint for a chordate embryo. Science 312, 1183–1187. Jennings, B.H., Pickles, L.M., Wainwright, S.M., Roe, S.M., Pearl, L.H., Ish-Horowicz, D., 2006. Molecular recognition of transcriptional repressor motifs by the WD domain of the Groucho/TLE corepressor. Mol. Cell 22, 645–655. Keys, D.N., Lee, B.I., Di Gregorio, A., Harafuji, N., Detter, J.C., Wang, M., Kahsai, O., Ahn, S., Zhang, C., Doyle, S.A., Satoh, N., Satou, Y., Saiga, H., Christian, A.T., Rokhsar, D.S., Hawkins, T.L., Levine, M., Richardson, P.M., 2005. A saturation screen for cis-acting regulatory DNA in the Hox genes of Ciona intestinalis. Proc. Natl. Acad. Sci. 102, 679–683.

142

Li, S., Zhou, D., Lu, M.M., Morrisey, E.E., 2004. Advanced cardiac morphogenesis does not require heart tube fusion. Science 305, 1619–1622. Morimoto, M., Kiso, M., Sasaki, N., Saga, Y., 2006. Cooperative Mesp activity is required for normal somitogenesis along the anterior– posterior axis. Dev. Biol. 300, 687–698. Olson, E.N., 2006. Gene regulatory networks in the evolution and development of the heart. Science 313, 1922–1927. Passamanec, Y.J., Di Gregorio, A., 2005. Ciona intestinalis: chordate development made simple. Dev. Dyn. 233, 1–19. Saga, Y., Kitajima, S., Miyagawa-Tomita, S., 2000. Mesp1 expression is the earliest sign of cardiovascular development. Trends Cardiovasc. Med. 10, 345–352. Satou, Y., Imai, K.S., Satoh, N., 2002. Fgf genes in the basal chordate Ciona intestinalis. Dev. Genes Evol. 212, 432–438.

PART  |  2  Cardiac Precursor Populations and Lineages

Satou, Y., Imai, K.S., Satoh, N., 2004. The ascidian Mesp gene specifies heart precursor cells. Development 131, 2533–2541. Schoenebeck, J.J., Yelon, D., 2007. Illuminating cardiac development: Advances in imaging add new dimensions to the utility of zebrafish genetics. Semin. Cell Dev. Biol. 18, 27–35. Takahashi, H., Hotta, K., Erives, A., Di Gregorio, A., Zeller, R.W., Levine, M., Satoh, N., 1999. Brachyury downstream notochord differentiation in the ascidian embryo. Genes Dev. 13, 1519–1523. Tsang, M., Dawid, I.B., 2004. Promotion and attenuation of FGF signaling through the Ras-MAPK pathway. Sci. STKE 228, pe17. Wodarz, A., 2005. Molecular control of cell polarity and asymmetric cell division in Drosophila neuroblasts. Curr. Opin. Cell Biol. 17, 475–481. Zaffran, S., Kelly, R., Munk, A., Brown, N., Buckingham, M.E., 2003. The mouse as a model for heart morphogenesis in mammals: the origin of myocytes and studies with cardiac explants. J. Soc Biol. 197, 187–194.

Chapter 2.2

The Second Heart Field Robert G. Kelly1 and Sylvia M. Evans2 1

Developmental Biology Institute of Marseilles-Luminy, UMR 6216 CNRS-Université de la Méditerranée, Campus de Luminy, Marseilles, France  2 Skaggs School of Pharmacy and Pharmaceutical Sciences, and Department of Medicine, University of San Diego, La Jolla, CA, USA

I.  Introduction The vertebrate heart is a complex four-dimensional s­tructure required to function throughout life from the earliest stages of embryonic development, which in adult amniotes directs parallel systemic and pulmonary circulatory systems. Heart development involves the orchestrated deployment of multiple cell lineages and differentiated cell types controlled by overlapping genetic regulatory networks and environmental influences, in particular haemodynamics. It has recently become apparent that cardiomyocytes of the adult heart are derived from multiple progenitor cell populations, including first and second heart field progenitor cells. The first heart field gives rise to the linear heart tube and definitive left ventricle, and part of the inflow region of the heart. In contrast, the second heart field, identified only in 2001, gives rise to right ventricular and outflow tract (OFT) myocardium, in addition to part of the inflow region. Cells of the second heart field are situated in pharyngeal mesoderm contiguous with the poles of the heart tube, where they are recruited during the concomitant and extremely rapid processes of heart tube elongation and looping. Discovery of the second heart field has had, in a short period of time, a major impact on our understanding of heart development, in particular concerning early heart morphogenesis, the etiology of congenital heart defects and the characterization of cardiac progenitor cells. The second heart field paradigm provides a framework for deconstruction of the complex events underlying early heart development, and in the process has unearthed a new series of questions. In this chapter we will discuss the f­indings that Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

preceded and led to the identification of this progenitor cell population, and different molecular and genetic experiments evaluating the overlap between the first and second heart fields and their contributions to the definitive heart. We will subsequently examine extensive new data that has led to molecular insights into the transcription factors and signaling pathways which regulate second heart field deployment. Finally, we will consider the biomedical significance of the second heart field for our understanding of common forms of congenital heart defects and the properties of cardiac progenitor cells in the postnatal heart.

II.  Discovery and initial characterization of the second heart field II.A.  The Avian Second Heart Field Three papers published in 2001 demonstrated that pharyngeal mesoderm contributes to the elongating outflow tract of the chick and mouse hearts during cardiac looping (Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001). These cells were termed the anterior heart-forming or secondary heart field. Although the source of the progenitor cells had not previously been identified, the fact that the heart tube elongates by addition of cells to the growing poles had been appreciated for many years. In 1973, in a seminal light and electron microscopy study of mouse heart development, Viragh and Challice described cells undergoing an epithelial to myocardial transition at the border between the heart tube and contiguous splanchnic 143

144

pharyngeal mesoderm. This process continues up to day 10.5 of mouse development, three days after differentiation of the first cardiomyocytes in the cardiac crescent (Viragh and Challice, 1973). Maria Victoria de la Cruz and colleagues, in an elegant series of experimental manipulations including lineage tracing in live embryos, demonstrated that the chick heart tube extends not only by addition of new cardiomyocytes at the posterior end, but also anterior to the point of initial fusion of the left and right cardiac primordia. Embryos bisected at the anterior end of the fusion point showed beating cardiomyocytes anterior to the initial heart tube (Arguello et al., 1975). Subsequently, de la Cruz and colleagues used iron oxide particle-labeling experiments to demonstrate that the future smooth walled component of the right ventricle, in addition to the proximal and distal regions of the outflow tract, are not present in the newly-formed heart tube, but appear progressively between stages HH9 and HH22 (de la Cruz et al., 1977). These experiments established the important, but initially not widely-appreciated, concept that not all regions of the definitive heart are present in the linear heart tube (Fig. 1A). The importance of an “extracardiac” splanchnic mesoderm contribution to the heart tube during the progressive formation of the proximal and distal outflow tract was established in an analysis of rat and human embryonic hearts by de Vries, who noted that heart tube elongation occurs concomitantly with posterior displacement of the arterial pole of the heart tube in the pharyngeal region (de Vries, 1981). Thus, these different studies demonstrated that the heart tube elongates by addition of cells to the arterial pole. The importance of this pre-existing evidence for a second heart field became apparent in 2001, with discovery of the source of these arterial pole progenitor cells in chick and mouse through the application of molecular and fluorescent labeling techniques. Directly inspired by the work of de la Cruz and colleagues, Mjaatvedt et al. (2001) demonstrated using viral infection and vital dye lineage analysis, together with ablation of the classically-defined heart forming fields, that the outflow tract of the chick heart grows not by expansion of the embryonic right ventricle, but by addition of cardiomyocytes from a progenitor cell population in cephalic mesoderm. These cells, which surround the aortic sac immediately anterior to the existing heart tube, were termed the anterior heart-forming or anterior heart field (Fig. 1B). Waldo et al. (2001) used gene and epitope expression, coupled with vital dye lineage analysis, to show that myocardium of the distal outflow tract in the chick heart forms by accretion of cells from a secondary heart field located in splanchnic mesoderm underlying the pharynx, contiguous and posterior to the arterial pole of the heart tube (Fig. 1C). The transcription factors Nkx2.5 and Gata4 were found to be expressed in these splanchnic mesoderm cells, and the carbohydrate epitope HNK-1 involved in cell–cell or cell–substrate interactions

PART  |  2  Cardiac Precursor Populations and Lineages

associated with cell motility is expressed in splanchnic mesoderm proximal to the arterial pole and in newly differentiated myosin positive cells in the outflow tract itself, at stages corresponding to the time that the outflow tract is added to the heart tube, as defined by de la Cruz and colleagues (Waldo et al., 2001). Both of these initial studies in the chick addressed the mechanism underlying arterial pole extension. Mjaatvedt et al. used microdissected tissue explants to show that anterior heart-forming field explants formed contractile myocytes, either in the presence of serum or myocardium from the distal rim, but not future trabeculated component, of the right ventricle, suggesting that myocardium at the arterial pole may be competent to recruit undifferentiated pharyngeal mesoderm cells into the growing heart tube. Waldo et al. (2001) identified a candidate pathway involved in this signaling process, showing that distal outflow tract myocardium expresses BMP2, and that the BMP inhibitor noggin blocks myocardial differentiation of secondary heart field explants. These authors argued that the dynamic accretion of myocardium to the elongating outflow tract may result from recruitment of second heart field cells via a BMP signal, whereas progenitor cell proliferation is maintained by fibroblast growth factor signaling in the pharyngeal region (Fig. 1C) (Waldo et al., 2001).

II.B.  The Mammalian Second Heart Field Evidence for the growth of the mammalian outflow tract from pharyngeal mesoderm progenitor cells came from the analysis of a fortuitous enhancer trap transgene expressed in arterial pole myocardium and contiguous pharyngeal mesoderm (Kelly et al., 2001). Transgene expression was initially observed in a crescent situated in splanchnic mesoderm medial to the lateral population of differentiated cardiomyocytes comprising the cardiac crescent and giving rise to the linear heart tube. Expression was subsequently observed in mesoderm dorsal and anterior to the heart tube in an expression domain that corresponded temporally with the time during which the arterial pole of the mouse heart tube extends, as defined by Viragh and Challice (Fig. 2A) (Kelly et al., 2001). DiI labeling confirmed a ventral movement of pharyngeal mesoderm cells into the elongating heart tube, and characterization of the transgene integration site led to the discovery of the first molecular marker of the second heart field, Fibroblast growth factor 10 (Fgf10) (Fig. 2B). Explant analysis using left and right ventricular restricted transgene expression, together with additional DiI labeling experiments, has confirmed that pharyngeal mesoderm gives rise to the right ventricle, in addition to outflow tract myocardium (Fig. 2C) (Zaffran et al., 2004). These findings were significantly extended by the identification of another gene expressed in the second heart field, Isl1, encoding the LIM homeodomain transcription

Chapter  |  2.2  The Second Heart Field

145

(A)

(B)

(C)

Figure 1  The avian second heart field. (A) Cartoon showing the results of in vivo labeling experiments in the chick embryonic heart carried out by de la Cruz and colleagues. Label c, placed at the anterior end of the linear heart tube at Hamburger Hamilton (HH) stage 9, is found at the border of the trabeculated region of the right ventricle at HH36, demonstrating that the non-trabeculated component of the right ventricle and outflow tract myocardium are not present in the linear heart tube. Reproduced from de la Cruz et al. (1991) Cardiol. Young 1, 123, with permission of Cambridge University Press. (B) After in vivo ablation of the classical heart fields (asterisk) at HH8, a beating outlet structure (arrow) is formed by HH11. Reproduced from Mjaatvedt et al. (2001) Dev. Biol. 238, 97–109, with permission of Elsevier. (C) Injection of mitotracker into the second heart field (asterisk) at HH14 reveals the addition of two new segments to the outflow tract by HH22, one fluorescent (asterisk) and one non-fluorescent (star). Right panel: model for dynamic accretion of myocardium from the second heart field to the elongating heart tube through BMP-mediated recruitment. Reproduced from Waldo et al. (2001) Development 128, 3179–3188, with permission of the Company of Biologists.

146

(A)

(B)

(C)

Figure 2  The mammalian second heart field: Fibroblast growth f­actor 10. (A) An Fgf10-nlacZ enhancer trap transgene labels outflow tract myocardium and contiguous pharyngeal mesoderm of the second heart field (arrowheads) at embryonic day E 9.5 in a ventral view with the heart removed. The contribution of the second heart field to the elongating outflow tract is almost complete by E10.5. (B) Fgf10 transcripts accumulate medially to differentiated cells of the cardiac crescent at E7.5. (C) DiI label placed at the arterial pole of the mouse heart tube prior to looping is found in left ventricular myocardium after 24 hours of embryo culture, suggesting that the linear heart tube gives rise to the left ventricle and that the right ventricle is a second heart field derivative. (B) and (C) Reproduced from Kelly et al. (2001) Dev. Cell 1, 435–40, with permission of Elsevier.

factor Islet-1 (Cai et al., 2003). Isl1 had previously been shown to be expressed in mesoderm surrounding the embryonic coelom in the chick (Yuan and Schoenwolf, 2000). Cai et al. (2003) used genetic lineage tracing in the mouse to demonstrate that Isl1 labels pharyngeal mesoderm progenitor cells situated medial to the cardiac crescent, whose descendants give rise to a majority of definitive cardiomyocytes, in particular, the right ventricle and outflow tract, but also part of the left ventricle and inflow region of the heart (Fig. 3A,B). The observation that Isl1 cells contribute to both poles of the elongating heart tube was a major advance in understanding the importance of the contribution of pharyngeal mesoderm progenitor cells to heart development. Isl1 is required for the process

PART  |  2  Cardiac Precursor Populations and Lineages

of heart tube elongation; in Isl1 mutant embryos the l­inear heart tube fails to extend and loop, and BMP and FGF gene expression is downregulated in the distal heart tube and pharyngeal region (Fig. 3C) (Cai et al., 2003). The definitive heart is therefore comprised of cardiomyocytes derived from two lineages: an Isl1 negative lineage contributing to the linear heart tube, definitive left ventricle and part of the inflow region; and an Isl1 positive lineage giving rise to the remainder of the heart, including the outflow tract, right ventricle, some cells within the left ventricle and a majority of the inflow region. Recent data have suggested that in addition to expression in the second heart field, Isl1 may be transiently expressed in the first heart field. A conditional reporter gene activated by a new Isl1-Cre line is more broadly expressed in the heart, although part of the left ventricle remains unlabeled (Sun et al., 2007), and Isl1 protein has been detected in the cardiac crescent at E7.5 (Prall et al., 2006). These observations suggest that Isl1 may be expressed in both the first and second heart fields. Support for this conclusion comes from a study of Xenopus Isl1 which is co-localized with Nkx2.5 at early stages of heart development (Brade et al., 2007). Nevertheless, the murine Isl1-null phenotype and Isl1 knockdown experiments in Xenopus show that Isl1 is not necessary for initial heart tube formation, revealing a critical requirement for Isl1 function in the second heart field, but not the first heart field. Genetic evidence for the existence of two lineages of cardiomyocytes has come from a retrospective clonal analysis of mouse heart development using a defective -galactosidase encoding reporter gene, nlaacZ (which carries a duplication rendering it inactive), targeted to the cardiac alpha-actin locus (Meilhac et al., 2004). The nlaacZ system relies on a random, low-frequency recombination event restoring nlacZ reporter gene activity, and permits unbiased analysis of cell behavior during development (Meilhac et al., 2003). Three classes of clones were observed at day 8.5 of mouse development: clones restricted to the embryonic left ventricle and inflow region; clones contributing to the outflow or both outflow and inflow regions; and a third class contributing to all regions of the heart tube. The first two classes of clones correspond to the first and second lineages, as defined by Isl1 expression (Fig. 4). The third class of clones are the only class contributing to both the outflow tract and left ventricle, and occur at a low frequency, suggesting that such clones arise early in development and that the first heart field and second heart field diverge from a common pancardiac lineage in the early embryo (Meilhac et al., 2004).

II.C.  The Second Heart Field Paradigm Together, these data have led to a new paradigm for heart morphogenesis by which genetically distinct cell lineages give rise to different components of the definitive heart

147

Chapter  |  2.2  The Second Heart Field

(A)

(B)

(C)

Figure 3  The mammalian second heart field: Islet1. (A) Isl1 transcripts accumulate in pharyngeal mesoderm at E8.5 (arrow in left panel), whereas transcripts for Myosin light chain 2a accumulate in differentiated cardiomyocytes (middle panel). Lineage tracing using an Isl1-Cre recombinase allele reveals that Isl1 descendants colonize the poles of the heart tube (right panel). (B) Transverse sections showing Isl1 transcripts (blue) in mesoderm contiguous, but medial to, differentiated Mlc2a-positive cells of the cardiac crescent and linear heart tube (red). (C) Isl1 descendants fail to colonize the heart of Isl1/ embryos; scanning electron microscopy reveals that the Isl1/ heart remains unlooped and fails to elongate. (A–C) reproduced from Cai et al. (2003) Dev. Cell 5, 877–889, with permission of Elsevier.

(A)

1 1

5

4 3

2

6

3

6 6

(B)

Secondary heart field Tbx1 CNC anterior heart field Foxh1 First heart field

Second heart field Isl1

Distal outflow tract Proximal outflow tract right ventricle Left ventricle atria inflow tract

Atria inflow tract

Figure 4  The two lineage model of heart development. (A) E8.5 nlacZ/nlaacZ chimeric hearts showing clonally-related b-galactosidase-positive cardiomyocytes in the future left ventricle (3) and inflow region (4, 6) of the heart (left panel), in both the arterial (1, 2) and venous poles (5, 6) of the heart (middle panel) or in all compartments of the embryonic heart (right panel). Reproduced from Meilhac et al. (2004) Dev. Cell 6, 685–698, with permission of Elsevier. (B) Cartoon showing the contribution of different myocardial lineages to the embryonic heart tube. The first lineage gives rise to the linear heart tube and future left ventricle (black) whereas the second lineage gives rise to myocardium at the venous pole of the heart (blue), in addition to cells of the anterior heart field which contribute to the right ventricle (red) and the secondary heart field which gives rise to outflow tract myocardium (orange). Adapted from Kelly, (2005) Trends Cardiovasc. Med. 15, 51–56, with permission of Elsevier.

148

PART  |  2  Cardiac Precursor Populations and Lineages

(reviewed in Kelly and Buckingham, 2002; Abu-Issa et al., 2004; Buckingham et al., 2005; Kelly, 2005). At the cardiac crescent stage these two lineages, the first heart field and second heart field, lie adjacent in lateral and medial splanchnic pharyngeal mesoderm respectively (Fig. 5A). Subsequent morphogenetic events spatially separate the second heart field from differentiated cardiomyocytes of the linear heart tube. As the linear heart tube forms, it is initially a trough open dorsally to pharyngeal endoderm (Fig. 5B). Second heart field cells, which remain closely apposed to ventral pharyngeal endoderm, contribute to the heart tube along its entire length, giving rise to the dorsal wall of the heart tube and the future inner curvature of the heart. This process terminates when the dorsal mesocardium, the mesentery by which the heart is suspended in the ventral region of the embryo, breaks down, thus isolating cells of the second heart field in the dorsal pericardial wall. Continuity is maintained at the poles of the heart through

which second heart field cells subsequently contribute to the rapidly elongating heart tube (Fig. 5C). Convergence of the inflow and outflow poles of the heart, likely driven by caudal displacement of the arterial pole of the heart in the pharynx, and rightward looping occur at the same stage (between days 8 and 9 of mouse development) during which period the heart tube increases its linear dimensions five-fold. It is important to note that, in addition to differential expression of regulatory genes such as Isl1 and spatial separation in the dorsal pericardial wall, second heart field cells are characterized by a differentiation delay with respect to cells giving rise to the linear heart tube. The second heart field thus provides the major part of the myocardium which is added to the pre-existing scaffold of the linear heart tube. There are two principal components to the second heart field: (1) cells of the anterior heart field which contribute to elongation of the arterial pole of the heart tube; and (2) cells which contribute

(A)

a b

3 – 4 somites, E7.5

a

b

6 – 8 somites, E8

c

d

11–13 somites, E8.5

e

f

(B)

c d

(C)

e f

Figure 5  The second heart field paradigm. (A) At the cardiac crescent stage cells of the second heart field (green) are positioned medially to differentiated cells of the cardiac crescent (red), with which they form a contiguous cell population. (B) At the linear heart tube stage, cells of the second heart field are observed in pharyngeal mesoderm lateral to and underlying foregut endoderm (yellow) in the dorsal pericardial wall up to the level of the dorsal mesocardium. (C) As the heart tube loops and elongates, cells from the second heart field contribute to myocardium at the arterial and venous poles (blue). Adapted from Cai et al. (2003) Dev. Cell 5, 877–889, with permission of Elsevier.

Chapter  |  2.2  The Second Heart Field

to elongation of the venous pole of the heart. Within the arterial pole progenitor cell population the secondary heart field, as defined by Waldo et al. (2001), represents that population of the anterior heart field which gives rise to distal outflow tract myocardium and the terminal phase of second heart field deployment, coincident with the movement of neural crest cells into the heart (Waldo et al., 2001; Abu-Issa et al., 2004). Each of these properties of the second heart field (differentiation delay, juxtaposition with pharyngeal endoderm and the differing cellular environments at the arterial and venous poles of the heart tube) are likely to be instructive in regulating the progressive deployment of the second heart field during heart formation.

III.  Evaluating the contribution   of the second heart field III.A.  The Contribution of the Second Heart Field at the Arterial Pole We are still at an early stage in our understanding of the second heart field. The cellular continuity with progenitors of the linear heart tube at the cardiac crescent stage, and later across the dorsal mesocardium, has led to uncertainty as to the precise contribution of the second heart field to the heart tube (Moorman et al., 2007). Similarly, the boundaries between the components of the second heart field contributing to the arterial and venous poles of the heart remain to be defined, as do the distal limits of the second heart field in pharyngeal mesoderm. In the absence of definitive first and second heart field markers it may be helpful to consider the above paradigm as a concept to facilitate understanding of heart morphogenesis, rather than as a precise definition of a homogeneous cell population. As pointed out by Moon and colleagues, the concept of the second heart field reflects the relationship between the position of a progenitor cell in splanchnic mesoderm (lateral or medial) and the ultimate location of descendants of such a cell in the definitive heart (Park et al., 2006). Nevertheless, gene expression and lineage analysis support the concept that genetic differences exist across the continuum of splanchnic mesoderm which prefigure the topographical separation of second heart field cells from the body of the heart tube that occurs during looping morphogenesis. Indeed, dorsomedial–ventrolateral embryonic patterning may contribute to partitioning the first and second heart field within splanchnic mesoderm (Kelly and Buckingham, 2002). Cre recombinase lineage tracing analysis has provided the most informative data on the extent to which the second heart field contributes to the definitive heart. Using Cre targeted to the Isl1 locus, Cai et al. (2003) demonstrated a minor contribution (less than 20% of cardiomyocytes) to the left ventricle, particularly in the dorsal inner curvature region, and a two-thirds contribution to right and

149

left atria, in addition to the outflow tract and right v­entricle contribution. Both Isl1-Cre lineage analysis and the retrospective lineage experiments of Meilhac et al. (2004) suggest that a small percentage of cells (10%) in the right ventricle are derived from the first heart field. Additional Cre lineage experiments addressing this issue include four different experiments using regulatory sequences from the Tbx1 locus, encoding the T-box containing transcriptional activator Tbx1 which is expressed in the second heart field and required for normal levels of progenitor cell proliferation (see below). While Tbx1 itself is not expressed in differentiated myocardium, using Cre driven by 6 kb upstream of the Tbx1 transcriptional start site, Brown et al. (2004) found that, like Isl1, Tbx1-expressing cells gave rise to a majority of cardiomyocytes in the heart, with the exception of the ventral region of the left ventricle; intriguingly the Cre-labeled population increased with developmental age, revealing either misregulation of the transgene within the fetal heart or an unprecedented degree of infiltration of the developing heart with cardiomyocytes from a Tbx1-expressing progenitor cell population. In a different transgenic experiment, Maeda et al. (2006) used a Tbx1 enhancer containing a critical forkhead transcription factor binding site to reveal descendants of transgene expressing cells in the outflow tract and right ventricle, in particular in the right ventricular infundibulum and dorsal atrial wall, with only a small contribution to the left ventricle. In contrast, Cre targeted to the endogenous Tbx1 locus revealed labeled cells in the outflow tract with a minor contribution to the right ventricle (Xu et al., 2004). This more restricted distribution may be due to lower levels of Cre accumulation in progenitor cells under transcriptional control of the endogenous locus compared to the situation with multicopy transgenes. However, recent analysis of a second Tbx1 Cre allele has uncovered a more extensive contribution of Tbx1-expressing cells to the outflow tract. Tbx1 descendants comprised the outflow tract with the exception of the most superior region, revealing subcompartments within the second heart field and outflow tract myocardium (Huynh et al., 2007). Mef2c, encoding a MADS-box-containing transcription factor, is expressed throughout the heart by a series of cis-acting modules; one of these regulatory elements has been shown to be active in the anterior heart field and outflow tract and right ventricular myocardium, including the right ventricular facing component of the interventricular septum (Dodou et al., 2004). Using this element to drive Cre expression, Verzi et al. (2005) have demonstrated that the ventricular septum and immediately adjacent left ventricular myocardium is derived from the anterior heart field (Fig. 6A). Thus, the entire interventricular septum forms within a second heart field-derived domain of ventricular myocardium. Intriguingly, despite a second heart field origin, the left and right facing sides of the interventricular septum share the transcriptional programs of the respective ventricular free walls, including right-restricted expression of

150

PART  |  2  Cardiac Precursor Populations and Lineages

(A) RA

OFT

LA

RV

LV

Ao

PM

PM

OFT

RA

LV

RV

PA

RA

LA

RV

Ao LA

PA

RA

RV

LV

LA

LV

(B)

OT RV

A LV

RV

A LV

(C) OFT

(D) ac lb

a v

a

v lb

a

Figure 6  The contribution of the second heart field to the heart. (A) Comparison of expression of a Mef2c enhancer active in the second heart field and right ventricle with a Cre lineage tracing experiment using the same regulatory element at E12 (left panels) and E14.5 (right panels). Note the second heart field contribution to the entire interventricular region and part of the left ventricle in the Cre line (second and fourth panels). Reproduced from Verzi et al. (2005) Dev. Biol. 287, 134–145, with permission of Elsevier. (B) Isl1 is expressed in progenitor cells of the second heart field at the arterial pole (left panel) and venous pole (right panel) in addition to the dorsal pericardial wall (arrow in central panel). Reproduced from Cai et al. (2003) Dev. Cell 5, 877–889, with permission of Elsevier. (C) The Fgf10-nlacZ-enhancer trap transgene is expressed in arterial pole myocardium and progenitor cells in the dorsal pericardial wall (arrowheads) but not at the venous pole of the heart. (D) In contrast, Odd-skipped related 1 (Osr1) is expressed in the venous pole of the heart and in progenitor cells in the dorsal pericardial wall (arrowheads). Reproduced from Wang et al. (2005) Dev. Biol. 288, 582–594, with permission of Elsevier.

the Mef2c anterior heart field enhancer (Verzi et al., 2005; Franco et al., 2006). In support of these observations in the mouse, recent lineage studies in avian embryos suggest that a large part of the right ventricle is derived from the outflow tract and, by extrapolation, the second heart field (Rana et al., 2007).

III.B.  The Contribution of the Second Heart Field at the Venous Pole At the venous pole of the heart the second heart field gives rise to approximately two-thirds of atrial myocytes, including myocardium of the dorsal atrial wall and the interatrial

septum (Cai et al., 2003; Xu et al., 2004). This contribution has recently been further investigated using explant analysis and Isl1 expression studies at the venous pole of the heart (Snarr et al., 2007; Galli et al., 2008; Goddeeris et al., 2008). Regulatory elements specific to the venous pole contribution of the second lineage remain to be defined. However, it is of note that Osr1, encoding a zincfinger transcription factor and the murine homolog of the Drosophila gene Oddskipped, a marker of pericardial cells associated with the dorsal vessel of the fly, is expressed in this population of pharyngeal mesoderm myocardial progenitor cells, in addition to a number of other sites in the developing embryo, including intermediate mesoderm

151

Chapter  |  2.2  The Second Heart Field

(Wang et al., 2005). Osr1 appears to be required in these cells as atrial morphogenesis and formation of the venous valves is perturbed in Osr1 mutant embryos (Wang et al., 2005). Tbx5, also required for normal development of the venous pole of the heart, is expressed in venous pole progenitor cells and is unaltered in Osr1 mutant embryos; the role of Tbx5 in inflow tract development appears to affect first heart field derivatives, including the left ventricle, in addition to the contribution of the second heart field to the venous pole (Bruneau et al., 2001). Expression of Osr1 and Tbx5, in addition to Isl1 and Tbx1, therefore distinguishes the venous pole contribution of the second heart field from the arterial pole progenitor population (Fig. 6B–D). An Isl1-negative population of venous pole myocardial progenitor cells has recently been identified by Christoffels et al. (2006). These progenitor cells are also Nkx2.5negative, but express the transcription factor Tbx18, and are situated more laterally to the Isl1 positive population, giving rise to myocardium of the sinus horns and caval veins (Christoffels et al., 2006). Tbx18 is required for development of this component of inflow tract myocardium, but not for formation of the atrial septum or dorsal atrial wall (Christoffels et al., 2006). Whether these cells represent a distinct field of progenitor cells, or an Isl1negative component of the second lineage, remains to be resolved by future lineage experiments. Distinct subpopulations of venous pole progenitor cells certainly exist, since within the Isl1-positive population of venous pole progenitor cells, Pitx2 plays a critical role in imposing a left atrial program at the venous pole of the heart (Mommersteeg et al., 2007a; Galli et al., 2008). In addition, a particular subpopulation of progenitor cells, distinct from other atrial progenitor cells and dependent on the transcription factors Nkx2.5 and Pitx2, gives rise to myocardium associated with the pulmonary veins (Mommersteeg et al., 2007b). Tbx1 and Pitx2 lineage analysis suggests that different subpopulations of Isl1-positive progenitor cells also contribute to the arterial pole of the heart, potentially giving rise to distinct subdomains of the definitive ventricular outlets (Liu et al., 2002; Maeda et al., 2006; Huynh et al., 2007).

IV.  The contribution of nonmyocardial cell types to development of the arterial pole of the heart and the limits of the second heart field IV.A.  Pharyngeal Endoderm and Ectoderm Although pharyngeal endoderm and ectoderm are not thought to make a cellular contribution to the second heart field, these epithelia are important sources of signaling molecules controlling second heart field development. As noted in Section II.B, second heart field cells remain

apposed to pharyngeal endoderm as the linear heart tube forms. This juxtaposition may be essential to provide traction as cells move towards the elongating poles of the heart tube, and is also likely to provide a source of signals controlling second heart field development. For example, prolonged exposure to endodermal signaling may contribute to the delay in differentiation of the second heart field. A number of genes expressed in the second heart field are also expressed in ventral pharyngeal endoderm, including Isl1 and Nkx2.5, while others, including Fgf8 and Fgf10, are expressed in lateral pharyngeal endoderm. Subsequently, neural crest cells may modulate the effect of endodermal-derived growth factor signaling on development of the second heart field (Hutson et al., 2006). Although spatially distant from the elongating heart tube, pharyngeal ectoderm is an additional source of signaling molecules required for normal pharyngeal development, and is closely apposed to cranial mesoderm upstream of the region of pericardial inflexion where it may influence second heart field progenitor cells prior to their contribution to splanchnic mesoderm (Macatee et al., 2003; Zhang et al., 2006).

IV.B.  Endocardium and Epicardium The lineage experiments using anterior heart field expressed Cre lines revealed that pharyngeal mesoderm gives rise not only to myocardial, but also to endocardial, lineages at the arterial pole of the heart. It has long been established that outflow tract endocardial cells are derived from cranial mesoderm, although a contribution to myocardium was not observed in these chick–quail transplantation experiments (Noden, 1991). Furthermore, epicardium in the outflow tract region has been shown to have a cranial origin, as opposed to the proepicardial origin of epicardial cells covering the majority of the heart (Perez-Pomares et al., 2003). Thus, diverse cardiac cell populations at the arterial pole of the heart originate from mesoderm anterior to the heart tube. The lineage relationships between these different cell types are unknown, although the Cre experiments suggest that endocardial and myocardial progenitor cells express common transcription factors, including Isl1, Tbx1 and Nkx2.5. These different cell types could potentially be derived from common progenitor cells in the second heart field or, alternatively, pharyngeal mesoderm could contain a mixed population of progenitor cells for different cell types which diverged at earlier developmental stages. Multipotential Isl1/Flk1/ Nkx2.5 positive progenitors have been identified within pharyngeal mesoderm of E8.75 embryos which, when clonally propagated, can give rise to myocardial, endothelial and smooth muscle lineages (Moretti et al., 2006). Flk1-positive cells clonally isolated from headfold stage embryos also give rise to these three cardiovascular lineages (Kattman et al., 2006). Intriguingly, it has recently

152

PART  |  2  Cardiac Precursor Populations and Lineages

been shown that Tbx18-expressing epicardial progenitor cells contribute to cardiomyocytes within atrial and v­entricular chambers. Cre experiments reveal that this newly-identified lineage makes a contribution to the developing ventricular septum which is complementary to that of Isl1-derived lineages (Cai et al., 2008).

(A)

IV.C.  Neural Crest Cells

(B)

A nonmesodermal mesenchymal cell population which plays a critical role in pharyngeal and heart development and develops in conjunction with the second heart field is the cardiac neural crest (see also Chapters 3.1 and 7.2). These cells, derived from ectomesenchymal transformation of the dorsal neural tube, are highly migratory and their influx into the pharyngeal region causes sequential swelling of the pharyngeal arches as neural crest cells surround the pharyngeal mesodermal core. At the levels of arches 3–6, a subpopulation of neural crest derived cells continues into the outflow tract of the heart, where they contribute to the aorticopulmonary septum and are essential for outflow tract septation (Kirby and Waldo, 1995). Cre lineage experiments in the mouse have clearly shown that neural crest cells do not contribute to the myocardial walls of the outflow tract (Fig. 7A,B) (Jiang et al., 2000). However, as discussed below in the section addressing congenital heart defects, the development of the cardiac neural crest and second heart field is intimately linked, and cardiac neural crest cells play an indirect role in outflow tract extension by regulating proliferation in the second heart field. Neural crest ablation is thought to impact on second heart field development by altering levels of fibroblast growth factor signaling, leading to second heart field overproliferation and failure to differentiate (Farrell et al., 1999; Hutson et al., 2006). Cardiac neural crest ablation thus leads not only to a failure of outflow tract septation, but also to shortening of the myocardial outflow tract and consequent defects in arterial pole alignment (Yelbuz et al., 2002, 2003; Waldo et al., 2005a). Neural crest ablation also leads to loss of expression of the gene encoding the transcriptional repressor Id2 in second heart field splanchnic mesoderm and outflow tract myocardium (Martinsen et al., 2004). In the zebrafish, unlike the situation in mouse and chick embryos, three recent reports have suggested that neural crest cells may give rise to a significant percentage of cardiomyocytes of the primary heart tube, and furthermore, that incorporation of these ectomesenchymal cells in the heart field is regulated by semaphorin signaling (Li et al., 2003; Sato et al., 2003, 2006). As discussed below, it is unclear whether a similar heart field paradigm to that seen in mouse and chick exists in zebrafish, in which there is a single ventricle with a short myocardial outflow component, and it is possible that other cell sources may have been subverted to the cardiac lineage in zebrafish to maximize cardiomyocyte numbers.

(C)

Figure 7  The borders of the second heart field. (A) Cells of the second heart field are closely apposed to neural crest-derived cells. The Fgf10nlacZ-enhancer trap transgene is expressed in the mesodermal core of the pharyngeal arches (arrow in left panel) and the myocardial wall of the outflow tract (arrow in right panel). (B) In contrast, neural crest-derived mesenchymal cells, visualized using Wnt1-Cre and R26R transgenes, constitute the mass of arch mesenchyme surrounding the mesodermal core and contribute to the aortiocopulmonary septum (arrow in right panel) but not myocardial wall of the outflow tract. Reproduced from Jiang et al. (2000) Development 127, 1607–1616, with permission of the Company of Biologists. (C) At E9.5 proximal core arch mesoderm (arrow in left panel) is contiguous with the second heart field and co-expresses the Fgf10-nlacZ enhancer trap transgene. However, these cells have a branchiomeric skeletal myogenic fate and are positive for transcripts encoding the skeletal muscle determination factor MyoD (arrow in right panel).

In addition to the aorticopulmonary septum, neural crest cells contribute to part of the smooth muscle component of the ascending aorta and pulmonary trunk (Jiang et al., 2000). However, a ring of smooth muscle at the base of the great arteries is non-neural crest in origin (Waldo et al., 2005b). Recent experiments in the chick have demonstrated that these cells are derived from the secondary heart field, and are likely to correspond to the terminal contribution of the second heart field to the arterial pole (Waldo et al., 2005b). Addition of this smooth muscle component appears to be critical to ensure a correct ventriculo– arterial transition, and for normal development of the proximal coronary arteries (Waldo et al., 2005b; Ward et al., 2005). Using the Mef2c-Cre transgene and Isl1-Cre mice, it has been shown that a minority of arterial pole smooth muscle cells in the mouse also have a second heart field

153

Chapter  |  2.2  The Second Heart Field

o­rigin, including certain smooth muscle cells a­ssociated with proximal coronary arteries (Verzi et al., 2005; Moretti et al., 2006; Sun et al., 2007). In addition, a contribution of Tbx1 descendants to the wall of the pulmonary trunk and ascending aorta has been recently documented by Huynh et al. (2007). The existence of a similar domain of smooth muscle giving rise to the entire bulbus arteriosus of the zebrafish heart has raised the possibility that related mechanisms underlie the development of the arterial pole of the heart in fish and amniotes (Grimes et al., 2006). Interestingly this domain has an atypical smooth muscle phenotype characterized by elevated NO synthesis at early stages of both chick and zebrafish heart development (Waldo et al., 2005b; Grimes et al., 2006).

IV.D.  The Edge of the Second Heart Field The anterior heart field is contiguous in pharyngeal mesoderm with a subset of craniofacial skeletal muscle progenitor cells giving rise to branchiomeric muscles involved in jaw opening and closing, facial expression, and laryngeal and pharyngeal function. The junction with these skeletal myogenic progenitor cells appears to define the distal border of the second heart field. Intriguingly, these skeletal muscle progenitor cells, situated in the mesodermal core of the proximal pharyngeal arches, share an overlapping genetic program with cells of the anterior heart field, including expression of Tbx1 and Fgf10 (Fig. 7C) (Kelly et al., 2004; Bothe and Dietrich 2006; Tirosh-Finkel et al., 2006; Grifone and Kelly, 2007; Nathan et al., 2008). Furthermore, lineage analysis using the Cre lines discussed above reveals that these skeletal muscle progenitor cells have previously expressed Isl1, Nkx2.5 and the Mef2c anterior heart field enhancer (Stanley et al., 2002; Cai et al., 2003; Verzi et al., 2005; Nathan et al., 2008). Thus, prior to activation of skeletal muscle determination genes of the MyoD family of basic helix-loop-helix transcription factors, these skeletal muscle progenitors share a common transcriptional program with cardiac progenitor cells. Furthermore, this genetic program diverges from that of all other skeletal muscle progenitor cells in the embryo. This difference in branchiomeric versus nonbranchiomeric myogenic specification is apparent from the phenotype of Tbx1 mutant embryos that fail to robustly activate MyoD and Myf5 in pharyngeal mesoderm, leading to the absence of branchiomeric muscles or the presence of hypoplastic asymmetric craniofacial muscles; myogenesis elsewhere in the embryo proceeds normally (Kelly et al., 2004). Understanding the genetic hierarchies regulating the divergent cardiac and skeletal myogenic fates of the adjacent progenitor cell populations in pharyngeal mesoderm is a challenge for future research. However, recent studies suggest that the divergent fates within the continuum of pharyngeal mesoderm are regulated by Bmp signaling which

plays a role in recruitment of second heart field cells to the growing arterial pole of the heart, promoting cardiac and inhibiting skeletal myogenesis (Waldo et al., 2001; Tirosh-Finkel et al., 2006). Wnt signals from the midline are also likely to block cardiomyogenesis in more medially-positioned skeletal muscle progenitor cells, and may also regulate the delayed differentiation of the second heart field which is positioned medially with respect to the first heart field (see section on Wnt signaling below) (Tzahor et al., 2003). At early developmental stages, skeletal muscle progenitor cells in cranial mesoderm express multiple transcriptional repressors of the myogenic program, including genes encoding the basic helix-loop-helix factors Capsulin, MyoR and Twist, raising the possibility that differentiation of these muscle cells may be delayed to allow recruitment of cells to the second heart field for construction of the rapidly-elongating heart tube (Bothe and Dietrich, 2006).

V.  Transcriptional networks controlling the second heart field V.A.  Transcriptional Circuits Driving Second Heart Field Differentiation Transcriptional differences between left and right ventricles had been observed prior to identification of the second heart field; in particular a number of cis-acting modules, often dissected from the regulatory regions of genes expressed throughout the heart, were found to drive reporter gene expression in a single ventricle of transgenic mice (Firulli and Olson, 1997; Kelly et al., 1999). These experiments provided the first indication that different transcriptional programs operate in left and right ventricles, presaging discovery of the second heart field. Since its discovery, extensive inroads have been made into understanding the transcriptional circuitry regulating second heart field development, in particular that of the anterior heart field component and its myocardial derivatives. Isl1 is essential for anterior heart field deployment, and is required for the regulation of multiple signaling molecules of the BMP and FGF families in the distal heart tube and pharyngeal region (Cai et al., 2003). Isl1 has been shown to directly regulate the anterior heart field specific enhancer of the Mef2c gene, encoding a MADS-box transcription factor essential for development of the outlet component of the heart (Lin et al., 1997; Dodou et al., 2004; Chapter 4.3). Isl1 acts together with cardiac Gata transcription factors to drive expression of this enhancer in pharyngeal mesoderm of the anterior heart field. Isl1 and Gata sites are also required for activation of an Nkx2.5 enhancer in the second heart field and arterial pole of the heart; in vitro experiments suggest that the T-box transcription factor Tbx20 synergizes with these transcription factors at both the Mef2c and Nkx2.5 enhancers, to

154

PART  |  2  Cardiac Precursor Populations and Lineages

confer maximal transcriptional activity (Takeuchi et al., 2005). A second enhancer at the Mef2c locus has also been shown to drive transgene expression in the arterial pole of the heart and adjacent anterior heart field (Von Both et al., 2004). This enhancer lies within 3 kb of that identified by Dodou et al. (2004), and the two regulatory elements flank an internal Mef2c promoter (Kelly, 2005). The second Mef2c enhancer is regulated by a forkhead transcription factor, Foxh1, which is required for anterior heart field deployment; furthermore, Foxh1 activates this enhancer together with Nkx2.5 (Von Both et al., 2004). Thus, a combination of second lineage transcription factors (Isl1 and Foxh1) and broadly-expressed cardiac transcription factors (including Gata factors and Nkx2.5) drive gene expression in the anterior heart field (Fig. 8 and Chapter 9.1). Of the cardiac Gata factors, the expression profiles of Gata5 and Gata6 suggest that these genes are more likely to contribute to transcriptional activity in the second heart field than Gata4 (Zeisberg et al., 2005). Although part of the anterior heart field-derived myocardium is added to the heart tube in the absence of Mef2c, this gene is required for normal morphogenesis of the outflow tract and right ventricle, in addition to the components of the first heart field, including the left ventricle. Mef2c mutant embryos appear to misallocate first heart field progenitor cells from the left ventricle to the inflow tract (Lin et al., 1997; Verzi et al., 2005; Vong et al., 2006). Bop, encoding a transcriptional repressor and potential histone methyltransferase, is a direct transcriptional target of Mef2c, and is also required for development of the outflow tract and right ventricle (Gottlieb et al., 2002; Phan et al., 2005). The basic helix-loop-helix transcription factor Hand2 (dHand) is, in turn, regulated by BOP and is essential for expansion of right ventricular progenitor cells and for outflow tract morphogenesis; Hand2 expression is also regulated by Gata4

β-catenin

Nkx2.5

Foxh1

Isl1

Mef2c Gata

Bop

in the Nkx2.5 expression domain (Srivastava et al., 1997; Gottlieb et al., 2002; Zeisberg et al., 2005; Chapter 9.1). A cascade of transcription factors has therefore been identified which regulates anterior heart field deployment and the development of anterior heart field derived components of the heart (Fig. 8). Indeed, the ultimate role of this cascade appears to be the control of differentiation of second heart field-derived cardiomyocytes after their addition to the heart tube. The downstream targets of this transcriptional cascade mediating outflow tract and right ventricular differentiation at the cellular level remain to be identified. Candidate genes include Cspg2, encoding the core protein for the chondroitin sulfate proteoglycan versican; mice carrying an insertional mutation at this locus exhibit a hypomorphic right ventricle and outflow tract (Mjaatvedt et al., 1998). Defective differentiation of right ventricular myocardium would be expected to perturb further recruitment of second heart field cells by, for example, interfering with Bmp ligand expression. This illustrates the important point that genes regulating differentiation of second heart field-derived cardiomyocytes after incorporation into the heart tube, or even genes regulating myocyte differentiation within the linear heart tube, may indirectly impact on second heart field deployment. Serial Cre analysis and detailed knowledge of expression profiles are therefore required to delineate the critical time and site of gene function for regulators of second heart field deployment.

V.B.  The Role of Pitx2c in the Second Heart Field Rightward looping of the ventricular segment takes place as second heart field cells are added to the early heart tube. Given the medial origin of the second heart field compared

Shh

Foxc1, Foxc2

RA

Tbx1

Pitx2c

Foxa2

Wnt11

Fgf8 Fgf10 BMP

Hand2 Differentiation

Proliferation Survival Migration

Laterality Polarized cell movement

Figure 8  Transcription factors and signaling pathways regulating second heart field development. Tbx1 regulates second heart field deployment and laterality by potentially activating signaling molecules of the FGF and BMP families, in addition to Pitx2c expression in the left second heart field. Tbx1 expression is regulated by Shh signaling via Forkhead-containing transcription factors. Isl1 regulates growth factor gene expression and Mef2c transcription in the second heart field, thus controlling both deployment and differentiation of this cell population. Mef2c lies in a central position within a cascade of transcription factors that regulate differentiation of outflow tract myocardium as second heart field-derived cells contribute to the elongating heart tube.

155

Chapter  |  2.2  The Second Heart Field

to the lateral origin of progenitors of the linear heart tube, and the midline origin of left–right laterality signaling in the early embryo, it is possible that the second heart field carries laterality information to the looping heart and actively drives rightward looping. Indeed, genes encoding the TGF family signaling molecule Nodal and the homeobox-containing transcription factor Pitx2c are expressed in left, but not right, splanchnic mesoderm dorsal to the early heart tube, including second heart field cells at both the venous and arterial poles of the heart tube (Fig. 9A) (Meno et al., 2001; Liu et al., 2002; Chapters 4.2 and 4.3). Nodal is directly required for rightward looping, whereas Pitx2c, downstream of Nodal, regulates atrial left–right identity and rotation of the arterial pole of the heart necessary for correct alignment of the aorta and pulmonary trunk with the left and right ventricles (Liu et al., 2002). The Pitx2c-independent mechanism by which Nodal controls rightward looping is unknown, but is likely to be mediated by the second heart field. Indeed, experiments in chick have revealed that asymmetric expression of genes encoding extracellular matrix components, including matrix metalloproteinases in the dorsal pericardial wall and dorsal mesocardium, in addition to torsion at the poles of the elongating heart tube, play critical roles in driving rightward looping (Linask et al., 2003, 2005; Manner, 2004). A series of Cre recombinase-mediated conditional mutations have recently shown that Pitx2c is directly required in myocardial progenitor cells of the second heart field for normal outflow tract development and atrial identity (Ai et al., 2006). The failure of imposing left identity on the left side of the anterior heart field results in defective proliferation of a subpopulation of proximal outflow tract myocardial cells, leading to outflow tract defects including double-outlet right ventricle and transposition of the great arteries (Ai et al., 2006). The myocardial wall of the outflow tract undergoes a counterclockwise rotation during great artery development, as shown originally by labeling experiments in the chick embryo using autoradiographic tattoos and computer reconstructions of a temporal series of histological sections of human embryos (Lomonico et al., 1986; Thompson et al., 1987). Defects in outflow tract rotation in Pitx2c mutant embryos have been observed using a transgene expressed in the dorsal wall of the outflow tract (Bajolle et al., 2006). This transgene is later expressed in myocardium around the base of the pulmonary trunk consistent with a counterclockwise rotation of the myocardial wall during outflow tract septation, further confirmed by DiI labeling experiments in embryo culture and analysis of a second transgene expressed in a complementary profile in subaortic myocardium (Fig. 9B–D) (Bajolle et al., 2006, 2008). Rotation of the avian outflow tract has also been observed by Ward et al. (2005), who demonstrated that fluorescent labeling of cells in the right second heart field of chick embryos at stage HH14 led to labeled cells being observed in the left side

of the ventricular outlet and base of the pulmonary trunk 48 hours later. The difference between these two datasets is that the mouse data suggests that the outflow tract elongates prior to rotation, whereas the chick data suggest that second heart field cells spiral into the elongating outflow tract (Ward et al., 2005; Bajolle et al., 2006). Whether or not this difference is one of experimental approach remains to be determined.

V.C.  Tbx1 Regulation of the Second Heart Field Another transcriptional circuit required for second heart field deployment lies downstream of the forkhead transcription factors Foxc1 and Foxc2 (Fig. 8) (Seo and Kume, 2006). Foxc1;Foxc2 double mutants fail to form a right ventricle and outflow tract, and have a proliferation deficit in the anterior heart field associated with downregulation of the T-box containing transcription factor Tbx1 (see Chapter 9.4) and the Fgf ligand encoding genes Fgf8 and Fgf10 (Seo and Kume, 2006). Defects are also observed in the venous pole of Foxc1;Foxc2 double mutants, suggesting a role for these transcription factors throughout the second heart field; analysis of embryos carrying single Foxc alleles suggests that Foxc2 is the major paralog required for second heart field development (Seo and Kume, 2006). Two Forkhead transcription factor target sites have been identified upstream of Tbx1 which bind Foxc1 and Foxc2, and together are required for expression in pharyngeal mesoderm (Yamagishi et al., 2003; Maeda et al., 2005). Hu et al. (2004) have shown that a third forkhead transcription factor encoding gene Foxa2 (originally HNF3b) is expressed in core pharyngeal mesoderm and can both bind and activate transcription through the distal Tbx1 Fox site. Foxa2 expression in this region is lost in Tbx1 mutant embryos, revealing the existence of a positive feedback loop controlling Tbx1 expression in the anterior heart field (Fig. 8) (Hu et al., 2004). Tbx1 expression in the second heart field is also regulated by the zinc-finger-containing transcription factor Blimp1 (Robertson et al., 2007). TBX1 is the major candidate gene for del22q11.2 or DiGeorge syndrome in man, associated with craniofacial and cardiovascular defects and a multigene heterozygous deletion involving approximately 30 genes on chromosome 22 (Lindsay, 2001; Chapter 7.2). Tbx1 haploinsufficient embryos display aortic arch patterning defects found in del22q11.2 patients, while homozygous mutant mice recapitulate the most severe defects observed in del22q11.2 patients and die at term with a common arterial trunk and ventricular septal defect (Jerome and Papaioannou, 2001; Lindsay et al., 2001; Merscher et al., 2001; Vitelli et al., 2002a). The defects of Tbx1 homozygous mutant embryos originate in severe caudal pharyngeal hypoplasia at midgestation. The distal outflow tract of the embryonic heart is shorter and

156

PART  |  2  Cardiac Precursor Populations and Lineages

(A)

(B)

(C)

(D)

Figure 9  The second heart field and cardiac laterality. (A) The second heart field is lateralized. At E8.5 the Fgf10-lacZ-enhancer trap transgene is expressed bilaterally in the second heart field, whereas a Nodal lacZ transgene is expressed in cells of the second heart field in the left (L) but not right (R) dorsal pericardial wall. Right hand panel reproduced from Meno et al., Dev. Cell 1, 127–138, with permission of Elsevier. (B) The myocardial wall of the outflow tract undergoes a counterclockwise rotation during great artery development. Cartoon representing expression of a b-galactosidase encoding transgene in the inferior wall of the outflow tract at E10.5, subsequently in the left facing OFT wall (E12.5), and finally in ventral myocardium at the base of the pulmonary trunk (E14.5). (C) Transverse sections through the outflow tracts of transgenic hearts supporting the cartoon in (B). (D) Outflow tract rotation is defective in Pitx2c mutant hearts; transgene expression remains in myocardium around the base of the pulmonary trunk, now in a more dorsal position, consistent with the transposition of the great arteries observed in Pitx2c mutant hearts. (B–D) reproduced with permission from Bajolle et al. (2006) Circ. Res. 98, 421–428.

narrower than that of wild-type embryos, and an elegant series of experiments by Baldini and colleagues have shown that the outflow tract defects result directly from hypoplasia of the second heart field due to reduced proliferation in pharyngeal mesoderm (Fig. 10). Using a conditional

Tbx1 mutant allele and a series of Cre recombinase lines, Baldini and colleagues have shown that the requirement for Tbx1 in the anterior heart field is a direct, though cell non-autonomous, requirement in pharyngeal mesoderm, that mesodermal Tbx1 is sufficient to restore normal heart

157

Chapter  |  2.2  The Second Heart Field

(A)

(B)

(C) RSA

RCA

RCA

LCA LSA

LCA LSA

RSA P

Ao

PTA pa pa

(D)

Ao PTA VSD

little Tbx1 results in apparently convergent phenotypes, is not yet known. Development of an allelic series at Tbx1 has allowed high-resolution variation of Tbx1 expression levels. This analysis has identified critical threshold requirements for Tbx1 that differ in mouse and man, providing a mouse model which more closely approximates TBX1 haploinsufficiency in DiGeorge syndrome (Zhang and Baldini, 2008). Second, the Cre experiments discussed above suggest that Tbx1 is expressed more broadly in the second heart field than the phenotype of Tbx1 mutant embryos would suggest. The possibility that other members of the Tbx gene family play overlapping roles with Tbx1 during early stages of heart development remains to be explored, although the recent implication of another member of the Tbx1 subfamily of Tbx genes, Tbx20, in anterior heart field deployment and outflow tract development suggest that this may be the case (Takeuchi et al., 2005). In support of such a role, Tbx20 is required to downregulate expression of the second heart field gene Isl1 on differentiation (Cai et al., 2005). Finally, Tbx1 has been shown to bind to an enhancer of Pitx2c required to maintain asymmetric Pitx2c expression in left lateral mesoderm; Nkx2.5 also binds to this enhancer and directly interacts with Tbx1 (Nowotschin et al., 2006). Furthermore, compound Pitx2 Tbx1 heterozygotes die perinatally, with a range of cardiac defects typical of Pitx2c mutant embryos, suggesting that normal levels of Tbx1 are required for Pitx2c function in the second heart field (Nowotschin et al., 2006).

RV LV

Figure 10  The role of Tbx1 in the second heart field. (A) At E10.5 the distal outflow tract of Tbx1 mutant hearts (arrow in right panel) is shorter and narrower than in control hearts (left). (B) The second heart field, visualized using the Fgf10-nlacZ-enhancer trap transgene, is hypoplastic and lacks bilateral streams of cells in the dorsal pericardial wall (asterisks). Reproduced with permission from Kelly and Papaioannou, Dev. Dyn. 236, 821–828. (C, D) Tbx1 is required in pharyngeal mesoderm for outflow tract development. The hearts on the left have a Tbx1 null phenotype with a common ventricular outlet (PTA in (C)) and ventricular septal defect (VSD in (D)). Restoration of Tbx1 exclusively in pharyngeal mesoderm using Cre-lox technology rescues the cardiac phenotype. Reproduced from Zhang et al. (2006) Development 133, 3587– 3595, with permission of the Company of Biologists.

development in a Tbx1 mutant background and that the temporal requirement for Tbx1 in heart development is between days 8.5 and 9.5 of development (Xu et al., 2004, 2005; Zhang et al., 2006). There are a number of intriguing questions concerning Tbx1 requirements during heart development. First, the level of Tbx1 required for normal heart development is critical, as transgenic mice carrying three copies of Tbx1 or patients with a duplication of the 22q11.2 region also have outflow tract defects (Liao et al., 2004). The question of how Tbx1 levels are titrated, and how too much or too

VI.  Signaling networks controlling the second heart field VI.A.  Fibroblast Growth Factor Signaling Tbx1 has been proposed to regulate anterior field deployment by controlling the expression of FGF ligand encoding genes (Vitelli et al., 2002b). The expression of Fgf8 and Fgf10 is downregulated in the pharyngeal region of Tbx1 homozygous mutant embryos, and Tbx1 responsive T-box target sites have been identified upstream of both genes, although the in vivo significance of these sites has yet to be evaluated (Hu et al., 2004; Xu et al., 2004). As noted above, FGF ligand expression is also reduced in Isl1 mutant embryos (Cai et al., 2003). The epistatic relationship of Isl1 and Tbx1 remains to be determined, and it is unclear how the activity of these transcription factors converges at the level of FGF ligand expression. How does FGF signaling impact on second heart field development? Fgf10 and Fgf8 have been most extensively studied in this context. Fgf10 mutant mice die at term, due to the absolute requirement for Fgf10 in lung morphogenesis (Min et al., 1998; Sekine et al., 1999). Fgf10 mutant mice also have limb aplasia and loss or hypoplasia of

158

multiple embryonic structures (Min et al., 1998; Sekine et al., 1999; Ohuchi et al., 2000). However, despite expression of Fgf10 and an Fgf10 enhancer trap transgene in the second heart field, the initial stages of heart development, including outflow tract elongation, appear to occur normally in Fgf10 mutant embryos (Marguerie et al., 2006; Chapter 3.1). Neither ventricular septal nor outflow tract defects are observed in Fgf10 mutant embryos, although Fgf10 plays a later role in ventricular growth and positioning of the heart in the thoracic cavity (Marguerie et al., 2006). In contrast, mice lacking the major Fgf10 receptor, Fgfr2-IIIb, display ventricular septal defects with overriding aorta or double-outlet right ventricle, suggesting that other FGF ligands signaling through this receptor are required to achieve maximal outflow tract length and correct alignment or, alternatively, are required at later stages of outflow tract septation (Marguerie et al., 2006). In contrast to Fgf10, Fgf8 is now recognized to be a central driver of anterior heart field deployment. Fgf8 is critically required for the movement of mesodermal cells out of the primitive streak during gastrulation, leading to an early post-implantation lethal phenotype (Sun et al., 1996). Mice which carry a hypomorphic over null allelic combination at the Fgf8 locus display aortic artery patterning defects and a shortened outflow tract, leading to ventricular septal defects and outflow tract alignment anomalies at later developmental stages (Abu-Issa et al., 2002; Frank et al., 2002). Fgf8 is expressed in pharyngeal mesoderm of the second heart field and two extensive series of experiments using conditional Fgf8 alleles and a range of Cre recombinase lines have recently shown that Fgf8 is required in the second heart field for normal anterior heart field deployment (Fig. 11A) (Ilagan et al., 2006; Park et al., 2006). Using Mesp1 and Isl1 Cre lines, Park et al. (2006) demonstrated that early loss of Fgf8 in the second heart field leads to a failure of heart tube extension associated with a reduction of Isl1 expression, albeit with variable severity. The role of Fgf8 in the venous pole component of the second heart field remains unclear, as does a potential overlap in function with Fgf10. The demonstration that murine Fgf8 drives second heart field deployment in pharyngeal mesoderm is consistent with the phenotype of the zebrafish Fgf8 mutant acere­bellar which displays a severe truncation of the ventricle, but not atrium, corresponding to the arterial pole of the fish heart (Fig. 11B–D) (Reifers et al., 2000; Chapter 1.4). Zebrafish Fgf8 is expressed in the entire ventricular segment, and at early stages of heart development accumulates in a region medial to the lateral mesodermal expression domain of Nkx2.5 (Fig. 11D) (Reifers et al., 2000). By analogy with observations in the mouse, these findings suggest that Fgf8 expression may identify the second heart field contribution to the zebrafish heart, and that this may include the entire outlet segment. Alternatively, as discussed above, extensive contribution of pharyngeal mesoderm to the heart may be a

PART  |  2  Cardiac Precursor Populations and Lineages

(A)

(B)

(C)

(D)

Dorsal view

Cross-section

Dorsal view

NT MHB

7 som

Cross-section

NT MHB

11 som

Figure 11  Fibroblast growth factor 8 drives second heart field deployment. (A) An Fgf8-lacZ allele reveals that Fgf8 is expressed in the second heart field and outflow tract and right ventricular myocardium in the mouse. Ablation of Fgf8 in the Nkx2.5 expression domain (right panel) blocks second heart field deployment, resulting in severe truncation of the arterial pole of the heart. Reproduced from Ilagan et al. (2006) Development 133, 2435–2445 with permission of the Company of Biologists. (B, C) Zebrafish acerebellar mutants lack Fgf8 and display a hypoplastic heart with a truncated ventricular (outlet) segment (red). (D) At early stages of heart development Fgf8 expressing cells (orange) lie medial to cells expressing cardiac markers such as Nkx2.5 (red) and Gata4 (blue), similar to the medial expression domain of Fgf10 and Isl1 in mice (see Fig. 5). (B–D) Reproduced from Reifers et al. (2006) Development 127, 125–135, with permission of the Company of Biologists.

feature of multichambered amniote heart development and a contribution in the fish, if any, may be limited to the distal ventricular lip or smooth muscle walled truncus (Grimes et al., 2006). Exactly how FGF ligands regulate second heart field deployment is unknown. FGF signaling is required for induction of cardiomyocyte differentiation at early stages of chick heart development, and it is unclear whether FGF signaling in the second heart field is playing an inductive role or a role in driving progenitor cell proliferation, as suggested by the Fgf8 conditional and Tbx1 mutant phenotypes (Alsan and Schultheiss, 2004; Xu et al., 2005; Park et al., 2006). A third possibility is suggested by the

Chapter  |  2.2  The Second Heart Field

phenotype of the Drosophila heartless mutant. Heartless encodes an FGF receptor required for the directed migration of newly-ingressed mesodermal cells to the dorsal position where Dpp (equivalent to vertebrate BMP) signaling activates the cardiomyogenic program (Beiman et al., 1996; Gisselbrecht et al., 1996; Shishido et al., 1997). There is an apparent analogy with the movement of second heart field cells towards the poles of the heart tube, where Bmp signaling initiates cardiomyogenesis, raising the possibility that FGF signaling may play a role in cell movement during second heart field deployment, as well as roles in differentiation and proliferation. Conditional analysis of FGF receptors in the second heart field and surrounding cell types, together with a dissection of downstream components of the FGF signaling pathway during heart tube extension will provide further insights into the mechanisms by which FGF signaling controls second heart field deployment. Decreased expression of both Isl1 and Wnt11 has been observed in tissue-specific Fgf8 mutants (Park et al., 2006), suggesting that these two genes may be critical effectors downstream of FGF signaling in the second heart field. Intriguingly, ablation of the cardiac neural crest results in overproliferation of second heart field cells, as opposed to the underproliferation observed in Fgf8 conditional and Tbx1 mutant mouse embryos. This is thought to result from overexposure of second heart field cells to endodermal Fgf8, normally absorbed by intervening neural crest cells prior to their entry into the heart (Farrell et al., 1999; Chapters 7.1 and 7.2). Neural crest ablation results in overproliferation of second heart field cells in the pharyngeal region, leading to a reduction in outflow tract elongation and consequent alignment defects, including overriding aorta and doubleoutlet right ventricle, associated with impaired ventricular function (Yebluz et al., 2002, 2003; Waldo et al., 2005a). Thus, both over- and underproliferation of second heart field progenitor cells impact negatively on outflow tract length. These experiments also suggest that precise levels of Fgf8 signaling are critical for second heart field development. Blocking FGF receptor activity pharmacologically can rescue the alignment defects of cardiac neural crest ablated embryos, whereas blocking receptor activity without neural crest ablation leads to outflow tract shortening (Hutson et al., 2006). However, reducing normal levels of FGF signaling without neural crest ablation leads to a different spectrum of outflow tract defects than those observed after neural crest ablation. These include tetralogy of Fallot and pulmonary stenosis and atresia, defects which are also observed after laser ablation of the right side of the second heart field (Ward et al., 2005; Hutson et al., 2006). The link between Tbx1 and Fgf ligand function in second heart field deployment has recently been questioned. While Tbx1 and Fgf8 genetically interact to control aortic arch artery development, a knockin of Fgf8 at the Tbx1 locus does not restore normal outflow tract development in Tbx1-null

159

embryos, as might be expected if reduction of Fgf8 expression were the major consequence of loss of Tbx1 (Vitelli et al., 2002b, 2006). Similarly, in contrast to the interaction documented between Fgf8 and Tbx1 in aortic arch artery development, no genetic interaction has been found in Fgf10 Tbx1 compound heterozygotes (Vitelli et al., 2002; Aggarwal et al., 2006; Kelly and Papaioannou, 2007). While other FGF ligands may also intervene in this process, in vivo evidence for direct regulation of Fgf8 and Fgf10 by Tbx1 is lacking, as discussed above (Hu et al., 2004; Xu et al., 2004). Together, these results suggest that a critical role of Fgf8 in driving second heart field development may not be directly dependent on Tbx1.

VI.B.  Bone Morphogenetic Protein Signaling BMP signaling is also critically required for outflow tract elongation. As discussed above, the initial experiments of Waldo et al. implicated Bmp2 as a candidate molecule involved in recruiting cells from the second heart field into the growing heart tube. BMP ligand expression is downregulated in Isl1 mutant embryos; in contrast, overexpression of BMP can drive additional cells into a cardiomyogenic phenotype (Cai et al., 2005; Tirosh-Finkel et al., 2006). Outflow tract elongation is abnormal in mouse embryos lacking both Bmp4 and Bmp7, suggesting that ligand redundancy may mask a potentially essential role for BMP signaling in recruitment of second heart field cells (Liu et al., 2004). Loss of Bmp4 expression in outflow tract myocardium is observed in Smarcd3-mutant embryos, and is associated with impaired anterior heart field deployment resulting in a hypoplastic right ventricle and shortened outflow tract (Lickert et al., 2004). Smarcd3 encodes Baf60c, a component of a chromatin remodeling complex, and is also required for Pitx2c expression in the left anter­ ior heart field and robust levels of Tbx20 expression in outflow tract myocardium (Lickert et al., 2004). Removal of the BMP receptor Bmpr1a in Isl1-expressing cells using a conditional receptor allele and an Isl1 Cre demonstrated that BMP signaling is required for normal right ventricular and outflow tract morphology at midgestation, leading to common ventricular outlet and ventricular septal defects (Yang et al., 2006). Deletion of the same receptor in neural crest cells also leads to a shortened outflow tract, suggesting an additional indirect role of BMP signaling in second heart field deployment (Stottmann et al., 2004). Using a Mesp1-Cre line active in the precursors of both the first and second heart fields has recently revealed an earlier role for this receptor in specification of the first heart field, but not the second heart field (Klaus et al., 2007). Consistent with the model that BMP signals recruit second heart field cells to a myocardial fate, it has recently been shown that overexpression of Bmp2 in Nkx2.5 mutants leads to an expanded domain of cardiac progenitor cell markers, including second heart field markers;

160

in addition abnormal expression of these progenitor cell genes is observed in the mutant heart tube (Prall et al., 2007; Chapter 9.1). Nkx2.5 mutant embryos therefore display cardiac overspecification, which, followed rapidly by proliferative failure, leads to a block in heart development soon after looping initiates and a failure of second heart field deployment; counterintuitively this key cardiac transcription factor may therefore act to brake cardiac specification during normal development (Prall et al., 2007).

VI.C.  Hedgehog Signaling In addition to FGF and BMP signaling, the hedgehog signaling pathway has been implicated in second heart field development. Sonic hedgehog, acting through Fox transcription factors, positively regulates Tbx1 expression (Garg et al., 2001). A shortened outflow tract at E8.75 is one of the first morphological manifestations of loss of Shh (Washington Smoak et al., 2005). At later developmental stages this results in a cranially positioned right ventricle and common ventricular outlet; Shh signaling is also required for normal distribution of the cardiac neural crest in the outflow tract (Washington Smoak et al., 2005). Both Shh- and Tbx1-mutant embryos present hypoplasia of the caudal pharynx and Shh-mutant hearts appear to represent an extreme case of pulmonary atresia, a phenotypic assessment which may also apply to the common ventricular outlet in Tbx1-mutant hearts. Embryos lacking the essential hedgehog receptor Smoothened have a more severe cardiac phenotype, and fail to develop beyond the linear heart tube stage, suggesting that hedgehog signaling is required in myocardial progenitor cells and that ligands other than Shh are implicated; indeed Shh;Ihh double mutant embryos display an indistinguishable phenotype to that of Smoothened mutant embryos (Zhang et al., 2001). Recently, deletion of Smoothened in the Isl1 expression domain has been shown to result in outflow tract shortening and common arterial trunk associated with ventricular septal defects (Lin et al., 2006), phenocopying the Shh mutant phenotype. Lin et al. (2006) demonstrated elevated cell death and downregulation of Tbx1 in pharyngeal mesoderm and Neuropilin2 in the outflow tract of conditional mutant embryos. Endodermal Shh signaling impacts on both the second heart field and neural crest cells, and separation of these phenotypes using Cre recombinase-mediated Smoothened deletion has allowed a fine dissection of the role of hedgehog signaling in each cell type (Goddeeris et al., 2007). Reception of Shh signaling is required in neural crest cells for survival, in second heart field cells for later outflow tract septation, and in the endoderm itself to regulate a secondary signal controlling second heart field addition to the elongating heart tube (Goddeeris et al., 2007). Outflow tract shortening

PART  |  2  Cardiac Precursor Populations and Lineages

due to over- or under-proliferation can produce divergent phenotypes and these experiments also show that different etiologies can lead to apparently convergent phenotypes, in this case a common ventricular outlet. Hedgehog signaling has recently been shown to be required for a second heart field-derived cellular contribution to the dorsal mesenchymal protusion at the venous pole of the heart. Addition of these cells is critical for atrioventricular septation, revealing a broad requirement for second heart field deployment in cardiac septation (Goddeeris et al., 2008).

VI.D.  Retinoic Acid Signaling Retinoic acid signaling in the posterior region of the heart field is a potential regulator of second heart field development. Raldh2 encodes a critical enzymatic component of the retinoic acid synthesis pathway, and is expressed in the caudal pharynx. Raldh2 mutants have severe heart defects and fail to extend the heart tube (Neidereither et al., 2001; Chapter 3.3). Loss of Raldh2 is associated with an expansion of the posterior expression domain of second heart field markers, suggesting that retinoic acid signaling limits the extent of the second heart field (Ryckebusch et al., 2008). Altered retinoic acid levels are observed in Tbx1-mutant embryos and the Raldh2 expression domain is shifted anterioraly, possibly as a result of caudal pharyngeal hypoplasia (Ivins et al., 2005; Guris et al., 2006). In parallel, retinoic acid-catabolizing enzymes of the Cyp26a family are downregulated in Tbx1-mutant embryos (Roberts et al., 2006). A complex feedback loop exists between Tbx1 and retinoic acid, since retinoic acid treatment has been shown to downregulate Tbx1 expression in vitro and in vivo (Roberts et al., 2005). Altering retinoic acid levels in chick embryos supports a role for this morphogen in promoting inflow tract development, suggesting that retinoic acid may be required for development of the venous pole component of the second heart field (Hochgreb et al., 2003).

VI.E.  Notch Signaling Cells of the second heart field are distinguished from cells of the linear heart tube by their delayed differentiation. Understanding the molecular regulation of this differentiation delay, and how differentiation is progressively activated, is an important focus of current research into this progenitor cell population. Insight into this process comes from analysis of the notch signaling pathway during Xenopus heart development (see Chapter 1.3). While the notch receptor is present throughout the heart tube, the Notch ligand serrate is expressed only in the dorsal pericardial wall and dorsal mesocardium (Rones et al., 2000). Activated Notch is thus restricted to the region of the dorsal mesocardium where it represses cardiomyocyte

161

Chapter  |  2.2  The Second Heart Field

differentiation, downstream of Nkx2.5 activation (Rones et al., 2000). Recent evidence that Xenopus Isl1 is expressed in the dorsal pericardial wall and dorsal mesocardium supports the conclusion that this is an analogous cell population to the amniote second heart field (Brade et al., 2007); whether or not Notch signaling plays a similar role in the regulation of second heart field development awaits further investigation.

VI.F.  Wnt Signaling and Intersections with TGF Signaling During Outflow Tract Morphogenesis B-catenin is the transcriptional effector of the canonical Wnt pathway. A series of recent studies have revealed a critical requirement for -catenin in second heart field deployment (Ai et al., 2007; Cohen et al., 2007; Klaus et al., 2007; Kwon et al., 2007; Lin et al., 2007; Qyang et al., 2007; reviewed in Tzahor, 2007). Ablation of -catenin with second heart field Cre lines has revealed a requirement for -catenin within the second heart field for multiple aspects of cardiac morphogenesis (Ai et al., 2007; Cohen et al., 2007; Lin et al., 2007; Chapter 4.3). Isl1-Cre; -catenin mutants die at approximately E13, exhibiting a common ventricular outlet, aortic arch artery defects, hypomorphic right ventricle, thin-walled myocardium, hypomorphic atrioventricular cushions and atrial septal defects (Lin et al., 2007). Hypomorphic mandibles are also observed. Hypomorphic atrioventricular cushions are consistent with a potential contribution of the second heart field to atrio­ ventricular cushion mesenchyme, and with previous studies demonstrating a requirement for -catenin within endothelial cell lineages for atrioventricular cushion morphogenesis (Liebner et al., 2004). Activation of -catenin signaling in the second heart field using a conditional stabilized form of -catenin leads to hyperproliferation, impaired differentiation of second heart field cells and outflow tract defects (Ai et al., 2007; Cohen et al., 2007; Qyang et al., 2007). Further analysis of Isl1-Cre;-catenin mutants revealed that Isl1 expression is dependent on -catenin signaling, with -catenin directly binding to the Isl1 promoter in vivo (Lin et al., 2007). Fgf10 is also a direct target of activated -catenin signaling (Cohen et al., 2007). Consistent with regulation of Isl1, -catenin is required for proliferation and survival of the second heart field. Pitx2 is also downregulated in Isl1-Cre;-catenin mutants, consistent with direct targeting of Pitx2 by -catenin (Kioussi et al., 2002; Zhou et al., 2007). Tbx1 is also upstream of Pitx2 in the second heart field (Nowotschin et al., 2006), suggesting cooperative regulation by Wnt signaling and Tbx1. In Isl1-derived lineages -catenin signaling appears to be upstream of a number of other genes required for outflow tract or other aspects of cardiac morphogenesis, as

expression of Tbx2, Tbx3, Shh and Wnt11 is significantly decreased in Isl1-Cre;-catenin mutants (Liu et al., 2002; Harrelson et al., 2004; Yang et al., 2006; Lin et al., 2007; Zhou et al., 2007). Identification of the key Wnt ligands controlling second heart field development and their sites of expression will be an important next step. The noncanonical Wnt pathway regulates cell polarity and polarized cell movements in a variety of contexts. Recent data suggest that polarized cell movements required for outflow tract development are mediated by noncanonical Wnt pathways. Looptail (Lp) mice which are mutant for the cell polarity gene Vangl2 (strabismus), exhibit double-outlet right ventricle and defective muscularization of the outflow tract septum (Phillips et al., 2005; Henderson et al., 2006). Myocardializing cells express Vangl2 and other components of the planar cell polarity pathway, RhoA, and its downstream mediator ROCK1. RhoA expression is disrupted in Vangl2 mutants. Mice which are mutant for Wnt11, a noncanonical Wnt ligand, exhibit similar outflow tract remodeling defects to those observed in Vangl2 mutants. Intriguingly, the Wnt11 gene is a direct target of Pitx2, and similar phenotypes of Wnt11 and Pitx2 mutants suggest that Wnt11 is a critical effector of Pitx2 in outflow tract morphogenesis (Zhou et al., 2007). Expression of the transforming growth factor beta ligand encoding gene TGF2 is downregulated in both Pitx2 and Wnt11 mutants, and TGF2 mutants also exhibit outflow tract defects (Bartram et al., 2001). Together, these data demonstrate the existence of a canonical Wnt/Tbx1Pitx2-noncanonical Wnt-TGFb pathway regulating outflow tract morphogenesis.

VII.  Outstanding questions concerning second heart field deployment How exactly cells move from the second heart field into the elongating heart tube is unknown, and much of the basic cell biology of second heart field development, including the molecular basis for differentiation delay of second heart field cells relative to first heart field cells, remains to be investigated. Here we discuss a number of outstanding questions. How splanchnic mesoderm in the dorsal pericardial wall is replenished as cells of the second heart field contribute to the heart tube is unclear. We favor a model by which cells in cranial mesoderm move laterally and ventrally, and contribute to splanchnic mesoderm at the border of the embryonic coelom (or pericardial flexure) which corresponds to the somatic/splanchnic mesodermal divide. It remains to be seen whether second heart field cells at this junction undergo a mesenchymal-to-epithelial

162

transition, a prediction of such a model. This replenishment of splanchnic mesoderm appears to be a continuum of the early lateral–ventral movement of anterior mesodermal cells which form the first heart field in the cardiac crescent (De Ruiter et al., 1992). High levels of proliferation in pharyngeal mesoderm medial to the somatic–splanchnic divide would provide a continuous source of new cells for both somatic and splanchnic mesoderm as the pericardial cavity grows. This is a major site of Fgf10 expression, and Fgf8 is also expressed in this mesoderm in addition to adjacent ectoderm and endoderm. A nonexclusive alternative is that proliferation also occurs in splanchnic mesoderm of the dorsal pericardial wall; outflow tract myocardium itself appears to have a low proliferative rate (Sedmera et al., 2003; Soufan et al., 2006). Is the second heart field prepatterned prior to its addition to the heart tube? Transgene expression in the mouse suggests that the inferior wall of the early outflow tract contributes to ventral subpulmonary myocardium at later stages of development, and conversely, the superior wall of the early outflow tract to myocardium at the base of the aorta (Bajolle et al., 2006, 2008). Expression analysis at earlier developmental stages suggests that these cells may originate from anterior and posterior subdomains of the anterior heart field. In support of such a model, the caudal pharyngeal hypoplasia characterized by Tbx1 and Shh mutant embryos is associated with failure of posterior aortic arch artery development and pulmonary trunk aplasia. An alternative origin of subpulmonary myocardium has been suggested by Ward et al. (2005), who demonstrated that ablation of the right second heart field led to hypoplasia of the pulmonary trunk (see Chapter 7.2); consistent with this data is the result of a Pitx2c-Cre lineage mapping experiment in the mouse where Pitx2c descendants are observed in myocardium at the base of the aorta (Ai et al., 2006). Whether left–right or anterior–posterior patterning dominate in prefiguring future subaortic and subpulmonary myocardial domains remains to be seen; however these experiments suggest that distinct myocardial domains of the definitive heart are prefigured in distinct subdomains of the second heart field.

VIII.  The biomedical significance   of the second heart field VIII.A.  The Second Heart Field and Congenital Heart Defects About 50% of congenital heart defects detected at birth affect the arterial pole of the heart, including defects in outflow tract septation, alignment and ventricular septation (Iserin et al., 1998). The multiplicity of signaling events between the diverse cellular components of the arterial pole of the heart complicate investigation of the etiology of

PART  |  2  Cardiac Precursor Populations and Lineages

such defects. However, it is now apparent that either direct or indirect perturbation of second heart field deployment or defects in the second heart field-derived myocardial outflow tract wall can contribute significantly to congenital heart defects (Fig. 12). At one extreme, total failure of second heart field deployment blocks heart development at the linear heart tube stage, as in the case of Isl1 mutants or Mesp1-Cre Fgf8 conditional mutants (Cai et al., 2003; Park et al., 2006). More subtle defects in second heart field development in mouse and chick models lead to a spectrum of morphological anomalies similar to congenital heart defects observed in human patients. The second heart field paradigm allows deconstruction of complex cardiac phenotypes, and enhances our understanding of the etiology of such anomalies. A central example is that of neural crest ablation, which leads to overproliferation of second heart field cells, a shortened outflow tract and alignment defects, including overriding aorta and double-outlet right ventricle; the common ventricular outlet frequently observed in neural crest ablated embryos is also associated with a dextraposed outlet (Kirby and Waldo, 1995; Yelbuz et al., 2002; Waldo et al., 2005a). While failure of neural crest cells to enter the outflow tract directly causes the septation defects, it is the shortened outflow tract that leads to altered looping and consequent alignment defects. Thus, neural crest ablation is now thought to result in a composite cardiac phenotype comprised of direct effects on outflow tract septation and indirect effects via perturbation of second heart field development on outflow tract elongation and alignment (Hutson and Kirby, 2007; Chapter 7.2). In contrast to the alignment defects observed in neural crest ablated embryos, del22q11.2 syndrome patients display tetralogy of Fallot and pulmonary atresia. These defects are also observed after direct ablation of the second heart field or inhibition of FGF signaling in chick embryos (Ward et al., 2005; Hutson et al., 2006; Chapter 7.2). In these cases, loss or underproliferation of second heart field progenitor cells appears to underlie the observed phenotypes. The del22q11.2 candidate gene Tbx1 not only regulates second heart field proliferation, via a direct role in pharyngeal mesoderm, but also septation of the outflow tract via a separate requirement for Tbx1 in the formation of the aorticopulmonary septum (Xu et al., 2004). Recent experiments have shown that outflow tract extension and septation phenotypes can also be separated on the basis of mesodermal versus endodermal ablation of Fgf8 (Park et al., 2006). The second heart field thus provides a new framework for the characterization of congenital heart defects, and advances our understanding of the etiology of these malformations. Venous pole malformations in second heart field-mutant models are only beginning to be investigated; the recent identification of a hedgehog-d­ependent contribution of second heart fieldderived cells to the dorsal mesocardial protusion suggests

163

Chapter  |  2.2  The Second Heart Field

(A)

SHF CNC Tbx1

Tbx1 Fgf8

Fgf8 Fgf10

AHF proliferation Differentiation

(B)

(C)

Figure 12  Perturbation of the second heart field results in congenital heart defects. (A) Cartoon showing the juxtaposition between cells of the second heart field and neural crest cells (purple arrows) in the pharyngeal region during arterial pole morphogenesis. Reproduced from Kelly (2005) Trends Cardiovasc. Med. 15, 51–56, with permission of Elsevier. (B) Cardiac neural crest ablation results in a failure of outflow tract septation and a common ventricular outlet (left panel). Secondary overproliferation of second heart field progenitor cells in neural crest ablated embryos results in outflow tract alignment defects including double-outlet right ventricle (right panel). (C) Genetic or mechanical ablation of the second heart field results in tetralogy of Fallot type phenotypes (left panel) and pulmonary atresia (right panel).

that the second heart field also plays an important role in atrial and atrioventricular septation (Goddeeris et al., 2008). Intriguingly, human pedigrees with linked outflow and inflow tract anomalies have recently been reported,

suggesting that mutations in genes regulating second heart field development may underlie a specific spectrum of congenital defects affecting both poles of the heart (Bajolle et al., 2009).

164

VIII.B.  The Second Heart Field and Resident Cardiac Progenitor Cells Isl1 has been shown to be a marker of second heart field progenitor cells, and to be downregulated as these cells differentiate and contribute to the elongating heart tube (Cai et al., 2003). Although Isl1 expression is downregulated in most cardiac progenitors as they differentiate, analysis of an Isl1-nlacZ mouse and co-immunostaining for Isl1 and lineage markers demonstrated that Isl1 is expressed in distinct subdomains of the heart, and in diverse cardiovascular lineages (Moretti et al., 2006; Sun et al., 2007). During embryogenesis, Isl1 expression persists in myocardial lineages of the outflow tract, the atrial septum, and in sinoatrial and atrioventricular nodes. The myocardialized septum of the outflow tract also continues to express Isl1. Isl1-expressing cells also contribute to endothelial and vascular smooth muscle lineages, including smooth muscle of the coronary vessels. The potential role of Isl1 within these cardiac lineages remains to be explored. Isl1 expression within these subdomains of the heart gradually diminishes throughout development. At postnatal stages, a small number of Isl1-positive cells can be detected within the heart of different mammalian species (Laugwitz et al., 2005). The distribution of these cells in the heart echoes the distribution of Isl1-Cre recombinase traced cells, with lowest numbers in the left ventricle (Laugwitz et al., 2005). These cells appear to represent a residual population of undifferentiated second heart field cells which may maintain their progenitor status. Indeed, using an Isl1-Cre allele to sort Isl1 descendants with a fluorescent -galactosidase substrate, Laugwitz et al. (2005) have shown that undifferentiated Isl1-positive cells can be isolated and expanded on cardiac mesenchymal feeder layers and spontaneously differentiate when cultured with myocytes. Furthermore, Isl1-positive cells which are also Flk1- and Nkx2.5-positive have recently been identified using embryonic stem cells; these cells are multipotent, and when clonally propagated can give rise to myocardial, endothelial and smooth muscle lineages (Moretti et al., 2006). Recent experiments have shown that the mesenchymal feeder layer environment can be reconstituted by Wnt3a-secreting feeder cells, supporting a critical requirement for Wnt/-catenin signaling in regulating renewal and differentiation of second heart field cells (Qyang et al., 2007). Ongoing evaluation of the properties of resident cardiac progenitor cells is likely to lead to important advances in myocardial repair. Central questions remain how long Isl1-positive cells persist in the postnatal heart, what is the lineage relationship, if any, between Isl1-positive cells and other resident stem cell populations in the postnatal heart, and the extent to which the second heart field genetic program is maintained in these cells. Understanding the mechanisms underlying second heart field deployment and the regulation of second heart field progenitor cell differentiation in the early embryo can

PART  |  2  Cardiac Precursor Populations and Lineages

therefore be expected to provide insights into the properties and biomedical application of progenitor cells in the postnatal heart.

Acknowledgments RK is supported by the Inserm Avenir program, the Fondation de France, Fondation pour la Recherche Medicale, Agence Nationale de la Recherche and the European Community’s Sixth Framework Program contract (“HeartRepair”) LSHM-CT-2005-018630 and Seventh Framework Program “CardioGeNet”. SE would like to acknowledge and thank the American Heart Association and the National Institutes of Health for their ongoing support of research in the Evans laboratory.

References Abu-Issa, R., Waldo, K., Kirby, M.L., 2004. Heart fields: one, two or more? Dev. Biol. 272, 281–285. Abu-Issa, R., Smyth, G., Smoak, I., Yamamura, K., Meyers, E.N., 2002. Fgf8 is required for pharyngeal arch and cardiovascular development in the mouse. Development 129, 4613–4625. Aggarwal, V.S., Liao, J., Bondarev, A., Schimmang, T., Lewandoski, M., Locker, J., Shanske, A., Campione, M., Morrow, B.E., 2006. Dissection of Tbx1 and Fgf interactions in mouse models of 22q11DS suggests functional redundancy. Hum. Mol. Genet. 15, 3219–3228. Ai, D., Liu, W., Ma, L., Dong, F., Lu, M.F., Wang, D., Verzi, M.P., Cai, C., Gage, P.J., Evans, S., Black, B.L., Brown, N.A., Martin, J.F., 2006. Pitx2 regulates cardiac left–right asymmetry by patterning second cardiac lineage-derived myocardium. Dev. Biol. 296, 437–449. Ai, D., Fu, X., Wang, J., Lu, M.F., Chen, L., Baldini, A., Klein, W.H., Martin, J.F., 2007. Canonical Wnt signaling functions in second heart field to promote right ventricular growth. Proc. Natl. Acad. Sci. USA 104, 9319–9324. Alsan, B.H., Schultheiss, T.M., 2002. Regulation of avian cardiogenesis by Fgf8 signaling. Development 129, 1935–1943. Arguello, C., de la Cruz, M.V., Gomez, C.S., 1975. Experimental study of the formation of the heart tube in the chick embryo. J. Embryol. Exp. Morphol. 33, 1–11. Bajolle, F., Zaffran, S., Kelly, R.G., Hadchouel, J., Bonnet, D., Brown, N.A., Buckingham, M.E., 2006. Rotation of the myocardial wall of the outflow tract is implicated in the normal positioning of the great arteries. Circ. Res. 98, 421–428. Bajolle, F., Zaffran, S., Meilhac, S., Dandonneau, M., Chang, T., Kelly, R.G., Buckingham, M.E., 2008. Myocardium at the base of the aorta and pulmonary trunk is prefigured in the outflow tract of the heart and in subdomains of the second heart field. Dev. Biol. 313, 21–34. Bajolle, F., Zaffran, S., Losay, J., Buckingham, M., and Bonnet, D., 2009. Conotruncal defects associated with anomalous pulmonary venous connections. Arch. Cardiovasc. Dis. 103, 105–110. Bartram, U., Molin, D.G., Wisse, L.J., Mohamad, A., Sanford, L.P., Doetschman, T., Speer, C.P., Poelmann, R.E., Gittenberger-de Groot, A.C., 2001. Double-outlet right ventricle and overriding tricuspid valve reflect disturbances of looping, myocardialization, endocardial cushion differentiation, and apoptosis in TGF-beta(2)-knockout mice. Circulation 103, 2745–2752.

Chapter  |  2.2  The Second Heart Field

Beiman, M., Shilo, B.Z., Volk, T., 1996. Heartless, a Drosophila FGF receptor homolog, is essential for cell migration and establishment of several mesodermal lineages. Genes Dev. 10, 2993–3002. Bothe, I., Dietrich, S., 2006. The molecular setup of the avian head mesoderm and its implication for craniofacial myogenesis. Dev. Dyn. 235, 2845–2860. Brade, T., Gessert, S., Kuhl, M., Pandur, P., 2007. The amphibian second heart: Xenopus islet-1 is required for cardiovascular development. Dev. Biol. 311, 297–310. Brown, C.B., Wenning, J.M., Lu, M.M., Epstein, D.J., Meyers, E.N., Epstein, J.A., 2004. Cre-mediated excision of Fgf8 in the Tbx1 expression domain reveals a critical role for Fgf8 in cardiovascular development in the mouse. Dev. Biol. 267, 190–202. Bruneau, B.G., Nemer, G., Schmitt, J.P., Charron, F., Robitaille, L., Caron, S., Conner, D.A., Gessler, M., Nemer, M., Seidman, C.E., Seidman, J.G., 2001. A murine model of Holt-Oram syndrome defines roles of the T-box transcription factor Tbx5 in cardiogenesis and disease. Cell 106, 709–721. Buckingham, M., Meilhac, S., Zaffran, S., 2005. Building the mammalian heart from two sources of myocardial cells. Nat. Rev. Genet. 6, 826–835. Cai, C.L., Liang, X., Shi, Y., Chu, P.H., Pfaff, S.L., Chen, J., Evans, S., 2003. Isl1 identifies a cardiac progenitor population that proliferates prior to differentiation and contributes a majority of cells to the heart. Dev. Cell 5, 877–889. Cai, C.L., Zhou, W., Yang, L., Bu, L., Qyang, Y., Zhang, X., Li, X., Rosenfeld, M.G., Chen, J., Evans, S., 2005. T-box genes coordinate regional rates of proliferation and regional specification during cardiogenesis. Development 132, 2475–2487. Cai, C.L., Martin, J.C., Sun, Y., Cui, L., Wang, L., Ouyang, K., Yang, L., Bu, L., Liang, X., Zhang, X., Stallcup, W.B., Denton, C.P., McCulloch, A., Chen, J., Evans, S.M., 2008. A myocardial lineage derives from Tbx18 epicardial cells. Nature 454 (7200), 104–108. Christoffels, V.M., Mommersteeg, M.T., Trowe, M.O., Prall, O.W., de Gier-de Vries, C., Soufan, A.T., Bussen, M., Schuster-Gossler, K., Harvey, R.P., Moorman, A.F., Kispert, A., 2006. Formation of the venous pole of the heart from an Nkx2-5-negative precursor population requires Tbx18. Circ. Res. 98, 1555–1563. Cohen, E.D., Wang, Z., Lepore, J.J., Lu, M.M., Taketo, M.M., Epstein, D.J., Morrisey, E.E., 2007. Wnt/beta-catenin signaling promotes expansion of Isl-1-positive cardiac progenitor cells through regulation of FGF signaling. J. Clin. Invest. 117, 1794–1804. de la Cruz, M.V., Sanchez Gomez, C., Arteaga, M.M., Arguello, C., 1977. Experimental study of the development of the truncus and the conus in the chick embryo. J. Anat. 123, 661–686. DeRuiter, M.C., Poelmann, R.E., VanderPlas-de Vries, I., Mentink, M.M., Gittenberger-de Groot, A.C., 1992. The development of the myocardium and endocardium in mouse embryos. Fusion of two heart tubes? Anat. Embryol. (Berl). 185, 461–473. deVries, P.A., 1981. Evolution of precardiac and splanchnic mesoderm in relationship to the infundibulum and truncus. Perspect. Cardiovasc. Res. 5, 31–48. Dodou, E., Verzi, M.P., Anderson, J.P., Xu, S.M., Black, B.L., 2004. Mef2c is a direct transcriptional target of ISL1 and GATA factors in the anterior heart field during mouse embryonic development. Development 131, 3931–3942. Farrell, M., Waldo, K., Li, Y.X., Kirby, M.L., 1999. A novel role for cardiac neural crest in heart development. Trends Cardiovasc. Med. 9, 214–220. Firulli, A.B., Olson, E.N., 1997. Modular regulation of muscle gene transcription: a mechanism for muscle cell diversity. Trends Genet. 13, 364–369.

165

Franco, D., Meilhac, S.M., Christoffels, V.M., Kispert, A., Buckingham, M., Kelly, R.G., 2006. Left and right ventricular contributions to the formation of the interventricular septum in the mouse heart. Dev. Biol. 294, 366–375. Frank, D.U., Fotheringham, L.K., Brewer, J.A., Muglia, L.J., TristaniFirouzi, M., Capecchi, M.R., Moon, A.M., 2002. An Fgf8 mouse mutant phenocopies human 22q11 deletion syndrome. Development 129, 4591–4603. Galli, D., Dominguez, J.N., Zaffran, S., Munk, A., Brown, N.A., Buckingham, M.E., 2008. Atrial myocardium derives from the poster­ ior region of the second heart field, which acquires left/right identity as Pitx2c is expressed. Development 135 (6), 1157–1167. Garg, V., Yamagishi, C., Hu, T., Kathiriya, I.S., Yamagishi, H., Srivastava, D., 2001. Tbx1, a DiGeorge syndrome candidate gene, is regulated by sonic hedgehog during pharyngeal arch development. Dev. Biol. 235, 62–73. Gisselbrecht, S., Skeath, J.B., Doe, C.Q., Michelson, A.M., 1996. Heartless encodes a fibroblast growth factor receptor (DFR1/DFGFR2) involved in the directional migration of early mesodermal cells in the Drosophila embryo. Genes Dev. 10, 3003–3017. Goddeeris, M.M., Schwartz, R., Klingensmith, J., Meyers, E.N., 2007. Independent requirements for Hedgehog signaling by both the anter­ ior heart field and neural crest cells for outflow tract development. Development 134, 1593–1604. Goddeeris, M.M., Rho, S., Petiet, A., Davenport, C.L., Johnson, G.A., Meyers, E.N., Klingensmith, J., 2008. Intracardiac septation requires hedgehog-dependent cellular contributions from outside the heart. Development 135, 1887–1895. Gottlieb, P.D., Pierce, S.A., Sims, R.J., Yamagishi, H., Weihe, E.K., Harriss, J.V., Maika, S.D., Kuziel, W.A., King, H.L., Olson, E. N., Nakagawa, O., Srivastava, D., 2002. Bop encodes a musclerestricted protein containing MYND and SET domains and is essential for cardiac differentiation and morphogenesis. Nat. Genet. 31, 25–32. Grifone, R., Kelly, R.G., 2007. Heartening news for head muscle development. Trends Genet. 23, 365–369. Grimes, A.C., Stadt, H.A., Shepherd, I.T., Kirby, M.L., 2006. Solving an enigma: arterial pole development in the zebrafish heart. Dev. Biol. 290, 265–276. Guris, D.L., Duester, G., Papaioannou, V.E., Imamoto, A., 2006. Dosedependent interaction of Tbx1 and Crkl and locally aberrant RA signaling in a model of del22q11 syndrome. Dev. Cell 10, 81–92. Harrelson, Z., Kelly, R.G., Goldin, S.N., Gibson-Brown, J.J., Bollag, R.J., Silver, L.M., Papaioannou, V.E., 2004. Tbx2 is essential for patterning the atrioventricular canal and for morphogenesis of the outflow tract during heart development. Development 131, 5041–5052. Henderson, D.J., Phillips, H.M., Chaudhry, B., 2006. Vang-like 2 and noncanonical Wnt signaling in outflow tract development. Trends Cardiovasc. Med. 16, 38–45. Hochgreb, T., Linhares, V.L., Menezes, D.C., Sampaio, A.C., Yan, C.Y., Cardoso, W.V., Rosenthal, N., Xavier-Neto, J., 2003. A caudorostral wave of RALDH2 conveys anteroposterior information to the cardiac field. Development 130, 5363–5374. Hu, T., Yamagishi, H., Maeda, J., McAnally, J., Yamagishi, C., Srivastava, D., 2004. Tbx1 regulates fibroblast growth factors in the anterior heart field through a reinforcing autoregulatory loop involving forkhead transcription factors. Development 131, 5491–5502. Hutson, M.R., Kirby, M.L., 2007. Model systems for the study of heart development and disease. Cardiac neural crest and conotruncal malformations. Semin. Cell Dev. Biol. 18, 101–110.

166

Hutson, M.R., Zhang, P., Stadt, H.A., Sato, A.K., Li, Y.X., Burch, J., Creazzo, T.L., Kirby, M.L., 2006. Cardiac arterial pole alignment is sensitive to FGF8 signaling in the pharynx. Dev. Biol. 295, 486–497. Huynh, T., Chen, L., Terrell, P., Baldini, A., 2007. A fate map of Tbx1 expressing cells reveals heterogeneity in the second cardiac field. Genesis 45, 470–475. Ilagan, R., Abu-Issa, R., Brown, D., Yang, Y.P., Jiao, K., Schwartz, R.J., Klingensmith, J., Meyers, E.N., 2006. Fgf8 is required for anterior heart field development. Development 133, 2435–2445. Iserin, L., de Lonlay, P., Viot, G., Sidi, D., Kachaner, J., Munnich, A., Lyonnet, S., Vekemans, M., Bonnet, D., 1998. Prevalence of the microdeletion 22q11 in newborn infants with congenital conotruncal cardiac anomalies. Eur. J. Pediatr. 157, 881–884. Ivins, S., Lammerts van Beuren, K., Roberts, C., James, C., Lindsay, E., Baldini, A., Ataliotis, P., Scambler, P.J., 2005. Microarray analysis detects differentially expressed genes in the pharyngeal region of mice lacking Tbx1. Dev. Biol. 285, 554–569. Jerome, L.A., Papaioannou, V.E., 2001. DiGeorge syndrome phenotype in mice mutant for the T-box gene, Tbx1. Nat. Genet. 27, 286–291. Jiang, X., Rowitch, D.H., Soriano, P., McMahon, A.P., Sucov, H.M., 2000. Fate of the mammalian cardiac neural crest. Development 127, 1607–1616. Kattman, S.J., Huber, T.L., Keller, G.M., 2006. Multipotent flk-1 cardio­ vascular progenitor cells give rise to the cardiomyocyte, endothelial, and vascular smooth muscle lineages. Dev. Cell 11, 723–732. Kelly, R.G., 2005. Molecular inroads into the anterior heart field. Trends Cardiovasc. Med. 15, 51–56. Kelly, R.G., Buckingham, M.E., 2002. The anterior heart-forming field: voyage to the arterial pole of the heart. Trends Genet. 18, 210–216. Kelly, R.G., Papaioannou, V.E., 2007. Visualization of outflow tract development in the absence of Tbx1 using an FgF10 enhancer trap transgene. Dev. Dyn. 236, 821–828. Kelly, R.G., Zammit, P.S., Buckingham, M.E., 1999. Cardiosensor mice and transcriptional subdomains of the vertebrate heart. Trends Cardiovasc. Med. 9, 3–10. Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1, 435–440. Kelly, R.G., Jerome-Majewska, L.A., Papaioannou, V.E., 2004. The del22q11.2 candidate gene Tbx1 regulates branchiomeric myogenesis. Hum. Mol. Genet. 13, 2829–2840. Kioussi, C., Briata, P., Baek, S.H., Rose, D.W., Hamblet, N.S., Herman, T., Ohgi, K.A., Lin, C., Gleiberman, A., Wang, J., Brault, V., RuizLozano, P., Nguyen, H.D., Kemler, R., Glass, C.K., Wynshaw-Boris, A., Rosenfeld, M.G., 2002. Identification of a Wnt/Dvl/beta-Catenin → Pitx2 pathway mediating cell-type-specific proliferation during development. Cell 111, 673–685. Kirby, M.L., Waldo, K.L., 1995. Neural crest and cardiovascular patterning. Circ. Res. 77, 211–215. Klaus, A., Saga, Y., Taketo, M.M., Tzahor, E., and Birchmeier, W., 2007. Distinct roles of Wnt/beta-catenin and Bmp signaling during early cardiogenesis. Proc. Natl. Acad. Sci. (USA) 104, 18531–18536. Kwon, C., Arnold, J., Hsiao, E.C., Taketo, M.M., Conklin, B.R., Srivastava, D., 2007. Canonical Wnt signaling is a positive regulator of mammalian cardiac progenitors. Proc. Natl. Acad. Sci. USA 104, 10894–10899. Laugwitz, K.L., Moretti, A., Lam, J., Gruber, P., Chen, Y., Woodard, S., Lin, L.Z., Cai, C.L., Lu, M.M., Reth, M., Platoshyn, O., Yuan, J.X., Evans, S., Chien, K.R., 2005. Postnatal Isl1 cardioblasts enter fully differentiated cardiomyocyte lineages. Nature 433, 647–653.

PART  |  2  Cardiac Precursor Populations and Lineages

Li, Y.X., Zdanowicz, M., Young, L., Kumiski, D., Leatherbury, L., Kirby, M.L., 2003. Cardiac neural crest in zebrafish embryos contributes to myocardial cell lineage and early heart function. Dev. Dyn. 226, 540–550. Liao, J., Kochilas, L., Nowotschin, S., Arnold, J.S., Aggarwal, V.S., Epstein, J.A., Brown, M.C., Adams, J., Morrow, B.E., 2004. Full spectrum of malformations in velo-cardio-facial syndrome/DiGeorge syndrome mouse models by altering Tbx1 dosage. Hum. Mol. Genet. 13, 1577–1585. Lickert, H., Takeuchi, J.K., Von Both, I., Walls, J.R., McAuliffe, F., Adamson, S.L., Henkelman, R.M., Wrana, J.L., Rossant, J., Bruneau, B.G., 2004. Baf60c is essential for function of BAF chromatin remodelling complexes in heart development. Nature 432, 107–112. Liebner, S., Cattelino, A., Gallini, R., Rudini, N., Iurlaro, M., Piccolo, S., Dejana, E., 2004. Beta-catenin is required for endothelial-mesenchymal transformation during heart cushion development in the mouse. J. Cell Biol. 166, 359–367. Lin, Q., Schwarz, J., Bucana, C., Olson, E.N., 1997. Control of mouse cardiac morphogenesis and myogenesis by transcription factor MEF2C. Science 276, 1404–1407. Lin, L., Bu, L., Cai, C.L., Zhang, X., Evans, S., 2006. Isl1 is upstream of sonic hedgehog in a pathway required for cardiac morphogenesis. Dev. Biol. 295, 756–763. Lin, L., Cui, L., Zhou, W., Dufort, D., Zhang, X., Cai, C.L., Bu, L., Yang, L., Martin, J., Kemler, R., Rosenfeld, M.G., Chen, J., Evans, S.M., 2007. Beta-catenin directly regulates Islet1 expression in cardiovascular progenitors and is required for multiple aspects of cardiogenesis. Proc. Natl. Acad. Sci. USA 104, 9313–9318. Linask, K.K., Han, M.D., Linask, K.L., Schlange, T., Brand, T., 2003. Effects of antisense misexpression of CFC on downstream flectin protein expression during heart looping. Dev. Dyn. 228, 217–230. Linask, K.K., Han, M., Cai, D.H., Brauer, P.R., Maisastry, S.M., 2005. Cardiac morphogenesis: matrix metalloproteinase coordination of cellular mechanisms underlying heart tube formation and directionality of looping. Dev. Dyn. 233, 739–753. Lindsay, E.A., 2001. Chromosomal microdeletions: dissecting del22q11 syndrome. Nat. Rev. Genet. 2, 858–868. Lindsay, E.A., Vitelli, F., Su, H., Morishima, M., Huynh, T., Pramparo, T., Jurecic, V., Ogunrinu, G., Sutherland, H.F., Scambler, P.J., Bradley, A., Baldini, A., 2001. Tbx1 haploinsufficieny in the DiGeorge syndrome region causes aortic arch defects in mice. Nature 410, 97–101. Liu, C., Liu, W., Palie, J., Lu, M.F., Brown, N.A., Martin, J.F., 2002. Pitx2c patterns anterior myocardium and aortic arch vessels and is required for local cell movement into atrioventricular cushions. Development 129, 5081–5091. Liu, W., Selever, J., Wang, D., Lu, M.F., Moses, K.A., Schwartz, R.J., Martin, J.F., 2004. Bmp4 signaling is required for outflow-tract septation and branchial-arch artery remodeling. Proc. Natl. Acad. Sci. USA 101, 4489–4494. Lomonico, M.P., Moore, G.W., Hutchins, G.M., 1986. Rotation of the junction of the outflow tract and great arteries in the embryonic human heart. Anat. Rec. 216, 544–549. Macatee, T.L., Hammond, B.P., Arenkiel, B.R., Francis, L., Frank, D.U., Moon, A.M., 2003. Ablation of specific expression domains reveals discrete functions of ectoderm- and endoderm-derived FGF8 during cardiovascular and pharyngeal development. Development 130, 6361–6374. Maeda, J., Yamagishi, H., McAnally, J., Yamagishi, C., Srivastava, D., 2006. Tbx1 is regulated by forkhead proteins in the secondary heart field. Dev. Dyn. 235, 701–710.

Chapter  |  2.2  The Second Heart Field

Manner, J., 2004. On rotation, torsion, lateralization, and handedness of the embryonic heart loop: new insights from a simulation model for the heart loop of chick embryos. Anat. Rec. A. Discov. Mol. Cell Evol. Biol. 278, 481–492. Marguerie, A., Bajolle, F., Zaffran, S., Brown, N.A., Dickson, C., Buckingham, M.E., Kelly, R.G., 2006. Congenital heart defects in Fgfr2-IIIb and Fgf10 mutant mice. Cardiovasc. Res. 71, 50–60. Martinsen, B.J., Frasier, A.J., Baker, C.V., Lohr, J.L., 2004. Cardiac neural crest ablation alters Id2 gene expression in the developing heart. Dev. Biol. 272, 176–190. Meilhac, S.M., Esner, M., Kelly, R.G., Nicolas, J.F., Buckingham, M.E., 2004. The clonal origin of myocardial cells in different regions of the embryonic mouse heart. Dev. Cell 6, 685–698. Meilhac, S.M., Kelly, R.G., Rocancourt, D., Eloy-Trinquet, S., Nicolas, J.F., Buckingham, M.E., 2003. A retrospective clonal analysis of the myocardium reveals two phases of clonal growth in the developing mouse heart. Development 130, 3877–3889. Meno, C., Takeuchi, J., Sakuma, R., Koshiba-Takeuchi, K., Ohishi, S., Saijoh, Y., Miyazaki, J., ten Dijke, P., Ogura, T., Hamada, H., 2001. Diffusion of nodal signaling activity in the absence of the feedback inhibitor Lefty2. Dev. Cell 1, 127–138. Merscher, S., Funke, B., Epstein, J.A., Heyer, J., Puech, A., Lu, M.M., Xavier, R.J., Demay, M.B., Russell, R.G., Factor, S., Tokooya, K., Jore, B.S., Lopez, M., Pandita, R.K., Lia, M., Carrion, D., Xu, H., Schorle, H., Kobler, J.B., Scambler, P., Wynshaw-Boris, A., Skoultchi, A.I., Morrow, B.E., Kucherlapati, R., 2001. TBX1 is responsible for cardiovascular defects in velo-cardio-facial/DiGeorge syndrome. Cell 104, 619–629. Min, H., Danilenko, D.M., Scully, S.A., Bolon, B., Ring, B.D., Tarpley, J.E., DeRose, M., Simonet, W.S., 1998. Fgf-10 is required for both limb and lung development and exhibits striking functional similarity to Drosophila branchless. Genes Dev. 12, 3156–3161. Mjaatvedt, C.H., Yamamura, H., Capehart, A.A., Turner, D., Markwald, R.R., 1998. The Cspg2 gene, disrupted in the hdf mutant, is required for right cardiac chamber and endocardial cushion formation. Dev. Biol. 202, 56–66. Mjaatvedt, C.H., Nakaoka, T., Moreno-Rodriguez, R., Norris, R.A., Kern, M.J., Eisenberg, C.A., Turner, D., Markwald, R.R., 2001. The outflow tract of the heart is recruited from a novel heart-forming field. Dev. Biol. 238, 97–109. Mommersteeg, M.T., Hoogaars, W.M., Prall, O.W., de Gier-de Vries, C., Wiese, C., Clout, D.E., Papaioannou, V.E., Brown, N.A., Harvey, R. P., Moorman, A.F., Christoffels, V.M., 2007a. Molecular pathway for the localized formation of the sinoatrial node. Circ. Res. 100, 354–362. Mommersteeg, M.T., Brown, N.A., Prall, O.W., de Gier-de Vries, C., Harvey, R.P., Moorman, A.F., Christoffels, V.M., 2007b. Pitx2c and Nkx2-5 are required for the formation and identity of the pulmonary myocardium. Circ. Res. 101, 902–909. Moorman, A.F., Christoffels, V.M., Anderson, R.H., van den Hoff, M.J., 2007. The heart-forming fields: one or multiple? Philos. Trans. R. Soc. Lond. B Biol. Sci. 362, 1257–1265. Moretti, A., Caron, L., Nakano, A., Lam, J.T., Bernshausen, A., Chen, Y., Qyang, Y., Bu, L., Sasaki, M., Martin-Puig, S., Sun, Y., Evans, S. M., Laugwitz, K.L., Chien, K.R., 2006. Multipotent embryonic isl1 progenitor cells lead to cardiac, smooth muscle, and endothelial cell diversification. Cell 127, 1151–1165. Nathan, E., Monovich, A., Tirosh-Finkel, L., Harrelson, Z., Rousso, T., Rinon, A., Harel, I., Evans, S.M., Tzahor, E., 2008. The contribution of Islet1-expressing splanchnic mesoderm cells to distinct

167

branchiomeric muscles reveals significant heterogeneity in head muscle development. Development 135, 647–657. Niederreither, K., Vermot, J., Messaddeq, N., Schuhbaur, B., Chambon, P., Dolle, P., 2001. Embryonic retinoic acid synthesis is essential for heart morphogenesis in the mouse. Development 128, 1019–1031. Noden, D.M., 1991. Origins and patterning of avian outflow tract endocardium. Development 111, 867–876. Nowotschin, S., Liao, J., Gage, P.J., Epstein, J.A., Campione, M., Morrow, B.E., 2006. Tbx1 affects asymmetric cardiac morphogenesis by regulating Pitx2 in the secondary heart field. Development 133, 1565–1573. Ohuchi, H., Hori, Y., Yamasaki, M., Harada, H., Sekine, K., Kato, S., Itoh, N., 2000. FGF10 acts as a major ligand for FGF receptor 2 IIIb in mouse multi-organ development. Biochem. Biophys. Res. Commun. 277, 643–649. Park, E.J., Ogden, L.A., Talbot, A., Evans, S., Cai, C.L., Black, B.L., Frank, D.U., Moon, A.M., 2006. Required, tissue-specific roles for Fgf8 in outflow tract formation and remodeling. Development 133, 2419–2433. Perez-Pomares, J.M., Phelps, A., Sedmerova, M., Wessels, A., 2003. Epicardial-like cells on the distal arterial end of the cardiac outflow tract do not derive from the proepicardium but are derivatives of the cephalic pericardium. Dev. Dyn. 227, 56–68. Phan, D., Rasmussen, T.L., Nakagawa, O., McAnally, J., Gottlieb, P.D., Tucker, P.W., Richardson, J.A., Bassel-Duby, R., Olson, E.N., 2005. BOP, a regulator of right ventricular heart development, is a direct transcriptional target of MEF2C in the developing heart. Development 132, 2669–2678. Phillips, H.M., Murdoch, J.N., Chaudhry, B., Copp, A.J., Henderson, D. J., 2005. Vangl2 acts via RhoA signaling to regulate polarized cell movements during development of the proximal outflow tract. Circ. Res. 96, 292–299. Prall, O.W., Menon, M.K., Solloway, M.J., Watanabe, Y., Zaffran, S., Bajolle, F., Biben, C., McBride, J.J., Robertson, B.R., Chaulet, H., Stennard, F.A., Wise, N., Schaft, D., Wolstein, O., Furtado, M.B., Shiratori, H., Hamada, H., Black, B., Saga, Y., Robertson, E.J., Buckingham, M.E., Harvey, R.P., 2007. A Nkx2.5/BMP2/Smad1 negative feedback loop orhestrates cardiac progenitor cell specification and proliferation in the second heart field. Cell 128 (5), 947–959. Qyang, Y., Martin-Puig, S., Chiravuri, M., Chen, S., Xu, H., Bu, L., Jiang, X., Lin, L., Granger, A., Moretti, A., Caron, L., Wu, X., Clarke, J., Taketo, M.M., Laugwitz, K., Moon, R.T., Gruber, P., Evans, S.M., Ding, S., Chien, K.R., 2007. The renewal and differentiation of isl1 cardiovascular progenitors are controlled by a Wnt/catenin pathway. Cell Stem Cell 1, 165–179. Rana, M.S., Horsten, N.C., Tesink-Taekema, S., Lamers, W.H., Moorman, A.F., van den Hoff, M.J., 2007. Trabeculated right ventricular free wall in the chicken heart forms by ventricularization of the myocardium initially forming the outflow tract. Circ. Res. 100, 1000–1007. Reifers, F., Walsh, E.C., Leger, S., Stainier, D.Y., Brand, M., 2000. Induction and differentiation of the zebrafish heart requires fibroblast growth factor 8 (Fgf8/acerebellar).. Development 127, 225–235. Roberts, C., Ivins, S., Cook, A.C., Baldini, A., Scambler, P.J., 2006. Cyp26 genes a1, b1 and c1 are down-regulated in Tbx1-null mice and inhibition of Cyp26 enzyme function produces a phenocopy of DiGeorge Syndrome in the chick. Hum. Mol. Genet. 15 (23), 3394–3410. Roberts, C., Ivins, S.M., James, C.T., Scambler, P.J., 2005. Retinoic acid down-regulates Tbx1 expression in vivo and in vitro. Dev. Dyn. 232, 928–938.

168

Robertson, E.J., Charatsi, I., Joyner, C.J., Koonce, C.H., Morgan, M., Islam, A., Paterson, C., Lejsek, E., Arnold, S.J., Kallies, A., Nutt, S.L., Bikoff, E.K., 2007. Blimp1 regulates development of the poster­ior forelimb, caudal pharyngeal arches, heart and sensory vibrissae in mice. Development 134, 4335–4345. Rones, M.S., McLaughlin, K.A., Raffin, M., Mercola, M., 2000. Serrate and Notch specify cell fates in the heart field by suppressing cardiomyogenesis. Development 127, 3865–3876. Ryckebusch, L., Wang, Z., Bertrand, N., Lin, S., Chi, X., Schwartz, R., Zaffran, S., Niederreither, K., 2008. Retinoic acid deficiency alters second heart field formation. Proc. Natl. Acad. Sci. (USA) 105 (8), 2913–2918. Sato, M., Tsai, H.J., Yost, H.J., 2006. Semaphorin3D regulates invasion of cardiac neural crest cells into the primary heart field. Dev. Biol. 298, 12–21. Sato, M., Yost, H.J., 2003. Cardiac neural crest contributes to cardiomyogenesis in zebrafish. Dev. Biol. 257, 127–139. Sedmera, D., Reckova, M., DeAlmeida, A., Coppen, S.R., Kubalak, S.W., Gourdie, R.G., Thompson, R.P., 2003. Spatiotemporal pattern of commitment to slowed proliferation in the embryonic mouse heart indicates progressive differentiation of the cardiac conduction system. Anat. Rec. A. Discov. Mol. Cell Evol. Biol. 274, 773–777. Sekine, K., Ohuchi, H., Fujiwara, M., Yamasaki, M., Yoshizawa, T., Sato, T., Yagishita, N., Matsui, D., Koga, Y., Itoh, N., Kato, S., 1999. Fgf10 is essential for limb and lung formation. Nat. Genet. 21, 138–141. Seo, S., Kume, T., 2006. Forkhead transcription factors, Foxc1 and Foxc2, are required for the morphogenesis of the cardiac outflow tract. Dev. Biol. 296, 421–436. Shishido, E., Ono, N., Kojima, T., Saigo, K., 1997. Requirements of DFR1/Heartless, a mesoderm-specific Drosophila FGF-receptor, for the formation of heart, visceral and somatic muscles, and ensheathing of longitudinal axon tracts in CNS. Development 124, 2119–2128. Snarr, B.S., O’Neal, J.L., Chintalapudi, M.R., Wirrig, E.E., Phelps, A.L., Kubalak, S.W., Wessels, A., 2007. Isl1 expression at the venous pole identifies a novel role for the second heart field in cardiac development. Circ. Res. 101, 971–974. Soufan, A.T., van den Berg, G., Ruijter, J.M., de Boer, P.A., van den Hoff, M.J., Moorman, A.F., 2006. Regionalized sequence of myocardial cell growth and proliferation characterizes early chamber formation. Circ. Res. 99, 545–552. Srivastava, D., Thomas, T., Lin, Q., Kirby, M.L., Brown, D., Olson, E.N., 1997. Regulation of cardiac mesodermal and neural crest development by the bHLH transcription factor, dHAND. Nat. Genet. 16, 154–160. Stanley, E.G., Biben, C., Elefanty, A., Barnett, L., Koentgen, F., Robb, L., Harvey, R.P., 2002. Efficient Cre-mediated deletion in cardiac progenitor cells conferred by a 3UTR-ires-Cre allele of the homeobox gene Nkx2-5. Int. J. Dev. Biol. 46, 431–439. Stottmann, R.W., Choi, M., Mishina, Y., Meyers, E.N., Klingensmith, J., 2004. BMP receptor IA is required in mammalian neural crest cells for development of the cardiac outflow tract and ventricular myocardium. Development 131, 2205–2218. Sun, X., Meyers, E.N., Lewandoski, M., Martin, G.R., 1999. Targeted disruption of Fgf8 causes failure of cell migration in the gastrulating mouse embryo. Genes Dev. 13, 1834–1846. Sun, Y., Liang, X., Najaf, N., Cass, M., Lin, L., Cai, C., Chen, J., Evans, S., 2007. Islet1 is expressed in distinct cardiovascular lineages,

PART  |  2  Cardiac Precursor Populations and Lineages

including pacemaker and coronary vascular cells. Dev. Biol. 304, 286–296. Takeuchi, J.K., Mileikovskaia, M., Koshiba-Takeuchi, K., Heidt, A.B., Mori, A.D., Arruda, E.P., Gertsenstein, M., Georges, R., Davidson, L., Mo, R., Hui, C.C., Henkelman, R.M., Nemer, M., Black, B.L., Nagy, A., Bruneau, B.G., 2005. Tbx20 dose-dependently regulates transcription factor networks required for mouse heart and motoneuron development. Development 132, 2463–2474. Thompson, R.P., Abercrombie, V., Wong, M., 1987. Morphogenesis of the truncus arteriosus of the chick embryo heart: movements of autoradiographic tattoos during septation. Anat. Rec. 218, 434–440, 394–395. Tirosh-Finkel, L., Elhanany, H., Rinon, A., Tzahor, E., 2006. Mesoderm progenitor cells of common origin contribute to the head musculature and the cardiac outflow tract. Development 133, 1943–1953. Tzahor, E., Kempf, H., Mootoosamy, R.C., Poon, A.C., Abzhanov, A., Tabin, C.J., Dietrich, S., Lassar, A.B., 2003. Antagonists of Wnt and BMP signaling promote the formation of vertebrate head muscle. Genes Dev. 17, 3087–3099. Tzahor, E., 2007. Wnt/beta-catenin signaling and cardiogenesis: timing does matter. Dev. Cell.13:10–13. Verzi, M.P., McCulley, D.J., De Val, S., Dodou, E., Black, B.L., 2005. The right ventricle, outflow tract, and ventricular septum comprise a restricted expression domain within the secondary/anterior heart field. Dev. Biol. 287, 134–145. Viragh, S., Challice, C.E., 1973. Origin and differentiation of cardiac muscle cells in the mouse. J. Ultrastruct. Res. 42, 1–24. Vitelli, F., Morishima, M., Taddei, I., Lindsay, E.A., Baldini, A., 2002a. Tbx1 mutation causes multiple cardiovascular defects and disrupts neural crest and cranial nerve migratory pathways. Hum. Mol. Genet. 11, 915–922. Vitelli, F., Taddei, I., Morishima, M., Meyers, E.N., Lindsay, E.A., Baldini, A., 2002b. A genetic link between Tbx1 and fibroblast growth factor signaling. Development 129, 4605–4611. Vitelli, F., Zhang, Z., Huynh, T., Sobotka, A., Mupo, A., Baldini, A., 2006. Fgf8 expression in the Tbx1 domain causes skeletal abnormalities and modifies the aortic arch but not the outflow tract phenotype of Tbx1 mutants. Dev. Biol. 295, 559–570. von Both, I., Silvestri, C., Erdemir, T., Lickert, H., Walls, J.R., Henkelman, R.M., Rossant, J., Harvey, R.P., Attisano, L., Wrana, J.L., 2004. Foxh1 is essential for development of the anterior heart field. Dev. Cell 7, 331–345. Vong, L., Bi, W., O’Connor-Halligan, K.E., Li, C., Cserjesi, P., Schwarz, J.J., 2006. MEF2C is required for the normal allocation of cells between the ventricular and sinoatrial precursors of the primary heart field. Dev. Dyn. 235, 1809–1821. Waldo, K.L., Hutson, M.R., Stadt, H.A., Zdanowicz, M., Zdanowicz, J., Kirby, M.L., 2005a. Cardiac neural crest is necessary for normal addition of the myocardium to the arterial pole from the secondary heart field. Dev. Biol. 281, 66–77. Waldo, K.L., Hutson, M.R., Ward, C.C., Zdanowicz, M., Stadt, H.A., Kumiski, D., Abu-Issa, R., Kirby, M.L., 2005b. Secondary heart field contributes myocardium and smooth muscle to the arterial pole of the developing heart. Dev. Biol. 281, 78–90. Waldo, K.L., Kumiski, D.H., Wallis, K.T., Stadt, H.A., Hutson, M.R., Platt, D.H., Kirby, M.L., 2001. Conotruncal myocardium arises from a secondary heart field. Development 128, 3179–3388. Wang, Q., Lan, Y., Cho, E.S., Maltby, K.M., Jiang, R., 2005. Odd-skipped related 1 (Odd 1) is an essential regulator of heart and urogenital development. Dev. Biol. 288, 582–594.

Chapter  |  2.2  The Second Heart Field

Ward, C., Stadt, H., Hutson, M., Kirby, M.L., 2005. Ablation of the secondary heart field leads to tetralogy of Fallot and pulmonary atresia. Dev. Biol. 284, 72–83. Washington Smoak, I., Byrd, N.A., Abu-Issa, R., Goddeeris, M.M., Anderson, R., Morris, J., Yamamura, K., Klingensmith, J., Meyers, E.N., 2005. Sonic hedgehog is required for cardiac outflow tract and neural crest cell development. Dev. Biol. 283, 357–372. Xu, H., Cerrato, F., Baldini, A., 2005. Timed mutation and cell-fate mapping reveal reiterated roles of Tbx1 during embryogenesis, and a crucial function during segmentation of the pharyngeal system via regulation of endoderm expansion. Development 132, 4387–4395. Xu, H., Morishima, M., Wylie, J.N., Schwartz, R.J., Bruneau, B. G., Lindsay, E.A., Baldini, A., 2004. Tbx1 has a dual role in the morphogenesis of the cardiac outflow tract. Development 131, 3217–3227. Yamagishi, H., Maeda, J., Hu, T., McAnally, J., Conway, S.J., Kume, T., Meyers, E.N., Yamagishi, C., Srivastava, D., 2003. Tbx1 is regulated by tissue-specific forkhead proteins through a common Sonic hedgehog-responsive enhancer. Genes Dev. 17, 269–281. Yang, L., Cai, C.L., Lin, L., Qyang, Y., Chung, C., Monteiro, R.M., Mummery, C.L., Fishman, G.I., Cogen, A., Evans, S., 2006. Isl1Cre reveals a common Bmp pathway in heart and limb development. Development 133, 1575–1585. Yelbuz, T.M., Waldo, K.L., Kumiski, D.H., Stadt, H.A., Wolfe, R.R., Leatherbury, L., Kirby, M.L., 2002. Shortened outflow tract leads to altered cardiac looping after neural crest ablation. Circulation 106, 504–510.

169

Yelbuz, T.M., Waldo, K.L., Zhang, X., Zdanowicz, M., Parker, J., Creazzo, T.L., Johnson, G.A., Kirby, M.L., 2003. Myocardial volume and organization are changed by failure of addition of secondary heart field myocardium to the cardiac outflow tract. Dev. Dyn. 228, 152–160. Yuan, S., Schoenwolf, G.C., 2000. Islet-1 marks the early heart rudiments and is asymmetrically expressed during early rotation of the foregut in the chick embryo. Anat. Rec. 260, 204–207. Zaffran, S., Kelly, R.G., Meilhac, S.M., Buckingham, M.E., Brown, N.A., 2004. Right ventricular myocardium derives from the anterior heart field. Circ. Res. 95, 261–268. Zeisberg, E.M., Ma, Q., Juraszek, A.L., Moses, K., Schwartz, R.J., Izumo, S., Pu, W.T., 2005. Morphogenesis of the right ventricle requires myocardial expression of Gata4. J. Clin. Invest. 115, 1522–1531. Zhang, X.M., Ramalho-Santos, M., McMahon, A.P., 2001. Smoothened mutants reveal redundant roles for Shh and Ihh signaling including regulation of L/R symmetry by the mouse node. Cell 106, 781–792. Zhang, Z., Baldini, A., 2008. In vivo response to high-resolution variation of Tbx1 mRNA dosage. Hum. Mol. Genet. 17, 150–157. Zhang, Z., Huynh, T., Baldini, A., 2006. Mesodermal expression of Tbx1 is necessary and sufficient for pharyngeal arch and cardiac outflow tract development. Development 133, 3587–3595. Zhou, W., Lin, L., Majumdar, A., Li, X., Zhang, X., Liu, W., Etheridge, L., Shi, Y., Martin, J., Van de Ven, W., Kaartinen, V., WynshawBoris, A., McMahon, A.P., Rosenfeld, M.G., amd Evans, S.M., 2007. Modulation of morphogenesis by noncanonical Wnt signaling requires ATF/CREB family-mediated transcriptional activation of TGFbeta2. Nat. Genet. 39, 1225–1234.

Chapter 2.3

Patterning and Development of the Conduction System of the Heart: Origins of the Conduction System in Development Vincent M. Christoffels, Willem M.H. Hoogaars and Antoon F.M. Moorman Heart Failure Research Center, Academic Medical Center, Amsterdam, The Netherlands

I.  Introduction The conduction system of the heart initiates and coordinates the electrical signal that causes the rhythmic and synchronized contractions of the atria and ventricles. In higher vertebrates this system is generally considered to comprise the nodes and the rapidly-conducting “wiring” of the ventricles. The sinoatrial node (SAN; sinus node) is the primary pacemaking component that generates the electrical impulse (Keith and Flack, 1907; Davies et al., 1983). In the mature heart, the node is located at the junction of the superior caval vein and right atrium (Fig. 1). From there, the impulse propagates rapidly through the fast-conducting atrial muscle. Then the electrical impulse travels slowly through the atrioventricular (AV) node, loc­ ated in the bottom of the right atrium just adjacent to the tricuspid valve in the triangle of Koch (Anderson and Ho, 2003). This slow-conducting node forms the only myocardial connection via the AV bundle between the atria and ventricles, which for the remainder are electrically insulated from each other by a fibrous skeleton (Fig. 1). The delay in propagation of the impulse to the ventricles is essential for heart function, as it allows the ventricles to be filled by atrial contraction. Moreover, the AV nodal delay provides protection from ventricular arrhythmias that may be triggered by the atria (Dobrzynski et al., 2003; Kreuzberg et al., 2006a). From there, the electrical impulse enters the fast-conducting atrioventricular bundle Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

(His bundle) and bundle branches at both sides of the ventricular septum and, subsequently, the Purkinje fiber network, which rapidly transmits the impulse to the ventricular working myocardium from apex to base. From this electrical configuration, an electrocardiogram (ECG) can be derived in which the atrial activation shows as the P–wave, the atrioventricular delay as the P–Q interval, and the ventricular activation pattern as the QRS complex (Fig. 2C). “Conduction system” is somewhat of a misnomer (Moorman and Christoffels, 2003a); it includes the slow-conducting SAN and AV node, as well as the fastconducting AV bundle, bundle branches and Purkinje fiber network, while excluding the fast-conducting atrial and ventricular working myocardium. It is therefore more appropriate to speak about the “pacemaking and conduction” system (Gourdie et al., 2003c). Indeed, as will be discussed below, these tissues have distinct origins and phenotypes. The components of the conduction system in mature mammalian hearts are morphologically well-defined, although species differences exist, ranging from welldeveloped structures in hoofed animals to a poorly developed system in rodents, with the human in between (Romer, 1962; Moorman and Christoffels, 2003b). In birds the system is also well-developed (Davies, 1930; Lu et al., 1993). In contrast, in lower vertebrates or mammalian embryos, this is not the case. Nevertheless, the activation and contraction pattern and derived ECG described above

171

172

PART  |  2  Cardiac Precursor Populations and Lineages

oft

oft

oft rv

ev

avn avc

lv ift

ift

E8

avb

E8.5

E9.5-10.5

scv

ra

la

oft

la lbb

ra lv

avn

lv

primary ring

san

sh

san

avc rv

a

rv avj

avb

avb

E10.5 -12.5

Pn

rbb Mature

Figure 1  Schematic overview of heart development in higher vertebrates. Chamber myocardium (red, ventricular; blue, atrial) expands from the outer curvatures of the primary heart tube, whereas nonchamber myocardium (gray) of the inflow tract (ift), sinus horns (sh), atrioventricular canal (avc), outflow tract (oft) and inner curvatures does not expand. Sinus horn myocardium gives rise to the sinoatrial node (san), atrioventricular canal myocardium to the atrioventricular node (avn) and atrioventricular junction. The first three panels show left-lateral view (a: atrium; avb: atrio­ ventricular bundle; avc: atrioventricular canal; avj: atrioventricular junction; avn: atrioventricular node; ev: embryonic ventricle; ift: inflow tract; la: left atrium; lbb: left bundle branch; lv: left ventricle; oft: outflow tract; Pn: purkinje fibers; ra: right atrium; rbb: right bundle branch; rv: right ventricle; san: sinoatrial node; scv: superior caval vein; sh: sinus horn).

(A)

(B)

is evolutionary conserved, and has been realized already in the fish heart (Randall, 1970; Moorman and Christoffels, 2003b; Chapter 1.1). Here, the impulse is generated at the blood intake side of the heart, rapidly propagated through the single atrium, delayed in the AV junction, and then spreads through the apex and base of the single ventricle (Randall, 1970). From the chick embryo a “mature” ECG can already be derived when the ventricular and atrial chambers start to develop in the embryonic heart tube (Fig. 2) (van Mierop, 1967; Paff et al., 1968; Seidl et al., 1981). As soon as a heart tube is formed in the early embryo, pacemaker activity can be recorded at the intake. The electrical impulse migrates very slowly through the myocardium, resulting in a sinusoidal ECG and a sluggish, peristaltic contraction pattern (van Mierop, 1967; Paff et al., 1968; Seidl et al., 1981). Fast-conducting atrial and ventricular compartments will subsequently form at discrete sites within the vertebrate heart tube, and a slow-conducting AV canal in between will be “left behind” and function to delay the impulse going from atrium to ventricle (de Jong et al., 1992; Moorman and Christoffels, 2003b; Chapter 3.2). The configuration of the distinct components of the embryonic heart is sufficient to maintain a synchronous activation cycle, the derived ECG of which starts to resemble an adult ECG from HH13 onwards in chicks (50 hours of incubation, 20 somites) (Paff et al., 1968; Seidl et al., 1981). These observations imply that conduction system function is already present in the early embryonic heart, and that this function does not require the formation of morphologically-defined conduction system components. Nonetheless, defined conduction system components are Figure 2  Development of chambers and the ECG. (A–B) Left-sided view of embryonic hearts of E8–8.5 (A) and E9.5–10.5 (B). Formation of the chambers is in accordance with the development of an adult-like ECG. Red arrows demarcate the ventricular and atrial inner curvatures. (C) Schematic overview showing the different ECG segments correlating with the different components of the embryonic heart. The black arrow depicts the flowthrough and the sequential mode of activation. Note the PQ interval that is caused by the slow conducting AV canal. (D) Serial sections of an E9.5 mouse heart. Cx40 is expressed in the developing chambers, while Tbx3 demarcates the AV canal (black arrowhead). (E) Expression of Cx40 is conserved in the developing (E3) chick heart.

(C)

oft

oft rv ev

avc

at

lv ift ift

(D)

(E)

oft

lv Tbx3

TSRQ

avc

lv

at ift

at

P

oft lv

avc at

Cx40

Cx40

ift

Chapter  |  2.3  Patterning and Development of the Conduction System of the Heart

required for proper functioning of the mature mammalian heart. Congenital malformations or surgical interventions affecting one or more of these components result in arrhythmias, such as bradycardia, or AV block. A number of excellent reviews on the development of the conduction system are available (Moorman et al., 1998; Pennisi et al., 2002; Gourdie et al., 2003c; Myers and Fishman, 2003; Mikawa and Hurtado, 2007) which cover various aspects of conduction system formation. In this chapter we focus on the cellular origin of the conduction system components and the molecular genetic mechanisms that may control their phenotype and position within the developing heart. We will try to find the connection between heart cell precursor pools, which in a temporal pattern form the heart, and the genesis of the conduction system components. Unless specified otherwise, the mouse has been used as the principle model.

II.  Myocardial origin and composition of the developing conduction system The distinct components of the cardiac conduction system of the heart are essentially myocardial (Fig. 1) (de Haan, 1961; Virágh and Challice, 1977a,b, 1980, 1982). They are innervated by cardiac ganglia largely derived from neural crest (Kirby and Stewart, 1983; Canale et al., 1986; Verberne et al., 1998). In addition, a large fraction of cells in the mature conduction system is noncardiac (i.e., fibroblasts), and insulating layers of fibrous tissue are found around conduction system components, such as the SAN and AV bundle (Anderson et al., 1976; Virágh and Challice, 1977b, 1980; Davies et al., 1983). These noncardiac cell types are derived from the epicardium (fibro­ blasts), endocardium, neural crest (neural innervations) and other sources, although their origins have not been defined in detail (Moorman and Christoffels, 2003b). Although these nerves and fibrous tissues are important, or even a prerequisite, for conduction system formation and function (Eralp et al., 2006; Gurjarpadhye et al., 2007), the cardiomyocytes are essential for impulse generation and propagation. Furthermore, in the embryo the functional myocardial conduction system is not yet innervated and interstitial fibroblast and fibrous tissues in association with the conduction system are sparse or absent (van Mierop and Gessner, 1970; Virágh and Challice, 1980, 1982). Studies of the development of the conduction system components strongly suggest that they originate from myocardial precursors, which in turn are derived from mesoderm and pericardial wall mesenchyme (Virágh and Challice, 1973, 1980, 1982; Buckingham et al., 2005). Nevertheless, the finding that myocardium of the conduction system expresses neural genes (e.g., neurofilament) and proteins also identifying neural crest (e.g., Gln2, HNK-1)

173

has invoked the hypothesis of a neural crest origin for the conduction system (Gorza et al., 1988). However, neural crest enters the chick heart at E4, far beyond the time when conduction system function and an adult fashion ECG can be registered. Furthermore, lineage analysis experiments in chick strongly support a myogenic origin of the conduction system. Retroviral-labeled single cardiac cells of the early embryonic heart (E3, HH15–17,  30 somites) were found to give rise to multicellular clones containing both working myocardium and myocardial conduction cells (Gourdie et al., 1995; McCabe et al., 1995). These analyses also suggested that the specification and differentiation of the conduction system is a local process occurring within single cell-derived clusters, because no progeny was found outside these clusters. Moreover, in both chick and mouse, labeled neural crest or proepicardial cells were never traced to the conduction system components, even though their descendants were found in close association with these components (Cheng et al., 1999; Gittenberger-de Groot et al., 2003; Nakamura et al., 2006). In conclusion, the cardiac myocytes can be regarded as the source and principal cell type of the conduction system.

III.  Markers for the forming conduction system A selective set of markers have been used to gain insight into the formation of the conduction system in various species. These markers have been of great value in visualizing the differentiation and morphogenetic processes underlying conduction system formation, but they sometimes cause confusion regarding the phenotype or origin of conduction system cells (see above). Some markers represent a phenotypic characteristic that is shared by all conduction system components, such as an underdeveloped contractile apparatus (pale cytoplasm), slow proliferation, or neurogenic gene expression. Phenotypic characteristics may also be very distinct between components. For example, conduction is slow in the SA and AV nodes, but fast in the AV bundle, bundle branches, Purkinje fibers and in the atrial and ventricular working myocardium, which is reflected in the expression profile of gap-junction encoding genes. Gap junctions importantly contribute to the conductivity of myocardium in an isoform-specific manner. It is therefore not surprising that their genes are among the best markers for the developing conduction system (see Table 1). The very low conductance (9 pS) gap junction Cx30.2 is specifically expressed in the SA and AV node, and is required to slow down the conduction in the AV node (Kreuzberg et al., 2006a). Cx40 is expressed in all atrial and venous myocardium (myocardium of the pulmonary vein and systemic veins) and in the ventricular conduction system of the mouse heart, but throughout development is selectively excluded from the SAN and AV node, allowing

174

Table 1  Markers of the Cardiac Conduction System Histochemistry

SH

SAN

A

AVC

AVN

AVB

BB

V

PF

OFT

remarks

references

AchE*





















*Acethylcholine esterase, demarcates CCS innervation

Lamers et al., 1987

PAS*





















*Periodic acid Schiff’s stains glycogen deposits

Virágh and Challice, 1977a,b, 1982

Transcription factors 





*

*









*

*Not detected at later stages (E16.5–adult)

Christoffels et al., 2004b

Tbx3





















Expressed in developing and adult proximal CCS

Hoogaars et al., 2004

Tbx5

















/



Enriched in adult CCS

Bruneau et al., 1999; Moskowitz et al., 2004

Id2





















Expressed in cushions

Moskowitz et al., 2007

Tbx18















*





* IVS / LV

Christoffels et al., 2006

Nkx2-5

/  *

/  *

















*Expression from E14.5 onward

Thomas et al., 2001; Mommersteeg et al., 2007

Msx2









ND











Expression in chick (HH15–37).

Chan-Thomas et al., 1993

Cx30.2





















Expressed in adult

Kreuzberg et al., 2005

Cx40

/  *













**





*Expression from E14.5 onward **Expression absent in adult

Delorme et al., 1995; Coppen et al., 2003

Cx43





















Absent from CCS

Delorme et al., 1997; Coppen et al., 2003

Cx45

ND



















Initially present in whole heart tube, later enriched in CCS and OFT

Coppen et al., 1999, 2001

Kcnk3/TASK1

ND

ND







*

*



*



*Expression restricted to PVSC at later stages (E13.5–adult)

Graham et al., 2006

Hcn4





















Expression in all conduction system components

Moosmang et al., 2001; Garcia-Frigola et al., 2003

Ionchannels

(Continued)

PART  |  2  Cardiac Precursor Populations and Lineages

Tbx2

PSA-NCAM





















Marks PVCS

Watanabe et al., 1992; Chuck and Watanabe, 1997

HNK-1/Leu7





















Protein expressed in human, rat and chick

Nakagawa et al., 1993; Chuck and Watanabe, 1997; Blom et al., 1999; Ikeda et al., 1990

Gln2





















Protein expressed in human

Wessels et al., 1992

NF160/NF-M





















Specific epitope for rabbit

Gorza and Vitadello, 1989; Vitadello et al., 1996

EAP-300

ND

ND

*

*







*



*

Chick

McCabe et al., 1995

*Expression absent at later stages

Transgenes cGata6-LacZ





















Marks AVCS

Davis et al., 2001b; Viswanathan et al., 2007

cTnI-LacZ





















Marks AVCS

Di Lisi et al., 2000

Des1-nLacZ









ND











Marks AVCS  BB

Li et al., 1993

CCS-LacZ





















Marks CCS

Rentschler et al., 2001; Viswanathan et al., 2007

MinK-LacZ





















Marks CCS

Kondo et al., 2003; Viswanathan et al., 2007

Hf-1b-LacZ





















Marks PVCS

Nguyen-Tran et al., 2000

Hop-LacZ



/

















Enriched in adult CCS

Ismat et al., 2005; Viswanathan et al., 2007

Chapter  |  2.3  Patterning and Development of the Conduction System of the Heart

Neuronal

175

176

PART  |  2  Cardiac Precursor Populations and Lineages

(A)

(B)

san

rvv lavj

ravj

la

ra

avb lv

rv

Tbx3

(C)

Cx43

pv

san

lvv as

rvv

la

lvv

rvv as

ra

san

avc

avc avb

E9.5

san

E12.5

ravrb avn

lavj lbb

rbb

E17.5

Figure 3  Tbx3 expression delineates the conduction system. (A) Reconstruction of the Tbx3 expression domain (red) in the atrioventricular canal, atrioventricular bundle and bundle branches at embryonic day (E12.5). The lumen of the right and left ventricle are depicted in green and yellow, respectively. (B) Serial sections of an E12.5 heart showing the expression of Tbx3 and Cx43 in relation to the 3-D reconstruction of panel (A). Red arrows indicate complementary expression of Tbx3 and Cx43. (C) Dorsal views of 3D reconstructions of the developing mouse heart at stage E9.5, E12.5 and E17.5, showing the expression profile of Tbx3 in the developing nodes, putative internodal tracts, atrioventricular bundle and bundle branches. The myocardium has been removed revealing the lumen of the atria in blue. The Tbx3-expressing myocardium is shown in red (as: atrial septum; avb: atrioventricular bundle; avc: atrioventricular canal; avn: atrioventricular node; la/ra: left/right atrium; lavj: left atrioventricular junction; l/rbb: left/righ bundle branches; pv: pulmonary vein; ravrb: right atrioventricular ring bundle; san: sinoatrial node; l/rvv: left/right venous valve). Adapted from Hoogaars et al. (2004).

their unambiguous delineation. Likewise, Cx43 discriminates between the AV bundle and bundle branches (Cx43negative) and ventricular chamber (working) myocardium (Cx43-positive), providing one of the earliest and most discriminating markers for these structures (Fig. 3B). In addition to endogenous markers, there are a number of transgenes that mark components of the conduction system. The CCS-lacZ strain (enhancer trap, lacZ cassette driven by Slco3A1 and additional unknown regulatory sequences (Stroud et al., 2007)) seems to mark all components of the conduction system from very early stages onwards (Rentschler et al., 2001). LacZ targeted to the minK locus also reveals conduction system components, while the endogenous gene does not show this typical expression pattern (Kondo et al., 2003). Likewise, while Gata6 and cTnI are broadly expressed in the heart, regulatory sequences derived from these genes drive reporter gene expression in the early AV canal and AV node (Di Lisi et al., 2000; Davis et al., 2001a). Crossed into mouse models with conduction system defects, these transgenes represent powerful tools to analyze the molecular genetics of conduction system formation. The patterns of these transgenes have not always been validated in the conduction system components they are supposed to mark. To understand what these transgenes

reveal, and whether they can be used to trace conduction system cells precisely and completely throughout development, a comparison of their expression patterns with those of endogenous genetic markers that distinguish between conduction system components and developing working myocardium, or between myocytes and other cell types (e.g., fibroblasts), is an essential prerequisite. For example, the CCS-lacZ pattern reveals the ventricular conduction system, as shown by functional studies (Rentschler et al., 2001). In contrast, its pattern in the atria marks large parts of the right and left atria not known to contribute to the conduction system, while it fails to mark the SAN (Myers and Fishman, 2003; Jongbloed et al., 2004; Viswanathan et al., 2007). In the next sections the development of each of the conduction system components are discussed, referring to markers provided in Table 1.

IV.  Pacemaker activity, polarity and the formation of the sinus node All cardiac myocytes within the heart are, in principle, capable of producing an intrinsic cycle of electrical activity, called automaticity, intrinsic rhythmicity or pacemaker

177

Chapter  |  2.3  Patterning and Development of the Conduction System of the Heart

(A) san

Desmin Nkx2-5

Desmin Hcn4

rsh

ra

ra

(B)

san

Hcn4

Nppa

avc avc raa vv cm

as

<E9.5

Nkx2-5+ Nkx2-5

Nkx2-5 pv

Pitx2

mesenchyme

laa

san rsh

lsh

Nkx2-5+ Cx40+ Hcn4+ Tbx3+ Hcn4+ SAN sinus horn

Pitx2

E9.5-E14.5

Figure 4  Development of the sinus horns and sinus node. (A) The panels at the left side show sister sections of a wild-type E11.5 embryo doublestained with antibodies against Desmin (blue) and Hcn4 or Nkx2-5 (pink). The panels at the right side show the patterns of Nppa and Hcn4 in the SAN and right atrium. Note the mutually-exclusive patterns of sinoatrial node expression (Hcn4) and atrial gene expression (Nkx2-5, Nppa). (B) Schematic representation of sinus horn development and localized formation of the sinus node during heart development showing patterns of expression until E9.5, and between E9.5 and E14.5. Dotted lines represent borders of the expression domains of Nkx2-5 and Pitx2c. The horizontal (Nkx2-5) dotted lines also depict lineage borders. Except for the yellow-colored mesenchyme, only myocardium is depicted (as: atrial septum; avc: atrioventricular canal; la/ra: left/right atrium; laa/raa: left/right atrial appendage; lsh/rsh: left/right sinus horn; pv: pulmonary vein; san: sinoatrial node; vv: venous valves). Adapted from Mommersteeg et al. (2007).

activity, which leads to contraction. Because all cardiac cells in the heart are electrically coupled, the cells with the highest intrinsic cycle rate (rate of depolarization) will take the lead. Normally, the cardiac cells within the SAN, positioned at the inflow of the right atrium, have the highest intrinsic rate. Therefore, this structure initiates depolarization of the heart and dictates the rhythm of contraction. In the embryo, the heart is a simple tubular structure consisting of embryonic (primitive) myocardial cells with an endocardial lining. All these myocardial cells are automatic. However, the cells at the intake have the fastest cycle, which function as the dominant pacemaker that dictates the heart rate (Patten, 1949; van Mierop, 1967). Therefore, pacemaker activity is the first function of the heart to arise. Several ion currents contribute to the automaticity of the cardiac cells. These include the hyperpolarization­activated cation or funny current (Ih or If), L-type Ca2 current (ICaL), T-type Ca2 current (ICaT), delayed rectifier K current (IK) and sustained inward current (Ist) (Schram et al., 2002; Satoh, 2003). Members of the hyperpolarization-activated, cyclic nucleotide-gated cation channel (HCN) family generate pacemaker current (If), involved in phase 4 depolarization. This current has been implicated in pacemaker activity of spontaneously firing brain and cardiac muscle cells (DiFrancesco, 1993). At early developmental stages the Hcn4 family member is prevalently expressed, in a strikingly venous pole-enriched pattern

(Fig. 4) (Garcia-Frigola et al., 2003; Mommersteeg et al., 2007). In the mature heart, Hcn4 is selectively expressed in the SA and AV nodes (Moosmang et al., 2001; Dobrzynski et al., 2003; Liu et al., 2007). Embryos lacking Hcn4 die before E11.5, and show strongly-reduced contraction rates (Stieber et al., 2003). Humans heterozygous for a mutation in HCN4 display bradycardia (Milanesi et al., 2006). These observations indicate that Hcn4 plays an important role in the dominant pacemaker activity of the venous pole of the developing heart. Other ion channel genes such as Cav3.1, Cav1.3 and ERG1, which encode the T-type calcium current, the L-type calcium current and the slow component of the delayed rectifier current (IKr), respectively, have also been implicated in pacemaker activity. Knockout models of these genes show sinus node dysfunction, bradycardia and/or delayed atrioventricular conduction (Lees-Miller et al., 2003; Mangoni et al., 2003, 2006). Although these currents are expressed in the SA and AV node, detailed spatio–temporal expression of these genes in the developing heart has not been established (Bohn et al., 2000; Efimov et al., 2004). The cells of the definitive SAN form in the right sinus horn at the junction with the atrium from E9.5 onward, shortly after the establishment of the “mature configuration” of the four-chambered heart (Fig. 4.) (van Mierop and Gessner, 1970; Virágh and Challice, 1980). Based on histological sections, Virágh and Challice (1980) described the formation from E11 onwards of sinus ­ muscle cells,

178

PART  |  2  Cardiac Precursor Populations and Lineages

including the SAN primordium, from loose mesenchymal cells of the pericardial wall. This is consistent with the notion that the entire heart tube forms by continuous recruitment of noncardiac precursors to the poles (Buckingham et al., 2005; Chapter 2.2). The SAN precursors co-express Isl1 and Tbx18, indicating that the SAN forms from overlapping second heart field and caudal heart field precursors (see Fig. 5) (Christoffels et al., 2006; Moretti et al., 2006; Sun et al., 2007; Mommersteeg et al., 2007). Therefore, while pacemaker activity is the first function to arise in the heart, the definitive SAN cells are among the last to be added to the heart. During the recruitment of noncardiac cells to the venous pole (Figs 4; 5), dominant pacemaker activity is always found at the venous pole (Patten, 1949; van Mierop, 1967). These observations indicate that dominant pacemaker activity continuously shifts to the most caudal (inflow tract side) part of the developing heart tube, i.e., to the cardiac cells that have been added most recently to the venous side. We recently uncovered a possible mechanism for this phenomenon (Fig. 4) (Mommersteeg et al., 2007). Subsequent to the formation of the Nkx2-5-positive heart tube, nonmyocardial mesenchymal cells bordering the inflow tract of the heart tube give rise to the Nkx2-5-negative myocardial cells of the SAN and the sinus horns. As the myocardium of the heart tube matures, Nkx2-5 was found to suppress Hcn4 and the gene for T-box transcription factor Tbx3, thereby enforcing a progressive confinement of the expression of these genes to the forming Nkx2-5-negative SAN and sinus horns (see Chapter 9.1). Thus, Nkx2-5 is essential for establishing a gene expression border between the Nkx25-expressing atrium and the Nkx2-5-negative SAN/venous return, providing a mechanism for how pacemaker activity becomes progressively relegated to the most recently added components of the venous pole of the heart (Fig. 4).

Early in development, the entire sinus venosus acts as pacemaker, but at some point during development pacemaker activity must become confined to the actual SAN structure. Somewhere after E10.5, the sinus horns, which will give rise to the sinus venarum and myocardial sleeves of the right and left caval veins (right caval vein and coronary sinus, respectively, in human), initiate Cx43 expression, and after E12.5–13.5 also initiate Cx40 expression. Between E14.5 and birth, Hcn4 expression is being lost from the sinus horns. Thus, the sinus venosus obtains an atrial working myocardial phenotype. The SAN, however, retains its original Hcn4-positive, Cx40/Cx43-negative signature. Thus, before birth, the sinus venosus gene program becomes restricted to the SAN. Tbx3 is the only gene we currently know to be expressed selectively in the SAN anlage from E10.5 onwards (Hoogaars et al., 2004; Mommersteeg et al., 2007) (Figs 3; 4; see Chapter 9.4). The SAN region runs a gene expression program that is distinct from that of the bordering atrial cells. We found lineage segregation of Tbx3-negative atrial and Tbx3-positive SAN precursor cells as soon as cardiac cells turn on the atrial gene expression program, indicating that the SAN is formed from an early specified Tbx3-positive precursor population (Hoogaars et al., 2007). Tbx3-deficiency results in expansion of expression of the atrial gene program into the SAN domain, and partial loss of SAN-specific gene expression. Ectopic expression of Tbx3 in the atria of mice revealed that Tbx3 represses the atrial genes, activates SAN genes, and induces the development of functional ectopic pacemakers. These findings have revealed a Tbx3-dependent pathway for the specification and formation of the SAN, and show that Tbx3 regulates the pacemaker gene expression program and phenotype (Hoogaars et al., 2007). The paired, related homeodomain transcription factor Shox2 is expressed in the sinus venosus (sinus horns, SAN,

oft

PFN caudalHF

ev

PFN rv

1st HF

lv ift

ra

la

AVB sh

2nd HF

Tbx18+

AVN SAN

E7.5

E7.5-8

E8-8.5

E8.5-10.5

Figure 5  Schematic overview of the different cardiac progenitor populations. The embryonic ventricle (ev), the future left ventricle (lv), is mainly derived from the first heart field (HF) (red), whereas the outflow tract (oft), right ventricle (rv) and atria (left atrium: la; right atrium: ra) are derived from the second heart field (green). The Tbx18 caudal heart field (yellow) forms the sinus horns (sh) of the venous pole of the developing heart. The Purkinje fiber network (PFN) derives from the embryonic ventricular walls, the atrioventricular bundle (AVB) from the interventricular region, the atrioventricular node (AVN) from the ventral (caudal) AV canal and the sinoatrial node (SAN) from the cells of the right sinus horn derived from second and caudal heart field precursors.

Chapter  |  2.3  Patterning and Development of the Conduction System of the Heart

venous valves). Shox2 deficiency results in hypoplastic sinus venosus and upregulation of Nkx2-5 and Cx43 in the presumptive SAN anlage (Blaschke et al., 2007), indicating that it acts upstream in the specification pathway of the sinus venosus and SAN. The mechanism of action of Shox2 remains to be defined. Interestingly, both Shox2 and Tbx3 expression require Tbx5 (Mori et al., 2006), which in turn is broadly expressed in all atrial and venous myocardium, including the SAN. The transcription factor Pitx2 provides asymmetric morphogenesis to the heart (Franco and Campione, 2003; chapters in Part 4). Pitx2c-deficient fetuses form SANs at both the right and left sinoatrial junction which have indistinguishable molecular signatures, including Tbx3 expression. These observations indicate that Pitx2c functions within the left–right pathway to suppress a default program for SAN formation on the left (Fig. 4) (Mommersteeg et al., 2007). Taken together, Tbx5, Nkx2-5, Tbx3, Shox2 and Pitx2c are involved in the localization of the SAN gene program and activity to the sinus venosus, and ultimately to the SAN at the junction of the right atrium and superior caval vein.

V.  Chamber differentiation, atrioventricular canal specification and the formation of the atrioventricular node The initial heart tube consists of precursors for the left ventricle and for the AV canal, the latter constituting the inflow tract at early tubular stages (Fig. 5) (Davis et al., 2001b; Buckingham et al., 2005). The tube myocardium has an “embryonic” phenotype, resembling the nodal phenotype in that it displays automaticity, poor contraction and slow transmission of the electrical impulse (conduction) (Moorman and Christoffels, 2003b). For example, the heart tube expresses Cx45, a relatively low conductance (30 pS) gap junction, and the sarcomeres and sarcoplasmic reticulum are not (yet) well-developed (de Haan, 1961; Canale et al., 1986; Alcolea et al., 1999; Moorman and Christoffels, 2003b). During further elongation of the heart tube, the ventricular and atrial chambers differentiate. Concomitant with its differentiation, several genes are activated that provide a more working myocardial phenotype to the chamber myocardium (Fig. 6). These include Cx40 and Cx43, encoding high conductance (180 and 115 pS, respectively (Kreuzberg et al., 2006b)) gap junctions that contribute to the acquisition of rapid transmission of electrical impulse typical for working myocardium (Moorman and Christoffels, 2003b). However, the myocardium of the sinus horns, of the AV canal, inner curvatures and of the outflow tract initially retains its original embryonic phenotype (Figs 1; 2). The result of this development is that the initially sluggish peristaltic contraction of the embryonic heart tube is converted to a pattern in which the atrial and

179

ventricular compartment(s) contract rapidly and, owing to the delay of impulse in the AV canal, serially (Fig. 2A–C). With the exception of some studies suggesting that in chicks the AV node may form in the dorsal part of the atrial septum (Arguëllo et al., 1988), the majority of morphological, functional and expression studies support the view that the AV canal contains the precursors of the AV node, of the AV ring bundle(s), of the support of the AV valves and of the lower rim of the atrium (Fig. 1) (Virágh and Challice, 1977b, 1982; Wessels et al., 1992; Davis et al., 2001b). Therefore, chamber differentiation and local prevention of differentiation are key events in the local formation and function of the AV node. It has not been established when the AV canal emerges from the heart fields. However, given the position of the AV canal in between the ventricles and atria, and given the observation that the atria emerge later than the ventricles, it is attractive to speculate that the AV canal emerges slightly later than the left ventricle. Nevertheless, lacZ expression driven by a chick Gata6 enhancer that is selectively active in the AV canal was found to mark a population of cells in the cardiac crescent, corresponding to the first heart field (Davis et al., 2001a). This finding indicates that AV canal precursors may be derived from the first heart field. During early stages of heart development, Tbx2 is expressed in part of the cardiac crescent, and subsequently marks the AV canal (Christoffels et al., 2004b; Harrelson et al., 2004). When we marked all cells that express or once expressed Tbx2, using Cre-loxP lineage analysis tools, we observed very selective recombination of the AV canal from the earliest stages onwards, again indicating a lineage relationship between the cardiac crescent and the AV canal. Taken together, the AV canal precursors are incorporated in the heart tube at early stages, possibly only slightly later than the left ventricular precursors, and may derive from the first heart field. There is considerable insight into the molecular mechanism underlying AV canal specification and formation. One of the earliest events is the expression of Bmp2 in the embryonic heart tube in the precursors of the AV canal. Bmp2 is necessary and, probably, sufficient to activate Tbx2 (Yamada et al., 2000; Ma et al., 2005). Tbx2, in turn, is required to inhibit chamber differentiation of AV canal precursors, and to suppress chamber-specific genes, which include the genuine T-box target genes Nppa, Cx40 and Cx43 (Habets et al., 2002; Christoffels et al., 2004b; Harrelson et al., 2004). Slightly later Tbx3 is also activated in the atrioventricular canal, in a subdomain of the expression domain of Tbx2 (Fig. 2D) (Hoogaars et al., 2004). Tbx2 and Tbx3 are functionally-equivalent with respect to their ability to inhibit chamber differentiation and target gene regulation (Habets et al., 2002; Lingbeek et al., 2002; Hoogaars et al., 2004). Indeed, a large fraction of Tbx2-deficient embryos form relatively normal AV canals, whereas Tbx3-deficient embryos only display

180

PART  |  2  Cardiac Precursor Populations and Lineages

(A)

(B) Nppa Cx40 Cx43

Tbx2/3

CS

(C) Nppa Cx40 Cx43

Tbx2/3

WM specification

Tbx2/3

Nppa Cx40 Cx43

CS WM specification and recruitment to WM

CS WM recruitment to CS

(D) A-P, D-V patterning, Tbx5, Nkx2-5, Mef2, Hand, Gata, and others

primary myocardium

Tbx2/3 Nkx2-5

ift, avc, oft primary ring

san, avn Tbx3

Tbx2, Tbx3

proximal bb, avb, avj

*

Cx40, Tbx3

NG ventricular

trabecular

Cx40, Cx43, ANF

ANF, Cx40, Cx43

compact atrial

ANF, Cx40, Cx43

Cx43

CS

ET distal bb, pvcs ANF, Cx40, Cx43

ventricular WM Cx43

atrial WM

WM

ANF, Cx40, Cx43

Figure 6  Lineage relationships and mechanisms of conduction system development. (A–C) Schemes showing three mechanisms of conduction system specification: (A) Growth of a pool of specified precursors (specification); (B) Recruitment of myocardial cells into the conduction system lineage (recruitment); (C) Recruitment of conduction system cells into the working myocardial lineage (specification and recruitment). (D) Scheme showing the developmental relationships between components of the heart. Based on data reviewed in Christoffels et al. (2004a). Many transcription factors are required to form and pattern the primary myocardium and, subsequently, to effect the localized formation of ventricles (cranioventral) and atria (dorsocaudal). A subpopulation of the primary myocardium retains its phenotype by the action of Tbx2/3/Nkx2-5. Cells of this pool may subsequently differentiate to chamber myocardium (*). This happens, for example, during the ventricularization of the outflow tract. We hypothesize that chamber myocardium is not allowed to dedifferentiate to primary myocardium (red cross). The nodal components of the conduction system are formed from the primary myocardium, whereas the working myocardial chambers and ventricular conduction system are formed from the early chamber myocardium. Marker genes expressed in each of the components are indicated (avb: atrioventricular bundle; A–P: antero–posterior; D–V: dorso–ventral; avc: atrioventricular canal; avj: atrioventricular junction; avn: atrioventricular node; bb: bundle branches; cs: conduction system; ift: inflow tract; oft: outflow tract; pvcs: peripheral ventricular conduction system (Purkinje); san: sinoatrial node; wm: working myocardium).

cardiac defects at sites where Tbx2 is not expressed (Harrelson et al., 2004; unpublished observations). Tbx20 is required for heart tube elongation and chamber development (Cai et al., 2005; Singh et al., 2005; Stennard et al., 2005; Takeuchi et al., 2005). Importantly, Tbx20 confines Tbx2 expression to the AV canal and outflow tract, as loss of Tbx20 leads to widespread ectopic expression of Tbx2 in the entire heart tube. The phenotype of Tbx20 mutants is reminiscent of that of embryos in which Tbx2 is ectopically expressed in the entire primitive heart tube (Christoffels et al., 2004b) suggesting that failure to suppress Tbx2 may underlie part of the cardiac phenotype in Tbx20deficient embryos. In chicks, a Notch-Hey-Bmp2-Tbx2 signaling cascade was proposed to delimit the AV canal domain (Rutenberg et al., 2006). In forming chamber myocardium, Notch2 activates the expression of Hey1 (Hesr1),

and Hey1 and Hey2 (Hesr2) in turn repress Bmp2. In contrast, Bmp2-activated Tbx2 represses Hey1 and Hey2 in the AV canal. This feedback loop enforcing repression of Notch signaling is thought to sharpen the chamber–AV border. In mouse, a similar pathway has been revealed (Kokubo et al., 2005). In a Notch-independent pathway, Tbx2 and Bmp2 are repressed by Hey2 in the ventricle and Tbx2 by Hey1 in the atrium, thereby confining the expression of Bmp2 and Tbx2 to the AV canal. In zebrafish, notch1b was reported to be required for formation of AV canal function (Milan et al., 2006). However, the data provided are also consistent with developmental arrest of the chambers in morpholino-treated embryos, the affected heart tubes resembling those of 12-hour-younger controls, in which uniform conduction velocity and absence of druginducible AV block is observed. In mammals, neuregulin

Chapter  |  2.3  Patterning and Development of the Conduction System of the Heart

signaling controls the formation of the trabecules and, hence, the ventricular wall (see below). In zebrafish, neuregulin was found to be expressed in the endocardium, with accentuated expression in the AV canal region (Milan et al., 2006). Morpholino-treated embryos failed to develop fast conduction in the chambers (consistent with a function in chamber development as seen in mammals), and furthermore did not develop terfenadine-induced AV block, indicating a role of neuregulin in AV canal function. Msx2 (Drosophila muscle segment homeobox related) expression marks the AV canal in chick and mouse (ChanThomas et al., 1993; Abdelwahid et al., 2001). Its function and ability to form a complex with Tbx2 and Tbx3 (unpublished observations), render Msx2 a possible component of the regulatory pathway controlling AV canal formation. However, mice that are deficient in Msx2 do not display conduction defects (Jay et al., 2005). Nkx2-5 and Tbx5 are broadly expressed in the embryonic heart, although Tbx5 displays a more restricted caudo–cranial graded pattern of expression (Lints et al., 1993; Bruneau et al., 1999). Both factors are crucial activators of chamber-specific genes, and are essential for chamber differentiation (Lyons et al., 1995; Tanaka et al., 1999; Bruneau et al., 2001). Nkx2-5 and Tbx5 are also expressed in the AV canal, where they play important roles. Nkx25 haploinsufficiency results in atrial septal and AV conduction defects (Schott et al., 1998; Benson et al., 1999). Moreover, in mice with only one functional Nkx2-5 allele, the Cx40-negative AV node is strongly hypoplastic (Jay et al., 2004). Tbx5 haploinsufficiency leads to the HoltOram syndrome in humans (Basson et al., 1997; Li et al., 2000). These patients often display septal defects and AV conduction defects (AV conduction block). Mice with only one functional copy of Tbx5 display postnatal failure in maturation of the AV canal component, as visualized by persistent minK-lacZ expression, a marker for this component of the conduction system (Moskowitz et al., 2004). The mechanism that leads from Nkx2-5 or Tbx5 haploinsufficiency to these AV conduction defects is unclear. While Nkx2-5 cooperates with Tbx5 to form an activation complex, Nkx2-5 also cooperates with Tbx2 and Tbx3 to form a repression complex (Bruneau et al., 2001; Hiroi et al., 2001; Habets et al., 2002; Hoogaars et al., 2004). Moreover, Tbx5 and Tbx2/3 compete for DNA binding to target genes. It is therefore attractive to speculate that mutation of Nkx2-5 or Tbx5 causes imbalance between these complexes specifically in the AV canal, providing a possible explanation for the localized defects caused by somatic mutation of these broadly expressed factors. As discussed above, the AV canal retains the slow conducting property of the embryonic myocardium, and now functions as a “delay generator” between the fast atrial and ventricular chamber(s). In this respect, the AV canal can be regarded as a leftover product of the process of chamber differentiation. Indeed, the expression profile of a number

181

of important genes is maintained during maturation of the AV node (Table 1). Nevertheless, the AV canal myocardium also changes during development, as the mature AV node expresses several genes not expressed in the embryonic myocardium. Further analysis in zebrafish embryos revealed not only a several-fold increase in conduction velocity in the atria and ventricles during development as expected, but also a decrease in conduction velocity in the AV canal (Milan et al., 2006). We conclude that a transcriptional regulatory mechanism is active that inhibits the AV canal cells to differentiate into chamber myocardium, allowing these cells to further differentiate into mature nodes. In the embryo, the entire AV canal conducts slowly, thus functioning as an AV node equivalent. The electrical signal can travel from the atria to the ventricles (and back) through the ring of AV myocardium connecting these chambers, with the ventral and dorsal aspects of the AV canal being preferential sites of conduction (de Jong et al., 1992; Rentschler et al., 2002; Valderrabano et al., 2006). After septation, fibrous tissue will form which physically separates and insulates the atria and ventricles. The only connection that remains is the AV bundle, which connects the AV node at the atrial side of the fibrous body with the ventricles. The mechanism that underlies the formation of this separation and the morphogenesis of the AV node from the AV canal has not been established in detail. Late in development, expression of Tbx3, minK-lacZ and other markers disappears from most of the AV canal, and becomes confined to the AV node (Kondo et al., 2003; Hoogaars et al., 2004; Moskowitz et al., 2004), indicating the disappearance of AV myocardium, e.g., by apoptosis or differentiation into working myocardium associated with loss of expression. However, remnants of AV canal myocardium are found surrounding the AV valves. This myocardium retains an embryonic and “nodal” phenotype. In the case of SAN dysfunction or loss of atrial conductivity, ectopic beats may originate from this AV myocardium (Morton et al., 2001; Kistler et al., 2003; Bagwe et al., 2005). This is more prevalent from the right (tricuspid valve) side than the left, correlating well with a more pronounced AV canal structure at the right (right AV ring bundle) (Wessels et al., 1992) compared to the left side.

VI.  Internodal tracts and outflow tract Several decades ago, the notion of internodal tracts was in vogue (Janse and Anderson, 1974). Conduction system marker HNK-1 readily recognized several strands in the dorsal atrial wall between the SAN and AV node, indicative of their existence (Ikeda et al., 1990; Blom et al., 1999). Arguments in favor or against these strands being fastconducting tracts between the nodes were put forward, but

182

proof for their existence or function was not forthcoming. The necessity of these tracts being fast-conducting is questionable, as the atrial myocardium conducts very rapidly, and has preferential conduction paths. From an embryological point of view, these tracts may be the remnants of the primary myocardial inflow tract and caudal AV region, from which the SA and AV node also develop. While atrial chamber myocardium differentiates, this myocardium does not take part in this differentiation, but continues to run the original “nodal” gene program. Indeed, the venous side of the left and right venous valves (common inflow, including the future orifice of the coronary sinus in human) and the dorsal aspect of the atrial septum express nodal marker Tbx3 and do not express Cx40 or Cx43 (Hoogaars et al., 2004; Soufan et al., 2004) (Fig. 3). Therefore, we would anticipate that the internodal tracts are slow-conducting remnants of the original primary myocardium. The outflow tract in the embryo, which forms between E8.5 and 10.5 from the second heart field (Fig. 5), has a primitive phenotype. Like the AV canal, it conducts the electrical impulse slowly, and its cells have poorly-developed sarcoplasmatic reticular and sarcomeric structures, hallmarks of embryonic or primary myocardium. Several “conduction system” markers, such as minK-lacZ and Cx45, are expressed in the outflow tract, which has led to the notion that the outflow tract is part of an extended conduction system (Coppen et al., 1999; Kondo et al., 2003). Nevertheless, the outflow tract myocardium is a transient structure; it disappears because its cells differentiate into (right) ventricular working myocardium (Rana et al., 2007), thereby developing working myocardial properties such as fast-conduction, well-organized myofibrils/sarcomeres, etc. A small portion of the cells are removed by apoptosis (Watanabe et al., 2001; Barbosky et al., 2006; Rana et al., 2007). Interestingly, some patients manifest with tachycardia that arises from the right ventricular outflow tract, suggesting that remnants of this embryonic outflow tract myocardium may still be present in these patients (Timmermans et al., 2003).

VI.A.  The Formation of the Atrioventricular Bundle and Proximal Bundle Branches The ventricular conduction system consists of the AV bundle (bundle of His), the bundle branches and the network of Purkinje fibers. A key feature shared by these components is the rapid conduction of electrical impulses from the AV node to the ventricular working myocardium of the apex and the remainder of the ventricle. Their myocardial cells are well-coupled by gap junctions, and in mouse and chick Cx40 serve as a very useful and specific marker that distinguishes the fast-conducting components (AV bundle, bundle branches, Purkinje fiber network) from the working myocytes of the ventricle (Delorme et al., 1995; Gourdie

PART  |  2  Cardiac Precursor Populations and Lineages

et al., 1995; Gros et al., 1995; Miquerol et al., 2004). A variety of origins of the AV bundle have been proposed that range from outgrowth of AV node precursor cells, separate formation to join later with the AV node, formation in the atrial septum, to formation from the ventricular septum (reviewed in Virágh and Challice, 1977b; Arguëllo et al., 1988). In mouse, the AV bundle cells were first observed at E10–11 in the top (crest) of the forming ventricular septum. Careful morphogenetic analysis indicated that these cells give rise to the definitive AV bundle (Virágh and Challice, 1977a,b). With further development of the ventricular septum, left and right bundle branches bifurcating from the AV bundle were seen to develop directly from the subendocardial myocytes. The AV node and ventricular conduction components have been suggested to connect subsequent to their formation (Chuck and Watanabe, 1997). However, from the outset the AV node primordium (AV canal) is in direct contact with the fast-conducting Cx40-positive trabecules of the ventricles (Purkinje precursors) (see Fig. 2D,E) (Virágh and Challice, 1977a, 1982) and with the crest of the forming interventricular septum, including the primordial AV bundle and bundle branches, which at a later stage will initiate expression of Cx40. The second heart field contributes to the heart tube, including the ventricular region (Fig. 5) (Buckingham et al., 2005). Depending on which marker one uses, the ventricular septum is partly (Kelly et al., 2001; Franco et al., 2006), or even largely (Verzi et al., 2005), derived from the second heart field that also gives rise to the outflow tract and right ventricular cells. Therefore, the AV bundle on top of the ventricular septum forms from precursors of the border zone of the first and second heart fields. Gln2, Leu-7 and HNK-1 share the same spatio–temporal pattern, including the cells of the AV node and AV bundle from early stages onward, e.g., from the moment the first signs of formation of the ventricular septum are visible (4–5 weeks human, rat E11). The pattern shows the primary or ventricular ring (Fig. 1) and the subendocardial myocardial cells draping over the ventricular septum (Ikeda et al., 1990; Wessels et al., 1992; Nakagawa et al., 1993; Blom et al., 1999). The myocardial cells of the primary ring include the inner curvature where the AV canal and the outflow tract are connected, the AV node primordium, the crest of the ventricular septum (AV bundle primordium), the future right AV ring bundle and the retroaortic root branch (Wessels et al., 1992). Whether the expressing cells are indeed conduction system cells in all species is unclear. For example, in fetal chick, HNK-1 does not detect AV bundle cells, but the cells in close association with the AV bundle (Chuck and Watanabe, 1997). Tbx3 is specifically expressed in the AV canal and the ventricular septum (Figs 2D; 5). This pattern includes the primary ring. It is already expressed shortly after looping at E9.5 (Hoogaars et al., 2004), which corresponds to the expression of the Leu-7/HNK-1 epitope at E11 in rat

Chapter  |  2.3  Patterning and Development of the Conduction System of the Heart

(Ito et al., 1992; Aoyama et al., 1993), thus providing a very early marker for the AV bundle. With formation of the ventricular septum, its pattern also includes the proximal bundle branches. From approximately E10 onward, Cx43 expression is initiated in the ventricular chamber myocardium, but not in the Tbx3-positive area including the AV bundle (Fig. 3B). The Tbx3-positive, Cx43-negative proximal bundle branches become visible after E10.5, when development of the ventricular septum is well underway (Fig. 3B). In the mature heart, the AV bundle and proximal bundle branches still run a Tbx3-positive, Cx43negative gene program. Therefore, Tbx3 and Cx43 (by its specific absence) represent excellent markers to discriminate these conduction system structures from the earliest moment of their formation onward. In chick, it was found that the Msx2 pattern is equivalent to the Gln2 pattern in human (Chan-Thomas et al., 1993), thus indicating that the mechanism of mammalian and chick AV conduction system development is similar. Also, in mouse Msx2 is expressed in the AV canal, but the expression in the AV bundle is very weak (Abdelwahid et al., 2001) (our unpublished observations). If anything, the expression patterns demonstrate that these conduction system components become distinct and recognizable at early stages of heart development, i.e., shortly after looping, during the differentiation of the ventricular chambers. Moreover, these patterns indicate that the AV bundle and proximal bundle branches initially have a primitive, embryonic phenotype very similar to that of AV canal-derived components, which is in line with the results of the PAS stainings performed by Virágh and Chalice (1977a, 1982). Although direct lineage relationships between the cells of these early conduction system components and the definitive counterparts have not been provided, the spatio–temporal pattern of expression of the various marker genes suggests that this relationship exists. Whether progeny of initially Cx43-positive, Tbx3/ Gln2-negative cells contribute to the conduction system, or whether progeny of Cx43-negative, Tbx3/Gln2-positive cells contribute to the working myocardium remains to be established (also see below). Initially, the AV bundle does not express Cx40, suggesting that the preferential pathway for the impulse is from the AV canal directly through the connected trabecules that do express Cx40 and Cx43 from the onset of their formation at E9.5 onwards (see Fig. 2D,E). This pathway has been suggested based on morphology (Virágh and Chalice, 1977a) and indeed, in the mouse, fast ventricular activation was observed at E9.5, starting from the slow-conducting dorsal AV canal wall (AV node anlage) (Rentschler et al., 2002; Tallini et al., 2006; Valderrabano et al., 2006). Only after E14.5, when septation has been completed is Cx40 seen to be upregulated in the Cx43-negative AV bundle, its expression domain reaching the AV node by the time of birth. By that time the atria and ventricles have become electrically insulated by formation of a connective tissue

183

layer, the now Cx40-positive AV bundle is the only fast myocardial connection between the AV node and the ventricles. Interestingly, this spatio–temporal pattern of Cx40 is conserved between murine and chick (Gourdie et al., 1995; Gros et al., 1995). Analysis of mice heterozygous for a mutation in Nkx2-5 or Tbx5 revealed their requirement for this process (Jay et al., 2004; Moskowitz et al., 2004). Analysis of Cx40 expression and the pattern of the minK-lacZ allele in Nkx2-5 heterozygous mutants revealed a hypoplastic AV bundle and bundle branches. Mice heterozygous for a mutation in Tbx5 develop postnatal defects of the AV bundle and left bundle branches, whereas the right bundle branch is severely affected or completely absent at birth. These results indicate loss, or shortage, of AV conduction cells. Moreover, compound haploinsufficiency of Nkx25 and Tbx5 prevents the specification or maintenance of an AV bundle (Moskowitz et al., 2007). The expression of transcriptional repressor Id2 is expressed in the AV bundle, and required for its formation. Id2 is cooperatively activated by Tbx5 and Nkx2-5. Compound haploinsufficiency of Id2 and Tbx5 also prevents the specification or maintenance of an AV bundle. These findings have established an Nkx2-5/Tbx5/Id2-dependent pathway for the development of the AV bundle (Moskowitz et al., 2007). Tbx3, which is expressed in the AV bundle (Fig. 3), is a repressor of Cx43 and Cx40 (Hoogaars et al., 2004), competes with Tbx5, and cooperates with Nkx2-5. Preliminary data from our laboratory have revealed that Tbx3-deficient embryos fail to specify the AV bundle, resulting in ectopic expression of Cx43 and precocious activation of Cx40 in the AV bundle anlage.

VI.B.  The Peripheral Ventricular Conduction System: Distal Bundle Branches and Purkinje Fiber Network In the mature heart, the distal bundle branches and Purkinje fibers are only one to a few cells thick, and are located directly below the endocardium (Canale et al., 1986; Moorman et al., 1997). In chick, an additional periarterial network of Purkinje-like cells is present (Lamers et al., 1991; Gourdie et al., 1995). Together, these components form the peripheral ventricular conduction system. An important function of this system is to transmit the electrical impulse rapidly from the AV bundle and proximal bundle branches to the ventricular working myocardium, in such a way that the signal is distributed over all regions of the right and left ventricular walls for synchronized contraction, and that the apex is activated first. As the name for this system already indicates, the cells of the distal branches and Purkinje network are well-coupled, and express high levels of Cx40 and other genes that promote rapid impulse transmission (Fig. 6) (Moorman et al., 1998).

184

VI.B.i.  How Does the Purkinje Fiber Network Develop? When the embryonic ventricular wall is just a few cells thick and starts to form trabecules, it initiates the expression of a gene program that is required for rapid conduction, including Cx40 and Cx43 (Moorman and Christoffels, 2003b) (Figs 2D,E; 7). Embryonic mammalian hearts display fast ventricular conduction at stages when trabecules have been formed (de Jong et al., 1992; Chuck et al., 1997; Rentschler et al., 2001; Tallini et al., 2006; Valderrabano et al., 2006). In hearts of lower vertebrates, which do not possess a Purkinje network, the ventricle is also activated from apex to base, indicating that the substrate for preferential conduction towards the apex is already present (Subramaniam et al., 1991; Sedmera et al., 2003b). The spongy trabecular myocardium of the fish heart is remarkably similar to that of mammalian and chick embryos. These observations indicate that the trabecules are the functional and structural precursors of the Purkinje fiber network. With further development, Cx40/Nppa activity becomes gradually restricted to the trabecules, while a Cx40/Nppa-low compact layer differentiates and expands at the epicardial side (Fig. 7).

(A) trab. cm

trab.

cm

trab.

trab. ecv

9.5

14.5

17.5

high proliferation rate

(B)

11.5

Cx40 EAP-300 low clone

Other than that, these cells are more primitive than the working myocardium, in that they have less well-developed contractile machinery, in that respect resembling the embryonic myocytes (de Haan, 1961; Canale et al., 1986). Relevant questions concerning the origin of this system are: where do these cells come from, when are they formed and how are they specified? The initial heart tube is only one or a few myocytes thick. As soon as the ventricular chamber begins to develop, the subendocardial cells will form trabecules and the proliferative activity in these cells decreases. In contrast, the outer layer will continuously proliferate, which requires signals from the epicardium (Smith and Bader, 2007). The progeny of these cells will contribute to the elongating trabecules, and later will form the compact working myocardium. Mikawa and co-workers demonstrated that marked single cells of the cardiogenic mesoderm form cone- or wedge-shaped clones which transmurally span the thickness of the ventricular wall, thus showing that the trabecules and compact working myocardium form by local proliferation and limited horizontal migration of progeny of a myocardial progenitor cell (Mikawa et al., 1996). The myocytes of the trabecules adjacent to the endocardium will form the Purkinje fibers. These trabecules are derived from the first layers of ventricular cells to be formed. The embryonic left ventricle is formed from the first heart field precursors, while the right ventricle forms from the second heart field (Buckingham et al., 2005). Therefore, the Purkinje fiber network in the left ventricle is derived from the first heart field, whereas the network in the right ventricle is derived from the second heart field (Fig. 5).

PART  |  2  Cardiac Precursor Populations and Lineages

∗ NG mouse 9.5 chick E3

ET-1 12.5

17.5 E14-18

Figure 7  Development of the Purkinje fiber network. (A) Expression of Cx40 in the ventricles from E9.5 to E17.5. Cx40 is expressed transmurally. After E11.5 the Cx40-negative compact layer forms and expands at the epicardial side. Note that between E11.5 and E17.5 the trabecular component is similar in thickness. After birth, a further maturation step takes place, remodeling the trabecular zone into the definitive Purkinje fiber network. (B) Schematic showing model for the Purkinje fiber network development. Expression of Cx40/EAP-300 is depicted in a gradient from white (no expression) to black (high expression). Neuregulin (NG) is important for the formation of the trabecules and Endothelin-1 (ET-1) plays a role in later maturation/induction of the Purkinje fiber network. “Clone” depicts a putative clone derived from a progenitor cell marked at E3 (*), that contains both Purkinje cells and working myocardial cells (cm: compact myocardium; ecv: endothelial cells of the coronary vessels; trab: trabecular myocardium).

The compact zone rapidly increases while the trabecular, Cx40-positive zone remains more or less constant (Fig. 7). After birth, a further maturation step takes place, remodeling the trabecular zone into the definitive Purkinje fiber network (Eralp et al., 2006; Meysen et al., 2006). Neuregulin signaling is important for the formation of trabecules (Figs 6; 7). Mice deficient in neuregulin, produced by endocardium or the cognate receptors erbB2 or erbB4, expressed in the myocardium, fail to form trabecules and to expand the ventricular chambers (Gassmann et al., 1995; Lee et al., 1995; Meyer and Birchmeier, 1995). Neuregulin-1 treatment of cultured embryonic mice induced ventricular trabeculation and upregulation of expression of acetylcholine esterase, a marker for trabecules (Hertig et al., 1999). Moreover, the expression of lacZ from the CCS-lacZ line, which marks the trabecules in the embryo and the Purkinje fibers in the adult, was found to be induced by neuregulin in embryo cultures (Rentschler et al., 2002). These results indicate that neuregulin signaling regulates the relative proportion of the embryonic ventricular wall that forms trabecular

Chapter  |  2.3  Patterning and Development of the Conduction System of the Heart

myocardium or compact myocardium, implicating its role in Purkinje fiber formation (Hertig et al., 1999). During trabecule formation, Notch signaling was found to independently regulate ventricular cardiomyocyte proliferation and differentiation (Grego-Bessa et al., 2007). Notch signaling regulates Bmp10 expression and independently regulates ephrinB2 and neuregulin signaling involved in the formation of the trabecules. Bmp10, in turn, is involved in ventricular maturation by repressing the negative cell-cycle regulator p57kip2 and promoting the expression of cardiogenic factors Nkx2-5 and Mef2c (Chen et al., 2004). Not much is known about the late steps in the process of Purkinje fiber specification and formation. Adult mice heterozygous for Nkx2-5 display hypoplasia of Purkinje fibers, and disorganization of the network in the apex of the ventricles (Meysen et al., 2006). The deficiency in Cx40 (EGFP)-positive Purkinje cells was observed only after birth. The underlying mechanism has not been revealed, and may involve postnatal loss of Purkinje cells (e.g., a reduction of ventricular cells that maintain the Purkinje fiber phenotype), or reduced recruitment to the Purkinje fiber phenotype. Initial studies to unravel the lineage origin and molecular mechanism of the formation of the conduction system focused on the periarterial component uniquely present in chicks (Gourdie et al., 2003b). Clones derived from single cells retrovirally-labeled at E3 (HH17) in the right ventricular free wall were found to contain both periarterial Purkinje cells and ventricular working myocytes at E14, and the proportion of clones containing periarterial Purkinje cells increased thereafter (Gourdie et al., 1995). These and additional studies indicated that the periarterial ventricular conduction system progressively forms, grows and ramifies between E8 and E18 by recruitment of ventricular cells adjacent to the newly-forming arterial branches of the coronary vasculature (Hyer et al., 1999; Gourdie et al., 2003a). The restriction of Purkinje fiber formation adjacent to arteries, and not veins, suggested a role for hemodynamic factors from the vasculature involved in the recruitment process. Exposure of cultured embryonic myocytes to endothelin-1 (ET-1), a shear-stress responsive cytokine, caused the induction of a Purkinje fiber program (upregulation of Cx40, EAP-300, slow twitch muscle myosin heavy chain, atrial myosin heavy chain and downregulation of cardiac myosin binding protein-C) (Gourdie et al., 1998; Takebayashi-Suzuki et al., 2000). In a number of successive studies this process was further explored, and key aspects of the underlying mechanism were revealed. Active ET-1 is converted from an inactive precursor by ET1 converting enzyme (ECE-1). Both ET-1 and ECE-1 were shown to be required for the induction of the Purkinje phenotype in ventricular cells in vivo (Takebayashi-Suzuki et al., 2000; Hall et al., 2004). ECE-1 is preferentially expressed in endocardial cells and endothelial cells of the coronary arteries at the time of Purkinje fiber formation,

185

providing a mechanism for the timing and location of Purkinje fiber formation. The sequence of activation of the ventricles shifts from a unidirectional pattern to a mature apex-to-base pattern around E7–9 (HH29–35) (Chuck et al., 1997; Reckova et al., 2003; Hall et al., 2004). This change has been attributed to the further development of the His–Purkinje system (Chuck et al., 1997; Reckova et al., 2003). Changes in hemodynamic load were found to alter the timing of this shift, indicating that it affects the development of the ventricular conduction system. ECE-1 expression and Cx40 were also found to be regulated by hemodynamic load, correlating with the changes in timing of the shift to the mature apex-to-base pattern (Hall et al., 2004). These data suggest that biomechanical forces (shear stress/pressure), acting through stretch/pressure-induced ET and ECE-1, play a role in the development of the ventricular conduction system in chicks. However, in mouse a functional Purkinje system is present well before the onset of high pressure circulation, suggesting hemodynamic factors are not required for murine His–Purkinje induction (Rentschler et al., 2001).

VII.  Formation of the conduction system components by recruitment or by early specification and outgrowth Despite considerable insight into the origin of the distinct components of the heart and the morphogenesis of the conduction system, important issues remain to be addressed concerning the specification and differentiation of the conduction system. There are basically two views of the formation of the conduction system components. One is that these components are specialized entities that are formed during heart development by recruiting (working) cardiomyocytes to the conduction system lineage, a process that occurs until hatching (“recruitment” model) (Fig. 6B). The other view is that the conduction system precursors are specified early in heart development, grow by (slow) proliferation, and develop further into the conduction system components (“early specification” model) (Fig. 6A). The spatio–temporal expression profiles of specification markers indicate that pacemaking and conduction system precursors have been specified already at heart looping and chamber forming stages (between approximately E8.5 and E11) (see Table 1). The differences between the views, and arguments that would favor recruitment over early specification (referred to as “outgrowth” model), were recently put forward (Mikawa and Hurtado, 2007). The recruitment hypothesis is based on retrospective clonal analyses in chicks, in which clones derived from ­replication-incompetent retrovirus-infected founder cells were analyzed (Gourdie et al., 1995; Cheng et al., 1999).

186

Cell clusters in the central conduction system (right AV ring bundle, AV bundle) or Purkinje fibers that were clonally derived from single cells always contained both conduction and working myocardium, indicating that the working myocardium and the conduction system cells are derived from a common progenitor present at the stage of infection, E3 (HH15–17) (Gourdie et al., 1995; Cheng et al., 1999). Because the conduction system components proliferate much less, recruitment of working myocytes to the conduction system lineage was put forward as the primary mode of conduction system expansion (Sedmera et al., 2003a). The central conduction system (AV ring bundle, AV bundle) would form by recruiting myocytes to an initial framework (e.g., Msx2, Gln2-expressing population) until eight days of development. The Purkinje fibers of the ventricular conduction system would form de novo between eight days and hatching (Gourdie et al., 2003a). However, as discussed above, the ECG derived from HH13 embryos (50 hours of incubation, 20 somites) (Paff et al., 1968; Seidl et al., 1981), implies the early presence of a functional pacemaking and ventricular conduction system. Moreover, morphological and histological analysis of the AV conduction system has indicated that “as the myoblasts begin to differentiate, taking on the characteristics of typical heart muscle, the conductive tissues become progressively more easily distinguishable” (de Haan, 1961), because the myocytes of the atrial and ventricular chambers differentiate faster than the myocytes of the conduction system (de Haan, 1961; Moorman and Christoffels, 2003b). The recruitment model is at odds with these observations, as it implies de novo formation of conduction system components after the observed establishment of the functional and morphologically-recognizable components. Moreover, recruitment to the AV conduction system requires further differentiated atrial or ventricular cells to dedifferentiate when they are recruited to the conduction system. In contrast, the observations are compatible with early specification, in which a subpopulation of embryonic myocytes is designated to slowly differentiate and proliferate to form conduction system components, whereas another subpopulation differentiates and proliferates much faster to form the working myocardium (Moorman and Christoffels, 2003b; Soufan et al., 2006). Going back to the labeling studies in chicks; at the stage myocytes were labeled, E3, they already express specification markers for central conduction system components (e.g., Msx2/Gln2/ Tbx2/Tbx3) or chamber myocardium/Purkinje fibers (e.g., Cx40/Nppb). However, because it was not defined what type of cell was infected at E3 in these lineage studies, they do not discriminate between recruitment of working myocytes to the conduction system lineage, or vice versa. In fact, labeling of the various cell populations that express specification markers will be required to solve this issue. Our genetic labeling studies have indicated that Tbx3negative cells at the atrial side of, and adjacent to, the

PART  |  2  Cardiac Precursor Populations and Lineages

SAN are not recruited to the SAN (Hoogaars et al., 2007). Moreover, when following the fate of cells that express or once expressed primary myocardial marker Tbx2, we observed that Tbx2-expressing myocytes of the E9.5 AV canal escape the nodal (Tbx2-positive) program, allowing them to differentiate to Tbx2-negative ventricular working myocardium (Fig. 6C) (our unpublished observations). This possibly provides an alternative explanation for the findings of Cheng et al. (1999) that single cell-derived clones containing conduction system cells of the AV canal also contain working myocytes. An issue that challenges the outgrowth model is based on the misconception that the primary interventricular ring cells were proposed to expand (migrate) into the ventricles to form the bundle branches and Purkinje fiber network (Mikawa and Hurtado, 2007). This is not compatible with growth and differentiation of the ventricular wall, which occurs within individual myocyte-derived growth columns, involving very limited horizontal migration (Mikawa et al., 1996). However, while the Gln2/Leu-7/Msx2/Tbx3expressing ring was proposed to form the AV node and AV bundle, consistent with its position, it is the trabecules of the ventricles astride the ventricular septum that were proposed to give rise to the bundle branches and Purkinje fibers, consistent with the concept of the local growth columns (Moorman et al., 1998). The expansion of the expression of the “ring markers” into the trabecules merely indicates that the trabecules and the ring myocytes obtain shared phenotypic properties, and not that the ring cells migrate or expand into the ventricles to form the branches and Purkinje fibers. An important observation that supports the recruitment model for Purkinje fiber formation is that both in vitro and in vivo ET-1 signaling induces a Purkinje fiber program (see above), implying de novo formation of Purkinje cells from ventricular cells. However, the Purkinje fiber markers Cx40 and TASK-1 (chick and mouse) and EAP300 (chick) are expressed transmurally in the embryonic ventricular wall before their expression becomes restricted to the trabecules and later the Purkinje fibers (Figs 2E; 7; Table 1) (Delorme et al., 1995; Gros et al., 1995; McCabe et al., 1995; Graham et al., 2006). These observations indicate that while the bundle branches and Purkinje fibers maintain the embryonic ventricular wall phenotype, it is the compact working myocardium that differentiates. Furthermore, chick embryonic ventricles develop rapid propagation of the impulse already after E2 (HH14) (a hallmark of Purkinje fiber network function), which is prior to the proposed Purkinje fiber differentiation between E8–18 (de Jong et al., 1992; Gourdie et al., 2003a). Finally, as is the case for the AV conduction components, Purkinje fibers and embryonic ventricular cells share relatively poorly-developed contractile and sarcoplasmic reticular apparatus, while that of mature working myocytes is well-developed (de Haan, 1961;

Chapter  |  2.3  Patterning and Development of the Conduction System of the Heart

Canale et al., 1986). Together, these observations suggest that the process of Purkinje fiber formation in both mammals and chicks involves differentiation at the epicardial side of Purkinje-like embryonic ventricular cells into compact working myocardium, while at the endocardial side the Purkinje-like phenotype is maintained and further induced. Therefore, Purkinje fiber formation may involve both neuregulin signaling-mediated induction and maintenance of the early trabecular ventricular phenotype, and reinforcement and maturation of the Purkinje phenotype later in development (Figs 6; 7).

References Abdelwahid, E., Rice, D., Pelliniemi, L.J., Jokinen, E., 2001. Overlapping and differential localization of Bmp-2, Bmp-4, Msx-2 and apoptosis in the endocardial cushion and adjacent tissues of the developing mouse heart. Cell Tissue Res. 305, 67–78. Alcolea, S., Theveniau-Ruissy, M., Jarry-Guichard, T., Marics, I., Tzouanacou, E., Chauvin, J.P., Briand, J.P., Moorman, A.F.M., Lamers, W.H., Gros, D.B., 1999. Downregulation of connexin45 gene products during mouse heart development. Circ. Res. 84, 1365–1379. Anderson, R.H., Becker, A.E., Wenink, A.C.G., Janse, M.J., 1976. The development of the cardiac specialized tissue. In: Wellens, H.J.J., Lie, K.I., Janse, M.J. (Eds.). The Conduction System of the Heart. H. E. Stenfert Kroese B.V, Leiden, pp. 3–28. Anderson, R.H., Ho, S.Y., 2003. The morphology of the cardiac conduction system. In: Chadwick, D.J., Goode, J. (Eds.). Development of the Cardiac Conduction System. John Wiley, Chichester, UK, pp. 6–24. Aoyama, N., Kikawada, R., Yamashina, S., 1993. Immunohistochemical study on the development of the rat heart conduction system using anti-Leu-7 antibody. Arch. Histol. Cytol. 56, 303–315. Arguëllo, C., Alanis, J., Valenzuela, B., 1988. The early development of the atrioventricular node and bundle of His in the embryonic chick heart. An electrophysiological and morphological study. Development 102, 623–637. Bagwe, S., Berenfeld, O., Vaidya, D., Morley, G.E., Jalife, J., 2005. Altered right atrial excitation and propagation in connexin40 knockout mice. Circulation 112, 2245–2253. Barbosky, L., Lawrence, D.K., Karunamuni, G., Wikenheiser, J.C., Doughman, Y.Q., Visconti, R.P., Burch, J.B., Watanabe, M., 2006. Apoptosis in the developing mouse heart. Dev. Dyn. 235, 2592–2602. Basson, C.T., Bachinsky, D.R., Lin, R.C., Levi, T., Elkins, J.A., Soults, J., Grayzel, D., Kroumpouzou, E., Traill, T.A., Leblanc-Straceski, J., Renault, B., Kucherlapati, R., Seidman, J.G., Seidman, C.E., 1997. Mutations in human TBX5 cause limb and cardiac malformation in Holt-Oram syndrome. Nat. Genet. 15, 30–35. Benson, D.W., Silberbach, G.M., Kavanaugh-McHugh, A., Cottrill, C., Zhang, Y., Riggs, S., Smalls, O., Johnson, M.C., Watson, M.S., Seidman, C.E., Plowden, J., Kugler, J.D., 1999. Mutations in the cardiac transcription factor NKX2.5 affect diverse cardiac developmental pathways. J. Clin. Invest. 104, 1567–1573. Blaschke, R.J., Hahurij, N.D., Kuijper, S., Just, S., Wisse, L.J., Deissler, K., Maxelon, T., Anastassiadis, K., Spitzer, J., Hardt, S.E., Scholer, H., Feitsma, H., Rottbauer, W., Blum, M., Meijlink, F., Rappold, G.,

187

Gittenberger-de Groot, A.C., 2007. Targeted mutation reveals essential functions of the homeodomain transcription factor Shox2 in sinoatrial and pacemaking development. Circulation 115, 1830–1838. Blom, N.A., Gittenberger-de Groot, A.C., deRuiter, M.C., Poelmann, R.E., Mentink, M.M., Ottenkamp, J., 1999. Development of the cardiac conduction tissue in human embryos using HNK- 1 antigen expression: possible relevance for understanding of abnormal atrial automaticity. Circulation 99, 800–806. Bohn, G., Moosmang, S., Conrad, H., Ludwig, A., Hofmann, F., Klugbauer, N., 2000. Expression of T- and L-type calcium channel mRNA in murine sinoatrial node. FEBS Lett. 481, 73–76. Bruneau, B.G., Logan, M., Davis, N., Levi, T., Tabin, C.J., Seidman, J.G., Seidman, C.E., 1999. Chamber-specific cardiac expression of Tbx5 and heart defects in Holt-Oram syndrome. Dev. Biol. 211, 100–108. Bruneau, B.G., Nemer, G., Schmitt, J.P., Charron, F., Robitaille, L., Caron, S., Conner, D.A., Gessler, M., Nemer, M., Seidman, C.E., Seidman, J.G., 2001. A murine model of Holt-Oram syndrome defines roles of the T-box transcription factor Tbx5 in cardiogenesis and disease. Cell 106, 709–721. Buckingham, M., Meilhac, S., Zaffran, S., 2005. Building the mammalian heart from two sources of myocardial cells. Nat. Rev. Genet. 6, 826–837. Cai, C.L., Zhou, W., Yang, L., Bu, L., Qyang, Y., Zhang, X., Li, X., Rosenfeld, M.G., Chen, J., Evans, S., 2005. T-box genes coordinate regional rates of proliferation and regional specification during cardiogenesis. Development 132, 2475–2487. Canale, E.D., Campbell, G.R., Smolich, J.J., Campbell, J.H., 1986. Cardiac Muscle. Springer-Verlag, Berlin. Chan-Thomas, P.S., Thompson, R.P., Robert, B., Yacoub, M.H., Barton, P.J.R., 1993. Expression of homeobox genes Msx-1 (Hox-7) and Msx-2 (Hox-8) during cardiac development in the chick. Dev. Dyn. 197, 203–216. Chen, H., Shi, S., Acosta, L., Li, W., Lu, J., Bao, S., Chen, Z., Yang, Z., Schneider, M.D., Chien, K.R., Conway, S.J., Yoder, M.C., Haneline, L.S., Franco, D., Shou, W., 2004. BMP10 is essential for maintaining cardiac growth during murine cardiogenesis. Development 131, 2219–2231. Cheng, G., Litchenberg, W.H., Cole, G.J., Mikawa, T., Thompson, R.P., Gourdie, R.G., 1999. Development of the cardiac conduction system involves recruitment within a multipotent cardiomyogenic lineage. Development 126, 5041–5049. Christoffels, V.M., Burch, J.B., Moorman, A.F., 2004a. Architectural plan for the heart: early patterning and delineation of the chambers and the nodes. Trends Cardiovasc. Med. 14, 301–307. Christoffels, V.M., Hoogaars, W.M., Tessari, A., Clout, D.E., Moorman, A.F., Campione, M., 2004b. T-box transcription factor Tbx2 represses differentiation and formation of the cardiac chambers. Dev. Dyn. 229, 763–770. Christoffels, V.M., Mommersteeg, M.T., Trowe, M.O., Prall, O.W., de Gier-de Vries, C., Soufan, A.T., Bussen, M., Schuster-Gossler, K., Harvey, R.P., Moorman, A.F., Kispert, A., 2006. Formation of the venous pole of the heart from an Nkx2-5-negative precursor population requires Tbx18. Circ. Res. 98, 1555–1563. Chuck, E.T., Freeman, D.M., Watanabe, M., Rosenbaum, D.S., 1997. Changing activation sequence in the embryonic chick heart: implications for the development of the His-Purkinje system. Circ. Res. 81, 470–476. Chuck, E.T., Watanabe, M., 1997. Differential expression of PSA-NCAM and HNK-1 Epitopes in the developing cardiac conduction system of the chick. Dev. Dyn. 209, 182–195.

188

Coppen, S.R., Gourdie, R.G., Severs, N.J., 2001. Connexin45 is the first connexin to be expressed in the central conduction system of the mouse heart. Exp. Clin. Cardiol. 6, 17–23. Coppen, S.R., Kaba, R.A., Halliday, D., Dupont, E., Skepper, J.N., Elneil, S., Severs, N.J., 2003. Comparison of connexin expression patterns in the developing mouse heart and human foetal heart. Mol. Cell Biochem. 242, 121–127. Coppen, S.R., Severs, N.J., Gourdie, R.G., 1999. Connexin45 (alpha 6) expression delineates an extended conduction system in the embryonic and mature rodent heart. Dev. Genet. 24, 82–90. Davies, F., 1930. The conducting system of the bird’s heart. J. Anat. 64, 129–146. Davies, M.J., Anderson, R.H., Becker, A.E., 1983. The Conduction System of the Heart. Butterworths, London. Davis, D.L., Edwards, A.V., Juraszek, A.L., Phelps, A., Wessels, A., Burch, J.B., 2001a. A GATA-6 gene heart-region-specific enhancer provides a novel means to mark and probe a discrete component of the mouse cardiac conduction system. Mech. Dev. 108, 105–119. Davis, D.L., Edwards, A.V., Juraszek, A.L., Phelps, A., Wessels, A., Burch, J.B.E., 2001b. A GATA-6 gene heart-region-specific enhancer provides a novel means to mark and probe a discrete component of the mouse cardiac conduction system. Mech. Dev. 108, 105–119. de Haan, R.L., 1961. Differentiation of the atrioventricular conducting system of the heart. Circulation 24, 458–470. de Jong, F., Opthof, T., Wilde, A.A., Janse, M.J., Charles, R., Lamers, W.H., Moorman, A.F., 1992. Persisting zones of slow impulse conduction in developing chicken hearts. Circ. Res. 71, 240–250. Delorme, B., Dahl, E., Jarry-Guichard, T., Briand, J.P., Willecke, K., Gros, D., Théveniau-Ruissy, M., 1997. Expression pattern of connexin gene products at the early developmental stages of the mouse cardiovascular system. Circ. Res. 81, 423–437. Delorme, B., Dahl, E., Jarry-Guichard, T., Marics, I., Briand, J.P., Willecke, K., Gros, D., Théveniau-Ruissy, M., 1995. Developmental regulation of connexin40 gene expression in mouse heart correlates with the differentiation of the conduction system. Dev. Dyn. 204, 358–371. DiFrancesco, D., 1993. Pacemaker mechanisms in cardiac tissue. Annu. Rev. Physiol. 55, 455–472. Di Lisi, R., Sandri, C., Franco, D., Ausoni, S., Moorman, A.F., Schiaffino, S., 2000. An atrioventricular canal domain defined by cardiac troponin I transgene expression in the embryonic myocardium. Anat. Embryol. (Berlin) 202, 95–101. Dobrzynski, H., Nikolski, V.P., Sambelashvili, A.T., Greener, I.D., Yamamoto, M., Boyett, M.R., Efimov, I.R., 2003. Site of origin and molecular substrate of atrioventricular junctional rhythm in the rabbit heart. Circ. Res. 93, 1102–1110. Efimov, I.R., Nikolski, V.P., Rothenberg, F., Greener, I.D., Li, J., Dobrzynski, H., Boyett, M., 2004. Structure-function relationship in the AV junction. Anat. Rec. A Discov. Mol. Cell Evol. Biol. 280, 952–965. Eralp, I., Lie-Venema, H., Bax, N.A., Wijffels, M.C., van der Laarse, A., deRuiter, M.C., Bogers, A.J., van den Akker, N.M., Gourdie, R.G., Schalij, M.J., Poelmann, R.E., Gittenberger-de Groot, A.C., 2006. Epicardium-derived cells are important for correct development of the Purkinje fibers in the avian heart. Anat. Rec. A Discov. Mol. Cell Evol. Biol. 288, 1272–1280. Franco, D., Campione, M., 2003. The role of Pitx2 during cardiac development. Linking left–right signaling and congenital heart diseases. TCM 13, 157–163.

PART  |  2  Cardiac Precursor Populations and Lineages

Franco, D., Meilhac, S.M., Christoffels, V.M., Kispert, A., Buckingham, M., Kelly, R.G., 2006. Left and right ventricular contributions to the formation of the interventricular septum in the mouse heart. Dev. Biol. 294, 366–375. Garcia-Frigola, C., Shi, Y., Evans, S.M., 2003. Expression of the hyperpolarization-activated cyclic nucleotide-gated cation channel HCN4 during mouse heart development. Gene. Expr. Patterns 3, 777–783. Gassmann, M., Casagranda, F., Orioli, D., Simon, H., Lai, C., Klein, R., Lemke, G., 1995. Aberrant neural and cardiac development in mice lacking the ErbB4 neuregulin receptor. Nature 378, 390–394. Gittenberger-de Groot, A.C., Blom, N.M., Aoyama, N., Sucov, H., Wenink, A.C.G., Poelmann, R.E., 2003. The role of neural crest and epicardium-derived cells in conduction system formation. In: Chadwick, D.J., Goode, J. (Eds.) Development of the Cardiac Conduction System. John Wiley, Chichester, UK, pp. 125–141. Gorza, L., Schiaffino, S., Vitadello, M., 1988. Heart conduction system: a neural crest derivative? Brain Res. 457, 360–366. Gorza, L., Vitadello, M., 1989. Distribution of conduction system fibres in the developing and adult rabbit heart, revealed by an antineurofilament antibody. Circ. Res. 65, 360–369. Gourdie, R.G., Mima, T., Thompson, R.P., Mikawa, T., 1995. Terminal diversification of the myocyte lineage generates Purkinje fibers of the cardiac conduction system. Development 121, 1423–1431. Gourdie, R.G., Wei, Y., Kim, D., Klatt, S.C., Mikawa, T., 1998. Endothelin-induced conversion of embryonic heart muscle cells into impulse-conducting Purkinje fibers. Proc. Natl. Acad. Sci. USA 95, 6815–6818. Gourdie, R.G., Harris, B.S., Bond, J., Edmonson, A.M., Cheng, G., Sedmera, D., O’Brien, T.X., Mikawa, T., Thompson, R.P., 2003a. His-Purkinje lineages and development. In: Chadwick, D.J., Goode, J. (Eds.) Development of the Cardiac Conduction System. John Wiley, Chichester, UK, pp. 110–124. Gourdie, R.G., Harris, B.S., Bond, J., Justus, C., Hewett, K.W., O’Brien, T.X., Thompson, R.P., Sedmera, D., 2003b. Development of the cardiac pacemaking and conduction system. Birth Defects Res. Part C Embryo Today 69, 46–57. Gourdie, R.G., Harris, B.S., Bond, J., Justus, C., Hewett, K.W., O’Brien, T.X., Thompson, R.P., Sedmera, D., 2003c. Development of the cardiac pacemaking and conduction system. Birth Defects Res. Part C Embryo Today 69, 46–57. Graham, V., Zhang, H., Willis, S., Creazzo, T.L., 2006. Expression of a two-pore domain K channel (TASK-1) in developing avian and mouse ventricular conduction systems. Dev. Dyn. 235, 143–151. Grego-Bessa, J., Luna-Zurita, L., del Monte, G., Bolos, V., Melgar, P., Arandilla, A., Garratt, A.N., Zang, H., Mukouyama, Y.S., Chen, H., Shou, W., Ballestar, E., Esteller, M., Rojas, A., Perez-Pomares, J.M., de la Pompa, J.L., 2007. Notch signaling is essential for ventricular chamber development. Dev. Cell 12, 415–429. Gros, D., van Kempen, M.J.A., Théveniau, M., Delorme, B., JarryGuichard, T., Ten Velde, I., Maro, B., Briand, J.P., Jongsma, H.J., Moorman, A.F., 1995. Expression and distribution of connexin-40 in mammal heart. Prog. Cell Res. 4, 181–186. Gurjarpadhye, A., Hewett, K.W., Justus, C., Wen, X., Stadt, H., Kirby, M.L., Sedmera, D., Gourdie, R.G., 2007. Cardiac neural crest ablation inhibits compaction and electrical function of conduction system bundles. Am. J. Physiol. Heart Circ. Physiol. 292, H1291–H1300. Habets, P.E., Moorman, A.F., Clout, D.E., van Roon, M.A., Lingbeek, M., Lohuizen, M., Campione, M., Christoffels, V.M., 2002. Cooperative action of Tbx2 and Nkx2.5 inhibits ANF expression in

Chapter  |  2.3  Patterning and Development of the Conduction System of the Heart

the atrioventricular canal: implications for cardiac chamber formation. Genes Dev. 16, 1234–1246. Hall, C.E., Hurtado, R., Hewett, K.W., Shulimovich, M., Poma, C.P., Reckova, M., Justus, C., Pennisi, D.J., Tobita, K., Sedmera, D., Gourdie, R.G., Mikawa, T., 2004. Hemodynamic-dependent patterning of endothelin converting enzyme 1 expression and differentiation of impulse-conducting Purkinje fibers in the embryonic heart. Development 131, 581–592. Harrelson, Z., Kelly, R.G., Goldin, S.N., Gibson-Brown, J.J., Bollag, R.J., Silver, L.M., Papaioannou, V.E., 2004. Tbx2 is essential for patterning the atrioventricular canal and for morphogenesis of the outflow tract during heart development. Development 131, 5041–5052. Hertig, C.M., Kubalak, S.W., Wang, Y., Chien, K.R., 1999. Synergistic roles of neuregulin-1 and insulin-like growth factor-I in activation of the phosphatidylinositol 3-kinase pathway and cardiac chamber morphogenesis. J. Biol. Chem. 274, 37362–37369. Hiroi, Y., Kudoh, S., Monzen, K., Ikeda, Y., Yazaki, Y., Nagai, R., Komuro, I., 2001. Tbx5 associates with Nkx2-5 and synergistically promotes cardiomyocyte differentiation. Nat. Genet. 28, 276–280. Hoogaars, W.M., Engel, A., Brons, J.F., Verkerk, A.O., de Lange, F.J., Wong, L.Y.E., Bakker, M.L., Clout, D.E., Wakker, V., Barnett, P., Ravesloot, J.H., Moorman, A.F., Verheijck, E.E., Christoffels, V.M., 2007. Tbx3 controls the sinoatrial node gene program and imposes pacemaker function on the atria. Genes Dev. 21, 1098–1112. Hoogaars, W.M., Tessari, A., Moorman, A.F., de Boer, P.A., Hagoort, J., Soufan, A.T., Campione, M., Christoffels, V.M., 2004. The transcriptional repressor Tbx3 delineates the developing central conduction system of the heart. Cardiovasc. Res. 62, 489–499. Hyer, J., Johansen, M., Prasad, A., Wessels, A., Kirby, M.L., Gourdie, R. G., Mikawa, T., 1999. Induction of purkinje fiber differentiation by coronary arterialization. Proc. Natl. Acad. Sci. USA 96, 13214–13218. Ikeda, T., Iwasaki, K., Shimokawa, I., Sakai, H., Ito, H., Matsuo, T., 1990. Leu-7 immunoreactivity in human and rat embryonic hearts, with special reference to the development of the conduction tissue. Anat. Embryol. 182, 553–562. Ismat, F.A., Zhang, M., Kook, H., Huang, B., Zhou, R., Ferrari, V.A., Epstein, J.A., Patel, V.V., 2005. Homeobox protein Hop functions in the adult cardiac conduction system. Circ. Res. 96, 898–903. Ito, H., Iwasaki, K., Ikeda, T., Sakai, H., Shimokawa, I., Matsuo, T., 1992. HNK-1 expression pattern in normal and bis-diamine induced malformed developing rat heart: three dimensional reconstruction analysis using computer graphics. Anat. Embryol. 186, 327–334. Janse, M.J., Anderson, R.H., 1974. Specialized internodal atrial pathways – fact or fiction? Eur. J. Cardiol 2, 117–136. Jay, P.Y., Harris, B.S., Maguire, C.T., Buerger, A., Wakimoto, H., Tanaka, M., Kupershmidt, S., Roden, D.M., Schultheiss, T.M., O’Brien, T.X., Gourdie, R.G., Berul, C.I., Izumo, S., 2004. Nkx2-5 mutation causes anatomic hypoplasia of the cardiac conduction system. J. Clin. Invest. 113, 1130–1137. Jay, P.Y., Maguire, C.T., Wakimoto, H., Izumo, S., Berul, C.I., 2005. Absence of Msx2 does not affect cardiac conduction or rescue conduction defects associated with Nkx2-5 mutation. J. Cardiovasc. Elec­ trophysiol. 16, 82–85. Jongbloed, M.R., Schalij, M.J., Poelmann, R.E., Blom, N.A., Fekkes, M.L., Wang, Z., Fishman, G.I., Gittenberger-de Groot, A.C., 2004. Embryonic conduction tissue: a spatial correlation with adult arrhythmogenic areas. J. Cardiovasc. Electrophysiol. 15, 349–355. Keith, A., Flack, M., 1907. The form and nature of the muscular connections between the primary divisions of the vertebrate heart. JAP 41, 172–189.

189

Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1, 435–440. Kirby, M.L., Stewart, D.E., 1983. Neural crest origin of cardiac ganglion cells in the chick embryo: identification and extirpation. Dev. Biol. 97, 433–443. Kistler, P.M., Sanders, P., Hussin, A., Morton, J.B., Vohra, J.K., Sparks, P.B., Kalman, J.M., 2003. Focal atrial tachycardia arising from the mitral annulus: electrocardiographic and electrophysiologic characterization. J. Am. Coll. Cardiol. 41, 2212–2219. Kokubo, H., Miyagawa-Tomita, S., Nakazawa, M., Saga, Y., Johnson, R.L., 2005. Mouse hesr1 and hesr2 genes are redundantly required to mediate Notch signaling in the developing cardiovascular system. Dev. Biol. 278, 301–309. Kondo, R.P., Anderson, R.H., Kupershmidt, S., Roden, D.M., Evans, S.M., 2003. Development of the cardiac conduction system as delineated by minK-lacZ. J. Cardiovasc. Electrophysiol. 14, 383–391. Kreuzberg, M.M., Sohl, G., Kim, J.S., Verselis, V.K., Willecke, K., Bukauskas, F.F., 2005. Functional properties of mouse connexin30.2 expressed in the conduction system of the heart. Circ. Res. 96, 1169–1177. Kreuzberg, M.M., Willecke, K., Bukauskas, F.F., 2006b. Connexinmediated cardiac impulse propagation: connexin 30.2 slows atrioventricular conduction in mouse heart. TCM 16, 266–272. Kreuzberg, M.M., Schrickel, J.W., Ghanem, A., Kim, J.S., Degen, J., Janssen-Bienhold, U., Lewalter, T., Tiemann, K., Willecke, K., 2006a. Connexin30.2 containing gap junction channels decelerate impulse propagation through the atrioventricular node. Proc. Natl. Acad. Sci. USA 103, 5959–5964. Lamers, W.H., de Jong, F., de Groot, I.J., Moorman, A.F., 1991. The development of the avian conduction system, a review. Eur. J. Morphol. 29, 233–253. Lamers, W.H., te Kortschot, A., Los, J.A., Moorman, A.F., 1987. Acetylcholinesterase in prenatal rat heart: a marker for the early development of the cardiac conductive tissue? Anat. Rec. 217, 361–370. Lee, K.F., Simon, H., Chen, H., Bates, B., Hung, M.C., Hauser, C., 1995. Requirement for neuregulin receptor erbB2 in neural and cardiac development. Nature 378, 394–398. Lees-Miller, J.P., Guo, J., Somers, J.R., Roach, D.E., Sheldon, R.S., Rancourt, D.E., Duff, H.J., 2003. Selective knockout of mouse ERG1 B potassium channel eliminates IKr in adult ventricular myocytes and elicits episodes of abrupt sinus bradycardia. Mol. Cell Biol. 23, 1856–1862. Li, Q.Y., Newbury-Ecob, R.A., Terret, J.A., Wilson, D.I., Curtis, A.R., Yi, C.H., Gebuhr, T., Bullen, P.J., Robson, S.C., Strachan, T., Bonnet, D., Lyonnet, S., Young, I.D., Raeburn, J.A., Buckler, A.J., Law, D.J., Brook, J.D., 2000. Holt-Oram syndrome is caused by mutations in TBX5, a member of the Brachyury (T) gene family. Nat. Genet. 15, 21–29. Li, Z., Marchand, P., Humbert, J., Babinet, C., Paulin, D., 1993. Desmin sequence elements regulating skeletal muscle-specific expression in transgenic mice. Development 117, 947–959. Lingbeek, M.E., Jacobs, J.J., van Lohuizen, M., 2002. The T-box repressors TBX2 and TBX3 specifically regulate the tumor suppressor gene p14ARF via a variant T-site in the initiator. J. Biol. Chem. 277, 26120–26127. Lints, T.J., Parsons, L.M., Hartley, L., Lyons, I., Harvey, R.P., 1993. Nkx-2.5: a novel murine homeobox gene expressed in early heart progenitor cells and their myogenic descendants. Development 119, 419–431.

190

Liu, J., Dobrzynski, H., Yanni, J., Boyett, M.R., Lei, M., 2007. Organisation of the mouse sinoatrial node: structure and expression of HCN channels. Cardiovasc. Res. 73, 729–738. Lu, Y., James, T.N., Yamamoto, S., Terasaki, F., 1993. Cardiac conduction system in the chicken: gross anatomy plus light and electron microscopy. Anat. Rec. 236, 493–510. Lyons, I., Parsons, L.M., Hartley, L., Li, R., Andrews, J.E., Robb, L., Harvey, R.P., 1995. Myogenic and morphogenetic defects in the heart tubes of murine embryos lacking the homeo box gene Nkx2-5. Genes Dev. 9, 1654–1666. Ma, L., Lu, M.F., Schwartz, R.J., Martin, J.F., 2005. Bmp2 is essential for cardiac cushion epithelial-mesenchymal transition and myocardial patterning. Development 132, 5601–5611. Mangoni, M.E., Couette, B., Bourinet, E., Platzer, J., Reimer, D., Striessnig, J., Nargeot, J., 2003. Functional role of L-type Cav1.3 Ca2 channels in cardiac pacemaker activity. Proc. Natl. Acad. Sci. USA 100, 5543–5548. Mangoni, M.E., Traboulsie, A., Leoni, A.L., Couette, B., Marger, L., Le, Q.K., Kupfer, E., Cohen-Solal, A., Vilar, J., Shin, H.S., Escande, D., Charpentier, F., Nargeot, J., Lory, P., 2006. Bradycardia and slowing of the atrioventricular conduction in mice lacking cav3.1/alpha1g ttype calcium channels. Circ. Res. 98, 1422–1430. McCabe, C.F., Gourdie, R.G., Thompson, R.P., Cole, G.J., 1995. Developmentally regulated neural protein EAP-300 is expressed by myocardium and cardiac neural crest during chick embryogenesis. Dev. Dyn. 203, 51–60. Meyer, D., Birchmeier, C., 1995. Multiple essential functions of neuregulin in development. Nature 378, 386–390. Meysen, S., Marger, L., Hewett, K.W., Jarry-Guichard, T., Agarkova, I., Chauvin, J.P., Perriard, J.C., Izumo, S., Gourdie, R.G., Mangoni, M.E., Nargeot, J., Gros, D., Miquerol, L., 2006. Nkx2.5 cell-autonomous gene function is required for the postnatal formation of the peripheral ventricular conduction system. Dev. Biol. 303, 740–753. Mikawa, T., Hurtado, R., 2007. Development of the cardiac conduction system. Semin. Cell Dev. Biol. 18, 90–100. Mikawa, T., Hyer, J., Itoh, N., Wei, Y., 1996. Retroviral vectors to study cardiovascular development. TCM 6, 79–86. Milan, D.J., Giokas, A.C., Serluca, F.C., Peterson, R.T., MacRae, C.A., 2006. Notch1b and neuregulin are required for specification of central cardiac conduction tissue. Development 133, 1125–1132. Milanesi, R., Baruscotti, M., Gnecchi-Ruscone, T., DiFrancesco, D., 2006. Familial sinus bradycardia associated with a mutation in the cardiac pacemaker channel. N. Engl. J. Med. 354, 151–157. Miquerol, L., Meysen, S., Mangoni, M., Bois, P., Van Rijen, H.V., Abran, P., Jongsma, H., Nargeot, J., Gros, D., 2004. Architectural and functional asymmetry of the His-Purkinje system of the murine heart. Cardiovasc. Res. 63, 77–86. Mommersteeg, M.T., Hoogaars, W.M., Prall, O.W., de Gier-de Vries, C., Wiese, C., Clout, D.E., Papaioannou, V.E., Brown, N.A., Harvey, R. P., Moorman, A.F., Christoffels, V.M., 2007. Molecular pathway for the localized formation of the sinoatrial node. Circ. Res. 100, 354–362. Moorman, A.F., de Jong, F., Lamers, W.H., 1997. Development of the conduction system of the heart. Pacing Clin. Electrophysiol. 20, 2087–2092. Moorman, A.F., de Jong, F., Denyn, M.M., Lamers, W.H., 1998. Development of the cardiac conduction system. Circ. Res. 82, 629–644. Moorman, A.F., Christoffels, V.M., 2003a. Development of the cardiac conduction system: a matter of chamber development. In: Chadwick,

PART  |  2  Cardiac Precursor Populations and Lineages

D.J., Goode, J. (Eds.) Development of the Cardiac Conduction System. John Wiley, Chichester, UK, pp. 25–34. Moorman, A.F., Christoffels, V.M., 2003b. Cardiac chamber formation: development, genes and evolution. Physiol. Rev. 83, 1223–1267. Moosmang, S., Stieber, J., Zong, X., Biel, M., Hofmann, F., Ludwig, A., 2001. Cellular expression and functional characterization of four hyperpolarization-activated pacemaker channels in cardiac and neuronal tissues. Eur. J. Biochem. 268, 1646–1652. Moretti, A., Caron, L., Nakano, A., Lam, J.T., Bernshausen, A., Chen, Y., Qyang, Y., Bu, L., Sasaki, M., Martin-Puig, S., Sun, Y., Evans, S. M., Laugwitz, K.L., Chien, K.R., 2006. Multipotent embryonic isl1 progenitor cells lead to cardiac, smooth muscle, and endothelial cell diversification. Cell 127, 1151–1165. Mori, A.D., Zhu, Y., Vahora, I., Nieman, B., Koshiba-Takeuchi, K., Davidson, L., Pizard, A., Seidman, C.E., Seidman, J.G., Chen, X.J., Henkelman, R.M., Bruneau, B.G., 2006. Tbx5-dependent rheostatic control of cardiac gene expression and morphogenesis. Dev. Biol. 297, 566–586. Morton, J.B., Sanders, P., Das, A., Vohra, J.K., Sparks, P.B., Kalman, J.M., 2001. Focal atrial tachycardia arising from the tricuspid annulus: electrophysiologic and electrocardiographic characteristics. J. Cardiovasc. Electrophysiol. 12, 653–659. Moskowitz, I.P., Pizard, A., Patel, V.V., Bruneau, B.G., Kim, J.B., Kupershmidt, S., Roden, D., Berul, C.I., Seidman, C.E., Seidman, J.G., 2004. The T-Box transcription factor Tbx5 is required for the patterning and maturation of the murine cardiac conduction system. Development 131, 4107–4116. Moskowitz, I.P., Kim, J.B., Moore, M.L., Wolf, C.M., Peterson, M.A., Shendure, J., Nobrega, M.A., Yokota, Y., Berul, C., Izumo, S., Seidman, J.G., Seidman, C.E., 2007. A molecular pathway including id2, tbx5, and nkx2-5 required for cardiac conduction system development. Cell 129, 1365–1376. Myers, D.C., Fishman, G.I., 2003. Molecular and functional maturation of the murine cardiac conduction system. TCM 13, 289–295. Nakagawa, M., Thompson, R.P., Terracio, L., Borg, T.K., 1993. Developmental anatomy of HNK-1 immunoreactivity in the embryonic rat heart: co-distribution with early conduction tissue. Anat. Embryol. 187, 445–460. Nakamura, T., Colbert, M.C., Robbins, J., 2006. Neural crest cells retain multipotential characteristics in the developing valves and label the cardiac conduction system. Circ. Res. 98, 1547–1554. Nguyen-Tran, V.T., Kubalak, S.W., Minamisawa, S., Fiset, C., Wollert, K.C., Brown, A.B., Ruiz-Lozano, P., Barrere-Lemaire, S., Kondo, R., Norman, L.W., Gourdie, R.G., Rahme, M.M., Feld, G.K., Clark, R.B., Giles, W.R., Chien, K.R., 2000. A novel genetic pathway for sudden cardiac death via defects in the transition between ventricular and conduction system cell lineages. Cell 102, 671–682. Paff, G.H., Boucek, R.J., Harrell, T.C., 1968. Observations on the development of the electrocardiogram. Anat. Rec. 160, 575–582. Patten, B.M., 1949. Initiation and early changes in the character of the heartbeat in vertebrate embryos. Physiol. Rev. 29, 31–47. Pennisi, D.J., Rentschler, S., Gourdie, R.G., Fishman, G.I., Mikawa, T., 2002. Induction and patterning of the cardiac conduction system. Int. J. Dev. Biol. 46, 765–775. Rana, M.S., Horsten, N.C., Tesink-Taekema, S., Lamers, W.H., Moorman, A.F., van den Hoff, M.J., 2007. Trabeculated right ventricular free wall in the chicken heart forms by ventricularization of the myocardium initially forming the outflow tract. Circ. Res. 100, 1000–1007.

Chapter  |  2.3  Patterning and Development of the Conduction System of the Heart

Randall, D.J., 1970. The circulatory system. In: Hoar, W.S., Randall, D.J. (Eds.) Fish Physiology. Academic Press, New York, NY, pp. 133–172. Reckova, M., Rosengarten, C., DeAlmeida, A., Stanley, C.P., Wessels, A., Gourdie, R.G., Thompson, R.P., Sedmera, D., 2003. Hemodynamics is a key epigenetic factor in development of the cardiac conduction system. Circ. Res. 93, 77–85. Rentschler, S., Vaidya, D.M., Tamaddon, H., Degenhardt, K., Sassoon, D., Morley, G.E., Jalife, J., Fishman, G.I., 2001. Visualization and functional characterization of the developing murine cardiac conduction system. Development 128, 1785–1792. Rentschler, S., Zander, J., Meyers, K., France, D., Levine, R., Porter, G., Rivkees, S.A., Morley, G.E., Fishman, G.I., 2002. Neuregulin-1 promotes formation of the murine cardiac conduction system. Proc. Natl. Acad. Sci. USA 99, 10464–10469. Romer, A.S., 1962. Circulatory system. In: The Vertebrate Body. Saunders Company, W.B. London, UK, pp. 410–457. Rutenberg, J.B., Fischer, A., Jia, H., Gessler, M., Zhong, T.P., Mercola, M., 2006. Developmental patterning of the cardiac atrioventricular canal by Notch and Hairy-related transcription factors. Development 133, 4381–4390. Satoh, H., 2003. Sino-atrial nodal cells of mammalian hearts: ionic currents and gene expression of pacemaker ionic channels. J. Smooth Muscle Res. 39, 175–193. Schott, J.-J., Benson, D.W., Basson, C.T., Pease, W., Silberbach, G.M., Moak, J.P., Maron, B.J., Seidman, C.E., Seidman, J.G., 1998. Congenital heart disease caused by mutations in the transcription factor NKX2-5. Science 281, 108–111. Schram, G., Pourrier, M., Melnyk, P., Nattel, S., 2002. Differential distribution of cardiac ion channel expression as a basis for regional specialization in electrical function. Circ. Res. 90, 939–950. Sedmera, D., Reckova, M., DeAlmeida, A., Coppen, S.R., Kubalak, S.W., Gourdie, R.G., Thompson, R.P., 2003a. Spatiotemporal pattern of commitment to slowed proliferation in the embryonic mouse heart indicates progressive differentiation of the cardiac conduction system. Anat. Rec. 274A, 773–777. Sedmera, D., Reckova, M., DeAlmeida, A., Sedmerova, M., Biermann, M., Volejnik, J., Sarre, A., Raddatz, E., McCarthy, R.A., Gourdie, R.G., Thompson, R.P., 2003b. Functional and morphological evidence for a ventricular conduction system in zebrafish and Xenopus hearts. Am. J. Physiol. Heart Circ. Physiol. 284, H1152–H1160. Seidl, W., Schulze, M., Steding, G., Kluth, D., 1981. A few remarks on the physiology of the chick embryo heart (Gallus gallus). Folia Morphol. (Prague) 29, 237–242. Singh, M.K., Christoffels, V.M., Dias, J.M., Trowe, M.O., Petry, M., Schuster-Gossler, K., Burger, A., Ericson, J., Kispert, A., 2005. Tbx20 is essential for cardiac chamber differentiation and repression of Tbx2. Development 132, 2697–2707. Smith, T.K., Bader, D.M., 2007. Signals from both sides: Control of cardiac development by the endocardium and epicardium. Semin. Cell Dev. Biol. 18, 84–89. Soufan, A.T., van den Berg, G., Ruijter, J.M., de Boer, P.A., van den Hoff, M.J., Moorman, A.F., 2006. Regionalized sequence of myocardial cell growth and proliferation characterizes early chamber formation. Circ. Res. 99, 545–552. Soufan, A.T., van den Hoff, M.J., Ruijter, J.M., de Boer, P.A., Hagoort, J., Webb, S., Anderson, R.H., Moorman, A.F., 2004. Reconstruction of the patterns of gene expression in the developing mouse heart reveals an architectural arrangement that facilitates the understanding of atrial malformations and arrhythmias. Circ. Res. 95, 1207–1215.

191

Stennard, F.A., Costa, M.W., Lai, D., Biben, C., Furtado, M.B., Solloway, M.J., McCulley, D.J., Leimena, C., Preis, J.I., Dunwoodie, S.L., Elliott, D.E., Prall, O.W., Black, B.L., Fatkin, D., Harvey, R.P., 2005. Murine T-box transcription factor Tbx20 acts as a repressor during heart development, and is essential for adult heart integrity, function and adaptation. Development 132, 2451–2462. Stieber, J., Herrmann, S., Feil, S., Loster, J., Feil, R., Biel, M., Hofmann, F., Ludwig, A., 2003. The hyperpolarization-activated channel HCN4 is required for the generation of pacemaker action potentials in the embryonic heart. Proc. Natl. Acad. Sci. USA 100, 15235–15240. Stroud, D.M., Darrow, B.J., Kim, S.D., Zhang, J., Jongbloed, M.R., Rentschler, S., Moskowitz, I.P., Seidman, J., Fishman, G.I., 2007. Complex genomic rearrangement in CCS-LacZ transgenic mice. Genesis 45, 76–82. Subramaniam, A., Jones, W.K., Gulick, J., Wert, S.E., Neumann, J., Robbins, J., 1991. Tissue-specific regulation of the a-myosin heavy chain gene promoter in transgenic mice. J. Biol. Chem. 266, 24613–24620. Sun, Y., Liang, X., Najafi, N., Cass, M., Lin, L., Cai, C., Chen, J., Evans, S., 2007. Islet 1 is expressed in distinct cardiovascular lineages, including pacemaker and coronary vascular cells. Dev. Biol. 304, 286–296. Takebayashi-Suzuki, K., Yanagisawa, M., Gourdie, R.G., Kanzawa, N., Mikawa, T., 2000. In vivo induction of cardiac Purkinje fiber differentiation by coexpression of preproendothelin-1 and endothelin converting enzyme-1. Development 127, 3523–3532. Takeuchi, J.K., Mileikovskaia, M., Koshiba-Takeuchi, K., Heidt, A.B., Mori, A.D., Arruda, E.P., Gertsenstein, M., Georges, R., Davidson, L., Mo, R., Hui, C.C., Henkelman, R.M., Nemer, M., Black, B.L., Nagy, A., Bruneau, B.G., 2005. Tbx20 dose-dependently regulates transcription factor networks required for mouse heart and moto­ neuron development. Development 132, 2463–2474. Tallini, Y.N., Ohkura, M., Choi, B.R., Ji, G., Imoto, K., Doran, R., Lee, J., Plan, P., Wilson, J., Xin, H.B., Sanbe, A., Gulick, J., Mathai, J., Robbins, J., Salama, G., Nakai, J., Kotlikoff, M.I., 2006. Imaging cellular signals in the heart in vivo: Cardiac expression of the highsignal Ca2 indicator GCaMP2. Proc. Natl. Acad. Sci. USA 103, 4753–4758. Tanaka, M., Chen, Z., Bartunkova, S., Yamasaki, N., Izumo, S., 1999. The cardiac homeobox gene Csx/Nkx2.5 lies genetically upstream of multiple genes essential for heart development. Development 126, 1269–1280. Thomas, P.S., Kasahara, H., Edmonson, A.M., Izumo, S., Yacoub, M.H., Barton, P.J.R., Gourdie, R.G., 2001. Elevated expression of Nkx-2.5 in developing myocardial conduction cells. Anat. Rec. 263, 307–313. Timmermans, C., Rodriguez, L.M., Crijns, H.J., Moorman, A.F., Wellens, H.J., 2003. Idiopathic left bundle-branch block-shaped ventricular tachycardia may originate above the pulmonary valve. Circulation 108, 1960–1967. Valderrabano, M., Chen, F., Dave, A.S., Lamp, S.T., Klitzner, T.S., Weiss, J.N., 2006. Atrioventricular ring reentry in embryonic mouse hearts. Circulation 114, 543–549. van Mierop, L.H., 1967. Localization of pacemaker in chick embryo heart at the time of initiation of heartbeat. Am. J. Physiol. 212, 407–415. van Mierop, L.H., Gessner, I.H., 1970. The morphologic development of the sinoatrial node in the mouse. Am. J. Cardiol. 25, 204–212. Verberne, M.E., Gittenberger-de Groot, A.C., Poelmann, R.E., 1998. Lineage and development of the parasympathetic nervous system of the embryonic chick heart. Anat. Embryol. (Berlin) 198, 171–184.

192

Verzi, M.P., McCulley, D.J., De, V.S., Dodou, E., Black, B.L., 2005. The right ventricle, outflow tract, and ventricular septum comprise a restricted expression domain within the secondary/anterior heart field. Dev. Biol. 287, 134–145. Virágh, S., Challice, C.E., 1973. Origin and differentiation of cardiac muscle cells in the mouse. J. Ultrastruct. Res. 42, 1–24. Virágh, S., Challice, C.E., 1977a. The development of the conduction system in the mouse embryo heart. I. The first embryonic A-V conduction pathway. Dev. Biol. 56, 382–396. Virágh, S., Challice, C.E., 1977b. The development of the conduction system in the mouse embryo heart. II. Histogenesis of the atrioventricular node and bundle. Dev. Biol. 56, 397–411. Virágh, S., Challice, C.E., 1980. The development of the conduction system in the mouse embryo heart. III. The development of sinus muscle and sinoatrial node. Dev. Biol. 80, 28–45. Virágh, S., Challice, C.E., 1982. The development of the conduction system in the mouse embryo heart. IV. Differentiation of the atrioventricular conduction system. Dev. Biol. 89, 25–40.

PART  |  2  Cardiac Precursor Populations and Lineages

Viswanathan, S., Burch, J.B., Fishman, G.I., Moskowitz, I.P., Benson, D.W., 2007. Characterization of sinoatrial node in four conduction system marker mice. J. Mol. Cell Cardiol. 42, 946–953. Vitadello, M., Vettore, S., Lamar, E., Chien, K.R., Gorza, L., 1996. Neu­ rofilament M mRNA is expressed in conduction system myocytes of the developing and adult rabbit heart. J. Mol. Cell Cardiol. 28, 1833–1844. Watanabe, M., Timm, M., Fallah-Najmabadi, H., 1992. Cardiac expression of polysialylated NCAM in the chicken embryo: correlation with the ventricular conduction system. Dev. Dyn. 194, 128–141. Watanabe, M., Jafri, A., Fisher, S.A., 2001. Apoptosis is required for the proper formation of the ventriculo-arterial connections. Dev. Biol. 240, 274–288. Wessels, A., Vermeulen, J.L., Verbeek, F.J., Virágh, S., Kálmán, F., Lamers, W.H., Moorman, A.F., 1992. Spatial distribution of “tissue-specific” antigens in the developing human heart and skeletal muscle: III. An immunohistochemical analysis of the distribution of the neural tissue antigen G1N2 in the embryonic heart; implications for the development of the atrioventricular conduction system. Anat. Rec. 232, 97–111.

Chapter 3.1

The Behavior of Cells that Form the Myocardial Compartments of the Vertebrate Heart Sigolène M. Meilhac and Margaret E. Buckingham Department of Developmental Biology, CNRS URA 2578, Pasteur Institute, Paris, France

I.  Introduction The precise shape of the heart reflects a highly controlled cellular organization. Cardiac malformations, such as transposition of the great arteries, hypoplastic left ventricle or myofiber disarray, have severe consequences for the organism. In this chapter we shall discuss the behavior of cells that form the myocardium, including their proliferation, migration, morphology and cellular interactions. The underlying question is how such aspects of cell biology, both in progenitor cells and in differentiated myocardium, contribute to the final shape of the heart with its morphologically distinct compartments.

II.  Progenitor cell migration Cardiac progenitor cells arise in the mesodermal layer which forms by ingression of cells into the primitive streak at gastrulation (Fig. 1). Cell tracing experiments in the amniote (chick and mouse) embryo have shown that cardiac progenitors ingress early, at the mid-streak stage, and are located in the anterior region of the primitive streak. They will contribute to the different tissues of the heart, including myocardium and endocardium. In the zebrafish, it has been shown that endocardial and myocardial cells share common progenitors in the early blastula (Lee et al., 1994), and that the lineages are almost fully segregated at 40% epiboly, just before gastrulation (Keegan et al., 2004). In the chick it has also been reported that endocardial and myocardial lineages are already segregated when gastrulation begins (Cohen-Gould and Mikawa, 1996; Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

Wei and Mikawa, 2000). However, this is controversial and another experimental approach led to the conclusion that segregation is a later event, when the cardiac crescent forms (Linask and Lash, 1993; Linask et al., 1997). Another issue concerns the organization of the progenitor cells that contribute myocardium to different parts of the heart. For the avian embryo a segmental model of cardiac development had been proposed in which myocardial progenitors for the different cardiac compartments that form on the anterior–posterior axis of the heart tube are already segregated anteriorly to posteriorly in the primitive streak (Brand, 2003). However, this is probably no more than a tendency, with outflow tract progenitors located most anteriorly (Garcia-Martinez and Schoenwolf, 1993; LopezSanchez et al., 2001). In the zebrafish embryo, on the other hand, atrial and ventricular myocardial lineages segregate at the midblastula stage (Stainier et al., 1993) and can be mapped to distinct regions at 40% epiboly, prior to gastrulation (Fig. 1A) (Keegan et al., 2004). In the mouse a retrospective clonal analysis has also shown the existence of two myocardial lineages which segregate early from a common progenitor (Meilhac et al., 2004a). Their contribution to the heart is more complex than in the zebrafish. The first lineage is specific to the left ventricle, whereas the second lineage is specific to the outflow tract. Both lineages have overlapping contributions to the other regions of the heart (Fig. 1C). How these myocardial lineages are distributed in the mesoderm of the mouse embryo remains to be understood. As they leave the primitive streak in the amniote embryo, myocardial progenitors migrate laterally and anter­ iorly to locate under the head folds on either side of the

195

196

PART  |  3  Patterning of the Early Heart Tube

midline (Kinder et al., 2001) (Fig. 1B–C). It was thought, based on grafting experiments in the chick embryo, that these cells behave as a coherent sheet (Rosenquist and De Haan, 1966; Stalsberg and De Haan, 1969; Rosenquist, 1970), however more recent cell labeling analyses (Redkar et al., 2001) have shown that heart progenitor cells initially intermingle during their migration from the streak. Similarly, retrospective clonal analysis of myocardial cells in the mouse embryo showed that their progenitor cells undergo an early phase of dispersive growth, followed by coherent growth (Meilhac et al., 2003); when this transition in cell behavior takes place has not been determined.

Regulation of the migration of cardiac progenitors remains poorly-understood in the amniote embryo. The transcription factor Mesp1 is an early marker of cells that will contribute to the heart, as well as to other mesodermal derivatives. Mutation of the gene in the mouse delays migration, resulting in cardia bifida (Fig. 4C; see Section V of this chapter) and in double Mesp1; Mesp2 mutant embryos myocardial progenitors do not emerge from the streak (Saga et al., 1999; Kitajima et al., 2000). Fgfr1 (Deng et al., 1994; Yamaguchi et al., 1994; Ciruna et al., 1997) is also expressed in the primitive streak, and in Fgfr1 mutants mesodermal cells do not migrate from it correctly,

(A) Zebrafish

BA A

V V

A

5 hpf

16.5 hpf

26 hpf

V

A

52 hpf

mature heart

(B) Chick DOFT

Anterior heart field Secondary heart field

RA

LV

RV

ps

St 7

St 10

LA

POFT

Heart tube

St 3

OFT

PhA

St 16

St 22

(C) Mouse Cardiac crescent (first heart field)

PhA PhA

Second heart field

OFT RV

LV

Heart tube

Heart tube

LA

OFT

RA RV

LV

ps

E7.0

E7.5

E8.0

E8.5

E10.5

Figure 1  Summary of heart development in vertebrates. (A) In the zebrafish heart, two cell lineages are already distinguished before gastrulation, at 5 hours postfertilization (hpf). These adopt a medio-lateral position at 16.5 hpf before the formation of the heart tube. In the heart tube at 26 hpf, they contribute respectively to the atrium (A) and to the ventricle (V). At 52 hpf the heart tube has looped (BA: bulbus arteriosus). (B) In the chick embryo, heart progenitor cells invaginate into the anterior half of the primitive streak (stage (st) 3 according to the nomenclature of Hamburger and Hamilton, 1951). By stage 7, they have migrated under the head folds and start expressing markers such as Nkx2.5 (black zone). The tube is straight at stage 10 and has looped at stage 16. Secondary and anterior heart fields have been defined at stage 16 and will contribute to portions of the outflow tract (OFT) at stage 22 (DOFT: distal outflow tract; LA: left atrium; LV: left ventricle; PhA: pharyngeal arches; POFT: proximal outflow tract; ps: primitive streak; RA: right atrium; RV: right ventricle). (C) In the mouse, heart progenitors have reached the head folds by embryonic day (E). Two heart fields have been distinguished at E7.5, with a medio–lateral distribution. These participate differentially in the straight (E8) and looped (E8.5) heart tube. They correspond to two distinct lineages which contribute exclusively to the outflow tract or left ventricle, and overlap in the other chambers. Abbreviations are as in (B). Gastrulation movements in (A–C) are shown by dashed arrows. Black arrows indicate cell movement into the heart.

Chapter  |  3.1  The Behavior of Cells that Form the Myocardial Compartments of the Vertebrate Heart

resulting in an abnormal heart tube. In chimeras, mutant cells fail to colonize the heart, showing that the genes are required cell-autonomously. Studies in the zebrafish show that the G-protein coupled receptor, encoded by the miles apart (mil) locus, which binds sphingosine-1-phosphate, provides a permissive environment for the migration of myocardial progenitor cells (Kupperman et al., 2000). Another G-protein coupled receptor, Agtrl1b, in the lateral plate mesoderm and its ligand, Apelin, are essential for the migration of myocardial progenitors at the time of gastrulation (Scott et al., 2007; Zeng et al., 2007). Apelin, which is expressed at the midline, has been implicated more broadly in the regulation of convergence–extension movements (Solnica-Krezel, 2005). The Wnt/planar cell polarity pathway (see Karner et al., 2006b) is required for these movements and mutations in wnt11/silberblick, wnt11r, wnt4a, disheveled (Matsui et al., 2005), diversin/diego (Moeller et al., 2006) and csrp1 (Miyasaka et al., 2007) lead to cardia bifida in the zebrafish embryo. In mouse and chick, perturbations of Rho-associated kinases (Rock), which regulate the reorganization of the cytoskeleton downstream of the planar cell polarity pathway, also lead to cardia bifida (Wei et al., 2001). A potential activator of the planar cell polarity pathway in this context is Wnt3a, which is expressed in the primitive streak and mediates

(A)

197

repulsion of cardiac progenitor cells, and thus their lateral migration (Yue et al., 2008). Once they have taken up their position under the head folds, myocardial progenitor cells in amniotes can be divided into first and second heart fields. The two fields correlate with the concept of two lineages in terms of their contribution to the heart (see Chapter 2.2). The second heart field initially lies medially to the first heart field, where differentiating myocardial cells first appear in an epithelial structure, referred to as the cardiac crescent in mammals (Fig. 1C).

III.  Formation of the myocardial epithelium At the time when the cardiac crescent forms, the mesoderm has split into splanchnic and somatic mesoderm surrounding the pericardial coelomic cavity. At this stage, cells that will contribute to the heart become morphologically distinguishable from other mesodermal cells. The somatic layer contains cells that will form the pericardium, whereas cells that will form the myocardium lie in the splanchnic layer (Fig. 2A). Laterally-located myocardial progenitors change shape and become cuboidal or columnar, whereas

Apical

Microvilli

nt p

p Endo

Basal membrane Basal

Apical

(B) nt

aPKCλ, α-catenin

Endo p

p YSL

Tjp1 β-catenin Fibronectin Basal

Pericardium Myocardium Endocardium Figure 2  Formation of the myocardial epithelium. (A) Scheme of a transverse section of a 1-somite stage mouse embryo, showing in blue the newly-formed myocardial epithelium of the cardiac crescent. Its apical–basal polarity is shown. Pericardial cells (in red) remain squamous. Endocardial cells (in yellow) form separate groups of cells ventral to the myocardium. The situation in chick is similar, except that different markers of apico–basal polarity have been used (see text). (B) Scheme of a transverse section of an 18-somite stage zebrafish embryo, showing the polarized myocardial epithelium, at the level of lateral plate mesoderm tubules which have not yet fused (endo: endoderm; nt: neural tube; p: pericardial coelomic cavity; PKC: protein kinase C; Tjp1: tight junction protein (also known as ZO1); YSL: yolk sac layer).

198

pericardial progenitors remain squamous. Progenitors of the endocardium become visible underneath the myocardial epithelium, initially as separate groups of cells surrounding small cavities. This distinction between progenitors of the three main heart tissues occurs around the one somite stage in the mouse embryo (Kaufman and Navaratnam, 1981). Visualization of differentiating myocardial cells under the electron microscope shows their apico–basal polarity, including a basal basement membrane overlying the cells that will form the endocardium and apical microvilli exposed to the pericardial coelomic cavity. They also display lateral cell junctions such as desmosomes, gap, adherent and tight junctions (Navaratnam et al., 1986). In the chick, apico–basal polarity of myocardial progenitor cells in the cardiac primordia (equivalent to the crescent in the mouse) has also been described from the time of formation of the pericardial coelom at stage St6–7 (Linask, 1992). N-cadherin marks the apical membrane and integrin1 is located basally, whereas Na/KATPase is lateral. N-cadherin is essential for maintaining the epithelial integrity of this myocardium, as shown by its disorganization after injection of an antibody against N-cadherin (Nakagawa and Takeichi, 1997). A similar disorganization is observed in mice (Radice et al., 1997) and zebrafish (Bagatto et al., 2006) mutants, respectively for N-cadherin and cadherin2/glass onion, although this is manifest slightly later in the heart tube. Polarity of the myocardial epithelium has been more extensively characterized in the zebrafish embryo. Before heart tube fusion in the 18-somite stage embryo myocardial cells, which already express the myosin cmlc2 marker, are localized in tubular stuctures in ventral and medial regions of the lateral plate mesoderm, whereas endocardial cells lie between the bilateral tubules (Fig. 2B). Molecular markers such as -catenin and aPKC or -catenin are restricted, respectively, to the basal and lateral membranes of myocardial cells (Trinh et al., 2005), thus showing that they have apico–basal polarity. In other systems, especially in Drosophila melanogaster and Caenorhabditis elegans, establishment of apico–basal polarity has been shown to depend on the Par complex of proteins which localize to the lateral membrane and function to segregate cytoplasmic determinants and orient the mitotic spindle (Karner et al., 2006a). Interaction with other protein complexes, such as the crumbs complex at the subapical tight junctions and the scribble complex located at the base of the lateral membrane, is necessary to establish and maintain cell polarity. Analysis of zebrafish mutants shows that polarity of myocardial cells is important for heart morphogenesis and implies orthologous proteins. The heart and soul (has) mutation affects the formation of the cardiac cone (a structure that precedes the heart tube during zebrafish cardiogenesis; see Chapter 1.4) and causes the ventricle to develop within the atrium (Peterson et al., 2001). The has gene encodes aPKC, an atypical protein kinase that is part

PART  |  3  Patterning of the Early Heart Tube

of the Par complex. The nagie oko (nok) mutation impairs heart tube elongation (Rohr et al., 2006). The nok gene encodes Pals1, a membrane-associated guanylate kinase (Wei et al., 2002) that is part of the Crumbs complex. Both mutations affect the polarized epithelial organization of myocardial cells before cardiac fusion and formation of the cone (Rohr et al., 2006). Fibronectin, a major component of the extracellular matrix that interacts with integrins, is also necessary for the epithelial organization of myocardial cells. Disruption of fibronectin in the natter mutant prevents formation of the heart tube (Trinh and Stainier, 2004). In addition to impaired myocardial cell migration, adherent junctions between myocardial cells do not form properly and apico–basal markers are disorganized. Increased mitotic activity has been observed in the mouse myocardial epithelium at the time of its formation (Kaufman and Navaratnam, 1981). It is usually thought that epithelial cells grow coherently, with little intermingling, because of the tight adhesion between cells, although it has been shown that this may not be the case in the mouse epiblast, where epithelial cells may intermingle (Gardner and Cockroft, 1998). The epithelialization of myocardial progenitor cells may cause the transition in mode of growth from dispersive to coherent. A related question concerns the differentiated status of proliferating cells in the epithelial structure of the crescent. The expression of sarcomeric proteins is first detected within this epithelium (Bisaha and Bader, 1991; Han et al., 1992) in a subset of cells, although the contractile apparatus becomes functional later (Satin et al., 1988; Nishii and Shibata, 2006). It is not clear whether the epithelial structure is a prerequisite for myocardial differentiation, both at the crescent stage and when cells are added later to the formed tube. It is also not clear what signals trigger epithelialization. In the zebrafish embryo the cardiogenic factor Hand2, a bHLH transcription factor, lies genetically upstream of fibronectin which is required for the organization of a myocardial epithelium (Trinh et al., 2005). This suggests that progenitor cells have entered the myocardial program as epithelialization is established, and indeed they already express the sarcomeric myosin gene, cmlc2. This is also probably the case in the mouse embryo, where Myl7 (also known as Mlc2a) is expressed early (Cai et al., 2003). There is also the question of whether progenitor cells continue to enter the crescent once it has formed, for example posteriorly or even medio-laterally. The localization of such progenitor cells, expressing cardiac markers, is discussed in the next section on the behavior of cells in the second heart field.

IV.  Behavior of cells in the second heart field Cells in splanchnic mesoderm lying medially to the crescent on either side of the midline constitute the second

Chapter  |  3.1  The Behavior of Cells that Form the Myocardial Compartments of the Vertebrate Heart

heart field of the mouse embryo. As the heart tube forms, the second heart field comes to lie behind the cardiac tube, as well as extending more anteriorly and posteriorly (Fig. 1C). Explant experiments and Di-I labeling indicate that the crescent mainly contributes left ventricular myocardium, while outflow tract myocardium and part of that of the right ventricle and atria derive from the second heart field (Kelly et al., 2001; Zaffran et al., 2004; Galli et al., 2008). A regulatory network of genes is expressed in these cells, and their mutation leads to phenotypes affecting the parts of the heart that derive from the second heart field (see Chapter 2.2). This would suggest that these progenitors have a genetic program which is distinct from those of the first heart field that form the crescent. However, markers that distinguish first heart field progenitors are not well-defined (see Buckingham et al., 2005). In the absence of Nkx2.5, cells in the crescent, as well as those in the heart tube, express Isl1, which had been thought to mark only progenitor cells in the second heart field (Prall et al., 2007). Interestingly in Xenopus, which has a heart with only one ventricle, a field of Isl1 positive cells has now been identified which at early stages includes the cardiac crescent; Isl1 expression is required for normal heart development (Brade et al., 2007). Detailed analysis at the cellular level in the mouse embryo will be required to characterize undifferentiated cells in the crescent, and also to trace potential second heart field contributions to this epithelium. As the heart tube forms, clones of cells extend from the dorsal mesocardium into the dorsal epithelium of the still open tube, indicating that at this stage the second heart field does not only contribute to the growing poles of the heart (Meilhac et al., 2004a). The contribution of the two fields to different parts of the heart correlates with that of the two myocardial cell lineages identified by retrospective clonal analysis; the fact that these segregate early means that cells in the first and second heart fields have distinct developmental histories. In the chick embryo, secondary (Waldo et al., 2001) and anterior (Mjaatvedt et al., 2001) heart fields have been described that contribute to outflow tract myocardium (Fig. 1B), and this contribution may extend to the right ventricle also (Kirby, 2007). In the chick, as in mammals, the second heart field will contribute other cardiovascular cell types, including smooth muscle cells (Waldo et al., 2005b). It is not yet established whether these different cell types derive from common or from distinct progenitors. Heterogeneity exists between myocardial progenitors, as indicated by gene expression patterns. Isl1, for example, appears to mark most of the second heart field and Isl1-expressing cells contribute to the venous, as well as the arterial, pole of the heart (Cai et al., 2003), whereas Fgf8 and Fgf10 are transcribed in an anterior subpopulation of Isl1 positive cells, which colonize the anterior pole of the heart. Within the Fgf-positive population, further subpopulations are indicated by transgene expression profiles, with the suggestion

199

that these contribute to subdomains of outflow tract myocardium, and ultimately to that of the aorta or pulmonary trunk (Bajolle et al., 2007). Progenitor cells that will contribute to the atria are restricted to the posterior part of the second heart field, within which cells on the left or right sides contribute to the left or right common atrium, respectively (Galli et al., 2008). Evidence for left–right patterning also comes from cell tracing experiments in the chick embryo, which show that progenitors of smooth muscle at the base of the aorta or pulmonary trunk are located on the left or right sides of the second heart field, respectively (Ward et al., 2005). Regionalization of progenitor cells that will contribute to different domains of the heart has important implications for cell behavior in the second heart field. Cell dispersion must be limited, and indeed observation of the distribution of clonally-related cells in outflow tract mocardium (Bajolle et al., 2007) would suggest that this is the case. Time-lapse imaging should provide direct information about cell movement in the second heart field. Although the epithelial nature of this splanchnic mesoderm is less well-characterized than that of the crescent, it may partly account for limited cell dispersion. However, the molecular basis for this restriction remains to be elucidated. Progenitor cells proliferate in the second heart field, and a number of signaling pathways have now been implicated in the regulation of this proliferation. This role is often linked to the regulation of myocardial differentiation. Classically, Bmps induce this process (Harvey, 2002). However, Bmp signaling also regulates progenitor cell proliferation, as demonstrated by the upregulation of Bmp2/ Smad1 signaling, which inhibits proliferation in the second heart field of Nkx2.5 mutant mice (Prall et al., 2007). Bmp signaling is also implicated in the recruitment of cells to form outflow tract myocardium at the arterial pole of the cardiac tube in the chick embryo (Waldo et al., 2001). The control of this process at the poles of the tube, or indeed dorsally before the tube closes when cardiac actinpositive cells extend into the dorsal mesocardium (Meilhac et al., 2004a), is an important issue. Fgf signaling also plays a dual role in myocardial differentiation (Alsan and Schultheiss, 2002) and progenitor cell proliferation, as shown by manipulation of Fgf signaling in the chick embryo (Waldo et al., 2005a). Conditional mutation of Fgf8 in the mouse results in outflow tract abnormalities, probably reflecting reduced proliferation in the second heart field (Ilagan et al., 2006; Park et al., 2006), although there may also be effects on cell movement, as documented for zebrafish fgf8/acerebellar-mutants, in which the ventricle (there is no outflow tract in the early fish embryo) (Fig. 1A) fails to form (Reifers et al., 2000). Canonical Wnt signaling has been shown to exert a negative effect on myocardial cell differentiation before heart tube formation in chick and frog (Marvin et al., 2001; Schneider and Mercola, 2001; Tzahor and Lassar, 2001), and such ­ signaling from

200

the midline may prevent differentiation in the progenitors of the mouse second heart field, which is more medially located than the first heart field. In keeping with a negative effect of Wnt signaling on myocardial differentiation, Mesp1 activation of Dkk1, that inhibits this pathway, has been shown to be part of a mesodermal regulatory cascade in Xenopus, leading to myocardial ­ differentiation (David et al., 2008). A series of recent papers (Ai et al., 2007; Cohen et al., 2007; Klaus et al., 2007; Kwon et al., 2007; Lin et al., 2007; Qyang et al., 2007) document the effects of -catenin gain- and loss-of-function mutations on the proliferation of cells in the mouse second heart field, leading to a change in the size of the right ventricle and outflow tract. Canonical Wnt signaling promotes the expansion of progenitor cells. Consistent with this, cyclin D2 expression depends on the presence of -catenin, as does that of Tgf 2 and Bmp4, suggesting interactions between different signaling pathways that affect proliferation in the second heart field. In some experiments, the expression of Fgf10 or Isl1 was shown to depend on canonical Wnt signaling. Apoptosis in these experiments, when Isl1 is reduced, may be due to an effect on Shh signaling, which has been shown to be required for survival in the second heart field (Washington Smoak et al., 2005; Lin et al., 2006). In the gain of function mutation, accumulation of Isl1 positive progenitors in the outflow tract was documented, suggesting that the onset of differentiation is also perturbed. Variations in the phenotypes reported using different Cre drivers probably reflect differences in timing and efficiency of the conditional mutation, and also potentially of its targeting to different cell types in the second heart field. In these experiments, there is no evidence for a direct effect on cardiomyocyte differentiation. This is in contrast to observations on in vitro stem cell systems; however, promotion of progenitor cell proliferation or differentiation by Wnt signaling in ES cells depends on the stage of embroid body maturation (Qyang et al., 2007; Ueno et al., 2007). Temporal differences in the role of Wnt signaling have been documented for cardiogenesis in the zebrafish (Ueno et al., 2007), where Wnt signaling is shown to promote early myocardial differentiation before gastrulation, whereas later it has an inhibitory effect (see also Rottbauer et al., 2002). It is notable that in the phenotypes described after manipulation of -catenin in the mouse embryo, no effect on the cardiac crescent or left ventricle was observed with Cre drivers that are also expressed in the crescent or its progenitors (Cohen et al., 2007; Klaus et al., 2007), again pointing to a different regulatory circuit in the first heart field. One caveat in interpreting the -catenin gainand loss-of-function experiments is that -catenin is also implicated in cadherin-mediated cell contacts, and therefore may have additional effects on cell organization. Another potentially important signaling pathway which has not been investigated specifically in the second heart field is that of Notch signaling. Experiments in Xenopus suggest

PART  |  3  Patterning of the Early Heart Tube

that it may play a role in controlling myocardial versus nonmyocardial cell fates in cells that have activated Nkx2.5 and are already destined to contribute to the heart (Rones et al., 2000). To date, only later phenotypes in the heart have been described after manipulation of Notch signaling in the mouse embryo (see Section VI of this chapter). In conclusion, there is now extensive information about the genetic network that regulates the contribution of cells from the second heart field to different regions of the heart (reviewed in Chapter 2.2). How these genes influence cell behavior and the signaling pathways that are implicated is less well-understood. Recent results begin to establish the broad outlines of how proliferation, for example, is regulated. Other major questions include the avoidance of premature differentiation, the restriction of extensive cell movement and the activation of differentiation as cells enter the heart tube, and the mechanisms that control this recruitment. Another major issue is the significance of heterogeneity in gene expression patterns, how this is regulated and what this means in terms of regionalization of myocardial potential. In the mouse embryo more extensive prospective lineage analysis will be required to clarify the progenitor cell fate map of the second heart field.

V.  Formation of the cardiac tube Formation of a single heart tube, also known as fusion, occurs rostrally under the head folds when the bilaterallylocated myocardial epithelia fuse at the midline (Fig. 1). This process starts at the cardiac crescent stage in amniotes (E7.5 in mouse with 1–2 somites and St8–9 in chick with 4–9 somites) and at 18 hours post-fertilization (hpf) in zebrafish with about 18 somites. Not only does the timing of fusion, in relation to the development of other organs such as the somites, differ between amniotes and fish, but the cellular processes involved are also different. Fusion in amniotes starts with presumptive ventricular cells and generates a straight cardiac tube. Fusion then extends both rostrally and caudally (Moreno-Rodriguez et al., 2006), concomitantly with looping of the heart. In zebrafish, bilaterally-located myocardial cells will first contact caudally and then rostrally to generate a cone structure which surrounds endocardial progenitor cells (Glickman and Yelon, 2002; see also Chapter 1.4). Extension of the tube occurs by protrusion of the cone. Ventricular cells, which express the myosin isoform vmhc and have a cuboidal shape, are more internal in the cone than atrial cells and protrude first (Fig. 3A). However, despite timing and mechanistic differences, there is a common plan for the formation of the heart tube in vertebrates, with a phase of cell migration followed by fusion. Heart fusion is closely linked to the formation of the foregut. In amniotes, folding and invagination of the endo­derm occurs on both the rostral-to-caudal and lateral-to-medial axes

Chapter  |  3.1  The Behavior of Cells that Form the Myocardial Compartments of the Vertebrate Heart

A1

A2

A3

17.5hpf

18hpf

Atrial cells

A5

19.5hpf

24hpf

Ventricular cells wt

B1

t=50 min First phase C1

A4

201

B2

wt

t=130 min Second phase

B3

t=90 min

clo B4

mil

t=120 min

C2

Figure 3  Formation of the cardiac tube in zebrafish. (A) In dorsal views, after convergence (A1), the cardiogenic domains fuse caudally (A2) then rostrally (A3). In lateral views, the cone (A4) elongates as a tube (A5). Ventricular cells express both cmlc2 (red) and vmhc (green) and thus appear yellow, whereas atrial cells only express cmlc2. Panels modified from Glickman et al. (2002), with permission. (B) Trajectories of cells show two phases of fusion of the cardiogenic domains. Red arrows indicate medio-lateral movements, whereas yellow arrows indicate angular movements towards the antero–posterior axis. The former characterize the first phase (B1) and the latter the second phase (B2). In cloche (clo) (B3) and miles apart (mil) (B4) mutants respectively, the second or the first phase is abolished. Asterisks in B4 indicate no net displacement. The elapsed time from the start of the movie is given. Panels modified from Glickman et al. (2007), with permission. (C) Examples of phenotypes of cardia bifida in 33 hpf casanova/ sox32 (C1) and faust/gata5 (C2) mutant embryos, in which the cardiogenic domains do not fuse. Myosin heavy chain expression in the myocardium (MF20) is in red and in the atrium (St46) in green. Panels modified from Reiter et al. (1999), with permission.

(Fig. 4A,B). As a consequence, the dorsal–ventral organization of the cardiac tissues is inverted. Initially, the pericardial progenitor cells are dorsal to the myocardial progenitor cells, which in turn are dorsal to the endocardial progenitor cells. In contrast, when the foregut has closed, endocardial cells which abut the foregut, these are more dorsal to the myocardial cells and pericardial cells. Closure of the foregut brings together the bilateral cardiogenic domains, thereby initiating fusion. In zebrafish, endodermal cells, localized between ectoderm and mesoderm, reach the midline shortly before heart fusion at 16 hpf (Matsui et al., 2005). The gut forms by rearrangement of cells rather than folding of the tissue, such that endoderm cells become radially polarized and a lumen forms later by cavitation (Wallace and Pack, 2003). In all vertebrates, proper formation of the gut is essential for heart fusion. Zebrafish mutants which fail to form endoderm, such as casanova/sox32 (Alexander et al., 1999; Dickmeis et al., 2001; Kikuchi et al., 2001) or one-eyed pinhead (Schier et al., 1997), or which are defective for endoderm differentiation, such as bonnie and clyde (Kikuchi et al., 2000) or faust/ gata5 (Reiter et al., 1999), display cardia bifida, a phenotype of inhibited heart fusion with the formation of two bilateral heart rudiments (Fig. 3C). These mutations affect Nodal signaling, indicating that it is essential for endoderm specification and migration. The cardia bifida phenotype of one-eyed

pinhead and casanova mutants can be rescued by transplantation of cells which have been induced to activate Nodal signaling, and thus commit to an endoderm fate (Peyrieras et al., 1998; David and Rosa, 2001). This indicates that the impairment of heart fusion is secondary to malformations of the endoderm. In mouse, cardia bifida is also linked to failure or delay of foregut formation, such as in Gata4/ (Kuo et al., 1997; Molkentin et al., 1997) or Foxp4/ mutants (Li et al., 2004) (Fig. 4C). In addition, studies with chimeras have shown that Gata4 is required in the endoderm for heart fusion (Narita et al., 1997) and Foxp4 is also likely to play its role in the endoderm, as it is not expressed in the myocardium but in the foregut, epicardium and endocardium. Taken together, these data indicate that the endoderm, in addition to its role in regulating cardiac differentiation (Lough and Sugi, 2000), is important for the morphogenesis of the heart. How endoderm influences fusion of the heart remains to be understood. Is it acting as a substratum for mesodermal cell migration? Is it providing directional cues? Or does it induce passive migration of mesodermal cells during its convergence to the midline? Proper migration of mesodermal cells from their site of gastrulation to their final anterior midline position (see Section II of this chapter) is critical for heart fusion, and mutation of many of the genes implicated in this process

202

PART  |  3  Patterning of the Early Heart Tube

(A)

a

a

endo

nt

(B)

nt f

nt

nt

endo

1so

f 0so

6so

Endocardial cells Pericardial cells C1

C2

Foxp4-/-

Mesp1-/-

nt

nt

Myocardial cells endo

4-5so

f 2so

6so

Figure 4  Formation of the cardiac tube in the mouse. (A) Schema of sagittal sections shows the rostral-to-caudal invagination of the endoderm (blue arrows). (B) Schema of transverse sections shows the lateral-to-medial invagination of the endoderm (green arrows) and the fusion of the cardiogenic domains (red). Stages are indicated in number of somites (so), from E7.5 (0so) to E8 (6so). The process is similar in chick (a: amnion; endo: endoderm; f: foregut; nt: neural tube). (C) Examples of phenotypes of cardia bifida, when the cardiogenic domains do not fuse. Arrowheads show the two bilateral heart rudiments. C1 is an E11.5 embryo and C2, E9.5. Expression of cardiac troponin I in the myocardium is in blue (C1). Panels modified from Li et al. (2004) and Saga et al. (1999), with permission.

leads to cardia bifida. This is the case, for example, for Mesp1 in the mouse embryo (Fig. 4C) (Saga et al., 1999). As discussed previously, the planar cell polarity pathway also affects progenitor cell migration, with consequences for the formation of the heart tube. In Xenopus, Wnt11, which signals through the planar cell polarity pathway, is sufficient to drive cardiogenesis, epithelialization and formation of tubular structures in ventral marginal zone explants (Pandur et al., 2002). Embryos with reduced Wnt11-R activity exhibit defects in fusion of the heart tube (Garriock et al., 2005). Fibronectin, a component of the extracellular matrix present between endoderm and mesoderm, but also around heart progenitors, is another important factor for heart fusion. In mouse (George et al., 1993, 1997; Astrof et al., 2007), chick (Linask and Lash, 1988) and zebrafish (Trinh and Stainier, 2004) disruption of fibronectin impairs formation of the cardiac tube and often leads to cardia bifida. In zebrafish, upstream regulators of fibronectin, such as Mtx1 in the yolk syncitial layer (Sakaguchi et al., 2006) or Hand2 in cardiac progenitors (Yelon et al., 2000; Trinh et al., 2005) are necessary for heart fusion. Therefore, fibronectin may provide a substratum for cardiac progenitor cell migration, and/or permit the proper epithelial organization of cardiac cells, as explained above. Indeed, inactivation of fibronectin disrupts the epithelial organization of cardiac cells, as

seen in the absence of proteins of the Par complex such as has/aPKC and nok/Pals1 (Rohr et al., 2006), which are also required for proper formation of the heart tube. Similarly, interference with N-cadherin, which is important for the epithelial organization of cardiac cells, leads to cardia bifida (Nakagawa and Takeichi, 1997; Zhang et al., 2003). These data indicate that fusion of the heart tube requires organization of cardiac cells into a polarized epithelium. Cardia bifida is also seen when the furin gene is mutated in mouse (Roebroek et al., 1998). This gene, expressed in endoderm and mesoderm in the foregut region, encodes a protease which has been shown to be required for the maturation of Nodal in its active form, and which acts as well on other Tgf proteins. It is not known which Tgf protein is a substrate for furin in regulating fusion of the heart. Candidates include Tgf2 in the endoderm, Bmp2 and Tgf2 in the cardiogenic mesoderm, Bmp4/7 in myocardial cells and Tgf1 in endocardial cells. Although cardia bifida phenotypes all correspond to a disruption of heart tube fusion, the degree of development of the two bilateral heart tubes varies. In zebrafish (Fig. 3C) bilateral heart rudiments show correct patterning of atrial and ventricular myocytes in casanova mutants, whereas this is not the case in faust mutants (Reiter et al., 1999). An extreme case in mouse is that of the Foxp4/ mutant (Fig. 4C), which develops two correctly-patterned

Chapter  |  3.1  The Behavior of Cells that Form the Myocardial Compartments of the Vertebrate Heart

heart tubes that undergo looping until the E11.5 stage (Li et al., 2004). Active migration of myocardial cells has been shown to contribute to the fusion process in zebrafish. It is not known whether this is also the case in amniotes. Following the movement of cmlc2:egfp-expressing cells by timelapse, Holtzman and colleagues (2007) (Fig. 3B) (see also Chapter 1.4) have shown that myocardial cells migrate in two phases at the time of heart fusion. In the first phase, there is rapid migration of myocardial cells directly to the midline, whereas in the second phase, cells in the anterior and posterior region, but not in the central region, move at an angle to encircle endocardial progenitor cells. These phases can be genetically uncoupled. The cloche mutation, in a yet unidentified gene, abrogates formation of endothelium including endocardium, delays heart fusion and leads to the formation of a dysmorphic cardiac tube (Stainier et al., 1995). In these mutants, only the second phase of myocardial migration is compromised. In contrast, the first phase of myocardial migration is affected in miles apart mutants. The mil gene encodes a sphingosine-1-phosphate receptor which is required for heart tube fusion in a tissue other than the myocardial progenitor (Kupperman et al., 2000). In mouse, sphingosine-1-phosphate plays a role in later aspects of heart morphogenesis (Wendler and Rivkees, 2006). After the convergence of myocardial cells to the midline, it has recently been shown, in the zebrafish embryo, that myocardial cells undergo a complex phase of rotation towards the left and the anterior, known as cardiac jogging, during which the future venous pole becomes positioned on the left side. This is visible at the 25-somite stage (21.5 hpf). From a dorsal view, myocardial cells migrate as a coherent population (Rohr et al., 2008; Smith et al., 2008). In addition, on the dorsal–ventral axis, cells of the right-posterior myocardial epithelium involute ventrally (Rohr et al., 2008), such that the right myocardial epithelium will give rise to the ventral heart tube. Cardiac jogging depends on left–right signaling. Its orientation is randomized in southpaw mutants that affect a Nodal-related protein, and is directed by asymmetric BMP signaling. An effector of left–right signaling is hyaluronan synthase has2, which is involved in the remodeling of the extracellular matrix. Hyaluronan synthase has2 is expressed in a subset of future atrial cells that undergo leftward displacement, and is required for rotation of the cardiac cone. It is notable that the patterning of the zebrafish heart tube differs from that in amniotes; the venous pole is anterior in the zebrafish but posterior in amniotes and the ventral heart tube derives from the right myocardial epithelium in the zebrafish but from the left in amniotes (Stalsberg, 1969b; Campione et al., 2001). However, in both zebrafish and amniotes, formation of the cardiac tube is tightly linked to its looping. In conclusion, information about the process of heart fusion in the zebrafish embryo provides insights which it should be possible to explore in amniotes. The importance of correct cell migration, required for the positioning of the

203

first myocardial epithelium, is clear, as well as the necessity for correct positioning of the gut endoderm. However, the cellular mechanisms that underlie the convergence of the epithelial layers to the midline and their fusion remain obscure. During this process, the second heart field comes to lie behind the forming heart tube. The morphogenetic effects that this has on cardiac progenitor cell behavior require investigation.

VI.  Expansion of the cardiac chambers As development proceeds, regions of the myocardium become distinguishable along the arterial–venous axis of the cardiac tube. Cardiac chambers form by expansion of portions of the tube, whereas the outflow tract and the atrio­ventricular canal grow less and remain tubular. A range of molecular markers are known to be regionalized in the myocardium (Harvey, 2002; see Chapters 3.2–3.4). The analysis of mutant phenotypes and of molecular interactions between DNA and transcription factors has led to the definition of transcriptional modules (Srivastava, 2006) which may control different cell behavior in regions of the myocardium. What are the cellular mechanisms underlying the regionalized expansion of cardiac chambers?

VI.A.  Cell Proliferation Classically it is believed that morphogenesis is driven by differential rates of proliferation. Many experiments have been performed in the heart to investigate variations in the proliferation rate of myocardial cells. Decreased proliferation characterizes the early chick cardiac tube at the time of fusion (Stalsberg, 1969a; Soufan et al., 2006). This correlates with an increase in cell area/size. Subsequently, differential proliferation was observed in the chick (Thompson et al., 1990) and mouse (Sedmera et al., 2003) in the tube, with less proliferation in the regions which do not expand out of the tube, such as the inner curvature, the atrioventricular canal and the outflow tract. In these regions the T-box transcription factor, Tbx2, is specifically expressed, and may exert a negative effect on proliferation through transcriptional repression of Mycn (also known as N-myc) (Cai et al., 2005). The outer curvature of the cardiac tube is marked by the expression of other transcription factors, such as Hand1, Tbx20 or Tbx5, and Tbx5 target genes (Bruneau et al., 2001) such as those encoding Nppa (also known as ANF), the gap junction protein Cx40, as well as the cytoskeletal protein Chisel, which potentially affects cell behavior. The expression of these genes presages differential cell proliferation and has led to the “ballooning model” for chamber expansion (Christoffels et al., 2004) (see Chapter 3.2). There is reciprocal repression between genes expressed on the inner and outer curvature,

204

such as Tbx2 and Tbx20, respectively (Singh et al., 2005). However, it is not clear how transcriptional regionalization is initiated, or how regulatory factors control proliferation and cell size. Once the chambers attain their correct size, the question of how proliferation is arrested is also important. One indication comes from the identification of the heart restricted helicase, Champ, which probably negatively-regulates cell-cycle progression (Liu et al., 2001; Liu and Olson, 2002). This gene is a Mef2c target, and its gradual accumulation in myocardial cells may limit growth. Mef2c is itself a target of signal transduction pathways which control its activity (McKinsey et al., 2002) and may therefore act as an intermediary in response to signals that restrict proliferation. In the more mature heart tube, thickening of the ventricular wall is accompanied by decreased cell proliferation in the trabeculated layer (Mikawa et al., 1992; Sedmera et al., 2003). Signaling between endocardium and myocardium is essential for the formation of the trabeculae. Neuregulin, secreted by the endocardium, activates the ErbB receptors in the myocardium and regulates the maturation of trabeculae (Gassmann et al., 1995; Lee et al., 1995; Meyer and Birchmeier, 1995). EphrinB2 and its receptor EphB4, although both expressed in the endocardium, have also been shown to participate in this signaling (Wang et al., 1998; Gerety et al., 1999). The molecular cascade has been deciphered recently in an elegant study, which has shown that Notch1 and Delta1/4 in the endocardium act upstream in the cascade (Grego-Bessa et al., 2007). EphrinB2 is a direct target of Notch, and intervenes upstream of Neuregulin to regulate the maturation of trabeculae. Bmp10, which is independently regulated by Notch and expressed in the myocardium, balances this effect by controlling cell proliferation (Chen et al., 2004). This regulation of transmural growth of the myocardium by an interaction between endocardium and myocardium appears to be common to the zebrafish embryo. A novel transmembrane protein, Heg, which is expressed in the endocardium, has indeed been shown to be required for the formation of several concentric layers of myocardium (Mably et al., 2003). In contrast, signaling between the epicardium and myocardium is required for the growth of the compact myocardial layer. This may be mediated by retionoic acid signaling, with the enzyme Raldh2 and the receptor RXR, expressed in the epicardium, required to secrete a yet unknown cardiomyotrophic signal, which is potentially Wnt 9b or an Fgf (reviewed in Manner, 2006; see Chapter 5.2). Another important aspect of chamber maturation is the establishment of boundaries, with the formation of septa between the different compartments of the heart. A number of models had been proposed to account for septa formation, including folding of the myocardial wall. However, experiments in the chick embryo showed, using radioactive nylon probes introduced into the myocardium,

PART  |  3  Patterning of the Early Heart Tube

that within 250 m on either side of the site of ventricular septation, trabeculae will aggregate and protrude into the ventricular cavity (Harh and Paul, 1975). This movement is accompanied by cell division and two proliferative centers were identified in the ventricular part of the looped heart tube (Rychter et al., 1979). More recently, Lysozyme M expression has been found to mark cells in the forming ventricular septum and targeting the gene with GFP has made it possible to follow their movement (Stadtfeld et al., 2007). Septal formation therefore involves directed cell movement and underlying proliferation of cells at specific sites in the myocardium. In the case of the interventricular septum, transgene expression patterns suggested varying contributions from both right and left ventricular myocardium along the septum (Franco et al., 2006), substantiated by retrospective clonal analysis which demonstrates that clones of cells in the septum can extend into the myocardium on either side. Clones tend to lie along the axis of the septum, revealing how it grows (Meilhac et al., 2004b; Zaffran et al., 2004). Mutation of the gene encoding tolloid-like-1 (Tll1), a metalloprotease involved in Bmp regulation, results in incomplete interventricular septation (Clark et al., 1999), suggesting that Bmp signaling, which can inhibit proliferation in other contexts (Prall et al., 2007) is involved in this process. Manipulation of genes associated with chamber identity also affects septum formation. This is exemplified by Tbx5 expressed in left ventricular myocardium. In the absence of Tbx5 in the mouse embryo, there is no interventricular septum, and indeed no left ventricle (Bruneau et al., 2001). Partial expression in the right ventricle of the chick will shift the boundary to give a bigger left ventricle and smaller right ventricle, and focal expression of Tbx5 generates ectopic septal boundaries (Takeuchi et al., 2003). Hand1, another left ventricular factor, also affects septum positioning (Togi et al., 2004). Another boundary in the myocardium lies in the atrioventricular canal separating atrial and ventricular tissues; this is also regulated by T-box genes. Tbx2 is a marker of the atrioventricular canal in mouse. It represses chamberspecific gene expression – both atrial and ventricular – (Christoffels et al., 2004; Harrelson et al., 2004), and also cell proliferation by direct binding to N-myc (Cai et al., 2005). Similarly, in the zebrafish, expression of tbx2b marks the atrioventricular canal, which is directly regulated by tbx5 and foxn4. Mutation of slipjig/foxn4 specifically ablates the specification of the atrioventricular canal, without disrupting myocardial development (Chi et al., 2008). Notch signaling is well-known to regulate the formation of boundaries between cell compartments (Bray, 1998). In the mouse heart, conditional expression of activated Notch, using Mesp1-Cre, does not appear to interfere with the onset of cardiogenesis, but leads to a right-ward shift in the interventricular septum, as well as affecting the maturation of ventricular myocardium (Watanabe et al., 2006). Notch signaling is also implicated in atrial–ventricular ­boundary

Chapter  |  3.1  The Behavior of Cells that Form the Myocardial Compartments of the Vertebrate Heart

formation (Timmerman et al., 2004; Rutenberg et al., 2006). As in the case of interventricular septum formation, interference with the regionalized expression of regulatory genes affects the atrioventricular boundary, as shown for Hey1 and Hey2 which repress Tbx2 (Kokubo et al., 2007). In the chick heart, Hey1 appears to be a direct effector of Notch signaling (Rutenberg et al., 2006). Cushion formation, which depends on Bmps and is affected by Notch signaling, is closely linked to septation, most evident in the case of atrial–ventricular septal formation (see Chapters 6.1, 6.2). While the regulatory genes and some of the signaling pathways involved in the positioning of boundaries in the heart have been identified, the downstream mechanisms responsible for directed proliferation of selected myocardial cells to form the septum remain obscure.

VI.B.  Other Aspects of Cell Behavior At the onset of cardiac chamber formation, differential proliferation provides an explanation for the bulging of this myocardium from the outer curvature of the heart tube. However, there is no apparent variation in proliferation rates of myocardial cells within the forming chambers (Sissman, 1966). Thus, differential proliferation is not sufficient to account for the particular geometry of individual cardiac chambers. Another aspect of cell growth, its orien­ tation, could explain chamber morphogenesis. Oriented cell growth is known to be involved in the morphogenesis of other systems, such as the Drosophila wing (Resino et al., 2002) or the Antirrhinum petal (Rolland-Lagan et al., 2003). In the mouse heart, specific orientation of myocardial cell growth has been shown to ­characterize regions of the heart tube with distinct morphologies (Meilhac et al., 2004b). Whereas the left ventricle, at E10.5, shows cluster orientations consistent with growth by ballooning out from the tube, oriented cell growth in the right ventricle points to enlargement of the tube by circumferential growth, emanating from the inner curvature (Fig. 5A,B). In the outflow tract, oriented growth is axial, in agreement with the elongation of this region. Oriented growth is already detectable in the primitive cardiac tube at E8.5, and is maintained after septation at E14.5 with additional modulations, such as novel orientations of small clones, or secondary orientations within large clones. These latter are perceptible as early as E10.5 in the ventricles, but only later in the atrioventricular canal and atrial appendages. Secondary orientation of clonal cell growth prefigures myofiber architecture of the heart (Fig. 5C) (Meilhac et al., 2003), and this has also been described in the chick heart (Mikawa et al., 1992). These results show that oriented growth of myocardial cells may underlie different aspects of mouse heart morphogenesis and that the myocardium, from the time of its formation, is a polarized tissue. However, the signals that polarize the myocardium on its surface remain to be investigated. Oriented growth

205

may be based on the orientation of the mitotic spindle or on oriented postmitotic rearrangements of cells. Both of these mechanisms have been involved in the morphogenesis of other systems. On the one hand, oriented mitoses shape clones in the Drosophila wing (Baena-Lopez et al., 2005); orientation of the mitotic spindle may be a consequence of cell shape (also known as “Hertwig’s rule,” see for example Strauss et al., 2006), of extracellular matrix distribution (Thery et al., 2005) or of signaling from the planar cell polarity pathway (Gong et al., 2004). On the other hand, oriented intercalation of cells underlies tissue elongation, such as in the Drosophila germ-band (Bertet et al., 2004), the Drosophila hindgut (Johansen et al., 2003) or the headto-tail body axis in Xenopus (Keller, 2002). Morphogenesis of the heart also involves cell shape changes. In the chick, looping of the cardiac tube is accompanied by the elongation of cells on the outer curvature. Orientation of cells remains random in the outer curvature, whereas it becomes circumferential in the inner curvature (Manasek et al., 1972). At the same time, cells of the outer curvature also become larger (Soufan et al., 2006). In zebrafish, the ventricular chamber is bean-shaped at 52 hpf with a concave face or outer curvature, and convex face or inner curvature (Auman et al., 2007) (see also Chapter 1.4). In the outer curvature cells become larger and elongated, whereas in the inner curvature cells remain cuboidal, as in the linear heart tube. Elongation of cells is oriented perpendicular to the arterial–venous axis. Taken together, these data suggest that curvature of cardiac tissue in the chick heart tube or in the zebrafish ventricle is associated with cell shape changes, including elongation and enlargement. Cell interactions in the heart are influenced by the extracellular matrix, known as cardiac jelly, which is abundant between the endocardium and myocardium in the early cardiac tube. It is composed of glycosaminoglycans (GAG) such as Hyaluronan, proteoglycans such as Vcan (also known as versican or Cspg2), Agc1 (aggrecan) and Hspg2 (perlecan) and proteins such as Fibronectin, Collagen and Laminin (Hurle et al., 1980). GAG osmotically attract water, and thus provide a compressive strength. The model of Manasek (1983) for cardiac chamber expansion is based on this property. The model was also inspired by an analogy with animals such as nematodes, which are shaped by a high internal hydrostatic pressure exerting tension on a fibrous skeleton. According to the model, the cardiac jelly, through its GAG, generates an internal pressure which repercusses on the myofibrillar structure of the myocardium. In the chick embryo, destruction of the cardiac jelly by enzymatic treatments may cause changes in the shape of the heart (Nakamura and Manasek, 1981). However, these crude treatments also impair the histotypical organization of cells and the availability of growth factors, which are normally bound to the cardiac jelly. Analyses of mutant mice have demonstrated the involvement of the cardiac jelly in chamber morphogenesis. Inactivation of

206

PART  |  3  Patterning of the Early Heart Tube

(A) OFT RA

BA

LA

AVC RV

LV IV

(B) OFF OFT LV RV E10.5-540

E10.5-496

E10.5-682

(C) RV LA

LV

E14.5-795

E14.5-696

P7-391

Figure 5  Oriented clonal growth during the expansion of the cardiac chambers. (A) Myocardial cell growth shows orientations specific to different cardiac regions, as schematized in ventral views of E10.5 mouse hearts represented with (left), or without (right), the outflow tract (OFT). Orientations along the arterial–venous axis of the cardiac tube are in blue, circumferential in green and other types in red. (B) Examples of -galactosidase-positive clones in E10.5 c-actinnlaacZ1.1/ hearts demonstrate oriented cell growth. Panels modified from Meilhac et al. (2004), with permission. (C) Examples of -galactosidase-positive clusters in E14.5 and postnatal (P)7 c-actinnlaacZ1.1/ hearts showing secondary orientations of cell growth (short red lines), which at P7 parallel myofiber orientations. The main orientation of clusters of labeled cells is represented by a dashed line (AVC: atrioventricular canal; BA: body of the atrium; IV: interventricular region; LA: left atrium; LV: left ventricle; RA: right atrium; RV: right ventricle). Panels modified from Meilhac et al. (2003, 2004), with permission.

the genes encoding the proteoglycan Vcan (Yamamura et al., 1997; Mjaatvedt et al., 1998) or Hyaluronan synthase-2 (Camenisch et al., 2000) leads to the formation of a truncated heart, with a marked reduction in the outflow tract and right ventricular regions, and a small common atrium. This is reminiscent of second heart field phenotypes raising the possibility that myocardium from this source may secrete distinct cell matrix components. The cardiac jelly has also been shown to provide a microenvironment necessary for the trabeculation of the myocardial wall. Derepression of ADAMTS1, a matrix metalloproteinase secreted by the endocardium which can cleave versican, causes premature breakdown of the cardiac jelly and leads to trabeculation defects (Stankunas et al., 2008). In this context, the cardiac jelly may influence the cellular movements of trabeculations or affect the diffusion or maturation of proliferation signals between the endocardium and the myocardium (see Section VI.A).

Finally, blood flow is an important modulator of the expansion of cardiac chambers. Blood flow is progressively set up in the cardiac tube. Pumping potential is evidenced by contractions of cardiomyocytes which are detected in the 3-somite mouse embryo (Nishii and Shibata, 2006), although they are not yet synchronized. Blood cells, which are initially produced in the extra-embryonic yolk sac, are present at significant levels in the heart later, around the 8-somite stage, indicating the presence of a systemic blood flow (McGrath et al., 2003). It is only around the 20-somite stage, with the appearance of valves and of the conduction system, that blood flow becomes unidirectional and more efficient. These observations suggest that blood flow is not a major component of the initial expansion of cardiac chambers (see also Zak et al., 1979). This is consistent with the phenotype of the Ncx1 mutant, a gene that encodes a Na/Ca2 exchanger, which abrogates cardiac contraction and blood flow (Wakimoto et al., 2000).

Chapter  |  3.1  The Behavior of Cells that Form the Myocardial Compartments of the Vertebrate Heart

(A)

207

(B)

Figure 6  Expansion of the cardiac chambers in the absence of blood flow. (A) Heterozygote Ncx1lacZ/ and homozygote Ncx1lacZ/lacZ mutant mice are shown at E9. (B) Sections of their hearts at E10. Gross morphology of homozygote hearts is preserved, in particular cardiac chamber expansion, indicating that blood flow which is disrupted in the absence of the Na/Ca2 exchanger Ncx1, is dispensable for early heart morphogenesis. Note that homozygote embryos die between E9 and E10; embryos stop growing and appear smaller at E10 (LA: left atrium; LV: left ventricle; pLA: primitive left atrium; pRA: primitive right atrium; pV: primitive ventricle; RA: right atrium; RV: right ventricle). Panels modified from Koushik et al. (2001), with permission.

These embryos, which die between E9 and E10, initially have a heart which appears normal, with expanding cardiac chambers (Fig. 6) (Koushik et al., 2001). However, later remodeling of the heart and refinement of the morphology of the cardiac chambers are influenced by blood flow. Cardiac hypertrophy, for example, is known to be a response to hemodynamic overload (Zak et al., 1979). Formation of the outflow region in the zebrafish heart (Hove et al., 2003), or that of cardiac chambers in the chick (Clark, 1969), is impaired when blood flow is experimentally occluded. The specific cell shape change in the outer curvature of the zebrafish ventricle is sensitive to alteration in blood flow (Auman et al., 2007). Gene expression and cytoskeletal organization of cells are sensitive to the shear stretch created by blood flow (Yamazaki et al., 1996). Indeed, Ncx1/ hearts have an abnormal myofibrillar organization (Koushik et al., 2001). In keeping with the role of oriented growth in heart morphogenesis, orientation of cardiomyocytes can be directed by stretch (Kada et al., 1999). In conclusion, differential proliferation and oriented cell growth probably underlie the shaping of the cardiac chambers. The signaling pathways that trigger these changes are not well-characterized and it is only at later stages, during trabeculation, for example, that the sophisticated mechanisms controlling this process and myocardial cell proliferation have been described. Pressure exerted by cardiac jelly or by the blood flow can affect myocardial cell shape and chamber morphogenesis. However, little is known about the signaling pathways involved. This is also the case for oriented cell growth. Elucidation of not only the source, but also the mechanisms that activate the processes that shape the heart, is a major challenge for the future.

VII.  Outflow tract morphogenesis During the fetal period in amniotes, the heart tube is remodeled such that the systemic and pulmonary blood flows become separated. This involves septation of the cardiac chambers and structural reorganization at the arterial and venous poles. In this section we shall comment on aspects of outflow tract morphogenesis that illustrate points about myocardial cell behavior. The migration of neural crest into the outflow tract (see Chapter 7.2), external influences such as left–right signaling (see Chapter 4.2), as well as the genetic networks that shape the arterial pole of the heart (see Chapter 2.2), are discussed elsewhere. As development proceeds, the outflow tract is divided into the base of the pulmonary trunk and the aorta. Remodeling of the outflow tract not only depends on the formation of a septum, but also on the rotation of the myocardial wall. Polarized behavior of cardiac cells is involved in these processes.

VII.A.  Rotation of Outflow Tract Myocardium Radioactive tattooing in the chick embryo had shown rotation of the outflow tract (Thompson et al., 1987) which was also suggested by anatomical observations on human embryos (Lomonico et al., 1986) where arterial pole abnormalities are associated with arrested rotation (Bostrom and Hutchins, 1988; Lomonico et al., 1988). More recently, labeled progenitor cells in the right side of the chick secondary heart field were shown to contribute

208

PART  |  3  Patterning of the Early Heart Tube

to the left or pulmonary side of the outflow tract, leading to the proposition that this spiraling movement may underlie the rotation of the outflow tract required for the correct alignment of the great arteries (Ward et al., 2005). In the mouse embryo, clockwise rotation of the myocardium has been visualized in chase experiments using transgenic markers and by DiI tracing (Fig. 7A,B) (Bajolle et al., 2006). It is also indicated by spiraling of myocardial clones in this region of the heart (Meilhac et al., 2004b). Rotation is detected by E10.5, and its perturbation in Pitx2

A1

or Pax3 mutant embryos suggests that the malpositioning of the aorta and pulmonary trunk, seen in these mutants (Fig. 7C), may be partly due to an earlier effect on myocardial rotation. Lack of Pax3 or Pitx2 affects cardiac neural crest migration and left–right signaling respectively, implicating these processes in outflow tract rotation. The cellular mechanism underlying rotation of the outflow tract myocardium is unknown. However, it parallels the spiraling of the underlying ridges of mesenchymal tissue, which are the primordia of the septum of the outflow tract.

A2

ra

la

rv

lv

A3

B1

ao pt

B2 ba1 ba1

lv

ra

la

96-16 C1

ao

ao

pt

rv

Sp, 96-16 C2

Pitx2δc, 96-16 C3

pt

PTA

PTA

ao pt

pt ao

Figure 7  Rotation of the outflow tract. (A) Expression of the 96-16 transgene during remodeling of the outflow tract (OFT). This Myf5-nlacZ transgene, which is ectopically expressed in the heart due to an insertion site effect, provides a marker of a subdomain of OFT myocardium. Ventral views of hearts at stage E10.5 (A1), E11.5 (A2) and E12.5 (A3), and schematic representation of the -galactosidase activity showing the rotation of the -galactosidase-positive myocardium. (B) DiI labeling of the left OFT myocardium at E9.5 (B1) is found in the right wall of the OFT after 24 hours of culture (B2), showing the rotation of the OFT myocardium. Analysis of the expression profile of the 96-16 transgene in wt (C1), Splotch (Sp) (C2) and Pitx2c (C3) mutant hearts at E14.5 (C1–2) and E15.5 (C3). Insets show labeled cells in a cranial view. In the Splotch (Pax3) mutant heart with persistent truncus arteriosus (PTA), and in the Pitx2c mutant heart with transposition of the great arteries, -galactosidase-positive cells are mislocated in the outflow tract, indicating impairment of the rotation of the outflow tract myocardium (ao: aorta; ba: branchial arches; la: left atrium; lv: left ventricle; oft: outflow tract; pt: pulmonary trunk; ra: right atrium; rv: right ventricle). Panels modified from Bajolle et al. (2006), with permission.

Chapter  |  3.1  The Behavior of Cells that Form the Myocardial Compartments of the Vertebrate Heart

These spiraling ridges, which are first visible at E11.5 in the mouse (Fananapazir and Kaufman, 1988), are thought to be initially induced by the spiraling of the blood flow. Subsequently, mesenchymal cells become reoriented along the ridges. Mutation of hspg2 (Costell et al., 2002), which encodes a component of the extracellular matrix also known as perlecan, leads to an overproduction of cushion mesenchymal cells in the outflow tract, which precludes the formation of the spiral ridges.

VII.B.  Septation The septum of the outflow tract derives from different populations of cells: cushion mesenchyme (see Chapters 6.1, 6.2); cardiac neural crest (see Chapter 7.2); and myocardium. Cushion mesenchyme arises from an epitheliomesenchymal transition of endocardial cells induced by myocardial signals such as TGF (Brown et al., 1999). It invades the cardiac jelly from E9 in the mouse (Gitler et al., 2003) and St17 in the chick (Markwald et al., 1977). Cardiac neural crest cells delaminate from the neural tube and migrate through branchial arches 3, 4 and 6. They invade cushion tissues of the outflow tract at E10 in the mouse (Jiang et al., 2000) and St22 in the chick (Waldo et al., 1999). Migration of cardiac neural crest cells is regulated by signals such as the secreted protein, semaphorin3C, produced by the myocardium of the outflow tract (Feiner et al., 2001) and its receptor, plexinA2, located in cardiac neural crest cells (Brown et al., 2001). From E12.5 in the mouse and St28 in the chick, myocardial cells also invade cushion tissues, a process known as myocardialization (van den Hoff et al., 2004). This movement of myocardial cells involves active cell migration with cellular protrusions (Phillips et al., 2005). The migration is polarized from the myocardial wall to the cushion tissues, and polarity seems to be established by the planar cell polarity pathway (Fig. 8A). Vangl2, the vertebrate ortholog of strabismus, encoding a transmembrane protein that participates in the noncanonical Wnt/planar cell polarity pathway, is expressed in the mouse outflow tract myocardium. Its spontaneous mutation in loop-tail (Lp) mice leads to cardiac defects, including outflow tract defects such as common arterial trunk and double-outlet right ventricle (Fig. 8B) (Henderson et al., 2001). In the mutant cardiac neural crest migration is not affected, whereas myocardialization of the cushion tissues is impaired. Myocardial cell protrusions are absent at E12.5 and transcription of RhoA, which encodes an effector of the planar cell polarity pathway involved in reorganization of the cytoskeleton, is downregulated (Phillips et al., 2005). These results show that Vangl2 is important for the polarized migration of myocardial cells during myocardialization of the outflow tract septum. Other genes encoding components of the noncanonical Wnt/planar cell polarity pathway are expressed

209

in the mouse outflow tract, including Wnt11 in the myocardium, Wnt5a in the cushion mesenchyme, Dvl2 in the myocardium, Rock1 in myocardial cells that extend into cushion tissues and RhoA in the cushion mesenchyme (Phillips et al., 2005). Wnt11/ (Zhou et al., 2007; Nagy et al., 2009), Wnt5a/ (Oishi et al., 2003) and Dvl2/ (Hamblet et al., 2002) mutant mice exhibit outflow tract defects, including transposition of the great arteries and double-outlet right ventricle. In addition, the cytoarchitecture of myocardial cells is affected in Wnt11/ mutants (Zhou et al., 2007). However, it is not clear whether Dvl2 in the outflow tract myocardium is acting in the noncanonical Wnt pathway (Henderson et al., 2006) and/or in the canonical Wnt pathway which regulates the proliferation of cardiac neural crest cells via Pitx2 (Hamblet et al., 2002; Kioussi et al., 2002). Nonmuscle Myosin II-B, an effector of the planar cell polarity pathway, is necessary for the formation of the outflow tract, as its mutation results in outflow tract defects such as tetralogy of Fallot and double-outlet right ventricle (Tullio et al., 1997). Another planar cell polarity pathway, including Dchs1, the ortholog of Dachsous, Fjx1, the ortholog of four-jointed and Fat-j, the ortholog of Fat, are expressed in the mouse outflow tract (Rock et al., 2005). Taken together, these data suggest that the planar cell polarity pathway plays an important role during the septation of the mouse outflow tract, at least in regulating the polarized migration of myocardial cells. This is the first well-characterized evidence for planar cell polarity in the heart, showing that the myocardium is not only polarized transmurally on an apico–basal axis, as discussed for the myocardial epithelium of the cardiac crescent (see Section III).

VIII.  Concluding remarks In this chapter we have discussed the behavior of myocardial progenitor cells, and of myocardial cells in the crescent, in the formation of the cardiac tube, and in chamber expansion, as well as later in chamber maturation and outflow tract morphogenesis. Other important steps in cardiogenesis, such as cardiac looping or valve formation, are presented in other chapters (see chapters in Parts 4 and 7). Progress in understanding the behavior of cells in the heart has depended on different approaches that include anatomical observations, cell labeling with markers for time-lapse experiments, cell fate analyses and genetic manipulations. Among the questions it will be important to answer in the future are those of how progenitor cells of the first and second heart fields are distributed in the early mouse embryo, how these cells migrate, how myocardial cells intercalate, how they orient their mitoses and what polarizes the myocardium during chamber formation. Obtaining answers to these questions will be facilitated by improvement in the techniques for generating clones of cells in the chick or for prospective ­lineage

210

PART  |  3  Patterning of the Early Heart Tube

A-Noncanonical Wnt /Planar cell polarity pathway Wnt5a/11 VangL

Celsr

Fz Dvl

Pk

Dgo

Daam1 RhoA

Rac

JNK

mDIA

ROCK

Gene transcription

Actin polymerization

Myosin II Myosin activity

Cytoskeletal changes Effects on cell adhesion, motility, polarity B1

B3

Lp/+

ao

ao

B5

pt

rv B2

B4

B6

pt

Lp/Lp

ao

ao

Figure 8  Septation of the outflow tract. (A) Schematic representation of the Wnt/ planar cell polarity pathway, which is conserved between vertebrates and invertebrates (Celsr: cadherin EGF LAG seven-pass Gtype receptor, flamingo homolog; Daam: disheveled associated activator of morphogenesis; mDIA: diaphanous homolog; Dgo: diego; Dvl: disheveled homolog; Fz: frizzled homolog (Wnt receptor); JNK: c-Jun Nterminal kinases; Pk: prickle homolog; ROCK: Rho-associated kinases; VangL: van gogh like, strabismus homolog). (B) The Wnt/planar cell polarity pathway is involved in mouse heart morphogenesis, and is required for the correct septation of the outflow tract. Normal spiraling of the great arteries can be seen in wildtype fetuses at E15.5 (B1), whereas the bases of the arteries are para­llel and both exit from the right ventricle in the Lp/Lp fetuses (B2), which exhibit a spontaneous mutation in the Vangl2 gene. (B3–4) Myocardial cells, stained in brown for -smooth muscle actin, extend from the outflow tract wall into the outlet septum in E13.5 wild-type fetuses (arrowheads). In contrast, fewer myocardial cells extend into the septum in Lp/Lp mutant fetuses. (B5–6) Myocardializing muscle cells in the outflow tract wall extend membrane protrusions in the direction of cell movement in wild-type fetuses at E12.5 (arrowheads). This process is inhibited in Lp/Lp mutant fetuses and the cells remain nonpolarized (ao: aorta; pt: pulmonary trunk; rv: right ventricle). Panels modified from Henderson et al. (2006), with permission.

rv

analysis in the mouse embryo. Genetic approaches have led to striking progress in defining regulatory networks that control cardiogenesis, and identification of target genes for transcriptional regulators is now beginning to reveal effector genes that are important for cardiogenesis. The comparison between different animal models is important to understanding how the underlying cellular mechanisms and intercellular signaling that are common to vertebrates, control the contribution of progenitor cells, and shape the myocardium of the heart. In this context, the zebrafish appears as a major model to characterize cell behavior in detail, since it is

more amenable to cell injection and time-lapse experiments and to the powerful combination of imaging and genetic approaches. Imaging technologies are fundamental to the study of cell behavior. Three-dimensional imaging of whole organs such as the heart, which has a complex geometry, is now available with optical projection tomography (Sharpe et al., 2002), selective plane illumination (Huisken et al., 2004; Scherz et al., 2008) and high-resolution episcopic microscopy (Weninger et al., 2006). To study the dynamics of cell behavior in the mouse heart, live imaging is becoming a reality with the improvement of confocal microscopy,

Chapter  |  3.1  The Behavior of Cells that Form the Myocardial Compartments of the Vertebrate Heart

and with the realignment of images taken from a moving object such as the beating heart (Liebling et al., 2005, 2006). Integration of three-dimensional shapes with molecular and cellular data will depend on computer modeling, which has proved to be a productive approach to decipher plant morphogenesis (Prusinkiewicz and Rolland-Lagan, 2006). The application of this technology to the heart opens new perspectives, and will be valuable in defining the causal relationship between observed aspects of cell behavior and their impact on the shape of the heart.

Acknowledgments MB’s laboratory is supported by the Pasteur Institute and the Centre National de la Recherche Scientifique (URA 2578). It also benefits from support from the European Community’s Sixth Framework Programme contract (“Heart Repair”) – SHM-CT-2005-018630. S.M.M. is a research fellow in the Institut National de la Santé et de la Recherche Médicale.

References Ai, D., Fu, X., Wang, J., Lu, M.F., Chen, L., Baldini, A., Klein, W.H., Martin, J.F., 2007. Canonical Wnt signaling functions in second heart field to promote right ventricular growth. Proc. Natl. Acad. Sci. USA 104, 9319–9324. Alexander, J., Rothenberg, M., Henry, G.L., Stainier, D.Y., 1999. Casanova plays an early and essential role in endoderm formation in zebrafish. Dev. Biol. 215, 343–357. Alsan, B.H., Schultheiss, T.M., 2002. Regulation of avian cardiogenesis by Fgf8 signaling. Development 129, 1935–1943. Astrof, S., Kirby, A., Lindblad-Toh, K., Daly, M., Hynes, R.O., 2007. Heart development in fibronectin-null mice is governed by a genetic modifier on chromosome four. Mech. Dev. 124, 551–558. Auman, H.J., Coleman, H., Riley, H.E., Olale, F., Tsai, H.J., Yelon, D., 2007. Functional modulation of cardiac form through regionally confined cell shape changes. PLoS Biol. 5, e53. Baena-Lopez, L.A., Baonza, A., Garcia-Bellido, A., 2005. The orientation of cell divisions determines the shape of Drosophila organs. Curr. Biol. 15, 1640–1644. Bagatto, B., Francl, J., Liu, B., Liu, Q., 2006. Cadherin2 (N-cadherin) plays an essential role in zebrafish cardiovascular development. BMC Dev. Biol. 6, 23. Bajolle, F., Zaffran, S., Kelly, R.G., Hadchouel, J., Bonnet, D., Brown, N.A., Buckingham, M.E., 2006. Rotation of the myocardial wall of the outflow tract is implicated in the normal positioning of the great arteries. Circ. Res. 98, 421–428. Bajolle, F., Zaffran, S., Meilhac, S.M., Dandonneau, M., Chang, T., Kelly, R.G., Buckingham, M.E., 2007. Myocardium at the base of the aorta and pulmonary trunk is prefigured in the outflow tract of the heart and in subdomains of the second heart field. Dev. Biol. 313, 25–34. Bertet, C., Sulak, L., Lecuit, T., 2004. Myosin-dependent junction remodelling controls planar cell intercalation and axis elongation. Nature 429, 667–671. Bisaha, J.G., Bader, D., 1991. Identification and characterization of a ventricular-specific avian myosin heavy chain, VMHC1: expression in differentiating cardiac and skeletal muscle. Dev. Biol. 148, 355–364.

211

Bostrom, M.P., Hutchins, G.M., 1988. Arrested rotation of the outflow tract may explain double-outlet right ventricle. Circulation 77, 1258–1265. Brade, T., Gessert, S., Kuhl, M., Pandur, P., 2007. The amphibian second heart field: Xenopus islet-1 is required for cardiovascular development. Dev. Biol. 311, 297–310. Brand, T., 2003. Heart development: molecular insights into cardiac specification and early morphogenesis. Dev. Biol. 258, 1–19. Bray, S., 1998. Notch signalling in Drosophila: three ways to use a pathway. Semin. Cell Dev. Biol. 9, 591–597. Brown, C.B., Boyer, A.S., Runyan, R.B., Barnett, J.V., 1999. Requirement of type III TGF-beta receptor for endocardial cell transformation in the heart. Science 283, 2080–2082. Brown, C.B., Feiner, L., Lu, M.M., Li, J., Ma, X., Webber, A.L., Jia, L., Raper, J.A., Epstein, J.A., 2001. PlexinA2 and semaphorin signaling during cardiac neural crest development. Development 128, 3071–3080. Bruneau, B.G., Nemer, G., Schmitt, J.P., Charron, F., Robitaille, L., Caron, S., Conner, D.A., Gessler, M., Nemer, M., Seidman, C. E., Seidman, J.G., 2001. A murine model of Holt-Oram syndrome defines roles of the T-box transcription factor Tbx5 in cardiogenesis and disease. Cell 106, 709–721. Buckingham, M., Meilhac, S., Zaffran, S., 2005. Building the mammalian heart from two sources of myocardial cells. Nat. Rev. Genet. 6, 826–835. Cai, C.L., Liang, X., Shi, Y., Chu, P.H., Pfaff, S.L., Chen, J., Evans, S., 2003. Isl1 identifies a cardiac progenitor population that proliferates prior to differentiation and contributes a majority of cells to the heart. Dev. Cell 5, 877–889. Cai, C.L., Zhou, W., Yang, L., Bu, L., Qyang, Y., Zhang, X., Li, X., Rosenfeld, M.G., Chen, J., Evans, S., 2005. T-box genes coordinate regional rates of proliferation and regional specification during cardiogenesis. Development 132, 2475–2487. Camenisch Jr, T.D., Spicer, A.P., Brehm-Gibson, T., Biesterfeldt, J., Augustine, M.L., Calabro, A., Kubalak, S., Klewer, S.E., McDonald, J.A., 2000. Disruption of hyaluronan synthase-2 abrogates normal cardiac morphogenesis and hyaluronan-mediated transformation of epithelium to mesenchyme. J. Clin. Invest. 106, 349–360. Campione, M., Ros, M.A., Icardo, J.M., Piedra, E., Christoffels, V.M., Schweickert, A., Blum, M., Franco, D., Moorman, A.F., 2001. Pitx2 expression defines a left cardiac lineage of cells: evidence for atrial and ventricular molecular isomerism in the iv/iv mice. Dev. Biol. 231, 252–264. Chen, H., Shi, S., Acosta, L., Li, W., Lu, J., Bao, S., Chen, Z., Yang, Z., Schneider, M.D., Chien, K.R., Conway, S.J., Yoder, M.C., Haneline, L.S., Franco, D., Shou, W., 2004. BMP10 is essential for maintaining cardiac growth during murine cardiogenesis. Development 131, 2219–2231. Chi, N.C., Shaw, R.M., De Val, S., Kang, G., Jan, L.Y., Black, B.L., Stainier, D.Y., 2008. Foxn4 directly regulates tbx2b expression and atrioventricular canal formation. Genes Dev. 22, 734–739. Christoffels, V.M., Hoogaars, W.M., Tessari, A., Clout, D.E., Moorman, A.F., Campione, M., 2004. T-box transcription factor Tbx2 represses differentiation and formation of the cardiac chambers. Dev. Dyn. 229, 763–770. Ciruna, B.G., Schwartz, L., Harpal, K., Yamaguchi, T.P., Rossant, J., 1997. Chimeric analysis of fibroblast growth factor receptor-1 (Fgfr1) function: a role for FGFR1 in morphogenetic movement through the primitive streak. Development 124, 2829–2841.

212

Clark, E.B., 1969. Effect of inflow stream alteration on the morphogenesis of the chick heart. Anat. Rec. 163, 170–176. Clark, T.G., Conway, S.J., Scott, I.C., Labosky, P.A., Winnier, G., Bundy, J., Hogan, B.L., Greenspan, D.S., 1999. The mammalian Tolloid-like 1 gene, Tll1, is necessary for normal septation and positioning of the heart. Development 126, 2631–2642. Cohen, E.D., Wang, Z., Lepore, J.J., Lu, M.M., Taketo, M.M., Epstein, D.J., Morrisey, E.E., 2007. Wnt/beta-catenin signaling promotes expansion of Isl-1-positive cardiac progenitor cells through regulation of FGF signaling. J. Clin. Invest. 117, 1794–1804. Cohen-Gould, L., Mikawa, T., 1996. The fate diversity of mesodermal cells within the heart field during chicken early embryogenesis. Dev. Biol. 177, 265–273. Costell, M., Carmona, R., Gustafsson, E., Gonzalez-Iriarte, M., Fassler, R., Munoz-Chapuli, R., 2002. Hyperplastic conotruncal endocardial cushions and transposition of great arteries in perlecan-null mice. Circ. Res. 91, 158–164. David, N.B., Rosa, F.M., 2001. Cell autonomous commitment to an endodermal fate and behaviour by activation of Nodal signalling. Development 128, 3937–3947. David, R., Brenner, C., Stieber, J., Schwarz, F., Brunner, S., Vollmer, M., Mentele, E., Muller-Hocker, J., Kitajima, S., Lickert, H., Rupp, R., Franz, W.M., 2008. MesP1 drives vertebrate cardiovascular differentiation through Dkk-1-mediated blockade of Wnt-signalling. Nat. Cell Biol. 10, 338–345. Deng, C.X., Wynshaw-Boris, A., Shen, M.M., Daugherty, C., Ornitz, D. M., Leder, P., 1994. Murine FGFR-1 is required for early postimplantation growth and axial organization. Genes Dev. 8, 3045–3057. Dickmeis, T., Mourrain, P., Saint-Etienne, L., Fischer, N., Aanstad, P., Clark, M., Strahle, U., Rosa, F., 2001. A crucial component of the endoderm formation pathway, CASANOVA, is encoded by a novel sox-related gene. Genes Dev. 15, 1487–1492. Fananapazir, K., Kaufman, M.H., 1988. Observations on the development of the aortico-pulmonary spiral septum in the mouse. J. Anat. 158, 157–172. Feiner, L., Webber, A.L., Brown, C.B., Lu, M.M., Jia, L., Feinstein, P., Mombaerts, P., Epstein, J.A., Raper, J.A., 2001. Targeted disruption of semaphorin 3C leads to persistent truncus arteriosus and aortic arch interruption. Development 128, 3061–3070. Franco, D., Meilhac, S.M., Christoffels, V.M., Kispert, A., Buckingham, M.E., Kelly, R.G., 2006. Left and right ventricular contributions to the formation of the interventricular septum in the mouse heart. Dev. Biol. 294, 366–375. Galli, D., Dominguez, J.N., Zaffran, S., Munk, A., Brown, N.A., Buckingham, M.E., 2008. Atrial myocardium derives from the posterior region of the second heart field, which acquires left–right identity as Pitx2c is expressed. Development 135, 1157–1167. Garcia-Martinez, V., Schoenwolf, G.C., 1993. Primitive-streak origin of the cardiovascular system in avian embryos. Dev. Biol. 159, 706–719. Gardner, R.L., Cockroft, D.L., 1998. Complete dissipation of coherent clonal growth occurs before gastrulation in mouse epiblast. Development 125, 2397–2402. Garriock, R.J., D’Agostino, S.L., Pilcher, K.C., Krieg, P.A., 2005. Wnt11-R, a protein closely related to mammalian Wnt11, is required for heart morphogenesis in Xenopus. Dev. Biol. 279, 179–192. Gassmann, M., Casagranda, F., Orioli, D., Simon, H., Lai, C., Klein, R., Lemke, G., 1995. Aberrant neural and cardiac development in mice lacking the ErbB4 neuregulin receptor. Nature 378, 390–394. George, E.L., Baldwin, H.S., Hynes, R.O., 1997. Fibronectins are essential for heart and blood vessel morphogenesis but are dispensable for initial specification of precursor cells. Blood 90, 3073–3081.

PART  |  3  Patterning of the Early Heart Tube

George, E.L., Georges-Labouesse, E.N., Patel-King, R.S., Rayburn, H., Hynes, R.O., 1993. Defects in mesoderm, neural tube and vascular development in mouse embryos lacking fibronectin. Development 119, 1079–1091. Gerety, S.S., Wang, H.U., Chen, Z.F., Anderson, D.J., 1999. Symmetrical mutant phenotypes of the receptor EphB4 and its specific transmembrane ligand ephrin-B2 in cardiovascular development. Mol. Cell 4, 403–414. Gitler, A.D., Lu, M.M., Jiang, Y.Q., Epstein, J.A., Gruber, P.J., 2003. Molecular markers of cardiac endocardial cushion development. Dev. Dyn. 228, 643–650. Glickman, N.S., Yelon, D., 2002. Cardiac development in zebrafish: coordination of form and function. Semin. Cell Dev. Biol. 13, 507–513. Gong, Y., Mo, C., Fraser, S.E., 2004. Planar cell polarity signalling controls cell division orientation during zebrafish gastrulation. Nature 430, 689–693. Grego-Bessa, J., Luna-Zurita, L., del Monte, G., Bolos, V., Melgar, P., Arandilla, A., Garratt, A.N., Zang, H., Mukouyama, Y.S., Chen, H., Shou, W., Ballestar, E., Esteller, M., Rojas, A., Perez-Pomares, J.M., de la Pompa, J.L., 2007. Notch signaling is essential for ventricular chamber development. Dev. Cell 12, 415–429. Hamblet, N.S., Lijam, N., Ruiz-Lozano, P., Wang, J., Yang, Y., Luo, Z., Mei, L., Chien, K.R., Sussman, D.J., Wynshaw-Boris, A., 2002. Dishevelled 2 is essential for cardiac outflow tract development, somite segmentation and neural tube closure. Development 129, 5827–5838. Han, Y., Dennis, J.E., Cohen-Gould, L., Bader, D.M., Fischman, D.A., 1992. Expression of sarcomeric myosin in the presumptive myocardium of chicken embryos occurs within six hours of myocyte commitment. Dev. Dyn. 193, 257–265. Harh, J.Y., Paul, M.H., 1975. Experimental cardiac morphogenesis. I. Development of the ventricular septum in the chick. J. Embryol. Exp. Morphol. 33, 13–28. Harrelson, Z., Kelly, R.G., Goldin, S.N., Gibson-Brown, J.J., Bollag, R.J., Silver, L.M., Papaioannou, V.E., 2004. Tbx2 is essential for patterning the atrioventricular canal and for morphogenesis of the outflow tract during heart development. Development 131, 5041–5052. Harvey, R.P., 2002. Patterning the vertebrate heart. Nat. Rev. Genet. 3, 544–556. Henderson, D.J., Conway, S.J., Greene, N.D., Gerrelli, D., Murdoch, J.N., Anderson, R.H., Copp, A.J., 2001. Cardiovascular defects associated with abnormalities in midline development in the Loop-tail mouse mutant. Circ. Res. 89, 6–12. Henderson, D.J., Phillips, H.M., Chaudhry, B., 2006. Vang-like 2 and noncanonical Wnt signaling in outflow tract development. Trends Cardiovasc. Med. 16, 38–45. Hove, J.R., Koster, R.W., Forouhar, A.S., Acevedo-Bolton, G., Fraser, S.E., Gharib, M., 2003. Intracardiac fluid forces are an essential epigenetic factor for embryonic cardiogenesis. Nature 421, 172–177. Huisken, J., Swoger, J., Del Bene, F., Wittbrodt, J., Stelzer, E.H., 2004. Optical sectioning deep inside live embryos by selective plane illumination microscopy. Science 305, 1007–1009. Hurle, J.M., Icardo, J.M., Ojeda, J.L., 1980. Compositional and structural heterogenicity of the cardiac jelly of the chick embryo tubular heart: a TEM, SEM and histochemical study. J. Embryol. Exp. Morphol. 56, 211–223. Ilagan, R., Abu-Issa, R., Brown, D., Yang, Y.P., Jiao, K., Schwartz, R.J., Klingensmith, J., Meyers, E.N., 2006. Fgf8 is required for anterior heart field development. Development 133, 2435–2445. Jiang, X., Rowitch, D.H., Soriano, P., McMahon, A.P., Sucov, H.M., 2000. Fate of the mammalian cardiac neural crest. Development 127, 1607–1616.

Chapter  |  3.1  The Behavior of Cells that Form the Myocardial Compartments of the Vertebrate Heart

Johansen, K.A., Iwaki, D.D., Lengyel, J.A., 2003. Localized JAK/STAT signaling is required for oriented cell rearrangement in a tubular epithelium. Development 130, 135–145. Kada, K., Yasui, K., Naruse, K., Kamiya, K., Kodama, I., Toyama, J., 1999. Orientation change of cardiocytes induced by cyclic stretch stimulation: time dependency and involvement of protein kinases. J. Mol. Cell Cardiol. 31, 247–259. Karner, C., Wharton, K.A., Carroll, T.J., 2006a. Apical-basal polarity, Wnt signaling and vertebrate organogenesis. Semin. Cell Dev. Biol. 17, 214–222. Karner Jr, C., Wharton, K.A., Carroll, T.J., 2006b. Planar cell polarity and vertebrate organogenesis. Semin. Cell Dev. Biol. 17, 194–203. Kaufman, M.H., Navaratnam, V., 1981. Early differentiation of the heart in mouse embryos. J. Anat. 133, 235–246. Keegan, B.R., Meyer, D., Yelon, D., 2004. Organization of cardiac chamber progenitors in the zebrafish blastula. Development 131, 3081–3091. Keller, R., 2002. Shaping the vertebrate body plan by polarized embryonic cell movements. Science 298, 1950–1954. Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1, 435–440. Kikuchi, Y., Agathon, A., Alexander, J., Thisse, C., Waldron, S., Yelon, D., Thisse, B., Stainier, D.Y., 2001. Casanova encodes a novel Soxrelated protein necessary and sufficient for early endoderm formation in zebrafish. Genes Dev. 15, 1493–1505. Kikuchi, Y., Trinh, L.A., Reiter, J.F., Alexander, J., Yelon, D., Stainier, D.Y., 2000. The zebrafish bonnie and clyde gene encodes a Mix family homeodomain protein that regulates the generation of endodermal precursors. Genes Dev. 14, 1279–1289. Kinder, S.J., Loebel, D.A., Tam, P.P., 2001. Allocation and early differentiation of cardiovascular progenitors in the mouse embryo. Trends Cardiovasc. Med. 11, 177–184. Kioussi, C., Briata, P., Baek, S.H., Rose, D.W., Hamblet, N.S., Herman, T., Ohgi, K.A., Lin, C., Gleiberman, A., Wang, J., Brault, V., Ruiz-Lozano, P., Nguyen, H.D., Kemler, R., Glass, C.K., WynshawBoris, A., Rosenfeld, M.G., 2002. Identification of a Wnt/Dvl/betaCatenin→Pitx2 pathway mediating cell-type-specific proliferation during development. Cell 111, 673–685. Kirby, M.L., 2007. Cardiogenic fields and heart tube formation. In: Cardiac Development. Oxford University Press, New York, NY, pp. 21–33. Kitajima, S., Takagi, A., Inoue, T., Saga, Y., 2000. MesP1 and MesP2 are essential for the development of cardiac mesoderm. Development 127, 3215–3226. Klaus, A., Saga, Y., Takedo, M.M., Tzahor, Z., Birchmeier, W., 2007. Distinct roles of Wnt/b-catenin and Bmp singaling during early cardiogenesis. Proc. Natl. Acad. Sci. USA 104, 18531–18536. Kokubo, H., Tomita-Miyagawa, S., Hamada, Y., Saga, Y., 2007. Hesr1 and Hesr2 regulate atrioventricular boundary formation in the developing heart through the repression of Tbx2. Development 134, 747–755. Koushik, S.V., Wang, J., Rogers, R., Moskophidis, D., Lambert, N.A., Creazzo, T.L., Conway, S.J., 2001. Targeted inactivation of the sodium–calcium exchanger (Ncx1) results in the lack of a heartbeat and abnormal myofibrillar organization. Faseb. J. 15, 1209–1211. Kuo, C.T., Morrisey, E.E., Anandappa, R., Sigrist, K., Lu, M.M., Parmacek, M.S., Soudais, C., Leiden, J.M., 1997. GATA4 transcription factor is required for ventral morphogenesis and heart tube formation. Genes Dev. 11, 1048–1060. Kupperman, E., An, S., Osborne, N., Waldron, S., Stainier, D.Y., 2000. A sphingosine-1-phosphate receptor regulates cell migration during vertebrate heart development. Nature 406, 192–195.

213

Kwon, C., Arnold, J., Hsiao, E.C., Taketo, M.M., Conklin, B.R., Srivastava, D., 2007. Canonical Wnt signaling is a positive regulator of mammalian cardiac progenitors. Proc. Natl. Acad. Sci. USA 104, 10894–10899. Lee, K.F., Simon, H., Chen, H., Bates, B., Hung, M.C., Hauser, C., 1995. Requirement for neuregulin receptor erbB2 in neural and cardiac development. Nature 378, 394–398. Lee, R.K., Stainier, D.Y., Weinstein, B.M., Fishman, M.C., 1994. Cardiovascular development in the zebrafish. II. Endocardial progenitors are sequestered within the heart field. Development 120, 3361–3366. Li, S., Weidenfeld, J., Morrisey, E.E., 2004. Transcriptional and DNA binding activity of the Foxp1/2/4 family is modulated by heterotypic and homotypic protein interactions. Mol. Cell Biol. 24, 809–822. Liebling, M., Forouhar, A.S., Gharib, M., Fraser, S.E., Dickinson, M.E., 2005. Four-dimensional cardiac imaging in living embryos via postacquisition synchronization of nongated slice sequences. J. Biomed. Opt. 10 (5) 054001. Liebling, M., Forouhar, A.S., Wolleschensky, R., Zimmermann, B., Ankerhold, R., Fraser, S.E., Gharib, M., Dickinson, M.E., 2006. Rapid three-dimensional imaging and analysis of the beating embryonic heart reveals functional changes during development. Dev. Dyn. 235, 2940–2948. Lin, L., Bu, L., Cai, C.L., Zhang, X., Evans, S., 2006. Isl1 is upstream of sonic hedgehog in a pathway required for cardiac morphogenesis. Dev. Biol. 295, 756–763. Lin, L., Cui, L., Zhou, W., Dufort, D., Zhang, X., Cai, C.L., Bu, L., Yang, L., Martin, J., Kemler, R., Rosenfeld, M.G., Chen, J., Evans, S.M., 2007. Beta-catenin directly regulates Islet1 expression in cardiovascular progenitors and is required for multiple aspects of cardiogenesis. Proc. Natl. Acad. Sci. USA 104, 9313–9318. Linask, K.K., 1992. N-cadherin localization in early heart development and polar expression of Na,K()-ATPase, and integrin during pericardial coelom formation and epithelialization of the differentiating myocardium. Dev. Biol. 151, 213–224. Linask, K.K., Lash, J.W., 1988. A role for fibronectin in the migration of avian precardiac cells. I. Dose-dependent effects of fibronectin antibody. Dev. Biol. 129, 315–323. Linask, K.K., Lash, J.W., 1993. Early heart development: dynamics of endocardial cell sorting suggests a common origin with cardiomyocytes. Dev. Dyn. 196, 62–69. Linask, K.K., Knudsen, K.A., Gui, Y.H., 1997. N-cadherin-catenin interaction: necessary component of cardiac cell compartmentalization during early vertebrate heart development. Dev. Biol. 185, 148–164. Liu, C., Liu, W., Lu, M.F., Brown, N.A., Martin, J.F., 2001. Regulation of left–right asymmetry by thresholds of Pitx2c activity. Development 128, 2039–2048. Liu, Z.P., Olson, E.N., 2002. Suppression of proliferation and cardiomyocyte hypertrophy by CHAMP, a cardiac-specific RNA helicase. Proc. Natl. Acad. Sci. USA 99, 2043–2048. Lomonico, M.P., Moore, G.W., Hutchins, G.M., 1986. Rotation of the junction of the outflow tract and great arteries in the embryonic human heart. Anat. Rec. 216, 544–549. Lomonico, M.P., Bostrom, M.P., Moore, G.W., Hutchins, G.M., 1988. Arrested rotation of the outflow tract may explain tetralogy of Fallot and transposition of the great arteries. Pediatr. Pathol. 8, 267–281. Lopez-Sanchez, C., Garcia-Martinez, V., Schoenwolf, G.C., 2001. Localization of cells of the prospective neural plate, heart and somites within the primitive streak and epiblast of avian embryos at inter­ mediate primitive-streak stages. Cells Tissues Organs. 169, 334–346. Lough, J., Sugi, Y., 2000. Endoderm and heart development. Dev. Dyn. 217, 327–342.

214

Mably, J.D., Mohideen, M.A., Burns, C.G., Chen, J.N., Fishman, M.C., 2003. Heart of glass regulates the concentric growth of the heart in zebrafish. Curr. Biol. 13, 2138–2147. Manasek, F.J., Burnside, M.B., Waterman, R.E., 1972. Myocardial cell shape change as a mechanism of embryonic heart looping. Dev. Biol. 29, 349–371. Manner, J., 2006. Extracardiac tissues and the epigenetic control of myocardial development in vertebrate embryos. Ann. Anat. 188, 199–212. Markwald, R.R., Fitzharris, T.P., Manasek, F.J., 1977. Structural development of endocardial cushions. Am. J. Anat. 148, 85–119. Marvin, M.J., Di Rocco, G., Gardiner, A., Bush, S.M., Lassar, A.B., 2001. Inhibition of Wnt activity induces heart formation from poster­ ior mesoderm. Genes Dev. 15, 316–327. Matsui, T., Raya, A., Kawakami, Y., Callol-Massot, C., Capdevila, J., Rodriguez-Esteban, C., Izpisua Belmonte, J.C., 2005. Noncanonical Wnt signaling regulates midline convergence of organ primordia during zebrafish development. Genes Dev. 19, 164–175. McGrath, K.E., Koniski, A.D., Malik, J., Palis, J., 2003. Circulation is established in a stepwise pattern in the mammalian embryo. Blood 101, 1669–1676. McKinsey, T.A., Zhang, C.L., Olson, E.N., 2002. MEF2: a calciumdependent regulator of cell division, differentiation and death. Trends Biochem. Sci. 27, 40–47. Meilhac, S.M., Kelly, R.G., Rocancourt, D., Eloy-Trinquet, S., Nicolas, J.F., Buckingham, M.E., 2003. A retrospective clonal analysis of the myocardium reveals two phases of clonal growth in the developing mouse heart. Development 130, 3877–3889. Meilhac, S.M., Esner, M., Kelly, R.G., Nicolas, J.F., Buckingham, M.E., 2004a. The clonal origin of myocardial cells in different regions of the embryonic mouse heart. Dev. Cell 6, 685–698. Meilhac, S.M., Esner, M., Kerszberg, M., Moss, J.E., Buckingham, M.E., 2004b. Oriented clonal cell growth in the developing mouse myocardium underlies cardiac morphogenesis. J. Cell Biol. 164, 97–109. Meyer, D., Birchmeier, C., 1995. Multiple essential functions of neuregulin in development. Nature 378, 386–390. Mikawa, T., Borisov, A., Brown, A.M., Fischman, D.A., 1992. Clonal analysis of cardiac morphogenesis in the chicken embryo using a replication-defective retrovirus: I. Formation of the ventricular myocardium. Dev. Dyn. 193, 11–23. Miyasaka, K.Y., Kida, Y.S., Sato, T., Minami, M., Ogura, T., 2007. Csrp1 regulates dynamic cell movements of the mesendoderm and cardiac mesoderm through interactions with Dishevelled and Diversin. Proc. Natl. Acad. Sci. USA 104, 11274–11279. Mjaatvedt, C.H., Yamamura, H., Capehart, A.A., Turner, D., Markwald, R.R., 1998. The Cspg2 gene, disrupted in the hdf mutant, is required for right cardiac chamber and endocardial cushion formation. Dev. Biol. 202, 56–66. Mjaatvedt, C.H., Nakaoka, T., Moreno-Rodriguez, R., Norris, R.A., Kern, M.J., Eisenberg, C.A., Turner, D., Markwald, R.R., 2001. The outflow tract of the heart is recruited from a novel heart-forming field. Dev. Biol. 238, 97–109. Moeller, H., Jenny, A., Schaeffer, H.J., Schwarz-Romond, T., Mlodzik, M., Hammerschmidt, M., Birchmeier, W., 2006. Diversin regulates heart formation and gastrulation movements in development. Proc. Natl. Acad. Sci. USA 103, 15900–15905. Molkentin, J.D., Lin, Q., Duncan, S.A., Olson, E.N., 1997. Requirement for the transcription factor GATA4 for heart tube formation and ventral morphogenesis. Genes Dev. 11, 1061–1072.

PART  |  3  Patterning of the Early Heart Tube

Moreno-Rodriguez, R.A., Krug, E.L., Reyes, L., Villavicencio, L., Mjaatvedt, C.H., Markwald, R.R., 2006. Bidirectional fusion of the heart-forming fields in the developing chick embryo. Dev. Dyn. 235, 191–202. Nagy, II., Railo, A., Rapila, R., Hast, T., Sormunen, R., Tav, i P., Räsänen, J., Vainio, S.J., 2010. Wnt-11 signalling controls ventricular myocardium development by patterning N-cadherin and beta-catenin expression. Cardiovasc. Res. 85, 100–109. Nakagawa, S., Takeichi, M., 1997. N-cadherin is crucial for heart formation in the chick embryo. Dev. Growth Differ. 39, 451–455. Nakamura, A., Manasek, F.J., 1981. An experimental study of the relation of cardiac jelly to the shape of the early chick embryonic heart. J. Embryol. Exp. Morphol. 65, 235–256. Narita, N., Bielinska, M., Wilson, D.B., 1997. Wild-type endoderm abrogates the ventral developmental defects associated with GATA-4 deficiency in the mouse. Dev. Biol. 189, 270–274. Navaratnam, V., Kaufman, M.H., Skepper, J.N., Barton, S., Guttridge, K.M., 1986. Differentiation of the myocardial rudiment of mouse embryos: an ultrastructural study including freeze-fracture replication. J. Anat. 146, 65–85. Nishii, K., Shibata, Y., 2006. Mode and determination of the initial contraction stage in the mouse embryo heart. Anat. Embryol. 211, 95–100. Oishi, I., Suzuki, H., Onishi, N., Takada, R., Kani, S., Ohkawara, B., Koshida, I., Suzuki, K., Yamada, G., Schwabe, G.C., Mundlos, S., Shibuya, H., Takada, S., Minami, Y., 2003. The receptor tyrosine kinase Ror2 is involved in non-canonical Wnt5a/JNK signalling pathway. Genes Cells 8, 645–654. Pandur, P., Lasche, M., Eisenberg, L.M., Kuhl, M., 2002. Wnt-11 ­activation of a non-canonical Wnt signalling pathway is required for cardiogenesis. Nature 418, 636–641. Park, E.J., Ogden, L.A., Talbot, A., Evans, S., Cai, C.L., Black, B.L., Frank, D.U., Moon, A.M., 2006. Required, tissue-specific roles for Fgf8 in outflow tract formation and remodeling. Development 133, 2419–2433. Peterson, R.T., Mably, J.D., Chen, J.N., Fishman, M.C., 2001. Convergence of distinct pathways to heart patterning revealed by the small molecule concentramide and the mutation heart-and-soul. Curr. Biol. 11, 1481–1491. Peyrieras, N., Strahle, U., Rosa, F., 1998. Conversion of zebrafish blastomeres to an endodermal fate by TGF-beta-related signaling. Curr. Biol. 8, 783–786. Phillips, H.M., Murdoch, J.N., Chaudhry, B., Copp, A.J., Henderson, D.J., 2005. Vangl2 acts via RhoA signaling to regulate polarized cell movements during development of the proximal outflow tract. Circ. Res. 96, 292–299. Prall, O.W., Menon, M.K., Solloway, M.J., Watanabe, Y., Zaffran, S., Bajolle, F., Biben, C., McBride, J.J., Robertson, B.R., Chaulet, H., Stennard, F.A., Wise, N., Schaft, D., Wolstein, O., Furtado, M.B., Shiratori, H., Chien, K.R., Hamada, H., Black, B.L., Saga, Y., Robertson, E.J., Buckingham, M.E., Harvey, R.P., 2007. An Nkx2-5/ Bmp2/Smad1 negative feedback loop controls heart progenitor specification and proliferation. Cell 128, 947–959. Prusinkiewicz, P., Rolland-Lagan, A.G., 2006. Modeling plant morphogenesis. Curr. Opin. Plant. Biol. 9, 83–88. Qyang, Y., Martin-Puig, S., Chiravuri, M., Chen, S., Xu, H., Bu, L., Jiang, X., Lin, L., Granger, A., Moretti, A., Caron, L., Wu, X., Clarke, J., Taketo, M.M., Laugwitz, K.-L., Moon, R.T., Gruber, T., Evans, S.M., Ding, S., Chien, K.R., 2007. The renewal and ­ differentiation of Isl1 cardiovascular progenitors are controlled by a Wnt/beta­catenin pathway. Cell Stem. Cell 1, 1–15.

Chapter  |  3.1  The Behavior of Cells that Form the Myocardial Compartments of the Vertebrate Heart

Radice, G.L., Rayburn, H., Matsunami, H., Knudsen, K.A., Takeichi, M., Hynes, R.O., 1997. Developmental defects in mouse embryos lacking N-cadherin. Dev. Biol. 181, 64–78. Redkar, A., Montgomery, M., Litvin, J., 2001. Fate map of early avian cardiac progenitor cells. Development 128, 2269–2279. Reifers, F., Walsh, E.C., Leger, S., Stainier, D.Y., Brand, M., 2000. Induction and differentiation of the zebrafish heart requires fibroblast growth factor 8 (fgf8/acerebellar). Development 127, 225–235. Reiter, J.F., Alexander, J., Rodaway, A., Yelon, D., Patient, R., Holder, N., Stainier, D.Y., 1999. Gata5 is required for the development of the heart and endoderm in zebrafish. Genes Dev. 13, 2983–2995. Resino, J., Salama-Cohen, P., Garcia-Bellido, A., 2002. Determining the role of patterned cell proliferation in the shape and size of the Drosophila wing. Proc. Natl. Acad. Sci. USA 99, 7502–7507. Rock, R., Schrauth, S., Gessler, M., 2005. Expression of mouse dchs1, fjx1, and fat-j suggests conservation of the planar cell polarity pathway identified in Drosophila. Dev. Dyn. 234, 747–755. Roebroek, A.J., Umans, L., Pauli, I.G., Robertson, E.J., van Leuven, F., Van de Ven, W.J., Constam, D.B., 1998. Failure of ventral closure and axial rotation in embryos lacking the proprotein convertase Furin. Development 125, 4863–4876. Rohr, S., Bit-Avragim, N., Abdelilah-Seyfried, S., 2006. Heart and soul/ PRKCi and nagie oko/Mpp5 regulate myocardial coherence and remodeling during cardiac morphogenesis. Development 133, 107–115. Rohr, S., Otten, C., Abdelilah-Seyfried, S., 2008. Asymmetric involution of the myocardial field drives heart tube formation in zebrafish. Circ. Res. 102, e12–e19. Rolland-Lagan, A.G., Bangham, J.A., Coen, E., 2003. Growth dynamics underlying petal shape and asymmetry. Nature 422, 161–163. Rones, M.S., McLaughlin, K.A., Raffin, M., Mercola, M., 2000. Serrate and Notch specify cell fates in the heart field by suppressing cardiomyogenesis. Development 127, 3865–3876. Rosenquist, G.C., 1970. Location and movements of cardiogenic cells in the chick embryo: the heart-forming portion of the primitive streak. Dev. Biol. 22, 461–475. Rosenquist, G.C., De Haan, R.L., 1966. Migration of precardiac cells in the chick embryo : a radioautographic study. Contrib. Embryol. 263, 111–121. Rottbauer, W., Saurin, A.J., Lickert, H., Shen, X., Burns, C.G., Wo, Z.G., Kemler, R., Kingston, R., Wu, C., Fishman, M., 2002. Reptin and pontin antagonistically regulate heart growth in zebrafish embryos. Cell 111, 661–672. Rutenberg, J.B., Fischer, A., Jia, H., Gessler, M., Zhong, T.P., Mercola, M., 2006. Developmental patterning of the cardiac atrioventricular canal by Notch and Hairy-related transcription factors. Development 133, 4381–4390. Rychter, Z., Rychterova, V., Lemez, L., 1979. Formation of the heart loop and proliferation structure of its wall as a base for ventricular septation. Herz 4, 86–90. Saga, Y., Miyagawa-Tomita, S., Takagi, A., Kitajima, S., Miyazaki, J., Inoue, T., 1999. MesP1 is expressed in the heart precursor cells and required for the formation of a single heart tube. Development 126, 3437–3447. Sakaguchi, T., Kikuchi, Y., Kuroiwa, A., Takeda, H., Stainier, D.Y., 2006. The yolk syncytial layer regulates myocardial migration by influencing extracellular matrix assembly in zebrafish. Development 133, 4063–4072. Satin, J., Fujii, S., De Haan, R.L., 1988. Development of cardiac beat rate in early chick embryos is regulated by regional cues. Dev. Biol. 129, 103–113.

215

Scherz, P.J., Huisken, J., Sahai-Hernandez, P., Stainier, D.Y., 2008. Highspeed imaging of developing heart valves reveals interplay of morphogenesis and function. Development 135, 1179–1187. Schier, A.F., Neuhauss, S.C., Helde, K.A., Talbot, W.S., Driever, W., 1997. The one-eyed pinhead gene functions in mesoderm and endoderm formation in zebrafish and interacts with no tail. Development 124, 327–342. Schneider, V.A., Mercola, M., 2001. Wnt antagonism initiates cardiogenesis in Xenopus laevis. Genes Dev. 15, 304–315. Scott, I.C., Masri, B., D’Amico, L.A., Jin, S.W., Jungblut, B., Wehman, A.M., Baier, H., Audigier, Y., Stainier, D.Y., 2007. The g-protein-coupled receptor agtrl1b regulates early development of myocardial progen­ itors. Dev. Cell 12, 403–413. Sedmera, D., Reckova, M., DeAlmeida, A., Coppen, S.R., Kubalak, S.W., Gourdie, R.G., Thompson, R.P., 2003. Spatiotemporal pattern of commitment to slowed proliferation in the embryonic mouse heart indicates progressive differentiation of the cardiac conduction system. Anat. Rec. A. Discov. Mol. Cell Evol. Biol. 274, 773–777. Sharpe, J., Ahlgren, U., Perry, P., Hill, B., Ross, A., Hecksher-Sorensen, J., Baldock, R., Davidson, D., 2002. Optical projection tomography as a tool for 3D microscopy and gene expression studies. Science 296, 541–545. Singh, M.K., Christoffels, V.M., Dias, J.M., Trowe, M.O., Petry, M., Schuster-Gossler, K., Burger, A., Ericson, J., Kispert, A., 2005. Tbx20 is essential for cardiac chamber differentiation and repression of Tbx2. Development 132, 2697–2707. Sissman, N.J., 1966. Cell multiplication rates during development of the primitive cardiac tube in the chick embryo. Nature 210, 504–507. Smith, K.A., Chocron, S., von der Hardt, S., de Pater, E., Soufan, A., Bussmann, J., Schulte-Merker, S., Hammerschmidt, M., Bakkers, J., 2008. Rotation and asymmetric development of the zebrafish heart requires directed migration of cardiac progenitor cells. Dev. Cell 14, 287–297. Solnica-Krezel, L., 2005. Conserved patterns of cell movements during vertebrate gastrulation. Curr. Biol. 15, R213–R228. Soufan, A.T., van den Berg, G., Ruijter, J.M., de Boer, P.A., van den Hoff, M.J., Moorman, A.F., 2006. Regionalized sequence of myocardial cell growth and proliferation characterizes early chamber formation. Circ. Res. 99, 545–552. Srivastava, D., 2006. Making or breaking the heart: from lineage determination to morphogenesis. Cell 126, 1037–1048. Stadtfeld, M., Ye, M., Graf, T., 2007. Identification of interventricular septum precursor cells in the mouse embryo. Dev. Biol. 302, 195–207. Stainier, D.Y., Lee, R.K., Fishman, M.C., 1993. Cardiovascular development in the zebrafish. I. Myocardial fate map and heart tube formation. Development 119, 31–40. Stainier, D.Y., Weinstein, B.M., Detrich, H.W., Zon, L.I., Fishman, M. C., 1995. Cloche, an early acting zebrafish gene, is required by both the endothelial and hematopoietic lineages. Development 121, 3141–3150. Stalsberg, H., 1969a. Regional mitotic activity in the precardiac mesoderm and differentiating heart tube in the chick embryo. Dev. Biol. 20, 18–45. Stalsberg, H., 1969b. The origin of heart asymmetry: right and left contributions to the early chick embryo heart. Dev. Biol. 19, 109–127. Stalsberg, H., De Haan, R.L., 1969. The precardiac areas and formation of the tubular heart in the chick embryo. Dev. Biol. 19, 128–159. Stankunas, K., Hang, C.T., Tsun, Z.Y., Chen, H., Lee, N.V., Wu, J.I., Shang, C., Bayle, J.H., Shou, W., Iruela-Arispe, M.L., Chang, C.P.,

216

2008. Endocardial Brg1 represses ADAMTS1 to maintain the microenvironment for myocardial morphogenesis. Dev. Cell 14, 298–311. Strauss, B., Adams, R.J., Papalopulu, N., 2006. A default mechanism of spindle orientation based on cell shape is sufficient to generate cell fate diversity in polarised Xenopus blastomeres. Development 133, 3883–3893. Takeuchi, J.K., Ohgi, M., Koshiba-Takeuchi, K., Shiratori, H., Sakaki, I., Ogura, K., Saijoh, Y., Ogura, T., 2003. Tbx5 specifies the left/right ventricles and ventricular septum position during cardiogenesis. Development 130, 5953–5964. Thery, M., Racine, V., Pepin, A., Piel, M., Chen, Y., Sibarita, J.B., Bornens, M., 2005. The extracellular matrix guides the orientation of the cell division axis. Nat. Cell Biol. 7, 947–953. Thompson, R.P., Abercrombie, V., Wong, M., 1987. Morphogenesis of the truncus arteriosus of the chick embryo heart: movements of autoradio­ graphic tattoos during septation. Anat. Rec. 218, 434–440, 394–435. Thompson, R.P., Lindroth, J.R., Wong, Y.M.M., 1990. Regional differences in DNA-synthetic activity in the preseptation myocardium of the chick. In: Clark, E.B., Takao, A. (Eds.), Developmental Cardiology: Morphogenesis and Function. Futura Publishing Co, Mount Kisco, NY, pp. 219–234. Timmerman, L.A., Grego-Bessa, J., Raya, A., Bertran, E., PerezPomares, J.M., Diez, J., Aranda, S., Palomo, S., McCormick, F., Izpisua-Belmonte, J.C., de la Pompa, J.L., 2004. Notch promotes epithelial-mesenchymal transition during cardiac development and oncogenic transformation. Genes Dev. 18, 99–115. Togi, K., Kawamoto, T., Yamauchi, R., Yoshida, Y., Kita, T., Tanaka, M., 2004. Role of Hand1/eHAND in the dorso–ventral patterning and interventricular septum formation in the embryonic heart. Mol. Cell Biol. 24, 4627–4635. Trinh, L.A., Stainier, D.Y., 2004. Fibronectin regulates epithelial organization during myocardial migration in zebrafish. Dev. Cell 6, 371–382. Trinh, L.A., Yelon, D., Stainier, D.Y., 2005. Hand2 regulates epithelial formation during myocardial diferentiation. Curr. Biol. 15, 441–446. Tullio, A.N., Accili, D., Ferrans, V.J., Yu, Z.X., Takeda, K., Grinberg, A., Westphal, H., Preston, Y.A., Adelstein, R.S., 1997. Nonmuscle myosin II-B is required for normal development of the mouse heart. Proc. Natl. Acad. Sci. USA 94, 12407–12412. Tzahor, E., Lassar, A.B., 2001. Wnt signals from the neural tube block ectopic cardiogenesis. Genes Dev. 15, 255–260. Ueno, S., Weidinger, G., Osugi, T., Kohn, A.D., Golob, J.L., Pabon, L., Reinecke, H., Moon, R.T., Murry, C.E., 2007. Biphasic role for Wnt/beta-catenin signaling in cardiac specification in zebrafish and embryonic stem cells. Proc. Natl. Acad. Sci. USA 104, 9685–9690. van den Hoff, M.J., Kruithof, B.P., Moorman, A.F., 2004. Making more heart muscle. Bioessays 26, 248–261. Wakimoto, K., Kobayashi, K., Kuro, O.M., Yao, A., Iwamoto, T., Yanaka, N., Kita, S., Nishida, A., Azuma, S., Toyoda, Y., Omori, K., Imahie, H., Oka, T., Kudoh, S., Kohmoto, O., Yazaki, Y., Shigekawa, M., Imai, Y., Nabeshima, Y., Komuro, I., 2000. Targeted disruption of Na/Ca2 exchanger gene leads to cardiomyocyte apoptosis and defects in heartbeat. J. Biol. Chem. 275, 36991–36998. Waldo, K., Zdanowicz, M., Burch, J., Kumiski, D.H., Stadt, H.A., Godt, R.E., Creazzo, T.L., Kirby, M.L., 1999. A novel role for cardiac neural crest in heart development. J. Clin. Invest. 103, 1499–1507. Waldo, K.L., Kumiski, D.H., Wallis, K.T., Stadt, H.A., Hutson, M.R., Platt, D.H., Kirby, M.L., 2001. Conotruncal myocardium arises from a secondary heart field. Development 128, 3179–3188.

PART  |  3  Patterning of the Early Heart Tube

Waldo, K.L., Hutson, M.R., Stadt, H.A., Zdanowicz, M., Zdanowicz, J., Kirby, M.L., 2005a. Cardiac neural crest is necessary for normal addition of the myocardium to the arterial pole from the secondary heart field. Dev. Biol. 281, 66–77. Waldo, K.L., Hutson, M.R., Ward, C.C., Zdanowicz, M., Stadt, H.A., Kumiski, D., Abu-Issa, R., Kirby, M.L., 2005b. Secondary heart field contributes myocardium and smooth muscle to the arterial pole of the developing heart. Dev. Biol. 281, 78–90. Wallace, K.N., Pack, M., 2003. Unique and conserved aspects of gut development in zebrafish. Dev. Biol. 255, 12–29. Wang, H.U., Chen, Z.F., Anderson, D.J., 1998. Molecular distinction and angiogenic interaction between embryonic arteries and veins revealed by ephrin-B2 and its receptor Eph-B4. Cell 93, 741–753. Ward, C., Stadt, H., Hutson, M., Kirby, M.L., 2005. Ablation of the secondary heart field leads to tetralogy of Fallot and pulmonary atresia. Dev. Biol. 284, 72–83. Washington Smoak, I., Byrd, N.A., Abu-Issa, R., Goddeeris, M.M., Anderson, R., Morris, J., Yamamura, K., Klingensmith, J., Meyers, E.N., 2005. Sonic hedgehog is required for cardiac outflow tract and neural crest cell development. Dev. Biol. 283, 357–372. Watanabe, Y., Kokubo, H., Miyagawa-Tomita, S., Endo, M., Igarashi, K., Aisaki, K., Kanno, J., Saga, Y., 2006. Activation of Notch1 signaling in cardiogenic mesoderm induces abnormal heart morphogenesis in mouse. Development 133, 1625–1634. Wei, L., Roberts, W., Wang, L., Yamada, M., Zhang, S., Zhao, Z., Rivkees, S.A., Schwartz, R.J., Imanaka-Yoshida, K., 2001. Rho kinases play an obligatory role in vertebrate embryonic organogenesis. Development 128, 2953–2962. Wei, L., Imanaka-Yoshida, K., Wang, L., Zhan, S., Schneider, M.D., DeMayo, F.J., Schwartz, R.J., 2002. Inhibition of Rho family GTPases by Rho GDP dissociation inhibitor disrupts cardiac morphogenesis and inhibits cardiomyocyte proliferation. Development 129, 1705–1714. Wei, Y., Mikawa, T., 2000. Fate diversity of primitive streak cells during heart field formation in ovo. Dev. Dyn. 219, 505–513. Wendler, C.C., Rivkees, S.A., 2006. Sphingosine-1-phosphate inhibits cell migration and endothelial-to-mesenchymal cell transformation during cardiac development. Dev. Biol. 291, 264–277. Weninger, W.J., Geyer, S.H., Mohun, T.J., Rasskin-Gutman, D., Matsui, T., Ribeiro, I., Costa Lda, F., Izpisua-Belmonte, J.C., Muller, G.B., 2006. High-resolution episcopic microscopy: a rapid technique for high detailed 3D analysis of gene activity in the context of tissue architecture and morphology. Anat. Embryol. 211, 213–221. Yamaguchi, T.P., Harpal, K., Henkemeyer, M., Rossant, J., 1994. Fgfr-1 is required for embryonic growth and mesodermal patterning during mouse gastrulation. Genes Dev. 8, 3032–3044. Yamamura, H., Zhang, M., Markwald, R.R., Mjaatvedt, C.H., 1997. A heart segmental defect in the anterior–posterior axis of a transgenic mutant mouse. Dev. Biol. 186, 58–72. Yamazaki, T., Komuro, I., Nagai, R., Yazaki, Y., 1996. Stretching the evidence in the case of cardiac growth. Cardiovasc. Res. 31, 493–498. Yelon, D., Ticho, B., Halpern, M.E., Ruvinsky, I., Ho, R.K., Silver, L.M., Stainier, D.Y., 2000. The bHLH transcription factor hand2 plays parallel roles in zebrafish heart and pectoral fin development. Development 127, 2573–2582. Yue, Q., Wagstaff, L., Yang, X., Weijer, C., Munsterberg, A., 2008. Wnt3amediated chemorepulsion controls movement patterns of cardiac progenitors and requires RhoA function. Development 135, 1029–1037. Zaffran, S., Kelly, R.G., Meilhac, S.M., Buckingham, M.E., Brown, N.A., 2004. Right ventricular myocardium derives from the anterior heart field. Circ. Res. 95, 261–268.

Chapter  |  3.1  The Behavior of Cells that Form the Myocardial Compartments of the Vertebrate Heart

Zak, R., Kizu, A., Bugaisky, L., 1979. Cardiac hypertrophy: its characteristics as a growth process. Am. J. Cardiol. 44, 941–946. Zeng, X.X., Wilm, T.P., Sepich, D.S., Solnica-Krezel, L., 2007. Apelin and its receptor control heart field formation during zebrafish gastrulation. Dev. Cell 12, 391–402. Zhang, H., Toyofuku, T., Kamei, J., Hori, M., 2003. GATA-4 regulates cardiac morphogenesis through transactivation of the N-cadherin gene. Biochem. Biophys. Res. Commun. 312, 1033–1038.

217

Zhou, W., Lin, L., Majumdar, A., Li, X., Zhang, X., Liu, W., Etheridge, L., Shi, Y., Martin, J., Van de Ven, W., Kaartinen, V., WynshawBoris, A., McMahon, A.P., Rosenfeld, M.G., Evans, S.M., 2007. Modulation of morphogenesis by noncanonical Wnt signaling requires ATF/CREB family-mediated transcriptional activation of TGFbeta2. Nat. Genet. 39, 1225–1234.

Chapter 3.2

Early Cardiac Growth and the Ballooning Model of Cardiac Chamber Formation Antoon F.M. Moorman1, Gert van den Berg1, Robert H. Anderson2 and Vincent M. Christoffels1 1

Heart Failure Research Center, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands Cardiac Unit, Institute of Child Health, University College, London, United Kingdom

2

I.  Introduction The development of the atrial and ventricular chambers is a key event in the formation of the building plan of the vertebrate heart. Insight into the process of chamber formation, and the underlying mechanisms, is of critical importance to understand the evolution of the vertebrate heart (Randall and Davie, 1980; Moorman and Christoffels, 2003a). In our vertebrate ancestors the evolution of advanced systems for filtration, such as the liver and the kidneys, has necessitated the concomitant transformation of a circulation functioning in conditions of low pressure but high volume into one characterized by higher pressures but lower volumes (Randall and Davie, 1980). These changes, in turn, required the development of ventricles with powerful pumping capacities. The vertebrate heart with its multiple chambers is presumed to have evolved from a simple form of the peristaltic contracting vessel positioned ventral to the intestinal or pharyngeal region, as seen in extant lower vertebrates (Randall and Davie, 1980). The early presence in embryos from higher vertebrates of a peristaltic contracting heart underscores this notion. In essence, the primitive heart as seen in ancestral chordates was no more than a myocardial mantle surrounding a ventral aorta. If the circulation is interrupted by visceral organs, such as the liver, it is said that multiple hearts or contractile vessels can be present. Such hearts are comparable with the dorsal vessel seen in Drosophila. The heart of Drosophila, therefore, has become a valuable model for the formation of cardiac myocytes, as reflected in the evolutionary conservation of many basic molecular regulatory mechanisms (Zaffran and Frasch, 2002; Olson, 2006). The heart of Drosophila, nonetheless, cannot possibly instruct us about the intricate morphogenesis of the hearts of higher vertebrates, since Heart Development and Regeneration 2010 Copyright © 2010,Elsevier Inc. All rights of reproduction in any form reserved.

these latter pumps possess multiple chambers along with a specialized myocardial system for conduction, features that are not present in the heart of the fruit fly. The development of the highly-specialized forceproducing working myocardium that forms the cardiac chambers required important adaptations. These adaptations involved the development of one-way valves and of a myocardial system of electrical wiring, thus permitting a coordinated pattern of contraction. The system of wiring is known as the cardiac conduction system (see Chapter 2.3, Vol. I), albeit the myocardium of the chambers themselves is an integral part of the overall pattern of conduction (Moorman and Christoffels, 2003b). Remarkably similar electrocardiograms can be recorded from animals as diverse as fish and man (Moorman and Christoffels, 2003a), indicating that both the essence of the cardiac building plan and the electrical connections between the cardiac chambers have been highly conserved during vertebrate evolution. This is notwithstanding the fact that fish have a heart supporting a solitary circulatory pattern, whereas the hearts of mammals support both systemic and pulmonary circulations. The development of the mammalian four-chambered heart also required an expansion of the complexity of the ancestral network of cardiac transcriptional and signaling pathways. Recent studies, for example, have revealed a major role for a network of multiple T-box transcription factors in the design of the modern four-chambered heart, including the chambers, septa and valves, and the conduction system (Moorman et al., 2004; Plageman and Yutzey, 2005; Hoogaars et al., 2007). In this chapter we outline the mechanisms of formation of the building plan of the vertebrate heart, placing emphasis on cardiac growth and chamber development. We attempt to set the scene for understanding the relationships

219

220

PART  |  3  Patterning of the Early Heart Tube

between the building plan and mechanisms of cardiac patterning in the various fields of cardiac precursor cells. The pulmonary circulation is the essential difference between the mammalian heart and that of our vertebrate ancestors, with their purely systemic circulatory function. Hence, we also discuss the significance of the development of the cardiac components required for the pulmonary circulation.

II.  Diversity of cardiac myocytes: the primary versus the chamber myocardium Cardiac myocytes share a number of characteristic features that distinguish them from other cells. All cardiac myocytes have sarcomeres and a sarcoplasmic reticulum. They share the capacity of producing an intrinsic cycle of electrical activity resulting in contraction. This phenomenon is called automaticity, or pacemaker activity. Varying degrees of differentiation can be seen among embryonic populations of cardiac myocytes, which can be categorized as primary, chamber, nodal and conducting myocardium (see Table 1). An absolute requirement for effective automaticity, or pacemaker activity, is poor electrical coupling of the cardiac cells which also implies slow conduction between these cells, and absence of inward rectifier currents (Joyner and van Capelle, 1986; Miake et al., 2002). The myocardium of the embryonic heart tube is called primary myocardium, characterized by automaticity and slow conduction. The slow conduction of the depolarizing impulse over the heart tube causes a peristaltic wave of contraction, which pushes the blood towards the arterial pole. Because unambiguous morphological markers are often lacking in the early embryonic heart to distinguish the various cell types, automaticity and the speed of conduction have become important functional parameters with which to describe the development of the different parts of the heart. The unidirectional propulsion of the blood shows

that not all cardiomyocytes of the primary heart tube are identical, and that there is polarity of pacemaking activity over the heart tube. Regulatory and myogenic proteins are also present in cranio–caudal gradients over the heart tube (Moorman and Christoffels, 2003a), and retinoic acid plays an important role in the establishment of these gradients (Osmond et al., 1991; Yutzey et al., 1994; XavierNeto et al., 1999). While the primary heart tube is still being formed by addition of cardiac myocytes at both ends of the tube, ventricular chambers expand at the ventral side of the tube, and atrial chambers at the dorsal side by a local increase in growth and proliferation of the myocytes (Soufan et al., 2006). Automaticity is being suppressed in these cardiac cells by increasing levels of high conductance gap junctions (in mammals Connexin 40 and 43), inward rectifier channel activity and fast sodium currents. Hence, the cells of the atrial and ventricular chamber myocardium display virtually no automaticity and are well-coupled. They have well-developed sarcomeres and sarcoplasmic reticular structures. The increased intercellular conduction allows rapid conduction of the depolarizing impulse and the development of synchronous contraction of the chamber myocardium. Later in development there is formation of the rapidly-conducting ventricular conduction system, the cells of which display an even higher conduction velocity (see Chapter 2.3, Vol. I). The categorization of the cardiac myocytes as described above does not imply that all cells belonging to one group are identical, but rather that they share certain functional features that distinguish them from the myocytes of other groups.

III.  The cardiac chambers III.A.  Evolutionary Conservation Birds and mammals, along with some reptiles such as the crocodiles and alligators, have four-chambered hearts, with

Table 1  Basic Phenotypes of Cardiac Muscle Cells Type of Myocardium Feature

primary

nodal

ventricular conduction

chamber

Automaticity

high

high

high

low

Conduction velocity

low

low

high

high

Contractility

low

low

low

high

SR activity

low

low

low

high

Proliferation

low

low

low

high

All cardiomyocytes share a number of key features mentioned in the first column. For the sake of simplicity they are ranked only as being high or low. The morphological data are largely based on Canale et al. (1986), while Kleber et al. (2001) is used as the basis for the electrophysiological data and Sissman (1966); Thompson et al. (1990); Soufan et al. (2006) for the proliferation data. The proliferation data refer to forming myocardium. See Moorman and Christoffels (2003a); Christoffels et al. (2004a) for reviews.

Chapter  |  3.2  Early Cardiac Growth and the Ballooning Model of Cardiac Chamber Formation

the so-called right- and left-sided chambers functioning in parallel circuits. In these animals, including man, the right atrium is exclusively connected to the right ventricle, and the left atrium exclusively to the left ventricle. The right half of the heart drives blood from the body through the lungs, while the left half drives the oxygenated blood from the lungs through the rest of the body. The crucial additional feature in those reptiles with four-chambered hearts is that a right ventricular aorta also contributes to the systemic circulation, with a communication between the left ventricular and right ventricular aortas just distal from the valves. Animals lacking a pulmonary circulation, such as fish, have two-chambered hearts, with solitary atrial and ventricular chambers. One could argue that fish possess fourchambered hearts, counting the “sinus venosus” and “conus arteriosus” in addition to the atrium and ventricle (SimoesCosta et al., 2005). This may well be justified if the emphasis is placed on the cavities (the sinus venosus and conus arteriosus of fish are real cavities and may be developmental compartments). However, we choose to restrict the term “chambers” to the atrial and ventricular compartment(s), because the cavity-based nomenclature ignores the fundamental differences between the myocardium of the atrial and ventricular chambers compared with that of the systemic venous sinus (sinus venosus), the atrioventricular canal and the outflow tract (conus). For example, the walls of the atrial and ventricular chambers are composed of rapidlyconducting working myocardium, whereas the walls of the flanking compartments are not (Moorman and Christoffels, 2003a). This situation is well seen in the heart of the shark, in which the walls of the conus arteriosus conduct at a speed comparable to those of the outflow tract of the developing chick heart (Satchell and Jones, 1967; Burggren, 1988; de Jong et al., 1992). It has also been demonstrated that the myocardium of the developing outflow tract displays long periods of contraction, owing to its slow speed of conduction, permitting the myocardium to support the action of the arterial valves, whereas the force-producing working myocardium of the ventricle displays synchronous contractions as required for the build-up of the ventricular pressures. The conus of teleost fish fulfill the same function as the developing outflow tract (Icardo, 2006). In this chapter, therefore, we will distinguish the myocardium of the developing chambers from that of the flanking compartments, and not define the latter as discrete chambers.

III.B.  Chamber Development is a Local Process The description of cardiac compartments in the embryonic heart is based historically, for the most part, on its luminal configurations or expansions of the lumen, which are generally not controversial. The description of the development of the walls of the chambers themselves, in contrast,

221

is more complicated. One of the complications is that the differentiation of the chamber myocardium from primary myocardium is a highly-localized and progressive process, in which more and more primary myocardium is gradually recruited to the chambers (Christoffels et al., 2000; Moorman et al., 2000). Parts of the initial myocardial walls differentiate into chamber myocardium and expand, whereas other parts do not. Because of this, in the walls of the developing heart tube, large areas lack the phenotype for the developing chamber. In the looping heart, for example, local differentiation and expansion of areas of the primary heart tube results in the development of chambers and chamber myocardium, as represented in Fig. 1. At 9.5 days of development, it is already an easy matter to distinguish the atrial and ventricular components of the mouse heart. These components, at this stage, are separated by an atrioventricular canal, in which the paired endocardial cushions can be seen. At this stage, nonetheless, about half of the myocardium of the atrial chamber has the phenotype of primary myocardium (Soufan et al., 2004) (Fig. 2). The atrial floor, or caudal wall, is made exclusively of primary myocardium and is covered with extensions from the endocardial cushions which spread across the mediastinal mesenchyme at the connection of the heart to the body. Despite these differences in myocardial phenotype, we tend to call the entire region the developing atrium. With ongoing development, the atrial chamber myocardium proliferates rapidly, whereas the slower-proliferating primary myocardium differentiates into chamber. Hence, the relative contribution of the primary myocardium to the atrium decreases significantly. A similar situation is seen in the development of the ventricles. In the early stages, the inner curvature of the heart tube has not (yet) differentiated into ventricular chamber myocardium. Hence, it still has a primary phenotype, with the myocardium of the primary right atrioventricular canal in continuity with that of the primary outflow tract in this region. Despite this phenotypic arrangement, the region is still usually considered to be part of the developing ventricle, a view based upon the so-called “segmental” concept of cardiac growth. An alternative view, of course, would be to consider this region as still part of the primary heart tube, not yet recruited to its ultimate fate. This distinction is critical if we are to understand how the proper connections are made between the chambers; the remodeling of this primary myocardium in the inner curvature permits appropriate connection of the right atrium to the right ventricle, and between the left ventricle and the aorta. These connections cannot be made if the inner curvature has already been committed to the left or right ventricular chamber fate at these stages of remodeling (Moorman and Christoffels, 2003a). Several genes have been identified whose expression pattern is restricted to either primary or chamber myocardium (Christoffels et al., 2004a). Most prominent examples

222

PART  |  3  Patterning of the Early Heart Tube

LA

RA OFT

AVC

LV

RSV

LSV

Figure 1  Cranial views of a three-dimensionally reconstructed mouse heart at 9.5 days of development, demonstrating the differentiating atrial and ventricular chambers (red: primary myocardium; blue: chamber myocardium; green: cardiovascular lumen). The reconstruction was made using serial sections incubated with probes for a myocardial marker and the chamber markers Gja5 (encoding connexin 40) and Nppa (encoding natriuretic propeptide A) (Soufan et al., 2004). Note that in the developing ventricular chambers, the trabeculations express the chamber markers connexin40 and natriuretic propeptide A, whereas the forming compact layer expresses these markers at a much lower level (not represented). Note that, via the inner curvature of the looped heart, the primary myocardium of the atrioventricular canal is contiguous with that of the outflow tract (AVC: atrioventricular canal; LA: left atrium; LSV: left systemic vein; LV: left ventricle; OFT: outflow tract; RA: right atrium; RSV: right systemic vein). (See interactive PDF file: http://www.elsevierdirect. com/downloads/PDF/Rosenthal_Interactive_10210.pdf.)

AVC

100 µm Working myocardium

Outflow tract (cut)

Primary myocardium

Lumen

Dorsal view

Ventral view

AVC

RA

LA PVP

RSV

LSV

Figure 2  Detailed view of the inflow of the reconstructed mouse heart, shown in Fig. 1 (red: primary myocardium (removed in the bottom panels); blue: chamber myocardium; green: cardiovascular lumen). Note that the entire floor of the atrium consists of primary myocardium, which is contiguous with the primary myocardium of the atrioventricular canal. The systemic veins enter the heart caudally into primary myocardium, while the pulmonary vein primordium enters the heart dorsally into chamber myocardium, as defined by the expression of Gja5 (encoding connexin 40) (Soufan et al., 2004; Anderson et al., 2006). Some of the molecular mechanisms underlying the formation of the pulmonary myocardium (Mommersteeg et al., 2007a) and the myocardium of surrounding the systemic veins (Christoffels et al., 2006) have been recently unraveled (AVC: atrioventricular canal; LA: left atrium; LSV: left systemic vein; PVP: pulmonary vein primordium; RA: right atrium; RSV: right systemic vein primordium). (See interactive PDF file: http://www. elsevierdirect.com/downloads/PDF/Rosenthal_Interactive_10210.pdf.)

100 µm Working myocardium

Atrioventricular canal (cut)

Primary myocardium

Lumen

are Nppa and Cx40, which in the developing heart are restricted to the differentiating chamber myocardium, and Tbx2 and Tbx3, transcription factor genes expressed exclusively in the primary myocardium. However, many genes have a broader domain of expression in the embryonic heart than subsequent to the completion of septation. For example, Mlc2v and Irx4 are expressed in the ventricular chamber and also in the flanking primary myocardial atrioventricular canal and outflow tract. Their expression profile resembles an anteriorly- and

posteriorly-restricted segment in the heart tube. Never­ theless, these genes are often used as a ventricular-specific marker during cardiac development. A noticeable example of this is the presumed “ventricular chamber-specific” deletion of the cardiac transcription factor gene Nkx2-5, mediated by an Mlc2v-Cre allele, which also results in impairment of the atrioventricular node. As a consequence, the node was assumed to derive from the “ventricular lin­ eage” (Pashmforoush et al., 2004). However, the atrioventricular node derives from the atrioventricular canal and

Chapter  |  3.2  Early Cardiac Growth and the Ballooning Model of Cardiac Chamber Formation

is positioned in the atrial dorsal wall (Davis et al., 2001). Therefore, accurate definition of the patterns of marker genes for one or other cell type is crucial for their use in developmental or lineage studies.

IV.  Fields, lineages and cardiac precursor cells The mesoderm, containing the progenitor cells that will give rise to the heart, has been classically allocated to the cranio-lateral aspect of the flat embryonic disc (Rawles, 1943; Rosenquist and de Haan, 1966; Tam et al., 1997). This mesoderm originates from the cranial part of the primitive streak (Rosenquist, 1970b; Garcia-Martinez and Schoenwolf, 1993), and by the formation of the coelomic cavity subdivides into a splanchnic epithelium lining the endoderm and a somatic epithelium lining the ectoderm. The heart-forming regions (HFR) reside inside the bilateral splanchnic epithelia (Fig. 3). These bilateral heartforming regions will fuse in the embryonic midline to form the so-called cardiac crescent which, concomitantly with the folding of the embryo, transforms into the primary heart tube (Fig. 4). This tube is initially a trough, dorsally not yet closed and attached to the body wall via the dorsal mesocardium. Closure of the myocardial trough, along with the associated rupture of the dorsal mesocardium, starts in the middle of the straight heart tube and proceeds in a

223

zipper-like fashion in a caudal and cranial direction (Argüello et al., 1975; Moreno-Rodriguez et al., 2006). The size of the heart-forming region has been assessed by fate mapping experiments (Ehrman and Yutzey, 1999; Redkar et al., 2001). These studies have almost certainly underestimated the contribution of the heart-forming region, because analysis of such fate maps ended at straight heart tube stages, when the formation of the heart is not yet complete (Moorman and Christoffels, 2003a). Recent lineage studies have revealed that the straight heart tube continues to grow by addition of cardiac progenitor cells (Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001). This additional pool of precursors is currently known as the second heart field which, on the embryonic disc, is localized centrally to the region that will form the cardiac crescent/ straight heart tube and is now called the primary heart field (see Chapter 2.2, Vol. I). In Fig. 3 we show that the pharyngeal mesoderm from the outset is directly confluent with the second heart field. This configuration also explains the observation in retrospective clonal analyses in the mouse of clones present in both the arterial and the venous pole, but not in between the prospective left ventricle, which further endorses the notion of multiple cardiac fields (Meilhac et al., 2003, 2004a,b). This elegant genetic approach visualizes the descendants of rare spontaneous and clonal events of intragenic recombination, which irreversibly activate a reporter gene, allowing the visualization of the genealogy of a single clone within the myocardium. Based

Extra-embryonic mesoderm Amnion Stomato-pharyngeal membrane

1st heart field 2nd heart field

HFR

Hepatogenic mesoderm Pharyngeal mesoderm Yolk sac

Figure 3  Dorsal view of the embryonic disc. The cardiogenic mesoderm is shown on the tri-laminar embryonic disc in the context of the amnion and the yolk sac (red: ectoderm; yellow: mesoderm; green: endoderm). In the region of the stomatopharyngeal membrane the ectoderm has been removed to show the heart-forming regions (HFR) in the splanchnic mesoderm. The periphery of the heart-forming regions is marked with a red line, the central boundary with a blue line. A similar color code is used in Fig. 4. The prospective pharyngeal arch mesoderm lies central to the heart-forming regions (Rosenquist, 1970a), whereas the hepatogenic mesoderm of the septum transversum is located caudo-peripherally (Rosenquist, 1971). Note that the outer edge of the embryonic disc will form the future navel of the embryo, as it becomes localized ventrally with embryonic folding (HFR: heart-forming regions; hn: Hensen’s node; ca: caudal; cr: cranial; L: left; R: right).

224

PART  |  3  Patterning of the Early Heart Tube

Ventral view

Dorsal view Figure 4  Formation of the cardiac tube from the heart-forming regions with embryonic folding, represented by reconstructions of three stages of chicken heart development (yellow: splanchnic mesoderm; gray: myocardium). Folding occurs when the periphery of the embryo moves inwards, with the stomatopharyngeal membrane as an initial hinge point (hatched arrows). The central edges of the heart-forming regions, indicated in blue, retain their original position, being centrally flanked by prospective pharyngeal arch mesoderm (Rosenquist, 1970a). The peripheral edges of the heart-forming regions (indicated in red) form a lumen that will be contiguous with the later vitelline veins (dotted lines). This lumenized periphery of the heart-forming regions fuses in the embryonic midline, recognizable by the fusion line that runs at the ventral side of the straight heart tube. With looping of the heart tube the ventral fusion line rotates towards the right (Manasek et al., 1972). At the cranial pole of the heart tube, splanchnic mesoderm that overlies the pericardial back wall differentiates into myocardium, as also shown in Fig. 6 (HFR: heart-forming regions; dm: dorsal mesocardium; pbw: pericardial back wall; ca: caudal; cr: cranial; L: left; R: right). (See interactive PDF file: http://www.elsevierdirect.com/downloads/PDF/Rosenthal_Interactive_10210. pdf.)

on these analyses, it was concluded that two cardiac lineages contribute to the formation of the heart. A first lineage whose descendants are enriched in the ventricular chamber as seen at E8.5, and a second lineage is particularly implicated in the formation of the primary myocardial flanks and primitive atrial compartment. An additional argument for the existence of a second heart field came from studies on the LIM homeodomain transcription factor Isl1 (Cai et al., 2003). Disruption of the expression of Isl1 arrested cardiac development at the straight tube stage, while analysis of the lineage of Isl1-expressing cells showed that only the forming left ventricle was devoid of Isl1 progeny (Cai et al., 2003). Based on these experiments, it was assumed that the straight heart tube contains little more than the precursors of the left ventricle, and that the remainder of the heart tube is spawned from a separate field of cardiac precursor cells. The concept of multiple heart-forming fields, however, is also debated (Abu-Issa et al., 2004; Moorman et al.,

2007). A cause for this debate lies in the intricate architectural changes of the splanchnic mesoderm, seriously interfering with the interpretation of the observations. Folding of the embryo causes the peripheral part of the heart-forming region (red line in Fig. 4) to swing inward and attain a ventral position. The central part of the heart-forming region (blue line in Fig. 4) retains its central position, in close association with the forming pharynx (Stalsberg and de Haan, 1969; Abu-Issa and Kirby, 2008). The peripheral part of the heart-forming region gradually lumenizes to form two bilateral vessels that are contiguous with the vitelline veins. Subsequent to their fusion, the wall of these vessels initiates the expression of sarcomeric proteins, and shortly later starts rhythmical contractions. The firstly-fused part of the vessels contributes to the future left ventricle (reviewed in van den Berg and Moorman, 2009). The central, “mediastinal” part of the heart-forming region, which stays in close contact with the endoderm of the forming pharynx, essentially forms the pericardial back wall, which is now said

Chapter  |  3.2  Early Cardiac Growth and the Ballooning Model of Cardiac Chamber Formation

to contain the second heart field (reviewed in Buckingham et al., 2005). To facilitate the understanding of the complex morphogenetic process that results in the formation of the heart tube, we propose that the reader imagines placing one’s hands on the heart-forming region as seen from a dorsal view (Fig. 3; bottom panel Fig. 4), with the tips of the fingers pointing cranially and the wrists pointing caudally. The little fingers are now the red peripheral line of the heartforming region. If one now moves the little fingers towards the midline the hands form the cardiac trough, with a fusion seam in the ventral midline. The thumbs point to the blue central line and form the dorsal part of the tube, which is connected to the pericardial wall via the dorsal mesocardium. The dorsal mesocardium is the contact between the pericardial back wall and the myocardium of the forming heart tube. After rupture of this mesocardium, the pericardial back wall (or second heart field), only contacts the myocardial heart tube at its venous and arterial poles, implying that the addition of cardiac precursor cells to the heart can only occur at these sites. This description of the formation of the heart tube suggests that the original heartforming region is not the equivalent of the first heart field, but also encompasses (at least part of) the second heart field (reviewed in van den Berg and Moorman, 2009). In line with the view of just a single heart field are several studies showing that Isl1 is expressed in the entire heart-forming mesoderm of different species or, in fact, in the entire splanchnic lining of the intraembryonic coelomic cavity, and that expression in the precardiac mesoderm was lost shortly after differentiation into myocardium (Yuan and Schoenwolf, 2000; Brade et al., 2007; Prall et al., 2007). Therefore, Isl1 should be considered a pan-myocardial marker, rather than a hallmark gene of the second heart field (Prall et al., 2007). The fact that the left ventricle develops from the initially fused heart-forming region might explain why this structure appeared negative in Cremediated Isl1 lineage analyses (Cai et al., 2007). The initially fused heart-forming region differentiates rapidly into myocardium, and thus only expresses Isl1 very briefly. This short period of Isl1-expression might not cause sufficient Cre-expression to cause recombination. Indeed, a more sensitive analysis of the Isl1 lineage recently showed all myocardium to originate from Isl1-expressing cells (Ma et al., 2008). To the best of our knowledge, no genes have been described that exclusively identify the first or second heart field, suggesting that these fields may initially not be distinct or, at least, may not be two distinctive entities. However, some precursors clearly differentiate earlier (first heart field) than others (second heart field), indicating differential control on loss of precursor state and timing of differentiation. Moreover, the precursors that differentiate later are much more sensitive to genetic perturbation of several genes than the first (Buckingham et al., 2005). If the supposed fields are not discrete territories of gene expression, do they represent distinct cellular ­ lin­eages

225

that give rise to distinct components of the formed heart that cannot be interchanged? The above-mentioned retrospective clonal analysis (Buckingham et al., 2005) consistently revealed clones spanning two or more adjacent cardiac compartments, as well as clones exclusively present in the compartments at the extremities of the tube and clones exclusively found in the middle part of the cardiac tube, the prospective ventricle. The interpretation of these findings is that two distinct lineages contribute to the formation of the heart (Meilhac et al., 2004a), the first lineage contributing to both ventricles, the atrioventricular canal and both atrial chambers, the second lineage excluded from the left ventricle. Clones observed to also contribute to the left ventricle were large, suggesting that the founder cell was labeled before the lineage segregation. It remains to be proven, however, that this lineagesplit indeed occurs, and whether the cells of these lineages have obtained and interpreted information that specifies or determines their fate. An alternative view of these results is that the first lineage represents descendants of labeled precursor cells that have differentiated first, forming the initial heart tube (prospective left ventricle and adjacent compartments). Descendants of precursor cells labeled subsequently will automatically give rise to clones in compartments exclusively at the extremities of the tube, because the breakthrough of the dorsal mesocardium no longer permits addition of cells to the heart tube via this structure. Irrespective of such differences in interpretation, the data certainly demonstrate that none of the cardiac compartments of the formed heart derive from an exclusive lineage, but rather support a gradual formation of the heart (van den Berg and Moorman, 2009). In this view the second heart-forming field can be considered as an expansion of the primary heart field to accommodate the formation of the cardiac components required to sustain pulmonary circulation (Moorman et al., 2007). Recent studies from our own laboratory, as yet unpublished, have demonstrated the existence of a third population of cardiac precursor cells which is involved in the formation of the sinus horn myocardium. This population is located at the most peripheral border of the heart-forming region (Hoogaars et al., 2007). These cells express Tbx18. If the expression of this gene is disrupted, the formation of the sinus horns does not occur (Christoffels et al., 2006).

V.  Cardiac growth V.A.  Growth by Addition As discussed above, heart tubes of higher vertebrates grow by addition of myocardium at both their poles. This was long recognized by the classic cardiac morphologists (Virágh and Challice, 1973; Argüello et al., 1975). Romanov writes in his magisterial textbook published in

226

1960 (Romanoff, 1960) that: “The heart tube at the 9- to 12-somite stage represents only the ventricle and the bulbus (the region between the ventricle and the truncus arteriosus). The material that is to provide the remainder of the heart lays along the roots of the paired omphalomesenteric (or vitelline) veins”. Molecular lineage studies, summarized in Buckingham et al. (2005), along with the recent demonstration of a caudal growth center at the roots of the vitelline veins (van den Berg et al., 2009), have provided convincing proof for this process of growth. The forming heart tube impresses as an inverted “Y”. The forming stem of the straight heart tube contains the precursors for the left ventricle, while the legs of the stem at the venous pole will contain the precursors for part of the atrial chambers (De la Cruz et al., 1989; De la Cruz and Sanchez-Gomez, 1998). The remainder that is to be added comes from the so-called second heart field, as discussed above. Cell­tracing experiments performed in our laboratory (Rana et al., 2007) revealed that in the chick, as in the mouse, the stem of the straight heart tube harbors the precursors of the left ventricle, rather than also containing the precursors for the trabeculated portion of the right ventricle, as previously believed (De la Cruz et al., 1989). These observations resolved a longstanding enigma and demonstrate that the process of early heart formation is conserved in evolutionary terms between birds and mammals. Regardless of the debate regarding multiple heart-forming fields, the concept of cardiac growth by recruitment is rapidly gaining grace and is replacing the conventional, but erroneous view that the primary heart tube from the outset possesses all the compartments of the definitive postnatal heart (Davis, 1927; Srivastava and Olson, 2000).

V.B.  Formation of the Heart Tube from a Single Caudal Growth Center Recent studies of growth of the early chick heart from our laboratory showed that proliferation of cardiac precursors decreased dramatically on overt differentiation into heart muscle and incorporation in the heart tube, which shows its first contractions at Hamburger and Hamilton stage (HH) 10 (Sabin, 1920; DeHaan, 1965). Nonetheless, in a period of about two hours of development, spanning stage 10-minus to stage 10, the number of cardiomyocytes in the primary heart tube increases almost 4-fold (Soufan et al., 2006). The cell-cycle time of these cardiomyocytes was, however, found to be approximately 5.5 days which, in the dynamic context of a 2-day-old chick embryo, can be considered too low to account for any significant growth of the heart tube (van den Berg et al., 2009). Comparable data, as yet unpublished, were obtained in the mouse embryo. These observations are fully compatible with the notion of growth by addition of progenitors to the forming heart tube, and indicate that addition of cells is the sole

PART  |  3  Patterning of the Early Heart Tube

mode of growth of the early heart tube. Previous studies also showed that the myocytes within the straight cardiac tube proliferate at an inherently slow rate (Sissman, 1966; Stalsberg and de Haan, 1969; Rychter et al., 1979). Together these findings disprove the hypothesis that the early heart tube is highly proliferative, with cells withdrawing from the cell-cycle to specialize into the conduction system while chamber-forming myocardium maintains its initial rate of proliferation (Thompson et al., 1990, 1995; Sedmera et al., 2003). It seems reasonable to propose that the primary myocardial heart tube contains the precursors for the myocardium of both the chambers and the conduction system (Christoffels and Moorman, 2009). This proposal, of course, must be placed within the context of the dynamics of the continuous addition of myocardium at both poles of the heart. To get a grasp of the recruitment of cells from a pool of external cardiac precursor cells to the heart tube, we made three-dimensional quantitative reconstructions of proliferation rate in the heart-forming mesoderm during the early stages of heart development (Fig. 5 and interactive PDF files) (van den Berg et al., 2009). The rapidly-proliferating bilateral heart-forming region displays slow proliferation at its point of fusion and, on luminization, in its peripheral edges. The caudal and inner mesoderm retained a high proliferation rate. Fluorescent tracing of cells from the caudal growth center showed that juxtaposed cells in this growth center take very different routes into the heart (Fig. 6). The inner mesoderm moved up via the pericardial back wall (also called pharyngeal mesoderm) into the cardiac outflow, while the outer mesoderm was incorporated into the cardiac inflow (Fig.7). Local inhibition of proliferation in this caudal growth center hampered growth at both poles of the heart tube (van den Berg et al., 2009). Based on these observations it is likely that the early heart forms a single growth center with different routes of addition as visualised in Figure 7. Whether cells that follow these different routes already constitute different lineages (Meilhac et al., 2004a) remains to be assessed. Arguing against such a lineage specification are transplantation studies of De Haan and co-workers, showing that the precardiac mesoderm is not irreversibly determined along the cranio–caudal axis (Satin et al., 1987, 1988).

V.C.  Growth of the Chambers Taking into account the concept of cardiac growth by addition of cardiac precursor cells, the developing heart could be represented as a transversally-segregated structure, to which additional parts are added in a manner as occurs during somitogenesis (Davis, 1927; De la Cruz et al., 1989; Markwald et al., 1998). Another option is to consider the developing heart as a tube with an inflow and an outflow, in which the caudally added cells have been under the control of retinoic acid to provide them with caudal, atrial identity (Harvey, 2002a; Simoes-Costa et al., 2005).

Chapter  |  3.2  Early Cardiac Growth and the Ballooning Model of Cardiac Chamber Formation

227

Figure 5  Formation of the heart tube from a single caudal growth center. Local proliferation rate, as measured by bromodeoxy-uridine positive nuclear fraction (after 1 hour of exposure to bromodeoxy-uridine) is mapped onto the splanchnic mesoderm and the developing myocardium and shown in pseudo colors, ranging from blue (0%) to yellow (100%) (van den Berg et al., 2009). Refer to Fig. 4 for details regarding the morphology and morphogenesis. Proliferation rate within the heart-forming regions regresses on fusion and lumenization of this mesoderm. In the caudal and inner part of the heart-forming regions, a center of rapid proliferation remains (arrowheads).

Figure 6  Cell tracing of caudal splanchnic mesoderm (van den Berg et al., 2009). Cells from the caudal proliferating growth center are dispatched into both poles of the developing heart tube. The luminizing outer mesoderm (red), originating from the rapidly-proliferating heart-forming regions, is incorporated into the inflow. The rapidly-proliferating inner part of the heart-forming regions dispatches cells to the pericardial back wall, from where these cells are incorporated into the outflow pole of the heart.

The latter model is essentially a bipartite model, comprising only atrial and ventricular components arranged along the cranio–caudal axis. In this respect, it is well-established that the cardiac tube is patterned along a caudal-to-cranial, or inflow-to-outflow, axis (Osmond et al., 1991; Yutzey et al., 1994; Xavier-Neto et al., 1999). Such patterning permits formation of the atrial chamber at the caudal inflow end of the tube, and the ventricular chamber at the cranial or outflow end. Patterns of gene expression, and the localized formation of atrial and ventricular trabeculated myocardium at the outer curvatures of the heart, do not support a model of cardiac growth that depends exclusively on the specification of cranio–caudally arranged transverse components (Fig. 7) (Christoffels et al., 2000; Moorman et al.,

2000; Moorman and Christoffels, 2003a). The definitive chambers are also patterned along the dorso–ventral axis (Moorman and Christoffels, 2003a). Subsequent to the formation of the primary tube, exclusively at the ventral side of the forming heart tube, a focus of cells increases in size. These cells then reinitiate proliferation (Fig. 8) (Soufan et al., 2006). This observation is in line with earlier studies (Rychter et al., 1979) in which growth from such a proliferative center had been proposed. With further differentiation of the ventricle, a transmural pattern of proliferation emerges. The cardiomyocytes in the compact layer display a very short cell-cycle time of 8.5 hours, whereas the trabeculated myocardium displays a much slower rate of proliferation (van den Berg

228

PART  |  3  Patterning of the Early Heart Tube

Figure 7  Illustration of early cardiac morphogenesis from a single growth center (van den Berg et al., 2009) (red: rapid proliferation; blue: slow proliferation; yellow: lumen; green: endoderm). The transition from caudal, early in development, to cranial, later in development is depicted. Outer heart-forming mesoderm luminizes and regresses its proliferation rate (1–2). With embryonic folding, this outer mesoderm swings towards the embryonic midline and fuses, forming the ventral wall of the heart tube (2–4). The inner mesoderm keeps proliferating and forms the pericardial back wall and its connection with the heart tube (5). The model on the right shows that expansion from the caudal growth center leads to a radial addition of caudal and inner located mesoderm to the heart tube. After regression of the dorsal mesocardium, addition to the arterial pole occurs via the pericardial back wall.

et al., 2009). This mode of proliferation, with the compact layer as the ventricular growth center, is in line with lin­ eage analyses that show how the trabecules grow out of the compact layer (Mikawa et al., 1992). A recent study used clonal analysis to address the mode of growth of the chambers in the mouse (Meilhac et al., 2004b). The presence of small and large elongated clones in the ventricular region indicated directional clonal growth of the chambers. The basis of this conclusion is that growth rates in the developing heart are regionally and temporally homogeneous (Meilhac et al., 2004b). Nonetheless, in embryonic chick hearts we observed remarkably patterned ventricular growth, with highest proliferation in the region where the endocardium approaches the myocardium, and with proliferation gradually tapering off towards the periphery of this focus (Fig. 8). Such inhomogeneous growth rates could well contribute to the formation of small and large elongated clones in retrospective clonal analysis, an issue which should be addressed in the future. Whereas the differentiation of cardiac precursors into cardiomyocytes shows an inverse relationship between

proliferation and differentiation which is remarkably reminiscent of the differentiation of skeletal muscle (Stockdale and Holtzer, 1961), the formation of the ventricle shows simultaneous proliferation and differentiation. The rapidlyproliferating region of ventricular expansion is also distinguished by the expression of ventricular chamber-specific genes, as is already well-documented in the mouse (O’Dell et al., 1994; van Kempen et al., 1996; Delorme et al., 1997; Dunwoodie et al., 1998; Christoffels et al., 2000; Palmer et al., 2001; Houweling et al., 2002). Together, these proliferation, lineage and gene expression data demonstrate the presence of a program for morphogenetic patterning of the definitive cardiac chambers along both the caudal–cranial and dorso–ventral axes (Christoffels et al., 2000; Moorman et al., 2000; Moorman and Christoffels, 2003a). This regional growth and differentiation of chamber myocardium is now well-recognized as the “ballooning” model, as opposed to the earlier model of segregated differentiation of chamber myocardium within transversely arranged compartments of the heart tube, each transverse part considered to represent a compartment as present in the formed heart.

Chapter  |  3.2  Early Cardiac Growth and the Ballooning Model of Cardiac Chamber Formation

229

Figure 8  Ventricular growth by localized reinitiation of proliferation (Soufan et al., 2006). Embryos were exposed to bromodeoxy-uridine for a longer period of time than the embryos shown in Fig. 5. Therefore, due to the addition of bromodeoxy-uridine-labeled precursors from the caudal growth center to the myocardial lineage, myocardium at the inflow of the heart tube is bromodeoxy-uridine positive. Otherwise, most of the heart tube shows a virtual absence of bromodeoxy-uridine incorporation. Proliferation is reinitiated at the outer curvature, the site of differentiation of the primitive ventricle (arrowheads).

V.D.  Fate of Remaining Primary Myocardium As soon as it is possible to recognize the left ventricle on the basis of its local expansion then, by definition, this developing ventricle also possesses an inflow and an outflow region. These regions are represented by the primary myocardium of the atrioventricular canal and the outflow tract, respectively. No sharp boundaries of proliferation or differentiation between these regions were observed. Proliferation rate, for instance, gradually tapers off from the forming ventricle towards the flanking primary myocardium (Soufan et al., 2006). Because these flanking regions maintain a low rate of proliferation during ventricular expansion they become recognizable as constrictions or rings, as was remarkably well-described by Benninghoff as long ago as 1923 (Benninghoff, 1923). From our ongoing lineage studies, as yet unpublished, we anticipate that cells from these flanking regions of primary myocardium still participate in the formation of the definitive ventricular and atrial chambers. In support of this notion, De la Cruz (De la Cruz et al., 1989; De la Cruz and Sanchez-Gomez, 1998), using iron oxide particles, already indicated that the presumptive atrioventricular and outflow regions contribute to the base of the ventricles. Be this as it may, there can be no question that these same regions also contribute to the formation of the components of the

definitive conduction system (see Chapter 2.3, Vol. I). In this respect, from a stance of lineage, naming a structure at one particular stage of development does not imply the structure to be identical in its cellular composition at later developmental stages. The obvious boundaries seen in the embryonic heart, such as the atrioventricular canal or the boundary between inflow and outflow components may not represent strict boundaries in terms of lineage. Indeed, recent lineage studies demonstrated that Tbx2-positive primary atrioventricular canal myocardium substantially contributes to the left ventricular free wall (Aanhaanen et al., 2009). In other words, the initial atrioventricular canal does not become in its entirety, the lower rims, or vestibules, of the definitive atrial chambers, as we have thus far assumed (Wessels et al., 1996). It also has become clear that cells located within the outflow tract at an early stage of development are fated to become part of a ventricle (Zaffran et al., 2004; Rana et al., 2007; Aanhaanen et al., 2009). Retrospective lineage studies did not reveal an exclusive clonal relationship between the cells of the future chambers, as opposed to a unique clonal relationship between the cells of the flanking compartments (Meilhac et al., 2004a). This means that clones are not confined to the borders of the supposed segments, which is not surprising given the intrinsic uncertainty in placing borders during development. These observations

230

underscore the view of a gradual differentiation of the cardiac chambers from primary myocardium, according to a well-defined spatial and temporal program of patterning. The differentiation of chamber myocardium will be the topic of the next paragraph.

VI.  Differentiation of chamber myocardium VI.A.  Patterning of the Compartments The specification of the cardiac myocytes that form the distinct cardiac chambers required an additional level of regulation over and above an evolutionary conserved core network for formation of cardiac myocytes (Olson, 2006). The formation of myocytes from the two heart fields may to be governed by similar cardiogenic signals (Buckingham et al., 2005). The formation of the cardiac chambers and the concomitant formation of the chamber myocardium from the primary myocardium of the heart tube is initiated at specific sites of the heart tube, indicating that positional information is imposed on the myocardium. The role of the retinoic acid pathway in the formation of the cranio–caudal axis has been firmly established (Osmond et al., 1991; Yutzey et al., 1994; Xavier-Neto et al., 1999; Harvey, 2002b; Moorman and Christoffels, 2003a), as well as the involvement of the retinoic acid responsive GATA transcription factors (Jiang et al., 1999; Kostetskii et al., 1999) and Tbx5 (Bruneau et al., 1999; Liberatore et al., 2000). Tbx5 also mediates the regulation of the progression of the cardiac cell-cycle (Goetz et al., 2006). The mode of specification of the chamber myocardium, however, is far from clear. It requires dorso–ventral information, because the chambers develop at the outer curvatures of the S-shaped heart tube (Christoffels et al., 2000; Moorman and Christoffels, 2003a). A number of transcription factors, including Nkx2-5 (Lyons et al., 1995), Foxh1 (von Both et al., 2004), Mef2c (Lin et al., 1997) and Tbx5 (Liberatore et al., 2000; Bruneau et al., 2001), are essential for formation of the chambers. Because expression of these genes is not restricted to the site of formation of the chambers, other factors must also be involved. Obvious candidates are the Hand1 and Hand2 genes. Ventricular expansion and gene expression were found to be dependent on the gene dose of Hand1 and Hand2 (McFadden et al., 2005). Hand1 is crucially important in the maintenance of the balance between proliferation and differentiation of the cardiac precursors (Risebro et al., 2006). Absence of expression of both Hand1 and Hand2 in mouse would appear to be an essential prerequisite for formation of the ventricular septum (Togi et al., 2004, 2006). There are, however, important differences between species. In chick the hand genes are homogenously expressed, and yet a septum is formed. The T-box transcription factors Tbx5 and Tbx20 were also proposed to determine the identity of the left and right ventricles,

PART  |  3  Patterning of the Early Heart Tube

and the position of the ventricular septum. Tbx20 is expressed in the right ventricle and Tbx5 in the left ventricle in the chick, but the situation differs in the mouse (Takeuchi et al., 2003). In mouse, Tbx20 is expressed in the cardiac crescent, and in the entire linear and looping heart (Stennard et al., 2003; Cai et al., 2005; Singh et al., 2005; Stennard and Harvey, 2005; Takeuchi et al., 2005; Hoogaars et al., 2007). Tbx20-null mice display severely hypoplastic hearts, resembling the phenotype seen in mice overexpressing Tbx2 (Christoffels et al., 2004b). Tbx2 and Bmp2, which are normally restricted to the atrioventricular canal and outflow tract, are expressed over the entire heart tube. Tbx20 directly represses Tbx2 in the chamber-forming regions, allowing proliferative expansion and differentiation of chamber myocardium. How Tbx2 escapes this repressive action of Tbx20 in the atrioventricular canal and outflow tract is unknown. Bmp2/4 mediated signaling in these regions is likely to be involved. Notch, Hey1 and Hey2 may provide a signaling cascade which delimits the expression of Bmp2 and Tbx2 to the atrioventricular canal and inner curvature (Rutenberg et al., 2006).

VI.B.  Patterning and Formation of the Trabecular Ventricles Although initial formation of the chambers involves little more than a local widening of the original primary heart tube, it is almost immediately followed by the formation of trabeculations which are most conspicuous at the apices of the forming ventricles; they are more limited in the forming atrial chambers, where they form the pectinate muscles. In the very early stages of chamber formation (at E9.0 in the mouse) the ventricle, including the thin-walled outer layer, is almost entirely composed of phenotypic trabecular myocardium. At this stage, markers like Cx40 and Natriuretic propeptide A (Nppa), are expressed in the entire ventricular myocardial component, but become confined to the trabeculations as the compact myocardium thickens. Cx43, on the other hand, remains expressed in both the trabecular and compact layers of the ventricular wall (Christoffels et al., 2004a). Many genes that are confined to the atrial myocardium later in development remain expressed in the trabecular component, whereas calciumhandling genes are upregulated in the compact myocardium and remain low in the trabecular component (Franco et al., 1997; Moorman et al., 1998). In many respects, therefore, the trabecular layer of the ventricular walls has an ambiguous phenotype; it resembles the phenotype of the primary myocardium, but the differentiation of the sarcolemma in cells of the trabeculations is highly-advanced and this is reflected by high conduction velocities owing to the high expression of genes such as Cx40 and Scn5a that provide fast conduction. An intriguing aspect of the formation of the ventricles is the appearance of the highly-intricate trabecular architecture.

Chapter  |  3.2  Early Cardiac Growth and the Ballooning Model of Cardiac Chamber Formation

This complex transmural architecture cannot simply be the result of careful regulation of cell differentiation and cell proliferation. It also requires the interplay between these processes and other mechanisms that impact on the shape and size of the myocytes. These, in turn, require a balance between extrinsic signaling and intrinsic physical forces, as demonstrated in the development of chambers in zebrafish (Auman et al., 2007). A remarkable feature of the formation of the ventricles, nonetheless, is the highly-patterned proliferative activity (Soufan et al., 2006). Whereas formation starts with a localized increase in cellular volume followed by high proliferative activity, proliferation rapidly ceases in the newly-formed trabeculations (Thompson et al., 1990, 1995; Sedmera et al., 2003) which will subsequently contribute to the peripheral ramifications of the ventricular conduction system. This interiorly-localized component of the ventricular wall, therefore, becomes terminally differentiated to form the cardiac specialized conducting tissues. In reality, it has done no more than maintain its embryonic phenotype, having as its main feature the high expression of Cx40, but retaining the other features of the primary myocardium (see chapter 2.3, Vol. I). The developing compact layer, in contrast, is highly-proliferative, acquiring its adult phenotype at birth when the proliferative activity ceases. Lineage studies in chicks demonstrated that single labeled cells of the cardiogenic mesoderm form cone-shaped clones spanning the entire thickness of the ventricular wall (Mikawa et al., 1991). The shared lineage of the trabecular and the compact layers of the developing ventricular walls imply that local cues are responsible for further differentiation into the peripheral ventricular conduction system on the one hand, and the compact working myocardium on the other. Indeed, endocardial–myocardial interactions regulate trabecule formation and withdrawal from the cell-cycle, whereas epicardial– myocardial interactions regulate proliferation, growth and compact wall formation (Grego-Bessa et al., 2007; Smith and Bader, 2007). Notch signaling is crucially involved in the interaction between endocardium and myocardium (GregoBessa et al., 2007). The process involves Bmp10-dependent proliferative activity in the trabeculations, and EphrinB2dependent activation of the neuregulin/ErbB pathway, which leads to the differentiation of the trabeculations.

VII.  The origin of the components of the chambers in the mature heart VII.A.  Contributions of the Primary Myocardium Flanking the Chambers: The Cranio–Caudal Axis By 9.5 days of development in the mouse, when the heart tube has already looped, the ventricular chambers differentiate and expand ventrally (outer curvature), while the atrial

231

chambers expand by latero–dorsal growth from the primary tube. At this stage, the developing atrial and ventricular components are separated by an atrioventricular canal, and myocardium is still being added to the heart at the outflow and inflow poles. Whether the definitive chambers are formed solely from the highly-proliferative chamber myocardium that is already formed at these early stages (E9.5), or whether recruitment of myocardium takes place from the slowly-proliferating flanking regions to the rapidlyproliferating chamber lineage is not yet known. The classical studies of De la Cruz (De la Cruz et al., 1989) suggest that the presumptive atrioventricular canal and outflow tract of stage 9 chick hearts contribute to the inlets and outlets of the ventricles. Lineage studies from our laboratory (Aanhaanen et al., 2009), demonstrated that the ventricular myocardium at stage E9.5 largely contributes to the ventricular septum, while the free walls of both ventricles are derived from the flanking regions. This indicates that the ventricular chambers develop by recruitment of flanking primary myocardium to the chamber lineage, followed by highly-organized growth of the recruited myocardium. These observations again match well with the classic lineage studies of de la Cruz and coworkers, who demonstrated that in chick, deposits of iron oxide particles placed on the ventro–medial fusion line of the heart-forming region (prospective embryonic ventricle) eventually marked the ventricular septum (De la Cruz et al., 1997). They also concur with recent lineage studies, again performed in chicks, which demonstrate formation of the trabeculated right ventricular free wall by ventricularization of the myocardium initially forming the outflow tract (Rana et al., 2007). The differences in morphology between the two ventricles are not a reflection of the left–right asymmetry influencing the development of the atrial chambers, because the left and right ventricles develop along the cranio–caudal axis of the heart tube. The ventricular outlets are derived from the most distal part of the heart, which is unequivocally formed by the secondary heart field. From the outset, this initially undivided outlet is contiguous at the inner curve with the cavity of the developing left ventricle, and at the ventral or outer curvature with the right ventricular lumen. Rotation of the myocardial wall of the outflow tract is necessary to produce the characteristic relationship of the intrapericardial arterial trunks (Bajolle et al., 2006). Thus, the basic arrangement of the ventricular components is along the cranio–caudal axis of the ventral side of the original heart tube, with the right ventricle positioned downstream relative to the left ventricle, and the outflow tracts downstream to both ventricles. The original cranio–caudal gradients of gene expression may still impact on the chamber phenotype in the formed heart. In this context, it is worthy of note that the right ventricle and outflow tract are much more prone to arrhythmias than the left ventricle (Chinushi et al., 2002; Morita et al., 2003; Kaplan et al., 2004; van Rijen

232

et al., 2004; van Veen et al., 2005). Whether this phenomenon reflects the original cranio–caudal gradients of gene expression could prove to be of great clinical interest. Unlike the ventricular chambers, both atrial chambers develop laterally from the same part of the primary heart tube. This difference underscores the effects of lateralization in the heart and explains why only the atrial appendages, which are the major components expanding from the primary atrial component, can be isomeric. Be this as it may, the contribution of the myocardium of the atrioventricular canal and of the inflow tract to the definitive atrial chambers has still to be assessed. The contribution of the atrioventricular canal to the formation of the atrial vestibules remains to be assessed in unambiguous fashion. Aside from the issues of lineage (Aanhaanen et al., 2009), the developmental expression pattern of GlN2 at the inner curvature of the human heart provides strong evidence to support the notion that the original atrioventricular canal contributes to the atrial side of the atrioventricular junctions and forms the atrial vestibules (Wessels et al., 1996). The development of the venous pole of the heart recently became a focus of attention. In lower vertebrates, such as the fish, there are no lungs. The only venous structures, therefore, are the systemic veins which enter the heart via a thin-walled muscular structure called the “sinus venosus” (Anderson et al., 2006). In mouse, this structure first develops after 9.5 days of development from an Nkx2-5-negative precursor population (Soufan et al., 2004; Christoffels et al., 2006). Its development is dependent on the T-box transcription factor Tbx18. Islet1 is not expressed in this precursor population, permitting this population to be distinguished genetically from the pulmonary myocardium which surrounds the entrance of the pulmonary veins. The pulmonary veins drain to the atrium via the dorsal mesocardium, via a solitary orifice in the mouse, and initially also via a solitary orifice in man. The systemic veins, in contrast, enter the atrium caudally. Thus, the systemic and pulmonary venous structures are anatomically and molecularly distinctive. An example of the distinctiveness is that Tbx18-deficient mice have severe defects of the sinus horns (systemic venous pole), whereas the pulmonary vein(s) are not affected (Christoffels et al., 2006). In addition, the systemic venous pole is intimately associated with the conduction system (e.g., sinus node), whereas the pulmonary vein adopts an atrial working myocardial phenotype readily after its formation (Mommersteeg et al., 2007a). Therefore, we feel that referring to the entire venous pole (systemic and pulmonary) of the heart as the “sinus venosus” (deRuiter et al., 1995) underexposes these differences which, in turn, may lead to the erroneous conclusion that these venous structures are one entity under the control of the same genetic regulatory programs, and that the entire region of the supposed “sinus venosus” is involved in the development of the conduction system (Blaschke et al., 2007). As mentioned above, the pulmonary,

PART  |  3  Patterning of the Early Heart Tube

or mediastinal, myocardium has its own distinct lin­eage which is derived from the second heart-forming field. It forms a wedge of myocardium, including the atrial septum that interposes between the atria. We speculate that this part of the atrium is added to the heart during evolution, concomitant with the need to establish a pulmonary circulation.

VIII.  Concluding thoughts The local reinitiation of proliferation in the primary myocardial heart tube indicates that the formation of the myocardium of the chambers, in contrast to the initial formation of the primary myocardium, represents a localized process. As yet, the precise cues governing this local process of growth and differentiation are unknown. As the working myocardium of the chambers is formed, the nondifferentiating and nonexpanding flanking components (atrioventricular canal, outflow tract) become visible. The subsequent temporal repression by Tbx2, Tbx3 and other factors of the formation of working myocardium in these flanking regions initially permits these regions to function as peristaltic valves prior to the formation of the definitive valvar leaflets. The subsequent fate of the flanking regions remains to be analyzed in sufficient detail to understand their eventual role in the formation and alignment of the atrial and ventricular chambers, in the formation of the valves and in the formation of the conduction system. It seems, nonetheless, that none of the definitive cardiac chambers derives from a single lineage. How the formation of the definitive chambers relates to the alleged cardiac fields remains an important issue for future research. It will be crucially important to unravel the regional function of cardiac transcription factors such as Nkx2-5, which are essential but, nonetheless, play divergent roles in the first and second heart fields and in distinct regions of the heart. For example, Nkx2-5 is critically involved in the regulation of proliferation of the cardiac precursor cells in the second heart field, and for chamber differentiation and conduction system patterning in the formed myocardium (Prall et al., 2007). But its absence is crucial for the formation of the venous pole of the heart, because in this region it represses the expression of nodal genes in the atria allowing the sinus venosus part of the forming heart to function as the leading pacemaker, thus initiating the waves of contraction that lead to unidirectional propulsion of the blood along the cardiac tube (Mommersteeg et al., 2007b). Subsequently, Nkx2-5 becomes activated more broadly as the main part of the sinus venosus becomes transformed into working atrial myocardium. Nkx2-5, therefore, is critical for the formation of both the atrial and the ventricular chambers. In conclusion, although we now understand the mechanisms of basic patterning of the chambers, much has still to be done to understand the detail of their transcriptional regulation.

Chapter  |  3.2  Early Cardiac Growth and the Ballooning Model of Cardiac Chamber Formation

Acknowledgments This work is supported by the Netherlands Heart Foundation, grant 1996M002, and by the European Union FP6 program HeartRepair LSHM-CT 018630. We thank Dr A. T. Soufan for critical reading the manuscript and for help with the preparation of the figures.

References Aanhaanen, W.T.J., Brons, J.F., Domínguez, J.N., Rana, M.S., Norden, J., Airik, R., Wakker, V., de Gier-de Vries, C., Brown, N.A., Kispert, A., Moorman, A.F.M., Christoffels, V.M., 2009. The Tbx2 primary myocardium of the atrioventricular canal forms the atrioventricular node and the base of the left ventricle. Circ. Res. 104, 1267–1274. Abu-Issa, R., Kirby, M.L., 2008. Patterning of the heart field in the chick. Dev. Biol. 319, 223–233. Abu-Issa, R., Waldo, K., Kirby, M.L., 2004. Heart fields: one, two or more? Dev. Biol. 272, 281–285. Anderson, R.H., Brown, N.A., Moorman, A.F., 2006. Development and structures of the venous pole of the heart. Dev. Dyn. 235, 2–9. Argüello, C., De la Cruz, M.V., Gomez, C.S., 1975. Experimental study of the formation of the heart tube in the chick embryo. J. Embryol. Exp. Morphol. 33, 1–11. Auman, H.J., Coleman, H., Riley, H.E., Olale, F., Tsai, H.J., Yelon, D., 2007. Functional modulation of cardiac form through regionally confined cell shape changes. PLoS. Biol. 5, e53. Bajolle, F., Zaffran, S., Kelly, R.G., Hadchouel, J., Bonnet, D., Brown, N.A., Buckingham, M.E., 2006. Rotation of the myocardial wall of the outflow tract is implicated in the normal positioning of the great arteries. Circ. Res. 98, 421–428. Benninghoff, A., 1923. Über die Beziehungen des Reizleitungssystems und der papillarmuskeln zu den Konturfasern des Herzschlauches. Anat. Anz. 57, 185–208. Blaschke, R.J., Hahurij, N.D., Kuijper, S., Just, S., Wisse, L.J., Deissler, K., Maxelon, T., Anastassiadis, K., Spitzer, J., Hardt, S.E., Scholer, H., Feitsma, H., Rottbauer, W., Blum, M., Meijlink, F., Rappold, G., Gittenberger-de Groot, A.C., 2007. Targeted mutation reveals essential functions of the homeodomain transcription factor Shox2 in sinoatrial and pacemaking development. Circulation 115, 1830–1838. Boukens, B.J.D., Christoffels, V.M., Coronel, R., Moorman, A.F.M., 2009. Developmental basis for electrophysiological heterogeneity in the ventricular and outflow tract myocardium as a substrate for lifethreatening ventricular arrhythmias. Circ. Res. 104, 19–31. Brade, T., Gessert, S., Kuhl, M., Pandur, P., 2007. The amphibian second heart field: Xenopus islet-1 is required for cardiovascular development. Dev. Biol. 311, 297–310. Bruneau, B.G., Logan, M., Davis, N., Levi, T., Tabin, C.J., Seidman, J.G., Seidman, C.E., 1999. Chamber-specific cardiac expression of Tbx5 and heart defects in Holt-Oram syndrome. Dev. Biol. 211, 100–108. Bruneau, B.G., Nemer, G., Schmitt, J.P., Charron, F., Robitaille, L., Caron, S., Conner, D.A., Gessler, M., Nemer, M., Seidman, C.E., Seidman, J.G., 2001. A murine model of Holt-Oram syndrome defines roles of the T-box transcription factor Tbx5 in cardiogenesis and disease. Cell 106, 709–721. Buckingham, M., Meilhac, S., Zaffran, S., 2005. Building the mammalian heart from two sources of myocardial cells. Nat. Rev. Genet. 6, 826–837. Burggren, W.W., 1988. Cardiac design in lower vertebrates: what can phylogeny reveal about ontogeny? Experientia 44, 919–930.

233

Cai, C.L., Liang, X., Shi, Y., Chu, P.H., Pfaff, S.L., Chen, J., Evans, S., 2003. Isl1 identifies a cardiac progenitor population that proliferates prior to differentiation and contributes a majority of cells to the heart. Dev. Cell 5, 877–889. Cai, C.L., Zhou, W., Yang, L., Bu, L., Qyang, Y., Zhang, X., Li, X., Rosenfeld, M.G., Chen, J., Evans, S., 2005. T-box genes coordinate regional rates of proliferation and regional specification during cardiogenesis. Development 132, 2475–2487. Canale, E.D., Campbell, G.R., Smolich, J.J., Campbell, J.H., 1986. Cardiac Muscle. Springer-Verlag, Berlin. Chinushi, M., Washizuka, T., Chinushi, Y., Higuchi, K., Toida, T., Aizawa, Y., 2002. Induction of ventricular fibrillation in Brugada syndrome by site-specific right ventricular premature depolarization. Pacing. Clin. Electrophysiol. 25, 1649–1651. Christoffels, V.M., Habets, P.E., Franco, D., Campione, M., de Jong, F., Lamers, W.H., Bao, Z.Z., Palmer, S., Biben, C., Harvey, R.P., Moorman, A.F., 2000. Chamber formation and morphogenesis in the developing mammalian heart. Dev. Biol. 223, 266–278. Christoffels, V.M., Burch, J.B., Moorman, A.F., 2004a. Architectural plan for the heart: early patterning and delineation of the chambers and the nodes. Trends Cardiovasc. Med. 14, 301–307. Christoffels, V.M., Hoogaars, W.M., Tessari, A., Clout, D.E., Moorman, A.F., Campione, M., 2004b. T-box transcription factor Tbx2 represses differentiation and formation of the cardiac chambers. Dev. Dyn. 229, 763–770. Christoffels, V.M., Mommersteeg, M.T., Trowe, M.O., Prall, O.W., de Gier-de Vries, C., Soufan, A.T., Bussen, M., Schuster-Gossler, K., Harvey, R.P., Moorman, A.F., Kispert, A., 2006. Formation of the venous pole of the heart from an Nkx2-5-negative precursor population requires Tbx18. Circ. Res. 98, 1555–1563. Christoffels, V.M., Moorman, A.F., 2009. Development of the cardiac conduction system, why are some regions of the heart more arrhythmogenic than others? Circulation 2, 195–207. Davis, C.L., 1927. Development of the human heart from its first appearance to the stage found in embryos of twenty paired somites. Contrib. Embryol. 19, 245–284. Davis, D.L., Edwards, A.V., Juraszek, AL., Phelps, A., Wessels, A., Burch, J.B., 2001. A GATA-6 gene heart-region-specific enhancer provides a novel means to mark and probe a discrete component of the mouse cardiac conduction system. Mech. Dev. 108, 105–119. de Jong, F., Opthof, T., Wilde, A.A., Janse, M.J., Charles, R., Lamers, W.H., Moorman, A.F., 1992. Persisting zones of slow impulse conduction in developing chicken hearts. Circ. Res. 71, 240–250. De la Cruz, M.V., Sánchez-Gómez, C., Palomino, M., 1989. The primitive cardiac regions in the straight tube heart (stage 9) and their anatomical expression in the mature heart: an experimental study in the chick embryo. J. Anat. 165, 121–131. De la Cruz, M.V., Castillo, M.M., Villavicenzio, G., Valencia, A., MorenoRodriguez, R.A., 1997. Primitive interventricular septum, its primordium and its contribution in the definitive interventricular septum: in vivo labelling study in the chick embryo heart. Anat. Rec. 247, 512–520. De la Cruz, M.V., Sanchez-Gomez, C., 1998. Straight tube heart. Primitive cardiac cavities vs. primitive cardiac segments. In: De la Cruz, M., Markwald, R.R. (Eds.), Living Morphogenesis of the Heart, Chapter 3, Vol. 1. Birkhäuser, Boston, MA, pp. 85–98. DeHaan, R.L., 1965. Morphogenesis of the vertebrate beart. In: DeHaan, R.L., Ursprung, H. (Eds.), Organogenesis. Holt, Reinhart and Winston, New York, NY, pp. 377–419. Delorme, B., Dahl, E., Jarry-Guichard, T., Briand, J.P., Willecke, K., Gros, D., Théveniau-Ruissy, M., 1997. Expression pattern of

234

connexin gene products at the early developmental stages of the mouse cardiovascular system. Circ. Res. 81, 423–437. deRuiter, M.C., Gittenberger-de Groot, A.C., Wenink, A.C., Poelmann, R.E., Mentink, M.M., 1995. In normal development pulmonary veins are connected to the sinus venosus segment in the left atrium. Anat. Rec. 243, 84–92. Dunwoodie, S.L., Rodriguez, T.A., Beddington, R.S., 1998. Msg1 and Mrg1, founding members of a gene family, show distinct patterns of gene expression during mouse embryogenesis. Mech. Dev. 72, 27–40. Ehrman, L.A., Yutzey, K.E., 1999. Lack of regulation in the heart forming region of avian embryos. Dev. Biol. 207, 163–175. Franco, D., Ya, J., Wagenaar, G.T., Lamers, W.H., Moorman, A.F., 1997. The trabecular component of the embryonic ventricle. In: Ost’Adal, B., Nagano, M., Takeda, N., Dhalla, N.S. (Eds.), The Developing Heart. Lippincott-Raven, Philadelphia, PA, pp. 51–60. Garcia-Martinez, V., Schoenwolf, G.C., 1993. Primitive streak origin of the cardiovascular system in avian embryos. Dev. Biol. 159, 706–719. Goetz, S.C., Brown, D.D., Conlon, F.L., 2006. TBX5 is required for embryonic cardiac cell cycle progression. Development 133 (13), 2575–2584. Grego-Bessa, J., Luna-Zurita, L., del Monte, G., Bolos, V., Melgar, P., Arandilla, A., Garratt, A.N., Zang, H., Mukouyama, Y.S., Chen, H., Shou, W., Ballestar, E., Esteller, M., Rojas, A., Perez-Pomares, J.M., de la Pompa, J.L., 2007. Notch signaling is essential for ventricular chamber development. Dev. Cell 12, 415–429. Harvey, R.P., 2002a. Molecular determinants of cardiac development and congenital disease. In: Rossant, J., Tam, P.P.L. (Eds.), Mouse Development, Patterning, Morphogenesis, and Organogenesis, Vol. 1. Academic Press, West mead, Toronto, Canada, pp. 331–370. Harvey, R.P., 2002b. Patterning the vertebrate heart. Nat. Rev. Genet. 3, 544–556. Hoogaars, W.M., Barnett, P., Moorman, A.F., Christoffels, V.M., 2007. T-box factors determine cardiac design. Cell Mol. Life Sci. 64, 646–660. Houweling, A.C., Somi, S., van den Hoff, M.J., Moorman, A.F., Christoffels, V.M., 2002. Developmental pattern of ANF gene expression reveals a strict localization of cardiac chamber formation in chicken. Anat. Rec. 266, 93–102. Icardo, J.M., 2006. Conus arteriosus of the teleost heart: dismissed, but not missed. Anat. Rec. A. Discov. Mol. Cell Evol. Biol. 288, 900–908. Jiang, Y., Drysdale, T.A., Evans, T., 1999. A role for GATA-4/5/6 in the regulation of Nkx2.5 expression with implications for patterning of the precardiac field. Dev. Biol. 216, 57–71. Joyner, R.W., van Capelle, F.J., 1986. Propagation through electrically coupled cells. How a small SA node drives a large atrium. Biophys. J. 50, 1157–1164. Kaplan, S.R., Gard, J.J., Protonotarios, N., Tsatsopoulou, A., Spiliopoulou, C., Anastasakis, A., Prost Squarcioni, C., McKenna, W.J., Thiene, G., Basso, C., Brousse, N., Fontaine, G., Saffitz, J.E., 2004. Remodeling of myocyte gap junctions in arrhythmogenic right ventricular cardiomyopathy due to a deletion in plakoglobin (Naxos disease). Heart. Rhythm. 1, 3–11. Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1, 435–440. Kleber, A.G., Janse, M.J., Fast, V.G., 2001. Normal and abnormal conduction in the heart. In: Page, E., Fozzard, H.A., Solaro, R.J. (Eds.), Handbook of Physiology, Vol. 2001. Oxford University Press, Chicago, IL, pp. 455–530.

PART  |  3  Patterning of the Early Heart Tube

Kostetskii, I., Jiang, Y., Kostetskaia, E., Yuan, S., Zile, M., 1999. Retinoid signaling required for normal heart development regulates GATA-4 in a pathway distinct from cardiomyocyte differentiation. Dev. Biol. 206, 206–218. Liberatore, C.M., Searcy-Schrick, R.D., Yutzey, K.E., 2000. Ventricular expression of Tbx5 inhibits normal heart chamber development. Dev. Biol. 223, 169–180. Lin, Q., Schwarz, J., Bucana, C., Olson, E.N., 1997. Control of mouse cardiac morphogenesis and myogenesis by transcription factor MEF2C. Science 276, 1404–1407. Lyons, I., Parsons, L.M., Hartley, L., Li, R., Andrews, J.E., Robb, L., Harvey, R.P., 1995. Myogenic and morphogenetic defects in the heart tubes of murine embryos lacking the homeo box gene Nkx2-5. Gene. Dev. 9, 1654–1666. Ma, Q., Zhou, B., Pu, W.T., 2008. Reassessment of Isl1 and Nkx2-5 cardiac fate maps using a Gata4-based reporter of Cre activity. Dev. Biol. 323, 98–104. Manasek, F.J., Burnside, M.B., Waterman, R.E., 1972. Myocardial cell shape change as a mechanism of embryonic heart looping. Dev. Biol. 29, 349–371. Markwald, R.R., Trusk, T., Moreno-Rodriguez, R.A., 1998. Formation and septation of the tubular heart: integrating the dynamics of morphology with emerging molecular concepts. In: De la Cruz, M.V., Markwald, R.R. (Eds.), Living Morphogenesis of the Heart. Birkhäuser, Boston, MA, pp. 43–84. McFadden, D.G., Barbosa, A.C., Richardson, J.A., Schneider, M.D., Srivastava, D., Olson, E.N., 2005. The Hand1 and Hand2 transcription factors regulate expansion of the embryonic cardiac ventricles in a gene dosage-dependent manner. Development 132, 189–201. Meilhac, S.M., Kelly, R.G., Rocancourt, D., Eloy-Trinquet, S., Nicolas, J.F., Buckingham, M.E., 2003. A retrospective clonal analysis of the myocardium reveals two phases of clonal growth in the developing mouse heart. Development 130, 3877–3889. Meilhac, S.M., Esner, M., Kelly, R.G., Nicolas, J.F., Buckingham, M.E., 2004a. The clonal origin of myocardial cells in different regions of the embryonic mouse heart. Dev. Cell 6, 685–698. Meilhac, S.M., Esner, M., Kerszberg, M., Moss, J.E., Buckingham, M.E., 2004b. Oriented clonal cell growth in the developing mouse myocardium underlies cardiac morphogenesis. J. Cell Biol. 164, 97–109. Miake, J., Marban, E., Nuss, H.B., 2002. Biological pacemaker created by gene transfer. Nature 419, 132–133. Mikawa, T., Borisov, A., Brown, A.M., Fischman, D.A., 1992. Clonal analysis of cardiac morphogenesis in the chicken embryo using a replication-defective retrovirus. I. Formation of the ventricular myocardium. Dev. Dyn. 193, 11–23. Mikawa, T., Fischman, D.A., Dougherty, J.P., Brown, A.M., 1991. In vivo analysis of a new lacZ retrovirus vector suitable for cell lineage marking in avian and other species. Exp. Cell Res. 195, 516–523. Mjaatvedt, C.H., Nakaoka, T., Moreno-Rodriguez, R.A., Norris, R.A., Kern, M.J., Eisenberg, C.A., Turner, D., Markwald, R.R., 2001. The outflow of the heart is recruited from a novel heart forming field. Dev. Biol. 238, 97–109. Mommersteeg, M.T., Brown, N.A., Prall, O.W., de Gier-de Vries, C., Harvey, R.P., Moorman, A.F., Christoffels, V.M., 2007a. Pitx2c and Nkx2-5 are required for the formation and identity of the pulmonary myocardium. Circ. Res. 101, 902–909. Mommersteeg, M.T., Hoogaars, W.M., Prall, O.W., de Gier-de Vries, C., Wiese, C., Clout, D.E., Papaioannou, V.E., Brown, N.A., Harvey, R.P., Moorman, A.F., Christoffels, V.M., 2007b. Molecular

Chapter  |  3.2  Early Cardiac Growth and the Ballooning Model of Cardiac Chamber Formation

pathway for the localized formation of the sinoatrial node. Circ. Res. 100, 354–362. Moorman, A.F., de Jong, F., Denyn, M.M., Lamers, W.H., 1998. Development of the cardiac conduction system. Circ. Res. 82, 629–644. Moorman, A.F., Schumacher, C.A., de Boer, P.A., Hagoort, J., Bezstarosti, K., van den Hoff, M.J., Wagenaar, G.T., Lamers, J.M., Wuytack, F., Christoffels, V.M., Fiolet, J.W., 2000. Presence of functional sarcoplasmic reticulum in the developing heart and its confinement to chamber myocardium. Dev. Biol. 223, 279–290. Moorman, A.F., Christoffels, V.M., 2003a. Cardiac chamber formation: development, genes and evolution. Physiol. Rev. 83, 1223–1267. Moorman, A.F., Christoffels, V.M., 2003b. Development of the cardiac conduction system: a matter of chamber development. In: Chadwick, D.J., Goode, J. (Eds.), Development of the Cardiac Conduction System, Vol. 250. John Wiley, Chichester, UK, pp. 25–34. Moorman, A.F., Soufan, A.T., Hagoort, J., de Boer, P.A., Christoffels, V.M., 2004. Development of the building plan of the heart. Ann. NY Acad. Sci. 1015, 171–181. Moorman, A.F., Christoffels, V.M., Anderson, R.H., van den Hoff, M.J., 2007. The heart-forming fields: one or multiple?. Phil. Trans. R. Soc. B. 362, 1257–1265. Moreno-Rodriguez, R.A., Krug, E.L., Reyes, L., Villavicencio, L., Mjaatvedt, C.H., Markwald, R.R., 2006. Bidirectional fusion of the heart-forming fields in the developing chick embryo. Dev. Dyn. 235, 191–202. Morita, H., Fukushima-Kusano, K., Nagase, S., Takenaka-Morita, S., Nishii, N., Kakishita, M., Nakamura, K., Emori, T., Matsubara, H., Ohe, T., 2003. Site-specific arrhythmogenesis in patients with Brugada syndrome. J. Cardiovasc. Electrophysiol. 14, 373–379. O’Dell, T.J., Huang, P.L., Dawson, T.M., Dinerman, J.L., Snyder, S.H., Kandel, E.R., Fishman, M.C., 1994. Endothelial NOS and the blockade of LTP by NOS inhibitors in mice lacking neuronal NOS. Science 265, 542–546. Olson, E.N., 2006. Gene regulatory networks in the evolution and development of the heart. Science 313, 1922–1927. Osmond, M.K., Butler, A.J., Voon, F.C., Bellairs, R., 1991. The effects of retinoic acid on heart formation in the early chick embryo. Development 113, 1405–1417. Palmer, S., Groves, N., Schindeler, A., Yeoh, T., Biben, C., Wang, C.C., Sparrow, D., Barnett, L., Jenkins, N.A., Copeland, N.G., Koentgen, F., Mohun, T., Harvey, R.P., 2001. The small muscle-specific protein Csl modifies cell shape and promotes myocyte fusion in an insulinlike growth factor 1-dependent manner. J. Cell Biol. 153, 985–997. Pashmforoush, M., Lu, J.T., Chen, H., Amand, T.S., Kondo, R., Pradervand, S., Evans, S.M., Clark, B., Feramisco, J.R., Giles, W., Ho, S.Y., Benson, D.W., Silberbach, M., Shou, W., Chien, K.R., 2004. Nkx2-5 pathways and congenital heart disease; loss of ventricular myocyte lineage specification leads to progressive cardiomyopathy and complete heart block. Cell 117, 373–386. Plageman, Jr., T.F., Yutzey, K.E., 2005. T-box genes and heart development: putting the “T” in heart. Dev. Dyn. 232, 11–20. Prall, O.W., Menon, M.K., Solloway, M.J., Watanabe, Y., Zaffran, S., Bajolle, F., Biben, C., McBride, J.J., Robertson, B.R., Chaulet, H., Stennard, F.A., Wise, N., Schaft, D., Wolstein, O., Furtado, M.B., Shiratori, H., Chien, K.R., Hamada, H., Black, B.L., Saga, Y., Robertson, E.J., Buckingham, M.E., Harvey, R.P., 2007. An Nkx2-5/ Bmp2/Smad1 negative feedback loop controls heart progenitor specification and proliferation. Cell 128, 947–959. Rana, M.S., Horsten, N.C., Tesink-Taekema, S., Lamers, W.H., Moorman, A.F., van den Hoff, M.J., 2007. Trabeculated right ventricular free

235

wall in the chicken heart forms by ventricularization of the myocardium initially forming the outflow tract. Circ. Res. 100, 1000–1007. Randall, D.J., Davie, P.S., 1980. The hearts of urochordates and cephalochordates. In: Bourne, G.H. (Ed.), Hearts and Heart-like Organs. Academic Press, New York, NY, pp. 41–59. Rawles, M.E., 1943. The heart-forming areas of the early chick blastoderm. Physiol. Zool. 16, 22–42. Redkar, A., Montgomery, M., Litvin, J., 2001. Fate map of early avian cardiac progenitor cells. Development 128, 2269–2279. Risebro, C.A., Smart, N., Dupays, L., Breckenridge, R., Mohun, T.J., Riley, P.R., 2006. Hand1 regulates cardiomyocyte proliferation versus differentiation in the developing heart. Development 133, 4595–4606. Romanoff, A.L., 1960. The Avian Embryo: Structural and Functional Development. The Macmillan Company, New York, NY. Rosenquist, G.C., 1970a. Aortic arches in the chick embryo: origin of the cells as determined by radioautographic mapping. Anat. Rec. 168, 351–359. Rosenquist, G.C., 1970b. Location and movements of cardiogenic cells in the chick embryo: the heart forming portion of the primitive streak. Dev. Biol. 22, 461–475. Rosenquist, G.C., 1971. The origin and movements of hepatogenic cells in the chick embryo as determined by radioautographic mapping. J. Embryol. Exp. Morphol. 25, 97–113. Rosenquist, G.C., de Haan, R.L., 1966. Migration of precardiac cells in the chick embryo: A radioautographic study. Carnegie Inst Washington Publ. 625, (Contributions to Embryology), pp. 111–121. Rutenberg, J.B., Fischer, A., Jia, H., Gessler, M., Zhong, T.P., Mercola, M., 2006. Developmental patterning of the cardiac atrioventricular canal by Notch and Hairy-related transcription factors. Development 133, 4381–4390. Rychter, Z., Rychterova, V., Lemez, L., 1979. Formation of the heart loop and proliferation structure of its wall as a base for ventricular septation. Herz 4, 86–90. Sabin, F.R., 1920. Studies on the origin of blood vessels and of red blood corpuscules as seen in the living blastoderm of chicks during the second day of incubation. Contrib. Embryol. 9, 213–262. Satchell, G.H., Jones, M.P., 1967. The function of the conus arteriosus in the Port Jackson shark, Heterodontus portusjacksoni. J. Exp. Biol. 46, 373–382. Satin, J., Bader, D., de Haan, R.L., 1987. Local cues influence atrial and ventricular differentiation of precardiac mesoderm. J. Mol. Cell Cardiol. 19, S16a. (abstr.). Satin, J., Fujii, S., de Haan, R.L., 1988. Development of cardiac heartbeat in early chick embryos is regulated by regional cues. Dev. Biol. 129, 103–113. Sedmera, D., Reckova, M., DeAlmeida, A., Coppen, S.R., Kubalak, S.W., Gourdie, R.G., Thompson, R.P., 2003. Spatiotemporal pattern of commitment to slowed proliferation in the embryonic mouse heart indicates progressive differentiation of the cardiac conduction system. Anat. Rec. 274A, 773–777. Simoes-Costa, M.S., Vasconcelos, M., Sampaio, A.C., Cravo, R.M., Linhares, V.L., Hochgreb, T., Yan, C.Y., Davidson, B., Xavier-Neto, J., 2005. The evolutionary origin of cardiac chambers. Dev. Biol. 277, 1–15. Singh, M.K., Christoffels, V.M., Dias, J.M., Trowe, M.O., Petry, M., Schuster-Gossler, K., Burger, A., Ericson, J., Kispert, A., 2005. Tbx20 is essential for cardiac chamber differentiation and repression of Tbx2. Development 132, 2697–2707. Sissman, J., 1966. Cell multiplication rates during development of the primitive cardiac tube in the chick embryo. Nature 210, 504–507.

236

Smith, T.K., Bader, D.M., 2007. Signals from both sides: Control of cardiac development by the endocardium and epicardium. Semin. Cell Dev. Biol. 18, 84–89. Soufan, A.T., van den Hoff, M.J., Ruijter, J.M., de Boer, P.A., Hagoort, J., Webb, S., Anderson, R.H., Moorman, A.F., 2004. Reconstruction of the patterns of gene expression in the developing mouse heart reveals an architectural arrangement that facilitates the understanding of atrial malformations and arrhythmias. Circ. Res. 95, 1207–1215. Soufan, A.T., van den Berg, G., Ruijter, J.M., de Boer, P.A., van den Hoff, M.J., Moorman, A.F., 2006. Regionalized sequence of myocardial cell growth and proliferation characterizes early chamber formation. Circ. Res. 99, 545–552. Srivastava, D., Olson, E.N., 2000. A genetic blueprint for cardiac development. Nature 407, 221–226. Stalsberg, H., de Haan, R.L., 1969. The precardiac areas and formation of the tubular heart in the chick embryo. Dev. Biol. 19, 128–159. Stennard, F.A., Costa, M.W., Elliott, D.A., Rankin, S., Haast, S.J., Lai, D., McDonald, L.P., Niederreither, K., Dolle, P., Bruneau, B.G., Zorn, A.M., Harvey, R.P., 2003. Cardiac T-box factor Tbx20 directly interacts with Nkx2-5, GATA4, and GATA5 in regulation of gene expression in the developing heart. Dev. Biol. 262, 206–224. Stennard, F.A., Harvey, R.P., 2005. T-box transcription factors and their roles in regulatory hierarchies in the developing heart. Development 132, 4897–4910. Stockdale, F.E., Holtzer, H., 1961. DNA synthesis and myogenesis. Exp. Cell Res. 24, 508–520. Takeuchi, J.K., Ohgi, M., Koshiba-Takeuchi, K., Shiratori, H., Sakaki, I., Ogura, K., Saijoh, Y., Ogura, T., 2003. Tbx5 specifies the left–right ventricles and ventricular septum position during cardiogenesis. Development 130, 5953–5964. Takeuchi, J.K., Mileikovskaia, M., Koshiba-Takeuchi, K., Heidt, A.B., Mori, A.D., Arruda, E.P., Gertsenstein, M., Georges, R., Davidson, L., Mo, R., Hui, C.C., Henkelman, R.M., Nemer, M., Black, B.L., Nagy, A., Bruneau, B.G., 2005. Tbx20 dose-dependently regulates transcription factor networks required for mouse heart and motoneuron development. Development 132, 2463–2474. Tam, P.P., Parameswaran, M., Kinder, S.J., Weinberger, R.P., 1997. The allocation of epiblast cells to the embryonic heart and other mesodermal lineages: the role of ingression and tissue movement during gastrulation. Development 124, 1631–1642. Thompson, R.P., Lindroth, J.R., Wong, Y.M., 1990. Regional differences in DNA-synthetic activity in the preseptation myocardium of the chick. In: Clark, E.B., Takao, A. (Eds.), Developmental Cardiology: Morphogenesis and Function. Futura Publishing Co, Mount Kisco, NY, pp. 219–234. Thompson, R.P., Kanai, T., Germroth, P.G., Gourdie, R.G., Thomas, P.C., Barton, P.J.R., Mikawa, T., Anderson, R.H., 1995. Organization and function of early specialized myocardium. In: Clark, E.B., Markwald, R.R., Takao, A. (Eds.), Developmental Mechanisms of Heart Disease. Futura Publishing Co, Armonk, NY, pp. 269–279. Togi, K., Kawamoto, T., Yamauchi, R., Yoshida, Y., Kita, T., Tanaka, M., 2004. Role of Hand1/eHAND in the dorso–ventral patterning and interventricular septum formation in the embryonic heart. Mol. Cell Biol. 24, 4627–4635.

PART  |  3  Patterning of the Early Heart Tube

Togi, K., Yoshida, Y., Matsumae, H., Nakashima, Y., Kita, T., Tanaka, M., 2006. Essential role of Hand2 in interventricular septum formation and trabeculation during cardiac development. Biochem. Biophys. Res. Commun. 343, 144–151. van den Berg, G., Moorman, A.F., 2009. Concepts of cardiac development in retrospect. Ped. Cardiol 30 (5), 580–587. van den Berg, G., Abu-Issa, R., de Boer, B.A., Hutson, M.R., de Boer, P.A., Soufan, A.T., Ruijter, J.M., Kirby, M.L., van den Hoff, M.J., Moorman, A.F., 2009. A caudal proliferating growth center contributes to both poles of the forming heart tube. Circ. Res. 104, 179–188. van Kempen, M.J., Vermeulen, J.L., Moorman, A.F., Gros, D., Paul, D.L., Lamers, W.H., 1996. Developmental changes of connexin40 and connexin43 mRNA distribution patterns in the rat heart. Cardiovasc. Res. 32, 886–900. van Rijen, H.V., Eckardt, D., Degen, J., Theis, M., Ott, T., Willecke, K., Jongsma, H.J., Opthof, T., De Bakker, J.M., 2004. Slow conduction and enhanced anisotropy increase the propensity for ventricular tachyarrhythmias in adult mice with induced deletion of connexin43. Circulation 109, 1048–1055. van Veen, T.A., Stein, M., Royer, A., Le Quang, K., Charpentier, F., Colledge, W.H., Huang, C.L., Wilders, R., Grace, A.A., Escande, D., De Bakker, J.M., van Rijen, H.V., 2005. Impaired impulse propagation in Scn5a-knockout mice: combined contribution of excitability, connexin expression, and tissue architecture in relation to aging. Circulation 112, 1927–1935. Virágh, S., Challice, C.E., 1973. Origin and differentiation of cardiac muscle cells in the mouse. J. Ultrastruct. Res. 42, 1–24. von Both, I., Silvestri, C., Erdemir, T., Lickert, H., Walls, J.R., Henkelman, R.M., Rossant, J., Harvey, R.P., Attisano, L., Wrana, J.L., 2004. Foxh1 is essential for development of the anterior heart field. Dev. Cell. 7, 331–345. Waldo, K.L., Kumiski, D.H., Wallis, K.T., Stadt, H.A., Hutson, M.R., Platt, D.H., Kirby, M.L., 2001. Conotruncal myocardium arises from a secondary heart field. Development 128, 3179–3188. Wessels, A., Markman, M.W., Vermeulen, J.L., Anderson, R.H., Moorman, A.F., Lamers, W.H., 1996. The development of the atrio­ ventricular junction in the human heart. Circ. Res. 78, 110–117. Xavier-Neto, J., Neville, C.M., Shapiro, M.D., Houghton, L., Wang, G.F., Nikovits, W., Stockdale, F.E., Rosenthal, N., 1999. A retinoic acidinducible transgenic marker of sino-atrial development in the mouse heart. Development 126, 2677–2687. Yuan, S., Schoenwolf, G.C., 2000. Islet-1 marks the early heart rudiments and is asymmetrically expressed during early rotation of the foregut in the chick embryo. Anat. Rec. 260, 204–207. Yutzey, K.E., Rhee, J.T., Bader, D., 1994. Expression of the atrialspecific myosin heavy chain AMHC1 and the establishment of antero­ posterior polarity in the developing chicken heart. Development 120, 871–883. Zaffran, S., Frasch, M., 2002. Early signals in cardiac development. Circ. Res. 91, 457–469. Zaffran, S., Kelly, R.G., Meilhac, S.M., Buckingham, M.E., Brown, N.A., 2004. Right ventricular myocardium derives from the anterior heart field. Circ. Res. 95, 261–268.

Chapter 3.3

Retinoids and Heart Development Karen Niederreither1 and Pascal Dollé2 1

Departments of Medicine and Molecular and Cellular Biology, Center for Cardiovascular Development, Baylor College of Medicine, Houston, TX, USA  2 Institut de Génétique et de Biologie Moléculaire et Cellulaire (IGBMC), Illkirch, France; Inserm U 964, Illkirch, France; CNRS UMR 7104, Illkirch, France; Université de Strasbourg, Faculté de Médecine, Strasbourg, France

I.  Introduction It has long been recognized that proper vitamin A levels are essential for the development of the mammalian embryo. Warkany and colleagues have described a wide, pleiotropic spectrum of abnormalities in newborn or fetal rats generated from vitamin A-deficient (VAD) mothers; these include abnormalities of the heart and large vessels (Wilson and Warkany, 1950; Wilson et al., 1953). Since that time, the phenotype of VAD quail embryos has been reported (Heine et al., 1985; Dersch and Zile, 1993). In this avian model, VAD results in even earlier embryonic abnormalities, among which severe defects of the heart tube and of the vascular inflow tract are responsible for death after only four days of development. Vitamin A is a liposoluble vitamin which in most species is mainly ingested as betacarotene. Following the enzymatic reactions involved in absorbtion and storage (for a review see Harrison, 2005), vitamin A mainly circulates in the form of retinol (Fig. 1) coupled to a specific carrier protein RBP4. Retinol enters target cells through an interaction between RBP4 and its receptor Stra6 (Kawaguchi et al., 2007). Its intracellular transformation into active compounds involves the oxidation of its alcohol moiety to generate retinaldehyde, which in photoreceptor cells of the eye is bound to opsins and acts in the phototransduction process (for a review, Lamb and Pugh, 2004). An additional oxidative step generates retinoic acid (RA), which represents the active vitamin A derivative in most, if not all, other biological processes. A large body of work has demonstrated that retinoic acid acts as a genuine signaling molecule controlling many events during embryonic patterning, morphogenesis and organogenesis (for reviews see McCaffery and Drager, Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

2000; Clagett-Dame and DeLuca, 2002; Maden, 2002; Niederreither and Dolle, 2006). Retinoic acid thus needs to be produced in embryonic cells from maternal sources of vitamin A (retinol delivered transplacentally for mammals, or carotenoids stored in the yolk for oviparous species). The spatio-temporal control of its synthesis relies on the action of specific retinaldehyde dehydrogenases (RALDH1, 2 and 3) (Fig. 1). All three enzymes exhibit complex expression profiles throughout development; however, so far only RALDH2 has been implicated in retinoic acid functions relating to the heart (see below) and vascular development (Lai et al., 2003; Bohnsack et al., 2004). An additional mechanism exists to restrict the extent of retinoic acid signaling in embryonic tissues which involves its oxidative metabolism by a subfamily of P450 cytochromes, the CYP26A1, B1 and C1 enzymes (Fig. 1) (see Sirbu et al., 2005). As for the RALDHs, these enzymes are strongly-conserved both in terms of structure and expression patterns throughout vertebrate species. Although there are clues suggesting a role for CYP26A1 and/or B1 in regulating retinoic acid levels during vascular development (Emoto et al., 2005; Reijntjes et al., 2005; Ribes et al., 2007) until recently there were no data implicating CYP26 function in heart development proper. However, reduced expression of the three Cyp26 genes was found in Tbx1-null mice, and inhibition of CYP26 enzymatic function produced a phenocopy of DiGeorge syndrome in the chick (Guris et al., 2006; Roberts et al., 2006) (see Section III.B). Unlike peptidic signaling molecules that trigger a family of complex, membrane receptor-mediated intracellular cascades, retinoic acid directly acts as a ligand for a family of nuclear receptors, the retinoic acid receptors (RARs)

237

238

PART  |  3  Patterning of the Early Heart Tube

,  and , which belong to the class of nuclear receptor functioning as heterodimers with RXRs (Fig. 1). Retinoic acid can thus directly regulate the transcriptional activity of target genes by changing the properties of DNA-bound RAR/ RXR heterodimers from repressors to activators. Indeed, the conformational change induced by ligand binding leads to the release of co-repressor complexes from the RAR/ RXR dimer, and to the recruitment of coactivator proteins that lead to chromatin decompaction via covalent histone modifications (acetylation, methylation) and/or facilitate assembly of the transcription preinitiation complex (for further details, see Bastien and Rochette-Egly, 2004). Being a small, lipophilic molecule (Fig. 1), retinoic acid is easily diffusible by transfer through membranes, and there is evidence that this molecule can diffuse over several hundred micrometers in embryonic tissues (e.g., Thaller and Eichele, 1987). Thus, once produced in embryonic cells it may act either in an autocrine manner or as a paracrine signal

by local diffusion towards neighboring cells or tissue layers. As discussed hereafter (Sections II.C and III.B), retinoic acid could act by diffusion from mesodermal cells to pattern the foregut endoderm adjacent to the developing heart (for discussion about other paracrine functions of retinoic acid, see Maden, 2007). The range of retinoic acid-regulated genes is poorly-characterized, both in the context of heart development and more generally during embryogenesis. More than 500 genes, encoding highly diverse classes of proteins, are known to be retinoic acid-responsive, among which several dozen have been confirmed to be direct targets (for reviews, see McCaffery and Drager, 2000; Balmer and Blomhoff, 2005). The actual targets of retinoic acid-mediated regulation are thus expected to be diverse and highly dependent on a given cell’s history. In this chapter we review the experimental evidence that has implicated retinoic acid in several events of cardiac development, from early specification and/or patterning of

Retinol Blood

RBP

Cell

Retinaldehyde

Retinol ADH/RoDH

RA

RALDH

CRBP

CRABP

RoDH

CYP26

all-trans RA

Coo-

4-OH RA 4-oxo RA

9-cis RA?

RXR RAR Target genes RARE

Figure 1  Scheme of the intracellular pathways leading to retinoic acid-mediated gene regulation. In placental species, retinol is transfered from maternal to embryonic/fetal blood and taken up by retinol binding protein (RBP). Two enzymatic reactions catalyzed, respectively, by specific alcohol/ retinol dehydrogenases (ADH/RoDH) and retinaldehyde dehydrogenases (RALDH) transform retinol into retinoic acid (RA). Specific binding proteins for retinol (the cellular retinol-binding proteins CRBP I–III) and retinoic acid (the cellular retinoic acid-binding proteins CRABP I and II) are present in the embryo, although their exact role in facilitating the enzymatic conversion of retinol, and/or the nuclear transfer of retinoic acid, is still debated. Retinoic acid eventually binds to RARs, thereby mediating a conformational change of RAR/RXR heterodimers bound to DNA response elements (retinoic acid-response element, RAREs) from a repressing to an activating state with respect to transcriptional activity of target genes. All-transretinoic acid is the main form of retinoic acid detected in mammalian embryos (e.g., Ulven et al., 2000), whereas some of its derivatives (especially 3,4didehydro-retinoic acid) are quite abundant in chick and Xenopus embryos. The 9-cis retinoic acid stereoisomer is a selective ligand for RXRs (Heyman et al., 1992; Levin et al., 1992). However, this isomer is undetectable or only detected at trace levels regardless of the species, so it is likely that RXRs are unliganded in the embryo. According to the presence of specific P450 enzymes (CYP26A1, B1 and/or C1), retinoic acid can also be transformed into more polar metabolites, such as 4-hydroxy and 4-oxo-retinoic acid. Although 4-oxo-retinoic acid can bind RARs (Pijnappel et al., 1993), in vivo evidence suggests that CYP26 products are merely catabolic intermediates (Niederreither et al., 2002a; Uehara et al., 2006; Ribes et al., 2007, and references therein).

239

Chapter  |  3.3  Retinoids and Heart Development

II.  Early heart morphogenesis and patterning

the cardiac field (Section II) to the control of myocardial differentiation (Section IV). A substantial amount of this data has been obtained through phenotypic studies performed on the VAD quail model, as well as in targeted mouse mutants for the various RARs and RXRs and the synthesizing enzyme RALDH2. In parallel, a battery of experimental approaches have been used in each of the vertebrate models for studying development, from zebrafish (which, interestingly, provided evidence that retinoic acid has been functioning in vertebrates prior to the appearance of the four-chambered heart) to mouse. Quite often, these studies have compared the effects of exposure to exogenous retinoic acid, either to the whole embryo or by local administration, to those of a pharmacological inhibition of endogenous signaling using synthetic RAR/RXR antagonists or inhibitors of retinoic acid synthesis. The major outcomes – and sometimes the discrepancies – obtained in various animal models and/or distinct experimental approaches are discussed hereafter.

(A)

(B)

II.A.  Dynamic Patterns of Retinoic Acid Synthesis during Early Heart Development The earliest roles of retinoic acid with respect to heart development rely on its region-specific patterns of synthesis by retinaldehyde dehydrogenase 2 (RALDH2). Hochgreb et al. (2003) have analyzed the dynamics of Raldh2 expression in most detail with respect to heart field specification (Fig. 2) (see also Berggren et al., 1999) for RALDH2 protein immunolocalization in chick). In the gastrulating chick embryo, Raldh2 is first expressed by mesodermal cells posterior to Hensen’s node (Fig. 2A). At Hamburger-Hamilton (HH) stages 6–7 (i.e., shortly before and during formation of the first somite pairs), two arches of expressing cells appear in the anterior lateral mesoderm (Hochgreb et al., 2003). These arches intensify, progress anteriorly and eventually merge

(C)

(D)

Precardiac stage/ Cardiac cell determination

Heart tube fusion

Chamber morphogenesis/ Looping heart tube

Chamber septation/ Myocardial differentiation

Late gastrula/ Headfold stages Chick HH4-6 - Mouse E7-7.5

Early somite stages Chick HH6-8 - Mouse E8-8.5

Late somite stages Chick HH 10-13 - Mouse E8.5-9.5

Chick HH18+ - Mouse E10.5+

Pre-conus

Outflow tract

Pre-conus Pre-ventricle

Pre-ventricle Pre-atrium

Pre-atrium

Atrium Raldh2

Raldh2

Restriction of cardiac progenitor pool (zebrafish) LR laterality AP patterning information

Ventricle

Raldh2 Sinus venous

LR laterality AP patterning information Regulation of posterior segments growth and development Regulation of molecules involved in morphogenetic movements

Regulation of posterior segments growth and development Regulation of molecules involved in morphogenetic movements

Septation of the outflow tract (intra/extra-cardiac roles?) Regulation of myocardial growth and differentiation

Figure 2  Sequential contributions of retinoic acid to heart patterning and differentiation. Schematically, four developmental steps are considered, from late gastrulation/precardiac stages to late events in myocardial differentiation and outflow tract septation. The corresponding stages in chick (Hamburger-Hamilton stages, HH) and mouse (embryonic day, E) are indicated, and the postulated functions of retinoic acid are summarized below (see main text for details). Drawings show the restricted distribution of retinoic acid-producing cells (Raldh2-expressing) in precardiac mesodermal populations and in the early heart tube during looping morphogenesis. Data at early stages have mainly been obtained in chick embryos (Hochgreb et al., 2003); however, analysis of mouse embryos harboring a retinoic acid-responsive reporter transgene (e.g., Moss et al., 1998) show a close correlation between retinoic acid-producing and retinoic acid-responding cells prior to and during heart tube formation. At later stages, Raldh2 is selectively expressed in pericardial and epicardial layers of the heart, although retinoic acid activity as detected with the retinoic acid-reporter transgene also occurs within the developing myocardium. Retinoic acid generated outside of the heart proper (e.g., in the posterior branchial region (see Niederreither et al., 2003) may also regulate events in cardiac development, such as septation of the outflow tract (see Section III.C). Photographs from Moss et al. (1998), with permission.

240

PART  |  3  Patterning of the Early Heart Tube

at the midline below the anterior intestinal portal (Fig. 2B). This dynamic phase of expression at HH stages 7–8 coincides with the critical period for anterior–posterior specification of cardiac precursor cells. Double labeling with Gata4 and DiI injection experiments confirmed the anterior expansion of Raldh2-expressing cells, which eventually encompass about half of the heart field at the level of prospective poster­ ior (sinus venosa and atrial) cells. Similar expression features were observed in the mouse embryo at corresponding stages of development (Moss et al., 1998; Hochgreb et al., 2003). Furthermore, two retinoic acid-inducible transgenes, the RARE-hsp68-lacZ reporter (Moss et al., 1998) and a slow myosin heavy chain 3 (sMyHC3-HAP) transgene (XavierNeto et al., 1999) display strikingly restricted expression in posterior segments of the early heart tube. More complex patterns of Raldh2 expression and RARE-reporter transgene activity are seen shortly thereafter, from embryonic day (E) 9.5 onwards (Fig. 2C,D). These findings provide strong support for the idea that endogenous retinoic acid may act as a regionally-restricted signal imparting a poster­ ior fate to cardiac precursor cells.

including the forelimb buds (Niederreither et al., 2002b), dorsal pancreas (Martin et al., 2005; Molotkov et al., 2005) or foregut derivatives (Wang et al., 2006) could in part result from an abnormal allocation of precursor cells during gastrulation. In zebrafish, retinoic acid signaling was shown to act on a retinoic acid-responsive target gene hoxb5b expressed within the forelimb field, in order to restrict the number of atrial cells arising from the adjacent heart field (Waxman et al., 2008). These data indicate a nonautonomous influence of retinoic acid signaling to limit heart chamber size. Further exploration of possible links between limb and cardiac development in mouse or avian embryos will be necessary to conclude whether retinoic acid acts in a similar manner to restrict the cardiac progenitor cell pool in other species. Evidence that retinoid signaling directly regulates the mouse Hoxb5, Hoxb6 and Hoxb8 genes at the promoter level (Oosterveen et al., 2003) suggests that this function may have been maintained during evolution.

II.B.  Retinoic Acid Signaling Restricts the Zebrafish Cardiac Progenitor Cell Pool

One of the functions of retinoic acid produced by the posterior embryonic mesoderm would be to act as a regional determinant imparting posterior cell fates (hindbrain and spinal cord) to the newly-formed neural plate (for reviews see Maden, 2002, 2007). A possible similar function in instructing posterior cell fates within the cardiogenic mesoderm has been inferred by studies which have analyzed the effects of exposing embryos of various species to exogenous retinoic acid. In zebrafish, retinoic acid treatment at the blastula stage results in heart truncation progressively affecting anterior (arterial) towards posterior (venous) developing chambers, in a dose-dependent manner (Stainier and Fishman, 1992). Exposure of chick embryos to retinoic acid from HH stages 3 to 7 also led to abnormal hearts with truncated anterior segments, contrasting with a large, swollen atrium and sinus venosus region (Osmond et al., 1991) (Fig. 3C,D). Retinoic acid treatments at the earliest stages and/or at high doses actually had more drastic effects, with a failure of the bilateral heart primordia to fuse (cardia bifida) or even a complete absence of heart. These defects have been postulated to reflect an inhibition of the normal migration of precardiac mesodermal cells along the fibronectin-containing extracellular matrix (Osmond et al., 1991). Yutzey et al. (1994) provided the first evidence that retinoic acid may induce changes in anterior–posterior (AP) cell fates, by analyzing expression of the chick atrial-specific myosin heavy chain (AMHC1) gene. Expression of this gene is regionally restricted in the early heart tube, where it normally occupies about 25% of the posterior-most region. Exposure of embryos to retinoic acid led to an anterior expansion of the AMHC1 expression domain within the developing heart. Relatively “mild”

Evidence for an early role of retinoic acid in restricting the heart lineage was obtained in zebrafish embryos (Keegan et al., 2005, Chapter 1.4). Mutant embryos with a loss of function of RALDH2 (the neckless mutants) were found to have increased numbers of cardiomyocytes expressing Nkx2.5 or cardiac myosin light chain 2 (cmlc2). Treatment of gastrulating wild-type zebrafish embryos with a panRAR antagonist leads to a similar expansion of Nkx2.5 and cmlc2-expressing cells, whereas exposure to exogenous retinoic acid results in a reduced number of cardiomyocytes (Fig. 3A,B). Fate-map experiments in RAR antagonisttreated embryos suggest that higher numbers of lateral marginal zone blastomeres have become myocardial progenitors, perhaps at the expanse of other (pharyngeal pouches, pectoral fin and/or pancreatic) prospective cell fates (Keegan et al., 2005). Thus, one of the earliest functions of retinoic acid during gastrulation would be to restrict specification of the cardiac progenitor pool and thus define the limit of the cardiac field. Alternatively, the observed phenotypes could reflect a role of retinoic acid in regulating the density of myocardial progenitor cells. A slightly enlarged Nkx2.5-expressing domain has also been observed in Xenopus embryos treated with a pan-RAR antagonist (Collop et al., 2006); however, so far no evidence for a comparable role has been obtained in amniote embryos, including through phenotypic analysis of mouse Raldh2null mutants (Niederreither et al., 1999, 2001; Mic et al., 2002). It should be mentioned, however, that defective development or lack of several structures in these mutants,

II.C.  Retinoic Acid and Anterior–Posterior Patterning of the Heart Tube

241

Chapter  |  3.3  Retinoids and Heart Development

(A)

(E)

(G)

(I)

(K)

(F)

(H)

(J)

(L)

(C)

(B)

(D)

Figure 3  Excess or deficiency in retinoic acid signaling affects early events in heart development in various species. In zebrafish, lack of retinoic acid signaling following treatment with a synthetic inhibitor of RARs prior to and during gastrulation leads to an increased number of NKx2.5expressing cells at early somite-stages (A, B). Treatment of chick embryos with all-trans-retinoic acid at HH3–7 stages leads to a dysmorphic heart tube with enlarged atria and sinus venosus, and a reduced ventricle, bulbus cordis and outflow tract region (C, D). Mouse embryos with a knockout of the Raldh2 gene develop a highly-abnormal heart with a dilated outflow tract and ventricle region, the latter with a poor trabeculation, as well as markedly hypoplastic atrial and sinus venosus cavities (E, F: sagittal sections, hematoxylin-eosin staining). Analysis of Tbx5 expression in Raldh2/ embryos confirmed the poor development and improper specification of putative atrial/sinus venosus cells (G, H, arrow). Scanning electron microscopy (K, L, arrows) and additional molecular analyses (not shown) clearly showed that the heart tube of Raldh2/ mutants had failed to undergo left–right looping morphogenesis. Wild-type mouse embryos cultured exo utero at headfold stages (E7.5–E8) in the presence of all-trans-RA exhibit an abnormal, bilateral expression of left-side specific determinants such at Lefty1 (light blue) and Nodal (purple), eventually leading to a randomized sidedness of heart looping (I, J) (aip: anterior intestinal portal; at: atrium; fp: floor plate; lpm: lateral plate mesoderm; lv: left ventricle; n: node; ot: outflow tract; rv: right ventricle; sv: sinus venosus; v: ventricle). (A, B) from Keegan et al. (2005); (C, D) from Osmond et al. (1991); (E–H) and (K, L) from Niederreither et al. (2001); (I, J) from Chazaud et al. (1999), with permission.

treatments that do not alter heart morphogenesis expanded the AMHC1 domain in over 50% of the heart tube and under more severe conditions led to cardia bifida; AMHC1positive cells occupied almost 100% of the heart. Additional studies investigating the effects of exogenous retinoic acid have been performed in other species. In Xenopus, excess retinoic acid consistently led to cardia bifida, abnormal looping and/or altered expression of myocardial differentiation markers (Drysdale et al., 1997; Collop et al., 2006). Treatment of neurula-stage Xeno­ pus embryos with retinoic acid also reduced the heartforming region expressing Nkx2.5 and expanded Gata4/5/6 domains into noncardiogenic lateral plate mesoderm (Jiang et al., 1999). Most studies performed in the mouse have focused on retinoic acid-induced effects on heart looping and expression of left–right patterning genes (see Section II.E). Exposure of gastrulating embryos to retinoic acid in culture (Chazaud et al., 1999) or through maternal administration (Xavier-Neto et al., 1999) resulted in hearts with signs of posteriorization, i.e., with enlarged atrium and reduced outflow tract and ventricle. There has been no detailed molecular study to confirm that retinoic acid

may alter spatial domains of gene expression in the heart tube. Interestingly, however, retinoic acid treatment led to a rostral expansion of the activity of a reporter transgene (the sMyHC3-HAP), which is normally restricted to the sinoatrial region of the heart (Xavier-Neto et al., 1999). One of the strategies used to investigate the roles of embryonic retinoic acid is to inhibit its signaling activity by exposing embryos to synthetic ligands acting as retinoic acid antagonists. Once bound, these molecules block the receptor(s) in a conformation which favors the binding of co-repressor complexes, leading to the stable formation of repressing RAR–RXR complexes. Such molecules have been used in several species, either by incubation of whole embryos or by local administration, and have yielded interesting data on the involvement of retinoic acid in processes such as craniofacial (Schneider et al., 2001), branchial arch (Wendling et al., 2000) or limb bud (Helms et al., 1996) development. Using this strategy, Xavier-Neto and colleagues obtained the clearest evidence for an involvement of retinoic acid in AP cell determination within the cardiac field (Chapter 1.1). Treatment of chick embryos with a pan-RAR antagonist from HH stages 4 to 7 led to hearts with an oversized

242

ventricle and reduced inflow compartment (Hochgreb et al., 2003). The results of DiI injection experiments were consistent with a change in the cardiac fate map, with ventricular precursor cells abnormally located in the posterior cardiac field of the antagonist-treated embryos. These data are consistent with retinoic acid normally acting as a signal determining posterior cardiac fates, and forced inhibition of its signaling converting sinoatrial precursors to a ventricular fate. Treatment of mouse embryos at corresponding stages led to consistent defects, with an oversized ventricle cavity and reduced atrium, although these morphological changes have not been assessed molecularly (Chazaud et al., 1999). Similar abnormalities have been reported in mouse embryos exposed in utero to disulphiram, an inhibitor of aldehyde dehydrogenases that interferes with the activity of retinoic acid-synthesizing enzymes (Xavier-Neto et al., 1999). The definitive proof for retinoic acid being instrumental in heart chamber specification comes from the study of animal models deficient for its endogenous synthesis. Dietary vitamin-A-deficient (VAD) quail embryos die at day 4 of development with severe cardiovascular abnormalities that include an enlarged heart closed at its posterior extremity due to lack of formation of an inflow tract (Heine et al., 1985; Dersch and Zile, 1993). Additional abnormalities include a lack of chamber septation and defective looping morphogenesis (see Section II.D). In agreement with these data, mice with a targeted disruption of the Raldh2 gene develop abnormal hearts with highly hypoplastic sinus venosa and atrial regions (Niederreither et al., 2001) (Fig. 3E,F). Molecular analysis of these mutants confirmed the defective specification of posterior heart structures as the Tbx5 gene, which in wild-type embryos is expressed at high levels in prospective atrial and sinus venosa cells, but here was expressed in only a few scattered cells in the posterior region of the Raldh2/ hearts (Fig. 3G,H). In contrast, most of the dilated Raldh2/ heart tube consists of regions with outflow tract, right and left ventricle molecular identities (Niederreither et al., 2001). Whether the abnormal heart phenotypes in avian or mouse retinoic acid-deficient embryos result from an abnormal specification of A–P cell fates, rather than a regional lack of expansion of posterior cardiac progenitors has not been explored by performing vital dye lineage tracing experiments. Interestingly, however, both animal models can be strikingly rescued by supplying exogenous retinoic acid during a critical developmental window. The 4–5 somite stage has been defined as the last stage at which exogenous retinoic acid can rescue heart morphogenesis and abnormal gene expression in VAD quail embryos (Kostetskii et al., 1998). Consistent with these data, a 24-hour maternal retinoic acid supplementation between E7.5 and E8.5, i.e., during late gastrulation and early somitogenesis, can rescue heart looping morphogenesis and posterior chamber development in Raldh2/ mouse mutants (Niederreither

PART  |  3  Patterning of the Early Heart Tube

et al., 2001). Gata4 has been proposed as a key molecular determinant of the abnormal inflow tract phenotype of VAD quail embryos. Indeed, its expression is severely reduced both in the heart-forming region and endoderm, and retinoic acid treatment conditions that rescue the heart phenotype also restore normal Gata4 expression in both tissues (Kostetskii et al., 1999). Gata4 expression was also reduced in the mouse Raldh2/ mutants (Niederreither et al., 2001). Other possible downstream genes are members of the Hox family, some of which are expressed in the early cardiac field (Searcy and Yutzey, 1998) and are bona fide retinoic acid-target genes (e.g., Hoxa1) (Dupe et al., 1997). Further work will be required to identify critical targets of retinoic acid regulation during A–P heart tube specification, and to identify how the retinoic acid signal intersects with other pathways involved in A–P regionalization. An important question regarding the origin of the abnormal heart phenotype in retinoic acid-deficient models is that of the target tissue(s) where retinoic acid normally acts. Being a diffusible, lipophilic molecule, retinoic acid could act in a paracrine manner outside the cells that produce it, e.g., across the cardiogenic mesoderm (maybe according to a diffusion gradient), and/or in another tissue layer such as the foregut endoderm, which directly contacts the cardiogenic mesoderm and is known to be required for the initiation and maintenance of cardiac gene expression (e.g., Lough et al., 1996). As described above, expression of the retinoic acid-synthesizing enzyme Raldh2 is specific to mesodermal cells; however, activity of a retinoic acid-responsive lacZ reporter transgene is observed in both the mesodermal and endodermal layers of the mouse embryonic foregut (Niederreither et al., 2003). An interesting observation has been made using the VAD quail model. Ghatpande et al. (2000) demonstrated that transplantation of anterior endoderm from normal chick embryos is sufficient to rescue normal heart tube morphogenesis in VAD embryos. Although both posterior and anterior (foregut) endoderm contain detectable levels of retinoic acid (Maden et al., 1998), only anterior endoderm can elicit such a rescue, suggesting a requirement for additional, anterior-restricted factor(s). BMP2 is likely to be one such factor, as it is normally secreted by the anterior endoderm, its expression is abnormally weak in VAD embryos, and experimentally-induced reexpression of Bmp2 can rescue the VAD heart phenotype (Ghatpande et al., 2006). These data point to a model in which retinoic acid, normally produced by RALDH2 in the posterior cardiogenic mesoderm (see Section II.A), would not only act in an autocrine or paracrine manner in myocardial precursor cells, but would also act as an inductive signal towards the adjacent endoderm, in order to regulate gene expression and production of additional signals by this tissue layer. It is noteworthy that several alterations of gene expression have been described in the foregut endoderm

243

Chapter  |  3.3  Retinoids and Heart Development

of Raldh2/ mouse mutants (Niederreither et al., 2003; Wang et al., 2006) (see Section III.B). An additional pathway which regulates posteriorrestricted production of retinoic acid is through microRNA-induced gene repression. MicroRNAs (miRs) are an evolutionarily conserved class of small regulatory molecules that have recently been shown to regulate cardiac development and pathology. Primary miRNAs are transcribed and processed so that in their final cytoplasmic form they target specific genes for RNA degradation and/or translational repression (reviewed in Callis and Wang, 2008). In zebrafish one such inhibitor, miR-138, is expressed in the developing heart (Morton et al., 2008; Chapter 10.3). One of the normal functions of miR-138 is to bind to the 3 untranslated region of Raldh2 mRNA, restricting Raldh2 and retinoic acid signaling to the atrioventricular canal. At a distinct temporal window (24–34 hours postfertilization) miR-138 knockdown produces ectopic Raldh2 expression in ventricular cardiomyocytes, disrupting heart morphogenesis and function (Morton et al., 2008).

II.D.  Retinoic Acid Deficiency Alters Murine Second Heart Field Formation Recent work performed on mouse Raldh2/ mutants has implicated the second heart field (SHF) as the main cardiac progenitor population affected by retinoic acid deficiency. This population of undifferentiated cardiac precursor cells originates from prospective pharyngeal mesoderm, which lies medial to the cardiac crescent or first heart field (FHF). While the cardiac crescent is undergoing differentiation, second heart field cells continue proliferating along the coelomic wall ventral to the foregut, and are progressively added to the poles of the heart tube. Thus they give rise to most, if not all, the myocardium of the outflow tract, and part of the right ventricle and atria (for a review see Buckingham et al., 2005) (see Chapters 2.2 and 3.1). By analyzing endogenous markers of the second heart field, and by using several lacZ reporter transgenic lines to allow tracing of second heart field or first heart field derivatives, Ryckebusch et al. (2008) characterized a number of second heart field defects in Raldh2/ embryos (see also Sirbu et al., 2008). In mutants, the anterior part of the second heart field is disorganized and extends to abnormal posterior levels beyond the posterior pole of the forming heart tube. Additionally, the left and right second heart field populations do not merge properly along the dorsal mesocardium. As a result the deployment of second heart field cells within the heart tube is highly compromised, especially in the outflow tract region. Analysis of second heart field explants from Raldh2/ mutants revealed that these cells are impaired in their ability to differentiate and form beating cardiomyocytes. This may result from improper growth factor signaling, with FGFs normally producing the

second heart field (FGF8 and 10) being abnormally located and failing to signal within the heart tube. Downregulation of several Bmp genes has also been observed in the mutant embryos (Ryckebusch et al., 2008; Sirbu et al., 2008). Thus, in the mouse, the second heart field is the cardiac precursor population most sensitive to retinoic acid deficiency. Interestingly, the cellular and molecular defects characterized by Ryckebusch et al. (2008) arise at the critical stage (4–5 somites) previously defined in an avian model as being retinoic acid-dependent (Kostetskii et al., 1998). These results are not incompatible with retinoic acid also having effects on the first heart field; however, it should be noted that lacZ reporter transgene analysis revealed an essentially intact contribution of first heart field cells to the heart tube of Raldh2/ mutants (Ryckebusch et al., 2008). To date, the embryonic and evolutionary origins of the second heart field remain unclear (Buckingham et al., 2005). Possibly this lineage appeared during evolution of terrestrial vertebrates in parallel with the evolution of a four-chambered heart. While retinoic acid clearly has a role during heart development of both marine and terrestrial vertebrates (Keegan et al., 2005), it may have been functionally coopted to act on the second heart field, especially because its spatially-restricted distribution may have allowed it to control the extent and posterior restriction of this cell population with respect to the heart field.

II.E.  Retinoic Acid and Left–Right Heart Looping Morphogenesis Another important phenomenon in which retinoids have been implicated is the laterality pathway. Both conditions of retinoic acid excess or defective signaling can affect the sidedness and/or the realization of left–right (L–R) heart looping. Exogenous retinoic acid treatment can induce laterality defects in various rodent species (Fujinaga, 1997). Several studies have investigated how exposure of embryos to exogenous retinoic acid, or to synthetic RAR antagonists, would affect the expression of molecular determinants of the left–right asymmetry pathway (see Chapter 4.1 for details of this pathway). Exposure of mouse embryos to excess RA, either in utero through maternal administration (Wasiak and Lohnes, 1999), or in whole embryo culture (Chazaud et al., 1999; Tsukui et al., 1999) leads to an ectopic, bilateral expression of several genes normally induced only in the left lateral plate mesoderm (LPM), including Lefty1 and 2, Nodal and Pitx2 (see Fig. 3I,J). Conversely, exposure of the embryos to a pan-RAR antagonist can lead to a complete downregulation of expression of these genes in the left LPM, as well as in the midline floor plate cells for Lefty1 (Chazaud et al., 1999; Tsukui et al., 1999). Both types of treatments result in a statistical randomization of the left–right sidedness of heart looping. These effects are stage-specific, and retinoids need to

244

be delivered at the head fold/presomite stage to elicit such gene responses. Tsukui et al. (1999) have extended these findings to Xenopus and zebrafish embryos. These authors also showed that a local administration of these molecules alongside Hensen’s node in chick embryos suffices to induce the aforementioned effects on Lefty1-2, Nodal and Pitx2 expression. That retinoic acid can affect cardiac laterality when locally administered next to Hensen’s node was first demonstrated by Smith and colleagues (Dickman and Smith, 1996; Smith et al., 1997). These authors identified two proteins, the heart-specific lectin-associated matrix protein 1 (hLAMP1) and the fibrillin-associated protein recognized by the JB3 monoclonal antibody, as possible mediators of the retinoic acid-induced effects on cardiac sidedness. With respect to the regulatory cascade leading to side-specific gene expression (see Chapter 4.1), retinoids seem to act downstream, or in parallel with Sonic Hedgehog (Shh), as no alteration of Shh expression was detected in retinoid-treated embryos (Chazaud et al., 1999; Tsukui et al., 1999). Lefty1 was thought to be a key player in the laterality defects induced by retinoic acid or RAR antagonists, as this gene is a direct retinoic acidtarget gene (Meno et al., 1996; Oulad-Abdelghani et al., 1998). The presence of Lefty1 in midline floor plate cells is necessary for a “midline barrier” function preventing expression of left-side determinants, including its homolog Lefty2, in the right-side LPM (Meno et al., 1998). A simple model involving the midline barrier function of Lefty1 cannot explain the outcome of the various retinoid treatments however, as pan-RAR antagonist treatments that prevent Lefty1 midline expression do not lead to a bilateral expression of Lefty2 and Nodal. It is likely that the retinoid treatments lead to complex feedback effects that (directly or indirectly) affect Nodal, Lefty2 and other uncharacterized gene products. How do the striking effects of pharmacological retinoid treatments correlate with findings on animal models deficient for endogenous retinoic acid signaling? A high incidence of reversed cardiac looping has been recognized as one of the features of vitamin-A-deficient (VAD) quail embryos (Heine et al., 1985; Dersch and Zile, 1993). However, the VAD hearts never develop as true mirror images of normal looping hearts, i.e., they do not represent genuine cases of situs inversus; rather, their abnormal morphology indicates a failure to properly realize the looping process (Zile et al., 2000). Molecularly, no alteration of Lefty1 expression (or of other early left–right determinants including Fgf8, Activin receptor IIa and Shh) has been observed (Zile et al., 2000). Nodal and Pitx2 expression was affected in the VAD quail embryos but, unlike in the case of RAR antagonist treatments that can completely downregulate their expression (Chazaud et al., 1999; Tsukui et al., 1999), here the downregulation was limited to the posterior heart-forming mesoderm (Zile et al., 2000). In no case was Nodal or Pitx2 expression ectopic

PART  |  3  Patterning of the Early Heart Tube

or randomized. The sidedness of heart looping cannot be assessed in Raldh2/-null mouse mutants, as these consistently display a dilated heart tube bulging ventrally along the embryonic A–P axis, likely reflecting an inability to undergo left–right looping (Niederreither et al., 1999, 2001) (Fig. 3K,L). Analysis of Lefty1-2, Nodal and Pitx2 expression was performed in these mutants, and no alteration of their expression patterns was reported (Niederreither et al., 2001). Possibly some subtle or stage-specific abnormalities such as those described in VAD quail embryos may have been overlooked. It is also possible that abnormalities cannot be observed in these mutants due to their lethality by E9.5. Another interesting mouse model was reported by Iulianella and Lohnes (2002), who generated embryonic chimeras with embryonic stem (ES) cells expressing a dominant negative version of RAR. These chimeras exhibited complex heart defects, including looping anomalies. At E10.5, Pitx2 expression was reduced in chimeras with abnormal heart looping (Iulianella and Lohnes, 2002). Altogether, these animal models indicate that a lack of endogenous retinoid signaling does not interfere with regulation of early left–right determinants (in particular Lefty1), and that expression of left-side specific genes (including Nodal and Pitx2) is normally induced. Accordingly, the retinoic acid-deficient embryos do not exhibit genuine situs inversus, with mirror-image reversal of the heart and other viscera. The heart looping defects observed in both mouse and quail are indicative of a failure to realize normal looping morphogenesis. Region-specific downregulation of Nodal and Pitx2 in the left heart-forming region may participate in this defect, although it is most likely that additional effectors implicated in the “unlooped” heart phenotype remain to be characterized. An important conclusion is that the consequences of an endogenous retinoic acid deficiency are not the same as those of a pharmacological inhibition of RARs, which leads to more drastic downregulation of genes involved in the left–right laterality pathway. Reasons for these discrepancies have been prev­ iously discussed (Niederreither et al., 2001). It is noteworthy that the heart looping defect can be rescued in the VAD quail embryos by providing retinoic acid specifically at the 4–5 somite-stages (Zile et al., 2000), i.e., later than the time window during which excess retinoic acid, or RAR antagonist treatments, affect left–right gene expression and heart laterality. This indicates that a crucial function of endogenous retinoids occurs at early somite-stages in order to maintain adequate levels of Nodal and Pitx2 in lateral mesoderm, and/or regulate other effectors of heart looping. There has been some debate about whether embryonic retinoic acid provides a symmetric, rather than a left–right asymmetric, signal along the embryonic mesoderm. Until recently, the consensual view was that retinoic acid acts in a symmetrical manner along the left and right presomitic and somitic mesoderm. Detailed expression studies have been performed in avian or mouse embryos, which

245

Chapter  |  3.3  Retinoids and Heart Development

did not provide any evidence for a left–right bias in the expression of retinoic acid receptors (e.g., Cui et al., 2003) and retinoic acid-producing enzymes (e.g., Blentic et al., 2003), or in the patterns of retinoic acid-response using the RARE-hsp68-lacZ reporter transgene (Sirbu and Duester, 2006). An alternative view to a model involving bilaterally symmetric retinoic acid signaling has emerged with the discovery in the mouse embryo of nodal vesicular parcels (NVPs), membrane-sheathed vesicles released by ventral cells located on the right side of the node and transported by “nodal flow” to cells on the left side of the node (Tanaka et al., 2005; for details, see Chapter 4.2). These vesicles were found to contain Sonic hedgehog protein, as well as retinoic acid (Tanaka et al., 2005). Transport of retinoic acid by the NVPs would provide a mechanism for its enrichment in left-side perinodal cells, without a need for asymmetric expression of synthesizing or metabolizing enzymes, or of retinoic acid receptors. The discovery of NVPs is clearly an important breakthrough, as it has led to the description of a mechanism for translating left–right molecular asymmetries through the effect of nodal flow. It can in particular explain how Sonic hedgehog may act as a left–right determinant in the mouse embryo, without being asymmetrically expressed in this species (Tsukui et al., 1999). Nonetheless, the hypothesis of a left–right transport of retinoic acid might be considered with caution, as presently it has been documented only by the use of retinoic acid antiserum that has otherwise seldom been used. A definitive demonstration of a possible transport and leftside enrichment of retinoic acid will require additional biological evidence, for instance by demonstrating a left– right bias in the initial activation of reporter transgene(s) in perinodal cells, a higher ability of left-side perinodal cells to activate retinoic acid-sensitive cell lines (Wagner et al., 1992), or possibly the demonstration of unequal retinoic acid levels by sensitive HPLC methods.

III.  Later events in outflow tract septation and patterning of the posterior pharyngeal arches Abnormalities of the outflow tract and the large vessels derived from the aortic arches are a hallmark of the vitamin-A-deficiency syndrome in rat (Wilson and Warkany, 1950; Wilson et al., 1953) and of RAR/RXR gene knockouts in mice. Lack of formation of the aortico–pulmonary (Ao–P) septum, leading to a persistent truncus arteriosus (PTA), was consistently observed in mice with combined mutations of Rar and Rar, or Rar and Rar (including in the case of isoform-specific compound mutants) (Mendelsohn et al., 1994; Lee et al., 1997; Ghyselinck et al., 1998). The RAR compound mutants also displayed variable aberrant patterns of the aorta, pulmonary artery and other aortic arch-derived arteries (Fig. 4A–C). None of the

corresponding single RAR mutants exhibited such abnormalities; however, single Rxr/ mutants can exhibit, in an incompletely penetrant manner, PTA or partial formation of the Ao–P septum, with associated conotruncal abnormalities (Gruber et al., 1996). The penetrance of these abnormalities is increased in compound mutant genotypes for any of the three RARs (Kastner et al., 1997a). RAR function seems to be especially critical for proper formation of the Ao–P septum, as the presence of a complete PTA was observed in all Rxr;Rar double-null mutants, whereas this abnormality was incompletely penetrant in Rxr ;Rar or Rxr;Rar mutants. On the other hand, the presence of RAR isoforms has been found to be particularly important for development of the conotruncal ridges (Ghyselinck et al., 1998). Abnormal patterning of the aortic arch-derived great vessels can also be observed in various RXR/RAR compound mutants (Kastner et al., 1997a). These genetic studies have shown that RAR and RAR function as heterodimers with RXR during events leading to septation of the outflow tract. Interestingly, while the first study of RAR compound mutants has suggested essentially redundant functions of the RARs (Mendelsohn et al., 1994), it was then found that in an Rxr/ background, there is little functional redundancy between RAR and RAR with respect to the mechanisms of aorticopulmonary and conotruncal septation, respectively (Kastner et al., 1997a; Ghyselinck et al., 1998).

III.A.  Investigations of Defective Outflow Tract Development Septation of the outflow tract is a complex process, involving interactions between several cell types of various embryological origins. Most, if not all, cardiomyocytes of the outflow tract are derived from the second heart field, a major target tissue of retinoic acid during early heart formation (Ryckebusch et al., 2008) (see Section II.D), therefore it is likely that the outflow tract defects in Rar-Rxr mutants result, at least in part, from abnormal gene regulation in this cell population. Aorticopulmonary septation also involves migration of post-otic hindbrain (“cardiac”) neural crest cells (NCCs), which contribute to the wall of the aortic arch-derived arteries, whereas conotruncal septation involves the formation of cushions through epithelial–mesenchymal transformation of the endocardium, which are eventually also colonized by neural crest cells (see chapters from Parts 7 and 8 for detailed descriptions of these processes). Few studies have attempted to unveil the cellular and molecular mechanism(s) leading to abnormal outflow tract septation in retinoid receptor mouse mutants. Kubalak et al. (2002) reported abnormally high apoptosis in outflow tract cushion tissue as an early event occurring in Rxr/ embryos. The cell type(s) undergoing apoptosis have not been determined, although their location would

246

PART  |  3  Patterning of the Early Heart Tube

(D)

(A)

(F)

(B)

(G)

(H)

(C)

(I)

(E)

(J)

(K)

(L)

(M)

Figure 4  Lack or excess of retinoic acid signaling affects septation of the outflow tract and patterning of the large vessels, due to defects of the posterior embryonic branchial arches. (A–C) Schemes of the normal anatomy of the aorta, pulmonary trunk and main cephalic and brachial arteries in wild-type mice (A), and of abnormal patterns of the outflow tract and large vessels observed in compound mutant fetuses for Rar/Rar, or Rar/ Rar (B, C). The latter pattern, with persistent truncus arteriosus (PTA) and dextroposition of the arch of the aorta, also occurs in fetuses with a hypomorphic mutation of Raldh2 (Raldh2neo/) (AA: aortic arch; AscA: ascending aorta; CA: carotid arteries; Duc: ductus arteriosus; LAA: left aortic arch; LPA: left pulmonary artery; LSA: left subclavian artery; PT: pulmonary trunk; PTA: persistent truncus arteriosus; RAA: right aortic arch; RPA: right pulmonary artery; RSA: right subclavian artery). (D, E) Histological cross-sections (hematoxylin-eosin staining) of the aorta and pulmonary trunk of a wild-type E14.5 fetus (D) and a Raldh2neo/ littermate with complete PTA (E). (F, G) Abnormal development of posterior branchial and aortic arches in an E9.5 Raldh2neo/ embryo (coronal sections, hematoxylin-eosin staining). While the maxillo–mandibular and second (b2) branchial arches are well-developed in the mutant, there is severe hypoplasia of third (b3) and more posterior arches, although grooves (arrowheads) indicate the presence of putative pharyngeal pouches (P2–P4). (H, I) Downregulation of Fgf8 expression in the pharyngeal endoderm (en) of an E9.5 Raldh2/ embryo. Expression in the ectodermal (ec) layer, and in first arch (b1) tissues, is however maintained (p2, p3: pharyngeal pouches). (J, K) Ectopic activation of the RARE-hsp68-lacZ reporter transgene in the pharyngeal region and posterior hindbrain of E9 Tbx1//Crkl/ double mutant embryos. Ectopic RA signaling extends in mutants up to the level of the second branchial arch (b2) and the otocyst (arrowheads). (L, M) Marked downregulation of Cyp26a1 expression in the branchial region of an E9.5 Tbx1//Crkl/ embryo. Note that expression in the tail bud (tb) is unaffected. Drawings adapted from Mendelsohn et al. (1994). (D–G) from Vermot et al. (2003); (H, I) from Niederreither et al. (2003); (J–M) from Guris et al. (2006), with permission.

be consistent with excessive apoptosis of neural crest cells. TGF2 has been implicated as a possible mediator of this phenomenon, as this molecule is widely upregulated in heart tissues of Rxr/ mutants (Kubalak et al., 2002). Furthermore, through a genetic cross, these authors showed that heterozygosity for a TGF2-null allele partly rescued the abnormal apoptosis, as well as outflow tract septation, in Rxr/ embryos. Sucov and colleagues have used an elegant genetic strategy to analyze the migration and fate of neural crest cells in Rar1;Rar double mutants. The combination of a Wnt1-Cre transgene activated in premigratory neural crest cells and a “floxed” Rosa26-lacZ reporter allele allows a selective and stable marking of neural crestderived cells through -galactosidase activity (Jiang et al., 2000). Introduction of these alleles in Rar1;Rar mutants revealed that neural crest cells undergo normal migration along the posterior branchial arches and colonize the outflow tract to the same extent as in wild-type embryos (Jiang et al., 2002). An additional genetic strategy provided evidence that RAR/RXR function is not required

cell-autonomously (Jiang et al., 2002). The authors conclude that it is the surrounding tissues, rather than the cardiac neural crest proper, that respond directly to retinoic acid signaling in order to induce neural crest cells to initiate aorticopulmonary septation. Embryonic retinoic acid required for the proper realization of outflow tract septation is likely produced by the RALDH2 enzyme. Raldh2/ mutants, when rescued from early lethality by transient maternal retinoic acid supplementation, consistently exhibit PTA (Niederreither et al., 2001). Due to the relative difficulty of obtaining viable mutants at fetal stages using this rescue procedure, no detailed analysis of the pattern of the large vessels has been attempted. An interesting observation was made from another Raldh2 mutation behaving as a hypomorphic allele, due to the presence of an intronic neo minigene (Vermot et al., 2003). When combined to a null allele, this mutation leads to a neonatal lethal phenotype due to PTA, associated with variable abnormal patterns of the ascending aorta and other great vessels (Fig. 4D–E). No other aspect of the Raldh2/ phenotype was recapitulated in

Chapter  |  3.3  Retinoids and Heart Development

this mutant, indicating a particular sensitivity of the mechanisms involved in aortic arch patterning and outflow tract septation to a decrease in retinoic acid levels.

III.B.  Retinoic Acid Deficiency Affects Posterior Branchial Arch Development During normal development, Raldh2 is expressed in var­ ious mesodermal derivatives, including the mesoderm of the posterior (3rd–6th) branchial arches. Its absence of expression in the hindbrain neuroepithelium suggested a lack of expression in the corresponding migratory neural crest cells, which has been confirmed in chick embryos by double immunolabeling (Berggren et al., 1999). The foregut and branchial arch endoderm are also devoid of Raldh2 expression (Niederreither et al., 1997, 2003). Embryological and molecular studies have been performed both on the retinoic acid-rescued Raldh2/ mutants (Niederreither et al., 2003) and the neo-containing hypomorphic mutants (Vermot et al., 2003). In both cases, a profound disorganization of the developing posterior branchial arches has been found. While the second branchial arches were well-developed, more posterior (3rd–6th) arches were hypoplastic or unrecognizable and only one or two, instead of the normal three pairs of aortic arches, were formed at the corresponding level (Fig. 4F,G). Analysis of neural crest cell-specific markers showed the presence of post-otic hindbrain neural crest cells migrating along the abnormal pharyngeal region, although the lack of organized segmental migratory patterns suggested that cardiac neural crest cells may not be able to achieve their proper migratory path fully (Niederreither et al., 2001, 2003). A genetic strategy to stably label the neural crest cell lineage (Jiang et al., 2000) has not been employed in these mutants; hence one cannot presently conclude whether the lack of outflow tract septation reflects a deficiency of post-migratory cells or an inability of these cells to mediate septation events, as described in Rar1;Rar mutants (Jiang et al., 2002). Molecularly, the most striking alterations were observed in the pharyngeal endoderm of the rescued Raldh2/ mutants and the hypomorphic mutants. Downregulation of Hoxa1 and Hoxb1 was seen in both the foregut mesoderm and endoderm, while Fgf8 expression was selectively missing in the endoderm, but not the surface ectoderm (Fig. 4H,I), of the posterior branchial region (Niederreither et al., 2003; Vermot et al., 2003). A study on the effects of an RAR antagonist on cultured wild-type mouse embryos during branchial arch development (early somite stages) yielded consistent results, as several genes were found to be affected within the foregut endoderm (Wendling et al., 2000). It therefore appears that the embryonic posterior pharyngeal region requires local retinoic acid synthesis by RALDH2. Retinoic acid signaling is indispensable

247

for the correct development of the posterior (3rd, 4th and 6th) branchial arches and the corresponding aortic arches. Although synthesized in the mesoderm, retinoic acid exerts at least some of its effects in a paracrine manner by inducing gene responses in the endodermal layer. Some of the retinoic acid-dependent signals, such as FGF8, are important regulators of outflow tract formation, being expressed throughout the second heart field (see Chapter 2.2) and pharyngeal endoderm, and required for its survival and proliferation (Park et al., 2006). The hypomorphic and null Raldh2 mutants displayed additional defects of organ derivatives of the posterior pharyngeal pouches, including hypoplasia of the thymus and lack of parathyroids which, together with outflow tract abnormalities, recapitulated the traits of the human DiGeorge syndrome (DGS) (Niederreither et al., 2003; Vermot et al., 2003). These abnormalities are also part of the more pleiotropic phenotype of compound RAR mutants (Mendelsohn et al., 1994). Human DGS is caused by heterozygous microdeletions of region q11 of chromosome 22, which most likely corresponds to a contiguous gene syndrome involving the T-box factor TBX1 and the adaptator-protein CRKL (for a review see Baldini, 2005). The expressivity of this syndrome is known to be highly susceptible to genetic and/or environmental modifiers (Driscoll et al., 1993). The fact that DGS-like abnormal­ ities are the only patent defects in mice with reduced retinoic acid-synthesizing activity (Vermot et al., 2003) led to the suggestion that genetic or dietary conditions leading to decreased embryonic retinoic acid levels could be among the modifiers of the human DGS (or possibly, on their own, lead to DGS-like abnormalities in families without 22q11 deletions). Two studies have investigated such possible relationships by analyzing the levels of retinoic acid signaling in mouse models for the DGS, i.e., knockout mutants for the Tbx1 and Crkl genes (Guris et al., 2006; Roberts et al., 2006). Both of these mutants were found to exhibit locally increased retinoic acid signaling activity in the posterior pharyngeal region where the DGS-like abnormalities arise. Expression of an endogenous retinoic acid-target gene (Rar) and of a RARE-lacZ reporter transgene was abnormally elevated and extended to an ectopic anterior location in single or compound null mutants for Tbx1 and Crkl (see Fig. 4J,K). This abnormal signaling has been correlated to locally ectopic expression of Raldh2, as well as downregulation of the RA-metabolizing enzymes Cyp26a1 and Cyp26b1, in pharyngeal tissues (Fig. 4L,M). Genetic proof that enhanced retinoic acid signaling contributed to the DGS-like defects was obtained by generating triple Tbx//Crkl//Raldh2/ heterozygous mutants, in which a subset of the Tbx//Crkl/ defects (the thymic hypoplasia) were decreased in penetrance (Guris et al., 2006). These findings were somewhat unexpected, considering that collectively all genetic models with decreased retinoic

248

acid signaling (or RAR function) lead to DGS-like abnormalities. They probably highlight an important regulatory interference between DGS-causing genes (Tbx1 in particular) and retinoid signaling, and indicate that an important function of these genes is to negatively-regulate the levels of retinoic acid signaling by controlling the proper extent of its synthesis and metabolism. Actually these findings are not paradoxical as it is known that excess retinoic acid when provided on gestational day 9 in mouse can affect development of the posterior branchial arches, leading to thymic hypoplasia and outflow tract defects (Mulder et al., 1998, 2000). Inhibition of CYP26 enzyme function produces loss of posterior pharyngeal arches and aortic arteries, similar to retinoic acid excess (Roberts et al., 2006). These various results underscore the importance of the proper regulation of retinoic acid levels in the branchial region, which is normally achieved by interplay between RALDH2-mediated synthesis and site-specific metabolism by the CYP26A1, B1 and C1 enzymes. Expression of these enzymes is under complex regulatory feedbacks, including by TBX1. A reciprocal regulation of Tbx1 by retinoic acid may also occur, as zebrafish embryos treated with retinoic acid show reduced or absent Tbx1 expression (Zhang et al., 2006). Augmented Tbx1 levels have not been observed, however, in retinoic acid-deficient mouse embryos (Vermot et al., 2003) (our unpublished data). Future studies will better delineate the regulatory relationships between retinoic acid signaling and other mediators of pharyngeal development.

IV.  Regulation of myocardial cell proliferation and differentiation IV.A.  Epicardial Retinoic Acid Induces Myocardial Growth The therapeutic value of retinoic acid and retinoid compounds to inhibit cell growth and induce differentiation, while better understood in some tumor cell models (for a review see Sun and Lotan, 2002), are less clear in regulation of myocardial proliferation. Clues to functional signaling networks that induce proliferation of myocardial cells during embryonic and fetal development and the possible involvement of retinoic acid have mainly been obtained from work performed in chick embryos and in RAR/RXR mouse mutants. A key to this understanding is the characterization of cross-talk occurring between two adjacent regions of the heart – the epicardium and myocardium. The epicardium, the epithelial outer lining of the developing heart, performs two main functions: (1) it regulates growth of the adjacent compact myocardial region; (2) through epithelial–mesenchymal transformation, it contributes to the formation of the coronary vasculature (see the chapters of Part 5 for details). The growth-promoting effect of the epicardium has been clearly established by surgical removal experiments in the chick,

PART  |  3  Patterning of the Early Heart Tube

resulting in a hypocellularity of the myocardial compact zone (a phenotype termed “spongy myocardium”). In mouse, mutations of various genes expressed in the epicardium and encoding structural proteins or growth and transcription factors (see Chapter 5.2 for references) result in similar spongy myocardial phenotypes, as well as coronary vascular defects. There are several lines of evidence supporting the hypothesis that retinoic acid would be produced in the epicardium in order to regulate rapid cardiac growth at late embryonic and fetal stages. First, RALDH2 is selectively expressed in the pericardial and epicardial layers of the mouse heart from about E11.5 onwards, although the response of the RARE-hsp68lacZ reporter transgene is also seen throughout the myocardial layers (Moss et al., 1998) (Fig. 2D). Most significantly, both the Rar;Rar compound mutants and Rxr/-null mutants display severe hypoplasia of the ventricular myocardium, especially in its compact zone, a phenotype highly reminiscent of other mutants with dysregulated epicardial function (Kastner et al., 1994, 1997b; Sucov et al., 1994; Merki et al., 2005). The Rxr-null mutants that die in utero between E12.5 and E16.5 also exhibit defects of the ventricular septum, atrioventricular cushions and conotruncal ridges (Gruber et al., 1996). When characterized on an ultrastructural level, the outermost, subepicardial developing cardiomyocytes appear to undergo premature differentiation, as evidenced by the presence of striated myofibrils, sarcoplasmic reticulum and intercalation discs. A reduction in the rate of proliferation occurs in the same cell layer (Kastner et al., 1997b). These data support a model in which retinoic acid originating from the epicardial layer would control differentiation of subepicardial cardiomyocytes (Xavier-Neto et al., 2000; Perez-Pomares et al., 2002). However, although intuitively one would expect retinoic acid to act by diffusion, i.e., as a paracrine signal to elicit gene responses in subepicardial cells, there is recent evidence indicating that it may rather act within epicardial cells in order to induce secondary signals regulating cardiomyocyte growth. The strongest evidence for such an “intracrine” function comes from an epicardialspecific knockout of Rxr. Indeed, this tissue-specific loss of function is able to phenocopy, albeit in a less severe manner, the myocardial growth defect of Rxr/-null mutants (Merki et al., 2005). An explanation for the lower severity is that placental defects occurring in the null mutants (Sapin et al., 1997), which may contribute to cardiovascular failure, are not recapitulated in the epicardial-specific mutants. Likewise, a dominant-negative Rar transgene selectively expressed in the epicardium also produces myocardial compact zone outgrowth defects (Chen et al., 2002). Consistent negative data have been obtained by inactivating Rxr in ventricular cardiomyocytes, using a Cre recombinase under the control of the myosin light chain Mlc2v promoter (Chen et al., 1998). This knockout produced no abnormal phenotype, even though Rxr was inactivated in at least 80% of the ventricular myocytes. Along the same lines, expression of an RXR transgene driven by the cardiac -MHC promoter

Chapter  |  3.3  Retinoids and Heart Development

failed to rescue the Rxr/-null phenotype, further indicating that the requirement for RXR function is not intrinsic to myocardial cells (Subbarayan et al., 2000). A chimera analysis performed by Tran and Sucov (1998) also indicated that RXR function is carried out by a nonmyocyte lin­eage of the heart. Neural crest cell-specific and endocardialspecific excision of Rxr also resulted in phenotypically normal mice (Chen et al., 2002; Merki et al., 2005). Several approaches have been employed to explore pathways and cross-tissue interactions by which retinoids regulate cardiomyocyte proliferation. Stuckmann et al. (2003) found that retinoic acid acts in a similar manner to erythropoietin (Epo), inducing cell proliferation via (a) secreted mitogen(s), and proposed a model in which both the retinoic acid and Epo pathways act in parallel for inducing cardiac mitogens. The physical nature of the retinoic acid-induced trophic signals have been elusive, although biochemical assays showed that retinoic acid signaling increases the PI3K and ERK kinase pathways (Kang and Sucov, 2005). Other likely candidates for retinoic acid-regulated trophic signals are members of the FGF family. In both chick (Mikawa, 1995) and mouse (Lavine et al., 2005) FGFs regulate an early phase of cardiomyoblast proliferation. While a number of FGFs are present in the developing heart, excess retinoic acid appears to specifically increase epicardial expression of Fgf9 (Lavine et al., 2005). However, in the epicardialspecific Rxr mouse mutants expression of Fgf2, but not Fgf9, was reduced (Merki et al., 2005). Wnt signaling was also affected in these mutants, as the levels of Wnt9b expression and of active Wnt signaling (assayed by catenin protein levels) were downregulated. The Wnt sig­ naling pathway cooperates in the epithelial-to-mesenchymal transition of epicardial cells, so that in the mutants the epicardium is hindered in its ability to contribute to cardiomyocytes and integrate into the coronary vessel wall (Merki et al., 2005). Reduced trophic signaling in Rxr/-null mutants results in energy deprivation and also reduction in metabolic target gene expression, not unlike a fetal form of cardiomyopathy (Ruiz-Lozano et al., 1998). The global alterations in gene expression may be due to lack of RXR interaction with specific chromatin remodeling factors. Among these factors, the Polybromo protein BAF180 is recruited to retinoic acid-response elements of potential target genes in heart development, and its mutation produces ventricular hypoplasia similar to the Rxr/ mutants (Wang et al., 2004).

IV.B.  Retinoid Regulation of Heart Differentiation has Implications for Regeneration and Progenitor Cell Specification A key issue in cardiac regenerative medicine is to understand how developmental pathways are recapitulated when mesenchymal or embryonic stem cells are routed

249

into a cardiac lineage. A variety of retinoic acid-regulated growth factor pathways such as FGF and Sonic hedgehog (Shh) (Schneider et al., 2001; Niederreither et al., 2002c; Mic et al., 2004; Kondo et al., 2005; Ribes et al., 2006) are required for directed differentiation of progenitor populations (Wichterle et al., 2002; Glaser and Brustle, 2005). While FGF signaling is clearly implicated in regulating the acquisition and maintenance of cardiac cell fate (for a review see Solloway and Harvey, 2003), conclusive evidence that retinoic acid directly regulates cardiac stem cell determination is lacking. In the limited number of systems in which the role of retinoic acid in directing stem cell differentiation was examined, there is mixed evidence indicating that it may act in a concentration-dependent manner in directing pluripotent cells towards a cardiac lineage. For example, Wobus et al. (1997) showed that retinoic acid at a relatively physiological (108 M) concentration accelerates cardiac differentiation of embryonic stem (ES) cells cultured as embryoid bodies. Similarly, RXR agonists increase the number of beating cardiomyocytes in embryoid bodies (Honda et al., 2005). Potentially, retinoic acid concentration may regulate lineage selection, as higher concentrations (106 M) drive undifferentiated embryonic stem and P19 cells toward a neuronal lineage. In embryonic regions with high retinoid signaling, a switch to lower retinoic acid levels might drive specification of cardiac cells. As an example, retinoic acid deprivation in zebrafish reduces the size of the cardiac progenitor pool (Keegan et al., 2005) (see Section II.B), suggesting a role for endogenous retinoic acid in inhibiting cardiac stem cell expansion, potentially by driving cells toward other lin­eages. Effects on the differentiation of progenitor cells may occur in a concentration-dependent manner. In the trabecular region of the midgestational heart, intermediate or low levels of retinoic acid would allow a limited number of stem cells to differentiate. Blocking retinoic acid signaling might be expected to expand the number of progenitors, but reduce differentiated cardiomyocytes, analogous to how retinoic acid given to adult primary cardiomyocytes in culture blocks their hypertrophic response (Zhou et al., 1995; Wu et al., 1996; Wang et al., 2002). A direct role of retinoic acid in driving stem cell differentiation has been seen in the hematopoietic system, where inhibition of retinoic acid signaling expands stem cell numbers (Chute et al., 2006). Low levels of retinoic acid may be required at an early stage to induce levels of FGF signaling required for stem cell survival. In addition, independent cell-specific roles of retinoic acid may exist. In particular, retinoic acid may play a sustained role in the epicardium, where it allows for cardiac growth and repair. During regeneration of the injured zebrafish heart, two markers of the epicardium (Tbx18 and Raldh2) are increased in their expression (Lepilina et al., 2006). Markers of the precardiac field (Hand2, Nkx2.5, Tbx20) are then expressed in the regenerating region. FGF receptor (FGFR) signaling

250

via Fgf17b appears critical, as its disruption through a dominant negative FGFR produces larger wounds. The machinery to optimize regeneration from progenitors may be present in mammals but lying dormant (see chapters of Parts 15 and 16 for more information). Reactivation of inherent regenerative capacity of the heart by stimulating downstream targets (such as RALDH2, FGFs or Shh) may be a feasible alternate approach to cardiac stem cell replacement therapies.

Acknowledgments Work in the authors’ laboratories is supported by the American Heart Association (0330265N), the National Institute of Health (R01 HL070733) (K.N.), and the Centre National de la Recherche Scientifique, Institut National de la Santé et de la Recherche Médicale, Université de Strasbourg, Institut Universitaire de France, Agence Nationale de la Recherche, Fondation pour la Recherche Médicale, (P.D.). The authors thank Professor P. Chambon for initiating some of the work described herein.

References Baldini, A., 2005. Dissecting contiguous gene defects: TBX1. Curr. Opin. Genet. Dev. 15, 279–284. Balmer, J.E., Blomhoff, R., 2005. A robust characterization of retinoic acid response elements based on a comparison of sites in three species. J. Steroid. Biochem. Mol. Biol. 96, 347–354. Bastien, J., Rochette-Egly, C., 2004. Nuclear retinoid receptors and the transcription of retinoid-target genes. Gene 328, 1–16. Berggren, K., McCaffery, P., Drager, U., Forehand, C.J., 1999. Differential distribution of retinoic acid synthesis in the chicken embryo as determined by immunolocalization of the retinoic acid synthetic enzyme, RALDH-2. Dev. Biol. 210, 288–304. Blentic, A., Gale, E., Maden, M., 2003. Retinoic acid signalling centres in the avian embryo identified by sites of expression of synthesising and catabolising enzymes. Dev. Dyn. 227, 114–127. Bohnsack, B.L., Lai, L., Dolle, P., Hirschi, K.K., 2004. Signaling hierarchy downstream of retinoic acid that independently regulates vascular remodeling and endothelial cell proliferation. Genes Dev. 18, 1345–1358. Buckingham, M., Meilhac, S., Zaffran, S., 2005. Building the mammalian heart from two sources of myocardial cells. Nat. Rev. Genet. 6, 826–835. Callis, T.E., Wang, D.Z., 2008. Taking microRNAs to heart. Trends Mol. Med. 14, 254–260. Chazaud, C., Chambon, P., Dolle, P., 1999. Retinoic acid is required in the mouse embryo for left–right asymmetry determination and heart morphogenesis. Development 126, 2589–2596. Chen, J., Kubalak, S.W., Chien, K.R., 1998. Ventricular muscle-restricted targeting of the RXRalpha gene reveals a non-cell-autonomous requirement in cardiac chamber morphogenesis. Development 125, 1943–1949. Chen, T.H., Chang, T.C., Kang, J.O., Choudhary, B., Makita, T., Tran, C. M., Burch, J.B., Eid, H., Sucov, H.M., 2002. Epicardial induction of fetal cardiomyocyte proliferation via a retinoic acid-inducible trophic factor. Dev. Biol. 250, 198–207.

PART  |  3  Patterning of the Early Heart Tube

Chute, J.P., Muramoto, G.G., Whitesides, J., Colvin, M., Safi, R., Chao, N.J., McDonnell, D.P., 2006. Inhibition of aldehyde dehydrogenase and retinoid signaling induces the expansion of human hematopoietic stem cells. Proc. Natl. Acad. Sci. USA 103, 11707–11712. Clagett-Dame, M., DeLuca, H.F., 2002. The role of vitamin A in mammalian reproduction and embryonic development. Annu. Rev. Nutr. 22, 347–381. Collop, A.H., Broomfield, J.A., Chandraratna, R.A., Yong, Z., Deimling, S.J., Kolker, S.J., Weeks, D.L., Drysdale, T.A., 2006. Retinoic acid signaling is essential for formation of the heart tube in Xenopus. Dev. Biol. 291, 96–109. Cui, J., Michaille, J.J., Jiang, W., Zile, M.H., 2003. Retinoid receptors and vitamin A deficiency: differential patterns of transcription during early avian development and the rapid induction of RARs by retinoic acid. Dev. Biol. 260, 496–511. Dersch, H., Zile, M.H., 1993. Induction of normal cardiovascular development in the vitamin A-deprived quail embryo by natural retinoids. Dev. Biol. 160, 424–433. Dickman, E.D., Smith, S.M., 1996. Selective regulation of cardiomyocyte gene expression and cardiac morphogenesis by retinoic acid. Dev. Dyn. 206, 39–48. Driscoll, D.A., Salvin, J., Sellinger, B., Budarf, M.L., McDonaldMcGinn, D.M., Zackai, E.H., Emanuel, B.S., 1993. Prevalence of 22q11 microdeletions in DiGeorge and velocardiofacial syndromes: implications for genetic counselling and prenatal diagnosis. J. Med. Genet. 30, 813–817. Drysdale, T.A., Patterson, K.D., Saha, M., Krieg, P.A., 1997. Retinoic acid can block differentiation of the myocardium after heart specification. Dev. Biol. 188, 205–215. Dupe, V., Davenne, M., Brocard, J., Dolle, P., Mark, M., Dierich, A., Chambon, P., Rijli, F.M., 1997. In vivo functional analysis of the Hoxa-1 3 retinoic acid response element (3RARE). Development 124, 399–410. Emoto, Y., Wada, H., Okamoto, H., Kudo, A., Imai, Y., 2005. Retinoic acid-metabolizing enzyme Cyp26a1 is essential for determining territories of hindbrain and spinal cord in zebrafish. Dev. Biol. 278, 415–427. Fujinaga, M., 1997. Development of sidedness of asymmetric body structures in vertebrates. Int. J. Dev. Biol. 41, 153–186. Ghatpande, S., Brand, T., Zile, M., Evans, T., 2006. Bmp2 and Gata4 function additively to rescue heart tube development in the absence of retinoids. Dev. Dyn. 235, 2030–2039. Ghatpande, S., Ghatpande, A., Zile, M., Evans, T., 2000. Anterior endoderm is sufficient to rescue foregut apoptosis and heart tube morphogenesis in an embryo lacking retinoic acid. Dev. Biol. 219, 59–70. Ghyselinck, N.B., Wendling, O., Messaddeq, N., Dierich, A., Lampron, C., Decimo, D., Viville, S., Chambon, P., Mark, M., 1998. Contribution of retinoic acid receptor beta isoforms to the formation of the conotruncal septum of the embryonic heart. Dev. Biol. 198, 303–318. Glaser, T., Brustle, O., 2005. Retinoic acid induction of ES-cell-derived neurons: the radial glia connection. Trends Neurosci. 28, 397–400. Gruber, P.J., Kubalak, S.W., Pexieder, T., Sucov, H.M., Evans, R.M., Chien, K.R., 1996. RXR alpha deficiency confers genetic susceptibility for aortic sac, conotruncal, atrioventricular cushion, and ventricular muscle defects in mice. J. Clin. Invest. 98, 1332–1343. Guris, D.L., Duester, G., Papaioannou, V.E., Imamoto, A., 2006. Dosedependent interaction of Tbx1 and Crkl and locally aberrant RA signaling in a model of del22q11 syndrome. Dev. Cell 10, 81–92. Harrison, E.H., 2005. Mechanisms of digestion and absorption of dietary vitamin A. Annu. Rev. Nutr. 25, 87–103.

Chapter  |  3.3  Retinoids and Heart Development

Heine, U.I., Roberts, A.B., Munoz, E.F., Roche, N.S., Sporn, M.B., 1985. Effects of retinoid deficiency on the development of the heart and vascular system of the quail embryo. Virchows Arch. B. Cell Pathol. Incl. Mol. Pathol. 50, 135–152. Helms, J.A., Kim, C.H., Eichele, G., Thaller, C., 1996. Retinoic acid sign­ aling is required during early chick limb development. Development 122, 1385–1394. Heyman, R.A., Mangelsdorf, D.J., Dyck, J.A., Stein, R.B., Eichele, G., Evans, R.M., Thaller, C., 1992. 9-cis retinoic acid is a high affinity ligand for the retinoid X receptor. Cell 68, 397–406. Hochgreb, T., Linhares, V.L., Menezes, D.C., Sampaio, A.C., Yan, C.Y., Cardoso, W.V., Rosenthal, N., Xavier-Neto, J., 2003. A caudorostral wave of RALDH2 conveys anteroposterior information to the cardiac field. Development 130, 5363–5374. Honda, M., Hamazaki, T.S., Komazaki, S., Kagechika, H., Shudo, K., Asashima, M., 2005. RXR agonist enhances the differentiation of cardiomyocytes derived from embryonic stem cells in serum-free conditions. Biochem. Biophys. Res. Commun. 333, 1334–1340. Iulianella, A., Lohnes, D., 2002. Chimeric analysis of retinoic acid receptor function during cardiac looping. Dev. Biol. 247, 62–75. Jiang, X., Choudhary, B., Merki, E., Chien, K.R., Maxson, R.E., Sucov, H.M., 2002. Normal fate and altered function of the cardiac neural crest cell lineage in retinoic acid receptor mutant embryos. Mech. Dev. 117, 115–122. Jiang, X., Rowitch, D.H., Soriano, P., McMahon, A.P., Sucov, H.M., 2000. Fate of the mammalian cardiac neural crest. Development 127, 1607–1616. Jiang, Y., Drysdale, T.A., Evans, T., 1999. A role for GATA-4/5/6 in the regulation of Nkx2.5 expression with implications for patterning of the precardiac field. Dev. Biol. 216, 57–71. Kang, J.O., Sucov, H.M., 2005. Convergent proliferative response and divergent morphogenic pathways induced by epicardial and endocardial signaling in fetal heart development. Mech. Dev. 122, 57–65. Kastner, P., Grondona, J.M., Mark, M., Gansmuller, A., LeMeur, M., Decimo, D., Vonesch, J.L., Dolle, P., Chambon, P., 1994. Genetic analysis of RXR alpha developmental function: convergence of RXR and RAR signaling pathways in heart and eye morphogenesis. Cell 78, 987–1003. Kastner, P., Mark, M., Ghyselinck, N., Krezel, W., Dupe, V., Grondona, J.M., Chambon, P., 1997a. Genetic evidence that the retinoid signal is transduced by heterodimeric RXR/RAR functional units during mouse development. Development 124, 313–326. Kastner, P., Messaddeq, N., Mark, M., Wendling, O., Grondona, J.M., Ward, S., Ghyselinck, N., Chambon, P., 1997b. Vitamin A deficiency and mutations of RXRalpha, RXRbeta and RARalpha lead to early differentiation of embryonic ventricular cardiomyocytes. Development 124, 4749–4758. Kawaguchi, R., Yu, J., Honda, J., Hu, J., Whitelegge, J., Ping, P., Wiita, P., Bok, D., Sun, H., 2007. A membrane receptor for retinol binding protein mediates cellular uptake of vitamin A. Science 315, 820–825. Keegan, B.R., Feldman, J.L., Begemann, G., Ingham, P.W., Yelon, D., 2005. Retinoic acid signaling restricts the cardiac progenitor pool. Science 307, 247–249. Kondo, T., Johnson, S.A., Yoder, M.C., Romand, R., Hashino, E., 2005. Sonic hedgehog and retinoic acid synergistically promote sensory fate specification from bone marrow-derived pluripotent stem cells. Proc. Natl. Acad. Sci. USA 102, 4789–4794. Kostetskii, I., Yuan, S.Y., Kostetskaia, E., Linask, K.K., Blanchet, S., Seleiro, E., Michaille, J.J., Brickell, P., Zile, M., 1998. Initial retinoid requirement for early avian development coincides with retinoid

251

receptor coexpression in the precardiac fields and induction of normal cardiovascular development. Dev. Dyn. 213, 188–198. Kostetskii, I., Jiang, Y., Kostetskaia, E., Yuan, S., Evans, T., Zile, M., 1999. Retinoid signaling required for normal heart development regulates GATA-4 in a pathway distinct from cardiomyocyte differentiation. Dev. Biol. 206, 206–218. Kubalak, S.W., Hutson, D.R., Scott, K.K., Shannon, R.A., 2002. Elevated transforming growth factor beta2 enhances apoptosis and contributes to abnormal outflow tract and aortic sac development in retinoic X receptor alpha knockout embryos. Development 129, 733–746. Lai, L., Bohnsack, B.L., Niederreither, K., Hirschi, K.K., 2003. Retinoic acid regulates endothelial cell proliferation during vasculogenesis. Development 130, 6465–6474. Lamb, T.D., Pugh, E.N., Jr., 2004. Dark adaptation and the retinoid cycle of vision. Prog. Retin. Eye. Res. 23, 307–380. Lavine, K.J., Yu, K., White, A.C., Zhang, X., Smith, C., Partanen, J., Ornitz, D.M., 2005. Endocardial and epicardial derived FGF signals regulate myocardial proliferation and differentiation in vivo. Dev. Cell 8, 85–95. Lee, R.Y., Luo, J., Evans, R.M., Giguere, V., Sucov, H.M., 1997. Compartment-selective sensitivity of cardiovascular morphogenesis to combinations of retinoic acid receptor gene mutations. Circ. Res. 80, 757–764. Lepilina, A., Coon, A.N., Kikuchi, K., Holdway, J.E., Roberts, R.W., Burns, C.G., Poss, K.D., 2006. A dynamic epicardial injury response supports progenitor cell activity during zebrafish heart regeneration. Cell 127, 607–619. Levin, A.A., Sturzenbecker, L.J., Kazmer, S., Bosakowski, T., Huselton, C., Allenby, G., Speck, J., Kratzeisen, C., Rosenberger, M., Lovey, A., et al., 1992. 9-cis retinoic acid stereoisomer binds and activates the nuclear receptor RXR alpha. Nature 355, 359–361. Lough, J., Barron, M., Brogley, M., Sugi, Y., Bolender, D.L., Zhu, X., 1996. Combined BMP-2 and FGF-4, but neither factor alone, induces cardiogenesis in non-precardiac embryonic mesoderm. Dev. Biol. 178, 198–202. Maden, M., 2002. Retinoid signalling in the development of the central nervous system. Nat. Rev. Neurosci. 3, 843–853. Maden, M., 2007. Retinoic acid in the development, regeneration and maintenance of the nervous system. Nat. Rev. Neurosci. 8, 755–765. Maden, M., Sonneveld, E., van der Saag, P.T., Gale, E., 1998. The distribution of endogenous retinoic acid in the chick embryo: implications for developmental mechanisms. Development 125, 4133–4144. Martin, M., Gallego-Llamas, J., Ribes, V., Kedinger, M., Niederreither, K., Chambon, P., Dolle, P., Gradwohl, G., 2005. Dorsal pancreas agenesis in retinoic acid-deficient Raldh2 mutant mice. Dev. Biol. 284, 399–411. McCaffery, P., Drager, U.C., 2000. Regulation of retinoic acid signaling in the embryonic nervous system: a master differentiation factor. Cytokine Growth Factor Rev. 11, 233–249. Mendelsohn, C., Lohnes, D., Decimo, D., Lufkin, T., LeMeur, M., Chambon, P., Mark, M., 1994. Function of the retinoic acid receptors (RARs) during development (II). Multiple abnormalities at various stages of organogenesis in RAR double mutants. Development 120, 2749–2771. Meno, C., Saijoh, Y., Fujii, H., Ikeda, M., Yokoyama, T., Yokoyama, M., Toyoda, Y., Hamada, H., 1996. Left–right asymmetric expression of the TGF beta-family member lefty in mouse embryos. Nature 381, 151–155. Meno, C., Shimono, A., Saijoh, Y., Yashiro, K., Mochida, K., Ohishi, S., Noji, S., Kondoh, H., Hamada, H., 1998. lefty-1 is required for

252

left–right determination as a regulator of lefty-2 and nodal. Cell 94, 287–297. Merki, E., Zamora, M., Raya, A., Kawakami, Y., Wang, J., Zhang, X., Burch, J., Kubalak, S.W., Kaliman, P., Belmonte, J.C., Chien, K. R., Ruiz-Lozano, P., 2005. Epicardial retinoid X receptor alpha is required for myocardial growth and coronary artery formation. Proc. Natl. Acad. Sci. USA 102, 18455–18460. Mic, F.A., Haselbeck, R.J., Cuenca, A.E., Duester, G., 2002. Novel retinoic acid generating activities in the neural tube and heart identified by conditional rescue of Raldh2 null mutant mice. Development 129, 2271–2282. Mic, F.A., Sirbu, I.O., Duester, G., 2004. Retinoic acid synthesis controlled by Raldh2 is required early for limb bud initiation and then later as a proximodistal signal during apical ectodermal ridge formation. J. Biol. Chem. 279, 26698–26706. Mikawa, T., 1995. Retroviral targeting of FGF and FGFR in cardiomyocytes and coronary vascular cells during heart development. Ann. NY. Acad. Sci. 752, 506–516. Molotkov, A., Molotkova, N., Duester, G., 2005. Retinoic acid generated by Raldh2 in mesoderm is required for mouse dorsal endodermal pancreas development. Dev. Dyn. 232, 950–957. Morton, S.U., Scherz, P.J., Cordes, K.R., Ivey, K.N., Stainier, D.Y., Srivastava, D., 2008. MicroRNA-138 modulates cardiac patterning during embryonic development. Proc. Natl. Acad. Sci. USA 105, 17830–17835. Moss, J.B., Xavier-Neto, J., Shapiro, M.D., Nayeem, S.M., McCaffery, P., Drager, U.C., Rosenthal, N., 1998. Dynamic patterns of retinoic acid synthesis and response in the developing mammalian heart. Dev. Biol. 199, 55–71. Mulder, G.B., Manley, N., Maggio-Price, L., 1998. Retinoic acid-induced thymic abnormalities in the mouse are associated with altered pharyngeal morphology, thymocyte maturation defects, and altered expression of Hoxa3 and Pax1. Teratology 58, 263–275. Mulder, G.B., Manley, N., Grant, J., Schmidt, K., Zeng, W., Eckhoff, C., Maggio-Price, L., 2000. Effects of excess vitamin A on development of cranial neural crest-derived structures: a neonatal and embryologic study. Teratology 62, 214–226. Niederreither, K., McCaffery, P., Drager, U.C., Chambon, P., Dolle, P., 1997. Restricted expression and retinoic acid-induced downregulation of the retinaldehyde dehydrogenase type 2 (RALDH-2) gene during mouse development. Mech. Dev. 62, 67–78. Niederreither, K., Subbarayan, V., Dolle, P., Chambon, P., 1999. Embryonic retinoic acid synthesis is essential for early mouse postimplantation development. Nat. Genet. 21, 444–448. Niederreither, K., Vermot, J., Messaddeq, N., Schuhbaur, B., Chambon, P., Dolle, P., 2001. Embryonic retinoic acid synthesis is essential for heart morphogenesis in the mouse. Development 128, 1019–1031. Niederreither, K., Abu-Abed, S., Schuhbaur, B., Petkovich, M., Chambon, P., Dolle, P., 2002a. Genetic evidence that oxidative derivatives of retinoic acid are not involved in retinoid signaling during mouse development. Nat. Genet. 31, 84–88. Niederreither, K., Vermot, J., Fraulob, V., Chambon, P., Dolle, P., 2002b. Retinaldehyde dehydrogenase 2 (RALDH2)-independent patterns of retinoic acid synthesis in the mouse embryo. Proc. Natl. Acad. Sci. USA 99, 16111–16116. Niederreither, K., Vermot, J., Schuhbaur, B., Chambon, P., Dolle, P., 2002c. Embryonic retinoic acid synthesis is required for forelimb growth and anteroposterior patterning in the mouse. Development 129, 3563–3574. Niederreither, K., Vermot, J., Le Roux, I., Schuhbaur, B., Chambon, P., Dolle, P., 2003. The regional pattern of retinoic acid synthesis

PART  |  3  Patterning of the Early Heart Tube

by RALDH2 is essential for the development of posterior pharyngeal arches and the enteric nervous system. Development 130, 2525–2534. Niederreither, K., Dolle, P., 2006. Molecular mediators of retinoic acid signaling during development. Adv. Dev. Biol. 16, 105–143. Oosterveen, T., Niederreither, K., Dolle, P., Chambon, P., Meijlink, F., Deschamps, J., 2003. Retinoids regulate the anterior expression boundaries of 5 Hoxb genes in posterior hindbrain. EMBO. J. 22, 262–269. Osmond, M.K., Butler, A.J., Voon, F.C., Bellairs, R., 1991. The effects of retinoic acid on heart formation in the early chick embryo. Development 113, 1405–1417. Oulad-Abdelghani, M., Chazaud, C., Bouillet, P., Mattei, M.G., Dolle, P., Chambon, P., 1998. Stra3/lefty, a retinoic acid-inducible novel member of the transforming growth factor-beta superfamily. Int. J. Dev. Biol. 42, 23–32. Park, E.J., Ogden, L.A., Talbot, A., Evans, S., Cai, C.L., Black, B.L., Frank, D.U., Moon, A.M., 2006. Required, tissue-specific roles for Fgf8 in outflow tract formation and remodeling. Development 133, 2419–2433. Perez-Pomares, J.M., Phelps, A., Sedmerova, M., Carmona, R., GonzalezIriarte, M., Munoz-Chapuli, R., Wessels, A., 2002. Experimental studies on the spatiotemporal expression of WT1 and RALDH2 in the embryonic avian heart: a model for the regulation of myocardial and valvuloseptal development by epicardially derived cells (EPDCs). Dev. Biol. 247, 307–326. Pijnappel, W.W., Hendriks, H.F., Folkers, G.E., van den Brink, C.E., Dekker, E.J., Edelenbosch, C., van der Saag, P.T., Durston, A.J., 1993. The retinoid ligand 4-oxo-retinoic acid is a highly active modulator of positional specification. Nature 366, 340–344. Reijntjes, S., Blentic, A., Gale, E., Maden, M., 2005. The control of morphogen signalling: regulation of the synthesis and catabolism of retinoic acid in the developing embryo. Dev. Biol. 285, 224–237. Ribes, V., Wang, Z., Dolle, P., Niederreither, K., 2006. Retinaldehyde dehydrogenase 2 (RALDH2)-mediated retinoic acid synthesis regulates early mouse embryonic forebrain development by controlling FGF and sonic hedgehog signaling. Development 133, 351–361. Ribes, V., Fraulob, V., Petkovich, M., Dolle, P., 2007. The oxidizing enzyme CYP26A1 tightly regulates the availability of retinoic acid in the gastrulating mouse embryo to ensure proper head development and vasculogenesis. Dev. Dyn. 236 (3), 644–653. Roberts, C., Ivins, S., Cook, A.C., Baldini, A., Scambler, P.J., 2006. Cyp26 genes a1, b1 and c1 are down-regulated in Tbx1 null mice and inhibition of Cyp26 enzyme function produces a phenocopy of DiGeorge Syndrome in the chick. Hum. Mol. Genet. 15, 3394–3410. Ruiz-Lozano, P., Smith, S.M., Perkins, G., Kubalak, S.W., Boss, G.R., Sucov, H.M., Evans, R.M., Chien, K.R., 1998. Energy deprivation and a deficiency in downstream metabolic target genes during the onset of embryonic heart failure in RXRalpha / embryos. Development 125, 533–544. Ryckebusch, L., Wang, Z., Bertrand, N., Lin, S.C., Chi, X., Schwartz, R., Zaffran, S., Niederreither, K., 2008. Retinoic acid deficiency alters second heart field formation. Proc. Natl. Acad. Sci. USA 105, 2913–2918. Sapin, V., Dolle, P., Hindelang, C., Kastner, P., Chambon, P., 1997. Defects of the chorioallantoic placenta in mouse RXRalpha null fetuses. Dev. Biol. 191, 29–41. Schneider, R.A., Hu, D., Rubenstein, J.L., Maden, M., Helms, J.A., 2001. Local retinoid signaling coordinates forebrain and facial morphogenesis by maintaining FGF8 and SHH. Development 128, 2755–2767.

Chapter  |  3.3  Retinoids and Heart Development

Searcy, R.D., Yutzey, K.E., 1998. Analysis of Hox gene expression during early avian heart development. Dev. Dyn. 213, 82–91. Sirbu, I.O., Gresh, L., Barra, J., Duester, G., 2005. Shifting boundaries of retinoic acid activity control hindbrain segmental gene expression. Development 132, 2611–2622. Sirbu, I.O., Duester, G., 2006. Retinoic-acid signalling in node ectoderm and posterior neural plate directs left-right patterning of somitic mesoderm. Nat. Cell Biol. 8, 271–277. Sirbu, I.O., Zhao, X., Duester, G., 2008. Retinoic acid controls heart anteroposterior patterning by down-regulating Isl1 through the Fgf8 pathway. Dev. Dyn. 237, 1627–1635. Smith, S.M., Dickman, E.D., Thompson, R.P., Sinning, A.R., Wunsch, A.M., Markwald, R.R., 1997. Retinoic acid directs cardiac laterality and the expression of early markers of precardiac asymmetry. Dev. Biol. 182, 162–171. Solloway, M.J., Harvey, R.P., 2003. Molecular pathways in myocardial development: a stem cell perspective. Cardiovasc. Res. 58, 264–277. Stainier, D.Y., Fishman, M.C., 1992. Patterning the zebrafish heart tube: acquisition of anteroposterior polarity. Dev. Biol. 153, 91–101. Stuckmann, I., Evans, S., Lassar, A.B., 2003. Erythropoietin and retinoic acid, secreted from the epicardium, are required for cardiac myocyte proliferation. Dev. Biol. 255, 334–349. Subbarayan, V., Mark, M., Messadeq, N., Rustin, P., Chambon, P., Kastner, P., 2000. RXRalpha overexpression in cardiomyocytes causes dilated cardiomyopathy but fails to rescue myocardial hypoplasia in RXRalpha-null fetuses. J. Clin. Invest. 105, 387–394. Sucov, H.M., Dyson, E., Gumeringer, C.L., Price, J., Chien, K.R., Evans, R.M., 1994. RXR alpha mutant mice establish a genetic basis for vitamin A signaling in heart morphogenesis. Genes Dev. 8, 1007–1018. Sun, S.Y., Lotan, R., 2002. Retinoids and their receptors in cancer development and chemoprevention. Crit. Rev. Oncol. Hematol. 41, 41–55. Tanaka, Y., Okada, Y., Hirokawa, N., 2005. FGF-induced vesicular release of Sonic hedgehog and retinoic acid in leftward nodal flow is critical for left–right determination. Nature 435, 172–177. Thaller, C., Eichele, G., 1987. Identification and spatial distribution of retinoids in the developing chick limb bud. Nature 327, 625–628. Tran, C.M., Sucov, H.M., 1998. The RXRalpha gene functions in a noncell-autonomous manner during mouse cardiac morphogenesis. Development 125, 1951–1956. Tsukui, T., Capdevila, J., Tamura, K., Ruiz-Lozano, P., RodriguezEsteban, C., Yonei-Tamura, S., Magallon, J., Chandraratna, R.A., Chien, K., Blumberg, B., et al., 1999. Multiple left–right asymmetry defects in Shh(/) mutant mice unveil a convergence of the shh and retinoic acid pathways in the control of Lefty-1. Proc. Natl. Acad. Sci. USA 96, 11376–11381. Ulven, S.M., Gundersen, T.E., Weedon, M.S., Landaas, V.O., Sakhi, A. K., Fromm, S.H., Geronimo, B.A., Moskaug, J.O., Blomhoff, R., 2000. Identification of endogenous retinoids, enzymes, binding proteins, and receptors during early postimplantation development in mouse: important role of retinal dehydrogenase type 2 in synthesis of all-trans-retinoic acid. Dev. Biol. 220, 379–391. Vermot, J., Niederreither, K., Garnier, J.M., Chambon, P., Dolle, P., 2003. Decreased embryonic retinoic acid synthesis results in a DiGeorge syndrome phenotype in newborn mice. Proc. Natl. Acad. Sci. USA 100, 1763–1768.

253

Wagner, M., Han, B., Jessell, T.M., 1992. Regional differences in retinoid release from embryonic neural tissue detected by an in vitro reporter assay. Development 116, 55–66. Wang, H.J., Zhu, Y.C., Yao, T., 2002. Effects of all-trans retinoic acid on angiotensin II-induced myocyte hypertrophy. J. Appl. Physiol. 92, 2162–2168. Wang, Z., Zhai, W., Richardson, J.A., Olson, E.N., Meneses, J.J., Firpo, M.T., Kang, C., Skarnes, W.C., Tjian, R., 2004. Polybromo protein BAF180 functions in mammalian cardiac chamber maturation. Genes Dev. 18, 3106–3116. Wang, Z., Dolle, P., Cardoso, W.V., Niederreither, K., 2006. Retinoic acid regulates morphogenesis and patterning of posterior foregut derivatives. Dev. Biol. 297 (2), 433–445. Wasiak, S., Lohnes, D., 1999. Retinoic acid affects left–right patterning. Dev. Biol. 215, 332–342. Waxman, J.S., Keegan, B.K., Roberts, R.W., Poss, K.D., Yelon, D., 2008. Hoxb5b acts downstream of retinoic acid signaling in the forelimb field to restrict heart field potential in zebrafish. Dev. Cell 15, 923–934. Wendling, O., Dennefeld, C., Chambon, P., Mark, M., 2000. Retinoid signaling is essential for patterning the endoderm of the third and fourth pharyngeal arches. Development 127, 1553–1562. Wichterle, H., Lieberam, I., Porter, J.A., Jessell, T.M., 2002. Directed differentiation of embryonic stem cells into motor neurons. Cell 110, 385–397. Wilson, J.G., Warkany, J., 1950. Congenital anomalies of heart and great vessels in offspring of vitamin A-deficient rats. Am. J. Dis. Child 79, 963. Wilson, J.G., Roth, C.B., Warkany, J., 1953. An analysis of the syndrome of malformations induced by maternal vitamin A deficiency. Effects of restoration of vitamin A at various times during gestation. Am. J. Anat. 92, 189–217. Wobus, A.M., Kaomei, G., Shan, J., Wellner, M.C., Rohwedel, J., Ji, G., Fleischmann, B., Katus, H.A., Hescheler, J., Franz, W.M., 1997. Retinoic acid accelerates embryonic stem cell-derived cardiac differentiation and enhances development of ventricular cardiomyocytes. J. Mol. Cell Cardiol. 29, 1525–1539. Wu, J., Garami, M., Cheng, T., Gardner, D.G., 1996. 1,25(OH)2 vitamin D3, and retinoic acid antagonize endothelin-stimulated hypertrophy of neonatal rat cardiac myocytes. J. Clin. Invest. 97, 1577–1588. Xavier-Neto Jr, J., Neville, C.M., Shapiro, M.D., Houghton, L., Wang, G.F., Nikovits, W., Stockdale, F.E., Rosenthal, N., 1999. A retinoic acid-inducible transgenic marker of sino-atrial development in the mouse heart. Development 126, 2677–2687. Xavier-Neto, J., Shapiro, M.D., Houghton, L., Rosenthal, N., 2000. Sequential programs of retinoic acid synthesis in the myocardial and epicardial layers of the developing avian heart. Dev. Biol. 219, 129–141. Zhang, L., Zhong, T., Wang, Y., Jiang, Q., Song, H., Gui, Y., 2006. TBX1, a DiGeorge syndrome candidate gene, is inhibited by retinoic acid. Int. J. Dev. Biol. 50, 55–61. Zhou, M.D., Sucov, H.M., Evans, R.M., Chien, K.R., 1995. Retinoiddependent pathways suppress myocardial cell hypertrophy. Proc. Natl. Acad. Sci. USA 92, 7391–7395. Zile St, M.H., Kostetskii, I., Yuan, S., Kostetskaia, E., Amand, T.R., Chen, Y., Jiang, W., 2000. Retinoid signaling is required to complete the vertebrate cardiac left/right asymmetry pathway. Dev. Biol. 223, 323–338.

Chapter 3.4

A System for Describing Congenital Cardiac Malformations and Correlating Them with Abnormal Cardiac Development* Robert H. Anderson1, Antoon F. M. Moorman2, Sandra Webb3 and Nigel A. Brown3 1

Cardiac Unit, Institute of Child Health, University College London, London, UK Heart Failure Research Center, Academic Medical Center, Amsterdam, The Netherlands 3 Department of Basic Sciences, Anatomy and Developmental Biology, St George’s University of London, London, UK 2

I.  Introduction It is an understandable goal to seek to establish the precise way in which abnormal cardiac development results in congenitally malformed hearts. It is equally understandable to seek to unravel the deficient genetic and molecular mechanisms that underscore the disordered development. The literature devoted to the morphology of congenital cardiac malformations is replete with polemics addressing nomenclature and classification (Van Mierop, 1970; Van Praagh, 1970, 1972, 2000; Shinebourne et al., 1976; Anderson, 2000). Significant strides have now been made in clarifying and demystifying these previous disagreements (Jacobs et al., 2006; Jacobs and Anderson, 2006). In this chapter, we will review the advances made in analyzing and describing congenitally malformed hearts. We seek to show how the lessons learned in reaching the current consensus can be used readily by those investigating the structure of the developing normal and abnormal heart to provide a synthesis that, hopefully, will be equally acceptable to clinicians, morphologists and developmental biologists.

II.  Describing the cardiac components Many of the difficulties that arose in describing congenitally malformed hearts had their basis in interpretations, * Illustrations for this chapter are copyright 2010 Robert H. Anderson. Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

rather than in differences in descriptions. The problems with interpretation, nonetheless, frequently reflected the words used in description, or else differences in the definitions used for the various components under discussion (Van Mierop, 1970; Van Praagh, 1970). The potential for semantic disagreement is magnified in the setting of the developing heart, since the cardiac components change their positions and relationships with time during the process of development. We now know, for example, that the cells seen at a given stage of development are themselves not necessarily subsequently fixed within the developing heart (Moorman et al., 2007). Ideally, therefore, the same words should be used by all who describe these changing scenes, particularly the interrelationships of the parts of the heart. Sadly, this is rarely the case. A significant problem persists in this respect with even the simplest of descriptions, since biologists and clinicians currently differ in the way they account for the orthogonal planes of the body. The biologist is used to dealing with mammalian species having a plantigrade posture. For the biologist, the head is at the front of the body, and hence is appropriately described as being anterior. For the clinician, in contrast, all structures within the body are described in terms of the anatomical position, in which the subject stands upright and faces the observer. In this posture, it is the sternum which is to the front and hence is anterior. The head, in contrast, is uppermost and is described as being superior. So as to circumvent 255

256

such problems with the use of “anterior” and “posterior,” we will avoid the use of these adjectives. The opposite ends of the vertebral column will be described as being cranial and caudal, these adjectives being equally applicable to species taking orthograde or plantigrade postures. The only exception to this rule will be when we describe the atrioventricular endocardial cushions, since these are now widely accepted as being positioned “superiorly” and “inferiorly”, and the leaflets to which they give rise in the setting of atrioventricular septal defect with common atrioventricular junction are recognized as the superior and inferior bridging leaflets. The sternal surface of the body will be described as being ventral, and structures closer to the spinal cord as being dorsal. Right and left, of course, retain their currency. Other problems in relating cardiac development to structure have reflected the description of the walls that interpose between the adjacent cardiac chambers or trunks. Currently, it is usual to describe any structure that interposes between adjacent chambers or channels as a septum. Is this appropriate? We prefer to define a septal structure as that part that can be removed without exiting from the cavities of the heart or the lumens of the great vessels (Anderson and Brown, 2000). Consider our definition relative to the most distal intrapericardial components of the heart. In the developing embryo, the distal part of the outflow tract is initially a solitary channel with myocardial walls (Fig. 1A). In the formed heart, separate channels exist for the aortic and pulmonary pathways, each with its own discrete walls (Fig. 1B). If we are to understand the changes that take place during cardiac development, therefore, and relate them to malformations of the definitive structures, it is necessary that we use words that distinguish the two situations shown in Fig. 1. It should be obvious that, in the normal postnatal heart (Fig. 1B), there is no such thing as an “aortopulmonary septum”. Nor is there any “septum” between the aortic and pulmonary roots, since each valve is enclosed within its own separate sinusal sleeve, nor between the muscular subpulmonary infundibulum and the remainder of the heart, the infundibulum being a free-standing muscular sleeve (Merrick et al., 2000). A similar situation pertains in the arrangement of the walls interposed between the right and left atriums. It is well-established that, during normal embryology, a structure grows from the dorsal atrial wall and interposes between right and left sides of the developing atrium (Fig. 2A). This structure is universally known as the primary atrial septum, or “septum primum”. In the definitive heart, this structure forms the floor, or flap valve, of the oval fossa. The cranial rim of the fossa (at least in the human heart), against which the flap valve abuts to close the oval foramen in postnatal life is no more than an infolding between the walls of the caval veins draining to the right atrium, and the right pulmonary veins draining to the left atrium (Fig. 2B).

PART  |  3  Patterning of the Early Heart Tube

Outflow tract

Aorta

Ventricle

(A)

Pulmonary trunk

(B)

Figure 1  The left-hand illustration (A) is from a human embryo at Carnegie stage 12, when the intrapericardial outflow tract has exclusively muscular walls. The middle panel (B) shows a slightly later stage, after cushions have formed in the distal part of the outflow tract. After the completion of septation, this distal part is transformed into the intrapericardial components of the arterial trunks. As can be seen, subsequent to the completion of septation, each arterial trunk has its own walls and there is no “septum” between them, albeit the two arterial trunks are contained within a common pericardial sleeve. Their walls, however, are separated by extramural connective tissue.

Does it help to describe this infolding as the secondary atrial septum or “septum secundum”? We have argued that it would be best to avoid the word “septum” to describe this fold, since we would prefer to restrict the term “septum” to the description of structures that can be removed without creating communications with extracardiac space (Anderson and Brown, 2000). We know from discussions with our closest colleagues that not all are comfortable with this suggestion. If we are to make appropriate correlations between developing and postnatal structures, nonetheless, it is crucial that we distinguish, in some way or other, the anatomic situations created by folds, as opposed to solitary walls, interposed between adjacent chambers or channels, this being exemplified by the difference between the flap valve, derived from the primary septum, and representing a true septum within our definition, and the so-called “septum secundum”, in reality the cranial interatrial fold (see Fig. 2). We should also remember that there are marked differences in anatomy between experimental animals, in which most studies of development are performed, and humans, to whom most accounts of cardiac malformations pertain. Although there are four pulmonary venous orifices in humans, each vein at a corner of the atrial roof, there is typically only a solitary pulmonary venous orifice in small experimental mammals which is situated adjacent to the atrioventricular junctions. It is also the case that the heart in chick lacks any “septum secundum”, with the secondary openings being fenestrations within the cranial portion of the primary septum. It will not have passed unnoticed, we hope, that our preference in writing is to describe the various anatomic components of the heart using English, rather than Latin, words. This may

Chapter  |  3.4  A System for Describing Congenital Cardiac Malformations

Secondary foramen

Mesenchymal cap

Primary atrial septum

Superior endocardial cushion

(A)

Cranial interatrial fold RPV

Left atrium Primary atrial septum (flap valve)

SCV

Right atrium (B) Figure 2  The upper panel (A) is from a human embryo at Carnegie stage 16. It shows how the primary atrial septum, with its mesenchymal cap, is actively growing towards the endocardial cushions which are septating the atrioventricular canal. The upper edge of the primary septum has already broken down to form the secondary atrial foramen. The lower panel (B) shows the structure of the definitive atrial septum, which is formed by the flap valve, derived from the primary septum, anchored on the muscular cranio–ventral rim of the oval fossa. The so-called “septum secundum” is a deep infolding between the attachments of the right pulmonary vein (RPV) to the left atrium and the superior caval vein (SCV) to the right atrium. Exiting through the walls of this “septum”, in reality the cranial interatrial fold, takes the prosector outside the heart.

also offend the sensibilities of those who have become accustomed to more classical descriptions. It is now an established fact, nonetheless, that English, or at least its American version, has become the “lingua franca” of scientific discourse. We would suggest that writers translate Latin terms into the American equivalents, as we will do, since very few now have a working knowledge of Latin grammar and syntax.

III.  Describing the congenitally malformed heart If we examine accounts given by the giants of the past when categorizing congenitally malformed hearts, such as

257

the great Maude Abbott (Abbott, 1936), we find that the descriptions of so-called “simple” lesions, such as inter­atrial or interventricular communications, have stood the test of time well. It was the so-called “complex” malformations that created problems for these initial pioneers of classification. The resolution of these problems was provided by the so-called “segmental” approach to analysis (Van Praagh, 1972). In this respect, the choice of “segment” to describe the parts of the malformed hearts was less than optimal, since the parts themselves are obviously markedly different in terms of their morphology, while to biologists segments are initially supposed to have identical structures. The notion that the mammalian heart is a segmental structure is in dispute (see Chapter 3.2), and therefore this likely misuse of “segment” is something we will need to live with, since the term “segmental analysis” is now firmly embedded in the lexicon of those who diagnose and treat congenital cardiac malformations. This approach, which divides the heart into its atrial, ventricular and arterial components, was also subsequently shown to be less than perfect. This was because it paid less attention to the junctions between the cardiac components than to the “segments” themselves. Those involved in diagnosing and treating the malformations have now appreciated the need to pay particular attention to the way the cardiac components are joined (or not joined) together. The system most frequently used now for describing patients with congenital cardiac disease, therefore, is known as sequential segmental analysis (Shinebourne et al., 1976). The sequential approach has also undergone marked changes (Tynan et al., 1979; Anderson et al., 1984; Anderson and Ho, 1997; Anderson, 2000) since its first appearance (Shinebourne et al., 1976). Because of this, it is important that those seeking to apply the system to analysis of experimental animals with congenital cardiac malformations should use the latest versions (Anderson and Ho, 1997; Anderson, 2000). The basis of sequential segmental analysis, as with the initial segmental approach, is that there are very limi­ ted ways in which the parts of the heart can be joined, or not joined, together (Van Praagh, 1972; Shinebourne et al., 1976). There are limitless combinations of the lesions found to be deforming the different parts. Description of the way in which the parts are joined together, therefore, provides the cardiac template for the given individual, making it an easy matter to list the lesions within this template. One additional anatomic principle has proved crucial in adjudicating the disputes that arose during the refinement of sequential segmental analysis, and bringing it to its current version. This is the so-called “morphological method” (Van Praagh et al., 1980). This states that one variable structure should not be used to define another cardiac component which is itself variable. Instead, the most constant part of the component should be used for the purposes of morphological definition. This is best exemplified by considering the distinction between the atrial chambers. It is the venous

258

PART  |  3  Patterning of the Early Heart Tube

connections that provide the most obvious means of distinguishing right from left atriums. The venous connections themselves, however, can be abnormal, as for example in the lesion known as totally anomalous pulmonary venous connection. As is shown in Fig. 3, where the pulmonary veins connect to an extracardiac site, it remains possible to identify the left atrium because of the remaining morphological components, principally because of the characteristic appearance of the appendage, which is the most constant atrial component in the setting of congenital cardiac disease. As we will show, depending on the structure of the appendage, it is always possible to distinguish the morphological nature of a given atrial chamber, irrespective of its position relative to its neighbor and the remainder of the heart. The same goes for the ventricles when analyzed on the basis of the structure of their apical trabecular components, these being the most constant ventricular building-block. Hence, when describing the atrial or ventricular chambers, we account for them as being morphologically right or morphologically left. As we will see, very frequently they are not right-sided or left-sided when the heart is congenitally malformed, nor are the ventricles. This fact renders the use in isolation of “right” or left” as adjectives for the description of congenitally malformed hearts less than optimal.

IV.  The starting point for analysis Our analysis of the heart, whether the organ is normal or abnormal, begins with examination of the atrial chambers. As explained above, accurately to distinguish between morphologically right and left atriums, we examine the structure of the appendages, specifically the extent of the pectinate muscles relative to the atrioventricular junctions

Appendage

(Uemura et al., 1995). In atriums with a morphologically right appendage, the pectinate muscles encircle the atrio­ ventricular junction to reach the so-called cardiac crux, the crossing point of the septal components with the atrioventricular junctions (Fig. 4B). In atriums of morphologically left type, the pectinate muscles are confined within the tubular appendage, and the dorsal atrial wall is smooth (Fig. 4A). On the basis that all hearts, whether normal or congenitally malformed, possess two atrial appendages and that, according to the extent of the pectinate muscles as shown in Fig. 4, appendages can be of only right or left morphology, it follows that there are only four possible patterns of arrangement for the atrial appendages. These are the usual pattern, often called “situs solitus”, and its mirror-image, frequently terms “situs inversus” (see Chapters 4.1–4.3). The other two patterns are those of right- and left-isomerism, in which the appendages are morphological enantiomers (Fig. 5). The isomeric patterns are almost always found in the setting of the clinical syndrome known as visceral heterotaxy, or “situs ambiguus” (Van Mierop et al., 1972). In reality, when each of the systems of thoracic and abdominal organs is analyzed in its own right, there is nothing ambiguous about these arrangements. Visceral heterotaxy is often stratified into the syndromes of asplenia and polysplenia (Van Mierop et al., 1972). While these splenic malformations usually exist with one or other isomeric arrangement, the combinations are not sufficiently precise to permit patients having isomerism of the right atrial appendages to be always described as “asplenic”, nor those with isomerism of the left appendages to be described as exhibiting “polysplenia” (Uemura et al., 1995). Any problems in description, therefore, are resolved simply by describing the arrangements in each of the systems separately, the

Body of left atrium

Appendage Pulmonary veins to portal venous system Tricuspid valve Mitral valve

Vestibule Figure 3  It may be thought that the attachment of the pulmonary veins provides the best means of identifying the morphologically left atrium. As shown in the figure, however, it is still possible to identify the left atrium on the basis of the characteristic morphology of its appendage, even when the pulmonary veins are all joined to an extracardiac site, in this example, to the hepatic portal venous system within the abdomen.

(A)

Smooth vestibule

(B)

Pectinates

Figure 4  These illustrations show the structure of the atrial appendages in the normal human heart. The distinguishing feature is the extent of the pectinate muscles relative to the atrioventricular junctions, the pectinate muscles encircling the entirety of the parietal part of the junction in the morphologically right appendage.

259

Chapter  |  3.4  A System for Describing Congenital Cardiac Malformations

appendages serving as the final arbiter in the description of atrial arrangement. In this respect, it should also be noted that “dextrocardia” is not a diagnosis. It simply accounts for the position of the heart. The location of the heart needs to be described separately from the direction of the cardiac apex, since these features are themselves mutually independent. Rather than describing “dextrocardia” or “levocardia”, therefore, it is preferable to state when the heart is abnormally located whether it occupies the left chest, the right chest, or the middle, and then separately to account for whether the apex points to the left, the right, or to the middle. We are now able to provide rational explanations as to why the heart and organs can develop in the various patterns described above. It is well-established that the development of the organs that exhibit morphological leftness is under the control of a sequence of genes including Nodal, Lefty 1, Lefty 2, cited 2 and Pitx2 (Bamforth et al., 2004) (see Chapters 4.1–4.3). Under the influence of these genes, the organs of the body develop different features on the left and right sides such that, in the human, the right lung has three lobes fed by a long bronchus, while the left lung has two lobes, fed by a short bronchus. Within the heart, it is only the atrial component of the primary tube that is able to respond to the genes producing leftness and rightness. This is because the atrial appendages balloon out from either side of the atrial component of the linear heart tube, and hence each appendage responds in different fashion to the various genes. In the ventricular part, the apical parts of the developing ventricles balloon from the primary tube in sequence, rather than in parallel. Hence, the genes may have

equal effect on both ventricles as they develop. Isomerism can involve the atrial appendages, as demonstrated in animal models by knocking out genes such as Pitx2 (Fig. 6), but not the ventricles. Formation of the ventricular loop is random in the setting of patients with isomeric atrial appendages, as in mouse models showing abnormal laterality, such as the iv/iv mouse (Seo et al., 1992). Ventricular isomerism is exceedingly rare in humans with congenital cardiac malformation (Rinne et al., 2000).

V.  Analysis of the atrioventricular junctions The key to analysis of the atrioventricular junctions, having first appreciated that there are two such junctions in the normal heart, is to distinguish between the union of the atrial and ventricular myocardial masses across the junctions, as opposed to the nature of the valve (or valves) that guard them. Thus, the atrial myocardium can be joined to the ventricular mass across a well-formed junction, even in the absence of the leaflets of the atrioventricular valves. This arrangement is seen in the congenitally unguarded tricuspid orifice (Anderson et al., 1990), and to a lesser extent in all examples of Ebstein’s malformation (Ho et al., 2000). When considering the union of the muscular components across the junctions, there are various ways in which the myocardium of the atrial chambers can connect, or not connect, with the ventricular myocardial mass. This feature of union of the atrial and ventricular musculatures across the junctions is described as the type of atrioventricular

Usual arrangement

Mirror-imagery

Right isomerism

Left isomerism

Figure 5  As based on the extent of the pectinate muscles relative to the atrioventricular junctions, there are only four possible arrangements for the atrial appendages, the usual one and its mirror-image, and the two isomeric variants.

260

PART  |  3  Patterning of the Early Heart Tube

Systemic venous sinus

Primary septum

Left atrium

Left sinus horn (A) Right appendages

Symmetrical systemic venous sinis (B)

Midline pulmonary vein

Figure 6  These scanning electron micrographs of the atrial chambers of the mouse heart at the same stage of gestation contrast the normal lateralized arrangement (A) with the situation of right isomerism (B) seen in the mouse with knockout of the Pitx2 gene.

connection. The morphological arrangement of the atrioventricular valves does no more than modify the type of connection. So as to appreciate the types of connection, first it is necessary to establish the arrangement of the atrial appendages, as discussed above, and then to determine the topological structure of the ventricular mass. Almost always there are two chambers within the ventricular mass. In the past, arguments raged as to how best to describe these chambers, and whether they should even be considered as ventricles when they were incomplete and rudimentary. Incomplete chambers are those that do not possess all their normal components, the normal chambers possessing inlet and outlet parts in addition to their apical trabecular components (Jacobs and Anderson, 2006). It is now accepted

that, in those malformations with one dominant ventricle and another incomplete and rudimentary chamber, the circulation is functionally univentricular, but the anatomic arrangement is biventricular, since the morphology of the incomplete and small chamber can still be determined according to the nature of its apical trabecular component. Congenitally malformed hearts do exist with a truly solitary chamber in the ventricular mass, but these are exceedingly rare. When found, the solitary ventricle is of indeterminate morphology. In all other situations, there are two ventricles within the ventricular mass and these are arranged on the basis of either right-hand or left-hand topology. Another important step in analysis, therefore, is to determine the topologic arrangement of the ventricular mass. The topology is dependent on the way the ventricular component of the embryonic heart loops during its development. When the linear heart tube bends to the right, this produces right-hand ventricular topology. After development in this fashion, the definitive right ventricle accepts only the palmar surface of the right hand when either hand is placed on the ventricular surface with the thumb in the inlet component and the fingers in the outlet (Fig. 7A). Looping of the heart tube to the left, in contrast, produces a mirror-image of the topological arrangement, where only the palmar surface of the left hand can be placed on the septal surface of the right ventricle with the thumb in the inlet and the fingers in the outlet (Fig. 7B). Knowing the ventricular topology and the arrangement of the atrial appendages, it is possible to account for all possible ways in which the atrial chambers can be joined to the ventricles, albeit the options are strictly limited. In most instances, the cavities of the atrial and ventricular chambers will be joined in the expected fashion, so that the blood from the right atrium drains to the right ventricle and that from the left atrium to the left ventricle. This is described as concordant atrioventricular connections. Such connections can be found either in the normally structured heart, or in totally mirror-imaged structure (Fig. 8A,B). These arrangements depend on the linear heart tube turning to the right in the setting of usual arrangement of the atrial appendages, or to the left when the atrial appendages are themselves mirror-imaged. However, if the heart tube loops to the left but the atrial appendages are formed in the usual fashion, then the right atrium will join to the morphologically left ventricle, and the left atrium to the morphologically right ventricle. Should the heart tube turn to the right in the setting of mirror-imaged atrial appendages, the end result will be comparable, namely to produce discordant atrioventricular connections (Fig. 8C,D). In the setting of isomeric atrial appendages, irrespective of the direction of looping of the linear heart tube, the atrioventricular connections must be biventricular but mixed, since one atrium will be concordantly connected to its underlying ventricle, but the other atrium will be discordantly connected. Such a mixed connection can be found with right- or

261

Chapter  |  3.4  A System for Describing Congenital Cardiac Malformations

(A)

(B)

Right handed topology

Left handed topology

Figure 7  The cartoon shows the two forms of ventricular topology produced by either rightward (A) or leftward bending (B) of the linear heart tube during cardiac development. Topology is described according to the way the palms of the hands can be laid on the septal surface of the morphologically right ventricle.

left-isomerism, and with right-hand or left-hand ventricular topology (Fig. 8E,F). When the appendages are isomeric, therefore, to provide a complete description it is always necessary to specify both the type of isomerism and the topology of the ventricular mass. The patterns described so far are the possible connections that can be found when each atrial chamber is joined to its own ventricle across the atrioventricular junctions, in other words the possible biventricular connections. There is a further set of possible connections. These are found when the atrial chambers are joined to only one chamber within the ventricular mass. It is well-established that, during normal development, the atrial component of the linear heart tube is supported above that part of the tube that will form the definitive left ventricle. As the right ventricle is added to the tube from the secondary lineage, there is rearrangement within the inner curvature of the ventricular loop so that the right atrium becomes joined to the newly-formed morphological right ventricle (Fig. 9). If this rearrangement does not occur, there are two possibilities – either the right atrioventricular connection fails to form, or both atriums retain their connection to the definitive left ventricle. It is also known that the rearrangement within the inner curvature can be overly exuberant, so that not only the right, but also the left, atrium achieves a connection with the right ventricle, or alternatively that the connection between the left atrium and the left ventricle fails to develop. All of these changes during development underscore the second grouping of atrioventricular connections, comprising the univentricular arrangements.

As shown in Fig. 10, it is possible for these univentricular patterns to exist with any of the four arrangements of the atrial appendages, with either double inlet ventricle or absence of the right or left atrioventricular connections, and with the atriums themselves connected to a dominant left ventricle when the right ventricle is incomplete, to a dominant right ventricle when the left ventricle is incomplete, or very rarely to a solitary and incomplete ventricle (Anderson et al., 1983). It is also possible of course, that the rearrangement within the inner heart curvature produces intermediate states between the biventricular and univentricular arrangement. When this occurs, there is straddling and overriding of one or other of the atrioventricular valves or, rarely, both valves. When describing these arrangements, we find it helpful to distinguish between the features of straddling and overriding (Milo et al., 1979). We account for straddling when the tension apparatus of the valve is attached to both sides of the ventricular septum (Fig. 11A). We describe overriding, in contrast, when the atrial myocardium is joined to the musculature of both right and left ventricles, the junction itself then overriding the crest of the ventricular septum (Fig. 11B). Such overriding and straddling of an atrioventricular valve can also be found when one atrioventricular connection is absent. This rare situation underscores a third category of connections, namely the uniatrial but biventricular arrangement (Fig. 12). This pattern, although rare, can also exist with any arrangement of the atrial appendages, with the absence of either the right or left atrioventricular

262

(A)

PART  |  3  Patterning of the Early Heart Tube

(B) Left atrium

Right atrium

Right ventricle

Concordant atrioventricular connections (C)

(D)

Left ventricle

(A)

Right atrium

Discordant atrioventricular connections (E)

(F)

Right ventricle

Left ventricle

(B) Figure 9  The sections show the stages before (A) and after (B) remolding of the primary heart tube so the atrioventricular junction, initially undivided and looking primarily into the left ventricle, becomes committed directly to the developing right ventricle. Biventricular and mixed atrioventricular connections Figure 8  The cartoon shows the possible connections across the atrio­ ventricular junctions when each atrium is joined to its own ventricle, in other words with biventricular atrioventricular connections. The situation with isomeric appendages (E, F) is shown only for right isomerism, but the same situation exists for left isomerism.

connection, and with either right-hand or left-hand ventricular topology (Ho et al., 1982). The arrangement produced by the absence of an atrioventricular connection gives rise to the commonest forms of atrioventricular valvar atresia, either tricuspid atresia or mitral atresia (Fig. 13). It was once thought that atrioventricular valvar atresia was produced by the presence of an imperforate atrioventricular valve blocking the connection between either the right atrium or the right ventricle to produce tricuspid atresia, or between the left atrium and the left ventricle to produce mitral atresia. Such variants can exist (Fig. 14), but they are the exception rather than the rule. The presence of imperforate valves, nonetheless, shows how the morphology of the atrioventricular valves can modify the type of atrioventricular connection. Thus, both of the examples shown in Fig. 15 exhibit biventricular atrioventricular connections, but in both

instances the hearts produced are functionally univentricular, since the small ventricle in both cases is incapable of independently supporting either the pulmonary or systemic circulation. The other example of how variation in valvar morphology is seen when there is a common atrioventricular valve guarding a common atrioventricular junction, rather than tricuspid and mitral valves guarding separate right and left atrioventricular junctions (Fig. 15). In this instance, the presence of the common valve does not change the type of atrioventricular connection, which can be concordant (Fig. 16A) or double inlet through a common atrioventricular valve (Fig. 16B). The common valve, of course, can also be shared unequally between the two ventricles, so that there is a spectrum of malformation, just as is seen with straddling and overriding of either the tricuspid or mitral valves (Milo et al., 1979). The common valve, however, can be divided into separate right and left orifices for the two ventricles when there is a common atrioventricular junction. Such separation of the valve into double orifices is the essence of the so-called “ostium primum” defect, which in reality is an atrioventricular septal defect with shunting exclusively at atrial level (Anderson et al., 1998). This is because the leaflets of the effectively

263

Chapter  |  3.4  A System for Describing Congenital Cardiac Malformations

Usual atrial arrangement

Mirror-imaged arrangement

Absent right AV connection

Dominant left ventricle

Right isomerism

Double inlet AV connection

Solitary and indeterminate ventricle

Left isomerism

Absent left AV connection

Dominant right ventricle

Figure 10  The cartoon shows the possible combinations of atrial chambers and ventricular morphology to give univentricular atrioventricular connections. Morphologically left ventricles are shown with purple cavities and right ventricles with green cavities. The incomplete ventricles can either be right-sided (rightward looping) or left-sided (leftward looping). The middle ventricle on the bottom row, shown in blue, is solitary and of indeterminate morphology.

common valve that bridge the ventricular septum are attached to each other, so as to produce dual orifices within the common junction, but are also attached to the crest of the ventricular septum so as to confine shunting at ventricular level. In the typical variant of atrioventricular septal defect with a common atrioventricular junction (Fig. 17A), the bridging leaflets of the common valve float within the atrioventricular septal defect, and hence there is the potential for shunting at both atrial and ventricular levels (Fig. 17B). In rarer examples the bridging leaflets can be attached to the underside of the atrial septum, and then shunting through the atrioventricular septal defect can take place only at ventricular level (Fig. 17C). The essence of all these variants, therefore, is the common atrioventricular

junction, and it is this feature which is the criterion for diagnosis (Anderson et al., 1998).

VI.  Analysis of the ventriculo– arterial junctions Since the majority of malformations seen in mouse models of congenital cardiac disease involve the outflow tracts, it is necessary to have a consistent system for analysis. This is provided by assessing the anatomical features of the malformations separately; in most instances this involves the ventriculo–arterial junctions. At the level of these junctions, the specific features are the fashion in

264

PART  |  3  Patterning of the Early Heart Tube

Left atrium

Tricuspid valve

Right atrium Right atrium

Septum (B)

(A)

Ventricular septum

Figure 11  The pictures show straddling of the tricuspid valve. The upper panel (A) shows how the tendinous cords of the malformed valve are attached in both ventricles – this is the feature known as straddling. The lower panel, reproduced by kind permission of Dr L. H. Van Mierop, shows how the orifice of the valve overrides the crest of the ventricular septum.

Absent atrioventricular connection

Straddling and overriding AV valve

Figure 12  The cartoon shows how straddling and overriding of the solitary atrioventricular valve found when there is absence of one atrio­ ventricular connection produces an arrangement that, in terms of the connection itself, is uniatrial but biventricular. The illustration shows absence of the right atrioventricular connection with right-hand ventricular topology, but variants can be found with left-hand topology and with absence of the left atrioventricular connection, with any combination possible.

which the arterial trunks origin from the ventricular mass, the relationships of the arterial valves at the junctions and the structure of the outflow tracts supporting the arterial valves. Another feature worthy of consideration is the relationships of the intrapericardial arterial trunks as they exit from the heart. In terms of the hemodynamics produced by malformations of the outflow tracts, easily the most important feature is the nature of ventricular origin of the arterial trunks. We call this the type of ventriculo–arterial connection.

There are four possibilities. In the usual situation, the trunks arise from their morphologically appropriate ventricles, representing concordant connections. In this setting, the arterial trunks almost always spiral, the aorta arising dorsally and to the right of the pulmonary trunk as it exits from the heart (Fig. 18A). In rare circumstances, the arterial trunks can exit in parallel fashion, with the aorta arising ventrally and in left-sided position from the left ventricle. This latter situation has created enormous problems in the past, and has been described with such terms as “anatomically corrected malposition” (Van Praagh, 1976), or “isolated ventricular inversion” when found with aortic-to-mitral valvar continuity (Fig. 18B). These cryptic terms create difficulties in understanding since, in the situation shown in Fig. 18B, the aorta would be left-sided and the arterial trunks would spiral in mirror-imaged fashion if they were “normally related” to the ventricular mass with left-hand topology. When described in terms of concordant connections with parallel arterial trunks, all such confusion is dispelled (Cavalle-Garrido et al., 2007). The second important type of ventriculo–arterial connection is found when the arterial trunks arise from morphologically inappropriate ventricles. This produces the discordant arrangement which can be found with any of the atrioventricular connections. This situation has also been the source of huge polemics since in most instances when the arterial trunks are discordantly connected the aorta arises ventrally from the morphologically right ventricle (Fig. 19A) with its valve supported by a complete muscular infundibulum. This is not always the case, and in rare instances the aorta can arise dorsally and to the right (Fig. 19B) with the leaflets of its valve in fibrous continuity with the leaflets of the mitral valve (Van Praagh, 1970). The discovery of this rare malformation created difficulties for those who used the ventral position of the aorta

265

Chapter  |  3.4  A System for Describing Congenital Cardiac Malformations

Right atrium

Left atrium

Left atrium

Incomplete left ventricle

Right atrium

Dominant right ventricle

Dominant left ventricle (A)

(B)

Figure 13  The illustrations show sections through malformed human hearts with (A) tricuspid and (B) mitral atresia, the section replicating the echocardiographic “four-chamber” cuts. They show that the essence of the malformations is absence of the right (A) or left (B) atrioventricular connections (green dotted lines).

Aorta

Imperforate mitral valve

Right atrium Left atrium

Left ventricle

Hypoplastic left ventricle

Hypoplastic right ventricle Imperforate tricuspid valve (A)

(B)

Figure 14  The illustrations show tricuspid atresia (A) and mitral atresia (B) produced as the consequence of imperforate valves in the setting of concordant atrioventricular connections. Such examples are much rarer than absence of an atrioventricular connection (Fig. 13) as the substrate for atrioventricular valvar atresia.

as their definition of “transposition” (Van Mierop, 1970), since obviously an aorta arising from the right ventricle in a dorsal position would not fulfill the diagnosis for “transposition”, even if arising from the morphologically right ventricle. The solution to this dilemma is to avoid the use of “transposition” in this setting, and describe discordant

connections when the arterial trunks arise from morphologically inappropriate ventricles, since this cannot lead to confusion. The third type of ventriculo–arterial connection is found when both arterial trunks are supported by the same ventricle. This is usually the right ventricle, but both trunks can

266

PART  |  3  Patterning of the Early Heart Tube

Figure 15  The presence of a common atrioventricular valve guarding a common atrioventricular junction (solitary brace) does not change the fact that the atrioventricular connections are concordant (arrows), as seen in the presence of separate right and left atrioventricular junctions (dual braces).

Common Atrioventricular Valve

Right atrium

Right ventricle

Separate Atrioventricular Valve

Left atrium

Left ventricle

(A)

Left atrium Right atrium

Right ventricle (B)

Left ventricle

Figure 16  The two illustrations, replicating echocardiographic fourchamber sections, show how a common atrioventricular valve can guard either atrioventricular junctions connected in concordant fashion (A), or when both are connected to one ventricle, in this instance (B) a dominant right ventricle. The left ventricle is incomplete and hypoplastic. The star shows the muscular ventricular septum.

rarely arise from the morphologically left ventricle (Brandt et al., 1976). Double outlet from the right ventricle has also been the source of many disagreements, since for many years it was argued that both arterial valves would need to be supported by complete muscular outflow tracts, or “conuses”, so as to fulfill the criterions for diagnosis (Fig. 20A). It is now recognized that there can be fibrous continuity between the leaflets of the arterial and atrioventricular valves (Fig. 20B) in many malformed hearts with both arterial trunks arising exclusively from the right ventricle. This fact underscores the need to describe the ventricular origin of the trunks separately, which produces the double-outlet, the structure of the outflow tracts supporting the arterial valves and the relationships of the arterial trunks as they exit from the heart. In many hearts there are indeed complete infundibulums bilaterally (Fig. 20A), and the aorta arises ventrally relative to the pulmonary trunk as it exits from the heart. For those who used the criterion of an aorta as the diagnostic feature of “transposition” (Van Mierop, 1970), this would justify the description of “double-outlet right ventricle with transposition”, but for those who diagnose transposition on the basis of discordant ventriculo–arterial connections (Van Praagh, 1970) this combination is clearly an anatomical impossibility. Both camps are arguing from internally consistent positions, hence there is no way of achieving consensus. The answer, again, is to avoid the use of “transposition” to describe either a specific ventriculo–arterial connection, or an anterior position of the aorta. The latter feature can be described directly, while the ventriculo–arterial connections are best described as being discordant. This then leaves “transposition” to be used in its clinical sense, accounting for the combination of concordant atrioventricular and discordant ventriculo–arterial connections, with congenitally corrected transposition accounting for the com­bination of discordant connections at both atrioventricular and ventriculo–arterial junctions.

267

Chapter  |  3.4  A System for Describing Congenital Cardiac Malformations

Shunting at atrial level

Shunting at atrial and ventricular level

Shunting at ventricular level

Figure 17  The cartoon shows how, in the presence of an atrioventricular septal defect (bracket), the potential of shunting depends on the relationship between the leaflets that bridge through the septal defect and the underside of the atrial septum as opposed to the crest of the ventricular septum. The common atrioventricular junction is the unifying feature of the malformations.

Arterial duct Aortic-mitral continuity

Pulm. Trunk From MRV Aorta

Aorta

Pulmonary trunk

Left ventricle (A)

(B)

Figure 18  The illustrations show the marked difference between concordant ventriculo–arterial connections with spiraling arterial trunks (A), this being the normal arrangement, and the very rare variant with parallel trunks in which the aorta arises ventrally from the morphologically left ventricle (B). In the example shown, the left ventricle is right-sided, so-called left-hand ventricular topology, or “l-loop”. The aorta, however, is concordantly connected to this ventricle, and there is aortic-to-mitral valvar continuity.

Another problem with hearts having double-outlet ventricles is found when one or other of the arterial trunks does not arise exclusively from the right ventricle, but rather overrides an interventricular communication. These variants are particularly important in developmental terms since spectrums exist, on the one hand, between the extremes of double-outlet right ventricle with subaortic interventricular communication and pulmonary stenosis (Fig. 21A) and tetralogy of Fallot with concordant ventriculo–arterial connections (Fig. 21B), and on the other hand, with a separate spectrum between double-outlet right ventricle and subpulmonary interventricular communication (Fig. 22A) and

discordant ventriculo–arterial connections (Fig. 22B). The key to the difference between the spectrums is the location and attachment of the muscular outlet septum. This structure is formed by muscularization of the proximal parts of the cushions that divide the outflow tract in the developing heart (Fig. 23A). If development proceeds normally, this structure fuses with the crest of the muscular ventricular septum so as to wall the aorta into the left ventricle, concomitant with further remodeling of the parts of the ventricles derived from the primary myocardium (Moorman and Christoffels, 2003). The muscularized proximal cushions then form the free-standing subpulmonary infundibulum

268

PART  |  3  Patterning of the Early Heart Tube

Spiralling arterial trunks

Aorta

Aorta

Pulm. trunk

Pulm. trunk

Parallel arterial trunks

(A)

(B)

Figure 19  Both of these hearts have discordant ventriculo–arterial connections, the arterial trunks arising from morphologically-inappropriate ventricles. In the heart shown in (A), the aorta arises anteriorly ventrally and to the right, which is the usual situation. In the heart shown in (B), however, the aorta is positioned dorsally and to the right, and the arterial trunks spiral as they exist from the heart. Hearts of this type create problems for those who define “transposition” in terms of a ventral aorta.

Aortic valve

Pulmonary valve Pulm. trunk Aorta

Interventricular communication (A)

(B)

Fibrous continuity in roof of IVC

Figure 20  Both of the hearts have double outlet from the right ventricle, with the interventricular communication (IVC) in subaortic location. In the heart shown in the left hand panel (A), each arterial valve is supported by a complete muscular infundibulum (red dotted areas), but in the heart shown in the right hand panel (B) there is fibrous continuity between the leaflets of the aortic and mitral valves in the roof of the defect.

(Fig. 23B). In the process of normal development, the entirety of the outflow tract is initially supported above the developing right ventricle (Anderson et al., 2003a,b). The spectrums seen in the setting of double-outlet right ventricle illustrate, on the one hand, the perturbation of the process of transfer of the aorta to the left ventricle and, on the other hand, they show the effects of abnormal formation of the septum that normally divided the developing outflow tracts. The cushions giving rise to this septum are now known to be located in an abnormal position when the heart is programed to produce either double-outlet right ventricle

with subpulmonary defect, or discordant ventriculo– arterial connections. Evidence is now emerging that, rather than spiraling as in normal development (Fig. 24, left hand panel A) the cushions are aligned in parallel in mice showing the spectrum of double-outlet or discordant ventriculo– arterial connections (Fig. 24, right hand panel B). The significance of muscularization of the proximal outflow cushions during normal development (van den Hoff et al., 1999) is shown by another variant of double-outlet from the right ventricle. Thus, in some patients, the outlet septum remains as a fibrous raphe, and the interventricular

269

Chapter  |  3.4  A System for Describing Congenital Cardiac Malformations

Pulm. trunk

Aorta

Narrowed subpulm. outflow tract

Aortic valve

Interventricular communication

(A)

(B)

Figure 21  These hearts show the extremes in the spectrum between double-outlet right ventricle with subaortic defect (A) and subpulmonary (subpulm.) stenosis (bracket), and tetralogy of Fallot with concordant ventriculo–arterial connections (B). Note the location of the muscular outlet septum (star).

Aorta

Aorta

Pulm. trunk

Interventricular communication

Interventricular communication (A)

(B)

Figure 22  The hearts illustrate the extremes in the spectrum between double-outlet right ventricle and subpulmonary defect (A) and discordant ventriculo–arterial connections (B). In the heart shown in B, which is viewed from the right ventricle, the greater part of the pulmonary valve (not seen) is supported within the left ventricle. Note the different orientation of the muscular outlet septum (star) compared to the hearts shown in Fig. 21.

communication opens beneath both arterial valves. A spectrum is then seen in this setting between double-outlet right ventricle (Fig. 25) and double-outlet left ventricle (Brandt et al., 1976). The final variant of double-outlet right ventricle seen in the setting of humans with congenital cardiac

malformations is the one with a so-called noncommitted interventricular communication. In this setting, the defect either opens between the ventricular inlets, rather than to the outlets (as shown in Figs. 20–22 and 25), or else it is a muscular defect in the apical part of the septum, the initial

270

PART  |  3  Patterning of the Early Heart Tube

Muscularizing proximal cushions

Cushions have formed subpulmonary infundibulum Pulmonary trunk

Aorta

Fusion with septum walls aorta into LV

(A)

Aorta

(B)

Figure 23  The sections from human embryos just at (A), and after (B), the completion of ventricular septation, show how the muscularized proximal outflow cushions wall the aorta into the left ventricle, and then become converted to the muscular subpulmonary infundibulum.

Aorta

Pulmonary trunk

Aorta

Pulmonary trunk

Parallel cushions Right ventricle Spiralling cushions Wild-type mouse

Right ventricle

Pitx2 Knock-out

Figure 24  The left-hand panel shows the location of the outflow cushions in the normal mouse. Note that the cushions are spiraling, as are the developing aorta and pulmonary trunk. In the mutant embryo with knockout of the Pitx2 gene (right hand panel), the cushions are orientated in parallel fashion and the arterial trunks develop to produce either double-outlet right ventricle with subpulmonary interventricular communication, or discordant ventriculo–arterial connections.

embryonic interventricular communication having closed during fetal gestation. The closure of the embryonic communication also explains the rare examples of double-outlet right ventricle found in the setting of an intact ventricular septum. The final variant involving abnormal formation of the outflow tracts is again frequent in mouse models of congenital cardiac disease. This is exemplified by the group of connections producing a single outlet from the heart. The commonest lesion within this group is common arterial trunk, in which a solitary arterial trunk leaves the base of the heart via a common ventriculo–arterial junction, and directly supplies the aortic, pulmonary and coronary arterial circulations. In humans, this malformation is typically found with the truncal valve overriding the crest of the muscular ventricular septum (Fig. 26A), but it can be found arising exclusively from the right ventricle (Fig. 26B), this latter variant being much more frequent in mouse models. Common arterial trunk is well-explained on the basis of complete failure of fusion of the cushions throughout the length of the outflow tract. These cushions initially extend to the margins of the pericardial cavity,

Pulm. trunk

Fibrous outlet septum

Aortic outflow tract

Interventricular communication

Figure 25  In this heart with double-outlet right ventricle, the outlet septum is a fibrous raphe, and the interventricular communication opens beneath both arterial valves. As can be seen, this lesion is part of a spectrum leading to double outlet from the left ventricle.

271

Chapter  |  3.4  A System for Describing Congenital Cardiac Malformations

Common arterial trunk

Pulm. arteries Aorta

Coronary artery

Left ventricle Right ventricle Interventricular communication

(A)

(B)

Figure 26  These illustrations show common arterial trunk in its usual guise (A), when the truncal valve (bracket) overrides the crest of the ventricular septum (star), and in a rarer variant when the trunk arises exclusively from the right ventricle, with a completely muscular subtruncal infundibulum.

where they fuse with the back wall of the aortic sac, this being the so-called “aorto–pulmonary septum” (Fig. 27). The “close cousin” of common arterial trunk, the aorto– pulmonary window, is best explained on the basis of failure of this fusion between the distal ends of the outflow cushions and the dorsal wall of the sac, the more so since the hole between the arterial trunks is frequently associated with overriding the origin of the right pulmonary artery (Fig. 28). The various forms of common trunk can then be accounted for on the basis of varying formation of the intrapericardial pulmonary arteries, albeit the mechanism whereby these structures are developed from the distal part of the outflow tract has still to be determined. The proximal part of the outflow tract remains undivided in the setting of the common trunk, but this part of the outflow tract gives rise to both the truncal valvar sinuses and the ventricular outflow tracts (Ya et al., 1998; Webb et al., 2003). Indeed, the morphology of the so-called doubly committed ventricular septal defect points to the fact that the more distal part of the proximal outflow tract, which gives rise to the arterial valvar leaflets and their supporting sinuses, can become separated into the aortic and pulmonary roots, while the outflow tracts themselves retain a common ventriculo–arterial junction (Fig. 29). It is no coincidence, therefore, that both common arterial trunk and the doubly committed defect are commonly seen in patients having deletions of chromosome 22q11, this being the typical cause of DiGeorge syndrome.

4th arch

Margins of pericardial cavity

Distal outflow cushions

6th arch and left pulmonary artery

Figure 27  The section is from a human fetus at Carnegie stage 16, and shows the distal part of the outflow tract (bracket). The cushions will fuse with the dorsal wall of the aortic sac (star) close to the margins of the pericardial cavity, separating the initially common lumen into aortic and pulmonary channels.

Common arterial trunk, however, is not the only lesion producing single outlet from the heart. A single outlet is also produced in the setting of aortic or pulmonary atresia, the outlet then being through the complementary patent arterial trunk. In almost all such instances, these malformations represent change acquired during fetal development, and it is possible to diagnose the specific connections as being concordant, discordant, or double-outlet with atresia

272

PART  |  3  Patterning of the Early Heart Tube

of the arterial trunk that had initially been patent during development. These are the patterns typically seen in malformations such as hypoplasia of the left heart, usually seen with the atretic aorta arising from the small left ventricle,

Desc. aorta

Aorta Duct

Pulmonary valve

Left pulm. artery

Figure 28  The illustration shows an aortopulmonary window (yellow dotted lines), with the right pulmonary artery overriding the window, arising in its greater part from the aorta. Note that there is also interruption of the aortic arch (star), the descending (desc.) aorta fed through the persistently patent arterial duct.

Common arterial trunk

Pulmonary valve

hypoplasia of the right heart with an intact ventricular septum, which is normally found with the atretic pulmonary trunk connected to the right ventricle, or tetralogy of Fallot when found with pulmonary atresia rather than pulmonary stenosis, the atretic pulmonary trunk arising ventrally from the right ventricle. It remains the case, nonetheless, that occasionally it may not be possible to trace the atretic trunk to its ventricular origin with certainty. In such circumstances, therefore, it is better to describe the arrangement as solitary aorta with pulmonary atresia, or solitary pulmonary trunk with aortic atresia. There remains one situation when even this description proves unsatisfactory. This is found when the intrapulmonary arteries are completely lacking, either because the supply to the lungs is derived through systemic to pulmonary collateral arteries, or because the pulmonary arteries themselves are discontinuous, being supplied on each side by a persistently patent arterial duct. Alternatively, there can be a combination of these findings, with the pulmonary blood supply to one lung derived from systemic-to-pulmonary collateral arteries and to the other side via a persistently patent duct, which often becomes ligamentous (Fig. 30A). In all of these situations, a solitary trunk leaves the base of the heart (Fig. 30B). There is no way of knowing, in this setting, had the intrapulmonary arteries developed, whether they would have taken their origin from the heart, in which case the patent arterial trunk would have been an aorta, or from the trunk itself, which would then have been a common structure. Thus, the only way to describe this arrangement accurately is to label the trunk as being solitary. Mention has been made in the preceding discussion of systemic-to-pulmonary collateral arteries. It is now wellestablished that, in the setting of tetralogy of Fallot with pulmonary atresia, these arteries can supply a significant proportion, if not all, of the blood to the lungs. They can also supplement the pulmonary arterial supply in tetralogy with pulmonary stenosis, albeit they can then prove a significant Fibrous raphe

Separate aortic valve

Subarterial interventricular communication

Subarterial interventricular communication

(A)

(B)

Figure 29  These illustrations show the remarkable morphologic similarity between common arterial trunk (A) and the doubly committed variant of ventricular septal defect (B). The difference is that, in the common trunk, the entirety of the outflow tract is a common structure, whereas the doubly committed defect has separate aortic and pulmonary roots and valves.

Chapter  |  3.4  A System for Describing Congenital Cardiac Malformations

nuisance during surgical repair. When they are found in tetralogy with pulmonary atresia, it is almost always necessary for the collateral arteries to be reunited with the intrapericardial pulmonary arteries. This process is known as unifocalization. The conundrum faced by the clinician during diagnosis, therefore, is to establish the proportion of the pulmonary parenchyma fed by collateral arteries as opposed to intrapericardial pulmonary arteries, and to establish the location of anastomoses between the two circulations, which can occur at hilar, lobar, or segmental levels (Rossi et al., 2002). A different conundrum faces the developmental biologist, namely the origin of the arteries. It has yet to be established whether the collateral arteries are persisting connections between the descending aorta and the lungs, or bronchial arteries recruited to supply pulmonary arterial blood in the setting of atresia of the intrapericardial pulmonary arterial supply. Additional study of animal models, such as those which resemble tetralogy of Fallot with pulmonary atresia (Jackson et al., 1995), could clarify this unsolved question.

VII.  Cataloging the associated malformations All the malformations we have discussed thus far either represent lesions resulting from abnormal lateralization (as seen in isomerism), or abnormal connections across the atrioventricular or ventriculo–arterial junctions. We have emphasized these lesions, since description of the segmental combinations provides the “flow template” of any given heart. Having established the segmental arrangements, it is essential to

273

identify and describe each and every associated malformation. In many instances the associated malformation will be the only abnormality, since in the majority of cases congenital cardiac lesions are found in the setting of an otherwise normally-structured heart. It could be, however, that any given combination of anomalies may never have been encountered previously. Providing they are described and cataloged as discussed thus far, the uniqueness will not pose problems since the combination will be readily understandable. It is our practise to list anomalous pulmonary or systemic venous connections as associated malformations, but it could be argued that the venous channels should be considered as an additional cardiac “segment.” We find it much easier simply to list these entities, such as persistent left superior caval vein draining to the coronary sinus, interruption of the inferior caval vein with continuation through the azygos venous system, or the many variants of anomalous pulmonary venous connection. Within the atrial segment, we also list lesions such as interatrial communications, or divided atriums. In terms of interatrial communications, only those deficiencies within the confines of the oval fossa represent true septal defects (Anderson et al., 1999). Several other holes permit interatrial shunting. The sinus venosus defects are found in the mouths of either the superior or inferior caval veins, and require anomalous connection of one or more pulmonary veins to produce the interatrial communication (Ettedgui et al., 1992; Al Zaghal et al., 1997). Since the pulmonary veins do not migrate to the atrial roof until after the completion of atrial septation in the human, and since there is a solitary pulmonary venous orifice in the mouse located

Ligamentous duct to left lung

Aorta

Solitary arterial trunk

Systemic-to-pulmonary arteries feeding right lung

(A)

(B)

Absence of intrapericardial pulmonary arteries

Figure 30  The left-hand panel (A) shows the lungs as seen from behind from a patient diagnosed with tetralogy of Fallot with pulmonary atresia. The left pulmonary artery was fed initially through a persistently patent arterial duct which has become ligamentous. The right lung is supplied exclusively through two systemic-to-pulmonary collateral arteries. The right-hand panel (B) shows the arrangement at the arterial pole. There is a solitary arterial trunk with absence of the intrapericardial arteries. There is no way of knowing, had the intrapericardial pulmonary arteries developed, whether they would have originated from the heart or from the trunk itself. Thus, the only accurate description is a solitary arterial trunk.

274

PART  |  3  Patterning of the Early Heart Tube

adjacent to the atrioventricular junction (Webb et al., 2001), it is questionable whether the sinus venosus defect will be found in mutant mice. Another defect is found at the mouth of the coronary sinus when the walls that usually separate this channel from the left atrium as it courses through the left atrioventricular groove are absent. When it was thought that a “party wall” separated the coronary sinus from the left atrium, it was an easy matter to account for the formation of this lesion on the basis of fenestration of the wall. We now know that the sinus possesses its own walls discrete from the wall of the left atrium, so the morphogenesis of the coronary sinus must be much more complex (Knauth et al., 2002). The other hole that permits shunting at atrial level, but remains outside the confines of the oval fossa, is the co-called “ostium primum” defect. As we have already explained (Fig. 17), this is no more than an atrioventricular septal defect with common atrioventricular junction, but with the valvar leaflets attached to each other and also to the crest of the ventricular septum (Anderson et al., 1998). The very fact that the leaflets can be joined to each other in the setting of a common atrioventricular junction, however, calls into question the long-held notion that these lesions represent failure of fusion of the atrioventricular endocardial cushions. It is far more likely that the causative lesion lies in the mechanism of separation of the embryonic atrioventricular canal (Webb et al., 1997). At ventricular level, the most important malformation is a hole between the ventricles. In terms of morphology such holes, if assessed as seen from the morphologically right ventricle, can take only three forms. The most common is probably the one within the substance of the ventricular

Remnant of membranous septum

septum. Evidence is beginning to accrue that these are the consequence of impairment of formation of the compact layer during development, but this concept has yet to be proven. Be that as it may, many of these defects, at least in humans, close either during late fetal or early postnatal development. The most frequent defect seen in clinical practise is that which represents the embryonic interventricular communication. Its defining morphological feature is the presence of fibrous continuity between the leaflets of the aortic and tricuspid valves in its dorso–caudal border (Fig. 31A). The area of fibrous tissue incorporates the atrioventricular component of the membranous septum, and often a remnant of the interventricular membranous septum reinforces the border. Such defects, which are described as being perimembranous (Soto et al., 1980), can also exist in the setting of overriding arterial valves, but then it is important to judge the morphology as seen from the right ventricle (Baker et al., 1988), since if the defect is considered to represent the cranial contin­uation of the deficient ventricular septum, then even those defects having a muscular dorso–caudal rim (Fig. 31B) will be found to have fibrous continuity between the mitral and aortic valves in their dorso–caudal margins. They will not, however, exhibit fibrous continuity between the leaflets of the aortic and tricuspid valves. The third type of interventricular communication is the one positioned within the outflow tracts, and roofed by fibrous continuity between the leaflets of the aortic and pulmonary valves. In most instances, these defects, which are doubly committed and juxta-arterial, possess a muscular dorso–caudal rim as illustrated in Fig. 29B. They can, however, also be found when there is fibrous continuity

Muscle bar between tricuspid and aortic valvar leaflets

Aortic to tricuspid continuity Aortic to mitral continuity

(A)

Aortic to mitral continuity

(B)

Figure 31  These hearts are both photographed from the left ventricle, showing the difference in structure between defects which (A) are perimembranous, having fibrous continuity between the leaflets of the aortic and tricuspid valves, and (B) those which have a muscular dorso–caudal rim. Note that there is mitral to aortic fibrous continuity in both hearts.

275

Chapter  |  3.4  A System for Describing Congenital Cardiac Malformations

between the leaflets of the aortic and tricuspid valves, making them both perimembranous and doubly committed. In developmental terms, the doubly committed defect exists because of failure in the formation of the muscular subpulmonary infundibulum. Such defects cannot exist in the otherwise normally-structured heart, having more in common in developmental terms with common arterial trunk. The categorization described above, designed initially for use in the clinical setting, works equally well for animal models of congenital heart disease (Ho et al., 1991). Most other lesions involving the ventricles have already been discussed, such as atrioventricular septal defect with common atrioventricular junction. Lesions involving the atrioventricular or arterial valves in isolation are readily recognized and can produce either stenosis or regurgitation, or a combination of the two. Other miscellaneous lesions, such as division of the ventricles to produce so-called “two-chambered” arrangements, are equally obvious and should give no trouble in recognition, albeit the substrate producing the partition within the ventricle is not always clear (Alva et al., 1999). Many of the lesions involving the outflow tracts have also been discussed. These lesions are often lumped together as “conotruncal malformations.” Since there is currently no agreement as to what constitutes the embryo­ nic conus or truncus, or which lesions should be placed

Esophagus

in the overall grouping, we find it much easier simply to describe malformations of the outflow tracts. These can then be considered in terms of those reflecting abnormal development of the proximal parts of the outflow tracts, such as the doubly committed and juxta-arterial interventricular communication, those involving the arterial valves and their supporting sinuses, such as valvar stenosis, or aneurysm of the sinuses of Valsalva, those involving the intrapericardial arterial trunks, such as aorto–pulmonary window or anomalous origin of the right pulmonary artery from the aorta, and those involving the entirety of the outflow tract, as exemplified by common arterial trunk. Then there are the lesions that involve the extrapericardial arterial pathways, such as aortic coarctation, rings and slings, and persistent patency of the arterial duct. It is beyond the scope of this review to give details of all these lesions, but suffice it to say that simple description provides the necessary information in all instances. Understanding of vascular rings, such as double aortic arch and all its variants, nonetheless, is greatly facilitated by an appreciation of the hypothetical double aortic arch proposed by Edwards and his colleagues (Edwards, 1953) (Fig. 32). In closing our account of associated malformations, we should emphasize again that we have made no mention of “dextrocardia” and its variants. This is because the position of the heart is not a malformation in itself. We describe an

Neutral descending aorta Trachea

Right subclavian artery

Left subclavian artery

Right carotid artery Right duct

Left carotid artery Left duct

Aorta

Pulmonary trunk

Figure 32  The cartoon shows the structure of the hypothetical double arch proposed by Edwards and his colleagues to provide an explanation of the many and varied vascular rings. Each arch gives rise to an arterial duct, a common carotid and a subclavian artery. The descending aorta is in the midline. The double arch can be broken at any of the points shown by the green lines, flow occurring through the persisting parts of the system. Thus far, the hypothesized arrangement has served its purpose, explaining all known malformations of the aortic arches.

276

abnormal cardiac location, taking note of first its position within the chest, and then the direction of its apex, since these two features are mutually independent. The normal heart can be positioned in the right side of the chest simply because of problems with other thoracic structures. It is also frequently stated that looping of the heart tube is the first instance of breaking symmetry in the developing individual. This is also untrue. In patients with isomerism, the direction of ventricular looping, and hence the resulting ventricular topology, is a random event. It is the atrial component of the heart which exhibits loss of symmetry, specifically with the formation of discrete morphologically right and left atrial appendages, and it is the structure of these appendages which serves as the distinguishing feature when arbitrating between rightness and leftness in the atrial segment of the heart.

VIII.  Conclusion In compiling our review, we have sought to present a simple means of describing congenital malformations anywhere within the heart. The system described is just as easy to use for complex malformations as for simple ones. It requires no prior knowledge of development, but use of the system will greatly facilitate the analysis of cardiac malformations increasingly encountered in mutant animals such as mice and rats. The system begins by analyzing the structure of the atrial appendages, recognizing that in most, but not all, instances the arrangement of these structures is in harmony with that of the other lateralized organs of the body. Whenever additional useful information is available, it is used. Having established the atrial arrangement, the next step is to analyze the atrioventricular junctions, taking note of ventricular topology and both the type and mode of atrioventricular connections. The same process is then repeated for the ventriculo–arterial junctions, in this instance taking note of connections, infundibular morph­o­logy and relationships of the arterial trunks. Only then, having established the cardiac template, is a catalog made of all the associated malformations, including information on abnormal cardiac position should this be encountered. In this way, a simple account can be provided for each and every malformation, even if they have never previously been encountered.

References Abbott, M. 1936. Atlas of Congenital Cardiac Disease. American Heart Association, New York, NY. Al Zaghal, A.M., Li, J., Anderson, R.H., Lincoln, C., Shore, D., Rigby, M.L., 1997. Anatomical criteria for the diagnosis of sinus venosus defects. Heart 78, 298–304. Alva, C., Ho, S.Y., Lincoln, C.R., Wright, A., Anderson, R.H., 1999. The nature of the obstructive muscular bundles in double-chambered right ventricle. J. Thorac. Cardiovasc. Surg. 117, 1180–1189.

PART  |  3  Patterning of the Early Heart Tube

Anderson, R.H., 2000. Nomenclature and classification: sequential segmental analysis. In: Moller, J.H., Hoffman, J.I.E. (Eds.), Pediatric Cardiovascular Medicine, Churchill Livingstone, New York, NY, pp. 263–274. Anderson, R.H., Ho, S.Y., 1997. Sequential segmental analysis – description and categorization for the millenium. Cardiol. Young 7, 98–116. Anderson, R.H., Brown, N.A., 2000. The anatomy of the heart revisited. Anat. Rec. 246, 1–7. Anderson, R.H., Macartney, F.J., Tynan, M., Becker, A.E., Freedom, R.M., Godman, M.J., Hunter, S., Quero Jimenez, M., Rigby, M.L., Shinebourne, E.A., Sutherland, G., Smallhorn, J.F., Soto, B., Thiene, G., Wilkinson, J.L., Wilcox, B.R., Zuberbuhler, J.R., 1983. Univentricular atrioventricular connection: the single ventricle trap unsprung. Pediatr. Cardiol. 4, 273–280. Anderson, R.H., Becker, A.E., Freedom, R.M., Macartney, F.J., QueroJimenez, M., Shinebourne, E.A., Wilkinson, J.L., Tynan, M., 1984. Sequential segmental analysis of congenital heart disease. Pediatr. Cardiol. 5, 281–288. Anderson, R.H., Ho, S.Y., Falcao, S., Daliento, L., Rigby, M.L., 1998. The diagnostic features of atrioventricular septal defect with common atrioventricular junction. Cardiol. Young 8, 33–49. Anderson, R.H., Silverman, N.H., Zuberbuhler, J.R., 1990. Congenitally unguarded tricuspid orifice: its differentiation from Ebstein’s malformation in association with pulmonary atresia and intact ventricular septum. Pediatr. Cardiol. 11, 86–90. Anderson, R.H., Webb, S., Brown, N.A., 1999. Clinical anatomy of the atrial septum with reference to its developmental components. Clin. Anat. 12, 362–374. Anderson, R.H., Webb, S., Lamers, W., Moorman, A., 2003a. Development of the heart: (2) Septation of the atriums and ventricles. Heart 89, 949–958. Anderson, R.H., Webb, S., Brown, N.A., Lamers, W., Moorman, A., 2003b. Development of the heart: (3) Formation of the ventricular outflow tracts, arterial valves, and intrapericardial arterial trunks. Heart 89, 1110–1118. Baker, E., Leung, M.P., Anderson, R.H., Fischer, D.R., Zuberbuhler, J.R., 1988. The cross-sectional anatomy of ventricular septal defects: a reappraisal. Br. Heart J. 59, 339–351. Bamforth, S.D., Bragança, J., Farthing, C.R., Schneider, J.E., Broadbent, C., Michell, A.C., Clarke, K., Neubauer, S., Norris, D., Brown, N.A., Anderson, R.H., Bhattacharya, S., 2004. Cited2 controls left–right patterning and heart development through a Nodal-Pitx2c pathway. Nat. Genet. 36, 1189–1196. Brandt, P.W., Calder, A.L., Barratt-Boyes, B.G., Neutze, J.M., 1976. Double outlet left ventricle. Morphology, cineangiography, diagnosis and surgical treatment. Am. J. Cardiol. 38, 897–909. Cavalle-Garrido, T., Bernasconi, A., Perrin, D., Anderson, R.H., 2007. Hearts with concordant ventriculo–arterial connections, but with parallel arterial trunks. Heart 93, 100–106. Edwards, J.E., 1953. Malformations of the aortic arch system manifested as “vascular rings”. Lab. Invest. 2, 56–75. Ettedgui, J.A., Siewers, R.D., Zuberbuhler, J.R., Anderson, R.H., 1992. Echocardiographic diagnosis of inferior sinus venosus defects. Cardiol. Young 2, 338–341. Ho, S.Y., Goltz, D., McCarthy, K., Cook, A.C., Connell, M.G., Smith, A., Anderson, R.H., 2000. The atrioventricular junctions in Ebstein’s malformation. Heart 83, 444–449. Ho, S.Y., Milo, S., Anderson, R.H., Macartney, F.J., Goodwin, A., Becker, A.E., Wenink, A.C.G., Gerlis, L.M., Wilkinson, J.L., 1982. Straddling atrioventricular valve with absent atrioventricular connection. Report of 10 cases. Br. Heart. J. 47, 344–352.

Chapter  |  3.4  A System for Describing Congenital Cardiac Malformations

Ho, S.Y., Thompson, R.P., Gibbs, S.R., Swindle, M.M., Anderson, R.H., 1991. Ventricular septal defects in a family of Yucatan miniature pigs. Int. J. Cardiol. 33, 419–426. Jackson, M., Connell, M.G., Smith, A., Drury, J., Anderson, R.H., 1995. Common arterial trunk and pulmonary atresia: close developmental cousins? Results from a teratogen induced animal model. Cardiovasc. Res. 30, 992–1000. Jacobs, J.P., Franklin, R.C., Colan, S.D., Jacobs, M.L., Tchervenkov, C.L., Maruszewski, B., Gaynor, J.W., Spray, T.L., Stellin, G., Aiello, V.D., Beland, M.J., Krogmann, O.N., Kurosawa, H., Weinberg, P.M., Elliott, M.J., Mavroudis, C., Anderson, R.H., 2006. Classification of the functionally univentricular heart: unity from mapped codes. Cardiol. Young 16 (Suppl. 1), 9–21. Jacobs, M.L., Anderson, R.H., 2006. Nomenclature of the functionally univentricular heart. Cardiol. Young 16 (Suppl. 1), 3–8. Knauth, A., McCarthy, K.P., Webb, S., Ho, S.Y., Allwork, S.P., Cook, A.C., Anderson, R.H., 2002. Interatrial communication through the mouth of the coronary sinus. Cardiol. Young 12, 364–372. Merrick, A.F., Yacoub, M.H., Ho, S.Y., Anderson, R.H., 2000. Anatomy of the muscular subpulmonary infundibulum with regard to the Ross procedure. Ann. Thorac. Surg. 69, 556–561. Milo, S., Ho, S.Y., Macartney, F.J., Wilkinson, J.L., Becker, A.E., Wenink, A.C.G., Gittenberger-de Groot, A.C., Anderson, R.H., 1979. Straddling and overriding atrioventricular valves morphology and classification. Am. J. Cardiol. 44, 1122–1134. Moorman, A.F., Christoffels, V.M., 2003. Cardiac chamber formation: development, genes and evolution. Physiol. Rev. 83, 1223–1267. Moorman, A.F. , Christoffels, V.M., Anderson, R.H., van den Hoff, M. B., 2007. The heart-forming fields: one or multiple? Phil. Trans. Roy. Soc. 362, 1257–1265. Rinne, K., Smith, A., Ho, S.Y., 2000. A unique case of ventricular isomerism. Cardiol. Young 10, 42–45. Rossi, R.N., Hislop, A., Anderson, R.H., Maymone Martins, F., Cook, A.C., 2002. Systemic-to-pulmonary blood supply in tetralogy of Fallot with pulmonary atresia. Cardiol. Young 12, 373–388. Seo, J.-W., Brown, N.A., Ho, S.Y., Anderson, R.H., 1992. Abnormal laterality and congenital cardiac anomalies. Relations of visceral and cardiac morphologies in the iv/iv mouse. Circulation 86, 642–650. Shinebourne, E.A., Macartney, F.J., Anderson, R.H., 1976. Sequential chamber localization – the logical approach to diagnosis in congenital heart disease. Br. Heart J. 38, 327–340. Soto, B., Becker, A.E., Moulaert, A.J., Lie, J.T., Anderson, R.H., 1980. Classification of ventricular septal defects. Br. Heart J. 43, 332–343.

277

Tynan, M.J., Becker, A.E., Macartney, F.J., Quero-Jimenez, M., Shinebourne, E.A., Anderson, R.H., 1979. Nomenclature and classification of congenital heart disease. Br. Heart J. 41, 544–553. Uemura, H., Ho, S.Y., Devine, W.A., Anderson, R.H., 1995. Analysis of visceral heterotaxy according to splenic status, appendage morpho­ logy, or both. Am. J. Cardiol. 76, 846–849. Uemura, H., Ho, S.Y., Devine, W.A., Kilpatrick, L.L., Anderson, R.H., 1995. Atrial appendages and venoatrial connections in hearts with patients with visceral heterotaxy. Ann. Thorac. Surg. 60, 561–569. van den Hoff, M.B., Moorman, A.F., Ruijter, J.M., Lamers, W.H., Bennington, R.W., Markwald, R.R., Wessels, A., 1999. Myocardial­ ization of the cardiac outflow. Dev. Biol. 212, 477–490. Van Mierop, L.H., 1970. Transposition of the great arteries. Clarification or further confusion? Am. J. Cardiol. 28, 735–738. Van Mierop, L.H., Gessner, I.H., Schlieber, G.L., 1972. Asplenia and polysplenia syndromes. In: Bergsma, D. (Ed.), Birth Defects. Original Article Series, vol. VIII. Williams and Wilkins, Baltimore, MD, pp. 74–82. Van Praagh, R., 1970. Transposition of the great arteries. II Transposition clarified. Am. J. Cardiol. 28, 739–741. Van Praagh, R., 1972. Segmental approach to diagnosis in congenital heart disease. In: Bergsma, D. (Ed.), Birth Defects. Original Article Series, vol. VIII. No. 5. Williams and Wilkins, Baltimore, MD, pp. 4–23. Van Praagh, R., 1976. The story of anatomically corrected malposition of the great arteries. Chest 69, 2–4. Van Praagh, R., 2000. Nomenclature and classification: morphologic and segmental approach to diagnosis. In: Moller, J.H., Hoffman, J.I.E. (Eds.), Pediatric Cardiovascular Medicine. Churchill Livingstone, New York, NY, pp. 275–288. Van Praagh, R., David, I., Wright, G.B., Van Praagh, S., 1980. Large RV plus small LV is not single RV. Circulation 61, 1057–1058. Webb, S., Brown, N.A., Anderson, R.H., 1997. Cardiac morphology at late fetal stages in the mouse with trisomy 16: consequences for different formation of the atrioventricular junction when compared to humans with trisomy 21. Cardiovasc. Res. 34, 515–524. Webb, S., Kanani, M., Anderson, R.H., Richardson, M.K., Brown, N.A., 2001. Development of the human pulmonary vein and its incorporation in the morphologically left atrium. Cardiol. Young 11, 632–642. Webb, S., Qayuum, S.R., Anderson, R.H., Lamers, W.H., Richardson, M.K., 2003. Septation and separation within the outflow tract of the developing heart. J. Anat. 202, 327–342. Ya, J., van den Hoff, M.J., de Boer, P.A., Tesink-Taekema, S., Franco, D., Moorman, A.F., Lamers, W.H., 1998. The normal development of the outflow tract in the rat. Circ. Res. 82, 464–472.

Chapter 4.1

Cardiac Left–Right Asymmetry Jeffrey D. Amack1 and H. Joseph Yost2 1

Department of Cell and Developmental Biology, SUNY Upstate Medical University, Syracuse, NY, USA Department of Neurobiology and Anatomy, University of Utah School of Medicine, Salt Lake City, UT, USA

2

I.  Introduction In the formation of the parallel systemic and pulmonary circulatory systems of the heart, left–right asymmetry is an essential aspect in determining both structure and function. It is becoming clear that traditional examples such as dextrocardia, the misplacement of the heart to the right, or heterotaxia, the alteration of heart orientation with respect to other organs, are only a subset of left–right cardiac defects. More commonly, subtle defects result in misalignment of the inflow and outflow tracts of the heart, resulting in a significant spectrum of congenital defects. Left–right patterning involves over 100 genes that encode components of several major cell–cell signaling pathways, as well as transcription factors and proteins required for formation of cilia. Strikingly, several embryonic tissues outside the cardiogenic fields are essential for normal cardiac left–right patterning, including cells in the embryonic node, caudal mesoderm, lateral plate mesoderm, notochord and ventral floorplate of the neural tube. Thus, defects in cardiac left– right development can often be part of a syndrome that also presents with cilia defects or subclinical defects in midline or neural tube patterning. Much of our current understanding of left–right development has come from embryological and genetic studies in model organisms that have left–right asymmetric organ orientation, or laterality, that is similar to humans. Here we review mechanisms that influence left–right asymmetry in vertebrate heart development.

II.  Overview of cardiac left–right development All vertebrates share analogous steps during early heart development, from bilateral cardiac primordia to asymmetric looping of the cardiac tube (Icardo, 1988; Lohr and Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

Yost, 2000; Wagner and Siddiqui, 2007). Differences in the geometry of the embryos, for example spherical Xenopus embryos compared with discoid avian embryos, result in distinct spatial positioning of the cardiac primordia in the embryo, but the general steps in cardiac morphogenesis appear to be similar (Fig. 1). Presumptive cardiac cells arise as a pair of lateral mesoderm primordia at the beginning of gastrulation. Explant and tissue recombination experiments in Xenopus have shown that signals derived from the dorsal midline mesoderm and endoderm cells induce dorsolateral mesoderm cells to become precardiac mesoderm cells (Sater and Jacobson, 1989, 1990; Muslin and Williams, 1991; Nascone and Mercola, 1995; see Chapter 1.3). Molecular studies in several vertebrates have identified a complex mix of positive signals that induce cardiogenic cell fate, and negative signals that likely set the boundaries of the cardiac lineage (reviewed in Harvey, 2002; Brand, 2003; see Chapters 5.1, 9.1). Several well-studied pathways have been implicated in this process, including components of TGF, BMP, FGF8, Wnt, Hedgehog and Serrate/Notch signaling cascades (see Chapters 4.2, 4.3). The interactions between these pathways are thought to control cardiac induction by regulating expression of heart identity genes, such as members of the Tbx, Nkx and Gata families of transcription factors. During gastrulation, the precardiac mesoderm involutes with other axial mesoderm to take up a dorsoanterior position in the embryo. Left–right asymmetric signals are present during gastrula stages, at least in Xenopus, so it is possible that prospective heart cells acquire left–right identities during their migration. In Xenopus, results with pharmacological inhibitors of PKC and dominant negative Syndecan-2 proteins that can be destroyed (by proteolytic cleavage of an engineered site in the extracellular domain) at specific stages of development suggest that asymmetric phosphorylation of Syndecan-2 transmits left–right information from the ectoderm to the migrating mesoderm during 281

282

(A)

PART  |  4  Asymmetry in Cardiac Development

(B)

A

(E)

(D)

(C)

a v R

v

L v

a

a P

a v Adult heart

Figure 1  Early steps of vertebrate heart development. During gastrula stages presumptive cardiac cells, induced in lateral mesoderm, form bilateral primary heart fields (A). These heart fields then fuse at the midline to form a symmetric cardiac tube (B). Next, the heart tube initiates rightward looping (C), which positions the future atria (a) dorsal and anterior to the prospective ventricles (v) and is essential for normal alignment of chambers, valves and septa (D) (A–P: anterior–posterior axis; L–R left–right axis).

gastrulation (Kramer et al., 2002; Kramer and Yost, 2002). At the end of gastrulation, precardiac cells are positioned in anterior lateral mesoderm and have formed bilateral primary heart fields separated by the embryonic midline (Fig. 1A). Next, the paired heart primordia fuse at the midline in an anterior-to-posterior progression to form a linear cardiac tube. In addition to the primary heart fields, mesodermal cells from a recently-described second (or anterior) heart field also contribute to the heart tube (Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001; see Chapters 2.2 and 3.1). The cardiac tube becomes morphologically divided into distinct regions along the anterioposterior axis that will give rise to the atria and ventricles, but remains relatively symmetric along the left–right axis (Fig. 1B). This apparent bilateral symmetry is broken by rightward looping of the heart tube (Fig. 1C), which occurs at approximately 30–36 hours post fertilization in zebrafish, 44–50 hours in Xenopus, 8.5 days in mice and 23 days in humans. The first known indication of cardiac left–right asymmetry is observed in the heart field well before cardiac looping, appearing as asymmetric gene expression. Several genes are expressed exclusively in either the left or right heart field (discussed further below). This molecular asymmetry is thought to impart “left-sidedness” and “right-sidedness” by controlling expression of target genes that subsequently carry out asymmetric heart morphogenesis. Consistent with this, left–right determination in the Xenopus heart occurs when the primordia is a simple sheet of mesoderm, shortly after fusion of the paired heart fields at the ventral midline (Danos and Yost, 1995; Chapter 1.4). Explanted cardiac primordia form cardiac tubes that loop in vitro. Explants made at the end of gastrulation have randomized orientation of cardiac looping. However, explants made a few hours later, as the neural tube closes, form a cardiac tube that loops in the normal left–right orientation in vitro. Thus, the cardiac primordium has established left– right identities by the time the neural tube closes. Similarly, bmp4 and pitx2 genes are asymmetrically expressed in the zebrafish heart tube before left–right morphogenesis (Chen et al., 1997; Campione et al., 1999; Essner et al., 2000;

Chapter 4.3). In zebrafish and mice the cardiac tube appears to rapidly and transiently shift to the left prior to rightward looping (Biben and Harvey, 1997; Chen et al., 1997). In zebrafish, the cardiac tube then shifts back to a central position before looping to the right. This transient shift has been termed cardiac “jogging”, and appears to be highly predictive of the subsequent orientation of cardiac looping in zebrafish (Chen et al., 1997; see Chapter 1.4). Cardiac looping, together with heart tube elongation, place the prospective atrium derived from the posterior (caudal) portion of the cardiac tube dorsal and anterior (cranial) to the prospective ventricle (Fig. 1D). By this stage the heart is circulating blood, and valve formation and septation are underway. In addition to controlling the direction of looping, evidence from mouse mutants indicates left–right information is necessary for several other aspects of heart development. For example, targeted disruption of Pitx2, a transcription factor involved in left–right development (discussed further below), does not alter the direction of heart looping in the mouse embryo but still results in a broad spectrum of cardiac defects (Kitamura et al., 1999; Lin et al., 1999; Lu et al., 1999; Liu et al., 2001; Shiratori et al., 2006). These include randomized positioning of the apex, atrial isomerism and valve and septal defects. It is not clear whether these defects are independent of looping, since the process of looping is important for proper valve and septa development (Ramsdell, 2005). It is possible that even if the direction of heart looping is normal (as in Pitx2 mutants) subtle defects in looping morphogenesis disrupt subsequent steps of heart development (see Chapter 4.3).

III.  Left–right nomenclature It is useful to make a distinction between “left–right axis formation” and “left–right development”. First, left–right axis formation refers to an early, embryo-wide (global) mechanism that establishes a difference between left body side and right body side, and directs the subsequent left– right asymmetric morphogenesis of organ systems in a

283

Chapter  |  4.1  Cardiac Left–Right Asymmetry

coordinated fashion. From an anatomical perspective, the left–right axis is orthogonal to the other body axes, rostral– caudal (anterior–posterior in embryology) and dorsal– ventral. The use of the term left–right axis formation should be limited to the initial mechanisms that establish global left–right asymmetries across an otherwise bilaterally symmetric body plan. Second, left–right development is the reception and playing out of this global axial information by individual organ primordia, such as groups of cells that form the heart, segments of the gut or brain. As outlined below, there are four outcomes of embryological manipulations or genetic disruptions that alter left–right development, each of which is consistent with having altered either left–right axis formation or the response of individual primordia to left–right axis information.

of the viscera in inversus viscerum (iv) mutant mice (Seo et al., 1992). These experimental conditions are reminiscent of humans affected with situs ambiguus, also referred to as visceral heterotaxy or heterotaxia (greek meaning “other arrangement”) (Kosaki and Casey, 1998). In contrast to individuals affected with situs inversus, patients with situs ambiguus suffer from severe health problems, including a high frequency of complex congenital heart defects. Cardiac defects associated with situs ambiguus, which arise from more subtle errors in left–right morphogenesis, include atrial and ventricular septal defects, transposition of the great arteries, and double-outlet right ventricle (Ramsdell, 2005).

III.A.  Inversion

A third outcome is randomization of the left–right orientation of an individual organ, without alteration of the normal orientation of surrounding organs, and perhaps without the alteration of normal asymmetric expression patterns of upstream genes. This is seen as a result of experimental manipulation of small patches of extracellular matrix which individual primordia migrate across in Xenopus embryos (Yost, 1992) and in human cases of isolated laterality defects, such as isolated cardiac reversals (Bowers et al., 1996). In these cases, it is likely that left–right axis formation and the production of global left–right signals are normal, but that the affected organ primordium has either lost its ability to receive left–right signals or its ability to use those signals during organ morphogenesis. Although there are fewer examples of this situation in model organisms, genetic screens to identify mutants with isolated cardiac laterality defects are likely to uncover genetic pathways that are downstream of left–right axis formation, but upstream of the mechanism that biases the direction of cardiac tube looping.

In rare situations, left–right axis information can be inverted, so that genes normally expressed on the left are now expressed only on the right, and all organs are the mirror image of normal. This biological enantiomer of normal organ orientation (situs solitus) is known as situs inversus. Interestingly, a complete, concordant reversal of organ laterality carries little clinical consequence in humans (Bisgrove et al., 2003). This indicates organs can function normally in reversed orientation, as long as their structure and position relative to each other remains intact. Several embryological manipulations and genetic mutations that disrupt left–right axis formation and early steps of left– right development (discussed below) have reproduced situs inversus in model organisms.

III.B.  Global Randomization Elimination of normal left–right axis information throughout the embryo leads to global randomization of the orientation of each organ with respect to the orthogonal axes (dorsoventral and anterioposterior) and with respect to each other. In the absence of left–right axis information, many organ primordia appear to retain an intrinsic ability to generate left–right asymmetry, but the orientation of asymmetric organs is no longer concordant. This suggests that left–right axis information is capable of strongly biasing the orientation of asymmetric development in organ primordia, but is not necessary for organ asymmetry per se. For experimental manipulations that appear to eliminate left–right axis information, individual embryos can have a heart in either normal, reversed or indeterminate orientation and intestinal coiling in normal, reversed or indeterminate orientation, with no correlation between the orientation of the heart and intestine in an individual embryo (Hyatt et al., 1996). Similarly, the orientation of the heart is not statistically concordant with the orientation

III.C.  Isolated Randomization

III.D.  Failure of Left–Right Morphogenesis A fourth outcome is failure to generate left–right asymmetry. The simplest example of this is a cessation of cardiac tube looping. For example, in Xenopus embryos treated with -xyloside to block proteoglycan synthesis during the neurula stages the cardiac tube fails to generate left–right asymmetry later in development (Yost, 1990). Embryological recombination of a wild-type cardiac primordium with a xyloside-treated cardiac primordium rescues looping in the chimeric heart tube, suggesting that the untreated cardiac primordium is capable of driving morphogenesis. Due to the developmental stages at which these manipulations have an effect, it is likely that these treatments block mechanisms of organ left–right morphogenesis rather than earlier steps of global left–right axis formation, although this has not been examined with molecular

284

markers. In humans, isomerisms, such as polysplenia syndrome and asplenia syndrome, are congenital disorders in which asymmetric organs (e.g., one spleen on the left) fail to show asymmetry (Capdevila et al., 2000). Polysplenia (multiple spleens) is considered a left isomerism and asplenia (absent spleen) is bilateral right-sidedness. Polysplenia and asplenia syndromes are almost always associated with cardiac laterality defects (Aylsworth, 2001).

IV.  Cardiac looping Asymmetric cardiac looping is an essential process that establishes relative spatial positions of the heart tube necessary for normal alignment of the chambers, valves and septa. Looping morphogenesis can be separated into two main steps (Manner, 2000; Taber, 2006). First, during c-looping (or dextral looping), the straight heart bends ventrally to form a c-shaped tube. There is a coordinated dextral (rightward) twist that places the ventral surface of the linear heart tube on the outer surface of the c-shaped heart. This looping event biases the position of the heart to the right of the midline. In the second step, s-looping brings the atrial and ventricular regions close together. These movements in combination with growth of the heart tube invert the initial alignment of the heart, such that the atrium is now (and permanently) superior to the ventricle. Although heart looping has been studied for decades, the mechanisms that underlie looping morphogenesis remain poorly-understood (Taber, 2006). However, these asymmetric morphogenetic events are likely controlled by asymmetries at the cellular and molecular levels. A handful of molecules have been identified that are asymmetrically distributed in the heart during looping, providing an entry point to further our understanding how the heart tube knows when and where to loop.

IV.A.  Mechanisms of Looping Several hypotheses have been put forth to explain the biomechanics of cardiac looping, but to date the mechanisms that control c-looping and s-looping morphogenesis are not well-defined. The observation that explanted cardiac tubes from frog (Yost, 1990; Danos and Yost, 1996) and chick (Manning and McLachlan, 1990) can loop in vitro indicates that this process is intrinsic to the heart tube. Consistent with this idea, embryological manipulations have ruled out spatial constraints (Manning and McLachlan, 1990), extracellular matrix of the cardiac jelly (Baldwin and Solursh, 1989; Linask et al., 2003) and myocardial contraction or blood flow (Manasek and Monroe, 1972; Sehnert et al., 2002) as looping determinants. However, in vitro looping only involves bending of the heart tube. Twisting or rotational movements during looping are influenced by extrinsic forces, such as pushing from the splanchnopleure,

PART  |  4  Asymmetry in Cardiac Development

a tissue adjacent to the heart tube (Voronov and Taber, 2002; Voronov et al., 2004). It has been recently proposed that multiple and redundant mechanisms are involved in ensuring proper heart looping (Taber, 2006). One can imagine several morphogenetic mechanisms by which a relatively symmetric tube of cells breaks symmetry. First, an asymmetric accumulation of cells on one side of the tube would cause it to bend. This could be accomplished by differential cell proliferation or cell death, coordinated cell movements or differential cell contributions from the two sides of the embryo. However, experiments in chick have found no significant regional differences in cell division or cell death capable of bending the heart tube (Sissman, 1966; Stalsberg, 1969b). In addition, there are no reports of directed cell movements within the cardiac tube during looping. In amphibians and chick, fate-map studies suggest that there is a differential contribution of cells from the left and right sides along the length of the cardiac primordium (Stalsberg, 1969a), but these findings have been associated with a slight rightward bend, rather than robust ventral bending (Voronov et al., 2004). More recently, cell lineage analysis in Xenopus has shown that particular regions of the heart are made up of myocytes unilaterally derived from either the left or right side of the embryo, and that these left–right cardiomyocyte lineages are altered in embryos with left–right defects (Ramsdell et al., 2006). As an alternative to an asymmetric distribution of cells, looping could be driven by changes in cell size or shape. Treatment of chick embryos with actin polymerization inhibitors cytochalasin B, cytochalasin D and latrunculin A blocks cardiac looping (Itasaki et al., 1991; Latacha et al., 2005), implicating actin polymerization and the cytoskeleton in cardiac loop morphogenesis. It has recently been proposed that differential polymerization of the actin cytoskeleton mediates bending the heart tube (Taber, 2006), perhaps by generating flattened cells at the outer curvature of the loop and constricted cells in the inner curvature. Investigation of the cellular and molecular biology of cardiac looping will continue to be an exciting area of research.

IV.B.  Asymmetry in the Heart Tube One approach to understanding the mechanisms of cardiac looping is to identify and characterize genes that control this process. A few genes have been implicated in looping morphogenesis based on their left–right asymmetric expression pattern in the looping heart. These include a transcription factor (Pitx2), a signaling molecule (BMP4) and extracellular matrix proteins (flectin, hLAMP1 and JB3). Pitx2, a paired-related bicoid-type homeobox transcription factor, has received the most attention (see Chapter 4.3). This is likely due to its striking asymmetric expression pattern on the left side of the heart tube in all vertebrate embryos examined, including mouse,

Chapter  |  4.1  Cardiac Left–Right Asymmetry

chick, Xenopus and zebrafish (Logan et al., 1998; Ryan et al., 1998; Yoshioka et al., 1998; Campione et al., 1999; Essner et al., 2000). Asymmetric pitx2 expression persists throughout heart looping stages, which led to the hypothesis that Pitx2 may regulate target genes that mediate looping morphogenesis. Consistent with this idea, ectopic right-sided expression of Pitx2 in Xenopus (Campione et al., 1999; Essner et al., 2000) and chick (Logan et al., 1998; Ryan et al., 1998) resulted in a high frequency of reversed heart orientation (situs inversus) and cardiac isomerism. However, ectopic Pitx2 also caused additional morphological defects in the heart, suggesting Pitx2 plays a role in looping and other aspects of heart development. Recent evidence indicates that in addition to patterning the primary heart field, pitx2 is also asymmetrically expressed in the secondary heart field (Liu et al., 2002; Ai et al., 2006) that contributes to the outflow tract and right ventricle. These studies add to a growing body of evidence that Pitx2 plays multiple roles during cardiac development. In addition to regulating heart development, Pitx2 is involved in a number of other processes during embryogenesis. Haploinsufficiency of the human Pitx2 gene (RIEG) causes Axenfled-Reiger syndrome (ARS) which is characterized by eye and tooth developmental defects and in severe cases by heart defects (Semina et al., 1996; Lines et al., 2002). Heterozygous pitx2/ knockout mice recapitulate several aspects of ARS, whereas pitx2/-null mice have body wall closure defects, specific organ laterality defects and die before birth (Gage et al., 1999; Kitamura et al., 1999; Lin et al., 1999). Laterality defects include right pulmonary isomerism, gut malrotation and cardiac abnormalities associated with human heterotaxy syndromes such as septation defects, transposition of the great arteries, and double-outlet right ventricle. Surprisingly, however, heart looping directionality is normal in homozygous pitx2/ mice. This suggests that the handedness of cardiac looping (i.e., rightward versus leftward loop) is separable from other aspects of cardiac left–right development. Experiments testing different pitx2 alleles in mice have revealed that embryonic development, including organ laterality, is exquisitely sensitive to pitx2 gene dosage (Gage et al., 1999; Liu et al., 2001, 2003). Recent studies show that partial depletion (“knockdown”) of pitx2 in Xenopus embryos results in cardiac disturbances similar to those seen in mouse models, including septation and outflow tract defects (Dagle et al., 2003). To date, relevant Pitx2 left–right target genes in the heart have not been identified. Studies in cultured cells have implicated Pitx2 in cell proliferation and cell migration (Kioussi et al., 2002), but the roles of Pitx2 during cardiac development remain poorly understood. In zebrafish (Chen et al., 1997) and Xenopus (Breckenridge et al., 2001) expression of BMP4, a member of the TGF superfamily of cell signaling molecules, is initially symmetric in the cardiac tube and then becomes stronger on the left side. Zebrafish BMP4 asymmetry correlates

285

with the direction of normal and aberrant cardiac looping. In Xenopus, BMP4 asymmetry was shown to depend on Nodal signaling in the heart field. Furthermore, forced bilateral BMP4 expression randomized the direction of heart looping, whereas inhibition of BMP signaling halted looping morphogenesis. Interestingly, bilateral BMP4 in the heart altered looping morphogenesis without disrupting left-sided expression of pitx2 (Breckenridge et al., 2001). These lines of evidence indicate BMP4 signaling acts downstream of Nodal – and independently of Pitx2 – to mediate looping of the heart tube. Identification of molecular targets of this pathway will be of great interest and may aid our understanding of looping mechanisms. Three extracellular matrix proteins are expressed asymmetrically in the cardiac jelly of the looping heart; Flectin (Tsuda et al., 1996, 1998) and hLAMP1 (Smith et al., 1997) on the left and Fibrillin-related JB3 on the right (Smith et al., 1997). The expression of these molecules is responsive to a retinoid-sensitive pathway. Retinoic acid (RA) treatment on the right results in randomized cardiac orientation and altered hLAMP1 and JB3 expression (Chen and Solursh, 1992; Smith et al., 1997), whereas treatment on the left only does so at high concentrations. Recent findings indicate RA may be involved in asymmetric signaling in the early embryo (Tanaka et al., 2005), prior to the appearance of the cardiac primordia (discussed below). In quail embryos that are deficient in vitamin A, a precursor of retinoids, cardiac looping is perturbed and Flectin expression in the heart is disorganized (Tsuda et al., 1996). Furthermore, in chick, Flectin antibodies randomize heart looping, misexpression of Pitx2 can alter Flectin expression and sidedness of Flectin directly correlates with looping directionality (Linask et al., 2002, 2003). In contrast, asymmetric hLAMP1 expression was found to be unaltered in heart tubes that undergo reversed looping (Linask et al., 2003). While the asymmetry of these extracellular matrix proteins is exciting, much work remains to uncover how these molecules influence heart laterality.

V.  An asymmetric signaling cascade controls cardiac left–right development One of the most significant findings in cardiac left–right development is that the earliest molecular events that determine cardiac orientation do not involve the cardiac primordia. Left–right asymmetry is established during early embryogenesis by a complex signaling cascade that is thought to pattern the entire embryo left versus right. Thus, genes that are not expressed within prospective heart cells have a profound effect on the formation of the heart. In the mid-1990s several signaling molecules were shown to be asymmetrically expressed either on the left or right side of the chick embryo during gastrulation. Most of these

286

genes are not expressed in cardiac precursor cells, yet experimental alterations of the expression patterns of these genes can alter cardiac orientation (Levin et al., 1995). This provided evidence that “molecular left–right asymmetry” in the embryo precedes and controls morphological left–right development of organ primordia. Over the past decade, a large number of genes have been described in various species that are asymmetrically expressed along the left–right axis or that perturb left–right development when ectopically expressed or when mutated. Some of these genes appear to be specific to a particular species or class of vertebrates, while others, including members of the nodal, lefty and pitx2 gene families, participate in a conserved left–right signaling pathway found in all vertebrate embryos studied to date.

V.A.  Asymmetric Signaling in Chick:   A Role for the Node The initial breakthrough reports of molecular left–right asymmetry described asymmetric mRNA expression of several signaling factors along the left–right axis in chick embryos during gastrula and neurula stages, predominantly either in or near a structure called Hensen’s Node. Hensen’s Node (HN) in chick embryos is analogous to the Organizer in frog embryos and the node in mouse embryos. This group of cells is located at the midline and drives both gastrulation and the progressive formation of the embryonic anterior–posterior (cranial–caudal) axis. It is striking that most asymmetrically expressed genes around HN are not specifically “left–right genes”, but encode cell–cell signaling proteins that are involved in other embryonic patterning events. In many cases, genes and pathways implicated in left–right asymmetry play additional role(s) in other aspects of cardiac development. Based on experimental manipulations of embryos and developmental timing of asymmetric expression patterns, a cascade of left–right signals has been elucidated in the chick. This pathway is described in detail in several reviews (Burdine and Schier, 2000; Mercola and Levin, 2001; Wagner and Siddiqui, 2007) and summarized here (see Fig. 2). Activin B, a member of the TGF superfamily, is expressed on the right side of HN from stage 3 to stage 5 (Levin et al., 1997) and induces asymmetric BMP4 expression in these cells (Monsoro-Burq and Le Douarin, 2001). BMP4 inhibits sonic hedgehog (shh, a member of the hedgehog family of cell signaling peptides) and nodal (nodal-related gene 1) expression on the right side of HN (Monsoro-Burq and Le Douarin, 2001). BMP4 also acts to upregulate right-sided FGF8 expression, which in turn induces the snail-related zinc-finger gene (cSnR) in right LPM (Isaac et al., 1997; Boettger et al., 1999). On the left side of HN at stage 7, shh induces expression of the nodal gene, which encodes a secreted TGF signaling molecule. Importantly, Nodal

PART  |  4  Asymmetry in Cardiac Development

Activinβb BMP4

FGF8

Hensen’s Node

Nodal

Shh

Shh Nodal

Lefty cSnR

Lefty Pitx2

LPM

LPM Midline R

L

Figure 2  Asymmetric signaling cascade in the chick embryo. Asymmetric gene expression is initiated in a structure called Hensen’s Node. On the right side ActivinB upregulates BMP4, that inhibits Shh and induces FGF8. On the left side, unimpeded Shh activates Nodal expression. Nodal signals induce more Nodal expression, and the asymmetric Nodal domain spreads to lateral plate mesoderm (LPM; light gray box). In left LPM cells, Nodal activates Lefty, a feedback inhibitor of Nodal, and Pitx2 transcription factor. Opposite to Nodal, cSnR is expressed in right LPM. The embryonic midline structures, which express Lefty proteins, serve as a “barrier” to prevent the spread of Nodal expression to the right side of the embryo.

signaling activates nodal expression via an autoregulatory loop (Osada et al., 2000; Saijoh et al., 2000; Norris et al., 2002). After being initiated in a small left-sided domain, asymmetric nodal expression expands to a broad domain strictly in left lateral plate mesoderm (LPM), likely via a self-enhancement and lateral inhibition mechanism (Nakamura et al., 2006; see Chapter 4.2). Nodal ultimately spreads to the left heart field, where it activates asymmetric expression of target genes, including pitx2 (Logan et al., 1998), in cells that contribute to the heart tube. Experimental results indicate establishing asymmetric left-sided nodal is critical for normal heart laterality in chick. First, ectopic expression of shh on the right side of HN, which induces nodal expression in the right LPM (giving bilateral nodal expression), eliminates right-sided cSnR and results in randomized cardiac left–right orientation (Levin et al., 1995; Isaac et al., 1997). Similarly, adding nodal expression directly to the right side results in a higher frequency of bilaterally symmetric or inverted hearts (Levin et al., 1997). An important link between

287

Chapter  |  4.1  Cardiac Left–Right Asymmetry

Nodal asymmetry and heart looping was provided with the discovery that pitx2 is a target gene of Nodal signaling (Logan et al., 1998). In contrast to nodal, asymmetric expression of pitx2 persists in the cardiac tube. Thus, Pitx2 is thought to transmit left–right asymmetric Nodal signaling to the looping heart, although Pitx2 downstream response genes involved in asymmetric morphogenesis are unknown. Taken together, results from the chick embryo have provided a framework for a left–right signaling pathway in which the node is a source of asymmetric signals that establish distinct molecular left and right identities in LPM cells and subsequently in cardiac progenitor cells prior to the onset of asymmetric heart looping morphogenesis.

V.B.  A Conserved Nodal-Lefty-Pitx2   Pathway Results from chick and a number of other vertebrates suggest that the crux of the left–right asymmetric signaling cascade lies in the initiation, expansion and regulation of asymmetric Nodal signaling. While some genes initially characterized in chick may not play a conserved role during left–right development, it is now clear that there is a core left–right signaling pathway in vertebrates that involves nodal and its interactions with pitx2 and lefty genes (Fig. 2). To date, nodal genes have the earliest known conserved asymmetric expression pattern; chick nodal-related 1 (cNR1) (Levin et al., 1995), Xenopus nodal-related 1(Xnr-1) (Hyatt et al., 1996; Lowe et al., 1996), southpaw in zebrafish (Long et al., 2003) and medakafish (Soroldoni et al., 2007), and nodal in mouse (Collignon et al., 1996; Lowe et al., 1996) and rabbit (Fischer et al., 2002). In Xenopus (Sampath et al., 1997) and chick (Levin et al., 1997), ectopic Nodal alters cardiac laterality. In mice, it has not been possible to assess left–right phenotypes in the complete absence of nodal, because null embryos arrest prior to the onset of left–right development due to the requirement for nodal during early embryogenesis (Zhou et al., 1993). However, genetic evidence for the involvement of nodal in mouse left–right development has come from mice heterozygous for conditional mutant allele of nodal that show an absence of pitx2 expression and organ left–right defects (Lowe et al., 2001). More specifically, reduction of Nodal expression only in the node results in loss of asymmetric nodal and pitx2 expression in LPM and defects in organ left–right orientation (Brennan et al., 2002; Saijoh et al., 2003). In the zebrafish embryo, antisense morpholino oligonucleotides that knockdown expression of the nodal homolog, southpaw, eliminate asymmetric pitx2 expression and alter left–right asymmetry in the heart (Long et al., 2003). Thus, work from several model systems has led to the conclusion that an asymmetric signaling cascade in the left LPM in which Nodal activates target genes, such as pitx2 and lefty,

is essential for normal cardiac left–right morphogenesis in vertebrates.

V.C.  Role of the Midline Vertebrate embryos go to great lengths to ensure Nodal signaling occurs exclusively in the left LPM. As exemplified around HN in the chick, this requires a tightly-regulated cascade to correctly initiate nodal in left LPM. However, since Nodal can induce its own expression in neighboring cells, there must also be a way to prevent left-sided nodal expression from spreading to the right side of the embryo. This is the proposed function of the transient embryonic midline (notochord and floorplate of the neural tube) and the lefty genes (Fig. 2). Embryological manipulations in Xenopus that reduced the midline or extirpations that removed the midline altered cardiac left–right development (Danos and Yost, 1995, 1996). At the molecular level, these manipulations resulted in bilateral expression of nodal (Xnr-1) (Lohr et al., 1997), suggesting the embryonic midline serves as a “barrier” to keep left and right sides separate. Consistent with this hypothesis, zebrafish mutants with compromised midline integrity, including the no tail (Halpern et al., 1993) and floating head (Talbot et al., 1995) mutants that lack notochord, show cardiac laterality defects and bilateral expression of left–right markers (Danos and Yost, 1996; Bisgrove et al., 2000). In humans, analysis of families with situs ambiguus has identified mutations in the zinc-finger transcription factor gene ZIC3 (Gebbia et al., 1997), which is thought to play a role in midline development. In both Xenopus and mouse, altered expression of ZIC3 alters nodal asymmetry and results in organ laterality defects (Kitaguchi et al., 2000; Purandare et al., 2002). Are the cells in midline structures sufficient to separate left from right, or do proteins in the midline establish a “molecular barrier?” Lefty proteins are divergent TGB superfamily ligands that respond to and antagonize Nodal signaling (Meno et al., 1996, 1998, 1999; Bisgrove, 1999; Bisgrove et al., 1999; Cheng et al., 2000; see Chapter 4.2). Nodal establishes a negative feedback loop by inducing the expression of lefty genes. Lefty1 is expressed in the embryonic midline, in the notochord in frogs and zebrafish (Bisgrove et al., 1999; Thisse and Thisse, 1999; Branford et al., 2000; Cheng et al., 2000), and in ventral floorplate in mice (Meno et al., 1996, 1997). Knockout mice deficient for lefty1 show bilateral expression of nodal in LPM and perturbed organ left–right orientation (Meno et al., 1998). This suggests Lefty1 is a molecular component of a midline barrier that prevents contralateral expression of nodal. A second Lefty gene, lefty2, is induced by Nodal asymmetrically in left LPM, and is thought to act similarly to Lefty1 as a feedback inhibitor of Nodal in these cells (Meno et al., 1999). In zebrafish, another Nodal antagonist, Charon, has been proposed to prevent Nodal from spreading to right LPM cells at the caudal end of the midline (Hashimoto et al., 2004).

288

PART  |  4  Asymmetry in Cardiac Development

VI.  Left–right cilia and asymmetric fluid flow Key questions in cardiac left–right development center on the mechanisms upstream of molecular left–right asymmetry. What signals activate asymmetric gene expression? What is the initial break of bilateral symmetry in the embryo? Are the mechanisms conserved? The node/ Organizer structure has received much attention, since the site of asymmetric gene expression in LPM is at the same cranial–caudal level as the node. An elegant model, first proposed from studies of the mouse embryo, suggests symmetry is broken by an asymmetric fluid flow across the ventral node, generated by motile monocilia within the node (Fig. 3). This “nodal flow” hypothesis is supported by observations that mice with mutations affecting ciliogenesis or cilia motility have left–right defects (see Chapter 4.2). In addition, situs inversus is seen in 50% of patients affected with a form of primary cilia dyskinesia called Kartagener syndrome, indicating that loss of cilia function randomizes left–right axis specification in humans (Afzelius, 1976).

VI.A.  Discovery of Asymmetric Nodal Flow in Mouse In the mouse embryo, the node is a transient structure that forms at the rostral end of the midline during gastrulation. Cells on the ventral surface of the node form a pit and assemble monocilia. The microtubule-based monocilia protrude into the pit and have been shown by videomicrosopy to rotate rapidly in a clockwise direction (Nonaka et al., 1998; see Chapter 4.2). This rotational beating of cilia in the node is distinct from other motile cilia and flagella that beat back and forth in a whip-like fashion. To determine whether rotating node cilia could generate a coordinated flow, fluorescent microbeads were added to the extra-embryonic fluid in the ventral node pit of wild-type embryos in culture. It was discovered that the microbeads were swept asymmetrically toward the left side of the node by “nodal flow” (Nonaka et al., 1998). Embryos with an engineered mutation in the kinesin gene Kif3b that regulates ciliogenesis lacked node cilia, lacked nodal flow and presented defects in left–right asymmetric gene expression and heart laterality. Based on these results, the nodal flow hypothesis proposed that cilia-driven fluid flow breaks bilateral symmetry by sending left–right signals to the left side of the node to upregulate Nodal signaling in these cells, and subsequently in left LPM and organ primordia (Fig. 3). Initially, it was unclear how a rotating cilium could produce directional laminar fluid flow (Cartwright et al., 2004). An intrinsic posterior tilt of the cilium relative to the cell membrane may be the answer. High speed video­ microscopy was used to track the trajectory of node cilia in mouse embryos, which are posteriorly tilted at an

Brain

Heart

Gut Pitx2 Lefty

Nodal

Midline barrier

Asymmetric flow LR cilia

Left-sided signal (Ca2+)

L

R

Figure 3  The nodal flow hypothesis of left–right development. Based on studies in the mouse embryo, it was proposed that specialized left– right ciliated cells in the ventral node break bilateral embryonic symmetry by generating a leftward fluid flow. This asymmetric flow activates signals on the left side of the node (which likely include Ca2 signals) that in turn trigger asymmetric Nodal expression. The Nodal signaling cascade then expands on the left side of embryo. Pitx2, a Nodal target gene, is thought to “transfer” left–right information generated by left– right cilia to developing organ primordia such as the heart, gut and brain.

angle of 40° (Okada et al., 2005). These observations revealed that the rightward rotational movement of the cilium occurs close to the cell surface, whereas the leftward motion is away from the surface (Fig. 4). Based on hydrodynamic modeling, it was proposed that shear resistance of the cell surface slows the rightward swing, but leaves the leftward stroke unimpeded to generate a leftward fluid flow (Buceta et al., 2005; Okada et al., 2005). Interestingly, the posterior positioning of the cilium may be controlled by the alignment of the base of the cilium, the basal body, within the cell (Fig. 4) (Okada et al., 2005). This suggests a role for planar cell polarity (PCP) signaling or a similar pathway in orienting left–right cilia relative to the previously established anterioposteror axis. Several experimental results support the nodal flow model. Positional cloning of the inversus viscerum (iv) mutation in mice identified an axonemal dynein gene named left-right dynein (lrd) that encodes a structural component of node cilia (Supp et al., 1997). The spontaneous iv mutation randomizes visceral organ laterality and alters patterns of asymmetric gene expression (Lowe et al., 1996). Analysis of node cilia in iv mutant embryos revealed that the cilia were present but paralyzed (Okada

289

Chapter  |  4.1  Cardiac Left–Right Asymmetry

Anterior

Right Asymmetric flow

Two-cilia model

Morphogen gradient model

Motile cilium Mechanosensory cilium

Basal body Left

Ciliated cell Posterior

Figure 4  Cilia-driven asymmetric flow. Motile left–right cilia beat in a vortical fashion and are tilted to the posterior. Hydrodynamic modeling suggests that the tilt results in a dampened rightward stroke close to the cell surface (blue arrows) and an unimpeded leftward stroke away from the surface (red arrows) capable of generating net leftward fluid flow. The posterior positioning of the cilium might be controlled by the alignment of the basal body. There are two prominent models that explain how asymmetric flow is translated into left–right signals. First, the “morphogen gradient” model proposes that leftward flow asymmetrically distributes secreted morphogens, which accumulate and activate signaling on the left. Second, the “twocilia” model suggests that in addition to motile cilia that generate fluid flow, there is a distinct set of mechanosensory cilia that sense the flow. Bending of mechanosensory cilia on the left side (but not the right) has been proposed to induce asymmetric Ca2 signaling and Nodal expression. The A–P (anterior–posterior) and LR (left–right) axes are indicated.

et al., 1999), providing an additional link between cilia motility and left–right axis specification. Perhaps the strongest evidence supporting a role for nodal flow in left– right development comes from cultured mouse embryos exposed to external fluid flow. In these experiments, an artificial rightward fluid flow reversed asymmetric gene expression in wild-type embryos, and leftward flow was able to rescue left–right development in mutant embryos with inverted left–right asymmetry (Nonaka et al., 2002). These data provided a potential molecular mechanism underlying left–right defects in cilia defective syndromes such as Kartagener’s syndrome. At the same time that experimental evidence was bolstering the case for nodal flow, a number of important questions about the model began to arise (reviewed in Wagner and Yost, 2000). First, is nodal flow a phenomenon found only in the mouse embryo? Given the conservation of the left-sided Nodal signaling cascade in vertebrates, it appeared likely that mechanisms that directly trigger this pathway are also conserved. However, there was no clear evidence that cilia were involved in left– right asymmetry in other model vertebrates. Second, while the connection between cilia defects and left–right defects in mouse mutants was indeed compelling, it was strictly correlative. Mutations in genes such as kif3b (Nonaka et al., 1998) or kif3a (Marszalek et al., 1999; Takeda et al., 1999) that disrupt node cilia and nodal flow also disrupt other developmental pathways, resulting in pleiotrophic phenotypes that include aberrant midline development.

Therefore, left–right defects in these embryos could be due to other developmental defects, not node cilia defects. Third, how exactly is asymmetric fluid flow translated into left–right “signals” capable of activating asymmetric gene expression in neighboring cells? The original proposal that fluid flow asymmetrically distributed a secreted morphogen was not entirely consistent with all data from different cilia mutants. The remainder of this section will review recent work designed to address these issues.

VI.B.  Asymmetric Fluid Flow in Zebrafish: Ciliated Cells are Necessary for Left–Right Development To address whether cilia-driven asymmetric flow is specific to the mouse embryo or if it is present in other vertebrates, genes homologous to the mouse lrd gene were cloned from chick, frog and zebrafish (Essner et al., 2002). Using these genes as markers to hunt for cells analogous to ventral node cells in mouse, monociliated cells expressing lrd were found in HN and the primitive streak in chick, the ventral side of the dorsal blastopore in frog and in an enigmatic structure called Kupffer’s vesicle in zebrafish (Essner et al., 2002). In each species, the presence of these ciliated cells preceded the onset of asymmetric gene expression. This was the first inkling that ciliary left–right asymmetric fluid flow may be part of a conserved mechanism to direct asymmetry in the developing vertebrate embryo.

290

The zebrafish was the first nonmammalian embryo in which the functionality of cilia-driven asymmetric flow was tested. In the zebrafish embryo, monociliated cells expressing lrd line the lumen of Kupffer’s vesicle (KV), a transient, spherical, fluid-filled organ that develops at the caudal end of the midline. Kupffer’s vesicle is a conserved structure in teleost fishes, and was first described well over a century ago (Kupffer, 1868). Until recently, the function of Kupffer’s vesicle was a mystery and the structure was referred to as an “organ of ambiguity” (Warga and Stainier, 2002). Fate mapping studies (Cooper and D’Amico, 1996; Melby et al., 1996) show that Kupffer’s vesicle is generated from a group of about two dozen specialized cells, known as dorsal forerunner cells (DFCs), that migrate just ahead of the embryonic shield (the zebrafish equivalent of the node/Organizer) during gastrulation. Laser ablation of DFCs or surgical disruption of Kupffer’s vesicle altered left–right asymmetric gene expression without inducing other gross developmental defects such as midline abnormalities, indicating a role for Kupffer’s vesicle cells in left– right development (Essner et al., 2005). Videomicroscopy of live embryos revealed that monocilia in Kupffer’s vesicle are motile, beat in a rotational fashion and generate a consistent asymmetric flow of fluorescent microbeads injected into the vesicle (Essner et al., 2005; KramerZucker et al., 2005). Antisense morpholino knockdown of zebrafish genes that affect ciliogenesis (polaris, hippi) or cilia motility (lrd) abrogated Kupffer’s vesicle fluid flow and altered left–right asymmetric gene expression (Essner et al., 2005; Kramer-Zucker et al., 2005). Based on these results, Kupffer’s vesicle was proposed to function as an embryonic “organ of asymmetry” (Essner et al., 2005) that directs left–right patterning via asymmetric fluid flow in a manner analogous to the ventral node in mouse. Genetic evidence from mouse and zebrafish suggested ciliated cells play an important role in left–right development. However, genes proposed to function in ciliated node cells in mouse and Kupffer’s vesicle cells in zebrafish are also expressed in other cell types during left–right development. As a result, mutations in these genes result in pleiotropic phenotypes that make it impossible to connect node cilia defects directly with left–right developmental defects. For example, mutations in several genes, including kif3b, kif3a and polaris that affect ciliated node cells also disrupt midline structures that are essential for normal left–right development. Thus, in these mutants it is unclear whether left–right abnormalities are due to cilia defects or midline barrier defects. To test directly whether gene function in ciliated cells is required for left–right development, a technique was developed in zebrafish to knockdown expression of a specific gene by placing antisense morpholinos exclusively in the DFC/KV cell lineage (Amack and Yost, 2004). This technique creates mosaic embryos in which function of a particular gene is reduced in DFC/KV cells, but remains normal in other embryonic cells. This approach

PART  |  4  Asymmetry in Cardiac Development

was first used to knockdown selectively DFC/KV expression of the T-box transcription factor no tail (homologous to mouse brachyury), a gene required for development of both the ciliated Kupffer’s vesicle and the embryonic midline (Halpern et al., 1993; Melby et al., 1996). Knockdown of no tail specifically in DFC/KV cells did not affect midline development, but disrupted Kupffer’s vesicle morphology and altered left–right asymmetric gene expression and heart laterality (Amack and Yost, 2004). This provided the first evidence in any organism that loss of gene function specifically and exclusively in ciliated cells disrupts left– right development. This approach has been subsequently used to show that additional genes necessary for Kupffer’s vesicle form or function are required in ciliated cells for normal left–right development. Targeted knockdown of these genes in DFC/KV cells, which include lrd (Essner et al., 2005), polaris and pkd2 (Bisgrove et al., 2005), tbx16 (Amack et al., 2007), Na,K-ATPase 2 and Ncx4a (Shu et al., 2007) and BMP4 (Chocron et al., 2007), demonstrate that disruption of Kupffer’s vesicle function – and in particular loss of Kupffer’s vesicle fluid flow – is sufficient to alter left–right development in the context of an otherwise wild-type embryo.

VI.C.  “Left–Right Cilia” and the   “Organ of Asymmetry” Recently, cilia-driven asymmetric fluid flow has been described in several other vertebrates: Kupffer’s vesicle in medakafish; the notochordal plate in rabbit; and the caudal gastrocoel roof plate in Xenopus (Okada et al., 2005; Schweickert et al., 2007). These observations indicate specialized ciliated cells play a conserved role in left– right development. Since the location of these apparently analogous groups of ciliated cells differs from embryo to embryo, nomenclature can appear confusing (e.g., “cil­ iated ventral node cells” versus “ciliated Kupffer’s vesicle cells” or “nodal flow” versus “Kupffer’s vesicle flow” versus “notochordal plate flow”). To avoid confusion, we propose unifying terms for the components of the ciliary mechanism that controls left–right development in the vertebrate embryo. We refer to the transient ciliated structure that generates fluid flow as the “organ of asymmetry,” the cilia within this structure as “left–right cilia” and the ciliadriven fluid flow as “asymmetric flow.” Ironically, asymmetric flow has not been reported in the chick embryo where asymmetric gene expression was first described.

VI.D.  How does Asymmetric Flow Send Left–Right Signals? Two models have been proposed to explain how asymmetric flow directs asymmetric gene expression and subsequent left–right development. There are arguments for and

291

Chapter  |  4.1  Cardiac Left–Right Asymmetry

against both models, which are not mutually exclusive. The “morphogen gradient” hypothesis came first, proposing that signaling molecules are asymmetrically distributed by fluid flow (Fig. 4). In its simplest form, this model requires that signaling molecules such as Shh, FGF or Nodal proteins are secreted into the flowing extracellular fluid and swept to the left side of the embryo to bind cell surface receptors and initiate a signaling cascade. Supporting this idea, fluorescently labeled dextrans (20–40 kD) added to the mouse ventral node pit or rabbit notochordal plate were asymmetrically distributed with a leftward bias (Okada et al., 2005). Although straightforward, morphogen flow does not encompass all available data. One would predict that mutants lacking flow would share a common molecular phenotype; random diffusion of morphogens could activate the Nodal signaling cascade on the left, right, both sides, or not at all. This phenotype is seen in lrd/ mice that have immotile nodal cilia, however, mice devoid of nodal cilia (e.g., kif3a/, kif3b/) have a different phenotype, showing either bilateral or absent Nodal signaling (reviewed in Wagner and Yost, 2000). Thus, concerns about how mutations in different cilia genes that similarly compromise fluid flow could result in different patterns of left–right gene expression have called this hypothesis into question. Moreover, an in vivo asymmetric gradient of an endogenously expressed morphogen has not been identified. A more sophisticated version of the morphogen gradient hypothesis that stems from observations in the mouse node (Tanaka et al., 2005) postulates that instead of freely bumping around in extracellular fluid, Shh and retinoic acid (RA) signals are distributed asymmetrically in membrane-bound nodal vesicular parcels (NVPs). These parcels, which are relatively large (0.3–5 microns in diameter), pinch off from node cells, are transported to the left side of the node by fluid flow, are then broken apart by contacting rotating cilia and finally absorbed by left-sided node cells (reviewed in (Hirokawa et al., 2006). Pharmacological experiments suggest nodal vesicular parcel-transported Shh and retinoic acid act to elevate the Ca2 concentration in an FGF-dependent manner on the left side of the node (Tanaka et al., 2005), which is in turn thought to influence the initiation of asymmetric gene expression in adjacent cells. Asymmetric Ca2 concentration gradients have been observed at the node in mouse (McGrath et al., 2003), HN in chick (Raya et al., 2004) and Kupffer’s vesicles in zebrafish (Sarmah et al., 2005). If nodal vesicular parcel flow concentrates Shh and retinoic acid on the left and these molecules function to activate downstream signaling pathways, one might expect to detect asymmetric expression of downstream Shh- or retinoic acid-response genes. However, asymmetric expression of Shh and retinoic acid pathway genes has not been detected in or near the mouse node. As an alternative to morphogen gradients, the “twocilia” model proposes that there are two distinct types of left–right cilia: motile cilia that generate fluid flow; and

immotile mechanosensory cilia that sense and respond to flow (Tabin and Vogan, 2003). In this model, asymmetric flow generated by motile cilia bends sensory cilia exclusively on the left side of the organ of asymmetry, which activates a Ca2 response that initiates asymmetric gene expression (Fig. 4). Two biochemically-distinct populations of left–right cilia could explain why mutant embryos with defects in motile cilia have different left–right phenotypes than embryos devoid of all cilia. The best support of this comes from the mouse node, where two populations of cilia have been identified using a transgene that encodes a GFPtagged left–right dynein (LrdGFP) fusion protein (McGrath et al., 2003). Cilia that express LrdGFP are thought to be motile cilia, and those that do not are predicted to be immotile sensory cilia. This study also reported an asymmetric Ca2 gradient across the node that correlates with asymmetric gene expression. However, inhibition of FGF signaling suppressed asymmetric Ca2 signals in the mouse node without altering asymmetric flow (Tanaka et al., 2005). This indicates cilia motility and asymmetric flow can be uncoupled from asymmetric Ca2 at the node. The link between cilia and Ca2 signaling comes primarily from work in kidney cells, where immotile cilia respond to fluid flow-induced bending by initiating an influx of extracellular calcium (Ca2) through mechanosensory channels (Praetorius and Spring, 2001). Two genes mutated in autosomal dominant polycystic kidney disease, Pkd1 and Pkd2, interact to form a Ca2-permeable cation channel (Hanaoka et al., 2000) that localizes to cilia in kidney cells (Yoder et al., 2002). Data implicating sensory cilia in left–right patterning come from pkd2 mouse mutants (Pennekamp et al., 2002), zebrafish MO studies (Bisgrove et al., 2005) and zebrafish mutant studies (Schottenfeld et al., 2007) and indicate Pkd2 is required for normal left– right development. Mechansensory cilia may serve to sense and respond to asymmetric flow by generating a Ca2 influx via the Pkd1/Pkd2 channel. Ca2 signals initiated in a few cells on the left side of the embryo could be transmitted to surrounding cells through gap junctions, which may explain the previously recognized requirement of gap junctions in left–right patterning (Levin and Mercola, 1998, 1999). A left-sided Ca2 signal could induce asymmetric gene expression in the LPM either directly or indirectly by activating a morphogen or other regulatory factor. While the involvement of mechanosensory cilia remains to be clarified, there indeed appears to be a conserved role for asymmetric Ca2 gradients, as mentioned above, in establishing the asymmetric Nodal signaling cascade.

VII.  Asymmetries that precede   cilia-dependent asymmetric flow Although the role of cilia in left–right development appears to be highly conserved in vertebrates, and there

292

PART  |  4  Asymmetry in Cardiac Development

pH/voltage gradients

Sydecan-2 signaling

?

?

LR cilia Asymmetric flow

Nodal

Lefty Pitx2

?

?

?

Heart looping Gut looping

Brain asymmetry

Figure 5  Working model of left–right development. The relationship between early left–right asymmetries (such as pH and voltage gradients and signaling by asymmetrically phosphorylated Syndecan-2) and subsequent left–right cilia-dependent asymmetric flow remains unclear. These pathways may independently influence the conserved Nodal/Lefty/Pitx2 signaling cascade. Alternatively, early asymmetries may direct the form or function of left–right cilia, such that asymmetric flow acts to relay or amplify previously-established left–right information. Pitx2 is asymmetrically expressed in the developing heart, gut and brain and likely transfers left–right information to these organs. However, the downstream genes that mediate this transfer and the mechanisms used by individual organs to receive and interpret this information remain poorly-understood.

are elegant models by which asymmetric fluid flow generates left–right asymmetry from bilateral symmetry, it is not clear whether cilia-dependent flow is the prime mover in left–right axis formation. There are quite a few examples of molecular left–right asymmetries that precede function of the ciliated organ of asymmetry (Fig. 5). First, there appear to be several molecules that are asymmetrically distributed in cleavage-stage Xenopus embryos, including 14-3-3ε protein (Bunney et al., 2003), H/K-ATPase (Levin et al., 2002) and H-V-ATPase (Adams et al., 2006). Inhibition of these molecules disrupts asymmetric nodal expression and organ laterality, suggesting that the left– right axis is established early in ectoderm lineages. It has been proposed that the H/K-ATPase and H-V-ATPase

pumps establish left–right gradients of pH and voltage across the embryo to drive asymmetric accumulations of signaling molecules, such as Serotonin (Fukumoto et al., 2005). Second, the transmembrane proteoglycan syndecan-2 is equally abundant on the left and right ectoderm in Xenopus embryos, but the cytoplasmic domain is phosphorylated on the right side and not the left (Kramer et al., 2002; Kramer and Yost, 2002). This asymmetry appears to transmit left–right information to migrating mesoderm during gastrulation. Third, perturbations of microtubule-dependent cytoplasmic rotation in the first cell-cycle after fertilization alter left–right axis formation in Xenopus (Yost, 1991). While each of these examples were first described in the Xenopus embryo, there is evidence that asymmetric H/ K-ATPase and H-V-ATPase activity is required for normal left-sided Nodal in chick and zebrafish (Levin et al., 2002; Kawakami et al., 2005; Adams et al., 2006). It has been proposed that H/K-ATPase activity leads to an increase in left-sided extracellular Ca2, which in turn activates Notch signaling upstream of nodal (Raya et al., 2004; Kawakami et al., 2005). Several experimental approaches indicate that these early left–right asymmetries are utilized before the formation of left–right cilia. However, it is not known whether these pathways are parallel to the roles of cilia in left–right development, or whether they directly impact on the formation or function of ciliated structures that contribute to left–right development. In one striking example of pathway convergence, mRNA encoding the proton pump H-V-ATPase subunit A is asymmetrically distributed early in Xenopus development, and perturbation of its function in zebrafish results in altered Kupffer’s vesicle formation (Adams et al., 2006), which presumably disrupts asymmetric flow. This suggests that early molecular asymmetries function by influencing the form and/or function of ciliated cells in the organ of asymmetry. However, it does not exclude the possibility that the role of early molecular asymmetries is to establish the left–right axis and the role of cilia-dependent fluid flow is to relay or amplify this left–right information.

VIII.  Conclusions and future perspectives Our understanding of the molecular basis of left–right development has greatly expanded in the last decade. Over 100 candidate genes have been implicated in the pathway. The mechanisms that generate left–right asymmetry are actively studied in several vertebrate model systems, including embryos of mice, zebrafish, frogs, medaka fish, chick and rabbit, and by genetic analysis for inherited laterality defects in humans. In addition, the noncardiac cells that are essential for cardiac left–right patterning, such as ciliated cells, midline cells and lateral plate mesoderm

Chapter  |  4.1  Cardiac Left–Right Asymmetry

cells, are becoming better understood. Thus, the earlier steps in generating the left–right axis in embryos are becoming clearer. Among the challenges for the next decade will be to determine what, if any, connections link early molecular left–right asymmetries with the generation of cilia-driven asymmetric flow, and at the other end of the process, to understand the cellular and molecular mechanisms by which cardiac cells receive left–right axis information and utilize it to build an asymmetric heart.

References Adams, D.S., Robinson, K.R., Fukumoto, T., Yuan, S., Albertson, R.C., Yelick, P., Kuo, L., McSweeney, M., Levin, M., 2006. Early, HV-ATPase-dependent proton flux is necessary for consistent left– right patterning of nonmammalian vertebrates.. Development 133, 1657–1671. Afzelius, B.A., 1976. A human syndrome caused by immotile cilia. Science 193, 119–317. Ai, D., Liu, W., Ma, L., Dong, F., Lu, M.F., Wang, D., Verzi, M.P., Cai, C., Gage, P.J., Evans, S., Black, B.L., Brown, N.A., Martin, J.F., 2006. Pitx2 regulates cardiac left–right asymmetry by patterning second cardiac lineage-derived myocardium. Dev. Biol. 296, 437–449. Amack, J.D., Yost, H.J., 2004. The T-box transcription factor no tail in ciliated cells controls zebrafish left–right asymmetry. Curr. Biol. 14, 685–690. Amack, J.D., Wang, X., Yost, H.J., 2007. Two T-box genes play independent and cooperative roles to regulate morphogenesis of ciliated Kupffer’s vesicle in zebrafish. Dev. Biol. 310, 196–210. Aylsworth, A.S., 2001. Clinical aspects of defects in the determination of laterality. Am. J. Med. Genet. 101, 345–355. Baldwin, H.S., Solursh, M., 1989. Degradation of hyaluronic acid does not prevent looping of the mammalian heart in situ. Dev. Biol. 136, 555–559. Biben, C., Harvey, R.P., 1997. Homeodomain factor Nkx2-5 controls left/ right asymmetric expression of bHLH gene eHand during murine heart development. Genes. Dev. 11, 1357–1369. Bisgrove, B.W., Essner, J.J., Yost, H.J., 1999. Regulation of midline development by antagonism of lefty and nodal signaling. Development 126, 3253–3262. Bisgrove, B.W., Essner, J.J., Yost, H.J., 2000. Multiple pathways in the midline regulate concordant brain, heart and gut left–right asymmetry. Development 127, 3567–3579. Bisgrove, B.W., Morelli, S.H., Yost, H.J., 2003. Genetics of human laterality disorders: insights from vertebrate model systems. Annu. Rev. Genomics. Hum. Genet. 4, 1–32. Bisgrove, B.W., Snarr, B.S., Emrazian, A., Yost, H.J., 2005. Polaris and Polycystin-2 in dorsal forerunner cells and Kupffer’s vesicle are required for specification of the zebrafish left–right axis. Dev. Biol. 287, 274–288. Boettger, T., Wittler, L., Kessel, M., 1999. FGF8 functions in the specification of the right body side of the chick. Curr. Biol. 9, 277–280. Bowers, P.N., Brueckner, M., Yost, H.J., 1996. The genetics of left–right development and heterotaxia. Semin. Perinatol. 20, 577–588. Brand, T., 2003. Heart development: molecular insights into cardiac specification and early morphogenesis. Dev. Biol. 258, 1–19. Branford, W.W., Essner, J.J., Yost, H.J., 2000. Regulation of gut and heart left–right asymmetry by context-dependent interactions between Xenopus lefty and BMP4 signaling. Dev. Biol. 223, 291–306.

293

Breckenridge, R.A., Mohun, T.J., Amaya, E., 2001. A role for BMP signalling in heart looping morphogenesis in Xenopus. Dev. Biol. 232, 191–203. Brennan, J., Norris, D.P., Robertson, E.J., 2002. Nodal activity in the node governs left–right asymmetry. Genes. Dev. 16, 2339–2344. Buceta, J., Ibanes, M., Rasskin-Gutman, D., Okada, Y., Hirokawa, N., Izpisua-Belmonte, J.C., 2005. Nodal cilia dynamics and the specification of the left/right axis in early vertebrate embryo development. Biophys. J. 89, 2199–2209. Bunney, T.D., De Boer, A.H., Levin, M., 2003. Fusicoccin signaling reveals 14-3-3 protein function as a novel step in left–right patterning during amphibian embryogenesis. Development 130, 4847–4858. Burdine, R.D., Schier, A.F., 2000. Conserved and divergent mechanisms in left–right axis formation. Genes. Dev. 14, 763–776. Campione, M., Steinbeisser, H., Schweickert, A., Deissler, K., van Bebber, F., Lowe, L.A., Nowotschin, S., Viebahn, C., Haffter, P., Kuehn, M.R., Blum, M., 1999. The homeobox gene Pitx2: mediator of asymmetric left–right signaling in vertebrate heart and gut looping. Development 126, 1225–1234. Capdevila, J., Vogan, K.J., Tabin, C.J., Izpisua Belmonte, J.C., 2000. Mechanisms of left–right determination in vertebrates. Cell 101, 9–21. Cartwright, J.H., Piro, O., Tuval, I., 2004. Fluid-dynamical basis of the embryonic development of left–right asymmetry in vertebrates. Proc. Natl. Acad. Sci. USA 101, 7234–7239. Chen, J.N., van Eeden, F.J., Warren, K.S., Chin, A., Nusslein-Volhard, C., Haffter, P., Fishman, M.C., 1997. Left–right pattern of cardiac BMP4 may drive asymmetry of the heart in zebrafish. Development 124, 4373–4382. Chen, Y., Solursh, M., 1992. Comparison of Hensen’s node and retinoic acid in secondary axis induction in the early chick embryo. Dev. Dyn. 195, 142–151. Cheng, A.M., Thisse, B., Thisse, C., Wright, C.V., 2000. The lefty-related factor Xatv acts as a feedback inhibitor of nodal signaling in mesoderm induction and L–R axis development in Xenopus. Development 127, 1049–1061. Chocron, S., Verhoeven, M.C., Rentzsch, F., Hammerschmidt, M., Bakkers, J., 2007. Zebrafish Bmp4 regulates left–right asymmetry at two distinct developmental time points. Dev. Biol. 305, 577–588. Collignon, J., Varlet, I., Robertson, E.J., 1996. Relationship between asymmetric nodal expression and the direction of embryonic turning. Nature 381, 155–158. Cooper, M.S., D’Amico, L.A., 1996. A cluster of noninvoluting endocytic cells at the margin of the zebrafish blastoderm marks the site of embryonic shield formation. Dev. Biol. 180, 184–198. Dagle, J.M., Sabel, J.L., Littig, J.L., Sutherland, L.B., Kolker, S.J., Weeks, D.L., 2003. Pitx2c attenuation results in cardiac defects and abnormalities of intestinal orientation in developing Xenopus laevis. Dev. Biol. 262, 268–281. Danos, M.C., Yost, H.J., 1995. Linkage of cardiac left–right asymmetry and dorsal–anterior development in Xenopus. Development 121, 1467–1474. Danos, M.C., Yost, H.J., 1996. Role of notochord in specification of cardiac left–right orientation in zebrafish and Xenopus. Dev. Biol. 177, 96–103. Essner, J.J., Branford, W.W., Zhang, J., Yost, H.J., 2000. Mesendoderm and left–right brain, heart and gut development are differentially regulated by pitx2 isoforms. Development 127, 1081–1093. Essner, J.J., Vogan, K.J., Wagner, M.K., Tabin, C.J., Yost, H.J., Brueckner, M., 2002. Conserved function for embryonic nodal cilia. Nature 418, 37–38.

294

Essner, J.J., Amack, J.D., Nyholm, M.K., Harris, E.B., Yost, H.J., 2005. Kupffer’s vesicle is a ciliated organ of asymmetry in the zebrafish embryo that initiates left–right development of the brain, heart and gut. Development 132, 1247–1260. Fischer, A., Viebahn, C., Blum, M., 2002. FGF8 acts as a right determinant during establishment of the left–right axis in the rabbit. Curr. Biol. 12, 1807–1816. Fukumoto, T., Kema, I.P., Levin, M., 2005. Serotonin signaling is a very early step in patterning of the left–right axis in chick and frog embryos. Curr. Biol. 15, 794–803. Gage, P.J., Suh, H., Camper, S.A., 1999. Dosage requirement of Pitx2 for development of multiple organs. Development 126, 4643–4651. Gebbia, M., Ferrero, G.B., Pilia, G., Bassi, M.T., Aylsworth, A., PenmanSplitt, M., Bird, L.M., Bamforth, J.S., Burn, J., Schlessinger, D., Nelson, D.L., Casey, B., 1997. X-linked situs abnormalities result from mutations in ZIC3. Nat. Genet. 17, 305–308. Halpern, M.E., Ho, R.K., Walker, C., Kimmel, C.B., 1993. Induction of muscle pioneers and floor plate is distinguished by the zebrafish no tail mutation. Cell 75, 99–111. Hanaoka, K., Qian, F., Boletta, A., Bhunia, A.K., Piontek, K., Tsiokas, L., Sukhatme, V.P., Guggino, W.B., Germino, G.G., 2000. Co-assembly of polycystin-1 and -2 produces unique cation-permeable currents. Nature 408, 990–994. Harvey, R.P., 2002. Patterning the vertebrate heart. Nat. Rev. Genet. 3, 544–556. Hashimoto, H., Rebagliati, M., Ahmad, N., Muraoka, O., Kurokawa, T., Hibi, M., Suzuki, T., 2004. The Cerberus/Dan-family protein Charon is a negative regulator of Nodal signaling during left–right patterning in zebrafish. Development 131, 1741–1753. Hirokawa, N., Tanaka, Y., Okada, Y., Takeda, S., 2006. Nodal flow and the generation of left–right asymmetry. Cell 125, 33–45. Hyatt, B.A., Lohr, J.L., Yost, H.J., 1996. Initiation of vertebrate left–right axis formation by maternal Vg1. Nature 384, 62–65. Icardo, J.M., 1988. Heart anatomy and developmental biology. Experientia 44, 910–919. Isaac, A., Sargent, M.G., Cooke, J., 1997. Control of vertebrate left–right asymmetry by a snail-related zinc finger gene. Science 275, 1301–1304. Itasaki, N., Nakamura, H., Sumida, H., Yasuda, M., 1991. Actin bundles on the right side in the caudal part of the heart tube play a role in dextro-looping in the embryonic chick heart. Anat. Embryol. (Berl) 183, 29–39. Kawakami, Y., Raya, A., Raya, R.M., Rodriguez-Esteban, C., Belmonte, J.C., 2005. Retinoic acid signalling links left–right asymmetric patterning and bilaterally symmetric somitogenesis in the zebrafish embryo. Nature 435, 165–171. Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1, 435–440. Kioussi, C., Briata, P., Baek, S.H., Rose, D.W., Hamblet, N.S., Herman, T., Ohgi, K.A., Lin, C., Gleiberman, A., Wang, J., Brault, V., Ruiz-Lozano, P., Nguyen, H.D., Kemler, R., Glass, C.K., WynshawBoris, A., Rosenfeld, M.G., 2002. Identification of a Wnt/Dvl/betaCatenin -- Pitx2 pathway mediating cell-type-specific proliferation during development. Cell 111, 673–685. Kitaguchi, T., Nagai, T., Nakata, K., Aruga, J., Mikoshiba, K., 2000. Zic3 is involved in the left–right specification of the Xenopus embryo. Development 127, 4787–4795. Kitamura, K., Miura, H., Miyagawa-Tomita, S., Yanazawa, M., KatohFukui, Y., Suzuki, R., Ohuchi, H., Suehiro, A., Motegi, Y., Nakahara, Y., Kondo, S., Yokoyama, M., 1999. Mouse Pitx2 deficiency leads to

PART  |  4  Asymmetry in Cardiac Development

anomalies of the ventral body wall, heart, extra- and periocular mesoderm and right pulmonary isomerism. Development 126, 5749–5758. Kosaki, K., Casey, B., 1998. Genetics of human left–right axis malformations. Semin. Cell. Dev. Biol. 9, 89–99. Kramer-Zucker, A.G., Olale, F., Haycraft, C.J., Yoder, B.K., Schier, A.F., Drummond, I.A., 2005. Cilia-driven fluid flow in the zebrafish pronephros, brain and Kupffer’s vesicle is required for normal organogenesis. Development 132, 1907–1921. Kramer, K.L., Yost, H.J., 2002. Ectodermal syndecan-2 mediates left– right axis formation in migrating mesoderm as a cell-nonautonomous Vg1 cofactor. Dev. Cell 2, 115–124. Kramer, K.L., Barnette, J.E., Yost, H.J., 2002. PKCgamma regulates syndecan-2 inside-out signaling during Xenopus left–right development. Cell 111, 981–990. Kupffer, C., 1868. Beobachtungea uber die Entwicklung der Knochenfische. Arch. Mikrob. Anat. 4, 209–272. Latacha, K.S., Remond, M.C., Ramasubramanian, A., Chen, A.Y., Elson, E.L., Taber, L.A., 2005. Role of actin polymerization in bending of the early heart tube. Dev. Dyn. 233, 1272–1286. Levin, M., Mercola, M., 1998. Gap junctions are involved in the early generation of left–right asymmetry. Dev. Biol. 203, 90–105. Levin, M., Mercola, M., 1999. Gap junction-mediated transfer of left– right patterning signals in the early chick blastoderm is upstream of Shh asymmetry in the node. Development 126, 4703–4714. Levin, M., Thorlin, T., Robinson, K.R., Nogi, T., Mercola, M., 2002. Asymmetries in H/K-ATPase and cell membrane potentials comprise a very early step in left–right patterning. Cell 111, 77–89. Levin, M., Johnson, R.L., Stern, C.D., Kuehn, M., Tabin, C., 1995. A molecular pathway determining left–right asymmetry in chick embryogenesis. Cell 82, 803–814. Levin, M., Pagan, S., Roberts, D.J., Cooke, J., Kuehn, M.R., Tabin, C.J., 1997. Left/right patterning signals and the independent regulation of different aspects of situs in the chick embryo. Dev. Biol. 189, 57–67. Lin, C.R., Kioussi, C., O’Connell, S., Briata, P., Szeto, D., Liu, F., Izpisua-Belmonte, J.C., Rosenfeld, M.G., 1999. Pitx2 regulates lung asymmetry, cardiac positioning and pituitary and tooth morphogenesis. Nature 401, 279–282. Linask, K.K., Yu, X., Chen, Y., Han, M.D., 2002. Directionality of heart looping: effects of Pitx2c misexpression on flectin asymmetry and midline structures. Dev. Biol. 246, 407–417. Linask, K.K., Han, M.D., Linask, K.L., Schlange, T., Brand, T., 2003. Effects of antisense misexpression of CFC on downstream flectin protein expression during heart looping. Dev. Dyn. 228, 217–230. Lines, M.A., Kozlowski, K., Walter, M.A., 2002. Molecular genetics of Axenfeld-Rieger malformations. Hum. Mol. Genet. 11, 1177–1184. Liu, C., Liu, W., Lu, M.F., Brown, N.A., Martin, J.F., 2001. Regulation of left–right asymmetry by thresholds of Pitx2c activity. Development 128, 2039–2048. Liu, C., Liu, W., Palie, J., Lu, M.F., Brown, N.A., Martin, J.F., 2002. Pitx2c patterns anterior myocardium and aortic arch vessels and is required for local cell movement into atrioventricular cushions. Development 129, 5081–5091. Liu, W., Selever, J., Lu, M.F., Martin, J.F., 2003. Genetic dissection of Pitx2 in craniofacial development uncovers new functions in branchial arch morphogenesis, late aspects of tooth morphogenesis and cell migration. Development 130, 6375–6385. Logan, M., Pagan-Westphal, S.M., Smith, D.M., Paganessi, L., Tabin, C.J., 1998. The transcription factor Pitx2 mediates situs-specific morphogenesis in response to left–right asymmetric signals. Cell 94, 307–317.

Chapter  |  4.1  Cardiac Left–Right Asymmetry

Lohr, J.L., Danos, M.C., Yost, H.J., 1997. Left–right asymmetry of a nodal-related gene is regulated by dorsoanterior midline structures during Xenopus development. Development 124, 1465–1472. Lohr, J.L., Yost, H.J., 2000. Vertebrate model systems in the study of early heart development: Xenopus and zebrafish. Am. J. Med. Genet. 97, 248–257. Long, S., Ahmad, N., Rebagliati, M., 2003. The zebrafish nodal-related gene southpaw is required for visceral and diencephalic left–right asymmetry. Development 130, 2303–2316. Lowe, L.A., Supp, D.M., Sampath, K., Yokoyama, T., Wright, C.V., Potter, S.S., Overbeek, P., Kuehn, M.R., 1996. Conserved left–right asymmetry of nodal expression and alterations in murine situs inversus. Nature 381, 158–161. Lowe, L.A., Yamada, S., Kuehn, M.R., 2001. Genetic dissection of nodal function in patterning the mouse embryo. Development 128, 1831–1843. Lu, M.F., Pressman, C., Dyer, R., Johnson, R.L., Martin, J.F., 1999. Function of Rieger syndrome gene in left–right asymmetry and craniofacial development. Nature 401, 276–278. Manasek, F.J., Monroe, R.G., 1972. Early cardiac morphogenesis is independent of function. Dev. Biol. 27, 584–588. Manner, J., 2000. Cardiac looping in the chick embryo: a morphological review with special reference to terminological and biomechanical aspects of the looping process. Anat. Rec. 259, 248–262. Manning, A., McLachlan, J.C., 1990. Looping of chick embryo hearts in vitro. J. Anat. 168, 257–263. Marszalek, J.R., Ruiz-Lozano, P., Roberts, E., Chien, K.R., Goldstein, L.S., 1999. Situs inversus and embryonic ciliary morphogenesis defects in mouse mutants lacking the KIF3A subunit of kinesin-II. Proc. Natl. Acad. Sci. USA 96, 5043–5048. McGrath, J., Somlo, S., Makova, S., Tian, X., Brueckner, M., 2003. Two populations of node monocilia initiate left–right asymmetry in the mouse. Cell 114, 61–73. Melby, A.E., Warga, R.M., Kimmel, C.B., 1996. Specification of cell fates at the dorsal margin of the zebrafish gastrula. Development 122, 2225–2237. Meno, C., Saijoh, Y., Fujii, H., Ikeda, M., Yokoyama, T., Yokoyama, M., Toyoda, Y., Hamada, H., 1996. Left–right asymmetric expression of the TGF beta-family member lefty in mouse embryos. Nature 381, 151–155. Meno, C., Ito, Y., Saijoh, Y., Matsuda, Y., Tashiro, K., Kuhara, S., Hamada, H., 1997. Two closely-related left–right asymmetrically expressed genes, lefty-1 and lefty-2: their distinct expression domains, chromosomal linkage and direct neuralizing activity in Xenopus embryos. Genes. Cells 2, 513–524. Meno, C., Shimono, A., Saijoh, Y., Yashiro, K., Mochida, K., Ohishi, S., Noji, S., Kondoh, H., Hamada, H., 1998. lefty-1 is required for left–right determination as a regulator of lefty-2 and nodal. Cell 94, 287–297. Meno, C., Gritsman, K., Ohishi, S., Ohfuji, Y., Heckscher, E., Mochida, K., Shimono, A., Kondoh, H., Talbot, W.S., Robertson, E.J., Schier, A.F., Hamada, H., 1999. Mouse Lefty2 and zebrafish antivin are feedback inhibitors of nodal signaling during vertebrate gastrulation. Mol. Cell 4, 287–298. Mercola, M., Levin, M., 2001. Left–right asymmetry determination in vertebrates. Annu. Rev. Cell. Dev. Biol. 17, 779–805. Mjaatvedt, C.H., Nakaoka, T., Moreno-Rodriguez, R., Norris, R.A., Kern, M.J., Eisenberg, C.A., Turner, D., Markwald, R.R., 2001. The outflow tract of the heart is recruited from a novel heart-forming field. Dev. Biol. 238, 97–109.

295

Monsoro-Burq, A., Le Douarin, N.M., 2001. BMP4 plays a key role in left–right patterning in chick embryos by maintaining Sonic Hedgehog asymmetry. Mol. Cell 7, 789–799. Muslin, A.J., Williams, L.T., 1991. Well-defined growth factors promote cardiac development in axolotl mesodermal explants. Development 112, 1095–1101. Nakamura, T., Mine, N., Nakaguchi, E., Mochizuki, A., Yamamoto, M., Yashiro, K., Meno, C., Hamada, H., 2006. Generation of robust left– right asymmetry in the mouse embryo requires a self-enhancement and lateral-inhibition system. Dev. Cell 11, 495–504. Nascone, N., Mercola, M., 1995. An inductive role for the endoderm in Xenopus cardiogenesis. Development 121, 515–523. Nonaka, S., Shiratori, H., Saijoh, Y., Hamada, H., 2002. Determination of left–right patterning of the mouse embryo by artificial nodal flow. Nature 418, 96–99. Nonaka, S., Tanaka, Y., Okada, Y., Takeda, S., Harada, A., Kanai, Y., Kido, M., Hirokawa, N., 1998. Randomization of left–right asymmetry due to loss of nodal cilia generating leftward flow of extraembryonic fluid in mice lacking KIF3B motor protein. Cell 95, 829–837. Norris, D.P., Brennan, J., Bikoff, E.K., Robertson, E.J., 2002. The Foxh1dependent autoregulatory enhancer controls the level of Nodal signals in the mouse embryo. Development 129, 3455–3468. Okada, Y., Nonaka, S., Tanaka, Y., Saijoh, Y., Hamada, H., Hirokawa, N., 1999. Abnormal nodal flow precedes situs inversus in iv and inv mice. Mol. Cell 4, 459–468. Okada, Y., Takeda, S., Tanaka, Y., Belmonte, J.C., Hirokawa, N., 2005. Mechanism of nodal flow: a conserved symmetry breaking event in left–right axis determination. Cell 121, 633–644. Osada, S.I., Saijoh, Y., Frisch, A., Yeo, C.Y., Adachi, H., Watanabe, M., Whitman, M., Hamada, H., Wright, C.V., 2000. Activin/nodal responsiveness and asymmetric expression of a Xenopus nodal-related gene converge on a FAST-regulated module in intron 1. Development 127, 2503–2514. Pennekamp, P., Karcher, C., Fischer, A., Schweickert, A., Skryabin, B., Horst, J., Blum, M., Dworniczak, B., 2002. The ion channel polycystin-2 is required for left–right axis determination in mice. Curr. Biol. 12, 938–943. Praetorius, H.A., Spring, K.R., 2001. Bending the MDCK cell primary cilium increases intracellular calcium. J. Membr. Biol. 184, 71–79. Purandare, S.M., Ware, S.M., Kwan, K.M., Gebbia, M., Bassi, M.T., Deng, J.M., Vogel, H., Behringer, R.R., Belmont, J.W., Casey, B., 2002. A complex syndrome of left–right axis, central nervous system and axial skeleton defects in Zic3 mutant mice. Development 129, 2293–2302. Ramsdell, A.F., 2005. Left–right asymmetry and congenital cardiac defects: getting to the heart of the matter in vertebrate left–right axis determination. Dev. Biol. 288, 1–20. Ramsdell, A.F., Bernanke, J.M., Trusk, T.C., 2006. Left–right lineage analysis of the embryonic Xenopus heart reveals a novel framework linking congenital cardiac defects and laterality disease. Development 133, 1399–1410. Raya, A., Kawakami, Y., Rodriguez-Esteban, C., Ibanes, M., RasskinGutman, D., Rodriguez-Leon, J., Buscher, D., Feijo, J.A., Izpisua Belmonte, J.C., 2004. Notch activity acts as a sensor for extracellular calcium during vertebrate left–right determination. Nature 427, 121–128. Ryan, A.K., Blumberg, B., Rodriguez-Esteban, C., Yonei-Tamura, S., Tamura, K., Tsukui, T., de la Pena, J., Sabbagh, W., Greenwald, J., Choe, S., Norris, D.P., Robertson, E.J., Evans, R.M., Rosenfeld, M.G., Izpisúa Belmonte, J.C., 1998. Pitx2 determines left–right asymmetry of internal organs in vertebrates. Nature 394, 545–551.

296

Saijoh, Y., Adachi, H., Sakuma, R., Yeo, C.Y., Yashiro, K., Watanabe, M., Hashiguchi, H., Mochida, K., Ohishi, S., Kawabata, M., Miyazono, K., Whitman, M., Hamada, H., 2000. Left–right asymmetric expression of lefty2 and nodal is induced by a signaling pathway that includes the transcription factor FAST2. Mol. Cell 5, 35–47. Saijoh, Y., Oki, S., Ohishi, S., Hamada, H., 2003. Left–right patterning of the mouse lateral plate requires nodal produced in the node. Dev. Biol. 256, 160–172. Sampath, K., Cheng, A.M., Frisch, A., Wright, C.V., 1997. Functional differences among Xenopus nodal-related genes in left–right axis determination. Development 124, 3293–3302. Sarmah, B., Latimer, A.J., Appel, B., Wente, S.R., 2005. Inositol polyphosphates regulate zebrafish left–right asymmetry. Dev. Cell 9, 133–145. Sater, A.K., Jacobson, A.G., 1989. The specification of heart mesoderm occurs during gastrulation in Xenopus laevis. Development 105, 821–830. Sater, A.K., Jacobson, A.G., 1990. The role of the dorsal lip in the induction of heart mesoderm in Xenopus laevis. Development 108, 461–470. Schottenfeld, J., Sullivan-Brown, J., Burdine, R.D., 2007. Zebrafish curly up encodes a Pkd2 ortholog that restricts left-side-specific expression of southpaw. Development 134, 1605–1615. Schweickert, A., Weber, T., Beyer, T., Vick, P., Bogusch, S., Feistel, K., Blum, M., 2007. Cilia-driven leftward flow determines laterality in Xenopus. Curr. Biol. 17, 60–66. Sehnert, A.J., Huq, A., Weinstein, B.M., Walker, C., Fishman, M., Stainier, D.Y., 2002. Cardiac troponin T is essential in sarcomere assembly and cardiac contractility. Nat. Genet. 31, 106–110. Semina, E.V., Reiter, R., Leysens, N.J., Alward, W.L., Small, K.W., Datson, N.A., Siegel-Bartelt, J., Bierke-Nelson, D., Bitoun, P., Zabel, B.U., Carey, J.C., Murray, J.C., 1996. Cloning and characterization of a novel bicoid-related homeobox transcription factor gene, RIEG, involved in Rieger syndrome. Nat. Genet. 14, 392–399. Seo, J.W., Brown, N.A., Ho, S.Y., Anderson, R.H., 1992. Abnormal laterality and congenital cardiac anomalies. Relations of visceral and cardiac morphologies in the iv/iv mouse. Circulation 86, 642–650. Shiratori, H., Yashiro, K., Shen, M.M., Hamada, H., 2006. Conserved regulation and role of Pitx2 in situs-specific morphogenesis of visceral organs. Development 133, 3015–3025. Shu, X., Huang, J., Dong, Y., Choi, J., Langenbacher, A., Chen, J.N., 2007. Na,K-ATPase alpha2 and Ncx4a regulate zebrafish left–right patterning. Development 134, 1921–1930. Sissman, N.J., 1966. Cell multiplication rates during development of the primitive cardiac tube in the chick embryo. Nature 210, 504–507. Smith, S.M., Dickman, E.D., Thompson, R.P., Sinning, A.R., Wunsch, A.M., Markwald, R.R., 1997. Retinoic acid directs cardiac laterality and the expression of early markers of precardiac asymmetry. Dev. Biol. 182, 162–171. Soroldoni, D., Bajoghli, B., Aghaallaei, N., Czerny, T., 2007. Dynamic expression pattern of Nodal-related genes during left–right development in medaka. Gene. Expr. Patterns 7, 93–101. Stalsberg, H., 1969a. The origin of heart asymmetry: right and left contributions to the early chick embryo heart. Dev. Biol. 19, 109–127. Stalsberg, H., 1969b. Regional mitotic activity in the precardiac mesoderm and differentiating heart tube in the chick embryo. Dev. Biol. 20, 18–45. Supp, D.M., Witte, D.P., Potter, S.S., Brueckner, M., 1997. Mutation of an axonemal dynein affects left–right asymmetry in inversus viscerum mice. Nature 389, 963–966.

PART  |  4  Asymmetry in Cardiac Development

Taber, L.A., 2006. Biophysical mechanisms of cardiac looping. Int. J. Dev. Biol. 50, 323–332. Tabin, C.J., Vogan, K.J., 2003. A two-cilia model for vertebrate left–right axis specification. Genes. Dev. 17, 1–6. Takeda, S., Yonekawa, Y., Tanaka, Y., Okada, Y., Nonaka, S., Hirokawa, N., 1999. Left–right asymmetry and kinesin superfamily protein KIF3A: new insights in determination of laterality and mesoderm induction by kif3A/ mice analysis. J. Cell Biol. 145, 825–836. Talbot, W.S., Trevarrow, B., Halpern, M.E., Melby, A.E., Farr, G., Postlethwait, J.H., Jowett, T., Kimmel, C.B., Kimelman, D., 1995. A homeobox gene essential for zebrafish notochord development. Nature 378, 150–157. Tanaka, Y., Okada, Y., Hirokawa, N., 2005. FGF-induced vesicular release of Sonic hedgehog and retinoic acid in leftward nodal flow is critical for left–right determination. Nature 435, 172–177. Thisse, C., Thisse, B., 1999. Antivin, a novel and divergent member of the TGFbeta superfamily, negatively regulates mesoderm induction. Development 126, 229–240. Tsuda, T., Philp, N., Zile, M.H., Linask, K.K., 1996. Left–right asymmetric localization of flectin in the extracellular matrix during heart looping. Dev. Biol. 173, 39–50. Tsuda, T., Majumder, K., Linask, K.K., 1998. Differential expression of flectin in the extracellular matrix and left–right asymmetry in mouse embryonic heart during looping stages. Dev. Genet. 23, 203–214. Voronov, D.A., Taber, L.A., 2002. Cardiac looping in experimental conditions: effects of extraembryonic forces. Dev. Dyn. 224, 413–421. Voronov, D.A., Alford, P.W., Xu, G., Taber, L.A., 2004. The role of mechanical forces in dextral rotation during cardiac looping in the chick embryo. Dev. Biol. 272, 339–350. Wagner, M., Siddiqui, M.A., 2007. Signal transduction in early heart development (I): cardiogenic induction and heart tube formation. Exp. Biol. Med. (Maywood) 232, 852–865. Wagner, M.K., Yost, H.J., 2000. Left–right development: the roles of nodal cilia. Curr. Biol. 10, R149–R151. Waldo, K.L., Kumiski, D.H., Wallis, K.T., Stadt, H.A., Hutson, M.R., Platt, D.H., Kirby, M.L., 2001. Conotruncal myocardium arises from a secondary heart field. Development 128, 3179–3188. Warga, R.M., Stainier, D.Y., 2002. The guts of endoderm formation. Results Probl. Cell Differ. 40, 28–47. Yoder, B.K., Hou, X., Guay-Woodford, L.M., 2002. The polycystic kidney disease proteins, polycystin-1, polycystin-2, polaris, and cystin, are co-localized in renal cilia. J. Am. Soc. Nephrol. 13, 2508–2516. Yoshioka, H., Meno, C., Koshiba, K., Sugihara, M., Itoh, H., Ishimaru, Y., Inoue, T., Ohuchi, H., Semina, E.V., Murray, J.C., Hamada, H., Noji, S., 1998. Pitx2, a bicoid-type homeobox gene, is involved in a leftysignaling pathway in determination of left–right asymmetry. Cell 94, 299–305. Yost, H.J., 1990. Inhibition of proteoglycan synthesis eliminates left– right asymmetry in Xenopus laevis cardiac looping. Development 110, 865–874. Yost, H.J., 1991. Development of the left–right axis in amphibians. Ciba. Found Symp. 162, 165–176. Yost, H.J., 1992. Regulation of vertebrate left–right asymmetries by extracellular matrix. Nature 357, 158–161. Zhou, X., Sasaki, H., Lowe, L., Hogan, B.L., Kuehn, M.R., 1993. Nodal is a novel TGF-beta-like gene expressed in the mouse node during gastrulation. Nature 361, 543–547.

Chapter 4.2

Molecular Mechanisms of Left–Right Development Hiroshi Hamada Developmental Genetics Group, Graduate School for Frontier Biosciences, Osaka University, and CREST, Japan Science and Technology Corporation (JST), Osaka, Japan

I.  Overview Establishment of the three body axes – anteroposterior (AP), dorsoventral (DV), and left–right (L–R) – is central to organization of the vertebrate body plan. The left–right axis is the last of these three axes to be established during development, and the process by which left–right asymmetry is generated can be divided into three steps (Fig. 1): (1) the initial breaking of left–right symmetry, which occurs in or near the node and at the late neural-fold stage; (2) transfer of a left–right-biased

e7.5

e10.5

Figure 1  Steps underlying the generation of left–right asymmetry. Three steps that contribute to the generation of left–right asymmetry are shown: (1) symmetry breaking; (2) molecular patterning of the lateral plate mesoderm (LPM); and (3) asymmetric organogenesis. The developmental stage (E: embryonic day) corresponding to each step in the mouse is indicated on the left.

Symmetry Breaking

e8.5

e9.5

signal (or signals) from the node to the lateral plate mesoderm (LPM), which leads to left–right asymmetric expression of signaling molecules such as the transforming growth factor--related (TGF) proteins Nodal and Lefty on the left side of the lateral plate mesoderm; (3) left–right asymmetric morphogenesis of visceral organs induced by these signaling molecules (Shiratori and Hamada, 2006). Here I describe the current understanding of the mechanism of left–right patterning during mouse development (see Chapters 4.1 and 4.3 for left–right asymmetry in other organisms).

R

L

Patterning

Morphogenesis lung

Nodal, Lefty

heart

Blood vessel

e11.5

Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

297

298

PART  |  4  Asymmetry in Cardiac Development

(A)

(B)

Figure 2  Cilia and fluid flow in the node. Scanning electron micrographs showing that each cell on the ventral side of the mouse node has a monocilium (A) and a ventral view of the mouse node at lower magnification. Left–right and anteroposterior (A–P) orientations are indicated, with the red arrow denoting the leftward flow of extraembryonic fluid in (B).

II.  Symmetry breaking by cilia and fluid flow The breaking of left–right symmetry, at least in the mouse, is thought to be achieved by the unidirectional flow of extraembryonic fluid in the node, an embryonic midline structure located at the anterior tip of the primitive streak in mouse embryos (Fig. 2). This leftward laminar flow of extraembryonic fluid in the node cavity, referred to as nodal flow (Nonaka et al., 1998), occurs at a speed of 15 to 20 m/s and is generated by the rotational movement of 9  0 monocilia (cilia with nine outer doublets of microtubules but no central pair) that protrude from cells located on the ventral side of the node into the node cavity (Sulik et al., 1994) (Fig. 2). These 200 to 300 cilia rotate in the same direction (clockwise as viewed from the ventral side) at a speed of 600 rpm (Nonaka et al., 1998). Nodal flow takes place for only a short period of time: it is first apparent at the one- to two-somite stage, persists for several hours and ends by the six-somite stage. The asymmetric expression of Nodal begins in the lateral plate mesoderm at the two-somite stage and disappears by the six-somite stage. Nodal flow may therefore occur specifically to initiate Nodal expression on the left side of the lateral plate mesoderm. There is a strong correlation between nodal flow and left–right patterning. Many mutant mice that lack nodal flow because the node cilia are either missing or immotile have been identified and all such mice exhibit aberrant left–right patterning of the lateral plate mesoderm. One example of such a mouse is the iv mutant (Supp et al., 1997), in which left–right asymmetry of body organs is randomized and in which node cilia were immotile and nodal flow was absent (Okada et al., 1999). The importance of nodal flow in left–right patterning has also been examined directly by imposition of an artificial rightward flow in the mouse embryo. Such an artificial rightward flow resulted in reversal of left–right patterning of the

embryo (Nonaka et al., 2002), establishing the role of fluid flow in the development of left–right asymmetry. How is the unidirectional flow of extraembryonic fluid generated by rotational movement of the node cilia? Hundreds of cilia rotating in the same direction would be expected to generate whirls rather than a unidirectional laminar flow. Observation by high-speed video microscopy revealed that the node cilia do not protrude vertically from the node cells. Instead, they are tilted posteriorly at an average angle of 30° (Nonaka et al., 2005; Okada et al., 2005). Hydrodynamic principles predict that cilia can generate a unidirectional flow if they are tilted toward a specific direction. When the cilia move closer to the surface, the movement of fluid near the surface will be restricted as a result of the “no-slip boundary effect”. Conversely, when the cilia move away from the surface, they move the neighboring fluid more effectively. If cilia are tilted toward the posterior side, they will be moving toward the right when they come close to the surface and toward the left when they are far from the surface, thus generating a leftward flow. The notion that the posterior tilt of the node cilia generates the leftward flow is key to the origin of the left–right axis (Fig. 3). The node cilia are thought to correspond to the chiral F-molecule proposed by Brown and Wolpert (Brown and Wolpert, 1990; Brown et al., 1991). Furthermore, two pre-existing positional cues are reflected in the cilia: the AP and DV axes are thus represented by the posterior tilt and ventral protrusion of the cilia, respectively. The node cilia thus generate the leftward flow by making use of both the pre-existing positional cues and their chiral structure. In other words, the origin of the left– right axis is both the pre-existing AP and DV positional information and the chirality of the cilia. How is AP information translated into the posterior tilt of the node cilia? Given its similarity to positioning of the hair in the Drosophila wing, a mechanism resembling the planar cell polarity (PCP) pathway involving

299

Chapter  |  4.2  Molecular Mechanisms of Left–Right Development

III.  Action of nodal flow V

P

A D

3 pre-existing Information in the cilia

1. A-P: posterior tilt 2. D-V: direction of protrusion 3. Chirality of rotation

Figure 3  Origin of the left–right axis. Each node cilium (red bars on left) is posteriorly tilted, likely because the basal body (green) is posteriorly shifted within the cell (blue). The cilium protrudes from the cell toward the ventral side of the embryo and rotates in a clockwise direction when viewed from the ventral side. Anteroposterior (A–P) and dorsoventral (D–V) orientations are indicated. A schematic representation of a transverse section of a cilium, revealing its chiral structure, is shown on the right. The cilium has nine pairs of microtubules (green) as well as inner and outer arms of dynein (pink).

noncanonical Wnt signaling that is responsible for this latter process (Klein and Mlodzik, 2005) may also underlie positioning of the node cilia. However, although several mouse mutants deficient in components of the noncanonical Wnt pathway have been described, none has been found to manifest left–right defects (Wang and Nathans, 2007). Clarification of whether these findings indicate that a mechanism other than the noncanonical Wnt pathway is responsible for the posterior shift of the node cilia may have to await identification of additional genes that play a role in planar cell polarity. In any event, it will be essential to observe and discover how the node cilia are formed during development. Live imaging of node cells during development may provide important information in this regard.

(A)

Determinant molecule

Although it remains unknown how nodal flow exerts its developmental action, there are two prevailing models (Fig. 4). First, the flow may transport a determinant molecule toward the left side (determinant-transporting model) (Fig. 4A). Sonic hedgehog and retinoic acid have each been proposed to function as such a determinant, but genetic evidence in the mouse does not fully support either notion. Another candidate for the determinant is Nodal, which is produced by the perinodal cells and may be transported by the flow. Genetic evidence has established that Nodal produced at the node is essential for subsequent Nodal expression in the left lateral plate (Brennan et al., 2002; Saijoh et al., 2003). However, a recent study (Oki et al., 2007) indicates that Nodal produced at the node acts not via an external route (after being secreted into the node cavity), but rather via an internal route (after being secreted within the embryo), suggesting that Nodal is not the putative determinant transported by nodal flow. Given that other genes encoding secreted proteins are expressed in the node or perinodal region, such proteins are also candidates for the determinant. The alternative model is the mechanosensory model (Fig. 4B), according to which the flow generates mechanical stress that is sensed by node cells, either pit cells (cells with primary cilia located in the node) or crown cells (cells located in the perinodal region that express Nodal). Physical force may be detected either by nonmotile cilia present in the perinodal region, as proposed by McGrath et al. (2003) and Tabin and Vogan (2003), or by the cell membrane per se. Given the slow speed of the flow and the viscosity of the extraembryonic fluid, the shear stress generated by nodal flow might be expected to be too small to be sensed by a cell. However, the small physical force

(B)

Mechanical stress

Left Left

Determinant molecule Nodal flow Cilium Left-determinant signal, intra- or inter-cells Generation of left side-specific character Figure 4  Two models for the mechanism of action of nodal flow. (A) Determinant-transporting model. (B) Mechanosensory model. Green arrows indicate the direction of nodal flow; yellow stars denote determinant molecules.

300

generated by the flow might be amplified by intracellular signaling triggered either by the bending of nonmotile cilia, or by the cell membrane directly. Regardless of whether the determinant-transporting model or the mechanosensory model is responsible for breaking left–right symmetry, the leftward flow is eventually translated into asymmetric Ca2 signaling around the node (McGrath et al., 2003). Consistent with the importance of such a mechanism, mutant mice lacking Pkd2 (Polycystin2), a putative Ca2 channel (Luo et al., 2003), do not manifest Nodal expression in the lateral plate mesoderm (Pennekamp et al., 2002).

IV.  Transfer of an asymmetric signal from the node to the   lateral plate mesoderm The asymmetric signal (or signals) generated in or near the node, whether it is mechanical stress or a molecular determinant, must be transferred to the lateral plate mesoderm, where it induces the asymmetric expression of Nodal. Several important questions remain unanswered about this process. How, and by which route, is the signal transferred to the lateral plate mesoderm? How does the signal activate Nodal expression in the left lateral plate mesoderm? What is the nature of the signal that reaches the left lateral plate mesoderm and activates Nodal expression there? Asymmetric elevation of the cytosolic Ca2 concentration may be an intermediate event in communication between the node and the lateral plate mesoderm, but how is it related to the asymmetric Nodal expression in the lateral plate mesoderm?

IV.A.  The Route of Signal Transfer Whether the cilium-generated asymmetric signal is molecular or mechanical in nature, it needs to be sensed by cells. It is currently unknown which type of cells sense the signal: the node pit cells with their rotating cilia; perinodal cells with immotile cilia as suggested by the two-cilia model (McGrath et al., 2003; Tabin and Vogan, 2003); or endoderm cells located distant from the node.

IV.B.  Expression of Nodal, a Left-Side Determinant, Begins at the Node Many of the signaling molecules expressed at the node are essential for left–right patterning of the lateral plate, and they may play a role in the transfer of the left–right signal. Nodal is expressed bilaterally at the node (in perinodal crown cells) before the onset of its expression in the left lateral plate mesoderm. Furthermore, genetic evidence has established that Nodal expression at the node is essential for

PART  |  4  Asymmetry in Cardiac Development

subsequent Nodal expression in the left lateral plate mesoderm (Brennan et al., 2002; Saijoh et al., 2003). The specific ablation of Nodal expression in the perinodal region thus prevents Nodal expression in the left lateral plate mesoderm (Brennan et al., 2002). Cerl2, which encodes an antagonist of Nodal, is also expressed in the perinodal region before Nodal expression begins in the left lateral plate mesoderm (Marques et al., 2004). Cerl2 is expressed in an asymmetric manner, with the level of expression on the right side being substantially higher than that on the left (Pearce et al., 1999). Mice that lack Cerl2 show bilateral or right-sided expression of Nodal in the lateral plate mesoderm (Marques et al., 2004), suggesting that this Nodal antagonist produced at the node regulates the asymmetric expression of Nodal in the lateral plate mesoderm. These observations thus indicate that Nodal may play a role in signal transfer from the node to the left lateral plate mesoderm M.

IV.C.  The Nodal Signal is Transferred Directly from the Node to the Lateral   Plate Mesoderm What is the precise role of Nodal produced at the node? The Nodal signal may be relayed to the lateral plate mesoderm. Nodal produced at the node may thus act on cells located between the node and the lateral plate mesoderm (such as perinodal cells, endodermal cells located distant from the node, paraxial mesoderm cells, or intermediate mesoderm cells) and induce a secondary signal in them that travels to the lateral plate mesoderm and activates Nodal expression. In the chick embryo (Rodriguez Esteban et al., 1999; Yokouchi et al., 1999), Sonic hedgehog produced in the node activates the expression of Caronte, which encodes an inhibitor of bone morphogenetic protein (BMP) in the paraxial mesoderm. Caronte, in turn, induces Nodal expression in the left lateral plate mesoderm. However, it is not known whether a similar signaling mechanism operates in other vertebrates. A BMP antagonist corresponding to Caronte has not been identified in mouse or zebrafish genomes. Nevertheless, a BMP signal may negatively regulate Nodal expression in the lateral plate mesoderm of the mouse embryo given that, in the absence of the BMP effector Smad5 (Chang et al., 2000), Nodal is expressed bilaterally in the lateral plate mesoderm. An alternative possibility is that Nodal itself is transported from the node to the left lateral plate mesoderm. Several lines of circumstantial evidence support this possibility: 1. Nodal expression in the lateral plate mesoderm can be induced by Nodal itself. The introduction of a Nodal expression vector in the right lateral plate mesoderm was thus found to induce ectopic expression of endogenous Nodal (Yamamoto et al., 2003).

301

Chapter  |  4.2  Molecular Mechanisms of Left–Right Development

2. A search for transcriptional regulatory sequences that control asymmetric Nodal expression identified two enhancers, both of which are able to confer asymmetric gene expression in the left lateral plate mesoderm (Adachi et al., 1999; Norris and Robertson, 1999; Saijoh et al., 2000). Importantly, both enhancers possess binding sequences for the transcription factor FoxH1 that are essential for enhancer activity and are responsive to Nodal (Saijoh et al., 2005). 3. Cryptic, an essential component of the Nodal signaling pathway, is required only in the lateral plate mesoderm for correct left–right patterning (Oki et al., 2007). 4. Nodal interacts with sulfated glycosaminoglycans, which are specifically localized to the basement membrane between the node and the lateral plate. Moreover, inhibition of sulfated glycosaminoglycan synthesis prevented Nodal expression in the lateral plate mesoderm (Oki et al., 2007). 5. Another TGF-related protein expressed at the node, growth-differentiation factor 1 (GDF1), appears to play a role in the transfer of the Nodal signal from the node to the lateral plate. Like Nodal, GDF1, which shares sequence similarity with Vg1 in Xenopus, is expressed bilaterally in the perinodal region. Mice that lack GDF1 lose asymmetric Nodal expression in the lateral plate mesoderm and exhibit right isomerism of the visceral organs (Rankin et al., 2000). The similarities in the expression domains of Gdf1 and Nodal, as

(A)

well as in the respective mutant phenotypes, suggested that GDF1 might interact with Nodal. This was indeed found to be the case (Tanaka et al., 2007). GDF1 alone does not activate signaling by components of the Nodal pathway under physiological conditions; rather, GDF1 physically interacts with Nodal and increases the efficiency of Nodal signaling through its receptor complex. These lines of circumstantial evidence thus indicate that Nodal produced at the node is transferred from the node to the lateral plate, where it activates Nodal expression. However, direct detection of secreted Nodal protein has not been successful because of technical difficulties.

V.  Molecular patterning by the asymmetric signals nodal   and lefty The genes for the TGF-related proteins Nodal and Lefty are expressed asymmetrically in the lateral plate mesoderm (Fig. 5A,B) and play a major role in patterning the lateral plate mesoderm (Levin et al., 1995; Collignon et al., 1996; Lowe et al., 1996; Meno et al., 1996; Chapter 4.1). Asymmetric expression of Nodal and Lefty is transient, lasting only several hours in the mouse embryo, from the two- to sixsomite stage. Although both genes are similarly expressed

(B)

Nodal

Lefty

R

(C)

L

(D)

Situs inversus

(E)

Left isomerism

Right isomerism

Figure 5  Left–right asymmetric expression of Nodal and Lefty. Expression of Nodal (A, C–E) and Lefty (B) in developing wild-type (A, B), inv mutant (C), Lefty1/ mutant (D), or Gdf1/ mutant (E) mouse embryos. Lefty1 and Lefty2 are expressed predominantly at the midline (1) and in the left lateral plate mesoderm (2), respectively (B).

302

on the left side, they have opposite functions. Genetic and biochemical studies have thus established that Nodal acts as a left side determinant whereas Lefty is an antagonist of Nodal. In the wild-type mouse, cells in the left lateral plate mesoderm which receive the Nodal signal contribute to the left side-specific morphology of various visceral organs, whereas cells in the right lateral plate mesoderm, which do not receive the Nodal signal, contribute to right side-specific morphology. Left–right reversal of the Nodal expression domain such as occurs in the inv mutant mouse (Yokoyama et al., 1993) (Fig. 5C), results in situs inversus; the complex reversal of the asymmetry of organs such as the lungs, stomach and heart. In Lefty1/ mice, in which Nodal is expressed bilaterally in the lateral plate mesoderm (Fig. 5D), visceral organs develop a left-sided identity on both left and right sides, referred to as left isomerism (Meno et al., 1998). Conversely, when Nodal expression is lost in the lateral plate mesoderm, as in Gdf1/ mice (Fig. 5E), visceral organs adopt right isomerism (Rankin et al., 2000). Nodal is a TGF family ligand that can trigger intracellular signaling through interaction with type I (ALK4 and ALK7) and type II (ActRIIa and ActRIIb) TGF receptors (Schier and Shen, 2000). Unlike other TGF family members, however, Nodal requires an EGF-CFC (epidermal growth factor-Cripto-FRL1-Cryptic) family protein (such as Cryptic or Cripto) as a coreceptor. Smad2 and Smad3, together with Smad4, are intracellular components of the Nodal signaling pathway. Although Smad2 and Smad3 exhibit distinct activities in certain biochemical assays, genetic evidence suggests that these two proteins perform the same function in vivo. Genetically-engineered mice that lack either Smad2 or Smad3 thus develop normally (Dunn et al., 2005). Smad2/3 may interact with a large number of sequence-specific transcription factors. However, with regard to Nodal signaling during development, FoxH1 is the major (although not necessarily only) factor that interacts with Smad2/3 and transduces the Nodal signal. Lefty actually refers to two related genes, Lefty1 and Lefty2, in mammals. These genes are expressed in the lateral plate mesoderm and prospective floor plate on the left side, with Lefty1 being expressed predominantly in the prospective floor plate and Lefty2 in the lateral plate mesoderm (Fig. 5B). The predicted amino acid sequences of the Lefty proteins are unusual for TGF family members, in that they lack a cysteine residue that is required for dimer formation. Biochemical data suggest that Lefty inhibits Nodal signaling by competitively interacting with the EGF-CFC coreceptor (Chen et al., 2004) and with the type II TGF receptor chain (Sakuma et al., 2002). Studies of the transcriptional regulation of Nodal and Lefty (Adachi et al., 1999; Norris et al., 1999; Saijoh et al., 1999, 2000) have revealed that Lefty is not a simple inhibitor of Nodal signaling, but is rather a feedback inhibitor (Fig. 6). Asymmetric expression of Nodal and Lefty is regulated by enhancers known as asymmetric

PART  |  4  Asymmetry in Cardiac Development

Lefty Nodal

EGF-CFC ActR p

Smad2 Smad4

FoxH1

Nodal

Lefty2 Figure 6  Positive and negative signaling loops mediated by Nodal and Lefty. Regulatory relationships between Nodal and Lefty are summarized. The red oval in the depictions of Nodal and Lefty2 genes indicate the Nodal-responsive asymmetric enhancer (ASE).

enhancers (ASEs), which are necessary and sufficient for the asymmetric expression patterns of the genes in vivo. The ASEs of Nodal and Lefty from several vertebrates share common sequence motifs, including two copies of the sequence ATT(A/C)(A/C)ACA, which is recognized by the transcription factor FoxH1. Given that FoxH1 is a component of the Nodal signaling pathway (Hoodless et al., 2001; Yamamoto et al., 2001), these findings suggest the existence of a regulatory network consisting of Nodal and Lefty (Fig. 6). The interaction of Nodal with a target cell will thus result in the transcriptional activation of both Nodal and Lefty. The Nodal thereby produced will amplify expression of Nodal and Lefty, whereas the Lefty produced will eventually terminate their expression. The Lefty­mediated negative loop would therefore be expected to restrict the duration and area of Nodal signaling in a highly precise manner, likely explaining the fact that asymmetric expression of Nodal and Lefty is indeed transient. The activities of Nodal and Lefty proteins and the transcriptional regulatory relationship between Nodal and Lefty genes described above suggest that the two proteins might constitute a self-enhancement lateral inhibition (SELI) system (Nakamura et al., 2006), a type of reaction– diffusion system (Turing, 1953; Meinhardt and Gierer, 2000). Reaction–diffusion systems involve two diffusible molecules, one of which is an activator that stimulates both its own synthesis and the synthesis of its partner, which is an inhibitor. The model stipulates that the inhibitor ­diffuses more rapidly than the activator. The outcome of

303

Chapter  |  4.2  Molecular Mechanisms of Left–Right Development

the reaction–diffusion mechanism depends on at least five parameters: the initial rate of synthesis of the activator; the diffusion rates of the activator and the inhibitor; and the rates of degradation of the activator and the inhibitor. How might such a SELI system underlie left–right determination? It may in fact explain the mechanistic basis of the generation of random asymmetry – the second component of the original Brown and Wolpert model (Brown and Wolpert, 1990). A SELI system has the potential to amplify an initially small difference in signaling between otherwise identical cells. Consider a SELI system in which an activator is induced at two points separated by the midline (Fig. 7), for example; if the level of the activator on one side of the midline is slightly higher than that on the other side, the former side will dominate. The activator level on this side will be amplified, whereas the activator will disappear on the opposite side. Although this mechanism is itself random, the existence of a bias that predetermines which side has the initially slightly higher level of the activator would result in that side always being dominant. Such a bias might be introduced by nodal flow, which is always toward the left. The combination of the Nodal–Lefty SELI system with nodal flow might thus be capable of generating a robust all-or-nothing molecular asymmetry such as left-sided Nodal expression. The simplest version of this model, in which Nodal itself would be the determinant that is transported by the flow, can be tested experimentally by specific ablation of Nodal expression at the node. It remains to be seen whether this conceptually attractive model will fit the experimental data.

VI.  The cellular basis of asymmetric morphogenesis Various visceral organs begin to develop anatomic asymmetries in distinct manners (such as directional looping, differential lobation, or unilateral regression) only after asymmetric Nodal expression in the lateral plate mesoderm has disappeared. A main player in the regulation of asymmetric organogenesis is the transcription factor Pitx2 (Logan et al., 1998; Yoshioka et al., 1998), whose asymmetric expression is induced by Nodal (Shiratori et al., 2001; Chapter 4.3). Like Nodal and Lefty2, Pitx2 is expressed asymmetrically in left lateral plate mesoderm, but its asymmetric expression persists until much later stages of development (Fig. 8). Mice deficient in Pitx2 (or, to be more precise, Pitx2c, the isoform that is asymmetrically expressed) exhibit laterality defects in most visceral organs (Liu et al., 2001; Shiratori et al., 2006). Thus, in the absence of Pitx2c, bilateral organs such as the lungs exhibit right isomerism. How does each organ primordium interpret left–right information? How does Pitx2 regulate situs-specific morphogenesis of various organs by seemingly different ­cellular mechanisms? We still do not know which genes are regulated by this transcription factor. Recent studies have revealed the cellular mechanisms of aortic arch development in the mouse (Yashiro et al., 2007) and of unilateral regression of the ovary in the chick (Guioli and LovellBadge, 2007; Ishimaru et al., 2007). The arterial system, including the dorsal aorta and branchial arch arteries (BAAs), is initially formed symmetrically with subsequent

Nodal

1) Generate a small difference by nodal flow

Lefty

Pitx2

E8.2

2) Contert to robust asymmetry by RD/SELI system

E9.5 R

L

Figure 7  Conversion of a small molecular difference to a robust downstream molecular asymmetry by a reaction-diffusion system. Robust left–right asymmetry (asymmetric Nodal expression) in the lateral plate mesoderm may be generated in two steps. First, nodal flow generates a small difference between the two sides of the lateral plate mesoderm. This small difference is then converted to a robust asymmetry by a reaction–diffusion (RD) system comprised of Nodal and Lefty proteins.

Figure 8  Asymmetric expression of Pitx2 persists during situs­specific organogenesis. Whole-mount in situ hybridization detection of Nodal, Lefty and Pitx2 expression in mouse embryos at E8.2 and E9.5. Lefty1 and Lefty2 expression domains (the midline and left LPM, respectively) are shown in the same embryo. Asymmetric expression of Nodal and Lefty is transient, whereas that of Pitx2 persists.

304

PART  |  4  Asymmetry in Cardiac Development

Remodeling of 6th branchial arch artery The side that persists after remodeling :

L

R

L&R

Ductus caroticus 3

3

4 AS

4 6

6

L

R

RV Normal LV

Dorsal aorta

Region that undergoes regression

Pitx2 mutant mouse

Figure 9  Remodeling of branchial arch arteries (BAAs). Patterns of BAA remodeling in normal and Pitx2 mutant embryos as represented by vascular lumen morphology at E11.5. Red indicates regions that undergo complete regression (RV and LV: right and left ventricle, respectively; AS: aortic sac). Numbers indicate BAAs.

asymmetric remodeling giving rise to the aortic arch. The left-sixth BAA thus persists and gives rise to the aortic arch, whereas the right-sixth BAA regresses (Fig. 9). Ablation of unilateral Pitx2 expression impairs asymmetric remodeling of the BAA system, resulting in randomized laterality of the aortic arch. Pitx2 functions in the myocardial cells of the secondary heart field and induces a dynamic morphological change in the outflow tract of the heart which results in the provision of an asymmetric blood supply to the sixth BAA (Yashiro et al., 2007) (Fig. 9); this uneven distribution of blood flow results in differential signaling by both the platelet-derived growth factor (PDGF) receptor and the vascular endothelial growth factor (VEGF) receptor 2. The consequent stabilization of the left-sixth BAA and the regression of its right counterpart underlie left-sided formation of the aortic arch. Hemodynamics, generated by a Pitx2-induced morphological change in the outflow tract, is therefore responsible for the asymmetric remodeling of the great arteries.

VII.  Diversity among vertebrates Although the relative positions and shapes of most of the visceral organs are conserved among vertebrates, there are differences such as those in the ovary. In birds, a pair of primordial ovaries and oviducts is formed during development; after hatching, however, the left components persist whereas the right ones regress. Blood vessels also undergo remodeling in distinct manners among vertebrates. Moreover, differences are apparent even within the same class of vertebrates,

as exemplified by the lobation pattern of the lungs. The mouse thus possesses a single lobe on the left and four lobes on the right, humans and horses have two lobes on the left and three lobes on the right, dogs and rabbits have three lobes on the left and four lobes on the right. What proportion of the entire left–right determination pathway is conserved among vertebrates (Tabin, 2005)? The patterning step regulated by Nodal and Lefty seems to be conserved among vertebrates including fish, frog, chick and mouse. Nodal and Lefty are also present in invertebrates such as ascidians and sea urchin (Imai et al., 2006; Duboc and Lepage, 2008). The morphogenesis step is also conserved in the sense that it is regulated by Pitx2 (Shiratori et al., 2006). However, an organ develops left– right asymmetry in different ways in different species. The unilateral regression of the right ovary in birds is under the control of Pitx2, but it is mediated by a signaling mechanism unique to birds (Guioli and Lovell-Badge, 2007; Ishimaru et al., 2007). Brain asymmetry is most obvious in the epithalamus of the zebrafish (Halpern et al., 2003). The epithalamic asymmetries include left-sided localization of the parapineal, as well as differences in size, gene expression, and target innervation pattern between the left and right habenular nuclei (Concha et al., 2003; Gamse et al., 2005; Aizawa et al., 2007). These asymmetries are induced by left-sided expression of Nodal (as well as of Lefty and Pitx2) in the developing dorsal encephalon. The symmetry-breaking step is variable among vertebrates. Given that the cilia-generated flow has been detected in the mouse and rabbit (Okada et al., 2005), it is likely conserved among mammals. Cilia-driven flow also

Chapter  |  4.2  Molecular Mechanisms of Left–Right Development

occurs in Kupffer’s vesicle of the zebrafish embryo (Essner et al., 2002, 2005; Kramer-Zucker et al., 2005) and in the organizer of Xenopus embryos (Schweickert et al., 2007), but it has not been detected in the chick. In Xenopus, however, certain gene products have been shown to be asymmetrically distributed at much earlier stages (as early as the two-cell stage) than is observed in mouse (Levin et al., 2002; Bunney et al., 2003). This finding raises the question as to whether cilia-driven flow is the symmetry-breaking event in Xenopus. Different mechanisms may thus have evolved for the symmetry-breaking event in vertebrates.

Acknowledgments I thank current and former members of my laboratory for discussion, as well as for providing illustrations. The work performed in my laboratory has been supported by CREST, Japan Science and Technology Corporation (JST) and by grants from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

References Adachi, H., Saijoh, Y., Mochida, K., Ohishi, S., Hashiguchi, H., Hirao, A., Hamada, H., 1999. Determination of left/right asymmetric expression of nodal by a left side-specific enhancer with sequence similarity to a lefty-2 enhancer. Genes Dev. 13, 1589–1600. Aizawa, H., Goto, M., Sato, T., Okamoto, H., 2007. Temporally regulated asymmetric neurogenesis causes left–right difference in the zebrafish habenular structures. Dev. Cell 12, 87–98. Brennan, J., Norris, D.P., Robertson, E.J., 2002. Nodal activity in the node governs left–right asymmetry. Genes Dev. 16, 2339–2344. Brown, N.A., Wolpert, L., 1990. The development of handedness in left/ right asymmetry. Development 109, 1–9. Brown, N.A., McCarthy, A., Wolpert, L., 1991. Development of handed body asymmetry in mammals. Ciba Found Symp. 162, 182–196. Bunney, T.D., De Boer, A.H., Levin, M., 2003. Fusicoccin signaling reveals 14-3-3 protein function as a novel step in left–right patterning during amphibian embryogenesis. Development 130, 4847–4858. Chang, H., Zwijsen, A., Vogel, H., Huylebroeck, D., Matzuk, M.M., 2000. Smad5 is essential for left–right asymmetry in mice. Dev. Biol. 219, 71–78. Chen, S.K., Olale, F., Brivanlou, A.H., Schier, A.F., 2004. Lefty blocks a subset of TGFbeta signals by antagonizing EGF-CFC coreceptors. PLoS Biol. 2, E30. Collignon, J., Varlet, I., Robertson, E.J., 1996. Relationship between asymmetric nodal expression and the direction of embryonic turning. Nature 381, 155–158. Concha, M.L., Russel, C., Regan, J.C., Tawk, M., Sidi, S., Gilmour, D.T., Kapsimali, M., Sumoy, L., Goldstone, K., Amayam, E., Wilson, S.W., 2003. Local tissue interactions across the dorsal midline of the forebrain establish CNS laterality. Neuron 39, 423–438. Duboc, V., Lepage, T., 2008. A conserved role for the nodal signaling pathway in the establishment of dorso–ventral and left–right axes in deuterostomes. J. Exp. Zoolog. B. Mol. Dev. Evol. 310, 41–53. Dunn, N.R., Koonce, C.H., Anderson, D.C., Islam, A., Bikoff, E.K., Robertson, E.J., 2005. Mice exclusively expressing the short isoform of Smad2 develop normally and are viable and fertile. Genes Dev. 19, 152–163.

305

Essner, J.J., Vogan, K.J., Wagner, M.K., Tabin, C.J., Yost, H.J., Brueckner, M., 2002. Conserved function for embryonic nodal cilia. Nature 418, 37–38. Essner, J.J., Amack, J.D., Nyholm, M.K., Harris, E.B., Yost, H.J., 2005. Kupffer’s vesicle is a ciliated organ of asymmetry in the zebrafish embryo that initiates left–right development of the brain, heart and gut. Development 132, 1247–1260. Gamse, J.T., Kuan, Y.S., Macurak, M., Brosamle, C., Thisse, B., Thisse, C., Halpern, M.E., 2005. Directional asymmetry of the zebrafish epithalamus guides dorsoventral innervation of the midbrain targets. Development 132, 4869–4881. Guioli, S., Lovell-Badge, R., 2007. Pitx2 controls asymmetric gonadal development in both sexes of the chick and can rescue the degeneration of the right ovary. Development 134, 4199–4208. Halpern, M.E., Liang, J.O., Gamse, J.T., 2003. Leaning to the left: laterality in the zebrafish forebrain. Trends Neurosci. 26, 308–313. Hoodless, P.A., Pye, M., Chazaud, C., Labbe, E., Attisano, L., Rossant, J., Wrana, J.L., 2001. FoxH1 (Fast) functions to specify the anterior primitive streak in the mouse. Genes Dev. 15, 1257–1271. Imai, K.S., Levine, M., Satoh, N., Satou, Y., 2006. Regulatory blueprint for a chordate embryo. Science 312, 1183–1187. Ishimaru, Y., Komatsu, T., Kasahara, M., Katoh-Fukui, Y., Ogawa, H., Toyama, Y., Maekawa, M., Toshimori, K., Chandraratna, R.A.S., Morohashi, K., Yoshioka, H., 2007. Mechanism of asymmetric ovarian development in chick embryos. Development 135, 677–685. Klein, T.J., Mlodzik, M., 2005. Planar cell polarization: an emerging model points in the right direction. Annu. Rev. Cell Dev. Biol. 21, 155–176. Kramer-Zucker, A.G., Olale, F., Haycraft, C.J., Yoder, B.K., Schier, A.F., Drummond, I.A., 2005. Cilia-driven fluid flow in the zebrafish pronephros, brain and Kupffer’s vesicle is required for normal organogenesis. Development 132, 1907–1921. Levin, M., Johnson, R.L., Stern, C.D., Kuehn, M., Tabin, C., 1995. A molecular pathway determining left–right asymmetry in chick embryogenesis. Cell 82, 803–814. Levin, M., Thorlin, T., Robinson, K.R., Nogi, T., Mercola, M., 2002. Asymmetries in H/K-ATPase and cell membrane potentials comprise a very early step in left–right patterning. Cell 111, 77–89. Liu, C., Liu, W., Lu, M.F., Brown, N.A., Martin, J.F., 2001. Regulation of left–right asymmetry by thresholds of Pitx2c activity. Development 128, 2039–2048. Logan, M., Pagan-Westphal, S.M., Smith, D.M., Paganessi, L., Tabin, C.J., 1998. The transcription factor Pitx2 mediates situs-specific morphogenesis in response to left–right asymmetric signals. Cell 94, 307–317. Lowe, L.A., Supp, D.M., Sampath, K., Yokoyama, T., Wright, C.V., Potter, S.S., Overbeek, P., Kuehn, M.R., 1996. Conserved left–right asymmetry of nodal expression and alterations in murine situs inversus. Nature 381, 158–161. Luo, Y., Vassilev, P.M., Li, X., Kawanabe, Y., Zhou, J., 2003. Native polycystin 2 functions as a plasma membrane Ca2-permeable cation channel in renal epithelia. Mol. Cell Biol. 23, 2600–2607. Marques, S., Borges, A.C., Silva, A.C., Freitas, S., Cordenonsi, M., Belo, J.A., 2004. The activity of the Nodal antagonist Cerl-2 in the mouse node is required for correct L/R body axis. Genes Dev. 18, 2342–2347. McGrath, J., Somlo, S., Makova, S., Tian, X., Brueckner, M., 2003. Two populations of node monocilia initiate left–right asymmetry in the mouse. Cell 114, 61–73. Meinhardt, H., Gierer, A., 2000. Pattern formation by local self-activation and lateral inhibition. Bioessays 22, 753–760.

306

Meno, C., Saijoh, Y., Fujii, H., Ikeda, M., Yokoyama, T., Yokoyama, M., Toyoda, Y., Hamada, H., 1996. Left–right asymmetric expression of the TGF beta-family member lefty in mouse embryos. Nature 381, 151–155. Meno, C., Shimono, A., Saijoh, Y., Yashiro, K., Mochida, K., Ohishi, S., Noji, S., Kondoh, H., Hamada, H., 1998. lefty-1 is required for left–right determination as a regulator of lefty-2 and nodal. Cell 94, 287–297. Nakamura, T., Mine, N., Nakaguchi, E., Mochizuki, A., Yamamoto, M., Yashiro, K., Meno, C., Hamada, H., 2006. Generation of robust left– right asymmetry in the mouse embryo requires a self enhancement lateral inhibition system. Dev. Cell 11, 495–504. Nonaka, S., Shiratori, H., Saijoh, Y., Hamada, H., 2002. Determination of left–right patterning of the mouse embryo by artificial nodal flow. Nature 418, 96–99. Nonaka, S., Tanaka, Y., Okada, Y., Takeda, S., Harada, A., Kanai, Y., Kido, M., Hirokawa, N., 1998. Randomization of left–right asymmetry due to loss of nodal cilia generating leftward flow of extraembryonic fluid in mice lacking KIF3B motor protein. Cell 95, 829–837. Nonaka, S., Yoshiba, S., Watanabe, D., Ikeuchi, S., Goto, T., Marshall, W.F., Hamada, H., 2005. De novo formation of left–right asymmetry by posterior tilt of nodal cilia. PLoS Biol. 3, e268. Norris, D.P., Robertson, E.J., 1999. Asymmetric and node-specific nodal expression patterns are controlled by two distinct cis-acting regulatory elements. Genes Dev. 13, 1575–1588. Okada, Y., Nonaka, S., Tanaka, Y., Saijoh, Y., Hamada, H., Hirokawa, N., 1999. Abnormal nodal flow precedes situs inversus in iv and inv mice. Mol. Cell 4, 459–468. Okada, Y., Takeda, S., Tanaka, Y., Belmonte, J.C., Hirokawa, N., 2005. Mechanism of nodal flow: a conserved symmetry breaking event in left–right axis determination. Cell 121, 633–644. Oki, S., Hashimoto, R., Otani, H., Shen, M., Saijoh, Y., Hamada, H., 2007. Sulfated glycosaminoglycan is necessary for Nodal signal transmission from the node to the left lateral plate in the mouse left– right patterning. Development 134, 3893–3904. Pearce, J.J., Penny, G., Rossant, J., 1999. A mouse cerberus/Dan-related gene family. Dev. Biol. 209, 98–110. Pennekamp, P., Karcher, C., Fischer, A., Schweickert, A., Skryabin, B., Horst, J., Blum, M., Dworniczak, B., 2002. The ion channel polycystin-2 is required for left–right axis determination in mice. Curr. Biol. 12, 938–943. Rankin, C.T., Bunton, T., Lawler, A.M., Lee, S.J., 2000. Regulation of left–right patterning in mice by growth/differentiation factor-1. Nat. Genet. 24, 262–265. Rodriguez Esteban, C., Capdevila, J., Economides, A.N., Pascual, J., Ortiz, A., Izpisua Belmonte, J.C., 1999. The novel Cer-like protein Caronte mediates the establishment of embryonic left–right asymmetry. Nature 401, 243–251. Saijoh, Y., Adachi, H., Mochida, K., Ohishi, S., Hirao, A., Hamada, H., 1999. Distinct transcriptional regulatory mechanisms underlie left– right asymmetric expression of lefty-1 and lefty-2. Genes Dev. 13, 259–269. Saijoh, Y., Adachi, H., Sakuma, R., Yeo, C.Y., Yashiro, K., Watanabe, M., Hashiguchi, H., Mochida, K., Ohishi, S., Kawabata, M., Miyazono, K., Whitman, M., Hamada, H., 2000. Left–right asymmetric expression of lefty2 and nodal is induced by a signaling pathway that includes the transcription factor FAST2. Mol. Cell 5, 35–47. Saijoh, Y., Oki, S., Ohishi, S., Hamada, H., 2003. Left–right patterning of the mouse lateral plate requires nodal produced in the node. Dev. Biol. 256, 160–172.

PART  |  4  Asymmetry in Cardiac Development

Saijoh, Y., Oki, S., Tanaka, C., Nakamura, T., Adachi, H., Yan, Y.T., Shen, M.M., Hamada, H., 2005. Two nodal-responsive enhancers control left–right asymmetric expression of Nodal. Dev. Dyn. 232, 1031–1036. Sakuma, R., Ohnishi, Y., Yi,Y., Meno, C., Fujii, H., Juan, H., Takeuchi, J., Ogura, T., Li, E., Miyazono, K., Hamada, H., 2002. Inhibition of Nodal signalling by Lefty mediated through interaction with common receptors and efficient diffusion. Genes Cells 7, 401–412. Schier, A., Shen, M.M., 2000. Nodal signalling in vertebrate development. Nature 403, 385–389. Schweickert, A., Weber, T., Beyer, T., Vick, P., Bogusch, S., Feistel, K., Blum, M., 2007. Cilia-driven leftward flow determines laterality in Xenopus. Curr. Biol. 17, 60–66. Shiratori, H., Sakuma, R., Watanabe, M., Hashiguchi, H., Mochida, K., Sakai, Y., Nishino, J., Saijoh, Y., Whitman, M., Hamada, H., 2001. Twostep regulation of left–right asymmetric expression of Pitx2: initiation by nodal signaling and maintenance by Nkx2. Mol. Cell 7, 137–149. Shiratori, H., Hamada, H., 2006. Left–right axis in the mouse: from its origin to organogenesis. Development 133, 2095–2104. Shiratori, H., Yashiro, K., Shen, M., Hamada, H., 2006. Conserved regulation and role of Pitx2 in situs-specific organogenesis. Development 133, 3015–3025. Sulik, K., Dehart, D.B., Iangaki, T., Carson, J.L., Vrablic, T., Gesteland, K., Schoenwolf, G.C., 1994. Morphogenesis of the murine node and notochordal plate. Dev. Dyn. 201, 260–278. Supp, D.M., Witte, D.P., Potter, S.S., Brueckner, M., 1997. Mutation of an axonemal dynein affects left–right asymmetry in inversus viscerum mice. Nature 389, 963–966. Tabin, C., 2005. Do we know anything about how left–right asymmetry is first established in the vertebrate embryo? J. Mol. Histol. 36, 317–323. Tabin, C.J., Vogan, K.J., 2003. A two-cilia model for vertebrate left–right axis specification. Genes Dev. 17, 1–6. Tanaka, C., Sakuma, R., Nakamura, T., Hamada, H., Saijoh, Y., 2007. Long-range action of Nodal requires interaction with GDF1. Genes and Dev. 21, 3272–3282. Turing, A.M., 1953. The chemical basis of morphogenesis. Phil. Trans. Soc. B237, 37–72. Wang, Y., Nathans, J., 2007. Tissue/planar cell polarity in vertebrates: new insights and new questions. Development 134, 647–658. Yamamoto, M., Meno, C., Sakai, Y., Shiratori, H., Mochida, K., Ikawa, Y., Saijoh, Y., Hamada, H., 2001. The transcription factor FoxH1 (FAST) mediates Nodal signaling during anterior–posterior patterning and node formation in the mouse. Genes Dev. 15, 1242–1256. Yamamoto, M., Mine, N., Mochida, K., Sakai, Y., Saijoh, Y., Meno, C., Hamada, H., 2003. Nodal signaling induces the midline barrier by activating Nodal expression in the lateral plate. Development 130, 1795–17804. Yashiro, K., Shiratori, H., Hamada, H., 2007. Haemodynamics determined by a genetic programme govern asymmetric development of the aortic arch. Nature 450, 285–288. Yokouchi Y., Vogan, K.J., Pearse, R.V., 2nd, Tabin, C.J., 1999. Antagonistic signaling by Caronte, a novel Cerberus-related gene, establishes left–right asymmetric gene expression. Cell 98, 573–583. Yokoyama, T., Copeland, N.G., Jenkins, N.A., Montgomery, C.A., Elder, F.F., Overbeek, P.A., 1993. Reversal of left–right asymmetry: a situs inversus mutation. Science 260, 679–682. Yoshioka, H., Meno, C., Koshiba, K., Sugihara, M., Itoh, H., Ishimaru, Y., Inoue, T., Ohuchi, H., Semina, E.V., Murray, J.C., Hamada, H., Noji, S., 1998. Pitx2, a bicoid-type homeobox gene, is involved in a lefty-signaling pathway in determination of left–right asymmetry. Cell 94, 299–305.

Chapter 4.3

Pitx2 in Cardiac Left–Right Asymmetry and Human Disease James F. Martin1, Brad A. Amendt1 and Nigel A. Brown2 1

Institute of Biosciences and Technology, Texas A&M System Health Science Center, Houston, TX, USA Department of Basic Medical Sciences, St. George’s University of London, London, UK

2

I.  Left–right asymmetry and   heart disease In vertebrates, internal organs show morphological differences between the left and right sides (see also Chapters 4.1 and 4.2). For example, the number of pulmonary lobes is different between the left and right lungs, and the intestines have a characteristic asymmetric rotation. For many organ systems, such as the guts, left–right asymmetric morphogenesis may be necessary for efficient organ packing in a defined body cavity. Although defective left–right asymmetric morphogenesis for the guts may be compatible with life, this is not the case with the heart. The heart is unique in that left–right asymmetric morphogenesis is essential for development of a functional heart that can sustain postnatal life. The initially bilaterally-symmetrical heart tube undergoes complex morphogenetic changes so that five cardiac morphological asymmetries develop (Fig. 1): (1) dextral looping, which places the inflow ventricle to the right and the outflow ventricle to the left; (2) the concomitant shift of the atrioventricular junction to the left, which subsequently expands to the right to become more centrally placed once again later in development; (3) the remolding of the bilateral systemic venous channels to the right atrium only; (4) the asymmetrical development of the right and left pulmonary ridges to form the primary interatrial septum, placing the orifice of the pulmonary vein in the left atrium; (5) the clockwise (viewed from the ventricle) spiraling of the aortic outflow. The end result of these morphological changes is the formation of distinct left and right anatomical characteristics that have critical functions in separating the systemic Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

(deoxygenated) and pulmonic (oxygenated) circulation that is essential for postnatal life. This is evident in the normal arrangement of the cardiac inflow. The pulmonary vein carrying oxygenated blood from the lungs drains into the left atrium, while the caval veins carry deoxygenated blood from the systemic circulation to the right atrium. In human patients and experimental mice with defects in left–right asymmetry, the normal venous atrial connections are abnormal, often resulting in postnatal lethality. Other aspects of normal cardiac physiology are also dependent on left–right asymmetry signaling pathways. A major component of the conduction system, the sinoatrial node or pacemaker, is found on the right and is critical for coordinated cardiac contractility. The asymmetric location of the sinoatrial node is important for initiation of normal, coordinated cardiac contractions, as ectopic foci often result in cardiac arrhythmias. Along these lines, the pulmonary vein is often a source of ectopic conduction impulses and it is known that normal pulmonary vein development is regulated by left–right asymmetry (Haissaguerre et al., 1998; Mommersteeg et al., 2007a). In the past, clinical studies have primarily focused on laterality defects that were part of a larger syndrome. As an example, a strong link between defects in cilia function or primary cilia dyskensia and laterality defects is well-established. When primary cilia dyskensia was associated with left–right asymmetry defects it is referred to as Kartegener’s syndrome. In addition to left–right asymmetry abnormalities, Kartegener’s patients often present with respiratory infections, male infertility and diminished smell, due to abnormalities in the dynein arms of the microtubules. Thus, left–right asymmetry defects were 307

308

PART  |  4  Asymmetry in Cardiac Development

5

RV 1

LV

Figure 1  Five morphological asymmetries develop from the initially bilaterallysymmetrical heart tube: (1) dextral looping placing the inflow ventricle to the right and the outflow ventricle to the left; (2) concurrent movement of the atrioventricular junction to the left (at later stages, the AV junction expands to the right and becomes symmetrical again); (3) remolding of the bilateral systemic venous channels to the right atrium only; (4) asymmetrical development of the right and left pulmonary ridges to form the primary interatrial septum, placing the orifice of the pulmonary vein in the left atrium; (5) clockwise (viewed from the ventricle) spiraling of the aortic outflow.

4 RA

LA

2

3

recognized to be a causative factor in congenital heart disease, but usually in the context of other syndromes. Recent experiments in human genetics mapping loci responsible for familial atrial fibrillation have suggested a connection between left–right asymmetry and atrial fibrillation, a common cardiac anomaly involving abnormal cardiac rhythm afflicting a large percentage of the adult population (Gudbjartsson et al., 2007). Thus, left–right asymmetry has important consequences for the development of a normal body plan and a correctly-functioning adult heart. During development, the first outward sign of left– right asymmetry is the rightward looping of the primitive heart tube, further highlighting the central importance of left–right asymmetry in cardiac morphogenesis (Harvey, 1998; Levin and Mercola, 1998; Capdevila et al., 2000). Although defects in left–right asymmetry are closely associated with a range of severe congenital heart anomalies, such as septal and valve defects (Icardo and Sanchez de Vega, 1991; Brown and Anderson, 1999), a firm understanding of the developmental and molecular mechanisms connecting left–right asymmetry to complex cardiac morphogenesis is lacking. Insight into the genetic pathways that regulate left–right asymmetry is fundamental for an understanding of normal heart development, and for any

efforts toward early diagnosis and therapeutic intervention for complex congenital cardiac anomalies. Recent work has begun to unravel the mechanisms that function at early, intermediate and late stages of left–right asymmetric morphogenesis. The remainder of this chapter will focus on how these new insights have provided novel information about cardiac morphogenesis, and in particular will consider the Pitx2 homeobox gene.

II.  Cardiac disease and the nodallefty-pitx2 left–right asymmetry pathway For heuristic purposes, left–right asymmetry can be categorized into at least three distinct developmental events: (1) initial break in symmetry; (2) transduction of an asymmetric signal; and (3) interpretation of the asymmetric signal by the individual organ primordia (Brown and Wolpert, 1990). Mouse models of defective left–right asymmetry have played a major role in our understanding of the early events of left–right asymmetry. One such mutant, the inversus viscerum (iv) mutant, displays randomized left–right asymmetry (including randomization of heart looping).

Chapter  |  4.3  Pitx2 in Cardiac Left–Right Asymmetry and Human Disease

A breakthrough in our understanding of the initial break in symmetry came with the observation that the gene mutated in the iv/iv mouse was a dynein referred to as left–right dynein (lrd, DnaH1) (Supp et al., 1997). Moreover, targeted mutations in kinesins, components of the microtubule complex, resulted in left–right asymmetry defects (Supp et al., 2000). Importantly, these defects were associated with failure of directional flow at the mouse node. In addition to being mutated in the murine iv/iv-mutant mice, mutations in Dnah11 have also been found in Kartegener’s patients (Guichard et al., 2001; Bartoloni et al., 2002). Although still debated, the mechanism regulating the initial break in symmetry likely involves an atypical morphogen gradient, controlled by directional movement of cilia in the embryonic node. This model is referred to as the “nodal flow” mechanism to induce embryonic asymmetry. Importantly, experiments that manipulate the direction and force of fluid flow around the mouse node provide very compelling functional evidence supporting the nodal flow model (Nonaka et al., 2002). One possibility is that the output of nodal flow results in asymmetric distribution of the Nodal signaling molecule, perhaps via calcium signaling, that then induces target genes in the left lateral plate mesoderm. Nodal auto-induces short-lived Nodal transcription in left lateral plate mesoderm. Interestingly, two inhibitory molecules, Lefty1 and Lefty2, are also induced by Nodal activity as part of a mechanism to fine tune Nodal activity (Meno et al., 1998, 1999). An effector of Nodal signaling in the lateral plate is the FoxH1 forkhead transcription factor. FoxH1 conditional mutants have right isomerism, consistent with the idea that FoxH1 is an effector of Nodal signaling. Moreover, activation of Nodal expression in the right lateral plate required FoxH1 activity, indicating that FoxH1 functions to transduce the Nodal signal (Yamamoto et al., 2003). Another critical downstream Nodal effector is the Pitx2 homeodomaincontaining transcription factor (see below).

II.A.  Nodal Signaling: Mutations in   Human Patients The relevance of Nodal signaling to human disease is evident from human genetic studies. Nodal and the two inhibitory molecules, Lefty1 and Lefty2, are members of the TGF superfamily of signaling molecules. Both Lefty genes act as competitive inhibitors of Nodal action, and limit the extent of Nodal signaling in the embryo (reviewed in Hamada et al., 2002). Also involved in Nodal signaling are the small cysteine-rich proteins that contain an EGFlike motif and a cysteine-rich domain termed the Cripto, FRL-1 and Cryptic (CFC) motif. This family of EGF–CFC transmembrane proteins are co-factors for Nodal signaling, which likely function by recruiting Nodal to the Activin receptor complex (Shen, 2007).

309

Defects in the Nodal signaling pathway have been identified in human patients with left–right asymmetry defects. Mutations in the EGF–CFC gene, Cryptic, result in a complex array of left–right asymmetry defects categor­ ized as heterotaxia (Bamford et al., 2000). Four mutations were identified that were proposed to result in sensitization to other disturbances in left–right development. The identified mutations resulted in phenotypes that were generally classified as right atrial isomerism or heterotaxia (except for one patient with mutations in both Nodal and Cryptic). The mutant proteins had defects in cellular localization, as shown by transfection assays, and failed to complement a zebrafish Nodal EGF–CFC mutant, one-eyed pinhead, in functional assays. This supports the notion that Nodal signaling in left–right asymmetry is conserved through evolution. Two patients with mutations in the human Lefty related gene, Lefty A, were described. One mutation was a premature stop codon and the second a nonconservative substitution in the cysteine knot region that were both likely to result in null alleles, indicating haploinsufficiency in humans for Lefty A (Kosaki et al., 1999). Afflicted patients displayed a spectrum of cardiac anomalies, including left ventricular hypoplasia, complete atrioventricular canal with common atrioventricular valve, left ventricular outflow obstruction and defective vein morphogenesis. Moreover, since patients had left pulmonary isomerism, the defects were categorized as left isomerism.

III.  PITX2 and cardiac morphogenesis A major downstream effector of the Nodal pathway is the Pitx2 homeobox gene. The Pitx (Pituitary homeobox) family of homeobox genes containing three genes (Pitx1, Pitx2 and Pitx3) is a Bicoid-related subfamily within the larger Paired-related superfamily of homeobox genes (Schneitz et al., 1993; Gage et al., 1999b). The Pitx group has been implicated in human development, disease and evolution (Semina et al., 1996, 1998; Shapiro et al., 2004). The most extensively-studied of the three genes, Pitx2, was identified as the gene mutated in Axenfeld–Rieger Syndrome that includes ocular, tooth and anterior body wall defects as cardinal features (Semina et al., 1996; Amendt, 2000, 2005). Subsequent work on Pitx2 revealed an important function for Pitx2 in left–right asymmetry, a fundamental component of organ morphogenesis in vertebrates. Mutations in Pitx3 have been shown to cause ocular defects in human patients (Semina et al., 1998), and Pitx3 also has a role in survival of the dopaminergic neurons within the substantia nigra of mice (Nunes et al., 2003). Although Pitx1 has yet to be identified as a disease-causing gene in humans, genetic studies have implicated Pitx1 as a major gene involved in morphogenetic change in the pelvic region during evolution (Shapiro et al., 2004). Thus,

310

PART  |  4  Asymmetry in Cardiac Development

it is clear that understanding the developmental pathways regulated by the Pitx family of transcription factors will be valuable with regards to human disease, development and evolution. Studies performed in chick, mouse, zebrafish and Xenopus embryos made a firm connection between Pitx2 and the Nodal-regulated left–right asymmetry pathway. Moreover, human genetic studies hinted at a role for Pitx2 in cardiac morphogenesis. The Axenfeld–Rieger Syndrome (Semina et al., 1996) also involves defects in cardiac morphogenesis such as atrial septal defects and valvular defects. These are infrequently associated with Axenfeld–Rieger Syndrome, but nonetheless suggest that Pitx2 has a role in cardiac morphogenesis. Alternatively, it was proposed that Axenfeld–Rieger Syndrome with cardiac manifestations may be a separate entity or contiguous gene syndrome (Cunningham et al., 1998; Mammi et al., 1998; Bekir and Gungor, 2000). Work from model organisms has clarified the direct function of Pitx2 in cardiac development. The Pitx2 gene encodes three isoforms, Pitx2a, Pitx2b and Pitx2c in mice, and a fourth Pitx2 isoform, Pitx2d, has

(A)

~3.8kb

0.5kb 2

1 233bp

~9.2kb

3

56bp 137bp

also been identified in humans (Cox et al., 2002) (Fig. 2). The different isoforms are generated by a combination of alternative splicing and alternative promoter usage (Shiratori et al., 2001) (Fig. 2). Recently, a novel mRNA splice variant of Pitx2b has been identified, and it has been shown that two protein products, Pitx2c and Pitx2c, are formed from Pitx2c mRNA by alternative translation initiation (Lamba et al., 2008), although the relative abundance of these two proteins during heart development has not yet been studied. All Pitx2 isoforms share a common C-terminus but have distinct N-termini (Fig. 2). Expression studies in mice have revealed both distinct and overlapping areas of isoform expression in embryos and adults. Pitx2c is the only asymmetrically-expressed isoform, while the Pitx2a, Pitx2b and Pitx2c isoforms are co-expressed in head meso­ derm, oral ectoderm, eye, body wall and central nervous system (Kitamura et al., 1999; Smidt et al., 2000; Liu et al., 2001; Schweickert et al., 2001). Pitx2c, but not Pitx2a or Pitx2b, is expressed in hematopoietic stem cells (Degar et al., 2001). Thus, co-expression of Pitx2 isoforms was found in the three developmental fields that are most

0.2kb

0.9kb

2.5kb

4a

4b

5

6

201bp

558bp

206bp

1258bp OAR domain

HD N

(B) PITX2A

C 38

1 1

PITX2A Exons

PITX2B

2

271 5

6

N

C 1 15 1

PITX2B Exons

PITX2C

98

61

84

2

144 3

317 5

6

N

C 1

PITX2C Exons

68 91

151

4b

PITX2D

324 5

6

C

N 1 32

PITX2D Exons

4a

205 5

6

Figure 2  Diagram of Pitx2c isoforms. The four Pitx2 isoforms are depicted. In (A) the Pitx2 genomic structure is shown. The Pitx2 gene contains six exons that encode four isoforms; (B) shows the exon usage for the four isoforms. The shaded box represents the homeodomain. The hatched box represents the conserved OAR domain that has been proposed to mediate protein–protein interactions.

311

Chapter  |  4.3  Pitx2 in Cardiac Left–Right Asymmetry and Human Disease

frequently affected in Axenfeld–Rieger Syndrome patients: eyes, teeth and anterior abdominal wall. Analysis of the Pitx2 expression pattern in mice and chick embryos revealed that Pitx2c was expressed on the left side of the anterior precardiac splanchnic mesoderm, and expression persisted in the left heart tube prior to looping morphogenesis. It is notable that Pitx2 expression in left heart tube is more apparent in the chick embryo than the mouse, and likely reflects a divergence of Pitx2 function between species (Campione et al., 2001). This early expression domain of Pitx2 suggested a role for Pitx2 in looping morphogenesis (Logan et al., 1998; Ryan et al., 1998; Campione et al., 2001). Gain-of-function studies in chick embryos showed that Pitx2, when overexpressed in right lateral plate mesoderm, resulted in hearts with reversed or ambiguous situs. Moreover, Nodal misexpression in the right lateral plate of chick embryos also resulted in induction of ectopic Pitx2 expression on the right (Logan et al., 1998; Ryan et al., 1998; Campione et al., 1999). These findings suggested the existence of a linear signaling cascade with Pitx2 serving as the final effector of the left–right asymmetry pathway within each organ primordium (Harvey, 1998). In addition to a potential role for Pitx2 in cardiac looping morphogenesis, other work suggested a direct role for Pitx2 in patterning a broad range of left-sided cardiac and vascular structures. Pitx2c was expressed in the left atrium and atrioventricular canal, left outflow tract, right ventricle and interventricular myocardium. Moreover, Pitx2 was also expressed in the primary and secondary interatrial septum, left atrial appendage, left superior caval vein and pulmonary vein myocardium (Franco et al., 2000; (A)

Right (F)

Right

(B)

(C)

Left Right

Left Right creneo

Studies performed in mice have begun to unravel the complex function of Pitx2 in heart and cardiovascular morphogenesis (Kioussi et al., 2002; Liu et al., 2002; Ai et al., 2006). Inactivation of Pitx2 in mice has partially supported and helped to refine the early models for Pitx2 function in left–right asymmetry and cardiac development. Pitx2null mutant mice had multiple defects in cardiac development, but the mutant heart tube looped correctly rightward (Gage et al., 1999a; Kitamura et al., 1999; Lin et al., 1999; Lu et al., 1999). Analysis of the Pitx2-null mutant embryos uncovered the C-loop phenotype, in which the heart is found lateral to the body in what likely represents an arrest in looping morphogenesis. This phenotype is also found in other mice with left–right asymmetry defects (Harvey, 2002). This result contrasts to the phenotypes observed in chick embryos that showed randomization of cardiac looping after misexpression of Pitx2 or a Pitx2 enr fusion construct as described above (Logan et al., 1998; Ryan et al., 1998; Campione et al., 1999; Yu et al., 2001). An explanation for (E)

Left Right

Left Right (H)

Left Right

III.A.  Pitx2 Function: Evidence from   Loss-of-Function Studies in Mice

(D)

Left Right

(G)

Liu et al., 2001, 2002). Lineage tracing with a Pitx2creneo knockin allele indicated that Pitx2 descendants contributed to multiple structure in the maturing heart, including the right ventricular myocardium, interventricular septum and interatrial septum (Fig. 3). Thus, the Pitx2 expression analysis and lineage tracing pointed to a role for Pitx2 in cardiac looping morphogenesis, and also in the morphogenesis of complex cardiac and vascular structures.

Left

(I)

Left

Left Right null

Figure 3  Lineage tracing using the Pitx2 knockin allele. (A, B) LacZ staining of 10.5 dpc control (A) and Pitx2 mutant (B) and R26R reporter trans-heterozygous embryo. Arrow shows lacZ positive cells in the outflow tract that have crossed the midline (dashed line). (C, D, E) LacZ staining of 12.5–16.5 dpc embryos showing the location of the Pitx2 lineage in the heart. (F, G) Coronal sections through a 16.0 dpc pitx2 abccreneo / and R26R reporter trans-heterozygotes at slightly different dorso–ventral planes. Arrows denote LacZ positive cells in myocardium (F) and in interatrial and interventricular septum (G). (H, I) Whole mount LacZ staining of 14.5 dpc wild-type (H) and pitx2 mutant (I) and rosa 26 reporter transheterozygous embryos. Arrows denote LacZ-positive cells. Circled area in L denotes region with fewer lacZ positive cells.

312

PART  |  4  Asymmetry in Cardiac Development

the inconsistencies in these data may be that the function of Pitx2 in chicks and mice may be incompletely conserved. This is plausible, since it is now clear that many components of the left–right signaling pathway have divergent functions between species (Levin and Palmer, 2007). Pitx2-null embryos also had severe defects in arterioventricular valve formation with complete AV canal (Fig. 4). There were also defects in sinuatrial morphogenesis, including failure of primary interatrial septum outgrowth and isomerized atrial appendages (Kitamura et al., 1999; Liu et al., 2001) (Fig. 4). Arterioventricular connections were abnormal with double-outlet right ventricle and transposition of the great vessels commonly observed (Liu et al., 2001; Kioussi et al., 2002) (Fig. 4). Growth of the ventricular myocardium was also defective, resulting

in right ventricular hypoplasia (Kitamura et al., 1999; Liu et al., 2001, 2002). Thus, analysis of the Pitx2-null mutant embryos revealed a wide range of defects in mutant hearts, pointing to many potential functions for Pitx2 in heart and vascular morphogenesis.

III.B.  Pitx2, the Second Heart Field and Outflow Tract Development The initial functional studies in Pitx2-mutant mice uncovered an important role for Pitx2 in outflow tract development (Gage et al., 1999a; Kitamura et al., 1999; Lin et al., 1999; Lu et al., 1999). Outflow tract defects were primarily abnormal alignment of the great vessels, such as double-outlet

(A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

(I)

Figure 4  The Pitx2c mutant heart phenotype – scanning electron micrographs. Atria: right isomerism, note bilateral venous valves (arrows in (B) and (C); normal structure: arrow in (A); absence of remodeling of the left superior caval vein to the right atrium (normal structure: arrowhead in (A)); absence of the primary interatrial septum (arrowhead in (B)). (Note also the indent caused by the primary interatrial septum in normal venous cast (F) and its absence in mutant cast (I) (the right superior caval vein fractured at its atrial junction in (I)). Pulmonary venous connection: note the normal connection to the left atrium in the dorsal view of the venous corrosion cast in (F) (arrow) and the anomalous connection to the left of bilateral inferior caval veins in cast (I) (arrow). Atrioventricular junctions: note the normal two separate AV junctions (arrowheads in (D)) with their respective AV valves and the mutant single junction with common valve (arrowhead in (E)). Outflow tract: “anterior” aorta (arrow in E and H), compared to normal “wedged” position, arrow in (D) and (G)). (E): outflow cut beneath level of the arterial valves to show double outlet from right ventricle. Note, in the arterial corrosion casts, the lack of spiraling of the ascending aorta (arrow in (G) and (H)) around the pulmonary trunk (A  E10.5; B  E11.5, C–I  E17.5). (A–C) Ventral views onto dorsal halves of atria (remainder of the heart removed). (D and E) Superior-right lateral view of base of the heart with atria and arterial trunks removed. (G and H) Inferior-right lateral view of arterial corrosion casts; note the coronary arteries from the right aortic sinus in (G) and the two pulmonary arteries from the pulmonary trunk in (H) (the left pulmonary artery is not visible in (G)).

Chapter  |  4.3  Pitx2 in Cardiac Left–Right Asymmetry and Human Disease

right ventricle and great vessel transposition, with less frequent persistent truncus arteriosus (Fig. 4). Previously, we provided evidence that Pitx2c patterns the second heart field (see Chapter 2.2) which invades the heart after looping and contri­butes to the outflow tract and right ventricular myocardium. This proposal was based on the phenotype of the Pitx2-null embryos that had correct dextral looping of the heart tube but severe defects in cardiac morphogenesis (Liu et al., 2002). In addition, Pitx2c expression was detected in the branchial arch and splanchnic mesoderm, as well as outflow tract and RV myocardium. These are in part SHF-derived cell populations. Subsequent work has also revealed that atrial myocardium that expresses Pitx2c is also derived from the second heart field (Cai et al., 2003; Buckingham et al., 2005). Other studies, performed prior to the full appreciation of the second heart field, suggested a role for Pitx2 in cardiac neural crest (Hamblet et al., 2002; Kioussi et al., 2002). However, recently published data indicate that Pitx2 functions in second heart field-derived outflow tract myocardium to regulate arterioventricular alignment. Conditional gene inactivation and fate mapping studies showed that Pitx2 functions in the outflow tract myocardium derived from second heart field, but was dispensable in the cardiac neural crest (Ai et al., 2006). The conditional gene-targeting data indicate that removal of Pitx2 from cardiac neural crest results in a normal outflow tract, while deletion from the second heart field phenocopies the Pitx2-null mutant outflow tract phenotype. Moreover, germline inactivation of Pitx2a, the proposed cardiac neural crest-specific Pitx2 isoform, also resulted in a normal outflow tract (Ai et al., 2006). Together, these data support the notion that Pitx2 functions in second heart field-derived cells to regulate outflow tract development (Buckingham et al., 2005). An important insight into the genetic pathways regulating Pitx2 in outflow tract development came from the observation of a strong genetic interaction between Pitx2 and Disheveled 2 (Dvl2), a component of Wnt signaling pathways (Hamblet et al., 2002; Kioussi et al., 2002). In addition, in vitro experiments indicate that -catenin interacts directly with Pitx2 at the protein–protein level to regulate gene expression (Kioussi et al., 2002; Amen et al., 2007). Moreover, in mice with a tissue-specific deletion of -catenin in the second heart field, expression of Pitx2c is reduced by approximately 50%, based on real time polymerase chain reaction assays (Lin et al., 2007). Together, these data strongly support the idea that Pitx2 intersects, perhaps at multiple levels, with Wnt signaling in the second heart field. Other data supporting this idea includes the finding that Pitx2 expression in the outflow tract is severely reduced in Dvl2-mutant embryos (Hamblet et al., 2002). The observation that Dvl2 is expressed in outflow tract myocardium, a second heart field derivative, provides supportive evidence that the Pitx2–Wnt pathway functions in outflow tract myocardium (Phillips et al., 2005).

313

III.C.  The Role of the Pitx2-Mediated   Left–Right Asymmetry Pathway in Outflow Tract Growth Pitx2-expressing cells are located in left second heart field precursor cells of the posterior pericardial wall and in left outflow tract myocardium (Liu et al., 2002; Lickert et al., 2004; Takeuchi et al., 2005). At early stages, the second heart field is correctly specified and second heart field-derived cells move into the outflow tract of Pitx2null mutant embryos, indicating that these processes are Pitx2-independent. Moreover, transgene rescue experiments indicated that Pitx2c organ expression was necessary to rescue the Pitx2-mutant phenotype (Shiratori et al., 2006). Transgenes that only expressed Pitx2c in lateral plate mesoderm failed to rescue the Pitx2-mutant phenotype. However, once the second heart field populates the outflow tract, Pitx2 is necessary for proper expansion of the proximal outflow tract myocardium. In the absence of Pitx2, the outflow tract is shorter and misaligned. Cell proliferation studies revealed that in Pitx2-null embryos the outflow tract myocardium had reduced proliferation compared to controls. These findings indicated that cell proliferation in the left outflow tract myocardium has a critical role in outflow tract lengthening. It should be noted that, although the current evidence suggests that Pitx2 is dispensable in second heart field progenitors, it is possible that Pitx2 has an unrecognized function in them. Further experiments are required to investigate this rigorously. Based on the observations described above, it is conceivable that the left–right asymmetry pathway provides an extra source of cells that function to increase outflow tract cellularity. However, it has also been shown that ablation of the right-sided second heart field in chick embryos also resulted in a shortened outflow tract, resulting in alignment defects and pulmonary atresia (Ward et al., 2005). Together, these data indicate that regulation of myocardial expansion is not specific for the left-sided Nodal-Lefty2Pitx2 genetic pathway. Rather, second heart field cells on both sides of the embryo move into the outflow tract and contribute to outflow tract lengthening. In addition to lengthening, the outflow tract undergoes a characteristic spiraling motion (Ward et al., 2005) (see Chapter 3.1). It may be that the Pitx2-mediated, left-specific pathway is important to direct this spiraling morphogenesis. Consistent with this idea, recent experiments have shown that failure of outflow tract spiraling in Pitx2-mutants results in hemodynamic alterations that disrupt branchial arch artery remodeling (Yashiro et al., 2007). While the underlying mechanisms controlling outflow tract spiraling are unknown, recent findings indicate that Pitx2c plays an important role in regulating the adhesive properties of left-sided cells during gut looping morphogenesis (Davis et al., 2008). Further experimentation is required to investigate the possibility that Pitx2 regulates similar cellular

314

events in outflow tract spiraling. The role of Pitx2 in spiraling morphogenesis cannot be rigorously evaluated in the Pitx2-null embryos because of the shortened outflow tract. It will be critical to develop mouse reagents to investigate the role of Pitx2 in spiraling outflow tract morphogenesis. One method to accomplish this would be to perform a stage-specific inactivation of Pitx2 to investigate Pitx2 during stages of spiraling morphogenesis. This strategy has been used effectively to carefully dissect Tbx1 function during development (Xu et al., 2005).

III.C.i.  The Role of Pitx2 in Outflow Tract Remodeling Previous work suggested sequential activities for Pitx2 in outflow tract remodeling (Liu et al., 2002). This is a difficult problem to investigate, due to the early function of Pitx2c in outflow tract growth. Nonetheless, there is evidence that Pitx2 has a function in outflow tract remodeling. Lineage tracing experiments in Pitx2-mutant embryos indicated that Pitx2-mutant daughter cells failed to populate the myocardium overlying the proximal aorta efficiently (Ai et al., 2006). This deficient contribution to the aortic myocardium suggested the idea that Pitx2 regulates circumferential expansion of outflow tract myocardium. Recent lineage tracing experiments identified oriented circumferential cell growth in the outflow tract and RV myocardium (Meilhac et al., 2004b). However, it is also possible that Pitx2 regulates local movement of outflow tract myocardium or outflow tract rotation, rather than growth of outflow tract myocardium (Lomonico et al., 1986; Bostrom and Hutchins, 1988). Indeed, recent findings using a novel transgene and Di-I labeling to mark the outflow tract myocardium have implicated Pitx2c in rotation of the outflow tract myocardium (Bajolle et al., 2006). To circumvent the early role of Pitx2 in outflow tract growth, chimera experiments were conducted. Analysis of Pitx2-mutant cells in chimeric embryos indicated that Pitx2-null mutant daughters were competent to contribute to the whole circumference of the proximal outflow tract (Ai et al., 2006). This finding contrasts with the data from the germline-null mutant, and has two likely interpretations: (1) Pitx2 may regulate the expression of a signaling molecule that is required for contribution of Pitx2-mutant daughters to the whole outflow tract myocardium. Support for this model has been obtained by the recent finding that Wnt11 is a direct target of Pitx2 in the outflow tract (Zhou et al., 2007); (2) it is possible that in germline-mutant embryos, Pitx2-mutant daughter cells contribute to the aortic myocardium inefficiently because of the early proliferative defect. In the context of a chimera, adequate numbers of Pitx2-mutant daughter cells may escape the early proliferative deficiency and then expand to contribute to the proximal outflow tract completely.

PART  |  4  Asymmetry in Cardiac Development

In addition to the findings regarding circumferential growth, the chimera data uncovered a novel phenotype that indicates a cell autonomous function for Pitx2. In both high- and low-percentage chimeras, Pitx2-mutant daughter cells accumulated in the ventral proximal myocardium between the great vessels at the ventriculo–arterial junction. Even in low percentage chimeras with normal outflow tract anatomy, the accumulation of Pitx2-mutant daughters was still observed in the proximal outflow tract between the great vessels (Ai et al., 2006). These findings indicate that Pitx2 has an autonomous function in remodeling of the outflow tract myocardium. It is known that apoptosis has an important role in outflow tract remodel­ ing (Sugishita et al., 2004). It is possible that Pitx2 may promote apoptosis at later stages of outflow tract development. Alternatively, it is possible that Pitx2 regulates changes in cell shape or adhesive properties of proximal outflow tract myocardium. Support for this comes from experiments performed in Hela cells, where overexpression of Pitx2 enhanced cell spreading through regulation of small GTPase activity by transcriptional regulation of the Trio guanine nucleotide exchange factor, and investigations of Pitx2c in gut looping (Wei and Adelstein, 2002; Davis et al., 2008).

III.C.ii.  Pitx2 and Noncanonical Wnt Signaling in the Outflow Tract The noncanonical Wnt signaling pathway is -catenin independent, but utilizes Dsh and Rho (Park and Moon, 2002). The Loop-tail (Lp) mouse mutant is a naturally occurring mouse mutant that was initially identified based on severe defects in neural tube closure (Phillips et al., 2005). The gene mutated in Lp is Vangl2, a homolog of the Drosophila planar cell polarity gene, Strabismus. Recent work has shown that Vangl2 functions in the outflow tract myocardium to regulate outflow tract septation. Oriented cell division has been observed in the outflow tract myocardium and, based on computer modeling, has been suggested to result from oriented mitosis (Meilhac et al., 2004a,b). Noncanonical Wnt signaling pathways are known to regulate convergent extension and cell intercal­ ation, as well as oriented cell division in the gastrulating zebrafish embryo (Gong et al., 2004). Along these lines, Pitx2 has been directly connected to the noncanonical Wnt signaling pathway. Recent findings showed that Wnt11 that primarily activates the noncanonical pathway is a direct target of Pitx2 in outflow tract (Zhou et al., 2007).

III.C.iii.  Summary: Pitx2 in Outflow Tract Development Substantial progress has recently been made in our understanding of outflow tract development (Webb et al., 2003;

Chapter  |  4.3  Pitx2 in Cardiac Left–Right Asymmetry and Human Disease

Buckingham et al., 2005). The second heart field functions as a source of myocardium for a rapidly-expanding outflow tract. Because the outflow tract undergoes a spiraling morphogenetic movement, defects in the second heart field or second heart field-derived myocardium would be predicted to result in abnormal great vessel alignment such as double-outlet right ventricle or great vessel transposition (Yutzey and Kirby, 2002; Ward et al., 2005). Recent experiments in chick embryos have also revealed that the second heart field also makes a late contribution to smooth muscle cells of the great vessels (Waldo et al., 2005). The second heart field contributes cells to the outflow tract endocardium, indicating that another function for the second heart field may be to regulate outflow tract valve morphogenesis. The extensive contribution of the second heart field to outflow tract development makes this group of cells a high priority for future study. Molecular analysis has defined both Isl1-Gata and Foxh1-Nkx2-5 pathways that regulate Mef2c expression in the second heart field subpopulation of the second lineage (Dodou et al., 2004; von Both et al., 2004). Moreover, there is evidence that Tbx1 interacts with Fgf and Pitx2-mediated pathways to regulate second heart field development (Xu et al., 2004; Nowotschin et al., 2006). Importantly, the Isl1-Mef2c pathway that is confined to the anterior-most region of the second lineage reveals that genetic regulation of the second lineage is subdivided into anterior and posterior fields.

III.D.  Pitx2 and Tbx1 in Second Heart Field Tbx1, the T-box containing gene that has been implicated in the DiGeorge microdeletion syndrome (see Chapter 9.4), is co-expressed with Pitx2 in the second heart field, but is also expressed in the pharyngeal endoderm. Recent work uncovered a strong genetic interaction between Pitx2 and Tbx1 in cardiac development (Nowotschin et al., 2006). Compound heterozygotes for Pitx2-null and Tbx1-null alleles had incompletely penetrant cardiac defects, including double-outlet right ventricle and atrial septal defects. This observation was supported with in situ data indicating that Pitx2 expression was downregulated in the Tbx1-null mutant second heart field. Moreover, Tbx1 was shown to be transiently expressed in the left second heart field, although this observation needs to be clarified by other groups. To dissect the mechanisms underlying these observations, the authors found that Tbx1 binds to a half recognition element within the Pitx2 gene, indicating that Tbx1 directly regulates Pitx2 transcription. The strength of this proposed model is the strong genetic interaction between Pitx2 and Tbx1. Moreover, Tbx1 and Pitx2 have been independently proposed to regulate cell proliferation in the second heart field. However, caution should be exercised before fully accepting the notion that Pitx2 is a direct target

315

of Tbx1. It is possible that reduced Pitx2 expression in Tbx1-null embryos may be indirect. The in vivo significance of the T-box half site in the Pitx2 gene will need to be tested using transgenic analysis, as has been done for the Nkx and Fox binding elements. Moreover, it is known that Fgf signaling is disrupted in Tbx1-mutants, and that Fgf signaling can induce Pitx2 expression in craniofacial development (St Amand et al., 2000). Also, there is other evidence that Tbx1 function resides completely in the pharyngeal endoderm (Arnold et al., 2006). Since Pitx2c function has been traced to second heart field rather than pharyngeal endoderm, a direct interaction between the two genes would seem less likely. Other models for the genetic and functional relationship between Tbx1 and Pitx2 have been proposed. In developing branchiomeric muscle, chromatin immunoprecipitation (ChIP) analysis indicated that Pitx2 directly regulated Tbx1 expression (Shih et al., 2007). Further experiments are required to address the genetic relationship between Pitx2 and Tbx1.

III.D.i.  Pitx2 in Branchiomeric Muscle:   A Subpopulation of the Second Heart Field Pitx2 is expressed in multiple muscle types including extraocular muscle, branchial arch or branchiomeric muscle, cardiac muscle and trunk skeletal muscle (Kitamura et al., 1999; Ai et al., 2006). The function of Pitx2 in branchiomeric muscle is relevant to a discussion of Pitx2 in cardiac development, because the second heart field contributes to branchiomeric muscle (Kelly et al., 2001). The in vivo function of Pitx2 in trunk skeletal muscle is poorly understood. Previous experiments investigating Pitx2 in the C2C12 myoblast cell line, derived from satellite cells of the adult leg, uncovered a direct role for Pitx2 in regulating myoblast proliferation through a mechanism mediated by the Nterminus of Pitx2a (Kioussi et al., 2002). In the heart, Pitx2 regulates proliferation of cardiomyocytes of the outflow tract (Ai et al., 2006). In extraocular muscle, it has been suggested that Pitx2 may directly regulate muscle-specific transcription factors such as Myogenin (Diehl et al., 2006). In branchiomeric muscle, Pitx2 regulates undifferentiated precursor cells and may control expression of genes that are involved in muscle expansion and survival. Recent experiments revealed a role for the bHLH encoding genes MyoR and Capsulin in the survival of a subset of first branchial muscle precursors (Lu et al., 2002). The Pitx2-null mutant branchiomeric muscle precursors fail to express MyoR and undergo apoptosis. It is notable that there is evidence in the pituitary that Pitx2 and the related factor, Pitx1, promote cell survival by regulating expression of Lhx3 (Charles et al., 2005; Zhao et al., 2006). In addition, the third member of the Pitx family, Pitx3, is required for postnatal survival of midbrain dopaminergic neurons (van den Munckhof et al., 2003). The requirement for Pitx2 in undifferentiated precursor cells contrasts to the

316

function of Pitx2 in asymmetric organ morphogenesis. In left–right organ morphogenesis, Pitx2 activity is needed in the organ primordium, rather than in undifferentiated precursors (Ai et al., 2006; Shiratori et al., 2006). This may reflect a difference in tissues that only express Pitx2c.

III.D.ii.  Pitx2 and Tbx1 in Branchiomeric Muscle Tbx1 mutants have sporadic failure of craniofacial muscle development with loss of Tlx1 and Fgf10 expression (Kelly et al., 2004). Moreover, Tbx1, as noted above, has been suggested to directly activate Pitx2 in the second cardiac lineage by binding to an element upstream of exon 6 (Nowotschin et al., 2006). Tbx1 was still expressed in undifferentiated cells of the Pitx2-null mutant branchial arch core mesoderm, consistent with the notion that Tbx1 is an upstream regulator of Pitx2. However, Pitx2 was still expressed in Tbx1-null mutants indicating that a simple epistatic relationship is unlikely. In addition, in contrast to Pitx2-null mutant embryos, Tbx1 mutants continue to express MyoR in the branchiomeric progenitors further arguing against a linear, epistatic relationship (Kelly et al., 2004). As noted above, ChIP analysis indicated that Pitx2 directly binds to the Tbx1 gene, suggesting that Pitx2 regulates Tbx1 transcription (Shih et al., 2007). Alternatively, it may be that Pitx2 and Tbx1 may converge on common target genes to regulate transcription. Pitx and Tbx genes have been shown to coordinately regulate gene expression in the pituitary. Pitx1 and Tpit (Tbx19) bind to closely spaced but distinct recognition elements in the POMC promoter (Lamolet et al., 2001). In this system, Pitx1 synergized with Tbx19, but failed to transcriptionally synergize with Tbx1, suggesting that cell type specific co-factors may be required for any potential synergism between Pitx2 and Tbx1 in branchiomeric muscle progenitors (Lamolet et al., 2001). In addition, recent experiments have uncovered a direct protein– protein interaction between Pitx2 and Tbx1. Tbx1 physically interacts with the Pitx2 C-terminal tail and acts to repress Pitx2 transcriptional activation of several gene promoters (Amendt, unpublished data). Thus, Tbx1 may actively repress Pitx2 function in specific cell lineages. It is notable that in the zebrafish mutant van gogh, that carries a mutant allele of Tbx1, muscle expression of endothelin 1 (edn1) is reduced (Piotrowski et al., 2003). In Pitx2-null mutants, edn1 expression is reduced in the oral ectoderm, suggesting the possibility that the Tbx1- and Pitx2-mediated pathway may converge on edn1 (Liu et al., 2003).

III.D.iii.  Splanchnic Mesoderm Contribution   to Branchiomeric Muscle Similar to that which has been described for the cardiac outflow tract, multiple lineages with distinct developmental

PART  |  4  Asymmetry in Cardiac Development

histories contribute to branchiomeric muscle. In the heart, the primary heart field contributes to the linear heart tube while the second lineage is sequestered and moves into the outflow tract at a later stage. The later addition of second heart field is required for proper outflow tract lengthening and morphogenesis (Buckingham et al., 2005). In branchio­ meric muscle, the addition of cells from multiple lineages may play a similar role in controlling the size and pattern of craniofacial muscle.

III.D.iv.  Summary: Pitx2 in Branchiomeric Muscle Recent work has uncovered a requirement for Pitx2 in branchiomeric muscle development and provided insight into the genetic pathways controlling the development of branchiomeric muscle. In Pitx2-null embryos, branchiomeric muscle precursors were initially present, but failed to expand and activate the myogenic program. Moreover, lack of MyoR expression and elevated apoptosis indicated a defect in survival of undifferentiated muscle progenitor cells. Conditional Pitx2 inactivation, overexpression and knockdown in chick primary cultures supported a direct role for Pitx2 in branchiomeric muscle development. A direct regulatory connection between Tbx1 and Pitx2 has been proposed, but further experiments are necessary to investigate this more thoroughly (Dong et al., 2006; Shih et al., 2007).

IV.  Pathways regulating pitx2 expression Recent data from the Hamada laboratory showed that Cripto homozygous mutant embryos have loss of asymmetric Pitx2 expression, solidifying the notion that Pitx2 is a target of Nodal signaling (Shiratori et al., 2006) (see Chapter 4.1 and 4.2). Moreover, transgenic analysis in mice indicated that asymmetric Pitx2 transcription was regulated by a conserved asymmetric element within the Pitx2 gene just upstream of the last exon (Shiratori et al., 2001). Interestingly, the asymmetric element is comprised of a bifunctional element containing Nkx and Fox recognition elements. The Fox element binds FoxH1, a Nodal signaling effector, and was necessary for induction of Pitx2 transcription in left lateral plate mesoderm. Maintenance of Pitx2 expression in individual organs was dependent on the Nkx element, indicating that two sequential events were required for complete, asymmetric Pitx2 transcription. Importantly, asymmetric element deletion largely recapitulated the phenotype of the Pitx2c-specific deletion, with the loss of most Pitx2c expression domains. The expression of Pitx2c in the left cardinal and vitelline veins is asymmetric element-independent, but is still Nodaldependent since these expression domains were lost in the

Chapter  |  4.3  Pitx2 in Cardiac Left–Right Asymmetry and Human Disease

Cripto homozygous mutant embryos. This indicates that there are other, redundant elements in the Pitx2 gene that respond to Nodal signaling. While the evidence strongly supports the conclusion that Pitx2c expression in the left lateral plate mesoderm is dependent on Nodal signaling through the Fox recognition element, this does not rule out the idea that other signaling pathways are also involved in asymmetric Pitx2 regulation. Indeed, the observation that the Nkx element is necessary for Pitx2 expression within left organ primordia implicates the Bmp signaling pathway in Pitx2 regulation, since there is evidence that Nkx2-5 is a direct target of Bmp signaling (Liberatore et al., 2002; Lien et al., 2002; Brown et al., 2004). It is notable that Nkx2-5-null mutant embryos still express a Pitx2 asymmetric element-driven transgene in the heart, although the levels of transgene expression were not investigated (Prall et al., 2007). This suggests that other Nkx factors may redundantly regulate Pitx2c expression in the heart. Alternatively, the Nkx element may be a quantitative element that is required to maintain adequate Pitx2 levels in the heart. Other findings revealed that Pitx2 and Nkx2-5 (see Chapter 9.1) cooperatively regulate common downstream target genes. For example, Pitx2 and Nkx2-5 increase Pitx2 transcriptional activation of the procollagen lysyl hydroxylase (PLOD) promoter (Ganga et al., 2003). Moreover, the Pitx2/Nkx2-5 complex synergistically regulated the atrial naturietic factor promoter. Furthermore, Pitx2 is a potent activator of its own promoter and in concert with Nkx2-5 can synergistically activate the Pitx2c promoter in multiple cell types, suggesting that Nkx2-5 and Pitx2 function in a feed-forward loop is required to maintain Pitx2 expression in the developing heart (Amendt, unpublished data). This potential mechanism may be general, since Gata4 can physically interact with Pitx2 to increase Pitx2 transcriptional activity (Amendt, unpublished data). Consistent with the notion that Bmp signaling regulates Pitx2 expression embryos with a conditional Bmp4 deletion had reduced Pitx2 expression in the outflow tract myocardium (Liu et al., 2004). Further experiments are required to determine if this regulation is direct, through Smads, or indirect through Nkx2-5. In embryos deficient for Baf 60, a component of the Swi/Snf chromatin remodeling complex, Pitx2 expression was also lost in outflow tract myocardium (see Chapter 10.1). Biochemical analysis indicated that the Swi/Snf remodeling complex failed to interact with Smad factors, but was able to enhance activation of a Pitx2 reporter by -catenin and TCF4 supporting the notion that Pitx2 outflow tract transcription is regulated by the canonical wnt signaling pathway, but may be independent of Smad factors (Lickert et al., 2004). Other data also indicate that Pitx2 is regulated by canonical Wnt signaling in pituitary, skeletal muscle and second heart field. Conditional inactivation of -catenin in second heart field using the Isl1cre driver resulted in a

317

50% reduction in Pitx2 transcription as determined by real time polymerase chain reaction assay (Lin et al., 2007). In skeletal muscle and pituitary cell lines, a Wnt-responsive element was identified in the Pitx2a 5 flanking region (Kioussi et al., 2002). Data from transgenic embryos uncovered a TCF/LEF recognition element, which would respond to canonical Wnt signaling in the Pitx2 3 flanking region that was necessary for Pitx2c expression in Rathke’s pouch (Ai et al., 2007). Interestingly, this work uncovered a cooperative regulation of Pitx2c transcription between Nuclear factor 1 (NF-1) and canonical Wnt signaling. The NF-1 family of transcription factors bind GC-rich elements and are thought to be broadly expressed (Gronostajski, 2000). Based on the transgenic analysis, it was proposed that NF-1 may potentiate the ability of Wnt signaling to activate Pitx2c transcription (Fig. 5). This suggests one mechanism to regulate the levels of Pitx2 transcription in a cell-type-specific manner. The Wnt signaling pathway also plays a critical role in regulating Pitx2 expression through a direct interaction of Pitx2 and -catenin. Pitx2 and -catenin can positively regulate gene expression through their direct physical interaction (Vadlamudi et al., 2005; Amen et al., 2007). The Pitx2/-catenin complex can synergistically activate the Pitx2c promoter at high levels (Amendt, unpublished data). Pitx2 also interacts with Lef-1 to synergistically activate the full-length Lef-1 isoform associated with cell proliferation (Amen et al., 2007). The Lef-1/Pitx2 complex can synergistically activate the Pitx2c promoter (Amendt, unpublished data). Thus, Pitx2 interactions with multiple factors allow it to auto-regulate its expression in different cell types.

V.  PITX Genes and transcriptional regulation To gain deeper insight into the molecular mechanisms underlying Pitx2 in left–right asymmetry, it is critical to have a detailed understanding of Pitx2 target genes. In work performed in C2C12 skeletal myoblasts, it was shown that Pitx2 is a target for Wnt signaling and a cofactor for -catenin that functions as an effector molecule for Wnt signaling (Kioussi et al., 2002). In the presence of growth factors, Pitx2 was recruited to the CyclinD2 regulatory elements, where it functioned as a repressor by recruiting HDAC1. Upon induction of Wnt signaling, -catenin was stabilized and recruited to the Cyclin D2 regulatory elements by interacting with Pitx2. Binding of -catenin by direct protein–protein interaction with Pitx2 resulted in HDAC1 dismissal and recruitment of histone acetyl transferases such as TIP60. This model provides insight into how Pitx2 may switch from a repressor to an activator, depending on cell context, by regulating covalent histone modification. Moreover, the idea that Pitx2 binds

318

PART  |  4  Asymmetry in Cardiac Development

Wnt

Cytoplasm Nucleus NF-1

Cytoplasm Nucleus

βcat TCF/ LEF

NF-1

TCF/ LEF

X

Pitx2c transcription

Wnt

Cytoplasm Nucleus NF-1

βcat TCF/ LEF

Pitx2c transcription attenuated

Figure 5  Model depicting a mechanism regulating cell-type-specific Pitx2c transcription. In the developing pituitary, NF-1 binding and canonical Wnt signaling are required for optimal Pitx2 expression. A possible function for NF-1 in remodeling the chromatin in close proximity to the TCF/ LEF binding element is proposed. In the presence of Wnt signaling and an intact NF-1 binding site, high level Pitx2c transcription is achieved. In the absence of the NF-1 site, Pitx2c can still be activated by Wnt signaling, but to a lesser extent.

directly to the CyclinD2 gene in vivo has also been shown in the pancreas, lending further support to the notion that Pitx2 regulates Cyclin D2 in vivo (Rulifson et al., 2007). A recent report has demonstrated a novel mechanism for the control of Pitx2 transcriptional activity and provides a model for the recruitment of homeodomain transcription factors to transcriptionally active chromatin. HMG-17 is a ubiquitously expressed chromatin-associated high mobility group protein that constantly shuttles in and out of the nucleus. HMG-17 interacts with Pitx2 and inhibits Pitx2 from binding DNA. Thus, HMG-17 recruits Pitx2 to active sites of chromatin by binding to acetylated histone H4, and sequesters Pitx2 in an inactive conformation. Upon Wnt/-catenin signaling, -catenin forms a complex with HMG-17 and Pitx2, which yields a potent transcriptional complex on the chromatin associated with acetylated histone H4. Thus, HMG-17 acts as a molecular switch to inhibit Pitx2 activity until -catenin interacts with the complex and changes the HMG-17/Pitx2 complex to a HMG17/Pitx2/-catenin activator complex (Amen et al., 2008).

The positioning of Pitx2 to active sites of chromatin transcription allows for the direct and quick activation of Pitx2 target genes required for the rapid increase in gene expression during development. The proposal that Pitx2 is a direct target for Wnt signaling in the pituitary has been validated in transgenic mouse experiments that uncovered a TCF/LEF binding site in a Pitx2 enhancer element that directs reporter activity in the pituitary (Ai et al., 2007). Similar experiments are required to determine whether Pitx2 is directly regulated by Wnt signaling in the heart. Recent work investigating the mechanism of Pitx1 transcription on the POMC regulatory elements provides further insight into possible mechanisms of Pitx2 transcriptional regulation in left–right asymmetry. Recent work has shown that Pitx1 directly interacts with Brg1, a component of the ATPase subunit of the Swi/Snf chromatin remodeling complex (Bilodeau et al., 2006). This work indicates that, in addition to covalent histone modification, Pitx2 may also regulate transcription by interacting with

Chapter  |  4.3  Pitx2 in Cardiac Left–Right Asymmetry and Human Disease

nuclesome remodeling complexes (Roberts and Orkin, 2004). This hypothesis is particularly interesting in view of the know importance of Swi/Snf complexes in outflow tract development (Lickert et al., 2004).

V.A.  Pitx2 in the Inflow Tract and Atrial Fibrillation The initial characterization of the Pitx2-mutant embryos revealed severe defects in left atrial morphogenesis that were subsequently characterized as right atrial isomerism (Kitamura et al., 1999; Lu et al., 1999; Liu et al., 2001). The Pitx2-mutant atria were morphologically right, with characteristics such as bilateral venous valves (Liu et al., 2001). This work has recently been extended to show that Pitx2c-mutant embryos had bilateral sinoatrial nodes at the sinoatrial junction (Mommersteeg et al., 2007b). Together these data suggest that Pitx2 functionally represses a default, right-sided genetic program. It will be interesting to determine if Pitx2 directly represses genes such as Shox2 and Tbx3 that are known to regulate sinoatrial node development (Blaschke et al., 2007; Hoogaars et al., 2007) (see Chapter 2.3). These observations fit nicely with the recent data implicating Pitx2 in atrial fibrillation, and indicate that left–right asymmetry pathways play a major role in atrial fibrillation (Gudbjartsson et al., 2007).

VI.  Conclusions and future considerations Remarkable advances in our understanding of cardiac disease and left–right asymmetry have been made in recent years. Early models held that a linear signaling cascade asymmetrically patterned all developing organs (Levin and Mercola, 1998). However, study of Pitx2 regulation has revealed that multiple pathways, such as Wnt and Bmp signaling, likely feed into left–right asymmetry morphogenesis. One challenge in the future will be to determine the mechanisms of signaling pathway convergence and over­-lap. Moreover, since Pitx2 is a major effector of the Nodal-mediated pathway and is critical for normal cardiac development, it will be important to gain more comprehensive insight into Pitx2 target genes. The finding that the second heart field is asymmetrically patterned by Pitx2 provides the opportunity to dissect the genetic pathways regulating second heart field development. It will be important to understand how Pitx2 regulates transcription in the heart by interacting with histone-modifying and chromatin-remodeling complexes.

References Ai, D., Liu, W., Ma, L., Dong, F., Lu, M.F., Wang, D., Verzi, M.P., Cai, C., Gage, P.J., Evans, S., Black, B.L., Brown, N.A., Martin, J.F., 2006.

319

Pitx2 regulates cardiac left-right asymmetry by patterning second cardiac lineage-derived myocardium. Dev. Biol. 296, 437–449. Ai, D., Wang, J., Amen, M., Lu, M.F., Amendt, B.A., Martin, J.F., 2007. Nuclear factor 1 and T-cell factor/LEF recognition elements regulate Pitx2 transcription in pituitary development. Mol. Cell Biol. 27, 5765–5775. Amen, M., Liu, X., Vadlamudi, U., Elizondo, G., Diamond, E., Engelhardt, J.F., Amendt, B.A., 2007. PITX2 and beta-catenin interactions regulate Lef-1 isoform expression. Mol. Cell Biol. 27, 7560–7573. Amen, M., Espinoza, H.M., Cox, C., Liang, X., Wang, J., Link, T.M., Brennan, R.G., Martin, J.F., Amendt, B.A., 2008. Chromatinassociated HMG-17 is a major regulator of homeodomain transcription factor activity modulated by Wnt/-catenin signaling. Nucleic Acids Res. 36 (2), 462–476. Amendt, B.A., 2005. The molecular and biochemical basis of AxenfeldRieger Syndrome. In: Amendt, B.A. (Ed.), The Molecular Mechanisms of Axenfeld-Rieger Syndrome. R. G. Landes Co, Austin, TX (Chapter 4). Amendt, B.A., Semina, E.V., Alward, W.L., 2000. Rieger syndrome: a clinical, molecular and biochemical analysis. Cell Mol. Life Sci. (Review) 57, 1652–1666. Arnold, J.S., Werling, U., Braunstein, E.M., Liao, J., Nowotschin, S., Edelmann, W., Hebert, J.M., Morrow, B.E., 2006. Inactivation of Tbx1 in the pharyngeal endoderm results in 22q11DS malformations. Development 133, 977–987. Bajolle, F., Zaffran, S., Kelly, R.G., Hadchouel, J., Bonnet, D., Brown, N.A., Buckingham, M.E., 2006. Rotation of the myocardial wall of the outflow tract is implicated in the normal positioning of the great arteries. Circ. Res. 98, 421–428. Bamford, R.N., Roessler, E., Burdine, R.D., Saplakoglu, U., dela Cruz, J., Splitt, M., Goodship, J.A., Towbin, J., Bowers, P., Ferrero, G.B., Marino, B., Schier, A.F., Shen, M.M., Muenke, M., Casey, B., 2000. Loss-of-function mutations in the EGF-CFC gene CFC1 are associated with human left–right laterality defects. Nat. Genet. 26, 365–369. Bartoloni, L., Blouin, J.L., Pan, Y., Gehrig, C., Maiti, A.K., Scamuffa, N., Rossier, C., Jorissen, M., Armengot, M., Meeks, M., Mitchison, H.M., Chung, E.M., Delozier-Blanchet, C.D., Craigen, W.J., Antonarakis, S.E., 2002. Mutations in the DNAH11 (axonemal heavy chain dynein type 11) gene cause one form of situs inversus totalis and most likely primary ciliary dyskinesia. Proc Natl Acad Sci USA 99, 10282–10286. Bekir, N.A., Gungor, K., 2000. Atrial septal defect with interatrial aneurysm and Axenfeld–Rieger syndrome. Acta Ophthalmol. Scand. 78, 101–103. Bilodeau, S., Vallette-Kasic, S., Gauthier, Y., Figarella-Branger, D., Brue, T., Berthelet, F., Lacroix, A., Batista, D., Stratakis, C., Hanson, J., Meij, B., Drouin, J., 2006. Role of Brg1 and HDAC2 in GR trans-repression of the pituitary POMC gene and misexpression in Cushing disease. Genes Dev. 20, 2871–2886. Blaschke, R.J., Hahurij, N.D., Kuijper, S., Just, S., Wisse, L.J., Deissler, K., Maxelon, T., Anastassiadis, K., Spitzer, J., Hardt, S.E., Schöler, H., Feitsma, H., Rottbauer, W., Blum, M., Meijlink, F., Rappold, G., Gittenberger-de Groot, A.C., 2007. Targeted mutation reveals essential functions of the homeodomain transcription factor Shox2 in sinoatrial and pacemaking development. Circulation 115, 1830–1838. Bostrom, M.P., Hutchins, G.M., 1988. Arrested rotation of the outflow tract may explain double-outlet right ventricle. Circulation 77, 1258–1265.

320

Brown 3rd, C.O., Chi, X., Garcia-Gras, E., Shirai, M., Feng, X.H., Schwartz, R.J., 2004. The cardiac determination factor, Nkx2-5, is activated by mutual cofactors GATA-4 and Smad1/4 via a novel upstream enhancer. J. Biol. Chem. 279, 10659–10669. Brown, N.A., Wolpert, L., 1990. The development of handedness in left/ right asymmetry. Development 109, 1–9. Brown, N.A., Anderson, R.H., 1999. Symmetry and laterality in the human heart: developmental implications. In: Harvey, R.P., Rosenthal, N. (Eds.), Heart Development. Academic Press, San Diego, CA, pp. 447–462. Buckingham, M., Meilhac, S., Zaffran, S., 2005. Building the mammalian heart from two sources of myocardial cells. Nat. Rev. Genet. 6, 826–835. Cai, C.L., Liang, X., Shi, Y., Chu, P.H., Pfaff, S.L., Chen, J., Evans, S., 2003. Isl1 identifies a cardiac progenitor population that proliferates prior to differentiation and contributes a majority of cells to the heart. Dev. Cell 5, 877–889. Campione, M., Steinbeisser, H., Schweickert, A., Deissler, K., van Bebber, F., Lowe, L.A., Nowotschin, S., Viebahn, C., Haffter, P., Kuehn, M.R., Blum, M., 1999. The homeobox gene Pitx2: mediator of asymmetric left–right signaling in vertebrate heart and gut looping. Development 126, 1225–1234. Campione, M., Ros, M.A., Icardo, J.M., Piedra, E., Christoffels, V.M., Schweickert, A., Blum, M., Franco, D., Moorman, A.F., 2001. Pitx2 expression defines a left cardiac lineage of cells: evidence for atrial and ventricular molecular isomerism in the iv/iv mice. Dev. Biol. 231, 252–264. Capdevila, J., Vogan, K.J., Tabin, C.J., Izpisua Belmonte, J.C., 2000. Mechanisms of left–right determination in vertebrates. Cell 101, 9–21. Charles, M.A., Suh, H., Hjalt, T.A., Drouin, J., Camper, S.A., Gage, P.J., 2005. PITX genes are required for cell survival and Lhx3 activation. Mol. Endocrinol. 19, 1893–1903. Cox, C.J., Espinoza, H.M., McWilliams, B., Chappell, K., Morton, L., Hjalt, T.A., Semina, E.V., Amendt, B.A., 2002. Differential regulation of gene expression by PITX2 isoforms. J. Biol. Chem. 277, 25001–25010. Cunningham Jr., E.T., Eliott, D., Miller, N.R., Maumenee, I.H., Green, W.R., 1998. Familial Axenfeld–Rieger anomaly, atrial septal defect, and sensorineural hearing loss: a possible new genetic syndrome. Arch. Ophthalmol. 116, 78–82. Davis, N.M., Kurpios, N.A., Sun, X., Gros, J., Martin, J.F., Tabin, C.J., 2008. Gut looping is initiated by left-right asymmetric changes in the architecture of the dorsal mesentery. Dev. Cell in press. Degar, B.A., Baskaran, N., Hulspas, R., Quesenberry, P.J., Weissman, S.M., Forget, B.G., 2001. The homeodomain gene Pitx2 is expressed in primitive hematopoietic stem/progenitor cells but not in their differentiated progeny. Exp. Hematol. 29, 894–902. Diehl, A.G., Zareparsi, S., Qian, M., Khanna, R., Angeles, R., Gage, P.J., 2006. Extraocular muscle morphogenesis and gene expression are regulated by Pitx2 gene dose. Invest. Ophthalmol. Vis. Sci. 47, 1785–1793. Dodou, E., Verzi, M.P., Anderson, J.P., Xu, S.M., Black, B.L., 2004. Mef2c is a direct transcriptional target of ISL1 and GATA factors in the anterior heart field during mouse embryonic development. Development 131, 3931–3942. Dong, F., Sun, X., Liu, W., Ai, D., Klysik, E., Lu, M.F., Hadley, J., Antoni, L., Chen, L., Baldini, A., Francis-West, P., Martin, J.F., 2006. Pitx2 promotes development of splanchnic mesoderm-derived branchiomeric muscle. Development 133, 4891–4899.

PART  |  4  Asymmetry in Cardiac Development

Franco, D., Campione, M., Kelly, R., Zammit, P.S., Buckingham, M., Lamers, W.H., Moorman, A.F., 2000. Multiple transcriptional domains, with distinct left and right components, in the atrial chambers of the developing heart. Circ. Res. 87, 984–991. Gage, P.J., Suh, H., Camper, S.A., 1999a. Dosage requirement of Pitx2 for development of multiple organs. Development 126, 4643–4651. Gage, P.J., Suh, H., Camper, S.A., 1999b. The bicoid-related Pitx gene family in development. Mamm. Genome 10, 197–200. Ganga, M., Espinoza, H.M., Cox, C.J., Morton, L., Hjalt, T.A., Lee, Y., Amendt, B.A., 2003. PITX2 isoform-specific regulation of atrial natriuretic factor expression: synergism and repression with Nkx2.5. J. Biol. Chem. 278, 22437–22445. Gong, Y., Mo, C., Fraser, S.E., 2004. Planar cell polarity signalling controls cell division orientation during zebrafish gastrulation. Nature 430, 689–693. Gronostajski, R.M., 2000. Roles of the NFI/CTF gene family in transcription and development. Gene 249, 31–45. Gudbjartsson, D.F., Arnar, D.O., Helgadottir, A., Gretarsdottir, S., Holm, H., Sigurdsson, A., Jonasdottir, A., Baker, A., Thorleifsson, G., Kristjansson, K., et al., 2007. Variants conferring risk of atrial fibrillation on chromosome 4q25. Nature 448, 353–357. Guichard, C., Harricane, M.C., Lafitte, J.J., Godard, P., Zaegel, M., Tack, V., Lalau, G., Bouvagnet, P., 2001. Axonemal dynein intermediatechain gene (DNAI1) mutations result in situs inversus and primary ciliary dyskinesia (Kartagener syndrome). Am. J. Hum. Genet. 68, 1030–1035. Haissaguerre, M., Jais, P., Shah, D.C., Takahashi, A., Hocini, M., Quiniou, G., Garrigue, S., Le Mouroux, A., Le Metayer, P., Clementy, J., 1998. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N. Engl. J. Med. 339, 659–666. Hamada, H., Meno, C., Watanabe, D., Saijoh, Y., 2002. Establishment of vertebrate left–right asymmetry. Nat. Rev. Genet. 3, 103–113. Hamblet, N.S., Lijam, N., Ruiz-Lozano, P., Wang, J., Yang, Y., Luo, Z., Mei, L., Chien, K.R., Sussman, D.J., Wynshaw-Boris, A., 2002. Dishevelled 2 is essential for cardiac outflow tract development, somite segmentation and neural tube closure. Development 129, 5827–5838. Harvey, R.P., 1998. Links in the left/right axial pathway. Cell 94, 273–276. Harvey, R.P., 2002. Patterning the vertebrate heart. Nat. Rev. Genet. 3, 544–556. Hoogaars, W.M., Engel, A., Brons, J.F., Verkerk, A.O., de Lange, F.J., Wong, L.Y., Bakker, M.L., Clout, D.E., Wakker, V., Barnett, P., Ravesloot, J.H., Moorman, A.F., Verheijck, E.E., Christoffels, V.M., 2007. Tbx3 controls the sinoatrial node gene program and imposes pacemaker function on the atria. Genes Dev. 21, 1098–1112. Icardo, J.M., Sanchez de Vega, M.J., 1991. Spectrum of heart malformations in mice with situs solitus, situs inversus, and associated visceral heterotaxy. Circulation 84, 2547–2558. Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1, 435–440. Kelly, R.G., Jerome-Majewska, L.A., Papaioannou, V.E., 2004. The del22q11.2 candidate gene Tbx1 regulates branchiomeric myogenesis. Hum. Mol. Genet. 13, 2829–2840. Kioussi, C., Briata, P., Baek, S.H., Rose, D.W., Hamblet, N.S., Herman, T., Ohgi, K.A., Lin, C., Gleiberman, A., Wang, J., Brault, V., Ruiz-Lozano, P., Nguyen, H.D., Kemler, R., Glass, C.K., Wynshaw-Boris, A., Rosenfeld, M.G., 2002. Identification of a Wnt/Dvl/beta-Catenin -- Pitx2 pathway mediating cell-type-specific proliferation during development. Cell 111, 673–685.

Chapter  |  4.3  Pitx2 in Cardiac Left–Right Asymmetry and Human Disease

Kitamura, K., Miura, H., Miyagawa-Tomita, S., Yanazawa, M., Katoh-Fukui, Y., Suzuki, R., Ohuchi, H., Suehiro, A., Motegi, Y., Nakahara, Y., Kondo, S., Yokoyama, M., 1999. Mouse Pitx2 deficiency leads to anomalies of the ventral body wall, heart, extra- and periocular mesoderm and right pulmonary isomerism. Development 126, 5749–5758. Kosaki, K., Bassi, M.T., Kosaki, R., Lewin, M., Belmont, J., Schauer, G., Casey, B., 1999. Characterization and mutation analysis of human LEFTY A and LEFTY B, homologues of murine genes implicated in left-right axis development. Am. J. Hum. Genet. 64, 712–721. Lamba, P., Hjalt, T.A., Bernard, D.J., 2008. Novel forms of Paired-like homeodomain transcription factor 2 (PITX2): generation by alternative translation initiation and mRNA splicing. BMC Mol. Biol. 28, 9–31. Lamolet, B., Pulichino, A.M., Lamonerie, T., Gauthier, Y., Brue, T., Enjalbert, A., Drouin, J., 2001. A pituitary cell-restricted T box factor, Tpit, activates POMC transcription in cooperation with Pitx homeoproteins. Cell 104, 849–859. Levin, M., Mercola, M., 1998. The compulsion of chirality: toward an understanding of left–right asymmetry. Genes Dev. 12, 763–769. Levin, M., Palmer, A.R., 2007. Left–right patterning from the inside out: widespread evidence for intracellular control. Bioessays 29, 271–287. Liberatore, C.M., Searcy-Schrick, R.D., Vincent, E.B., Yutzey, K.E., 2002. Nkx-2.5 gene induction in mice is mediated by a Smad consensus regulatory region. Dev. Biol. 244, 243–256. Lickert, H., Takeuchi, J.K., Von Both, I., Walls, J.R., McAuliffe, F., Adamson, S.L., Henkelman, R.M., Wrana, J.L., Rossant, J., Bruneau, B.G., 2004. Baf60c is essential for function of BAF chromatin remodelling complexes in heart development. Nature 432, 107–112. Lien, C.L., McAnally, J., Richardson, J.A., Olson, E.N., 2002. Cardiacspecific activity of an Nkx2-5 enhancer requires an evolutionarily conserved Smad binding site. Dev. Biol. 244, 257–266. Lin, C.R., Kioussi, C., O’Connell, S., Briata, P., Szeto, D., Liu, F., Izpisua-Belmonte, J.C., Rosenfeld, M.G., 1999. Pitx2 regulates lung asymmetry, cardiac positioning and pituitary and tooth morphogenesis. Nature 401, 279–282. Lin, L., Cui, L., Zhou, W., Dufort, D., Zhang, X., Cai, C.L., Bu, L., Yang, L., Martin, J., Kemler, R., Rosenfeld, M.G., Chen, J., Evans, S.M., 2007. Beta-catenin directly regulates Islet1 expression in cardiovascular progenitors and is required for multiple aspects of cardiogenesis. Proc. Natl. Acad. Sci. USA 104, 9313–9318. Liu, C., Liu, W., Lu, M.F., Brown, N.A., Martin, J.F., 2001. Regulation of left–right asymmetry by thresholds of Pitx2c activity. Development 128, 2039–2048. Liu, C., Liu, W., Palie, J., Lu, M.F., Brown, N.A., Martin, J.F., 2002. Pitx2c patterns anterior myocardium and aortic arch vessels and is required for local cell movement into atrioventricular cushions. Development 129, 5081–5091. Liu, W., Selever, J., Lu, M.F., Martin, J.F., 2003. Genetic dissection of pitx2 in craniofacial development uncovers new functions in branchial arch morphogenesis, late aspects of tooth morphogenesis, and cell migration. Development 130, 6375–6385. Liu, W., Selever, J., Wang, D., Lu, M.F., Moses, K.A., Schwartz, R.J., Martin, J.F., 2004. Bmp4 signaling is required for outflow-tract septation and branchial-arch artery remodeling. Proc. Natl. Acad. Sci. USA 101, 4489–4494. Logan, M., Pagan-Westphal, S.M., Smith, D.M., Paganessi, L., Tabin, C.J., 1998. The transcription factor Pitx2 mediates situs-specific morpho­ genesis in response to left–right asymmetric signals. Cell 94, 307–317.

321

Lomonico, M.P., Moore, G.W., Hutchins, G.M., 1986. Rotation of the junction of the outflow tract and great arteries in the embryonic human heart. Anat. Rec. 216, 544–549. Lu, J.R., Bassel-Duby, R., Hawkins, A., Chang, P., Valdez, R., Wu, H., Gan, L., Shelton, J.M., Richardson, J.A., Olson, E.N., 2002. Control of facial muscle development by MyoR and capsulin. Science 298, 2378–2381. Lu, M.F., Pressman, C., Dyer, R., Johnson, R.L., Martin, J.F., 1999. Function of Rieger syndrome gene in left–right asymmetry and craniofacial development. Nature 401, 276–278. Mammi, I., De Giorgio, P., Clementi, M., Tenconi, R., 1998. Cardiovascular anomaly in Rieger Syndrome: heterogeneity or contiguity? Acta. Ophthalmol. Scand. 76, 509–512. Meilhac, S.M., Esner, M., Kelly, R.G., Nicolas, J.F., Buckingham, M.E., 2004a. The clonal origin of myocardial cells in different regions of the embryonic mouse heart. Dev. Cell 6, 685–698. Meilhac, S.M., Esner, M., Kerszberg, M., Moss, J.E., Buckingham, M.E., 2004b. Oriented clonal cell growth in the developing mouse myocardium underlies cardiac morphogenesis. J. Cell Biol. 164, 97–109. Meno, C., Shimono, A., Saijoh, Y., Yashiro, K., Mochida, K., Ohishi, S., Noji, S., Kondoh, H., Hamada, H., 1998. Lefty-1 is required for left–right determination as a regulator of lefty-2 and nodal. Cell 94, 287–297. Meno, C., Gritsman, K., Ohishi, S., Ohfuji, Y., Heckscher, E., Mochida, K., Shimono, A., Kondoh, H., Talbot, W.S., Robertson, E.J., Schier, A.F., Hamada, H., 1999. Mouse Lefty2 and zebrafish antivin are feedback inhibitors of nodal signaling during vertebrate gastrulation. Mol. Cell 4, 287–298. Mommersteeg, M.T., Brown, N.A., Prall, O.W., de Gier-de Vries, C., Harvey, R.P., Moorman, A.F., Christoffels, V.M., 2007a. Pitx2c and Nkx2-5 are required for the formation and identity of the pulmonary myocardium. Circ. Res. 101, 902–909. Mommersteeg, M.T., Hoogaars, W.M., Prall, O.W., de Gier-de Vries, C., Wiese, C., Clout, D.E., Papaioannou, V.E., Brown, N.A., Harvey, R.P., Moorman, A.F., Christoffels, V.M., 2007b. Molecular pathway for the localized formation of the sinoatrial node. Circ. Res. 100, 354–362. Nonaka, S., Shiratori, H., Saijoh, Y., Hamada, H., 2002. Determination of left–right patterning of the mouse embryo by artificial nodal flow. Nature 418, 96–99. Nowotschin, S., Liao, J., Gage, P.J., Epstein, J.A., Campione, M., Morrow, B.E., 2006. Tbx1 affects asymmetric cardiac morphogenesis by regulating Pitx2 in the secondary heart field. Development 133, 1565–1573. Nunes, I., Tovmasian, L.T., Silva, R.M., Burke, R.E., Goff, S.P., 2003. Pitx3 is required for development of substantia nigra dopaminergic neurons. Proc. Natl. Acad. Sci. USA 100, 4245–4250. Park, M., Moon, R.T., 2002. The planar cell-polarity gene stbm regulates cell behaviour and cell fate in vertebrate embryos. Nat. Cell Biol. 4, 20–25. Phillips, H.M., Murdoch, J.N., Chaudhry, B., Copp, A.J., Henderson, D.J., 2005. Vangl2 acts via RhoA signaling to regulate polarized cell movements during development of the proximal outflow tract. Circ. Res. 96, 292–299. Piotrowski, T., Ahn, D.G., Schilling, T.F., Nair, S., Ruvinsky, I., Geisler, R., Rauch, G.J., Haffter, P., Zon, L.I., Zhou, Y., Foott, H., Dawid, I.B., Ho, R.K., 2003. The zebrafish van gogh mutation disrupts tbx1, which is involved in the DiGeorge deletion syndrome in humans. Development 130, 5043–5052. Prall, O.W., Menon, M.K., Solloway, M.J., Watanabe, Y., Zaffran, S., Bajolle, F., Biben, C., McBride, J.J., Robertson, B.R., Chaulet, H.,

322

et al., 2007. An nkx2-5/bmp2/smad1 negative feedback loop controls heart progenitor specification and proliferation. Cell 128, 947–959. Roberts, C.W., Orkin, S.H., 2004. The SWI/SNF complex – chromatin and cancer. Nat. Rev. Cancer 4, 133–142. Rulifson, I.C., Karnik, S.K., Heiser, P.W., ten Berge, D., Chen, H., Gu, X., Taketo, M.M., Nusse, R., Hebrok, M., Kim, S.K., 2007. Wnt signaling regulates pancreatic beta cell proliferation. Proc. Natl. Acad. Sci. USA 104, 6247–6252. Ryan, A.K., Blumberg, B., Rodriguez-Esteban, C., Yonei-Tamura, S., Tamura, K., Tsukui, T., de la Pena, J., Sabbagh, W., Greenwald, J., Choe, S., Norris, D.P., Robertson, E.J., Evans, R.M., Rosenfeld, M.G., Izpisúa Belmonte, J.C., 1998. Pitx2 determines left–right asymmetry of internal organs in vertebrates. Nature 394, 545–551. Shen, M.M., 2007. Nodal signaling: developmental roles and regulation. Development 134, 1023–1034. Schneitz, K., Spielmann, P., Noll, M., 1993. Molecular genetics of aristaless, a prd-type homeo box gene involved in the morphogenesis of proximal and distal pattern elements in a subset of appendages in Drosophila. Genes Dev. 7, 114–129. Schweickert, A., Steinbeisser, H., Blum, M., 2001. Differential gene expression of Xenopus Pitx1, Pitx2b and Pitx2c during cement gland, stomodeum and pituitary development. Mech. Dev. 107, 191–194. Semina, E.V., Reiter, R., Leysens, N.J., Alward, W.L., Small, K.W., Datson, N.A., Siegel- Bartelt, J., Bierke-Nelson, D., Bitoun, P., Zabel, B.U., Carey, J.C., Murray, J.C., 1996. Cloning and characterization of a novel bicoid-related homeobox transcription factor gene, RIEG, involved in Rieger syndrome. Nat. Genet. 14, 392–399. Semina, E.V., Ferrell, R.E., Mintz-Hittner, H.A., Bitoun, P., Alward, W.L., Reiter, R.S., Funkhauser, C., Daack-Hirsch, S., Murray, J.C., 1998. A novel homeobox gene PITX3 is mutated in families with autosomaldominant cataracts and ASMD. Nat. Genet. 19, 167–170. Shapiro, M.D., Marks, M.E., Peichel, C.L., Blackman, B.K., Nereng, K.S., Jonsson, B., Schluter, D., Kingsley, D.M., 2004. Genetic and developmental basis of evolutionary pelvic reduction in threespine sticklebacks. Nature 428, 717–723. Shih, H.P., Gross, M.K., Kioussi, C., 2007. Cranial muscle defects of Pitx2 mutants result from specification defects in the first branchial arch. Proc. Natl. Acad. Sci. USA 104, 5907–5912. Shiratori, H., Sakuma, R., Watanabe, M., Hashiguchi, H., Mochida, K., Sakai, Y., Nishino, J., Saijoh, Y., Whitman, M., Hamada, H., 2001. Twostep regulation of left–right asymmetric expression of Pitx2: initiation by nodal signaling and maintenance by Nkx2. Mol. Cell 7, 137–149. Shiratori, H., Yashiro, K., Shen, M.M., Hamada, H., 2006. Conserved regulation and role of Pitx2 in situs-specific morphogenesis of visceral organs. Development 133, 3015–3025. Smidt, M.P., Cox, J.J., van Schaick, H.S., Coolen, M., Schepers, J., van der Kleij, A.M., Burbach, J.P., 2000. Analysis of three Ptx2 splice variants on transcriptional activity and differential expression pattern in the brain. J. Neurochem. 75, 1818–1825. St Amand, T.R., Zhang, Y., Semina, E.V., Zhao, X., Hu, Y., Nguyen, L., Murray, J.C., Chen, Y., 2000. Antagonistic signals between BMP4 and FGF8 define the expression of Pitx1 and Pitx2 in mouse toothforming anlage. Dev. Biol. 217, 323–332. Sugishita, Y., Watanabe, M., Fisher, S.A., 2004. The development of the embryonic outflow tract provides novel insights into cardiac differentiation and remodeling. Trends Cardiovasc. Med. 14, 235–241. Supp, D.M., Witte, D.P., Potter, S.S., Brueckner, M., 1997. Mutation of an axonemal dynein affects left–right asymmetry in inversus viscerum mice. Nature 389, 963–966. Supp, D.M., Potter, S.S., Brueckner, M., 2000. Molecular motors: the driving force behind mammalian left–right development. Trends Cell Biol. 10, 41–45.

PART  |  4  Asymmetry in Cardiac Development

Takeuchi, J.K., Mileikovskaia, M., Koshiba-Takeuchi, K., Heidt, A.B., Mori, A.D., Arruda, E.P., Gertsenstein, M., Georges, R., Davidson, L., Mo, R., Hui, C.C., Henkelman, R.M., Nemer, M., Black, B.L., Nagy, A., Bruneau, B.G., 2005. Tbx20 dose-dependently regulates transcription factor networks required for mouse heart and motoneuron development. Development 132, 2463–2474. Vadlamudi, U., Espinoza, H.M., Ganga, M., Martin, D.M., Liu, X., Engelhardt, J.F., Amendt, B.A., 2005. PITX2, beta-catenin and LEF1 interact to synergistically regulate the LEF-1 promoter. J. Cell Sci. 118, 1129–1137. van den Munckhof, P., Luk, K.C., Ste-Marie, L., Montgomery, J., Blanchet, P.J., Sadikot, A.F., Drouin, J., 2003. Pitx3 is required for motor activity and for survival of a subset of midbrain dopaminergic neurons. Development 130, 2535–2542. von Both, I., Silvestri, C., Erdemir, T., Lickert, H., Walls, J.R., Henkelman, R.M., Rossant, J., Harvey, R.P., Attisano, L., Wrana, J.L., 2004. Foxh1 is essential for development of the anterior heart field. Dev. Cell 7, 331–345. Waldo, K.L., Hutson, M.R., Ward, C.C., Zdanowicz, M., Stadt, H.A., Kumiski, D., Abu-Issa, R., Kirby, M.L., 2005. Secondary heart field contributes myocardium and smooth muscle to the arterial pole of the developing heart. Dev. Biol. 281, 78–90. Ward, C., Stadt, H., Hutson, M., Kirby, M.L., 2005. Ablation of the secondary heart field leads to tetralogy of Fallot and pulmonary atresia. Dev. Biol. 284, 72–83. Webb, S., Qayyum, S.R., Anderson, R.H., Lamers, W.H., Richardson, M.K., 2003. Septation and separation within the outflow tract of the developing heart. J. Anat. 202, 327–342. Wei, Q., Adelstein, R.S., 2002. Pitx2a expression alters actin-myosin cytoskeleton and migration of HeLa cells through Rho GTPase signaling. Mol. Biol. Cell 13, 683–697. Xu, H., Morishima, M., Wylie, J.N., Schwartz, R.J., Bruneau, B.G., Lindsay, E.A., Baldini, A., 2004. Tbx1 has a dual role in the morphogenesis of the cardiac outflow tract. Development 131, 3217–3227. Xu, H., Cerrato, F., Baldini, A., 2005. Timed mutation and cell-fate mapping reveal reiterated roles of Tbx1 during embryogenesis, and a crucial function during segmentation of the pharyngeal system via regulation of endoderm expansion. Development 132, 4387–4395. Yamamoto, M., Mine, N., Mochida, K., Sakai, Y., Saijoh, Y., Meno, C., Hamada, H., 2003. Nodal signaling induces the midline barrier by activating Nodal expression in the lateral plate. Development 130, 1795–1804. Yashiro, K., Shiratori, H., Hamada, H., 2007. Haemodynamics determined by a genetic programme govern asymmetric development of the aortic arch. Nature 450, 285–288. Yu St, X., Amand, T.R., Wang, S., Li, G., Zhang, Y., Hu, Y.P., Nguyen, L., Qiu, M.S., Chen, Y.P., 2001. Differential expression and functional analysis of Pitx2 isoforms in regulation of heart looping in the chick. Development 128, 1005–1013. Yutzey, K.E., Kirby, M.L., 2002. Wherefore heart thou? Embryonic origins of cardiogenic mesoderm. Dev. Dyn. 223, 307–320. Zhao, Y., Morales, D.C., Hermesz, E., Lee, W.K., Pfaff, S.L., Westphal, H., 2006. Reduced expression of the LIM-homeobox gene Lhx3 impairs growth and differentiation of Rathke’s pouch and increases cell apoptosis during mouse pituitary development. Mech. Dev. 123 (8), 605–613. Zhou, W., Lin, L., Majumdar, A., Li, X., Zhang, X., Liu, W., Etheridge, L., Shi, Y., Martin, J., Van de Ven, W., Kaartinen, V., WynshawBoris, A., McMahon, A.P., Rosenfeld, M.G., Evans, S.M., 2007. Modulation of morphogenesis by noncanonical Wnt signaling requires ATF/CREB family-mediated transcriptional activation of TGFbeta2. Nat. Genet. 39, 1225–1234.

Chapter 5.1

Epicardial Lineage: Origins and Fates Takashi Mikawa1 and Thomas Brand2 1

Cardiovascular Research Institute, University of California, San Francisco, CA, USA Heart Science Centre, National Heart & Lung Institute, Imperial College London, United Kingdom

2

I.  Introduction The epicardium is a mesothelium forming the outermost layer of the heart. Epicardial cells arise from an extra­cardiac embryonic tissue (the proepicardium), migrate towards the developing heart and envelop the myocardium. Over the last few years, a wealth of new information has been collected on the unique genetic and phenotypic characteristics of the proepicardial and epicardial cells. Retroviral lineage studies, targeted transgenesis and mutational approaches have enabled more precise analyses on the cell fate, gene expression and differentiation of proepicardial and epicardial cells. In particular, there is mounting evidence that a subpopulation of proepicardial cells undergoes epithelial– to-mesenchymal transformation, migrate into the subepicardial and intramyocardial space, and differentiate into coronary endothelial and smooth muscle cells and cardiac fibroblasts. Furthermore, signals from the epicardium have important morphogenetic roles in promoting and maintaining the mitotic activity of heart muscle cells during heart formation. The epicardium may even harbor a progenitor cell population in the postnatal heart that potentially could be exploited for cardiac regeneration purposes. The mature vertebrate heart consists of three different tissue layers: the myocardium; the endocardium, which covers the myocardium from the inside; and the epicardium, which lines the myocardial layer on the outside. The epicardium is an epithelial layer, which in the mature vertebrate heart consists of a single-layered flat mesothelium connected to the myocardium by subepicardial connective tissue. One of its functions in adult life is to generate a smooth surface which enables the heart to freely move within the pericardial coelom. Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

Cardiac morphogenesis begins with the formation of the primitive heart tube that consists of two epithelial layers, the outer myocardium and inner endocardium (Manasek, 1969). Soon after the double-walled heart tube forms, a third epithelial layer begins to envelop the heart, forming the epicardium (Ho and Shimada, 1978; Virágh and Challice, 1981; Hiruma and Hirakow, 1989). In addition to these three layers, cardiac neural crest cells migrate to the heart and participate in cardiac patterning and remodeling (reviewed in Creazzo et al., 1998). There is mounting evidence that indicates that reciprocal interactions between the epicardium, myocardium and cardiac neural crest cells play critical regulatory roles in differentiation and patterning of the heart. Several animal models in which epicardial formation is genetically and surgically blocked, show various secondary cardiac defects, including thinning of the myocardial wall, and interventricular septal defects (Wilson and Warkany, 1949; Kreidberg et al., 1993; Männer, 1993; Kastner et al., 1994; Sucov et al., 1994; Kwee et al., 1995; Yang et al., 1995; Moore et al., 1999; Gittenberger-de Groot et al., 2000; Pennisi et al., 2003; Männer et al., 2005). Thus, an epicardium-derived signal(s) is essential for myocardial patterning during heart development (for further discussion of this issue see Chapter 5.2). The epicardium also provides cells of the coronary vasculature and connective tissues of the heart (Mikawa and Fischman, 1992; Mikawa and Gourdie, 1996; Dettman et al., 1998; Perez-Pomares et al., 1998; Li et al., 2002) (Fig. 1). The cardiac neural crest cells appear to play a critical supportive role in the normal development of coronary vessels and outflow tract (Creazzo et al., 1998; Hyer et al., 1999). This chapter will provide a brief overview of anatomical, embryological and molecular aspects of epicardial 325

326

PART  |  5  Epicardial Development

Endocardial endothelial cell

Cushion cell

Atrial myocyte

Cardiogenic mesoderm

Ventricular myocyte Purkinje fiber

Aortic smooth muscle Cardiac neural crest

Neuron

Smooth muscle Endothelial cell Proepicardium Fibroblast cell Epicardial cell

Blood cell

Figure 1  The origin and lineage relationships of cardiac cell types (modified from Mikawa, 1999). The diagram demonstrates that each cardiac cell type is established by lineage diversification of embryonic cells, which arise from one of three distinct origins: cardiogenic mesoderm, neural crest or proepicardium. These data define the chronology and distribution for the development of all cell lineages in the avian heart.

development. In recent years several excellent reviews have appeared on proepicardial and epicardial development (for further reading see these papers: Männer et al., 2001; Bernanke and Velkey, 2002; Poelmann et al., 2002; Reese et al., 2002; Majesky, 2004; Olivey et al., 2004; Wessels and Perez-Pomares, 2004; Mu et al., 2005; Tomanek, 2005; Männer, 2006; Smith and Bader, 2007).

II.  Epicardial development Epicardium and epicardium-derived coronary vasculature develops through several distinct morphogenetic steps during early cardiogenesis. This section provides a brief overview on several steps critical for epicardium formation, including the specification of epicardial progenitors, entry of the progenitors to the heart, formation of the epicardial sheet, epithelial-to-mesenchymal transformation (EMT) of a subpopulation of epicardial progenitors, and fate diversification into the epicardial and coronary vascular cell lineages.

II.A.  Induction and Specification of the Epicardial Anlagen Cells of the epicardium arise from an extracardiac mesodermal cell population that forms a protrusion, called the proepicardium, which develops immediately posterior to the sinoatrium (presumptive atrium) at the ventral lip of the anterior intestinal portal (Ho and Shimada, 1978; Virágh and Challice, 1981; Hiruma and Hirakow, 1989; Virágh et al., 1993; Männer et al., 2001; Nahirney et al., 2003). The proepicardium is composed of at least two distinct cell types, an external mesothelial epithelium and a mesenchymal core, which is also rich in extracellular matrix (Nahirney et al., 2003). In addition, there is a stratified cuboidal cell layer in the basal region of the proepicardium adjacent to the sinus venosus endothelium. Thin cellular extensions link the mesenchymal cells with the overlying cuboidal epithelium and with the deep stratified epithelium (Nahirney et al., 2003). With the help of marker genes (see Section V.C for a discussion of marker genes expressed in the proepicardium),

327

Chapter  |  5.1  Epicardial Lineage: Origins and Fates

(A)

(B)

(C)

A�

B�

C�

Figure 2  Expression of proepicardium marker genes prior to and during proepicardium formation. Whole mount in situ hybridization of chick embryos at HH stage 13 with probes for (A) TBX18, (B) WT1 and (C) CFC. In each case an L–R asymmetric expression of proepicardial marker gene expression is seen, being present on the right sinus horn (RS) and absent from the left (LS). (A–C) Transverse sections through the inflow tract shown in (A–C). The plane of sectioning is indicated in the individual panel. Black arrows point to the right-sided expression domain of the different proepicardium marker genes, red arrows point to the absence of proepicardium marker gene expression on the left side (IM: intermediate mesoderm; LS: left sinus horn; OT: outflow tract; RS: right sinus horn venosus; V: ventricle). Reprinted from Schlueter, J., Männer, J., and Brand, T. (2006). Dev. Biol. 295, 546–558, copyright 2006, with permission from Elsevier.

proepicardium formation can first be visualized at HH stage 11 in the chick and 8.5 dpc in the mouse (Schulte et al., 2007). In both species, proepicardium marker gene expression appears bilaterally, on the left and right sinus horns. Shortly thereafter in the chick expression becomes asymmetric, being stronger on the right side (Schlueter et al., 2006; Schlueter and Brand, 2009) (Fig. 2). The left and right side differs not only with regard to gene expression, but there is also evidence for differential levels of cell proliferation and apoptosis (Torlopp et al. in preparation). By scanning electron microscopy a vestigial proepicardium can be seen on the left side; however, no villi are formed (Schulte et al., 2007). Asymmetric morphogenesis could be the result of tissue interactions. For example, as a consequence of heart looping, the left sinus in the chick is exposed to the yolk sac which may suppress its development, while the right proepicardium develops opposite to the heart. Since proepicardium development on the right side is more advanced, it is also possible that the rightsided proepicardium might suppress development on the left side. Embryological experiments showed that neither of these assumptions is correct, suggesting that asymmetry of proepicardium development is under the control of the left– right (L–R) asymmetry pathway (Schlueter et al., 2006; Schulte et al., 2007; Schlueter and Brand, 2009) (Fig. 3). Since left–right asymmetry in vertebrates is governed to a large extent by the Nodal/Pitx2 pathway (see Chapter 4.1–4.3), it will be interesting to analyze its involvement in proepicardium development. Recent experiments indicate that at least asymmetric proepicardium-specific gene expression is under the control of an inductive right-sided pathway

including FGF8 and Snai1, whereas the Nodal/Pitx2 pathway seems not to be involved (Schlueter and Brand, 2009). Pitx2 might, however, control apoptosis of the vestigial proepicardium that is formed on the left side at later stages of development. Left–right asymmetry of proepicardium development appears to be an ancient property that is also found in lower vertebrates such as lamprey, dogfish, axolotl and Xenopus (Fransen and Lemanski, 1990; Pombal et al., 2008; Jahr et al., 2008). In contrast, mouse and zebrafish proepicardium does not display asymmetry (Schulte et al., 2007; Serluca, 2008). This may relate to the different strategies used by the chick and the mouse to transfer cells from the proepicardium to the heart, as described in Section II.B. Little is known about the embryonic origin of the proepicardium. Until now cell fate analysis in the chick or mouse embryos has not specifically addressed this tissue. However, since the proepicardium develops adjacent to the sinus myocardium, there is the possibility that it arises from the lateral margins of the heart fields (Redkar, 2002; Moorman et al., 2007). Using Cre-recombinase-mediated cell fate analysis in the mouse it has been established that proepicardium cells derive from cells that express Mesp1, which is also expressed by myocardial and endocardial cells (Saga et al., 2000). Moreover, Cre-mediated cell fate analysis in the mouse established transient expression of Isl1 and Nkx2.5 in the cell lineage forming the mouse proepicardium (Zhou et al., 2008). Consistent with these observations, double labeling experiments indicate that the pericardial mesoderm in the chick embryo co-expresses both myocardial and

328

PART  |  5  Epicardial Development

(A)

(B)

BMP4

BMP2

BMP2

RS

LS ?

Pitx2

Pitx2 Nkx3.2

Snai1

Nodal FGF8 FGF18 BMP4

R

Shh

Nodal Car

CFC

BMP2

L

Figure 3  Proepicardium formation in the chick embryo is under the control of the left–right pathway. (A) Double-color whole mount in situ hybridization of a chick embryo at HH stage 13 with probes for TBX18 (blue staining) and PITX2 (red staining) in order to visualize asymmetric expression domains on the right and left sinus horns, respectively. (B) The pathway of left–right asymmetry formation is potentially linked to asymmetric proepicardium formation on the right sinus horn. During early gastrulation in the chick embryo left–right asymmetric expression domains of BMP4 and SHH are established on the right and left side of Hensen’s node, respectively. An asymmetric NODAL expression domain adjacent to the left side of Hensen’s node is established by several independently-acting signaling pathways, in addition to Shh, and includes, for example, Notch/Delta and Wnt signals (not depicted). Nodal diffuses into the left lateral plate mesoderm with the help of Car, and establishes another expression domain of NODAL with the help of the competence factor CFC, which is controlled by BMP2. In the left LPM NODAL induces PITX2 and NKX3.2 (BapX1). Possibly PITX2 acts as a repressor of proepicardium development, since only a thin layer of mesothelial cells (labeled in yellow) is formed on the left sinus horn (LS). On the right side of Hensen’s node BMP4 induces FGF18 and FGF8. FGF8 induces SNAI1 in the right left lateral plate mesoderm, which prevents PITX2 expression on the right side. SNAI1 is part of the pathway that promotes proepicardium formation in the right sinus horn (RS) (Schlueter and Brand, 2009). The putative right-sided inducer of proepicardium development might act through the establishment of an asymmetric BMP4 expression domain in the proepicardium. BMP2, on the other hand, is symmetrically expressed in the myocardium of both sinus horns. The inflow tract of an HH stage 16 embryo is depicted here schematically. The sinus myocardium is labeled in blue, while the pericardial and proepicardial mesothelial cell layer are labeled in yellow. (A)

(B) Epicardial differentiation Proepicardium

Myocardium

Pericardial mesoderm

Myocardial differentiation Figure 4  Cardiac myocytes and proepicardial cells are probably generated from a common precursor pool. (A) Double staining of the tubular heart region of chick embryo at HH stage 16 with cytokeratin and MF20. While the heart is strongly labeled by MF20 and the proepicardium by cytokeratin (red), there is apparently a zone of overlap where both marker proteins are co-expressed, suggesting that they might originate from a common progenitor pool of mesenchymal cells that are derivatives of the posterior heart field mesoderm. (B) Schematic diagram of the proepicardial region is shown in which expression of myocardial marker (SERCA2) is depicted in green and a pericardial/epicardial marker (cytokeratin) in red. Co-expression of the myocardial and pericardial/epicardial marker at the base of the proepicardium suggests that a common pool of pericardial mesodermal cells contributes to both the IFT myocardium and proepicardium. (A) Reprinted from Kruithof et al. (2006) and (B) redrawn after Kruithof et al. (2006). Dev. Biol. 295, 507–522, copyright 2006, with permission from Elsevier.

­mesothelial marker proteins (Kruithof et al., 2006) (Fig. 4). The mouse proepicardium is believed to originate from the mesoderm in close proximity to the septum transversum. Prior to heart tube formation, the septum transversum

precursors are found rostral to the heart field and are labeled by Cited2 (Dunwoodie et al., 1998). Cell lineage studies independent of gene expression patterns in several vertebrate embryos are needed to establish their origin firmly.

Chapter  |  5.1  Epicardial Lineage: Origins and Fates

Inducing signal(s)

Liver bud endoderm

PE

Naive mesothelium Figure 5  Model for the specification of proepicardial fate. The proepicardium fate is specified within the mesoderm by inductive cues from a neighboring embryonic tissue(s), such as the liver bud endoderm. The proepicardium consists of a villous epithelium and a mesenchymal core.

Recent clonal analysis of FACS-sorted embryonic stem cells shows that after differentiation cells that became positive either for the VEGF receptor-2 (Flk-1), Isl1, or Nkx2.5 can give rise in vitro to myocytes, smooth muscle and endothelial cells (Garry and Olson, 2006; Kattman et al., 2006; Moretti et al., 2006; Wu et al., 2006). Moreover in vivo lineage tracings suggest that a minor population of Isl1-positive precursors contribute to smooth muscle and endothelial cells of the proximal part of the coronary tree (Moretti et al., 2006). Thus, these data would suggest that a lateral plate progenitor cell population exists that is directed towards different sublineages of the heart (potentially including the proepicardium) by an unknown set of signaling clues. However, another Cre recombinasemediated cell fate analysis has previously demonstrated that in the murine heart, cardiac neural crest cells also differentiate into smooth muscle cells of the proximal part of coronary tree (Jiang et al., 2000). The solution of this issue would require a “true” cell lineage study, in addition to those based on gene expression patterns, which often change dynamically during embryogenesis. The exact mechanism that induces and specifies the proepicardium fate in the mesoderm at the specific region of the embryo is currently unknown. However, it has been known that the proepicardium develops from mesodermal cells that overlay the liver bud endoderm (Männer, 1992; Virágh et al., 1993; Männer et al., 2001; Nahirney et al., 2003). Interestingly, when a donor quail liver bud is implanted in the posterior–lateral regions of host chick embryo, the expression of some proepicardium marker genes can be induced in host mesodermal cells ectopically at the implantation site (Ishii et al., 2007) (Fig. 5). The work suggests the potential role of a liver bud endodermderived paracrine cue(s) in proepicardium induction. Since the proepicardium is closely apposed not only to the liver, but also to the sinoatrial myocardium, both of these tissues

329

have been suggested as playing a role in inducing proepicardium development (Männer et al., 2001; Bernanke and Velkey, 2002; Poelmann et al., 2002; Reese et al., 2002; Majesky, 2004; Olivey et al., 2004; Wessels and Perez-Pomares, 2004; Mu et al., 2005; Tomanek, 2005; Männer, 2006; Smith and Bader, 2007). However, the above study also shows that induction of ectopic proepicardium marker gene expression can be triggered by the liver implant alone in mesoderm without co-implantation of a myocardial tissue, suggesting that a liver-derived signal(s) is sufficient for the induction of some proepicardium marker genes without a myocardium-derived signal(s). Induction of cardiogenic mesoderm depends on paracrine signals from the underlying foregut endoderm (Schultheiss et al., 1995, 1997; Lough et al., 1996; Andrée et al., 1998; Schlange et al., 2000b; Brand, 2003). Signals from the resulting cardiogenic mesoderm in turn initiate hepatic cell differentiation and liver bud formation in the ventral endoderm (Douarin, 1975; Fukuda-Taira, 1981; Gualdi et al., 1996; Duncan, 2003). The signals from cardiogenic mesoderm might include FGF (Jung et al., 1999; Zhang et al., 2004; Serls et al., 2005; Calmont et al., 2006) and BMP4 (Rossi et al., 2001). In addition to these two reciprocal ­mesoderm–endoderm interactions in the induction/differentiation of cardiomyocytes and hepatocytes, a tissue–­tissue interaction between the liver endoderm and overlaying mesoderm may play a role in induction and/or specification of the fate within the mesothelium. Currently, molecules that mediate the ­ proepicardiuminducing activity of the liver primordium are undefined. Recent studies have shown, however, that in the chick embryo both BMP2 and BMP4 are expressed in the inflow tract myocardium (BMP2) or in the proepicardium itself (BMP4) (Kruithof et al., 2006; Schlueter et al., 2006). Other BMPs (BMP5, BMP7 and BMP10) are also expressed in the inflow tract myocardium (Kruithof et al., 2006). In vitro and in vivo data suggest that BMP signaling is required to maintain proepicardium marker gene expression (Schlueter et al., 2006). Significantly BMP4 is expressed in the proepicardium itself, and blocking BMP signaling in proepicardium explant cultures leads to a loss of proepicardium marker gene expression, which suggests an important autocrine function of BMP for maintaining proepicardium identity (Schlueter et al., 2006). The expression patterns of Bmp2 and Bmp4 are not conserved in the mouse. Bmp4 is not expressed in the proepicardium and Bmp2 expression is not present in the inflow tract myocardium adjacent to the proepicardium (Schulte et al., 2007). Nonetheless, BMP signals do play a role in mouse proepicardium development, since the enhancer controlling Gata4 expression in the lateral plate mesoderm and in the proepicardium is BMP-dependent (Rojas et al., 2005). Possibly other BMP genes are expressed in the murine proepicardium and inflow tract and substitute for BMP2 and BMP4 action seen in the avian.

330

Several FGF genes are expressed in the inflow tract myocardium, and in the proepicardium itself (Kruithof et al., 2006; Schlueter et al., 2006). FGF8 is expressed in the inflow tract myocardium, while FGF2 and FGF10 are present in the proepicardium. These FGFs have complex activities when added to proepicardium explant cultures (Kruithof et al., 2006). Inhibition of FGF signaling results in impaired epithelial outgrowth (Torlopp et al. in preparation). Similarly, blocking FGF signaling in vivo inhibits proepicardium villi formation, while proepicardial gene expression is maintained. These data suggest that FGF has an autocrine role and maintains proepicardial survival and epithelial outgrowth, but is not involved in proepicardium cell specification. The EGF–CFC family encodes proteins required for receptor interaction of Nodal and related members of the TGF superfamily such as Gdf1, Gdf3 and Vg1 (Shen, 2007). Chick CFC1 is expressed in the proepicardium beginning at HH stage 11, and together with TBX18 is one of the earliest expressed marker genes for proepicardium development (Schlueter et al., 2006) (Fig. 2). Interestingly expression is maintained late into cardiac development and is still found in the epicardium at day 7 of chick development (Schlange et al., 2001). There are no reports of the presence of EGF–CFC family members in the proepicardium of other vertebrates. However, both Cripto (Tdgf1) in mouse and XCR2 in Xenopus are expressed in the cardiac primordium, and possibly are also involved in proepicardium development (Onuma et al., 2006). EGF–CFC proteins might be required for an uncharacterized Nodal-like signal involved in proepicardium induction. However, it is also possible that EGF–CFC proteins function in a Nodalindependent context (Adamson et al., 2002; Gray et al., 2003, 2006; Strizzi et al., 2004; Tao et al., 2005). Another candidate signaling molecule involved in proepicardium development is retinoic acid, which is required for inflow tract formation (Kostetskii et al., 1999; Ghatpande et al., 2000, 2006). Raldh2, one of the rate­limiting enzymes of retinoic acid biosynthesis is specifically expressed in the proepicardium and is also present in the pericardium (Xavier-Neto et al., 2000). Retinoic acid functions as a survival factor for proepicardium cells, since increased apoptosis in proepicardium cells and an impaired epicardialization of the myocardium was observed in the RXR-receptor null mutant (Jenkins et al., 2005). Angiopoietin 1 (Ang1), a growth regulator implicated in vascular development which binds to the receptor tyrosine kinase Tie2 has been implicated previously in ­endocardial– myocardial interactions (Dumont et al., 1994; Davis et al., 1996). Recently, however, overexpression of Ang1 revealed a function for this ligand in coronary arteriogenesis and/or epicardium formation (Ward et al., 2004). Despite all these data, at present the molecular signals that are involved in proepicardium induction are only poorly-characterized.

PART  |  5  Epicardial Development

II.B.  Proepicardial Growth toward the Heart As soon as the proepicardium fate is specified at the sinus horn region, proepicardium cells undergo a dramatic morphogenesis and growth phase. Two distinct processes of proepicardium cell growth and translocation across the pericardial cavity to the heart have been identified in a species-specific manner. Studies in avian embryos have demonstrated that proepicardium mesothelial cells develop multiple finger-like protusions, or villi, into the pericardial coelomic cavity directed toward the looping stage heart (Ho and Shimada, 1978; Hiruma and Hirakow, 1989; Virágh et al., 1993) (Fig. 6). In the chick embryo infrequent villi projections, reminiscent of a nascent proepicardium, also develop at other sites on the coelomic wall; however, they do not grow further. Studies in fish and mouse embryos suggest that in those species the proepicardium does not develop well-extended villi; instead it generates short protrusions or blebs that ultimately transform into proepicardium cysts that detach from the pericardial mesothelium as free-floating, multicellular aggregates or vesicles (Virágh and Challice, 1981; Komiyama et al., 1987) (Fig. 7). The proepicardium cysts released into the coelomic space make random contacts with the myocardial surface. Recent evidence suggests that the cyst-like transfer of proepicardial cells onto the surface of the murine heart is limited to the areas with the longest distance between the proepicardium and the developing heart, while in areas where the proepicardium and the heart surface are adjacent, proepicardium cells migrate towards the heart predominantly through the use of multicellular villi (Rodgers et al., 2008). Through proepicardium expansion, heart growth, and the motion created by the beating heart, the proepicardium cells detach from

Chicken

(A)

(B) Ventricle Pericardium Mesenchyme

PE

PE R

L

Epithelium

Sinus

Figure 6  Development and transfer mode of proepicardial cells in the chick. (A) At HH stage 16 in the chicken embryo, a proepicardium has fully-developed on the right sinus horn; while on the left side, a proepicardium anlage in a rudimentary state is seen that disappears at HH stage 18. (B) The right proepicardium establishes contact to the heart at HHstage 17. The proepicardium-derived epicardium spreads over the cardiac surface to form a continuous epithelial sheet.

331

Chapter  |  5.1  Epicardial Lineage: Origins and Fates

villi once in contact with the myocardium by a “Velcro type” mechanism. Since the proepicardium of the mouse is a fragile structure, this type of cell transfer might have been missed in the past due to preparation artifacts. In this regard it is noteworthy that a tissue-bridge-like transfer of proepicardium cells to the heart was also observed in the rat embryo (Nesbitt et al., 2006). Regardless of whether the mechanisms are proepicardium bridge-mediated or cyst-mediated, the epicardium and extracellular matrixrich subepicardial zone are ultimately formed on the surface of the myocardium. It should be noted that there are other terms in use to describe the proepicardium, including proepicardial organ (Reese et al., 2002) and proepicardial serosa (Männer et al., 2001). Historically, Virágh used the term proepicardium to describe the cauliflower-like structure that forms initially, and the term proepicardial organ for the subsequent stage where the proepicardial tissue bridge forms between the pericardial mesothelium and the myocardium (Virágh et al., 1993). However, since a tissue bridge does not form in fish and mice, and is only transiently present in avians and rat, the “proepicardial organ” does not appear to be an appropriate term as a common name for the epicardial anlagen across all species. Therefore, this review utilizes the term “proepicardium” for the initial structure formed in mammals, avian, amphibian and fish embryos, the term “proepicardium tissue bridge” for the proepicardium that makes contact to the cardiac surface in amphibians, avians and rat, and “proepicardium cysts” for fish and mouse embryos. While the proepicardium forms adjacent to the sinoatrium, proepicardium cells do not enter the sinoatrium. Instead, the proepicardium villi preferentially target the dorsal surface of the atrial–ventricular

(AV) junction on the lesser curvature of the looping stage heart. The mechanism(s) governing this precisely-oriented proepicardium tissue bridge extension and guidance of ­proepicardium cysts to the myocardium is currently uncertain. In the chick, most proepicardium cells reach the myocardial surface via a tissue bridge that extends across the coelomic cavity (Sejima et al., 2001). This tissue bridge forms along strands of extracellular matrix spanning the pericardial coelomic cavity between the proepicardium and myocardial surfaces (Nahirney et al., 2003) (Fig. 8). The injection of heparinase into the pericardial coelom results in the failure of proepicardium tissue bridge formation, suggesting there is a potential role for the extracellular matrix in extension and targeting of the proepicardium tissue bridge to the myocardium (Nahirney et al., 2003). In addition to a structural guidance for proepicardium extension to the lesser curvature, there is a possibility that the extracellular matrix bridge may serve as a rich store of growth factors that may play a direct role in paracrine signaling for proepicardium extension. For example, the heparan sulfate proteoglycans syndecan and glypican function as low affinity coreceptors for basic fibroblast growth factor (FGF), and this interaction is required for the high-affinity binding of basic FGF to its cellular receptor (Aviezer et al., 1994). In the chick, the transcription factor TBX5 is expressed in the proepicardium and retrovirus-mediated overexpression of human TBX5, as well as antisense-mediated knockdown of chick TBX5, diminishes proepicardium extension to the heart and their incorporation into the nascent epicardium and coronary vasculature (Hatcher et al., 2004). While proepicardium cysts may passively float across the pericardial cavity to eventually attach as independent patches on the myocardium in the mouse embryo (Virágh and Challice, 1981; Komiyama et al., 1987), it has been

Mouse (A)

(B)

(C) Ventricle

PE

PE

PE

PE Sinus

Figure 7  Transfer mode of proepicardial cells in the mouse. (A) A proepicardium anlage becomes morphologically identifiable at E8.5. In contrast to the chick, the proepicardium in the mouse is present on both sinus horns (blue-labeled cell aggregates). (B) By E9.5 the two anlagen have merged to form a fully-developed proepicardium anlage at the embryonic midline. (C) Between E9.5 and E10.5, the majority of proepicardium vesicular cell aggregates are released into the free pericardial cavity. These proepicardium vesicles adhere to the naked myocardial surface of the heart loop where they contribute to the formation of the primitive epicardium. However, the atria are not epicardialized by this mode of cell transfer; rather they are covered by epicardium that migrates as a sheet from the sinus to the atrium.

332

PART  |  5  Epicardial Development

OT

ECM bridge V PE

AV

chemoattractant SA

Figure 8  Model for the instructive role of myocardium-derived paracrine signal in oriented projection of proepicardium cell development and migration toward the heart. The myocardium-derived paracrine factors (arrow and dots) bind to the heparin sulfate-rich extracellular matrix bridge (ECMB, yellow lines) that expands between the proepicardium (blue) and the inner curvature of the looping stage heart (red). The locally-concentrated paracrine factors on the ECMB serve as a directional chemoattractant and provide a cue for oriented outgrowth and migration of proepicardium cells toward the specific site of the embryonic heart (PE: proepicardium; OT: outflow tract; SA: sinoatrium; V: presumptive ventricle).

shown that mutations of the 4-integrin (Itga4) gene encoding a cell surface receptor for fibronectin (Guan and Hynes, 1990; Yang et al., 1995) and Vcam1 (Osborn et al., 1989; Elices et al., 1990) in mouse embryos give rise to defects in attachment of the proepicardium cysts to the heart (Sengbusch et al., 2002). Thus, more active processes such as binding of proepicardium cells to extracellular matrix and cell–cell adhesion may be involved in translocation of proepicardium cysts to the heart.

II.C.  Epicardial Sheet Formation and Patterning of the Myocardium Epicardial sheet formation begins on contact of a proepicardium tissue bridge or cysts with the myocardium. In the chick the initial contacts of the proepicardium villi with the myocardium occur at HH stage 16, giving rise to the proepicardium tissue bridge (Ho and Shimada, 1978; Shimada et al., 1981; Virágh and Challice, 1981; Hiruma and Hirakow, 1989; Männer, 1992). On this contact, the tips of proepicardium villi open up, allowing proepicardium cells to spread out to form the epicardial epithelial sheet on the surface of the myocardium, with its original apical surface facing toward the pericardial coelomic cavity (Nahirney et al., 2003). A continuous epicardial sheet expansion takes place over the atria first, and then over the

ventricles, eventually covering the entire myocardium and pericardial cavity by stage 23 in chick embryos (Hiruma and Hirakow, 1989). In the mouse, epicardial sheet formation begins as proepicardium cysts attach on the myocardial surface at 9 dpc (Komiyama et al., 1987). Cells of each cyst flatten out, forming an island of a simple squamous epithelial sheet on the surface of the myocardium. Fusion of these epithelial islands to complete the entire epicardium occurs by 11 dpc (Komiyama et al., 1987). Paracrine cue(s) originating from the developing epicardium play a critical role in myocardial growth during heart development (Pennisi et al., 2003; Lavine et al., 2005, 2006). Both microsurgical and genetic inhibition of epicardium formation gives rise to a decrease in myocyte proliferation, accounting for a thinner compact myocardium (Kreidberg et al., 1993; Männer, 1993; Kastner et al., 1994; Sucov et al., 1994; Kwee et al., 1995; Yang et al., 1995; Moore et al., 1999; Gittenberger-de Groot et al., 2000; Pennisi et al., 2003; Männer et al., 2005). Early myocardium consists of an outer highly-mitotic compact zone and an inner trabecular zone with lower mitotic activity. It was originally shown in the chick embryonic heart that fibroblast growth factor (FGF) receptor-mediated signaling is central to myocyte proliferation in the developing heart (Mikawa, 1995; Mima et al., 1995). Consistent with this, knockout of FGF receptors in the myocardium of the mouse embryo has demonstrated that receptor-mediated FGF signaling is critical for proper proliferation and differentiation of the murine cardiac muscle (Lavine et al., 2005, 2006). Further, it has recently been shown that retinoic acid signaling induces epicardial secretion of FGFs, a key mitogen for myocardial proliferation (Chen et al., 2002; Merki et al., 2005). These studies suggest that epicardium-derived signals are essential for myocyte proliferation in the compact myocardium by means of producing mitogens, as well as regulating levels of receptors for mitogens in the myocardium (Pennisi et al., 2003; Lavine et al., 2005, 2006).

II.D.  Epithelial-to-Mesenchymal Transformation During proepicardium extension and epicardial sheet formation, a subpopulation of proepicardium and epicardial cells undergo an epithelial-to-mesenchymal transformation and migrate into the subepicardial space (Virágh and Challice, 1981). The epicardial epithelial-to-mesenchymal transformation is observed at the AV junction, in the ventricular epicardium, and in the epicardium at the junction between the ventricles and the outflow tract, but not in the atrial epicardium (Wessels and Perez-Pomares, 2004). Follow­ ing epicardial epithelial-to-mesenchymal transformation, ­epicardium-derived cells migrate into the myocardium and subsequently differentiate into a variety of myocardial cell

Chapter  |  5.1  Epicardial Lineage: Origins and Fates

types (Fig. 1) including subepicardial mesenchyme, interstitial fibroblast, coronary endothelium, coronary smooth muscle cells and hemangioblast (see also Chapter 5.2). The extent of epithelial-to-mesenchymal transformation is probably affected by the extracellular matrix in the subepicardial space, which is especially prominent in the AV groove and also generates epicardium-derived cells in large amounts. While the exact mechanism that regulates the epithelial-to-mesenchymal transformation remains to be determined, studies of ­epithelial-to-mesenchymal transformation in explanted proepicardium and epicardium have identified candidate factors that regulate epithelial-to-mesenchymal transformation. Both vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF) stimulate epithelial-to-mesenchymal transformation of epicardial cells in vitro (Morabito et al., 2001). Using a rat proepicardial cell line FGF2 and VEGF were found to induce epithelial-tomesenchymal transformation, but not PDGF-AA, PDGFAB, or FGF1 (Wada et al., 2003b). In the case of TGF contradictory results have been reported. Morabito et al. (2001) reported that TGF in epicardial explants was inhibitory towards epithelial-to-­mesenchymal transformation. In contrast, proepicardium explants showed an induction of epithelial-to-mesenchymal transformation on TGF stimulation (Compton et al., 2006; Olivey et al., 2006). Genetic studies have shown that disruption of a FOG-2 (Zfpm2)-dependent signal produced in the myocardium does not affect epicardium formation, but leads to a loss of mesenchyme production and lack of coronary blood vessel formation (Tevosian et al., 2000). Two members of the Ets family of transcription factors are involved in the control of epithelial-tomesenchymal transformation. Antisense inhibition of Ets-1 and Ets-2 resulted in a loss of subepicardial mesenchyme formation, as well as in a loss of coronary artery formation (Lie-Venema et al., 2003). Likewise, the histone acetyl transferase p300 is essential for epithelial-to-mesenchymal transformation in the epicardium. Mice harboring a single mutant allele of p300 have a normal epicardium, yet few subepicardial mesenchymal cells are present, and the coronary vessel network is defective (Shikama et al., 2003). Taken together, these results are consistent with the idea that the epicardial epithelial-to-mesenchymal transformation is regulated by paracrine signaling from the myocardium. It remains to be determined, however, how only a small subpopulation of epicardial cells respond to the myocardial signal(s) and undergo epithelial-to-mesenchymal transformation, and how the majority of epicardial cells remain epithelial.

II.E.  Fate Diversity of Proepicardium and Epicardial Cells Historically, there was controversy over the origin of the coronary vascular bed (Baldwin, 1996) as to whether there was outgrowth from the aortic root, i.e., ­ angiogenesis

333

(Aikawa and Kawano, 1982; Hirakow, 1983; Hutchins et al., 1988), or ingrowth by convergence of angio­blasts formed in situ, i.e., vasculogenesis (Bogers et al., 1989; Waldo et al., 1990). Over a decade ago, retroviral­mediated single cell tagging and tracking of epicardial cells provided the first direct evidence that epicardial cells produce coronary vascular smooth muscle cells, cardiac fibroblasts and endothelial cells (Mikawa and Fischman, 1992). The work also demonstrates that coronary vessels are established by vasculogenic mechanisms, not by angiogenic sprouting from the root of the aorta. Subsequent lineage studies on individual proepicardium cells have further revealed that all coronary vascular cell types arise from the proepicardium and begin to enter the developing heart along with the growing epicardial mantle (Mikawa and Gourdie, 1996). These retroviral cell lineage data have been confirmed by chick-quail chimera studies (Poelmann et al., 1993; Männer, 1999; Vrancken Peeters et al., 1999) and with implantation of adenovirally-tagged proepicardial cells (Dettman et al., 1998). Recent studies have further shown that proepicardium-derived cells transiently form blood islands in the subepicardial space (Tomanek et al., 2006). In contrast, lymph vessels which form alongside the coronary vessels do not originate from the proepicardium (Wilting et al., 2007). While the above cell lineage and chimera studies have mapped the origin of coronary vasculature to the proepicardium, it remains undetermined how the fate of individual proepicardium cells is diversified either into epicardial, cardiac endothelial, smooth muscle or fibroblast cells. The retroviral single cell lineage studies have demonstrated that individual proepicardium cells differentiate into only one of these cell types in vivo (Mikawa and Fischman, 1992; Mikawa and Gourdie, 1996) (Fig. 9). However, the data do not provide a definitive answer as to whether cell fate is predetermined in the proepicardium, or regulated by environmental cues provided by the myocardium after migration to the heart. In particular, it is controversial whether the fate specification and differentiation of coronary endothelial cells are induced prior to or only after proepicardium cells enter the heart (Mikawa and Gourdie, 1996; Perez-Pomares et al., 1998; Gourdie et al., 1999). Two independent studies observed that proepicardial explants are able to undergo cardiac myocyte differentiation in vitro. One study reported spontaneous formation of myocytes when proepicardium explants were cultured in the presence of chick serum (Kruithof et al., 2006). In contrast, under serum-free culture conditions cardiac myocyte differentiation was only observed after the addition of BMP2 or noggin (Schlueter et al., 2006). Not all proepicardium-derived cells differentiate into cardiac myocytes; only a small subpopulation of cells is converted. These data suggest that cells within the proepicardium are not fully committed to a mesothelial or coronary vessel cell

334

PART  |  5  Epicardial Development

PE induction

Cell differentiation

Oriented EMT and outgrowth migration

Endothelial cell

Proepicardium

Artery formation and patterning

Coronary progenitors

Smooth muscle cell

Mesothelium

Capillary plexus formation

Capillary plexi

Heart tube

Vasculogenesis

Figure 9  Diagram illustrating coronary arteriogenesis from proepicardium-derived cells. A mesothelium overlaying the liver primordium differentiates into the proepicardium, which is composed of multiple epithelial villi protruding toward the looping stage heart tube. Independent endothelial (red) and smooth muscle (green) progenitors begin to express their cell type-specific markers within the proepicardium. Together with epicardial cells and cardiac fibroblasts (gray), coronary progenitors (smooth muscle and endothelial cells) migrate to the heart precisely at the inner curvature of the looping heart tube. Mesenchymal cells in the core of the proepicardium develop into the subepicardial mesenchyme after transfer to the heart. The proepicardium-derived endothelial cells first form sinusoidal sacs which then fuse, eventually forming capillary plexi. Intracardiac smooth muscle progenitors migrate to particular vascular plexi and establish the coronary artery network. Modified from Mikawa, T. (1999); and Reese et al. (2002).

fate. Moreover, it suggests that BMP might be a signaling molecule that allocates cells either to the mesothelial or myocardial cell lineages. As pointed out already, the mesoderm that borders proepicardium, as well as sinus myocardium, likely has a bi-potential fate to either become cardiac myocytes or undergo proepicardium development (Fig. 4). In the mesoderm underneath the proepicardium markers for both lineages are co-expressed. It is probably this cell population that can undergo cardiac myocyte formation if signaling levels of BMP and FGF are modulated (Kruithof et al., 2006). In contrast, in quail-chick chimeras, no evidence for cardiac myocyte formation in vivo has been obtained (Männer, 1999). Likewise, retroviral cell labeling was unable to reveal any contribution of proepicardium cells to the cardiac myocyte cell population (Mikawa and Gourdie, 1996). Thus, it appears that cell fate allocation in vivo is stricter than under in vitro culture conditions. Interestingly, recent fate-map studies in the mouse embryo employing Wt1-Cre or Tbx18-Cre both showed a significant contribution of epicardium-derived cells to the myocardial cell lineage (Cai et al., 2008; Zhou et al., 2008). These data would suggest that chick and mice differ in regard to the contribution of epicardial cells to the myocardial lineage. It has, however, been recently observed that low levels of Tbx18 expression persist in the myocardium (Christoffels et al., 2009). Thus, Tbx18-Cremediated lineage tracing is possibly noninformative with regard to epicardial cell fate analysis. Quail-chick chimeras are made by grafting the quail proepicardium into the pericardial cavity of HH stage 16 chick embryos (Dettman et al., 1998; Gittenberger-de Groot et al., 1998; Männer, 1999). When investigated at stage 45 (E18), the contribution of the quail proepicardium-derived cells to the chick heart included epicardium,

the subepicardial connective tissue in the atrioventricular grooves, the coronary endothelial and smooth ­ muscle cells and the intramyocardial fibroblasts. Although ­proepicardium-derived mesenchymal cells have been shown to populate the atrioventricular cushions during development (Dettman et al., 1998; Gittenberger-de Groot et al., 1998; Männer, 1999), very few such cells are present in late stage hearts in the tricuspid and mitral valves, whereas the aortic and pulmonary valves do not receive any proepicardium cells during development (de Lange et al., 2004). Thus, there is little evidence for a material contribution of this cell lineage to valvular tissue; however, these cells could still have an important patterning role, being subsequently lost through apoptosis.

III.  Epicardial marker genes Essential for further insight into the process of proepicardium formation, cell fate diversity, proepicardium tissue bridge or cyst formation, etc., is the identification of molecular markers to distinguish different cell lineages, and states of determination and differentiation in this important cell population. In this section several transcription factors and other genes that are specifically expressed in the proepicardium are introduced, along with what is known about their function in the context of proepicardium formation and epicardial development.

III.A.  Wilms Tumor Gene 1 (Wt1) The Wilms tumor gene Wt1 encodes a zinc-finger protein which has been implicated in tumor development and as a critical regulator of embryonic development (Scholz and

335

Chapter  |  5.1  Epicardial Lineage: Origins and Fates

Kirschner, 2005). In the kidney, Wt1 is upregulated once the ureter epithelium has formed. Similarly, in the chick PE Wt1 is strongly-expressed in the epithelial cells, while being absent from the mesenchymal core of the proepicardium (Schlueter et al., 2006) (Fig. 2). Wt1 expression is maintained in the epicardium, in the subepicardial mesenchyme, as well as in migratory epicardium-derived cells (Carmona et al., 2001). After cells reach their final destination and start differentiation, Wt1 is downregulated (Perez-Pomares et al., 2002). Wt1 also serves as an excellent marker of early proepicardium formation, being strongly expressed in the chick proepicardium already by HH stage 13. In the absence of Wt1 activity in the mouse embryo, the epicardium does not form correctly, resulting in large gaps at the cranial end of the heart and a complete absence of the epicardium over the ventral surface of the aorta (Moore et al., 1999). The animals die showing a thin ventricle phenotype and probably have an embryonic form of heart failure. Transmural bleeding into the pericardial cavity due to rupture of the myocardium was also observed in the Wt1/ mutants. Recently, several putative target genes have been identified, including E-cadherin (Cdh1) (Hosono et al., 2000), Itga4 (Kirschner et al., 2006), erythropoietin (Epo) (Dame et al., 2006) and the neurotrophin receptor TrkB (Ntrk2) (Wagner et al., 2005). These genes have all been implicated in various aspects of epicardium formation (Itga4, Cdh1), epicardial cell differentiation (Ntrk2), or epicardium–myocardium interactions (Epo) (Sengbusch et al., 2002). Thus, Wt1 is an important regulator for epicardium formation. Another putative target of Wt1 is the intermediate filament gene Nestin (Nes) (Wagner et al., 2006). Nes is co-expressed with Wt1 in the epicardium, and in the forming coronary artery cells and interestingly, both genes are upregulated in coronary arteries bordering the infarct in rat heart (Wagner et al., 2002). It is an important question whether Nestin, which is often expressed in the context of stem cells, is expressed in the proepicardium. In conclusion, Wt1 appears to be of central importance for various aspects of epicardium formation and cell differentiation.

III.B.  T-box Genes T-box genes are involved in many different aspects of heart development (Plageman and Yutzey, 2005). Two T-box genes, Tbx5 and Tbx18, are expressed during epicardium formation in zebrafish, chick, mouse and man (Hatcher et al., 2004). The Tbx5-null mutant has a severe phenotype with aberrant morphogenesis of the posterior heart, and is embryonic lethal at E9.5, making it impossible to study its phenotype with regard to epicardium formation (Bruneau et al., 2001). Recently, a Tbx5 hypomorphic allele has been generated; however, in this case the myocardial phenotype was also too severe to be informative with regard

to epicardium formation (Mori et al., 2006). Knockdown experiments utilizing the chick explant culture system suggest that Tbx5 has a role in proepicardial cell migration (Hatcher et al., 2004). Overexpression of TBX5 by virus infection of proepicardial cells in the chick embryo also displayed impaired cell migratory activity, suggesting that a well-defined level of TBX5 protein is required for proper epicardialization of the heart. Interestingly, TBX5 is not only regulated at the transcriptional level, but nuclear localization of TBX5 protein is controlled through interaction with the PDZ–LIM domain protein LMP4 (Krause et al., 2004; Camarata et al., 2006; Bimber et al., 2007). LMP4 is expressed during the entire period of chick proepicardial development, and might serve as a marker for this cell population. LMP4 gets downregulated after epicardium-derived cells start to invade the myocardium (Krause et al., 2004; Camarata et al., 2006; Bimber et al., 2007). In contrast to LMP4, TBX5 is only found at HH stage 20 in proepicardial cells that have made contact with the myocardium, suggesting that either cytoskeletal rearrangement due to myocyte-proepicardial cell–cell interaction, or some sort of a short-range signaling molecule secreted by the myocardium affects proepicardial TBX5 expression. TBX5 expression becomes lost after the epicardium has fully formed, supporting the concept that TBX5 is required to keep epicardial cells in a migratory state (Krause et al., 2004; Camarata et al., 2006; Bimber et al., 2007). TBX5 may also play a role in pericardium formation, a cell layer that is formed by mesothelial cells which probably have the same embryonic origin as proepicardial cells (Fig. 4). A TBX5 missense mutation was reported in a patient that had, among other cardiac malformations, an agenesis of the left pericardium (Dias et al., 2007). TBX18 is specifically and strongly expressed in the epicardium (Fig. 2), and serves as a good marker of even the earliest stages of proepicardium formation in the zebrafish, chick and mouse (Schlueter et al., 2006). However, the loss of Tbx18 in the mouse does not affect epicardium formation, whereas sinus pole formation is abnormal (Christoffels et al., 2006). Whether T-box proteins serve any function other than regulating migratory activity during epicardial development is an important yet presently unsolved issue.

III.C.  Serum Response Factor (SRF) In the chick proepicardium SRF is not expressed, however it becomes rapidly upregulated after culturing proepicardial explants in serum, or after the addition of growth factors such as PDGF (Landerholm et al., 1999). Thus, SRF expression is upregulated after epithelial-to-­mesenchymal transformation initiation. Once epicardium formation is complete (HH stage 26 in the chick), cells within the forming subepicardial mesenchyme upregulate SRF. Differentiation of these cells to the smooth muscle cell

336

lineage depends on functional SRF (Landerholm et al., 1999). Other transcription factors that are implicated in smooth muscle differentiation, such as MEF2B, GATA5, HAND1 and CRIP2 are present in the proepicardium before smooth muscle differentiation has begun, suggesting that SRF is one of a limited number of factors that initiate smooth muscle differentiation. Srf expression in mice is upregulated in late-stage proepicardium, and subsequent to myocardial invasion is expressed in vascular smooth muscle cells, but not in the epicardium itself. An enhancer of the Srf gene that harbors E-box and Ets binding sites is important for driving proepicardial gene expression in transgenic mice (Nelson et al., 2006).

III.D.  GATA Genes All three of the cardiac GATA genes, Gata4, Gata5 and Gata6 are expressed in the septum transversum in mice, and subsequently are also present in the proepicardium and epicardium (Nemer and Nemer, 2003; Watt et al., 2004) along with the GATA co-factor Zfpm2 (Svensson et al., 1999). It is likely that some GATA-mediated functions during epicardium formation may be controlled in a redundant fashion. Nonetheless, Gata4 is essential for septum transversum formation and proepicardial development (Watt et al., 2004). In Gata4-null mutants with a tetraploid cell complementation of the early Gata4 functions in extraembryonic development, proepicardium cells are formed, however their number is reduced and the cells are misplaced. Gata4 is also an essential component of the development of the venous pole of the heart (Kostetskii et al., 1999) and thus, the observed effect on proepicardium formation may be secondary due to an essential role of Gata4 for inflow tract formation (Ghatpande et al., 2006). A Gata5 enhancer that drives Cre-recombinase in the epicardium (MacNeill et al., 2000) was recently used to ablate the Rxra gene in the epicardial cell lineage (Merki et al., 2005). GATA transcription factors also play a critical role during coronary artery formation. Two mouse mutants were generated that either harbor a Zfpm2-null mutation or a knockin mutation encoding a Gata4 variant unable to interact with Zfpm2 (Tevosian et al., 2000; Crispino et al., 2001). In both cases a strong impairment of coronary artery development was seen.

III.E.  Epicardin The basic helix-loop-helix gene epicardin/capsulin/Pod1 (official gene symbol: Tcf21) is specifically expressed in the chick and mouse proepicardium, epicardium and in the pericardium (Robb et al., 1998). It is believed to act as a repressor of cell differentiation (Funato et al., 2003). Null mutants die at birth from multiple organ defects that include lung hypoplasia, asplenism and renal dysplasia (Quaggin

PART  |  5  Epicardial Development

et al., 1998; Lu et al., 1998, 2000, 2002). Despite the ­specific expression pattern of Tcf21 in the epicardium, no epicardial phenotype was noted in the null mutant, although a hemopericardium was observed in the neonate, possibly related to defective coronary vessel maturation (Quaggin et al., 1998; Lu et al., 1998, 2000, 2002). Other related helix-loop-helix genes might functionally substitute for Tcf21.

III.F.  Forkhead Transcription Factors The forkhead transcription factor Foxf1a is expressed during early mouse development in the extraembryonic and lateral plate mesoderm (Peterson et al., 1997). Null mutants die at E8.0 due to defects in extraembryonic tissue formation (Mahlapuu et al., 2001). At later stages of development Foxf1a is expressed in the septum transversum mesenchyme, and in the adult heart is found in the coronary arteries and in the pericardium (Kalinichenko et al., 2003). Specific functions during proepicardium formation or epicardialization are unexplored. Other members of the forkhead family of genes expressed in the proepicardium in the mouse embryo are Foxc1 and Foxc2, which are both expressed in a subset of cells in the PE at E9.0 (18–20 somites) (Seo and Kume, 2006). Compound Foxc1//Foxc2/ mutants displayed abnormal formation of the epicardium at the conoventricular junction, characterized by premature epithelial-to-mesenchymal transformation and precocious differentiation into smooth muscle and endothelial cells (Seo and Kume, 2006).

III.G.  Cited2 and Pbx Genes The Cited2 gene in mice has been reported to be expressed in the septum transversum mesenchyme (Dunwoodie et al., 1998). Cited2 is a co-factor for the transcription factor TFAP2, which was recently implicated as a regulator of Pitx2 expression in the lateral plate mesoderm (Bamforth et al., 2004; Weninger et al., 2005). Analysis of mouse embryos between the early heart field and tubular heart stages suggest that the septum transversum precursors are labeled by Cited2 early in development (Dunwoodie et al., 1998). At head fold stage in the mouse, Cited2-­expressing cells are present rostral to the heart fields and they appear to move posterior into the forming septum transversum simultaneously with foregut and heart tube formation (Dunwoodie et al., 1998). These data would indicate that Cited2 is an early marker for precursors of the forming proepicardium. However, the avian homolog does not display this pattern of expression (Schlange et al., 2000a). Moreover, the Cited2-null mutant has no phenotype associated with epicardium formation (Bamforth et al., 2004; Weninger et al., 2005). Thus, at present the role of the

Chapter  |  5.1  Epicardial Lineage: Origins and Fates

Cited family of transcription factors with respect to proepicardium development is not fully clarified. Pbx homeodomain factors interact with Hox genes to modify their function. Two of the three Pbx genes, Pbx1 and Pbx3, display a strong and distinct expression in the septum transversum and pericardium (Di Giacomo et al., 2006). Loss of Pbx1 results in multi-organ hypoplasia (Selleri et al., 2001). Unfortunately, a specific analysis of cardiac development has not been performed. Possibly a double-null mutant of Pbx1 and Pbx3 will reveal whether the Pbx family has an essential function during epicardial development.

III.H.  The Bves/Popdc Gene Family The Bves/Popdc1 gene encodes a membrane protein belonging to the Popeye domain-containing gene family (Osler et al., 2006). These proteins are believed to function in the context of cell–cell interaction, and recently Bves/ Popdc1 was found to interact with the PDZ adapter protein ZO-1 (Osler et al., 2005). Using antibodies directed against conserved peptides of the Bves/Popdc1 protein, Bader and colleagues have shown that the proepicardium is a major expression domain of this factor, and in subsequent stages expression is also seen in the smooth muscle layer of the forming coronary arteries (Reese et al., 1999). Similarly it was found that an epicardial cell line of the rat expresses Bves/Popdc1, as revealed by antibody staining and reverse transcription-polymerase chain reaction (RTPCR) (Wada et al., 2003a). In contrast, Duncan and coworkers were unable to corroborate these findings with an independently-generated monoclonal antibody (Vasavada et al., 2004), and RT-PCR utilizing microdissected proepicardium explants did not provide evidence for the presence of the Popeye genes in the proepicardium (Torlopp et al., 2006).

IV.  Evo–devo aspects An epicardium probably forms in all vertebrate hearts. While its formation has been well-studied in avians (Virágh et al., 1993) and mammals (Virágh and Challice, 1981; Komiyama et al., 1987) much less is known in the case of amphibian and fish hearts. In Xenopus, evidence for asymmetric proepicardium formation has been recently reported (Jahr et al., 2008). The proepicardium forms as a cone-shaped accumulation of Tbx18-labeled mesothelial cells on the surface of the right sinus horn. Subsequently, a secondary tissue bridge is established facilitating the transfer of proepicardium cells to the heart. As in the chick embryo, proepicardium development in Xenopus proceeds in a bilaterally-asymmetric pattern. Several marker genes of proepicardium development have been

337

cloned in Xenopus, and expression in the proepicardium or epicardium has been described (Horb and Thomsen, 1999; Simrick et al., 2005). A morphological description of epicardium formation is also available for the axolotl (Fransen and Lemanski, 1990). Here, a tissue bridge is found (ventricular pericardial stalk) at stage 43, which suggests that in this amphibian species a cell transfer mode also exists that is similar to the one observed in the avian embryo. However, the authors also found evidence for proepicardium cyst formation. Thus, in axolotl cell transfer occurs probably both by proepicardium cysts, and by tissue bridges. Both cell transfer modes are also found in mammals and avian embryos. In the chick, a minor population of cells are transferred via proepicardium cysts (Sejima et al., 2001), and in mammals the atria are colonized by cell migration independent of proepicardium cyst formation (Hirose et al., 2006). In the dogfish, proepicardium cell transfer occurs through cyst formation on the left side, while the right proepicardium forms a tissue bridge (Munoz-Chapuli et al., 1994; Pombal et al., 2008). Recently, proepicardium development has also been studied in zebrafish (Begemann et al., 2002; Serluca, 2008). Bilateral proepicardium development occurs in zebrafish mutants with cardia bifida, suggesting that both heart fields have equal competence to generate a proepicardium. This is similar to what can be observed in mice, but is different from the chick where bifid embryos only develop a proepicardium on the right side (Schulte et al., 2007; Maretto et al., 2008; Serluca, 2008). The heart of the tunicate Ciona intestinalis (see Chapter 2.1) lacks an epicardium (Davidson, 2006), suggesting that a strategy evolved to cover the myocardium with an additional tissue layer probably only during vertebrate heart evolution, with the progressive requirement for a functional coronary artery system. In the lamprey, the epicardium originates from the pronephric external glomerula cells (PEG), which are formed on either side of the midline. The right-sided PEG contacts the ventricular surface and forms a tissue bridge (Pombal et al., 2008). This ancient origin of the proepicardium as part of the pronephric tissue might also explain why the proepicardium and the intermediate mesoderm share the expression of several transcription factor genes, such as Wt1, Tcf21 and Tbx18. In Drosophila the heart is composed of two different tissue layers (see Chapter 1.2), an inner layer of cardiac myocytes and an outer layer of pericardial cells. Whether this pericardium has any molecular or functional homology to the epicardium in vertebrates is unclear. Significantly, the pericardium and myocardium in Drosophila are formed from the same pool of precursor cells, and both cell types share several transcription factors (Cripps and Olson, 2002; Fujioka et al., 2005; Reim and Frasch, 2005). This situation in the insect heart is reminiscent of vertebrate epicardium and the myocardium, which during early development have an overlapping pattern of gene expression (Schlueter et al.,

338

2006). One of the functions of the fly’s pericardium is to serve as a mechanical support for the pumping activity of the heart (Yi et al., 2006). A similar function has been proposed for the avian epicardium (Männer et al., 2005). Thus, pericardial development in Drosophila might serve as a model for vertebrate proepicardium formation.

V.  The epicardium: a progenitor cell population Mesothelial cells line the surface of all inner organs and body cavities. Many of the genes that are expressed in the proepicardium and epicardium are also present in other mesothelial cells. This is, for example, true for Wt1, which is expressed in nearly all mesothelial cells (Carmona et al., 2001), and in the amniote heart is present in the epicardium and pericardium (Robb et al., 1998; Watt et al., 2004; Schlueter et al., 2006). This is also true for CFC1 and TBX18 in the chick and Raldh2 in mice (Robb et al., 1998; Schlueter et al., 2006). However, expression in these cases is often weaker in the pericardium than in the epicardium. There are also genes such as epicardin that are expressed in the proepicardium and epicardium, but are absent from the pericardium (Robb et al., 1998; Schlueter et al., 2006). Given that there are apparently specific patterns of gene expression in mesothelial cells of different origin, it is likely that different types of mesothelial cells exist (Herrick and Mutsaers, 2004). The property of being able to generate smooth muscle cells is apparently more widespread and has been demonstrated not only for the epicardium, but also found in the case of gut mesothelium (Wilm et al., 2005). As mentioned already, a subpopulation of proepicardium cells have the ability to form cardiac myocytes after the addition of BMP2 or serum (Kruithof et al., 2006; Schlueter et al., 2006). This ability of proepicardium cells in culture is lost once the proepicardium cells have made contact to the myocardium, suggesting that cell–cell contact might alter the spectrum of cell types that can be formed by cardiac mesothelial cells. In the mouse embryo, the epicardium also contributes to the myocardial cell lin­ eage (Cai et al., 2008; Zhou et al., 2008). How much stemness remains present in the adult epicardium? The ability to generate smooth muscle cells is not only present during embryogenesis, but seems to be retained in the adult, as demonstrated for epicardial cells of human atrial origin (van Tuyn et al., 2007), or rat heart (Eid et al., 1992, 1994). Thymosin 4 is able to induce the formation of smooth muscle cells, endothelial cells and fibroblasts from adult epicardium (Smart et al., 2007). Thus, the adult epicardium under certain circumstances can be activated and may act as a stem cell population. Possibly this can be exploited by regenerative medicine in order to restore the coronary vasculature.

PART  |  5  Epicardial Development

VI.  The role of the epicardium in heart regeneration The epicardium also plays an important role during adult heart regeneration. While the mammalian heart is not able to regenerate myocardium after myocardial infarction, the zebrafish or newt heart can fully regenerate myocardial tissue after surgical amputation of the ventricular apex (Poss et al., 2002) (for further discussion of this topic see Chapter 12.2). It has been proposed that cardiac regeneration in zebrafish involves myocyte dedifferentiation and redifferentiation (Raya et al., 2003). However, recent analysis using two different fluorescent reporter genes indicated that the majority of newly-regenerated myocardium is derived from a yet-unidentified progenitor cell population (Lepilina et al., 2006). The epicardium is apparently involved in the regeneration process, and even atrial epicardium far distant from the wound after myocardial amputation becomes activated and re-expresses embryonic proepicardium marker genes such as raldh2 and tbx18 (Lepilina et al., 2006). The question of whether epicardium is the site of origin of the progenitor cell population involved in zebrafish myocardial regeneration has not been fully-explored. Recruitment of epicardial cells to the blastema tissue involves FGF (fgf17b) signaling, which can be blocked by heat-shockinduced expression of a ­ dominant-negative FGF receptor (Lepilina et al., 2006). In the absence of epicardial cell recruitment no coronary arteries are formed, resulting in scar formation, however, initial stages of myocardial regeneration were undisturbed. It is therefore unlikely that the epicardium harbors a progenitor cell population involved in myocardial regeneration of the zebrafish heart. In mammals there is one report which observed re-expression of Wt1 in cells of the coronary vasculature of infarcted rat hearts (Wagner et al., 2002). However, whether this expression domain represents newly-recruited cells that have been incorporated into existing coronary arteries is presently unknown. Recent analysis of the transcriptional response of regenerating myocardium in zebrafish revealed a variety of transcripts that are differentially-expressed in the blastema which may be involved in modulating myocardial cell proliferation (Lien et al., 2006). Interestingly, in this study Thymosin 4 was found to be induced in the amputated zebrafish heart and may be required for vascularization of the regenerating heart analogous to the situation in mammals (Smart et al., 2007).

VII.  Outlook Over the last decade, a wealth of physiological, cell biological and molecular information has been accumulated on the role of the epicardium in heart development, and it is now well-established that nonmuscle cells present

339

Chapter  |  5.1  Epicardial Lineage: Origins and Fates

in the myocardium are to a large extent derived from the proepicardium. However, we still do not understand how these different cell lineages are generated. In the avian embryo an early separation of endothelial and smooth muscle cell precursors is observed, however we lack knowledge about underlying molecular mechanisms. Can proepicardium cells that are destined to form endothelial or smooth muscle cells be distinguished at the molecular level? There are data suggesting that at least in vitro a subpopulation of proepicardium cells have the ability to modify their cell fate, depending on the environment. For example, proepicardium cells can differentiate into cardiac myocytes in vitro. This has not been observed by cell fate analysis to occur in the chick embryo; however, differentiation of epicardial cells to the myocardial lineage was recently reported in the mouse embryo. What restricts proepicardium cell fate decisions in vivo, and why this is more relaxed in vitro? Identification of further genes that are expressed in the proepicardium may provide hints for gaining better understanding of these unresolved issues. Additionally, promoter analysis of proepicardium marker genes would facilitate the identification of novel signaling pathways and transcription factors that are involved in specifying this important cell lineage. While paracrine signals, such as BMP and FGF, are important regulators of proepicardium development, it is unclear if any of these signaling factors would be sufficient to induce proepicardium cells in vivo. The recent identification of the liver endoderm as a potential proepicardium inducer will make it possible to identify further molecular signals that are involved in its induction. Surprisingly, liver-induced ectopic induction of proepicardium marker gene expression did not differ between the left and right side of the embryo. Thus, there may be an additional mechanism for a side-specific responsiveness to induction of proepicardium marker gene expression seen in endogenous development. How and when left–right asymmetry is established during proepicardium development remains to be determined. Finally, the observed activation of epicardial cells during zebrafish heart regeneration and in the postinfarcted rat heart will obviously attract further investigation with regard to the function of epicardial cells in this process; the recent observation of the ability of the embryonic mouse epicardium to deliver cardiac myocytes points to the possible existence of a dormant ability of the mammalian epicardium to deliver cardiac myocytes during cardiac regeneration. The presence of an epicardium in the zebrafish, and the use of similar molecular pathways during epicardium formation and cardiac regeneration, will allow utilization of the power of zebrafish genetics to decipher further details of epicardiogenesis, cell fate allocation and the role of the epicardium in cardiac regeneration.

References Adamson, E.D., Minchiotti, G., Salomon, D.S., 2002. Cripto: a tumor growth factor and more. J. Cell Physiol. 190, 267–278. Aikawa, E., Kawano, J., 1982. Formation of coronary arteries sprouting from the primitive aortic sinus wall of the chick embryo. Experientia 38, 816–818. Andrée, B., Duprez, D., Vorbusch, B., Arnold, H.-H., Brand, T., 1998. BMP-2 induces ectopic expression of cardiac lineage markers and interferes with somite formation in chicken embryos. Mech. Dev. 70, 119–131. Aviezer, D., Levy, E., Safran, M., Svahn, C., Buddecke, E., Schmidt, A., David, G., Vlodavsky, I., Yayon, A., 1994. Differential structural requirements of heparin and heparan sulfate proteoglycans that promote binding of basic fibroblast growth factor to its receptor. J. Biol. Chem. 269, 114–121. Baldwin, H.S., 1996. Early embryonic vascular development. Cardiovasc. Res. 31, E34–E45. Bamforth, S.D., Braganca, J., Farthing, C.R., Schneider, J.E., Broadbent, C., Michell, A.C., Clarke, K., Neubauer, S., Norris, D., Brown, N.A., Anderson, R.H., Bhattacharya, S., 2004. Cited2 controls left–right patterning and heart development through a Nodal-Pitx2c pathway. Nat. Genet. 36, 1189–1196. Begemann, G., Gibert, Y., Meyer, A., Ingham, P., 2002. Cloning of zebrafish T-box genes tbx15 and tbx18 and their expression during embryonic development. Mech. Dev. 114, 137–141. Bernanke, D.H., Velkey, J.M., 2002. Development of the coronary blood supply: changing concepts and current ideas. Anat. Rec. 269, 198–208. Bimber, B., Dettman, R.W., Simon, H.G., 2007. Differential regulation of Tbx5 protein expression and sub-cellular localization during heart development. Dev. Biol. 302, 230–242. Bogers, A.J., Gittenberger-de Groot, A.C., Poelmann, R.E., Peault, B.M., Huysmans, H.A., 1989. Development of the origin of the coronary arteries, a matter of ingrowth or outgrowth? Anat. Embryol. (Berl.) 180, 437–441. Brand, T., 2003. Heart development: molecular insights into cardiac specification and early morphogenesis. Dev. Biol. 258, 1–19. Bruneau, B.G., Nemer, G., Schmitt, J.P., Charron, F., Robitaille, L., Caron, S., Conner, D.A., Gessler, M., Nemer, M., Seidman, C.E., Seidman, J.G., 2001. A murine model of Holt-Oram syndrome defines roles of the T-box transcription factor Tbx5 in cardiogenesis and disease. Cell 106, 709–721. Cai, C.L., Martin, J.C., Sun, Y., Cui, L., Wang, L., Ouyang, K., Yang, L., Bu, L., Liang, X., Zhang, X., Stallcup, W.B., Denton, C.P., McCulloch, A., Chen, J., Evans, S.M., 2008. A myocardial lineage derives from Tbx18 epicardial cells. Nature 454, 104–108. Calmont, A., Wandzioch, E., Tremblay, K.D., Minowada, G., Kaestner, K.H., Martin, G.R., Zaret, K.S., 2006. An FGF response pathway that mediates hepatic gene induction in embryonic endoderm cells. Dev. Cell 11, 339–348. Camarata, T., Bimber, B., Kulisz, A., Chew, T.L., Yeung, J., Simon, H.G., 2006. LMP4 regulates Tbx5 protein subcellular localization and activity. J. Cell Biol. 174, 339–348. Carmona, R., Gonzalez-Iriarte, M., Perez-Pomares, J.M., Munoz-Chapuli, R., 2001. Localization of the Wilms’ tumour protein WT1 in avian embryos. Cell Tissue Res. 303, 173–186. Chen, T.H., Chang, T.C., Kang, J.O., Choudhary, B., Makita, T., Tran, C.M., Burch, J.B., Eid, H., Sucov, H.M., 2002. Epicardial induction of fetal

340

cardiomyocyte proliferation via a retinoic acid-inducible trophic factor. Dev. Biol. 250, 198–207. Christoffels, V.M., Mommersteeg, M.T., Trowe, M.O., Prall, O.W., de Gier-de Vries, C., Soufan, A.T., Bussen, M., Schuster-Gossler, K., Harvey, R.P., Moorman, A.F., Kispert, A., 2006. Formation of the venous pole of the heart from an Nkx2-5-negative precursor population requires Tbx18. Circ. Res. 98, 1555–1563. Christoffels, V.M., Grieskamp, T., Norden, J., Mommersteeg, M.T., Rudat, C., Kispert, A., 2009. Tbx18 and the fate of PE cells. Nature 485, E8–E9. Compton, L.A., Potash, D.A., Mundell, N.A., Barnett, J.V., 2006. Transforming growth factor-beta induces loss of epithelial character and smooth muscle cell differentiation in epicardial cells. Dev. Dyn. 235, 82–93. Creazzo, T.L., Godt, R.E., Leatherbury, L., Conway, S.J., Kirby, M.L., 1998. Role of cardiac neural crest cells in cardiovascular development. Annu. Rev. Physiol. 60, 267–286. Cripps, R.M., Olson, E.N., 2002. Control of cardiac development by an evolutionarily conserved transcriptional network. Dev. Biol. 246, 14–28. Crispino, J.D., Lodish, M.B., Thurberg, B.L., Litovsky, S.H., Collins, T., Molkentin, J.D., Orkin, S.H., 2001. Proper coronary vascular development and heart morphogenesis depend on interaction of GATA-4 with FOG cofactors. Genes Dev. 15, 839–844. Dame, C., Kirschner, K.M., Bartz, K.V., Wallach, T., Hussels, C.S., Scholz, H., 2006. Wilms’ tumor suppressor, Wt1, is a transcriptional activator of the erythropoietin gene. Blood 107, 4282–4290. Davidson, B., 2007. Ciona intestinalis as a model for cardiac development. Semin. Cell Dev. Biol. 18, 16–26. Davis, S., Aldrich, T.H., Jones, P.F., Acheson, A., Compton, D.L., Jain, V., Ryan, T.E., Bruno, J., Radziejewski, C., Maisonpierre, P.C., Yancopoulos, G.D., 1996. Isolation of angiopoietin-1, a ligand for the TIE2 receptor, by secretion-trap expression cloning. Cell 87, 1161–1169. de Lange, F.J., Moorman, A.F., Anderson, R.H., Männer, J., Soufan, A.T., de Gier-de Vries, C., Schneider, M.D., Webb, S., van den Hoff, M.J., Christoffels, V.M., 2004. Lineage and morphogenetic analysis of the cardiac valves. Circ. Res. 95, 645–654. Dettman Jr, R.W., Denetclaw, W., Ordahl, C.P., Bristow, J., 1998. Common epicardial origin of coronary vascular smooth muscle, perivascular fibroblasts, and intermyocardial fibroblasts in the avian heart. Dev. Biol. 193, 169–181. Di Giacomo, G., Koss, M., Capellini, T.D., Brendolan, A., Popperl, H., Selleri, L., 2006. Spatio-temporal expression of Pbx3 during mouse organogenesis. Gene Expr. Patterns 6, 747–757. Dias, R.R., Albuquerque, J.M., Pereira, A.C., Stolf, N.A., Krieger, J.E., Mady, C., Oliveira, S.A., 2007. Holt-Oram syndrome presenting as agenesis of the left pericardium. Int. J. Cardiol. 114, 98–100. Douarin, N.M., 1975. An experimental analysis of liver development. Med. Biol. 53, 427–455. Dumont, D.J., Gradwohl, G., Fong, G.H., Puri, M.C., Gertsenstein, M., Auerbach, A., Breitman, M.L., 1994. Dominant-negative and targeted null mutations in the endothelial receptor tyrosine kinase, tek, reveal a critical role in vasculogenesis of the embryo. Genes Dev. 8, 1897–1909. Duncan, S.A., 2003. Mechanisms controlling early development of the liver. Mech. Dev. 120, 19–33. Dunwoodie, S.L., Rodriguez, T.A., Beddington, R.S., 1998. Msg1 and Mrg1, founding members of a gene family, show distinct patterns

PART  |  5  Epicardial Development

of gene expression during mouse embryogenesis. Mech. Dev. 72, 27–40. Eid, H., Larson, D.M., Springhorn, J.P., Attawia, M.A., Nayak, R.C., Smith, T.W., Kelly, R.A., 1992. Role of epicardial mesothelial cells in the modification of phenotype and function of adult rat ventricular myocytes in primary coculture. Circ. Res. 71, 40–50. Eid, H., de Bold, M.L., Chen, J.H., de Bold, A.J., 1994. Epicardial mesothelial cells synthesize and release endothelin. J. Cardiovasc. Pharmacol. 24, 715–720. Elices, M.J., Osborn, L., Takada, Y., Crouse, C., Luhowskyj, S., Hemler, M.E., Lobb, R.R., 1990. VCAM-1 on activated endothelium interacts with the leukocyte integrin VLA-4 at a site distinct from the VLA-4/fibronectin binding site. Cell 60, 577–584. Fransen, M.E., Lemanski, L.F., 1990. Epicardial development in the axolotl, Ambystoma mexicanum. Anat. Rec. 226, 228–236. Fujioka, M., Wessells, R.J., Han, Z., Liu, J., Fitzgerald, K., Yusibova, G.L., Zamora, M., Ruiz-Lozano, P., Bodmer, R., Jaynes, J.B., 2005. Embryonic even skipped-dependent muscle and heart cell fates are required for normal adult activity, heart function, and lifespan. Circ. Res. 97, 1108–1114. Fukuda-Taira, S., 1981. Hepatic induction in the avian embryo: specificity of reactive endoderm and inductive mesoderm. J. Embryol. Exp. Morph. 63, 111–125. Funato, N., Ohyama, K., Kuroda, T., Nakamura, M., 2003. Basic helixloop-helix transcription factor epicardin/capsulin/Pod-1 suppresses differentiation by negative regulation of transcription. J. Biol. Chem. 278, 7486–7493. Garry, D.J., Olson, E.N., 2006. A common progenitor at the heart of development. Cell 127, 1101–1104. Ghatpande, S., Ghatpande, A., Zile, M., Evans, T., 2000. Anterior endoderm is sufficient to rescue foregut apoptosis and heart tube morphogenesis in an embryo lacking retinoic acid. Dev. Biol. 219, 59–70. Ghatpande, S., Brand, T., Zile, M., Evans, T., 2006. Bmp2 and Gata4 function additively to rescue heart tube development in the absence of retinoids. Dev. Dyn. 235, 2030–2039. Gittenberger-de Groot, A.C., Vrancken Peeters, M.P., Mentink, M.M., Gourdie, R.G., Poelmann, R.E., 1998. Epicardium-derived cells contribute a novel population to the myocardial wall and the atrioventricular cushions. Circ. Res. 82, 1043–1052. Gittenberger-de Groot, A.C., Vrancken Peeters, M.P., Bergwerff, M., Mentink, M.M., Poelmann, R.E., 2000. Epicardial outgrowth inhibition leads to compensatory mesothelial outflow tract collar and abnormal cardiac septation and coronary formation. Circ. Res. 87, 969–971. Gourdie, R.G., Kubalak, S., Mikawa, T., 1999. Conducting the embryonic heart: orchestrating development of specialized cardiac tissues. Trends Cardiovasc. Med. 9, 18–26. Gray, P.C., Harrison, C.A., Vale, W., 2003. Cripto forms a complex with activin and type II activin receptors and can block activin signaling. Proc. Natl. Acad. Sci. U.S.A. 100, 5193–5198. Gray, P.C., Shani, G., Aung, K., Kelber, J., Vale, W., 2006. Cripto binds transforming growth factor beta (TGF-beta) and inhibits TGF-beta signaling. Mol. Cell Biol. 26, 9268–9278. Gualdi, R., Bossard, P., Zheng, M., Hamada, Y., Coleman, J.R., Zaret, K.S., 1996. Hepatic specification of the gut endoderm in vitro: cell signaling and transcriptional control. Genes Dev. 10, 1670–1682. Guan, J.L., Hynes, R.O., 1990. Lymphoid cells recognize an alternatively spliced segment of fibronectin via the integrin receptor alpha 4 beta 1. Cell 60, 53–61.

Chapter  |  5.1  Epicardial Lineage: Origins and Fates

Hatcher, C.J., Diman, N.Y., Kim, M.S., Pennisi, D., Song, Y., Goldstein, M.M., Mikawa, T., Basson, C.T., 2004. A role for Tbx5 in proepicardial cell migration during cardiogenesis. Physiol. Genomics 18, 129–140. Herrick, S.E., Mutsaers, S.E., 2004. Mesothelial progenitor cells and their potential in tissue engineering. Int. J. Biochem. Cell Biol. 36, 621–642. Hirakow, R., 1983. Development of the cardiac blood vessels in staged human embryos. Acta Anat. (Basel) 115, 220–230. Hirose, T., Karasawa, M., Sugitani, Y., Fujisawa, M., Akimoto, K., Ohno, S., Noda, T., 2006. PAR3 is essential for cyst-mediated epicardial development by establishing apical cortical domains. Development 133, 1389–1398. Hiruma, T., Hirakow, R., 1989. Epicardial formation in embryonic chick heart: computer-aided reconstruction, scanning, and transmission electron microscopic studies. Am. J. Anat. 184, 129–138. Ho, E., Shimada, Y., 1978. Formation of the epicardium studied with the scanning electron microscope. Dev. Biol. 66, 579–585. Horb, M.E., Thomsen, G.H., 1999. Tbx5 is essential for heart development. Development 126, 1739–1751. Hosono, S., Gross, I., English, M.A., Hajra, K.M., Fearon, E.R., Licht, J.D., 2000. E-cadherin is a WT1 target gene. J. Biol. Chem. 275, 10943–10953. Hutchins, G.M., Kessler-Hanna, A., Moore, G.W., 1988. Development of the coronary arteries in the embryonic human heart. Circulation 77, 1250–1257. Hyer, J., Johansen, M., Prasad, A., Wessels, A., Kirby, M.L., Gourdie, R.G., Mikawa, T., 1999. Induction of Purkinje fiber differentiation by coronary arterialization. Proc. Natl. Acad. Sci. U.S.A. 96, 13214–13218. Ishii, Y., Langberg, J.D., Hurtado, R., Lee, S., Mikawa, T., 2007. Induction of proepicardial marker gene expression by the liver primordium. Development 134, 3627–3637. Jahr, M., Schlueter, J., Brand, T., Männer, J., 2008. Development of the proepicardium in Xenopus laevis. Dev. Dyn. 237, 3088–3096. Jenkins, S.J., Hutson, D.R., Kubalak, S.W., 2005. Analysis of the ­proepicardium-epicardium transition during the malformation of the RXRalpha/ epicardium. Dev. Dyn. 233, 1091–1101. Jiang, X., Rowitch, D.H., Soriano, P., McMahon, A.P., Sucov, H.M., 2000. Fate of the mammalian cardiac neural crest. Development 127, 1607–1616. Jung, J., Zheng, M., Goldfarb, M., Zaret, K.S., 1999. Initiation of mammalian liver development from endoderm by fibroblast growth factors. Science 284, 1998–2003. Kalinichenko, V.V., Gusarova, G.A., Shin, B., Costa, R.H., 2003. The forkhead box F1 transcription factor is expressed in brain and head mesenchyme during mouse embryonic development. Gene Expr. Patterns 3, 153–158. Kastner, P., Grondona, J.M., Mark, M., Gansmuller, A., LeMeur, M., Decimo, D., Vonesch, J.L., Dolle, P., Chambon, P., 1994. Genetic analysis of RXR alpha developmental function: convergence of RXR and RAR signaling pathways in heart and eye morphogenesis. Cell 78, 987–1003. Kattman, S.J., Huber, T.L., Keller, G.M., 2006. Multipotent flk-1 cardio­ vascular progenitor cells give rise to the cardiomyocyte, endothelial, and vascular smooth muscle lineages. Dev. Cell 11, 723–732. Kirschner, K.M., Wagner, N., Wagner, K.D., Wellmann, S., Scholz, H., 2006. The Wilms’ tumor suppressor Wt1 promotes cell adhesion through transcriptional activation of the alpha4integrin gene. J. Biol. Chem. 281, 31930–31939.

341

Komiyama, M., Ito, K., Shimada, Y., 1987. Origin and development of the epicardium in the mouse embryo. Anat. Embryol. (Berl.) 176, 183–189. Kostetskii, I., Jiang, Y., Kostetskaia, E., Yuan, S., Evans, T., Zile, M., 1999. Retinoid signaling required for normal heart development regulates GATA-4 in a pathway distinct from cardiomyocyte differentiation. Dev. Biol. 206, 206–218. Krause, A., Zacharias, W., Camarata, T., Linkhart, B., Law, E., Lischke, A., Miljan, E., Simon, H.G., 2004. Tbx5 and Tbx4 transcription factors interact with a new chicken PDZ-LIM protein in limb and heart development. Dev. Biol. 273, 106–120. Kreidberg, J.A., Sariola, H., Loring, J.M., Maeda, M., Pelletier, J., Housman, D., Jaenisch, R., 1993. WT-1 is required for early kidney development. Cell 74, 679–691. Kruithof, B.P., van Wijk, B., Somi, S., Kruithof-de Julio, M., Perez Pomares, J.M., Weesie, F., Wessels, A., Moorman, A.F., van den Hoff, M.J., 2006. BMP and FGF regulate the differentiation of multipotential pericardial mesoderm into the myocardial or epicardial lineage. Dev. Biol. 295, 507–522. Kwee, L., Baldwin, H.S., Shen, H.M., Stewart, C.L., Buck, C., Buck, C.A., Labow, M.A., 1995. Defective development of the embryonic and extraembryonic circulatory systems in vascular cell adhesion molecule (VCAM-1) deficient mice. Development 121, 489–503. Landerholm, T.E., Dong, X.R., Lu, J., Belaguli, N.S., Schwartz, R.J., Majesky, M.W., 1999. A role for serum response factor in coronary smooth muscle differentiation from proepicardial cells. Development 126, 2053–2062. Lavine, K.J., Yu, K., White, A.C., Zhang, X., Smith, C., Partanen, J., Ornitz, D.M., 2005. Endocardial and epicardial derived FGF signals regulate myocardial proliferation and differentiation in vivo. Dev. Cell 8, 85–95. Lavine, K.J., White, A.C., Park, C., Smith, C.S., Choi, K., Long, F., Hui, C.C., Ornitz, D.M., 2006. Fibroblast growth factor signals regulate a wave of hedgehog activation that is essential for coronary vascular development. Genes Dev. 20, 1651–1666. Lepilina, A., Coon, A.N., Kikuchi, K., Holdway, J.E., Roberts, R.W., Burns, C.G., Poss, K.D., 2006. A dynamic epicardial injury response supports progenitor cell activity during zebrafish heart regeneration. Cell 127, 607–619. Li, W.E., Waldo, K., Linask, K.L., Chen, T., Wessels, A., Parmacek, M.S., Kirby, M.L., Lo, C.W., 2002. An essential role for connexin43 gap junctions in mouse coronary artery development. Development 129, 2031–2042. Lie-Venema, H., Gittenberger-de Groot, A.C., van Empel, L.J., Boot, M.J., Kerkdijk, H., de Kant, E., DeRuiter, M.C., 2003. Ets-1 and Ets-2 transcription factors are essential for normal coronary and myocardial development in chicken embryos. Circ. Res. 92, 749–756. Lien, C.L., Schebesta, M., Makino, S., Weber, G.J., Keating, M.T., 2006. Gene expression analysis of zebrafish heart regeneration. PLoS Biol. 4, e260. Lough, J., Barron, M., Brogley, M., Sugi, Y., Bolender, D.L., Zhu, X., 1996. Combined BMP-2 and FGF-4, but neither factor alone, induces cardiogenesis in non-precardiac embryonic mesoderm. Dev. Biol. 178, 198–202. Lu, J., Richardson, J.A., Olson, E.N., 1998. Capsulin: a novel bHLH transcription factor expressed in epicardial progenitors and mesenchyme of visceral organs. Mech. Dev. 73, 23–32. Lu, J., Chang, P., Richardson, J.A., Gan, L., Weiler, H., Olson, E.N., 2000. The basic helix-loop-helix transcription factor capsulin

342

controls spleen organogenesis. Proc. Natl. Acad. Sci. U.S.A. 97, 9525–9530. Lu, J.R., Bassel-Duby, R., Hawkins, A., Chang, P., Valdez, R., Wu, H., Gan, L., Shelton, J.M., Richardson, J.A., Olson, E.N., 2002. Control of facial muscle development by MyoR and capsulin. Science 298, 2378–2381. MacNeill, C., French, R., Evans, T., Wessels, A., Burch, J.B., 2000. Modular regulation of cGATA-5 gene expression in the developing heart and gut. Dev. Biol. 217, 62–76. Mahlapuu, M., Enerback, S., Carlsson, P., 2001. Haploinsufficiency of the forkhead gene Foxf1, a target for sonic hedgehog signaling, causes lung and foregut malformations. Development 128, 2397–2406. Majesky, M.W., 2004. Development of coronary vessels. Curr. Topics Dev. Biol. 62, 225–259. Manasek, F.J., 1969. Embryonic development of the heart II. Formation of the epicardium. J. Embryol. Exp. Morphol. 22, 333–348. Männer, J., 1992. The development of pericardial villi in the chick embryo. Anat Embryol. (Berl.) 186, 379–385. Männer, J., 1993. Experimental study on the formation of the epicardium in chick embryos. Anat Embryol. 187, 281–289. Männer, J., 1999. Does the subepicardial mesenchyme contribute myocardioblasts to the myocardium of the chick embryo heart? A quailchick chimera study tracing the fate of the epicardial primordium. Anat Rec. 255, 212–226. Männer, J., 2006. Extracardiac tissues and the epigenetic control of myocardial development in vertebrate embryos. Ann Anat. 188, 199–212. Männer, J., Perez-Pomares, J.M., Macias, D., Munoz-Chapuli, R., 2001. The origin, formation and developmental significance of the epicardium: a review. Cells Tissues Organs 169, 89–103. Männer, J., Schlueter, J., Brand, T., 2005. Experimental analyses of the function of the proepicardium using a new microsurgical procedure to induce loss-of-proepicardial-function in chick embryos. Dev. Dyn. 233, 1454–1463. Maretto, S., Muller, P.S., Aricescu, A.R., Cho, K.W., Bikoff, E.K., Robertson, E.J., 2008. Ventral closure, headfold fusion and definitive endoderm migration defects in mouse embryos lacking the fibronectin leucine-rich transmembrane protein FLRT3. Dev. Biol. 318, 184–193. Merki, E., Zamora, M., Raya, A., Kawakami, Y., Wang, J., Zhang, X., Burch, J., Kubalak, S.W., Kaliman, P., Belmonte, J.C., Chien, K.R., Ruiz-Lozano, P., 2005. Epicardial retinoid X receptor alpha is required for myocardial growth and coronary artery formation. Proc. Natl. Acad. Sci. U.S.A. 102, 18455–18460. Mikawa, T., 1995. Retroviral targeting of FGF and FGFR in cardiomyocytes and coronary vascular cells during heart development. Ann. N.Y. Acad. Sci. 752, 506–516. Mikawa, T., Fischman, D.A., 1992. Retroviral analysis of cardiac morphogenesis: discontinuous formation of coronary vessels. Proc. Natl. Acad. Sci. U.S.A. 89, 9504–9508. Mikawa, T., Gourdie, R.G., 1996. Pericardial mesoderm generates a population of coronary smooth muscle cells migrating into the heart along with ingrowth of the epicardial organ. Dev. Biol. 174, 221–232. Mima, T., Ueno, H., Fischman, D.A., Williams, L.T., Mikawa, T., 1995. Fibroblast growth factor receptor is required for in vivo cardiac myocyte proliferation at early embryonic stages of heart development. Proc. Natl. Acad. Sci. U.S.A. 92, 467–471. Moore, A.W., McInnes, L., Kreidberg, J., Hastie, N.D., Schedl, A., 1999. YAC complementation shows a requirement for Wt1 in the development of epicardium, adrenal gland and throughout nephrogenesis. Development 126, 1845–1857.

PART  |  5  Epicardial Development

Moorman, A.F., Christoffels, V.M., Anderson, R.H., van den Hoff, M.J., 2007. The heart-forming fields: one or multiple? Philos. Trans. R. Soc. Lond., B. Biol. Sci. 362, 1257–1265. Morabito, C.J., Dettman, R.W., Kattan, J., Collier, J.M., Bristow, J., 2001. Positive and negative regulation of epicardial–mesenchymal transformation during avian heart development. Dev. Biol. 234, 204–215. Moretti, A., Caron, L., Nakano, A., Lam, J.T., Bernshausen, A., Chen, Y., Qyang, Y., Bu, L., Sasaki, M., Martin-Puig, S., Sun, Y., Evans, S.M., Laugwitz, K.L., Chien, K.R., 2006. Multipotent embryonic isl1 progenitor cells lead to cardiac, smooth muscle, and endothelial cell diversification. Cell 127, 1151–1165. Mori, A.D., Zhu, Y., Vahora, I., Nieman, B., Koshiba-Takeuchi, K., Davidson, L., Pizard, A., Seidman, J.G., Seidman, C.E., Chen, X.J., Henkelman, R.M., Bruneau, B.G., 2006. Tbx5-dependent rheostatic control of cardiac gene expression and morphogenesis. Dev. Biol. 297, 566–586. Mu, H., Ohashi, R., Lin, P., Yao, Q., Chen, C., 2005. Cellular and molecular mechanisms of coronary vessel development. Vasc. Med. 10, 37–44. Munoz-Chapuli, R., Macias, D., Ramos, C., de Andres, V., Gallego, A., Navarro, P., 1994. Cardiac development in the dogfish (Scyliorhinus canicula): a model for the study of vertebrate cardiogenesis. Cardioscience 5, 245–253. Nahirney, P.C., Mikawa, T., Fischman, D.A., 2003. Evidence for an extracellular matrix bridge guiding proepicardial cell migration to the myocardium of chick embryos. Dev. Dyn. 227, 511–523. Nelson, T.J., Ge, Z.D., Van Orman, J., Barron, M., Rudy-Reil, D., Hacker, T.A., Misra, R., Duncan, S.A., Auchampach, J.A., Lough, J.W., 2006. Improved cardiac function in infarcted mice after treatment with pluripotent embryonic stem cells. Anat. Rec. 288, 1216–1224. Nemer, G., Nemer, M., 2003. Transcriptional activation of BMP-4 and regulation of mammalian organogenesis by GATA-4 and -6. Dev. Biol. 254, 131–148. Nesbitt, T., Lemley, A., Davis, J., Yost, M.J., Goodwin, R.L., Potts, J.D., 2006. Epicardial development in the rat: a new perspective. Microsc. Microanal. 12, 390–398. Olivey, H.E., Compton, L.A., Barnett, J.V., 2004. Coronary vessel development: the epicardium delivers. Trends Cardiovasc. Med. 14, 247–251. Olivey, H.E., Mundell, N.A., Austin, A.F., Barnett, J.V., 2006. Transforming growth factor-beta stimulates epithelial-mesenchymal transformation in the proepicardium. Dev. Dyn. 235, 50–59. Onuma, Y., Yeo, C.Y., Whitman, M., 2006. XCR2, one of three Xenopus EGF-CFC genes, has a distinct role in the regulation of left–right patterning. Development 133, 237–250. Osborn, L., Hession, C., Tizard, R., Vassallo, C., Luhowskyj, S., Chi-Rosso, G., Lobb, R., 1989. Direct expression cloning of vascular cell adhesion molecule 1, a cytokine-induced endothelial protein that binds to lymphocytes. Cell 59, 1203–1211. Osler, M.E., Chang, M.S., Bader, D.M., 2005. Bves modulates epithelial integrity through an interaction at the tight junction. J. Cell Sci. 118, 4667–4678. Osler, M.E., Smith, T.K., Bader, D.M., 2006. Bves, a member of the Popeye domain-containing gene family. Dev. Dyn. 235, 586–593. Pennisi, D.J., Ballard, V.L., Mikawa, T., 2003. Epicardium is required for the full rate of myocyte proliferation and levels of expression of myocyte mitogenic factors FGF2 and its receptor, FGFR-1, but not for transmural myocardial patterning in the embryonic chick heart. Dev. Dyn. 228, 161–172. Perez-Pomares, J.M., Macias, D., Garcia-Garrido, L., Munoz-Chapuli, R., 1998. The origin of the subepicardial mesenchyme in the avian

Chapter  |  5.1  Epicardial Lineage: Origins and Fates

embryo: an immunohistochemical and quail-chick chimera study. Dev. Biol. 200, 57–68. Perez-Pomares, J.M., Phelps, A., Sedmerova, M., Carmona, R., GonzalezIriarte, M., Munoz-Chapuli, R., Wessels, A., 2002. Experimental studies on the spatiotemporal expression of WT1 and RALDH2 in the embryonic avian heart: a model for the regulation of myocardial and valvuloseptal development by epicardially derived cells (EPDCs). Dev. Biol. 247, 307–326. Peterson, R.S., Lim, L., Ye, H., Zhou, H., Overdier, D.G., Costa, R.H., 1997. The winged helix transcriptional activator HFH-8 is expressed in the mesoderm of the primitive streak stage of mouse embryos and its cellular derivatives. Mech. Dev. 69, 53–69. Plageman Jr., T.F., Yutzey, K.E., 2005. T-box genes and heart development: putting the “T” in heart. Dev. Dyn. 232, 11–20. Poelmann, R.E., Gittenberger-de Groot, A.C., Mentink, M.M., Bokenkamp, R., Hogers, B., 1993. Development of the cardiac coronary vascular endothelium, studied with antiendothelial antibodies, in chicken-quail chimeras. Circ. Res. 73, 559–568. Poelmann, R.E., Lie-Venema, H., Gittenberger-de Groot, A.C., 2002. The role of the epicardium and neural crest as extracardiac contributors to coronary vascular development. Tex. Heart Inst. J. 29, 255–261. Pombal, M., Carmona, R., Megias, M., Ruiz, A., Pérez-Pomares, J., Munoz-Chapuli, R., 2008. Evolutionary origin of the vertebrate epicardium from an ancestral pronephric external glomerulus. Evol. Dev. 10, 210–216. Poss, K.D., Wilson, L.G., Keating, M.T., 2002. Heart regeneration in zebrafish. Science 298, 2188–2190. Quaggin, S.E., Vanden Heuvel, G.B., Igarashi, P., 1998. Pod-1, a ­mesoderm-specific basic-helix-loop-helix protein expressed in mesenchymal and glomerular epithelial cells in the developing kidney. Mech. Dev. 71, 37–48. Raya, A., Koth, C.M., Buscher, D., Kawakami, Y., Itoh, T., Raya, R.M., Sternik, G., Tsai, H.J., Rodriguez-Esteban, C., Izpisua-Belmonte, J.C., 2003. Activation of Notch signaling pathway precedes heart regeneration in zebrafish. Proc. Natl. Acad. Sci. U.S.A. 100 (Suppl. 1), 11889–11895. Redkar, A., Montgomery, M., Litvin, J., 2002. Fate map of early avian cardiac progenitor cells. Development 128, 2269–2279. Reese, D.E., Mikawa, T., Bader, D.M., 2002. Development of the coronary vessel system. Circ. Res. 91, 761–768. Reese, D.E., Zavaljevski, M., Streiff, N.L., Bader, D., 1999. Bves: a novel gene expressed during coronary blood vessel development. Dev. Biol. 209, 159–171. Reim, I., Frasch, M., 2005. The Dorsocross T-box genes are key components of the regulatory network controlling early cardiogenesis in Drosophila. Development 132, 4911–4925. Robb, L., Mifsud, L., Hartley, L., Biben, C., Copeland, N.G., Gilbert, D.J., Jenkins, N.A., Harvey, R.P., 1998. Epicardin: a novel basic helixloop-helix transcription factor gene expressed in epicardium, branchial arch myoblasts, and mesenchyme of developing lung, gut, kidney, and gonads. Dev. Dyn. 213, 105–113. Rodgers, L.S., Lalani, S., Runyan, R.B., Camenisch, T.D., 2008. Differential growth and multicellular villi direct proepicardial translocation to the developing mouse heart. Dev Dyn. 237, 145–152. Rojas, A., De Val, S., Heidt, A.B., Xu, S.M., Bristow, J., Black, B.L., 2005. Gata4 expression in lateral mesoderm is downstream of BMP4 and is activated directly by Forkhead and GATA transcription factors through a distal enhancer element. Development 132, 3405–3417. Rossi, J.M., Dunn, N.R., Hogan, B.L., Zaret, K.S., 2001. Distinct mesodermal signals, including BMPs from the septum transversum

343

mesenchyme, are required in combination for hepatogenesis from the endoderm. Genes Dev. 15, 1998–2009. Saga, Y., Kitajima, S., Miyagawa-Tomita, S., 2000. Mesp1 expression is the earliest sign of cardiovascular development. Trends Cardiovasc. Med. 10, 345–352. Schlange, T., Andrée, B., Arnold, H., Brand, T., 2000a. Expression analysis of the chicken homologue of CITED2 during early stages of embryonic development. Mech. Dev. 98, 157–160. Schlange, T., Andree, B., Arnold, H.H., Brand, T., 2000b. BMP2 is required for early heart development during a distinct time period. Mech. Dev. 91, 259–270. Schlange, T., Schnipkoweit, I., Andree, B., Ebert, A., Zile, M.H., Arnold, H.H., Brand, T., 2001. Chick CFC controls Lefty1 expression in the embryonic midline and nodal expression in the lateral plate. Dev. Biol. 234, 376–389. Schlueter, J., Männer, J., Brand, T., 2006. BMP is an important regulator of proepicardial identity in the chick embryo. Dev. Biol. 295, 546–558. Schlueter, J., Brand, T., 2009. A right-sided pathway involving FGF8/ Snai1 controls asymmetric development of the proepicardium in the chick embryo. Proc. Natl. Acad. Sci. U.S.A. 106, 7485–7490. Scholz, H., Kirschner, K.M., 2005. A role for the Wilms’ tumor protein WT1 in organ development. Physiology (Bethesda) 20, 54–59. Schulte, I., Schlueter, J., Abu-Issa, R., Brand, T., Männer, J., 2007. Morphological and molecular left–right asymmetries in the development of the proepicardium: a comparative analysis on mouse and chick embryos. Dev. Dyn. 236, 684–695. Schultheiss, T.M., Xydas, S., Lassar, A.B., 1995. Induction of avian cardiac myogenesis by anterior endoderm. Development 121, 4203–4214. Schultheiss, T., Burch, J., Lassar, A., 1997. A role for bone morphogenetic proteins in the induction of cardiac myogenesis. Genes Dev. 11, 451–462. Sejima, H., Isokawa, K., Shimizu, O., Morikawa, T., Ootsu, H., Numata, K., Fukai, M., Kubota, S., Toda, Y., 2001. Possible participation of isolated epicardial cell clusters in the formation of chick embryonic epicardium. J. Oral Sci. 43, 109–116. Selleri, L., Depew, M.J., Jacobs, Y., Chanda, S.K., Tsang, K.Y., Cheah, K.S., Rubenstein, J.L., O’Gorman, S., Cleary, M.L., 2001. Requirement for Pbx1 in skeletal patterning and programming chon­drocyte proliferation and differentiation. Development 128, 3543–3557. Sengbusch, J.K., He, W., Pinco, K.A., Yang, J.T., 2002. Dual functions of a4b1 integrin in epicardial development: initial migration and longterm attachment. J. Cell Biol. 157, 873–882. Seo, S., Kume, T., 2006. Forkhead transcription factors, Foxc1 and Foxc2, are required for the morphogenesis of the cardiac outflow tract. Dev. Biol. 296, 421–436. Serls, A.E., Doherty, S., Parvatiyar, P., Wells, J.M., Deutsch, G.H., 2005. Different thresholds of fibroblast growth factors pattern the ventral foregut into liver and lung. Development 132, 35–47. Serluca, F.C., 2008. Development of the proepicardial organ in the zebrafish. Dev. Biol. 315, 18–27. Shen, M.M., 2007. Nodal signaling: developmental roles and regulation. Development 134, 1023–1034. Shikama, N., Lutz, W., Kretzschmar, R., Sauter, N., Roth, J.F., Marino, S., Wittwer, J., Scheidweiler, A., Eckner, R., 2003. Essential function of p300 acetyltransferase activity in heart, lung and small intestine formation. EMBO J. 22, 5175–5185. Shimada, Y., Ho, E., Toyota, N., 1981. Epicardial covering over myocardial wall in the chicken embryo as seen with the scanning electron microscope. Scan. Electron. Microsc. Pt 2, 275–280.

344

Simrick, S., Masse, K., Jones, E.A., 2005. Developmental expression of Pod 1 in Xenopus laevis. Int. J. Dev. Biol. 49, 59–63. Smart, N., Risebro, C.A., Melville, A.A., Moses, K., Schwartz, R.J., Chien, K.R., Riley, P.R., 2007. Thymosin beta4 induces adult epicardial progenitor mobilization and neovascularization. Nature 445, 177–182. Smith, T.K., Bader, D.M., 2007. Signals from both sides: control of cardiac development by the endocardium and epicardium. Semin. Cell Dev. Biol. 18, 84–89. Strizzi, L., Bianco, C., Normanno, N., Seno, M., Wechselberger, C., Wallace-Jones, B., Khan, N.I., Hirota, M., Sun, Y., Sanicola, M., Salomon, D.S., 2004. Epithelial mesenchymal transition is a characteristic of hyperplasias and tumors in mammary gland from MMTVCripto-1 transgenic mice. J. Cell Physiol. 201, 266–276. Sucov, H.M., Dyson, E., Gumeringer, C.L., Price, J., Chien, K.R., Evans, R.M., 1994. RXR alpha mutant mice establish a genetic basis for vitamin A signaling in heart morphogenesis. Genes Dev. 8, 1007–1018. Svensson, E.C., Tufts, R.L., Polk, C.E., Leiden, J.M., 1999. Molecular cloning of FOG-2: a modulator of transcription factor GATA-4 in cardiomyocytes. Proc. Natl. Acad. Sci. U.S.A. 96, 956–961. Tao, Q., Yokota, C., Puck, H., Kofron, M., Birsoy, B., Yan, D., Asashima, M., Wylie, C.C., Lin, X., Heasman, J., 2005. Maternal wnt11 activates the canonical wnt signaling pathway required for axis formation in Xenopus embryos. Cell 120, 857–871. Tevosian, S., Deconinck, A., Tanaka, M., Schinke, M., Litovsky, S., Izumo, S., Fujiwara, Y., Orkin, S., 2000. FOG-2, a cofactor for GATA transcription factors, is essential for heart morphogenesis and development of coronary vessels from epicardium. Cell 101, 729–739. Tomanek, R.J., 2005. Formation of the coronary vasculature during development. Angiogenesis 8, 273–284. Tomanek, R.J., Ishii, Y., Holifield, J.S., Sjogren, C.L., Hansen, H.K., Mikawa, T., 2006. VEGF family members regulate myocardial tubulo­genesis and coronary artery formation in the embryo. Circ. Res. 98, 947–953. Torlopp, A., Breher, S.S., Schluter, J., Brand, T., 2006. Comparative analysis of mRNA and protein expression of Popdc1 (Bves) during early development in the chick embryo. Dev. Dyn. 235, 691–700. van Tuyn, J., Atsma, D.E., Winter, E.M., van der Velde-van Dijke, I., Pijnappels, D.A., Bax, N.A., Knaan-Shanzer, S., Gittenberger-de Groot, A.C., Poelmann, R.E., van der Laarse, A., van der Wall, E.E., Schalij, M.J., de Vries, A.A.F., 2007. Epicardial cells of human adults can undergo an epithelial-to-mesenchymal transition and obtain characteristics of smooth muscle cells in vitro. Stem Cells 25, 271–278. Vasavada, T.K., DiAngelo, J.R., Duncan, M.K., 2004. Developmental expression of Pop1/Bves. J. Histochem. Cytochem. 52, 371–377. Virágh, S., Challice, C.E., 1981. The origin of the epicardium and the embryonic myocardial circulation in the mouse. Anat. Rec. 201, 157–168. Virágh, S., Gittenberger-de Groot, A.C., Poelmann, R.E., Kalman, F., 1993. Early development of quail heart epicardium and associated vascular and glandular structures. Anat. Embryol. (Berl.) 188, 381–393. Vrancken Peeters, M.P., Gittenberger-de Groot, A.C., Mentink, M.M., Poelmann, R.E., 1999. Smooth muscle cells and fibroblasts of the coronary arteries derive from epithelial–mesenchymal transformation of the epicardium. Anat. Embryol. (Berl.) 199, 367–378. Wada, A.M., Smith, T.K., Osler, M.E., Reese, D.E., Bader, D.M., 2003. Epicardial/mesothelial cell line retains vasculogenic potential of embryonic epicardium. Circ. Res. 92, 525–531.

PART  |  5  Epicardial Development

Wagner, K.D., Wagner, N., Bondke, A., Nafz, B., Flemming, B., Theres, H., Scholz, H., 2002. The Wilms’ tumor suppressor Wt1 is expressed in the coronary vasculature after myocardial infarction. FASEB J. 16, 1117–1119. Wagner, N., Wagner, K.D., Theres, H., Englert, C., Schedl, A., Scholz, H., 2005. Coronary vessel development requires activation of the TrkB neurotrophin receptor by the Wilms’ tumor transcription factor Wt1. Genes Dev. 19, 2631–2642. Wagner, N., Wagner, K.D., Scholz, H., Kirschner, K.M., Schedl, A., 2006. Intermediate filament protein nestin is expressed in developing kidney and heart and might be regulated by the Wilms’ tumor suppressor Wt1. Am. J. Physiol. 291, R779–R787. Waldo, K.L., Willner, W., Kirby, M.L., 1990. Origin of the proximal coronary artery stems and a review of ventricular vascularization in the chick embryo. Am. J. Anat. 188, 109–120. Ward, N.L., Van Slyke, P., Sturk, C., Cruz, M., Dumont, D.J., 2004. Angiopoietin 1 expression levels in the myocardium direct coronary vessel development. Dev. Dyn. 229, 500–509. Watt, A.J., Battle, M.A., Li, J., Duncan, S.A., 2004. GATA4 is essential for formation of the proepicardium and regulates cardiogenesis. Proc. Natl. Acad. Sci. U.S.A. 101, 12573–12578. Weninger, W.J., Floro, K.L., Bennett, M.B., Withington, S.L., Preis, J.I., Barbera, J.P., Mohun, T.J., Dunwoodie, S.L., 2005. Cited2 is required both for heart morphogenesis and establishment of the left–right axis in mouse development. Development 132, 1337–1348. Wessels, A., Perez-Pomares, J.M., 2004. The epicardium and epicardially derived cells (EPDCs) as cardiac stem cells. Anat. Rec. 276A, 43–57. Wilm, B., Ipenberg, A., Hastie, N.D., Burch, J.B., Bader, D.M., 2005. The serosal mesothelium is a major source of smooth muscle cells of the gut vasculature. Development 132, 5317–5328. Wilson, J.G., Warkany, J., 1949. Aortic-arch and cardiac anomalies in the offspring of vitamin A deficient rats. Am. J. Anat. 85, 113–155. Wilting, J., Buttler, K., Schulte, I., Papoutsi, M., Schweigerer, L., Männer, J., 2007. The proepicardium delivers hemangioblasts but not lymphangioblasts to the developing heart. Dev. Biol. 305, 451–459. Wu, S.M., Fujiwara, Y., Cibulsky, S.M., Clapham, D.E., Lien, C.L., Schultheiss, T.M., Orkin, S.H., 2006. Developmental origin of a bipotential myocardial and smooth muscle cell precursor in the mammalian heart. Cell 127, 1137–1150. Xavier-Neto, J., Shapiro, M.D., Houghton, L., Rosenthal, N., 2000. Sequential programs of retinoic acid synthesis in the myocardial and epicardial layers of the developing avian heart. Dev. Biol. 219, 129–141. Yang, J.T., Rayburn, H., Hynes, R.O., 1995. Cell adhesion events mediated by alpha 4 integrins are essential in placental and cardiac development. Development 121, 549–560. Yi, P., Han, Z., Li, X., Olson, E.N., 2006. The mevalonate pathway controls heart formation in Drosophila by isoprenylation of Ggamma1. Science 313, 1301–1303. Zhang, W., Yatskievych, T.A., Baker, R.K., Antin, P.B., 2004. Regulation of Hex gene expression and initial stages of avian hepatogenesis by Bmp and Fgf signaling. Dev. Biol. 268, 312–326. Zhou, B., Ma, Q., Rajagopal, S., Wu, S.M., Domian, I., Rivera-Feliciano, J., Jiang, D., von Gise, A., Ikeda, S., Chien, K.R., Pu, W.T., 2008. Epicardial progenitors contribute to the cardiomyocyte lineage in the developing heart. Nature 454, 109–113.

Chapter 5.2

The Epicardial Signaling Center in Development and Disease Kory J. Lavine and David M. Ornitz Department of Developmental Biology, Washington University School of Medicine, St. Louis, MO, USA

I.  Introduction I.A.  Stages of Cardiac Development As introduced and discussed in previous chapters, cardiac development can be subdivided into early, midgestation and late or growth and maturation stages. Cardiac development begins at approximately embryonic day 8.0 (E8.0) in the mouse and includes such events as cardiac progenitor cell specification (primary and secondary heart fields), morphogenesis of the linear heart tube and cardiac looping. These processes are discussed in Volume I, Parts 1–3 and have been reviewed in great detail elsewhere (Mohun and Sparrow, 1997; Cripps and Olson, 2002). Midgestational heart development begins shortly after looping of the linear heart tube, commencing at approximately E10.5 and ending at approximately E16.5 in the mouse. Late gestational and neonatal stages ensue at E16.5 and at birth, respectively and, for the most part, involve growth and maturation of the myocardium, as well as other cardiac structures formed during previous stages.

I.B.  Midgestational Heart Development Important processes that occur during midgestational heart development include formation of the epicardium, myocardial growth, coronary vascular development, endocardial cushion formation, ventricular septation and trabeculation (Gittenberger-de Groot et al., 2005). Beginning at E10.5 in the mouse, the developing heart becomes engulfed by a population of mesothelial cells that will give rise to the third and most outer cardiac layer, the epicardium. As will be described in detail below, proper formation and Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

function of the epicardium is essential for several of the proceeding developmental events that occur during mid­ gestation, including myocardial growth and coronary vascu­ lar development. Other processes that occur in the midgestational heart include development of the atrial–ventricular (AV) and outflow tract (OFT) endocardial cushions. These structures will give rise to the atrioventricular valves (tricuspid and mitral) and outflow tract valves (pulmonary and aortic), respectively. Proper endocardial cushion development is essential for septation of both the atrial and ventricular chambers, as defects in this important process can lead to atrial and ventricular septal defects (ASD and VSD) both in animal models and in several human genetic syndromes. The development of the endocardial cushions and resultant defects from abrogation of this process are discussed in Chapters 6.1 and 6.2 and reviewed elsewhere (Kelly, 2005; Smith and Bader, 2007). Ventricular septation is further aided by the growth of the muscular interventricular septum, a process that occurs between E14.5 and E16.5 in the mouse. Lastly, trabeculation of the ventricular chambers occurs throughout midgestational heart development. This process patterns the myocardial wall into an outer compact layer and an inner trabecular layer. Little is known about the signals that regulate the formation of these finger-like extensions of myocardial tissue (trabeculae) that project into both the right and left ventricular chambers. However, an entity termed isolated ventricular noncompaction, which in many individuals leads to fulminant heart failure, results from the overproduction of these finger-like projections at the expense of compact myocardial tissue (Junga et al., 1999; Jenni et al., 2007). Of these critical processes occurring during midgestation, myocardial growth and coronary vascular development 345

346

PART  |  5  Epicardial Development

are the best characterized both on the morphological and molecular level. Between E11.5 and E13.5, the developing heart undergoes a dramatic increase in size (Fig. 1). This growth is attributable to cardiomyoblast proliferation, as mutants that display defects in cardiac growth demonstrate decreased proliferation (Sucov et al., 1994; Merki et al., 2005). Formation of the coronary vascular system

(A)

(B)

E11.5

(C)

E12.5

(D)

E13.5

(E)

Endocardium

Epicardium

coincides with this rapid increase in cardiac size. Coronary development can be divided into two stages, vascular plexus formation and vascular remodeling. The coronary plexus is formed between E11.5 and E13.5 and is remodeled between E14.5 and E16.5 to give rise to the mature coronary vascular system (Fig. 2). In the following sections, we will discuss in detail the mechanisms by which

Epicardium

Myocardium

RA

Proliferation

FGF9/16/20 ?RTK ligand

FGF9/1/2c ?RTK

Myocardium

Figure 1  Growth of the myocardium is regulated by the epicardium. (A–C) Between murine E11.5 and E13.5, the developing heart undergoes a robust increase in size. (D) Cardiomyoblast proliferation is responsible for this dramatic period of growth and is governed by signals produced by the epicardium. (E) Retinoic acid (RA) signaling within the epicardium results in the production of epicardially-derived mitogens such as fibroblast growth factor 9 (FGF9), which directly promotes cardiomyoblast proliferation via activation of FGFR1 and FGFR2c. In addition to FGF9 other FGFs, such as FGF16 and FGF20, and other receptor tyrosine kinase (RTK) signaling pathways may also participate in the control of myocardial growth. Adapted from Lavine and Ornitz (2008), with permission from Elsevier.

Figure 2  Development of the coronary vasculature. Platelet/endothelial cell adhesion molecule-1 (PECAM)/CD31 immunostaining of murine E11.5 to E13.5 and E16.5 hearts demonstrates the progression of coronary vascular growth during development. Between E11.5 and E13.5, the developing coronary vascular plexus emerges from the atrial–ventricular groove and grows in a wave-like pattern to cover both ventricles by E13.5. By E16.5, the coronary vascular plexus has remodeled, giving rise to the mature coronary vasculature. Adapted from Lavine and Ornitz (2007), with permission from Elsevier.

Chapter  |  5.2  The Epicardial Signaling Center in Development and Disease

347

midgestational myocardial growth and coronary vascular development occur, and the signaling pathways that coordinately control these two critical processes.

(VCAM1, 41-integrin) (Kwee et al., 1995; Yang et al., 1995; Davies et al., 1999; Moore et al., 1999; Shikama et al., 2003; Watt et al., 2004; Shao et al., 2005).

II.  Epicardium and myocardial proliferation

II.B.  Derivatives of the Epicardium

Following cardiac looping, the cardiomyoblast population undergoes a robust expansion resulting in marked growth of the embryonic heart. During these earlier stages of cardiac development (specification and looping stages), cardiomyoblast proliferation is dependent on endogenous cardiac-specific transcriptional regulators, such as NKX2-5, GATA4, MEFF2C, HOP, BOP, TBX5 and several other transcription factors (Cripps and Olson, 2002), and potentially exogenous signals derived from the secondary heart field (Meyers et al., 1998; Abu-Issa et al., 2002). Further details describing the role of these cardiac specific transcription regulators can be found in Volume II, Part 9. In contrast to these earlier stages of cardiac development, myocardial growth during midgestation is regulated via exogenous factors derived from the epicardium and endocardium. The epicardium is a single cell layer of mesothelial and possibly neural crest origin that overlays the myocardium. The epicardium gives rise to components of the coronary vasculature and interstitial cells of the myocardium, and also serves as a source of signals that control myocardial development (Mikawa and Gourdie, 1996; Dettman et al., 1998; Stottmann et al., 2004). Interestingly, it has recently been reported that epicardial progenitor cells may also contribute cells to the cardiomyocyte lin­ eage (Cai et al., 2008; Zhou et al., 2008). The endocardium is comprised of a single layer of cells situated on the inner side of the myocardium and is thought to produce several factors controlling the growth of the myocardium and cardiac trabeculae (Smith and Bader, 2007).

II.A.  Development of Epicardium Prior to stage 14 in the chick (E9.5 in the mouse), the heart consists of two layers, an outer myocardial layer and an inner endocardial layer. At stage 18 in the chick (E10.5 in the mouse), the third cardiac layer, the epicardium, migrates to and envelops the embryonic heart. Lineage tracing studies in avian species demonstrate that epicardial progenitors are derived from the proepicardial organ (an epithelium associated with the septum transversum), which travels as either individual cells or as a sheet to engulf the developing heart (Nesbitt et al., 2006). Several gene products have been implicated in this process, including factors required for proepicardial organ formation (GATA4, -catenin), epicardial progenitor cell survival (WT1), as well as epicardial progenitor cell migration and attachment

In addition to forming the outermost cell layer of the developing and adult heart, it is clear that the epicardium physically contributes perivascular and interstitial cell types to the heart through a process of epithelial mesenchymal transformation (EMT). Epicardial EMT initiates at approximately E11.5 and progresses in a spatio–temporal manner beginning at the base of the heart and proceeding in a wave-like pattern towards the cardiac apex. Epicardialderived cells subsequently either reside within the subepicardial space forming the subepicardial mesenchyme (a layer of mesenchymal tissue located between the epicardium and myocardium) or migrate into the myocardium in a perivascular distribution (Fig. 3). Lineage analysis of the proepicardial organ in chicken and quail embryos demonstrate that cells originating within the proepicardial organ give rise to perivascular fibroblasts, vascular smooth muscle and endothelial cells of the coronary blood vessels (Mikawa and Fischman, 1992; Mikawa and Gourdie, 1996; Perez-Pomares et al., 2002). However, both cell labeling and genetic lineage analysis of the epicardium reveal contributions to perivascular fibroblast and smooth muscle cell lineages, but not to the coronary endothelial cell lineage (Dettman et al., 1998; Vrancken Peeters et al., 1999; Merki et al., 2005). Together these lineage studies suggest that the proepicardial organ contains a mixed pool of vascular smooth muscle/fibroblast progenitors and endothelial progenitors, both of which migrate to the heart. On reaching the heart, these two progenitor populations segregate such that vascular smooth muscle cell/fibroblast progenitors are resident within the epicardium, while endothelial progenitors occupy another space. This notion is supported by work showing that these two lineages can be independently identified by immunohistochemical analysis in the migrating proepicardial organ (Dettman et al., 1998).

II.C.  Function of the Epicardium In addition to physically contributing cells to the developing heart, the epicardium serves as a critical structure for several events that occur during midgestation. In avian species, removal of the epicardium leads to an arrest in both cardiomyoblast proliferation and coronary development (Morabito et al., 2002; Reese et al., 2002; Wada et al., 2003). Moreover, deletion of genes necessary for epicardial development in the mouse (Gata4, -catenin, Wt1, V-cam1 and 41-integrin) similarly lead to a profound defect in both cardiomyoblast proliferation and coronary

348

PART  |  5  Epicardial Development

(A) Epicardial formation (E11.5)

(B) Epicardial EMT (E12.5)

(C) Epicardial formation (E13.5)

Endo Myo SEM

Epi

(D)

(E)

(F)

(G) Coronary artery

Epicardium

Coronary vein

Endocardium

Myocardium

Perivascular cell Mesenchymal cell

Figure 3  Morphologic events during coronary vascular development. (A) By murine E11.5, the embryonic heart is encased by the outermost cell layer, the epicardium. (B–C) Between E12.5 and E13.5, a subset of epicardial cells undergo an epithelial–mesenchymal transformation (EMT). These mesenchymal cells will either take residence in the subepicardial space along with subepicardial blood vessels and form the subepicardial mesenchyme, or migrate into the myocardium and encase intramyocardial blood vessels as pericytes. (D–E) Histology of E12.5 and E13.5 hearts showing the formation of the subepicardial mesenchyme (marked by green arrowheads). (F) PECAM/CD31 immunostaining at E13.5 shows the relative locations of the subepicardial (black arrowhead) and intramyocardial blood vessels (black arrow). (G) A three-dimensional reconstruction of the coronary vascular plexus shows that the subepicardial (white arrowhead) and intramyocardial (white arrow) blood vessels form connections via small projecting vessels (red arrow) reminiscent of mature arterial–venous vascular networks. Adapted from Lavine and Ornitz (2008), with permission from Elsevier.

vascular development (Kwee et al., 1995; Yang et al., 1995; Davies et al., 1999; Moore et al., 1999; Shikama et al., 2003; Watt et al., 2004; Shao et al., 2005). These data suggest that the epicardium supplies either a critical factor or cell type that is absolutely essential for these important processes. Several key experiments (described below) have implicated the epicardium in promoting myocardial proliferation and coronary vascular development via an epicardially-based paracrine signaling mechanism.

II.D.  Epicardial Control of Myocardial Growth Epicardial-derived signals are proposed to be important regulators of cardiomyoblast proliferation (Fig. 1).

Retinoic acid (RA) and erythropoietin (Epo) signaling within the epicardium are necessary for cardiomyoblast proliferation during midgestational heart development. Mice deficient in epicardial retinoic acid or erythropoietin signaling display profound myocardial hypoplasia. Furthermore, treatment of epicardial cell cultures with either retinoic acid or erythropoietin leads to the expression of mitogens that stimulate cardiomyoblast proliferation (Chen et al., 1998a; Wu et al., 1999; Chen et al., 2002; Stuckmann et al., 2003). Consistently, it has been hypothesized that epicardial retinoic acid and erythropoietin signaling regulate epicardial-derived mitogens that act on cardiomyoblasts which, until recently, had yet to be identified. Moreover, retinoic acid and erythropoietin are thought to induce distinct epicardial-derived factors (Stuckmann et al., 2003).

Chapter  |  5.2  The Epicardial Signaling Center in Development and Disease

II.E.  Fibroblast Growth Factor 9 (FGF9) as a Retinoic Acid Inducible Factor Retinoic acid and erythropoietin signaling are known to control the expression of distinct epicardial-derived mitogens that act on cardiomyoblasts. FGF9 is expressed in the epicardium and is necessary for the growth of the myocardium during midgestational cardiac development. Consistent with this model, Fgf9 expression is upregulated by retinoic acid (Lavine et al., 2005). Factors regulated by Epo have yet to be identified. In addition to FGF9, FGF16 and FGF20 are also expressed in the epicardium, and together may constitute an epicardial-derived signal that regulates myocardial proliferation. Furthermore, FGF9, FGF16 and FGF20 signal to the myocardium specifically through FGF receptor (FGFR) 1c and FGFR2c, directly promoting cardiomyoblast proliferation. Mice deficient in FGF9 (Fgf9/) and mice concurrently lacking myocardial FGFR1 and FGFR2 expression, through conditional inactivation of both genes with the Myosin light chain 2v-Cre driver (Fgfr1/2Mlc2v), display decreased cardiomyoblast proliferation and subsequent ventricular hypoplasia (Chen et al., 1998b; Lavine et al., 2005). Similar to Fgf9/ and Fgfr1/2Mlc2v hearts, RXR and Raldh2 knockout hearts, which fail to make retinoic acid in the epicardium, display profound myocardial hypoplasia, albeit significantly more severe than that seen for Fgf signaling knockouts (Sucov et al., 1994; Mic et al., 2002; Lavine et al., 2005). The greater extent of hypoplasia seen in RXR and Raldh2 knockouts compared to Fgfr1/2Mlc2v hearts is likely a result of the timing of loss of FGFR1 and FGFR2 protein in the myocardium, compared to the loss of the epicardial signal. That is, retinoic acid induces FGFs in the epicardium prior to deletion of FGFR1 and FGFR2 by the myocardial specific Cre driver, Mlc2v-cre, which is first activated at E10.5. Alternatively, it is possible that in addition to FGFs retinoic acid induces other epicardialderived mitogens (Fig. 1). Potential candidates include other receptor tyrosine kinase (RTK) ligands, which may function redundantly with FGF signaling in the developing myocardium. Consistent with this, ErbB2 (Egfr2) and Pdgfr knockout embryos have hypoplastic hearts and reduced myocardial proliferation (Price et al., 2001; Chen et al., 2002). The timing of action and stages of cardiac development that are affected by loss of ErbB2 and PDGFRa signaling have not yet been defined.

II.F.  FGF Signaling Regulates the Progression of Myocardial Differentiation FGF signals are known to promote the acquisition of the cardiac fate (cardiogenesis). Since cardiogenesis occurs well before the activation of Mlc2v-cre expression (E10.5), the unaltered expression of key cardiac markers (Nkx2-5,

349

Nppa and Myh7) in Fgfr1/2Mlc2v hearts is not surprising. However, Fgf9/ and Fgfr1/2Mlc2v hearts also displayed both increased cell size and increased amounts of sarcomeric actin (Lavine et al., 2005). Given the observation that cardiomyoblasts accumulate increasing amounts of cardiac (sarcomeric) actin as they differentiate (Carrier et al., 1992; Sugi and Lough, 1992), this finding suggests that FGF signaling also functions to delay cardiomyoblast differentiation. Given these data, we have postulated that FGF signaling in the developing myocardium functions to spatially and temporally coordinate cardiomyoblast proliferation with differentiation. Consistent with this notion, deletion of RXRa from the epicardium similarly leads to accelerated cardiomyoblast differentiation (Chen et al., 1998a).

III.  Coronary vascular development Vascular development is governed by two sequentially acting processes, vasculogenesis and angiogenesis. Vasculogenesis refers to the de novo formation of blood vessels via differentiation of either angioblast or hemangioblast precursors, whereas angiogenesis is defined as the growth or remodeling of established blood vessels. In general, vascular systems undergo a stereotyped pattern of development, beginning with the formation of a primary capillary plexus that is later remodeled, giving rise to the mature vasculature. It is thought that the primary capillary plexus forms by vasculogenesis and is remodeled via angiogenesis (Flamme et al., 1997; Risau, 1997). Similar to other vascular systems, coronary development begins with the formation of a vascular network that is later remodeled, giving rise to the mature coronary tree (Morabito et al., 2002; Kattan et al., 2004) (Fig. 2). Interestingly, the initial coronary vascular plexus consists of two sets of blood vessels located within different positions, the subepicardial mesenchyme and the myocardial wall. Intriguingly, these two sets of blood vessels are interconnected via small branching vascular structures, reminiscent of arterial–venous vascular beds (Fig. 3). Further discussion regarding the identities of subepicardial and intramyocardial blood vessels can be found below. Many of the morphological events involved in the formation of the coronary vascular system have been welldescribed, especially in avian systems (Morabito et al., 2002; Reese et al., 2002; Wada et al., 2003). Coronary development ensues following the investment of the heart by the epicardium. A subset of epicardial cells undergoes an epithelial–mesenchymal transformation (EMT) beginning at stage 26 in the chick (embryonic day 11.5 in the mouse). Epicardial epithelial–mesenchymal transformation leads to the formation of a mesenchyme situated between the epicardium and myocardium (subepicardial space). Within the subepicardial mesenchyme, endothelial

350

PART  |  5  Epicardial Development

cells coalesce to form vascular channels that are later ensheathed by smooth muscle cells and perivascular fibroblasts. Vascular channel formation leads to the development of a vascular plexus that covers the heart. At later stages of development, this vascular plexus undergoes remodeling, giving rise to the mature coronary vascular system. While much is now known regarding the early events in coronary development, little is known about this remodeling process.

III.A.  Epicardium and Coronary Development Initial studies in chicks and quails demonstrated that the proepicardial organ and the epicardium are required for coronary development, as removal of either structure severely perturbs coronary vessel formation (Gittenbergerde Groot et al., 2000). More recently, genetic analysis in the mouse has confirmed this observation (described below). It has been previously proposed that the epicardium enables coronary vascular development by physically contributing vascular cell types (specifically endothelial cells) to the heart through an epithelial–mesenchymal transformation (Mikawa and Fischman, 1992; Mikawa and Gourdie, 1996; Perez-Pomares et al., 2002). However, as mentioned above, lineage analysis, both in avian and mouse models, does not support an epicardial contribution to the coronary endothelium. Instead, the epicardium contributes smooth muscle and fibroblast cell fates to the developing coronary vasculature, while endothelial cells are derived from another source (Dettman et al., 1998; Vrancken Peeters et al., 1999). An alternative explanation as to why the epicardium is essential for coronary development is that the epicardium acts as a signaling center. Support for this hypothesis

E11.5

E12.5

stems from the initial observation that the epicardium promotes cardiomyoblast proliferation by secreting mitogens (Chen et al., 2002; Stuckmann et al., 2003). Interestingly, not only is retinoic acid signaling in the epicardium necessary for the secretion of such mitogens, but it is also required for coronary vascular development (Sucov et al., 1994; Merki et al., 2005).

III.B.  An FGF-HH-VEGF/ANG Signaling Pathway Controls Coronary Development Conditional gene targeting and organ culture assays have been used to define a signaling network that operates between the epicardium and the myocardium which is essential for coronary vascular development. These studies demonstrated that an FGF and HH signaling pathway controls the formation of the coronary vascular system by regulating the expression of vascular endothelial growth factors (Vegf) and angiopoietins (Ang) (Fig. 4) (Lavine et al., 2006). Epicardial and endocardial sources of FGF ligands control coronary development by signaling to the cardiomyoblast through redundant function of FGFR1 and FGFR2c. Myocardial FGF signaling regulates coronary vascular development by triggering a wave of Hedgehog (HH) activation that progresses from the atrial–ventricular groove (at E12.5) to the apex of the ventricles (at E13.5), tracking the progression of the developing coronary vascular plexus. Epicardial-derived HH ligands signal to the cardiomyoblast and perivascular mesenchymal cell, and induce the expression of Vegf-A, Vegf-B, Vegf-C and Ang2, resulting in the formation of the coronary vascular plexus. Importantly, organ culture experiments revealed that both a VEGF and ANG2 signal are required to support coronary vessel growth. Intriguingly, this FGF-HH-VEGF/ANG signaling cascade coordinately controls the growth of both

E13.5

Vascular plexus

SHH

SHH

SHH

PTC1, VEGF-A,-B,-C and ANG2

Figure 4  Hedgehog signaling controls the development of the coronary vasculature. Schematic highlighting the simultaneous wave-like expansion of the coronary vascular plexus and HH activation between E11.5 and E13.5. The blue-colored area denotes the expression domain of Ptc1, Vegf-A, Vegf-B, Vegf-C and Ang2. Adapted from Lavine and Ornitz (2007), with permission from Elsevier.

351

Chapter  |  5.2  The Epicardial Signaling Center in Development and Disease

subepicardial and intramyocardial blood vessels (Lavine et al., 2006). This is discussed in further detail below.

III.C.  The Epicardium Acts as a Signaling Center for Heart Development Removal of the epicardium leads to defects in both cardiomyoblast proliferation and coronary vascular development (Morabito et al., 2002; Reese et al., 2002; Wada et al., 2003). We have shown that epicardial-derived FGF signals are essential mediators of myocardial growth (Lavine et al., 2005). In addition to this mitogenic activity, epicardialand endocardial-derived FGF signals also control coronary vascular development. Examination of the Fgf9/ and Fgfr1/2Mlc2v mouse models revealed that FGF9 (and likely FGF16 and FGF20) is essential for coronary vascular formation and growth. These findings expand on the more general notion that FGF signaling is a strong promoter of blood vessel growth (Seghezzi et al., 1998; Auguste et al., 2003; Kanda et al., 2004). Thus, the epicardium acts as a source of signaling molecules critical for multiple aspects of heart development (Fig. 5). The overwhelming majority of mutations that result in coronary vascular defects are due to failures in epicardial formation or loss of epicardial integrity, highlighting the critical role of the epicardium in this process. Removal

of genes required for formation (Gata4) and migration or attachment (Vcam1, 41-integrin) of epicardial precursors to the myocardium leads to severe defects in coronary development (Kwee et al., 1995; Yang et al., 1995; Watt et al., 2004), and deletion of genes necessary for survival and/or integrity (p300, Wt1) of the epicardium lead to similar phenotypes (Davies et al., 1999; Moore et al., 1999; Shikama et al., 2003; Shao et al., 2005). One exception to this trend was unveiled by genetic analysis of Fog2 (Friend of GATA). Mutations in Fog2 (deletion or inability to interact with GATA4) led to severe defects in myocardial proliferation and coronary development, despite a normal appearing epicardium (Tevosian et al., 2000; Crispino et al., 2001). Fog2 mutant hearts failed to undergo epicardial epithelial–mesenchymal transformation. Furthermore, these phenotypes could be rescued by expression of Fog2 in the myocardium, demonstrating that signals from the myocardium regulate epicardial epithelial–mesenchymal transformation. Interestingly, the myocardial hypoplasia and coronary defects seen in Fog2/ hearts are reminiscent of the phenotypes seen in hearts lacking myocardial FGF signaling. Since FOG2 activity in the myocardium regulates these same processes, it is intriguing to postulate that FOG2 and FGF signaling may be functionally related. For example, FOG2 may be necessary for the cardiomyoblast to receive FGF signaling. Alternatively, it is possible that FOG2 and

Proliferation Epicardial signaling center

FGF9 FGF16 FGF20

Myocardial FGFR1 and 2 Differentiation

SHH

X? Perivascular cell VEGFC Myocardial VEGFA VEGFB ANG2

Endothelial cell VEGFR1 VEGFR2 TIE1, TIE2

Figure 5  Signaling pathways operating between the epicardium, myocardium and vasculature that are active in regulating myocardial growth and differentiation and coronary vascular development. The epicardium serves as a signaling center in which retinoic acid (RA), and potentially other molecules, regulate the expression of FGF9 (and potentially FGF16 and 20). FGFs released from the epicardium signal through myocardial-expressed FGFRs 1 and 2 to regulate midgestation myocardial growth and suppress differentiation. Unknown myocardial factors (X?) are thought to signal back to the epicardium to regulate the expression of SHH. Myocardial HH signaling results in the production of vasculogenic factors, including VEGFA, VEGFB and ANG2. SHH also acts on the perivascular cell where it regulates the expression of VEGFC and possibly other angiogenic factors. Both VEGFs and ANG2 are required for normal coronary vascular development. Adapted from Lavine and Ornitz (2008), with permission from Elsevier.

352

GATA4 may function either in conjunction with or downstream of FGF signaling to regulate the expression of key myocardial-derived signals that control epicardial function and signaling. Consistent with this notion, several interactions between FGF and GATA signaling have been identified (Xu et al., 1999; Fossett and Schulz, 2001; Bertrand et al., 2003; Iwahori et al., 2004; Murakami et al., 2004).

III.D.  A Ventricular Wave of HH Activity Mediates Coronary Development Previous work in avian systems has shown that coronary vessel development proceeds in a wave-like pattern, originating from the atrial–ventricular groove and extending toward the ventricular apices (Morabito et al., 2002; Reese et al., 2002; Wada et al., 2003). Consistent with this, HH signaling is activated in a similar pattern and is required for the wave-like growth of the coronary vascular plexus. Strikingly, epicardial HH ligand expression does not appear to travel in a wave, and thus the mechanism by which HH activation progresses toward the cardiac apex is not immediately obvious. One potential mechanism by which a gradient of sig­ naling can be formed in the face of uniform ligand expression is via altering the competency of cells to receive such a signal (a competency gradient). In other words, the target of signaling is not uniformly competent to receive or respond to the signal. In this model, the heart would gain competence to receive HH signaling in a wave-like progression. Consistent with this notion, experiments describing the effects of expressing a dominant active form of GLI2 in the myocardium revealed evidence consistent with the presence of a competency gradient. Specifically, despite uniform expression of the transgene throughout the ventricle, both Ptc1 (reporter of HH signaling) and Vegf-A expression were only increased within their normal areas of expression (Lavine et al., 2006). Thus, even though constitutively-active GLI2 was expressed throughout the ventricle, it only activated HH signaling within its normal wave-like distribution.

III.E.  Mechanism by which HH Controls Vegf and Angiopoietin2 Expression Genetic experiments utilizing a constitutively active GLI2 transgene demonstrated that HH signaling induces Vegf ligands and Ang2 in a cell-autonomous manner (Lavine et al., 2006). Based on this, it is reasonable to posit that Vegf-A, Vegf-B, Vegf-C and Ang2 may be direct targets of HH signaling. However, no GLI response elements are readily identifiable through computational analysis of the enhancer and promoter regions of these molecules, suggesting that they may be indirect targets of HH signaling (Zhao et al., 2004). Despite the lack of potential GLI response elements, these HH responsive genes share a number of putative

PART  |  5  Epicardial Development

transcription factor binding sites. Common sites included those for HIF-1/AHR (Hypoxia inducible factor-1), Gata, serum response factor (SRF), SMAD2/3 and ETS domain transcription factors. None of these putative transcription factor binding sites were present in the Vegf-D promoter, a Vegf gene not controlled by HH signaling during embryonic heart development (Zhao et al., 2004). Potential interactions between HH signaling and the HIF-1a pathway are of particular interest. HIF-1a is a critical regulator of the hypoxic response, functioning to upregulate a number of genes including Vegf-A, and to promote new vascular growth (Covello and Simon, 2004). In addition to its role in hypoxia, HIF-1a is also important for carcinogenesis. HIF-1a is activated in van Hippel Lindau disease, causing formation of angiosarcomas, as well as in a number of other cancers (Vaupel, 2004). Interestingly, HIF-1a is also essential for vascular growth in the embryo, regulating both neural and bone development (Schipani et al., 2001; Tomita et al., 2003), demonstrating that like the HH pathway, HIF-1a is essential for several aspects of embryonic vascular development. Although there are no reports of a direct interaction between HH and HIF-1a pathways, it is of interest to note that SHH is upregulated in response to skeletal muscle and cardiac ischemia (Pola et al., 2003; Kusano et al., 2005). These observations raise the question of whether HIF-1a and HH signaling somehow function together in the response to tissue hypoxia. Moreover, the presence of HIF-1/AHR response elements in the regulatory regions of Vegf-A, VegfB, Vegf-C and Ang2 suggests that HIF-1a may potentially be involved in coordinating the expression of angiogenic factors in response to HH signaling in both the adult and embryo. Moreover, it is also possible that HIF-1a may control HH ligand expression. These are important issues worthy of further investigation, especially given the recent development of pharmaceutical antagonists against HIF-1a (Tan et al., 2005). In addition to HIF-1a, binding sites for Gata and SRF transcription factors were also present in the regulatory regions of Vegf-A, Vegf-B, Vegf-C and Ang2. Conditional deletion of SRF in the developing myocardium leads to profound defects in heart development, including deficits in coronary vascular growth (Miano et al., 2004). Moreover, loss of FOG-2 also results in impairments in coronary development (Tevosian et al., 2000; Crispino et al., 2001), highlighting the importance of SRF and Gata activity in the formation of the coronary vasculature system. Thus, like HIF-1a, SRF and Gata may likely be involved in HH-mediated coronary vascular growth. Potential interactions of Gata and SRF with HH signaling should be explored in the future.

III.F.  FGF Regulation of HH Signaling FGF signaling to the myocardium promoted expression of SHH in the epicardium, suggesting that FGF signals

Chapter  |  5.2  The Epicardial Signaling Center in Development and Disease

control SHH expression via an indirect mechanism. Specifically, FGF signaling must regulate the expression of a myocardial-derived factor that likely signals back to the epicardium in a reciprocal manner. Despite the large number of molecules that could fulfill this role, several published observations point toward a few potential candidates, the first of which involves a similar FGF–HH cascade that functions to control early limb patterning and outgrowth. Genetic analysis of the limb deformity (ld) mutation has provided some insight into one mechanism by which FGF signaling could control Shh expression. Mice harboring the limb deformity (ld) mutation are unable to maintain Shh expression in the limb bud apical epidermal ridge (Chan et al., 1995; Haramis et al., 1995). The gene responsible for this function has been identified as the bone morphogenic protein (BMP) antagonist, gremlin (Zuniga et al., 2004). From this work and others, it has been suggested that FGF signaling may promote SHH expression by antagonizing BMP activity (Zuniga et al., 1999; Khokha et al., 2003). Another paradigm that has been established is reciprocal FGF signaling. In the limb bud and lung (as well as many other tissues), reciprocal FGF signaling occurs between epithelial, mesothelial and mesenchymal tissues. Generally, epithelial or mesothelial cells secrete FGF ligands (such as FGF8 and FGF9) that bind c-type FGFR splice forms (such as FGFR1c and FGFR2c) that are expressed in mesenchymal cells. FGF signaling to the mesenchyme, in turn, induces expression of FGFs (such as FGF7 and FGF10) that bind to b-type FGFR splice forms (such as FGFR2b) expressed in epithelial cells (Ornitz et al., 1996; Zhang et al., 2006). In both the limb bud and lung, reciprocal FGF signaling between epithelial and mesenchymal tissues is absolutely required for organ development (Xu et al., 1998; Colvin et al., 2001). Consistent with reciprocal FGF signaling in the developing heart, FGF10 has been reported to be expressed in the embryonic heart (Kelly et al., 2001) and phosphorylated ERK1/2 can be detected in epicardial cells (Liberatore and Yutzey, 2004). Examination of Fgfr2b, Fgf7 and Fgf10 knockout mice will be required to evaluate whether reciprocal FGF signaling occurs in this context.

III.G.  Development of Coronary Arteries and Veins A critical component of vascular development and remodel­ ing is the emergence of a vascular tree composed of larger proximal and smaller distal blood vessels. In addition, it is thought that arterial and venous vessels differentiate during this remodeling process. Thus, the remodeling process yields many of the components of the mature circulatory system, including larger arteries and veins, medium sized arterioles and venules, and smaller capillaries (Risau,

353

1997). Interestingly, it has been reported that capillaries also acquire arterial and venous fates (Wang et al., 1998; Gerety et al., 1999; Shin et al., 2001). The molecular mechanism underlying acquisition of arterial versus venous cell fates has been recently described. These analyses were based on the observation that arterial and venous endothelial cells differentially expressed ephrinB2 and its receptor, ephB4, respectively (Wang et al., 1998; Gerety et al., 1999; Shin et al., 2001) (see Chapters 8.2 and 8.3). Intriguingly, while deletion of either ephrinB2 or ephB4 in mice severely affected vascular development, the differential expression of ephrinB2 and ephB4 was not affected, indicating that factors acting upstream of these genes control arterial versus venous identity (Lawson and Weinstein, 2002). Further work has identified these factors as components of the Notch signaling pathway. Notch1, Notch3, Notch4 and the Notch ligand, Dll4, are expressed in arterial endothelial cells, are required for vascular development, and control ephrinB2 expression (Krebs et al., 2000; Domenga et al., 2004; Duarte et al., 2004; Fischer et al., 2004). Activation of Notch signaling is sufficient to promote acquisition of the arterial cell fate, and in the absence of Notch signaling blood vessels initially form, but all express the venous markers ephB4 and flt4 (Lawson et al., 2001). Similar to other vascular systems, coronary development begins with the formation of a vascular network that is later remodeled, giving rise to the mature coronary tree (Morabito et al., 2002; Kattan et al., 2004). Interestingly, the initial coronary vascular plexus consists of two sets of blood vessels located in different positions, the subepicardial mesenchyme and the myocardial wall. Others, and ourselves, have referred to these blood vessels as subepicardial and intramyocardial blood vessels, respectively. Both sets of coronary vessels grow in a wave-like progression and are regulated by FGF and HH signaling (Lavine et al., 2006). Interestingly, growth of these blood vessels appears to be regulated by HH signaling to distinct cell types, cardiomyoblasts and perivascular cells (Lavine et al., submitted). Recent data suggests that subepicardial and intramyocardial blood vessels represent or give rise to distinct blood vessel subtypes. Corrosion casting of developing coronary vessels in the rat demonstrated that veins grow within the subepicardial space, while arteries grow within the myocardial wall (Ratajska et al., 2003). Given this, it is interesting to consider the possibility that subepicardial and intramyocardial blood vessels represent or give rise to veins and arteries, respectively. If this is true, coincident HH signaling to the cardiomyoblast and perivascular cell may coordinate venous and arterial growth. In the embryonic heart, we have observed that coronary arterial and venous fates are established either during or prior to the vascular plexus stage (Lavine et al., 2008). Thus, the coronary vascular plexus is not merely a network

354

of capillaries, but rather consists of two distinct and organized sets of arterial and venous blood vessels. Specifically, blood vessels growing within the subepicardial space (subepicardial vessels) express markers of venous identity, while blood vessels growing within the myocardial wall (intramyocardial blood vessels) express markers of arterial identity. Interestingly, the spatial relationship between these two sets of blood vessels is conserved in the adult heart, indicating that processes which pattern the vascular plexus may influence the organization of the mature vasculature (Lavine et al., 2008). It is unclear whether the vascular systems of other organs develop in a similar manner. Studies analyzing spatio–temporal patterns of ephrinB2 and ephB4 expression have produced contrasting results. In the yolk sac, arterial and venous cell fates appear to be established during the vascular plexus stage. However, in the embryonic head region, arterial and venous gene expression is not observed until the remodeling phase (Wang et al., 1998; Gerety et al., 1999). Future work will undoubtedly determine whether the establishment of arterial and venous lineages during the vascular plexus stage is a general mechanism governing vascular development.

IV.  Role of signaling pathways governing midgestational heart development in adult heart disease Interestingly, several of the transcription factors and signaling pathways that control midgestational myocardial growth and coronary vascular development function in the adult heart and are upregulated in cardiac disease. Examples include Gata4, FGF, EGF, PDGF, Ras/MAPK and several proangiogenic growth factors such as HH, VEGF, and angiopoietins. Many of these factors, including Gata4, FGF, EGF, PDGF and the Ras/MAPK signaling pathways, have been implicated in the pathogenesis of cardiac hypertrophy and heart failure (Simm et al., 1997; Ozcelik et al., 2002; Detillieux et al., 2003; Oka et al., 2006). In contrast, FGF, VEGF, angiopoietin and most recently HH, have been demonstrated to trigger coronary neoangiogenesis, and are currently under investigation as targets of pharmacological therapy aimed at promoting coronary growth following myocardial infarction (Lavine and Ornitz, 2007).

IV.A.  Developmental Signaling Pathways in the Treatment of Ischemic Heart Disease Myocardial infarction and ischemic heart disease are the leading causes of death in the industrial world. Therapies employed for treating these diseases are aimed at restoring and/or improving blood flow to ischemic cardiac tissue.

PART  |  5  Epicardial Development

Coronary artery bypass grafting (CABG) and percutaneous coronary interventions (PCI) are the mainstays of treatment for these patients, and significantly reduce morbidity and mortality (Caines et al., 2004). Unfortunately, roughly 10% of all patients are ineligible for these procedures owing to the presence of diffuse or intractable lesions. Furthermore, a significant number of diabetic patients that undergo these procedures show insufficient improvement. Without treatment, these patients do poorly, with mortality rates of up to 8–10% per year (Syed et al., 2004). Recently, a noninvasive approach aimed at promoting the growth of new coronary blood vessels has been proposed to treat patients who are ineligible for CABG and PCI. This strategy, termed pharmacological revascular­ ization or angiogenesis, involves either systemic or local administration of proangiogenic agents that promote growth of the established vasculature and/or formation of new coronary blood vessels. Over the past 10 years, it has been realized that expression of a number of known proangiogenic factors can promote coronary vessel growth in animal models. Overexpression of fibroblast growth factor-2 (FGF2), vascular endothelial growth factor-A (VEGF-A) and angiopoietin-2 (ANG2) in the myocardium of adult mice leads to significant increases in coronary artery numbers (Landau et al., 1995; Uchida et al., 1995; Rajanayagam et al., 2000; Visconti et al., 2002; House et al., 2003; Syed et al., 2004; Tammela et al., 2005). Since these initial observations, the effects of Fgf2 and Vegf-A in the adult heart have been intensively investigated and proposed as candidates for the treatment of ischemic heart disease (Scheinowitz et al., 1997; Syed et al., 2004). However, despite their ability to promote new blood vessel growth in both normal and ischemic hearts, clinical trials utilizing either protein or gene therapy have been disappointing (Henry et al., 2000; Losordo et al., 2002; Simons et al., 2002; Syed et al., 2004). An alternative approach to identify new molecules that can effectively promote coronary vascular growth is to identify signaling pathway systems that are essential for the formation of the coronary vasculature, and determine whether these signaling cascades can be reactivated in the adult heart. Moreover, as developmental programs are often recapitulated in adult physiology and tissue repair, this approach offers an opportunity to develop novel or improved therapeutic agents. We and others have postulated that an understanding of how these pathways control coronary blood vessel formation in the embryonic heart may provide important information to guide the development of novel therapeutics (Kusano et al., 2005; Lavine et al., 2006). To this end, we have sought to identify sig­ naling pathways that coordinately control the expression of proangiogenic factors during coronary development, and determine whether reactivation of these signaling pathways can promote coronary vessel growth in the adult heart.

355

Chapter  |  5.2  The Epicardial Signaling Center in Development and Disease

and large animal models (Kusano et al., 2005). These studies directly implicate the HH signaling pathway as a potential therapeutic target for pharmacological revascularization, and make a compelling case for the potential therapeutic use of HH agonists in patients with ischemic heart disease. HH signaling orchestrates both coronary development and adult coronary neovascularization by controlling the expression of multiple proangiogenic genes including Vegf-A, Vegf-B, Vegf-C and Ang2 (Fig. 6). Based on the findings that HH signaling regulates coronary vascular formation, promotes neovascularization in the adult heart,

IV.B.  HH Signaling Mediates Coronary Vessel Growth in the Adult Heart Activation of HH signaling is both necessary for coronary development and sufficient to promote formation of new coronary vessels in the embryonic and adult heart. We and others have discovered that activation of HH signaling in the adult heart promotes coronary neovascularization and protects against ischemia (Kusano et al., 2005; Lavine et al., 2006). Moreover, Shh gene therapy can promote coronary neovascularization and protect from ischemic injury in rodent

(A) Embryo FGF9, 16,20 ?RTK ligands

?

FGFR1c/2c

SHH ?IHH/DHH

VEGF-A VEGF-B VEGFR1/2 VEGF-C TIE1/2 Ang2

PTC1

Myocardium

Mesenchymal cell

Coronary vein

Coronary artery

Adult Ischemia

SHH ?IHH/DHH

?Ras/MAPK

Myocardium

PTC1

Smooth muscle cell

Control

VEGF-A Ang2 VEGF-B VEGF-C

VEGFR1/2 TIE1/2

Blood vessel

HH activation (C)

(D)

(E)

Lectin

PECAM

(B)

Figure 6  Activation of HH signaling in the adult heart promotes coronary vessel growth. (A) Comparison between HH signaling during embryonic and adult coronary vascular growth. Top: Model displaying the source and target tissues of FGF, HH, VEGF and angiopoietin signals. Dashed arrow and question mark represent the unidentified myocardial to epicardial signal controlling SHH expression. Bottom: Model describing the mechanism by which HH signaling controls coronary growth in the adult heart. Question marks represent components proposed to be involved in this signaling cascade. (B–C) PECAM staining of (B) control and (C) cardiac specific dominant active GLI2 expressing (HH activation) hearts revealing increased coronary vessel density in the interstitial space. (D–E) Streptavidin-HRP staining (arrows) following intravascular injection of biotinylated Tomato Lectin into (D) control and (E) cardiac specific dominant active GLI2 expressing animals demonstrates that HH induced blood vessels are connected to the systemic vasculature. Adapted from Lavine and Ornitz (2007) (A); Lavine et al. (2006) (B–E), with permissions from Elsevier (A) and Cold Spring Harbor Laboratory Press (B–E).

356

and induces expression of numerous signaling molecules, it has become evident that the HH pathway represents a powerful regulator capable of orchestrating coronary vascular growth in numerous settings.

References Abu-Issa, R., Smyth, G., Smoak, I., Yamamura, K., Meyers, E.N., 2002. Fgf8 is required for pharyngeal arch and cardiovascular development in the mouse. Development 129, 4613–4625. Auguste, P., Javerzat, S., Bikfalvi, A., 2003. Regulation of vascular development by fibroblast growth factors. Cell Tissue Res. 314, 157–166. Bertrand, V., Hudson, C., Caillol, D., Popovici, C., Lemaire, P., 2003. Neural tissue in ascidian embryos is induced by FGF9/16/20, acting via a combination of maternal GATA and Ets transcription factors. Cell 115, 615–627. Cai, C.L., Martin, J.C., Sun, Y., Cui, L., Wang, L., Ouyang, K., Yang, L., Bu, L., Liang, X., Zhang, X., Stallcup, W.B., Denton, C.P., McCulloch, A., Chen, J., Evans, S.M., 2008. A myocardial lineage derives from Tbx18 epicardial cells. Nature 454 (7200), 104–108. Caines, A.E., Massad, M.G., Kpodonu, J., Rebeiz, A.G., Evans, A., Geha, A.S., 2004. Outcomes of coronary artery bypass grafting versus percutaneous coronary intervention and medical therapy for multivessel disease with and without left ventricular dysfunction. Cardiology 101, 21–28. Carrier, L., Boheler, K.R., Chassagne, C., de la Bastie, D., Wisnewsky, C., Lakatta, E.G., Schwartz, K., 1992. Expression of the sarcomeric actin isogenes in the rat heart with development and senescence. Circ. Res. 70, 999–1005. Chan, D.C., Wynshaw-Boris, A., Leder, P., 1995. Formin isoforms are differentially expressed in the mouse embryo and are required for normal expression of fgf-4 and shh in the limb bud. Development 121, 3151–3162. Chen, J., Kubalak, S.W., Chien, K.R., 1998a. Ventricular muscle-restricted targeting of the RXRalpha gene reveals a non-cell-autonomous requirement in cardiac chamber morphogenesis. Development 125, 1943–1949. Chen Jr., J., Kubalak, S.W., Minamisawa, S., Price, R.L., Becker, K.D., Hickey, R., Ross, J., Chien, K.R., 1998b. Selective requirement of myosin light chain 2v in embryonic heart function. J. Biol. Chem. 273, 1252–1256. Chen, T.H., Chang, T.C., Kang, J.O., Choudhary, B., Makita, T., Tran, C.M., Burch, J.B., Eid, H., Sucov, H.M., 2002. Epicardial induction of fetal cardiomyocyte proliferation via a retinoic acid-inducible trophic factor. Dev. Biol. 250, 198–207. Colvin, J.S., White, A.C., Pratt, S.J., Ornitz, D.M., 2001. Lung hypoplasia and neonatal death in Fgf9-null mice identify this gene as an essential regulator of lung mesenchyme. Development 128, 2095–2106. Covello, K.L., Simon, M.C., 2004. HIFs, hypoxia, and vascular development. Curr. Top. Dev. Biol. 62, 37–54. Cripps, R.M., Olson, E.N., 2002. Control of cardiac development by an evolutionarily conserved transcriptional network. Dev. Biol. 246, 14–28. Crispino, J.D., Lodish, M.B., Thurberg, B.L., Litovsky, S.H., Collins, T., Molkentin, J.D., Orkin, S.H., 2001. Proper coronary vascular development and heart morphogenesis depend on interaction of GATA-4 with FOG cofactors. Genes Dev. 15, 839–844.

PART  |  5  Epicardial Development

Davies, R., Moore, A., Schedl, A., Bratt, E., Miyahawa, K., Ladomery, M., Miles, C., Menke, A., van Heyningen, V., Hastie, N., 1999. Multiple roles for the Wilms’ tumor suppressor, WT1. Cancer Res. 59, 1747s– 1750s; discussion 1751s. Detillieux, K.A., Sheikh, F., Kardami, E., Cattini, P.A., 2003. Biological activities of fibroblast growth factor-2 in the adult myocardium. Cardiovasc. Res. 57, 8–19. Dettman Jr., R.W., Denetclaw, W., Ordahl, C.P., Bristow, J., 1998. Common epicardial origin of coronary vascular smooth muscle, perivascular fibroblasts, and intermyocardial fibroblasts in the avian heart. Dev. Biol. 193, 169–181. Domenga, V., Fardoux, P., Lacombe, P., Monet, M., Maciazek, J., Krebs, L.T., Klonjkowski, B., Berrou, E., Mericskay, M., Li, Z., Tournier-Lasserve, E., Gridley, T., Joutel, A., 2004. Notch3 is required for arterial identity and maturation of vascular smooth muscle cells. Genes Dev. 18, 2730–2735. Duarte, A., Hirashima, M., Benedito, R., Trindade, A., Diniz, P., Bekman, E., Costa, L., Henrique, D., Rossant, J., 2004. Dosagesensitive requirement for mouse Dll4 in artery development. Genes Dev. 18, 2474–2478. Fischer, A., Schumacher, N., Maier, M., Sendtner, M., Gessler, M., 2004. The Notch target genes Hey1 and Hey2 are required for embryonic vascular development. Genes Dev. 18, 901–911. Flamme, I., Frolich, T., Risau, W., 1997. Molecular mechanisms of vasculogenesis and embryonic angiogenesis. J. Cell Physiol. 173, 206–210. Fossett, N., Schulz, R.A., 2001. Conserved cardiogenic functions of the multitype zinc-finger proteins: U-shaped and FOG-2. Trends Cardiovasc. Med. 11, 185–190. Gerety, S.S., Wang, H.U., Chen, Z.F., Anderson, D.J., 1999. Symmetrical mutant phenotypes of the receptor EphB4 and its specific transmembrane ligand ephrin-B2 in cardiovascular development. Mol. Cell 4, 403–414. Gittenberger-de Groot, A.C., Vrancken Peeters, M.P., Bergwerff, M., Mentink, M.M., Poelmann, R.E., 2000. Epicardial outgrowth inhibition leads to compensatory mesothelial outflow tract collar and abnormal cardiac septation and coronary formation. Circ. Res. 87, 969–971. Gittenberger-de Groot, A.C., Bartelings, M.M., Deruiter, M.C., Poelmann, R.E., 2005. Basics of cardiac development for the understanding of congenital heart malformations. Pediatr. Res. 57, 169–176. Haramis, A.G., Brown, J.M., Zeller, R., 1995. The limb deformity mutation disrupts the SHH/FGF-4 feedback loop and regulation of 5 HoxD genes during limb pattern formation. Development 121, 4237–4245. Henry, G.D., Byrne, R., Hunyh, T.T., Abraham, V., Annex, B.H., Hagen, P.O., Donatucci, C.F., 2000. Intracavernosal injections of vascular endothelial growth factor protects endothelial dependent corpora cavernosal smooth muscle relaxation in the hypercholesterol­ emic rabbit: a preliminary study. Int. J. Impot. Res. 12, 334–339. House, S.L., Bolte, C., Zhou, M., Doetschman, T., Klevitsky, R., Newman, G., Schultz Jel, J., 2003. Cardiac-specific overexpression of fibroblast growth factor-2 protects against myocardial dysfunction and infarction in a murine model of low-flow ischemia. Circulation 108, 3140–3148. Iwahori, A., Fraidenraich, D., Basilico, C., 2004. A conserved enhancer element that drives FGF4 gene expression in the embryonic myotomes is synergistically activated by GATA and bHLH proteins. Dev. Biol. 270, 525–537.

Chapter  |  5.2  The Epicardial Signaling Center in Development and Disease

Jenni, R., Oechslin, E.N., van der Loo, B., 2007. Isolated ventricular noncompaction of the myocardium in adults. Heart 93, 11–15. Junga, G., Kneifel, S., Von Smekal, A., Steinert, H., Buersfeld, U., 1999. Myocardial ischaemia in children with isolated ventricular noncompaction. Eur. Heart. J. 20, 910–916. Kanda, S., Miyata, Y., Kanetake, H., 2004. Fibroblast growth factor-2mediated capillary morphogenesis of endothelial cells requires signals via Flt-1/vascular endothelial growth factor receptor-1: possible involvement of c-Akt. J. Biol. Chem. 279, 4007–4016. Kattan, J., Dettman, R.W., Bristow, J., 2004. Formation and remodeling of the coronary vascular bed in the embryonic avian heart. Dev. Dyn. 230, 34–43. Kelly, R.G., 2005. Molecular inroads into the anterior heart field. Trends Cardiovasc. Med. 15, 51–56. Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1, 435–440. Khokha, M.K., Hsu, D., Brunet, L.J., Dionne, M.S., Harland, R.M., 2003. Gremlin is the BMP antagonist required for maintenance of Shh and Fgf signals during limb patterning. Nat. Genet. 34, 303–307. Krebs, L.T., Xue, Y., Norton, C.R., Shutter, J.R., Maguire, M., Sundberg, J.P., Gallahan, D., Closson, V., Kitajewski, J., Callahan, R., Smith, G.H., Stark, K.L., Gridley, T., 2000. Notch signaling is essential for vascular morphogenesis in mice. Genes. Dev. 14, 1343–1352. Kusano, K.F., Pola, R., Murayama, T., Curry, C., Kawamoto, A., Iwakura, A., Shintani, S., Ii, M., Asai, J., Tkebuchava, T., Thorne, T., Takenaka, H., Aikawa, R., Goukassian, D., von Samson, P., Hamada, H., Yoon, Y.S., Silver, M., Eaton, E., Ma, H., Heyd, L., Kearney, M., Munger, W., Porter, J.A., Kishore, R., Losordo, D.W., 2005. Sonic hedgehog myocardial gene therapy: tissue repair through transient reconstitution of embryonic signaling. Nat. Med. 11, 1197–1204. Kwee, L., Baldwin, H.S., Shen, H.M., Stewart, C.L., Buck, C., Buck, C.A., Labow, M.A., 1995. Defective development of the embryonic and extraembryonic circulatory systems in vascular cell adhesion molecule (VCAM-1) deficient mice. Development 121, 489–503. Landau, C., Jacobs, A.K., Haudenschild, C.C., 1995. Intrapericardial basic fibroblast growth factor induces myocardial angiogenesis in a rabbit model of chronic ischemia. Am. Heart. J. 129, 924–931. Lavine, K.J., Ornitz, D.M., 2007. Rebuilding the coronary vasculature: hedgehog as a new candidate for pharmacological revascularization. Trends Cardiovasc. Med. 17, 77–83. Lavine, K.J., Long, F., Choi, K., Smith, C., Ornitz, D.M., 2008. Hedgehog signaling to distinct cell types differentially regulates coronary artery and vein development. Development 135, 3161–3171. Lavine, K.J., Yu, K., White, A.C., Zhang, X., Smith, C., Partanen, J., Ornitz, D.M., 2005. Endocardial and epicardial derived FGF signals regulate myocardial proliferation and differentiation in vivo. Dev. Cell 8, 85–95. Lavine, K.J., White, A.C., Park, C., Smith, C.S., Choi, K., Long, F., Hui, C., Ornitz, D.M., 2006. Fibroblast growth factor signals regulate a wave of Hedgehog activation that is essential for coronary vascular development. Genes Dev. 20, 1651–1661. Lawson, N.D., Scheer, N., Pham, V.N., Kim, C.H., Chitnis, A.B., Campos-Ortega, J.A., Weinstein, B.M., 2001. Notch signaling is required for arterial–venous differentiation during embryonic vascular development. Development 128, 3675–3683. Lawson, N.D., Weinstein, B.M., 2002. Arteries and veins: making a difference with zebrafish. Nat. Rev. Genet. 3, 674–682.

357

Liberatore, C.M., Yutzey, K.E., 2004. MAP kinase activation in avian cardiovascular development. Dev. Dyn. 230, 773–780. Losordo, D.W., Vale, P.R., Hendel, R.C., Milliken, C.E., Fortuin, F.D., Cummings, N., Schatz, R.A., Asahara, T., Isner, J.M., Kuntz, R.E., 2002. Phase 1/2 placebo-controlled, double-blind, dose-escalating trial of myocardial vascular endothelial growth factor 2 gene transfer by catheter delivery in patients with chronic myocardial ischemia. Circulation 105, 2012–2018. Merki, E., Zamora, M., Raya, A., Kawakami, Y., Wang, J., Zhang, X., Burch, J., Kubalak, S.W., Kaliman, P., Belmonte, J.C., Chien, K.R., Ruiz-Lozano, P., 2005. Epicardial retinoid X receptor alpha is required for myocardial growth and coronary artery formation. Proc. Natl. Acad. Sci. USA 102, 18455–18460. Meyers, E.N., Lewandoski, M., Martin, G.R., 1998. An Fgf8 mutant allelic series generated by Cre- and Flp-mediated recombination. Nat. Genet. 18, 136–141. Miano Jr, J.M., Ramanan, N., Georger, M.A., de Mesy Bentley, K.L., Emerson, R.L., Balza, R.O., Xiao, Q., Weiler, H., Ginty, D.D., Misra, R.P., 2004. Restricted inactivation of serum response factor to the cardiovascular system. Proc. Nat. Acad. Sci. USA 101, 17132–17137. Mic, F.A., Haselbeck, R.J., Cuenca, A.E., Duester, G., 2002. Novel retinoic acid generating activities in the neural tube and heart identified by conditional rescue of Raldh2 null mutant mice. Development 129, 2271–2282. Mikawa, T., Fischman, D.A., 1992. Retroviral analysis of cardiac morphogenesis: discontinuous formation of coronary vessels. Proc. Natl. Acad. Sci. USA 89, 9504–9508. Mikawa, T., Gourdie, R.G., 1996. Pericardial mesoderm generates a population of coronary smooth muscle cells migrating into the heart along with ingrowth of the epicardial organ. Dev. Biol. 174, 221–232. Mohun, T., Sparrow, D., 1997. Early steps in vertebrate cardiogenesis. Curr. Opin. Gen. Dev. 7, 628–633. Moore, A.W., McInnes, L., Kreidberg, J., Hastie, N.D., Schedl, A., 1999. YAC complementation shows a requirement for Wt1 in the development of epicardium, adrenal gland and throughout nephrogenesis. Development 126, 1845–1857. Morabito, C.J., Kattan, J., Bristow, J., 2002. Mechanisms of embryonic coronary artery development. Curr. Opin. Cardiol. 17, 235–241. Murakami, A., Shen, H., Ishida, S., Dickson, C., 2004. SOX7 and GATA-4 are competitive activators of Fgf-3 transcription. J. Biol. Chem. 279, 28564–28573. Nesbitt, T., Lemley, A., Davis, J., Yost, M.J., Goodwin, R.L., Potts, J.D., 2006. Epicardial development in the rat: A new perspective. Microsc. Microanal. 12, 390–398. Oka, T., Maillet, M., Watt, A.J., Schwartz, R.J., Aronow, B.J., Duncan, S.A., Molkentin, J.D., 2006. Cardiac-specific deletion of Gata4 reveals its requirement for hypertrophy, compensation, and myocyte viability. Circ. Res. 98, 837–845. Ornitz, D.M., Xu, J., Colvin, J.S., McEwen, D.G., MacArthur, C.A., Coulier, F., Gao, G., Goldfarb, M., 1996. Receptor specificity of the fibroblast growth factor family. J. Biol. Chem. 271, 15292–15297. Ozcelik, C., Erdmann, B., Pilz, B., Wettschureck, N., Britsch, S., Hubner, N., Chien, K.R., Birchmeier, C., Garratt, A.N., 2002. Conditional mutation of the ErbB2 (HER2) receptor in cardiomyocytes leads to dilated cardiomyopathy. Proc. Natl. Acad. Sci. USA 99, 8880–8885. Perez-Pomares, J.M., Carmona, R., Gonzalez-Iriarte, M., Atencia, G., Wessels, A., Munoz-Chapuli, R., 2002. Origin of coronary endothelial

358

cells from epicardial mesothelium in avian embryos. Int. J. Dev. Biol. 46, 1005–1013. Pola, R., Ling, L.E., Aprahamian, T.R., Barban, E., Bosch-Marce, M., Curry, C., Corbley, M., Kearney, M., Isner, J.M., Losordo, D.W., 2003. Postnatal recapitulation of embryonic hedgehog pathway in response to skeletal muscle ischemia. Circulation 108, 479–485. Price, R.L., Thielen, T.E., Borg, T.K., Terracio, L., 2001. Cardiac defects associated with the absence of the platelet-derived growth factor alpha receptor in the patch mouse. Microsc. Microanal. 7, 56–65. Rajanayagam, M.A., Shou, M., Thirumurti, V., Lazarous, D.F., Quyyumi, A.A., Goncalves, L., Stiber, J., Epstein, S.E., Unger, E.F., 2000. Intracoronary basic fibroblast growth factor enhances myocardial collateral perfusion in dogs. J. Am. Coll. Cardiol. 35, 519–526. Ratajska, A., Ciszek, B., Sowinska, A., 2003. Embryonic development of coronary vasculature in rats: corrosion casting studies. Anat. Rec. A. Discov. Mol. Cell Evol. Biol. 270, 109–116. Reese, D.E., Mikawa, T., Bader, D.M., 2002. Development of the coronary vessel system. Circ. Res. 91, 761–768. Risau, W., 1997. Mechanisms of angiogenesis. Nature 386, 671–674. Scheinowitz, M., Abramov, D., Eldar, M., 1997. The role of insulin-like and basic fibroblast growth factors on ischemic and infarcted myocardium: a mini review. Int. J. Cardiol. 59, 1–5. Schipani, E., Ryan, H.E., Didrickson, S., Kobayashi, T., Knight, M., Johnson, R.S., 2001. Hypoxia in cartilage: HIF-1alpha is essential for chondrocyte growth arrest and survival. Genes Dev. 15, 2865–2876. Seghezzi, G., Patel, S., Ren, C.J., Gualandris, A., Pintucci, G., Robbins, E.S., Shapiro, R.L., Galloway, A.C., Rifkin, D.B., Mignatti, P., 1998. Fibroblast growth factor-2 (FGF-2) induces vascular endothelial growth factor (VEGF) expression in the endothelial cells of forming capillaries: an autocrine mechanism contributing to angiogenesis. J. Cell Biol. 141, 1659–1673. Shao, Y., Lu, J., Zhang, G., Liu, C., Huang, B., 2005. Histone acetyltransferase p300 promotes the activation of human WT1 promoter and intronic enhancer. Arch. Biochem. Biophys. 436, 62–68. Shikama, N., Lutz, W., Kretzschmar, R., Sauter, N., Roth, J.F., Marino, S., Wittwer, J., Scheidweiler, A., Eckner, R., 2003. Essential function of p300 acetyltransferase activity in heart, lung and small intestine formation. EMBO. J. 22, 5175–5185. Shin Jr., D., Garcia-Cardena, G., Hayashi, S., Gerety, S., Asahara, T., Stavrakis, G., Isner, J., Folkman, J., Gimbrone, M.A., Anderson, D.J., 2001. Expression of ephrinB2 identifies a stable genetic difference between arterial and venous vascular smooth muscle as well as endothelial cells, and marks subsets of microvessels at sites of adult neovascularization. Dev. Biol. 230, 139–150. Simm, A., Nestler, M., Hoppe, V., 1997. PDGF-AA, a potent mitogen for cardiac fibroblasts from adult rats. J. Mol. Cell Cardiol. 29, 357–368. Simons, M., Annex, B.H., Laham, R.J., Kleiman, N., Henry, T., Dauerman, H., Udelson, J.E., Gervino, E.V., Pike, M., Whitehouse, M.J., Moon, T., Chronos, N.A., 2002. Pharmacological treatment of coronary artery disease with recombinant fibroblast growth factor-2: double-blind, randomized, controlled clinical trial. Circulation 105, 788–793. Smith, T.K., Bader, D.M., 2007. Signals from both sides: Control of cardiac development by the endocardium and epicardium. Semin. Cell Dev. Biol. 18, 84–89. Stottmann, R.W., Choi, M., Mishina, Y., Meyers, E.N., Klingensmith, J., 2004. BMP receptor IA is required in mammalian neural crest cells for development of the cardiac outflow tract and ventricular myocardium. Development 131, 2205–2218.

PART  |  5  Epicardial Development

Stuckmann, I., Evans, S., Lassar, A.B., 2003. Erythropoietin and retinoic acid, secreted from the epicardium, are required for cardiac myocyte proliferation. Dev. Biol. 255, 334–349. Sucov, H.M., Dyson, E., Gumeringer, C.L., Price, J., Chien, K.R., Evans, R.M., 1994. RXR alpha mutant mice establish a genetic basis for vitamin A signaling in heart morphogenesis. Genes Dev. 8, 1007–1018. Sugi, Y., Lough, J., 1992. Onset of expression and regional deposition of alpha-smooth and sarcomeric actin during avian heart development. Dev. Dyn. 193, 116–124. Syed, I.S., Sanborn, T.A., Rosengart, T.K., 2004. Therapeutic angiogenesis: a biologic bypass. Cardiology 101, 131–143. Tammela, T., Enholm, B., Alitalo, K., Paavonen, K., 2005. The biology of vascular endothelial growth factors. Cardiovasc. Res. 65, 550–563. Tan, C., de Noronha, R.G., Roecker, A.J., Pyrzynska, B., Khwaja, F., Zhang, Z., Zhang, H., Teng, Q., Nicholson, A.C., Giannakakou, P., Zhou, W., Olson, J.J., Pereira, M.M., Nicolaou, K.C., Van Meir, E.G., 2005. Identification of a novel small-molecule inhibitor of the hypoxia-inducible factor 1 pathway. Cancer Res. 65, 605–612. Tevosian, S.G., Deconinck, A.E., Tanaka, M., Schinke, M., Litovsky, S.H., Izumo, S., Fujiwara, Y., Orkin, S.H., 2000. FOG-2, a cofactor for GATA transcription factors, is essential for heart morphogenesis and development of coronary vessels from epicardium. Cell 101, 729–739. Tomita, S., Ueno, M., Sakamoto, M., Kitahama, Y., Ueki, M., Maekawa, N., Sakamoto, H., Gassmann, M., Kageyama, R., Ueda, N., Gonzalez, F.J., Takahama, Y., 2003. Defective brain development in mice lacking the Hif-1alpha gene in neural cells. Mol. Cell Biol. 23, 6739–6749. Uchida, Y., Yanagisawa-Miwa, A., Nakamura, F., Yamada, K., Tomaru, T., Kimura, K., Morita, T., 1995. Angiogenic therapy of acute myocardial infarction by intrapericardial injection of basic fibroblast growth factor and heparin sulfate: an experimental study. Am. Heart J. 130, 1182–1188. Vaupel, P., 2004. The role of hypoxia-induced factors in tumor progression. Oncologist 9 (Suppl. 5), 10–17. Visconti, R.P., Richardson, C.D., Sato, T.N., 2002. Orchestration of angiogenesis and arteriovenous contribution by angiopoietins and vascular endothelial growth factor (VEGF). Proc. Natl. Acad. Sci. USA 99, 8219–8224. Vrancken Peeters, M.P., Gittenberger-de Groot, A.C., Mentink, M.M., Poelmann, R.E., 1999. Smooth muscle cells and fibroblasts of the coronary arteries derive from epithelial-mesenchymal transformation of the epicardium. Anat. Embryol. (Berlin) 199, 367–378. Wada, A.M., Willet, S.G., Bader, D., 2003. Coronary vessel development: a unique form of vasculogenesis. Arterioscler. Thromb. Vasc. Biol. 23, 2138–2145. Wang, H.U., Chen, Z.F., Anderson, D.J., 1998. Molecular distinction and angiogenic interaction between embryonic arteries and veins revealed by ephrin-B2 and its receptor Eph-B4. Cell 93, 741–753. Watt, A.J., Battle, M.A., Li, J., Duncan, S.A., 2004. GATA4 is essential for formation of the proepicardium and regulates cardiogenesis. Proc. Natl. Acad. Sci. USA 101, 12573–12578. Wu, H., Lee, S.H., Gao, J., Liu, X., Iruela-Arispe, M.L., 1999. Inactivation of erythropoietin leads to defects in cardiac morphogenesis. Development 126, 3597–3605. Xu, R.H., Ault, K.T., Kim, J., Park, M.J., Hwang, Y.S., Peng, Y., Sredni, D., Kung, H., 1999. Opposite effects of FGF and BMP-4 on embryonic blood formation: roles of PV.1 and GATA-2. Dev. Biol. 208, 352–361.

Chapter  |  5.2  The Epicardial Signaling Center in Development and Disease

Xu, X., Weinstein, M., Li, C., Naski, M., Cohen, R.I., Ornitz, D.M., Leder, P., Deng, C., 1998. Fibroblast growth factor receptor 2 (FGFR2)-mediated reciprocal regulation loop between FGF8 and FGF10 is essential for limb induction. Development 125, 753–765. Yang, J.T., Rayburn, H., Hynes, R.O., 1995. Cell adhesion events med­ iated by alpha 4 integrins are essential in placental and cardiac development. Development 121, 549–560. Zhang, X., Stappenbeck, T.S., White, A.C., Lavine, K.J., Gordon, J.I., Ornitz, D.M., 2006. Reciprocal epithelial-mesenchymal FGF signaling is required for cecal development. Development 133, 173–180. Zhao, T., Chang, L.W., McLeod, H.L., Stormo, G.D., 2004. PromoLign: a database for upstream region analysis and SNPs. Hum. Mutat. 23, 534–539.

359

Zhou, B., Ma, Q., Rajagopal, S., Wu, S.M., Domian, I., Rivera-Feliciano, J., Jiang, D., von Gise, A., Ikeda, S., Chien, K.R., Pu, W.T., 2008. Epicardial progenitors contribute to the cardiomyocyte lineage in the developing heart. Nature 454, 109–113. Zuniga, A., Haramis, A.P., McMahon, A.P., Zeller, R., 1999. Signal relay by BMP antagonism controls the SHH/FGF4 feedback loop in vertebrate limb buds. Nature 401, 598–602. Zuniga, A., Michos, O., Spitz, F., Haramis, A.P., Panman, L., Galli, A., Vintersten, K., Klasen, C., Mansfield, W., Kuc, S., Duboule, D., Dono, R., Zeller, R., 2004. Mouse limb deformity mutations disrupt a global control region within the large regulatory landscape required for Gremlin expression. Genes Dev. 18, 1553–1564.

Chapter 6.1

Molecular Regulation of Cushion Morphogenesis Todd D. Camenisch1, Raymond B. Runyan2 and Roger R. Markwald3 1 Departments of Pharmacology and Toxicology, Molecular and Cellular Biology, Steele Children’s Research Center, Bio5 Institute, University of Arizona, Tucson, AZ, USA 2 Department of Cell Biology and Anatomy, University of Arizona, Tucson, AZ, USA 3 Department of Anatomy and Cell Biology & Developmental Cardiovascular Biology Center, Medical University of South Carolina, USA

I.  Origins and morphogenetic stages of valve–septal development I.A.  Introduction The endocardial cushions are some of the most unique regions in the forming heart. These transient structures give rise to the geometric center of the functioning heart. The process of how such critical structures form from limited progenitor cells is one of the most complex and fascinating stories in developmental biology. The complexity of this aspect of heart development is reflected in the common observation of valvular and septal defect involvement in many of the forms of congenital heart disease found in newborns. This chapter will discuss where and how the cardiac cushions form, the constituency of the cushions, progress in deciphering the molecular regulatory networks governing the functional formation of the valve leaflets, and partitioning of a four-chambered heart. Immediately following looping of the primitive heart tube, regional expansion of the extracellular matrix (cardiac jelly) separating the outer myocardial lining and the inner endothelial lining of the heart occurs to initiate endocardial cushion morphogenesis by a process called epithelial– mesenchymal transition (EMT). This process occurs initially in the atrioventricular (AV) canal region and later in the outflow tract (OFT) to produce cushion mesenchyme (Fig. 1). In the atrioventricular canal two main masses of cushion tissue, the dorsal or inferior atrioventricular cushion (associated with the greater curvature of the looped heart) and the ventral or superior atrioventricular

Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

endocardial cushion (associated with the inner curvature of the heart) originate from the epithelial–mesenchymal transition. There are two lateral cushions that contribute to the mural leaflets of the atrioventricular valves. The atrioventricular canal becomes divided into two distinct right and left orifices when the two main cushion masses fuse to form the atrioventricular septum intermedium and divide the common atrioventricular canal into a right-sided tricuspid orifice and left-sided, mitral orifice. These cushions successfully establish unidirectional blood flow even before they become mesenchymalized. The septum intermedium provides a “target” for insertion of the atrial primary septum and the interventricular muscular septum. Partitioning is completed when all three septa are fused in line, the only communications between right and left sides are through foramina located above and below the atrioventricular septum. This gives the structural appearance of a stretched figure eight containing the septum primum to the atrial aspect, the inferior and superior endocardial cushions of the atrioventricular canal, and the primitive interventricular septum (Webb et al., 1998). This structure serves simultaneously to segregate the atria and the apical trabeculated region of both ventricles, while dividing the single primitive inlet into right and left sides (Wessels et al., 1996). This establishes directional communication between the right atria with the right ventricle and the left atrium with the primitive left ventricle. This partitioning is vital for the efficient function of the future four-chambered heart, which not only operates to support the rapidly growing embryo, but also sustains the organism throughout its life.

363

364

PART  |  6  Cushions, Valves and Septa

(B)

(A)

A

OFT A

AVC

CM

Ec

V

V Myo

(C)

SA Node Septum secundum SVC

Foramen ovale RA

Septum primum

LA

Ec-Septum primum

IVC

Mitral valve

Membranous IVS

Tricuspid valve Chordae tendineae

Trabeculae RV

LV

Figure 1  Views of the heart during cushion development. (A) Perspective of the looped heart showing the curvature and overlap of the atrioventricular (AV) canal and the outflow tract. A single atrium and ventricle are separated by the atrioventricular canal. There is a yellow extracellular matrix (ECM) between the pink myocardium and the red endothelium. The extracellular matrix becomes expanded into cushions in the atrioventricular canal by increased synthesis of matrix molecles from the myocardium. The cushions are populated by mesenchymal cells as a result of epithelial–mesenchymal transition and subsequent cell proliferation. (B) A stylized sagital section of the heart that removes the overlap and emphasizes the epithelial–mesenchymal transition process that takes place in the atrioventricular canal and the proximal portion of the outflow tract. The blue portion of the diagram reflects the change in the extracellular matrix produced by secretion products of the endothelium and the newly-formed mesenchyme. Outflow tract epithelial–mesenchymal transition forms the progenitors of the aortic and pulmonary valves. (C) Stylized sagital section of the four-chambered heart showing structures (pink) derived from atrioventricular canal mesenchyme. Structures include the leaflets and chordae tendinae of the mitral and tricuspid valves, the membranous portion of the interventricular septum and the portion of the septum intermedium where the septum primum attaches.

365

Chapter  |  6.1  Molecular Regulation of Cushion Morphogenesis

I.B.  Structure–Function Relationships of Early Cushions

mechanism previously described (Forouhar et al., 2006). Based on mesomechanical testing (Butcher and Markwald, 2007), atrioventricular cushions increase in viscoelastic properties (stiffness) between HH17 and HH25 in chicken embryos concomitant with changes in function (Fig. 3). Enzymatic digestion to remove major structural extracellular matrix (ECM) constituents (e.g., collagens versus glycosaminoglycans) resulted in distinctly different stress–strain curves suggestive of their individual material contributions to the properties of endocardial cushions. These changes imply that embryonic cushion structure and atrioventricular function are intimately linked; specifically, structure signals function and vice versa.

The atrioventricular and outflow tract cushions of the heart are exposed to a constant barrage of hemodynamic and mechanical forces that increase over time during development. Early investigations using chick embryos highlighted the motions of these cushions in concert with the contracting myocardium, suggesting that these serve a valve-like function before actual valves form (Patten et al., 1948). However, a recent study showed that the atrioventricular canal in the early tubular heart of zebrafish functions like a suction pump, in contrast to the peristaltic

Figure 2  Developmental series on formation of the atrioventricular and semilunar valves. The top row shows the early expansion of the cardiac cushions by myocardial synthesis of extracellular matrix, the formation of mesenchyme by epithelial–mesenchymal transition, the cavitation and remodeling of the populated cushion into a valvular structure and the connection of valves to the papillary muscle by cushion-derived chordae tendinae. The lower row shows that a similar process occurs in the semi­ lunar valves of the aorta and pulmonary trunk, but without leaflet formation of chordae tendinae.

Atrioventricular valves

Early

EMT

Remodeling

Mature

Excavation

Mature

50

0.5

50

0.5

40

0.4

40

0.4

30

0.3

30

0.3

20

0.2

20

0.2

10

0.1

10

0.1

0

0

0.2

0.4 0.6 0.8 Cycle fraction Myocard Cushion

1

0

0

0

0.2

0.4 0.6 0.8 Cycle fraction Myocard Cushion

1

0

50

0.5

40

0.4

30

0.3

20

0.2

10

0.1

0

0

0.2

0.4 0.6 0.8 Cycle fraction Myocard Cushion

1

0

Cushion thickness, mm

EMT

Thickness, mm % Contraction

% Contraction

Early

Cushion thickness, mm % Contraction

Semilunar valves

Figure 3  Function of the atrioventricular cushions changes during the developmental sequence. Images are from videos of chick embryos photographed at Hamburger-Hamilton stage 17 (approximately 64 hours, start of epithelial–mesenchymal transition), stage 21 (approximately 84 hours, well-populated cushions) and stage 25 (approximately 108 hours). Lines are drawn on the figures to approximate the outlines of the cushions. M-mode diagrams below each figure illustrate the contraction of the myocardium (black line) and the thickness of the cushion (gray line). See Butcher et al. (2007a) for details.

366

II.  Cushion formation II.A.  The Extracellular Matrix The extracellular matrix of the endocardial cushions is a unique and dynamic mixture of proteins, glycosaminoglycans and proteoglycans that, at its origin, is devoid of any cells. The constituents of the matrix, prior to the epithelial– mesenchymal transition process, are primarily derived from the highly secretory myocardium lining the cushion regions (Krug et al., 1985). Major components include hyaluronan (HA), the proteoglycans aggrecan and versican, and glycoproteins including laminin, collagens (I, II, IV), fibronectin, fibulins, fibrillins and periostin (Kitten et al., 1987; Wunsch et al., 1994; Bouchey et al., 1996; Kern et al., 2005). This can be considered a highly-hydrated expanded basement membrane extending from the myocardium across to the endothelium lining these specialized heart segments. Recent studies also indicate that the endothelium contributes to the make-up of the cardiac jelly (Klewer et al., 2006). Importantly, the production and expansion of the cardiac jelly precedes epithelial– mesenchymal transition and production of mesenchyme (see section below). The importance of cardiac jelly in the morphology and development of the endocardial cushions is highlighted in the Trisomy 16 mouse. Murine trisomy 16 (Ts16) is accepted as a model for Trisomy 21 in humans (Down Syndrome) as there is genetic synteny in these regions and similar phenotypes. Ts21 in humans has a high incidence of heart defects (~40%) from abnormal endocardial cushion development. In Ts16 embryonic hearts, there is greatly elevated cushion volume due to increase in cardiac jelly content, but attenuated production of mesenchyme (Webb et al., 1996; Chapter 6.2). Ts16 embryos have a very high incidence of AVSD as a result of abnormal cushion formation (almost 100%). Thus the coordinated production and reduction in matrix within the cushions are equally important during the creation of valve and septal structures from the cardiac cushions. The regulatory mechanisms governing the selective production of cushion extracellular matrix are not clearly known, but some insights have been acquired through mouse functional genomic studies, and here we will discuss hyaluronan (HA) and versican as two prime examples of this biology. Glycosaminoglycans are also critical components of the extracellular milieu and are required for normal development, being conserved among species from invertebrates through higher vertebrate organisms. The principal glycosaminoglycan within the extracellular matrix of most mammalian tissues is hyaluronan (HA) including that present in cardiac jelly (Manasek et al., 1973; Bernanke and Markwald, 1979). This linear polysaccharide of repeating D-glucuronic acid and N-acetyl glucosamine residues is produced at the plasma membrane by specific enzymes termed HA-synthases (Has). Since HA

PART  |  6  Cushions, Valves and Septa

is not synthesized in vesicular compartments, it is unique in that it is not sulfated nor does it contain a protein core. HA functions during heart morphogenesis (Manasek et al., 1973). HA participates in creating a water-enriched macromolecular environment, and is believed to regulate embryonic development by establishing hydrated lowresistance matrix that promotes loss of cell contact inhibition and increased cell motility processes critical for endocardial cushion formation. HA is the principle matrix component responsible for the rapid expansion, hydration and organization of the cardiac jelly in the cushion regions. One key HA synthase, Has2, is responsible for the production of hyaluronan (HA) during morphogenesis of the cardiovascular system, in particular formation of the endocardial cushions (Baldwin and Solursh, 1990; Camenisch et al., 2000a). In general, during rapid expansion of extracellular space, there is temporal elevation in the deposition of HA supporting the formation of the cardiac cushions. There is dynamic expression of Has2, initially in the myocardium, then switching to the endocardium and mesenchyme before eventual downregulation as the valves mature. This pattern coincides with the deposition of HA between the myocardium and endocardium to expand the space into the cushion-like structure. Studies by Baldwin and Solursh (1990) showed that degradation of HA with hyaluronidase treatment of whole conceptus cultures results in abnormal endocardial cushion formation. Definitive evidence for HA functioning in endocardial cushion biology came from the genetic disruption of the principle enzyme, Has2 (Camenisch et al., 2000). Homozygous Has2/ embryos die at E10.0 from heart defects including disrupted cushion morphogenesis. Studies in zebrafish also support a role for HA in the production of endocardial cushion mesenchyme. The zebrafish mutant Jekyll exhibits hypoplastic cushions (Walsh and Stainier, 2001) similar to mouse embryos lacking Has2/. Jekyll mutants lack a functional uri­ dine 5-diphosphate(UDP)-glucose dehydrogenase gene whose enzymatic activity is required for HA biosynthesis. Thus, the defects in Jekyll mutants are likely related to deficits in HA production and activity. These shared phenotypes underscore the importance in the deposition of HA during cushion formation. Additional studies with the Has2/ mutants show that HA has more biological activity than previously recognized. Specifically, primary atrioventricular canal explants from Has2/ embryos do not show recovery of epithelial–mesenchymal transition as do versicandeficient (heart defect mouse model, hdf) cushion cultures. The Has2/ defect in epithelial–mesenchymal transition is rescued by exogenous HA or by expressing Has2 cDNA in Has2/ cushion explant cultures (Camenisch et al., 2000). The requirement for Has2-dependent HA induction of epithelial–mesenchymal transition can be circumvented by constitutively active Ras (caRas), thereby emphasizing the

Chapter  |  6.1  Molecular Regulation of Cushion Morphogenesis

importance of this signaling effector in the inductive pathway. Further evidence comes from detailed work examining the ErbB receptors during heart development (see section below). Versican is a large chondroitin-sulphate proteoglycan that can bind HA and is also present in cardiac jelly. There are four isoforms (V0, V1, V2, V3) with the core protein having amino and carboxy globular domains that can bind HA, 1 integrin and fibulin-1-2, respectively (Kern et al., 2006). The V1 isoform can upregulate EGFR to stimulate cell growth, while the V2 isoform has opposite inhibitory effects (Sheng et al., 2005). The hdf mouse model is a result of insertional-disruption by a LacZ reporter transgene into the Cspg2 gene (versican) on chromosome 13. LacZ expression is observed in the myocardium during endocardial cushion formation. Versican is deposited within endocardial cushions, with less detected on the endocardial side often associated with mesenchymal cells (Kern et al., 2006). Homozygote hdf-mutants do not show regional expansion of the cardiac cushions in the conus-truncus segment or in the atrioventricular canal. Hdf embryonic hearts do not show transformation of the endothelium into mesenchyme in vivo; however, endocardial cushion explants from the atrioventricular canal cultured in collagen gel invasion assay in vitro do show production of mesenchymal cells, despite the absence of versican proteoglycan. Thus, in a permissive context, versican-deficient cushion endothelium retains competency to execute epithelial–mesenchymal transition. This suggests that versican serves crucial structural support functions as a proteoglycan component of the cardiac jelly in the endocardial cushions.

II.B.  Dynamics of the Matrix The matrix constituents of the cardiac jelly within the endocardial cushions must be removed or redistributed during the remodeling phase of valve and septal development. The change in morphology of the cushions correlates with their transition from a peristaltic-like motion to a more defined piston-type of motion (Butcher et al., 2007a) to prevent retrograde flow as the heart enlarges. This remodeling and maturation of the valves coincides with reduction of matrix, due to upregulation of proteases and hyaluronidase activity (McGuire and Orkin, 1992). HA deposition is known to be critical as noted above, but its removal and change in the matrix environment is just as critical in developing the functional valve. Failure to do so may contribute to myxmatous types of valve disorders. Importantly, hyaluronidases are significantly upregulated in the areas with increased mesenchymal cell migration (Bernanke and Orkin, 1984). Hyaluronan was previously shown in in vitro studies to take on new biological roles, once broken down into small oligosaccharide fragments

367

(o-HA) (Noble et al., 1996; Slevin et al., 1998); these o-HA forms have proangiogenic effects. It was recently reported that o-HA induces increased levels of VEGF, which is inhibitory to epithelial–mesenchymal transition (see section below). In murine in vitro cushion explant assays, VEGF levels are markedly increased at both the mRNA and protein levels following treatment with o-HA. Production of mesenchyme is significantly decreased as a result of a block in epithelial–mesenchymal transition following treatment with o-HA (Rodgers et al., 2006). This phenotype is reversed by treatment with a soluble VEGF receptor (sFlt) with complete restoration of epithelial– mesenchymal transition. In addition, o-HA injected in vivo into endocardial cushions not only increases and extends the pattern of VEGF expression, but also decreases epithelial–mesenchymal transition within the endocardial cushion regions compared to controls. This data shows a direct functional link between depolymerized forms of HA and VEGF in the regulation of developmental epithelial–mesenchymal transition. This is in contrast to the proepithelial–mesenchymal transition activity of the naïve high molecular weight HA (HMW-HA) initially deposited and responsible for the expansion of the cardiac cushions. This highlights the dynamic environment of the extracellular matrix in the cushions, and supports the concept that matrix is not just a space-filling substance but can exhibit and support bioactivities. There is also dynamic processing of proteoglycans including versican by specific proteases such as ADAMTS, and the MMPs-1, -2, -3, -7 and -9. ADAMTS-1 and -9 are expressed during endocardial cushion development, with the latter demonstrated to produce cleavage products of versican in the embryonic heart. Specific DPEAAE versican fragments are located to the endocardium and newly-formed mesenchymal cells. The processing of versican appears to occur simultaneously or immediately following the epithelial–mesenchymal transition process (Kern et al., 2006), suggestive of a role for this versican peptide in mesenchymal cell biology. Another prime example of matrix dynamics is observed with type XVIII collagen. This collagen is a multi-domain protein that can be cleaved into fragments, one of which is endostatin. Collagen XVIII is encoded on chromosome 21 in the region linked with Down syndrome (Trisomy 21) and associated heart defects. Collagen XVIII is detected in the cushion and valve forming regions within the atrio­ ventricular canal (Carvalhaes et al., 2006). It is tightlyassociated with the mesenchymal cells as they migrate into the cardiac jelly. By late gestation, it becomes restricted to the basement membrane of the endothelial cells of the valve leaflets, similar to type VI collagen which has two of its collagen chains encoded on chromosome 21 as well. Since endostatin has potent anti-angiogenic activity, this matrix component may have additional bioactivity during endocardial cushion formation.

368

PART  |  6  Cushions, Valves and Septa

II.C.  The Epithelial–Mesenchymal Transition Paradigm Epithelial to mesenchymal transition occurs in the cardiac cushions of the atrioventricular canal and outflow tract regions. Here we will focus on the epithelial–mesenchymal transition for the atrioventricular canal. The process of epithelial–mesenchymal transition occurs around E9.5 in the mouse (21–28 somites) and at stages HH15–17 (25–28 somites) in the chick (Camenisch et al., 2002a; Delot et al., 2003). The physical process of epithelial–mesenchymal transition is not completely understood in the heart. Epithelial–mesenchymal transition in the atrioventricular canal becomes visible when a subset of endocardial cells loses cell-to-cell contact, hypertrophy, polarize the

golgi apparatus and extend filopodia into the underlying cushion matrix (Markwald et al., 1977; Kinsella and Fitzharris, 1980; Krug et al., 1985). Analysis of the epithelial–mesenchymal transition process was enabled by the development of a collagen gel epithelial–mesenchymal transition assay (Fig. 4). This assay first demonstrated that epithelial–mesenchymal transition is a regulated process induced by a signal produced by the adjacent myocardium of the atrioventricular canal (Runyan and Markwald, 1983). Examination of micrographs suggests that all of the endothelial cells lining the atrioventricular canal can undergo cellular hypertrophy and loss of cell–cell adhesions (Markwald et al., 1977). Measurement of a calcium flux within the endothelium of the atrioventricular canal supports common activation by all of the endothelia

A

V

Ventricle explant

AV Canal explant

Remove myocardium No EMT Time in culture

No EMT

EMT

EMT with Myo-conditioned medium or sufficient factors

Figure 4  Summary of the collagen gel assay. Tissue explants derived from the atrioventricular canal and ventricular portions of the pre-epithelial– mesenchymal transition are explanted onto the surface of collagen gels and cultured. Ventricle tissue, when cleanly-dissected from AV canal or outflow tract produces no mesenchymal cells. Atrioventricular canal explants recapitulate epithelial–mesenchymal transition in vitro. Experiments published in Runyan and Markwald (1983) demonstrated a requirement for a myocardially-derived signal to induce endothelial epithelial–mesenchymal transition.

369

Chapter  |  6.1  Molecular Regulation of Cushion Morphogenesis

(Runyan et al., 1990). Importantly, only a subset of activated endocardium executes the full program of epithelial– mesenchymal transition to become cushion mesenchyme. Markwald and colleagues (Markwald et al., 1984) showed the polarized expression of heparan sulfate proteoglycans to the basolateral membrane of endothelial cells just prior to invasion, suggestive of the transitioning into a mesenchyme phenotype. The loss of PECAM and VE-cadherin are also required for the endothelial to fully activate and move into the cushion matrix (Baldwin et al., 1994; Enciso et al., 2003; Liebner et al., 2004). The images of Kinsella and Fitzharris (1980) demonstrate a very active probing of the underlying extracellular matrix by long filopodia from the de novo mesenchyme. These newly-formed mesenchymal cells migrate away from the endocardial lining and invade the extracellular matrix (cardiac jelly) of the prominent cushion tissue. These cells serve as the progenitors for the constituent cells of the mitral and tricuspid valves and membranous

portion of the ventricular septum. The invasion process is mediated by matrix metalloproteinases (Song et al., 2000), extracellular matrix receptors (Loeber and Runyan, 1990), and is accompanied by the deposition and removal of extracellular matrix molecules (Sinning et al., 1988; Wunsch et al., 1994; Bouchey et al., 1996; Klewer et al., 1998). In some instances, mesenchymal cells have been observed to follow the tracts of leading cells through the extracellular matrix. Collectively, these observations indicate that the morphological events of epithelial– mesenchymal transition follow a semi-linear progression of transition. There is cell-to-cell separation from an epithelial arrangement. Single and rounded cells elongate and extend filopodia. The transitioned cells invade and migrate into the extracellular matrix of the cardiac jelly or into type I collagen-based matrix in vitro. A large number of regulatory pathways have been implicated in epithelial– mesenchymal transition, either by expression and/or functional studies (Figs 5; 6). We highlight only a few select

DSCR1 CD31 Prolif

EMT

VE-Cad TGFß2 EMT

CD31

Snail/Slug CA2+

Cx45

NOTCH1 BMP4

NFATc1

DSCR1

Endo

ß-Cat

ER

Calcn

Smad 6 ?

IP3

Smad 4

Smad 1/5

Smad 2/3

DLL4

Ras

BMP-RI/RII

Flk-1

ErbB2/3

Jag1

NOTCH1

VEGFE

Migration EMT

TGFß-RII/RIII

BMP4

Col I, VI

Wnt

Fibulins I, II Fibrillins

HA

VEGFM

TGFß2

BMP-2

TGFß2

Versican

Myoc MSX

VEGF NFATc1

Has2

Hhex UDP-GD

+ Ca2 /Calcn

Figure 5  Molecular regulation of epithelial–mesenchymal transition in mouse. Studies of epithelial–mesenchymal transition in the mouse model have identified a wide number of molecules and pathways involved in the regulation of epithelial–mesenchymal transition in the atrioventricular canal. Signal transduction pathways of note include TGF2, BMP2, VEGF, Hyaluronan (HA) and other molecules secreted by the myocardium and additional pathways identified in the endothelium including Notch and Wnt. The diagram here attempts to place known mediators in context. The most obvious difference, to date, between the mouse and chick model systems appears to be the later expression of TGF3 in the mouse (after epithelial–mesenchymal transition) and the expression of TGF1 in the mouse, but not in the chick.

370

PART  |  6  Cushions, Valves and Septa

EC Prolif Apoptosis Prolif

NFATc1 ERK 1/2

Smad 2/3 Calcn

PLCε

DSCR1

Smad 6

Rac1

+ PAI-1

Smad 4

+

CA2 IP3

Raf

ER

Smad 1/5

Nf1

shp2 Ras

Flk-1

Id1

PI-3K

Erb2/3

+

GIPC + ALK5

Ras

BMP-R1/RII

EGFR

ALK1,2

TBRII/RIII

EGF

VEGFE

TACE HB-EGF

Neur-1

IGF1

BMP2/4

TGFß2

Fibulin-1

Figure 6  Molecular regulation of post-cushion valvulogenesis. Numerous studies have identified molecules critical to normal valvulogenesis, but either not involved or distinguishable from roles in epithelial–mesenchymal transition. These molecules have been diagrammed as part of the signal transduction processes that regulate remodeling and maturation within the cardiac cushions.

pathways in this section as examples of regulatory circuits governing epithelial–mesenchymal transition during endocardial cushion morphogenesis (see also Chapter 6.2).

II.D.  TGFs as Mediators of Epithelial– Mesenchymal Transition Substantial evidence has been obtained from both the chick and mouse heart developmental systems that show TGF ligands, receptors and signaling mediate epithelial– mesenchymal transition during endocardial cushion morphogenesis. Early studies in the avian heart by Runyan and colleagues (Potts and Runyan, 1989; Potts et al., 1991, 1992; Runyan et al., 1992) first identified TGF as an inductive agent mediating cardiac cushion epithelial– mesenchymal transition. These studies using endocardial cushions from avian embryos in collagen gel assays demonstrated a requirement for TGF to form valvular mesenchyme (Potts and Runyan, 1989). Subsequent work showed isoform-specific roles in the chick for two TGF isoforms, TGF2 and TGF3, in this process (Potts et al., 1991; Boyer et al., 1999). Loss of TGF2 activity with a blocking

antibody inhibits separation and activation of atrioventricular canal endothelia in vitro, while treatment of collagen gel cultures with a blocking antibody against TGF3 prevents cell invasion into the collagen matrix (Boyer et al., 1999). These chick studies stimulated the Doetschman laboratory to examine heart development in knockout mice produced for each of the TGF isoforms (Sanford et al., 1997; Azhar et al., 2003). They observed that only isoform-specific TGF2-null mice have heart defects (Sanford et al., 1997). TGF1 or TGF3 deficient mice do not have overt heart defects (Shull et al., 1992; Dickson et al., 1993; Kaartinen et al., 1995). TGF2-deficient mice have variable penetrance for heart defects including double-outlet right ventricle (DORV), atrial and ventricular septal defects, double-inlet left ventricle and malformed cardiac cushions (Sanford et al., 1997; Bartram et al., 2001b). However, a surprising aspect to these studies was that the TGF2-null mice have hyperplastic cushions rather than a failure in mesenchyme formation. Collagen gel assays using endocardial cushions from TGF2-deficient embryos showed that the loss of TGF2 signaling results in a reduction of mesenchyme formation (Azhar et al., 2009). Thus, in the complex in vivo environment, TGF2-deficient embryos

Chapter  |  6.1  Molecular Regulation of Cushion Morphogenesis

may exhibit hyperproliferative cushions, due to elevation in TFG1 and TGF3 isoforms causing continuous proliferation during post-epithelial–mesenchymal transition remodeling. It will be important to analyze endocardial cushion morphogenesis in compound TGF-deficient lines to determine whether there is functional compensation.

II.D.i.  TGF Receptor Activity in Mediating Arterioventricular Canal Epithelial– Mesenchymal Transition The general paradigm of TGF signal transduction is nicely summarized by Wrana (Wrana et al., 1994). TGF isoforms bind to TGF type II receptors (TRII) that are serine-threonine kinases. This binding promotes association with a type I receptor (TRI) and the type II receptor transphosphorylates the type I receptor. The phosphorylated type I receptor is then able to initiate signaling through interactions with SMAD proteins or several other downstream effectors including Rho and TAK1 (Hoffmann et al., 2005). While there is a single type II receptor, several type I receptors, referred to as activin receptor-like kinases (Alk), have been identified. Three Alks, (Alk1, Alk2 and Alk5) are known to associate with the type II receptor and mediate TGF signaling (Tendijke et al., 1994; Lai et al., 2000). Additional TGF receptors include the type III receptor (TRIII, betaglycan), and a related molecule, endoglin. The activities of these two receptors are less well-defined. Betaglycan has been demonstrated to enhance the affinity of TGF2 for TRII (Lopez-Casillas et al., 1991) and is required for the high affinity binding of TGF2. Endoglin, interacting with TRII, appears to bind only TGF1 and TGF3 (Lopez-Casillas et al., 1994). Using a blocking antibody against the avian TRII, Brown et al. (1996) showed that this receptor was required for epithelial–mesenchymal transition in chick atrioventricular canal cushion explants. Jiao et al. (2006) recently deleted the gene for TRII in mouse endothelium, and showed that epithelial–mesenchymal transition occurs normally in vivo but that cushion remodeling is defective. Further, endocardial explants from TRII-deficient embryos revealed a loss of epithelial–mesenchymal transition (Jiao et al., 2006). These data are similar to that obtained by comparing loss of TGF2 in vitro and in vivo. Overall, these data suggest that TGF signaling is required for epithelial–mesenchymal transition in the more simple and controlled setting of the in vitro collagen gel invasion assay. In vivo, the environment of a more complex extracellular matrix and presence of other endogenous growth factors may circumvent the loss of TGF and permit mesenchyme production, but remodeling of the cellular cushions is impaired. Therefore, TGF may play several roles during valvulogenesis – that of facilitating the differentiation into mesenchyme, and equally importantly supporting the remodeling phase to create mature valves.

371

Similarly, a neutralizing antibody to TRIII blocked epithelial–mesenchymal transition, while forced expression of TRIII in ventricle endothelium caused epithelial– mesenchymal transition in a region where epithelial– mesenchymal transition does not normally take place (Brown et al., 1999). Correlation of the progression of epithelial–mesenchymal transition (i.e., activation, separation and invasion) with specific interventions lead to the suggestion that TGF2 activity is mediated in avian cushion development by TRIII and TGF3 activity by TRII (Boyer and Runyan, 2001). These data suggested a unique and nonredundant role for TBRIII receptors in TGF signaling. Further, TRIII is encoded in a region on chromosome 1 linked to endocardial cushion defects in humans (Sheffield et al., 1997). Collectively, these data indicate that TRIII activity has a substantial contribution to cushion and valve formation, a theory which is being validated by further experimentation. Studies by Barnett and colleagues (Lai et al., 2000; Desgrosellier et al., 2005) showed that the type I receptor, Alk2, is required and sufficient for epithelial–mesenchymal transition in avian explant cultures, and suggested that Alk5 is not sufficient. Alk2 is similarly required in mouse endocardial cushion epithelial–mesenchymal transition (Wang et al., 2005a). The evidence to suggest that Alk5 was not involved was obtained using an antibody against the extracellular domain of Alk5. Although proven to be effective in perturbing gene regulation in an assay system, it failed to block epithelial–mesenchymal transition. Further, these investigators found that transfection of ventricular endothelial explants with a constitutively active form of Alk5 failed to induce epithelial–mesenchymal transition. In contrast, recent work by Mercado-Pimentel et al. (2007) shows that epithelial–mesenchymal transition in vitro and expression of several markers of epithelial– mesenchymal transition is inhibited by antisense DNA and siRNA targeted towards Alk5. A kinase inhibitor designed to block Alk5 activity also inhibits epithelial–mesenchymal transition in atrioventricular canal cushion explants in the same study. Phenotypic examination of treated cushionendothelial cells suggests a role in the activation/separation step of epithelial–mesenchymal transition by Alk5. Goumans et al. (2002) showed that a small amount of Alk5 activity is required for Alk1 function, which may also be the case for Alk2. Overall these data demonstrate that Alk5 is not sufficient to induce epithelial–mesenchymal transition, but is required for full mesenchyme production. Endoglin is a type III-related receptor that is widely, but not exclusively, associated with endothelial cells. Mutations in either Alk1 or endoglin are associated with forms of human hereditary hemorrhagic telangiectasia (HHT2, Alk1 mutations, OMIM#600376) (HHT-1, OMIM #187300) (McAllister et al., 1994; Johnson et al., 1996). HHT syndrome has many arterio–venous malformations and vascular dysplasia. Sorensen et al. (2003) showed that

372

mouse embryos lacking Endoglin fail to undergo normal mesenchymal cell formation in vivo, but that endocardial cushion explants from homozygote embryos show normal numbers of mesenchymal cells in culture. Antisense DNA treatment to Endoglin in chick explant cultures showed inhibition of epithelial–mesenchymal transition at one day in culture, but a recovery by 48 hours. This temporal recovery suggests the presence of redundant mechanisms, or adaptation to the collagen gel matrix to produce the mesenchyme, as was seen in the type II receptor mutant mouse (Jiao et al., 2006). Thus, endoglin may be a required accessory receptor to fully engage the TGF signaling program during epithelial–mesenchymal transition. The requirements of downstream mediators of TGF signaling in the atrioventricular canal are less clear. Detailed biochemical studies established that Smad2 and Smad3 appear specific for TGF and activin signals, while Smad1, Smad5 and Smad8 are specific to BMP signaling. Smad4 appears to be the common Smad in both signaling pathways. Smad6 and Smad7 act as inhibitory mediators of TGF/BMP signal transduction. Defects in atrioventricular canal development are present in both Smad6- and Smad7-deficient mice (Galvin et al., 2000; Liu et al., 2007; Chen et al., 2009). Smad6/ embryos have hyperplastic cardiac valves and associated defects in septation. Smad7DMH2/ mice display increased apoptosis in the atrioventricular canal, and both atrial and ventricular septal defects. Both Smad6 and Smad7 molecules are inhibitory Smads, implying that normal receptor-Smad signaling is critical to atrioventricular canal development. Although Smad4 likely plays an essential role in cardiac cushion formation, embryos deficient for Smad4 die very early, due to severe defects in mesoderm formation (Sirard et al., 1998) similar to the phenotype of Bmp4-null embryos (Winnier et al., 1995; Chapter 6.2). The Smad3-null mouse has normal embryonic development, but displays epithelial–mesenchymal transition-related pathologies in postnatal life (Roberts et al., 2006). Smad5-deficient embryos have defective vasculogenesis retarding normal development of the heart (Chang et al., 1999). Similar to Smad5/ lethality, other Smad mutants are lethal prior to atrioventricular canal development (Dunker and Krieglstein, 2000). Snail1 and Slug (Snail2) are both downstream effectors of TGF signaling (Romano and Runyan, 2000; Wang et al., 2005a). Inactivating Snail2 attenuates epithelial–mesenchymal transition in the primitive streak and in the neural crest, indicating its activity is critical to developmental epithelial–mesenchymal transition (Nieto et al., 1994). Importantly, these transcription factors can act to downregulate the expression of cadherins, including VE-cadherin, a requisite activity to allow migration of endocardial cells during epithelial–mesenchymal transition. In support of this, Snail2 expression in the chick model can promote the loss of VE-cadherin and PECAM in the endothelium

PART  |  6  Cushions, Valves and Septa

(Baldwin et al., 1994; Romano and Runyan, 1999, 2000; Mercado-Pimentel et al., 2007). Loss of VE-cadherin and PECAM is required to allow cushion endothelial cells to separate and execute the full epithelial–mesenchymal transition process. The subsequent invasion event of epithelial– mesenchymal transition in the atrioventricular canal is distinctly regulated in the chick by TGF3 signaling and, likely, by BMP2/4 in the mouse (Boyer et al., 1999; Sugi et al., 2004). In the chick, TGF3 regulates several markers of the epithelial–mesenchymal transition process, and loss of this TGF isoform blocks activated cells from entering the matrix of collagen gels (Boyer et al., 1999). There appears to be some endogenous specificity of TGF receptors to match the isoform-specific effects (Boyer and Runyan, 2001). Rho and Rho kinase activities are involved in the reorganization of the cytoskeleton required for epithelial–mesenchymal transition (Hall, 1998; Kaartinen et al., 2002). Recent work confirms the TGF-mediated regulation of RhoA and requirements of RhoA and Rho kinase in cardiac epithelial–mesenchymal transition (Sakabe et al., 2006; Tavares et al., 2006). Integration of the morphological processes of epithelial–mesenchymal transition with signaling and transcriptional regulation remains an active area of investigation.

II.E.  BMP Signaling in Epithelial– Mesenchymal Transition Cushion-forming epithelial–mesenchymal transition also correlates spatially and temporally with the expression of bone morphogenetic protein-2 (BMP-2) and 2a (BMP-4) in the atrioventricular myocardium of the mouse (Lyons et al., 1990; Jones et al., 1991; Yamada et al., 1999; Yamagishi et al., 1999; Somi et al., 2004; Sugi et al., 2004). BMPs are a subgroup of the transforming growth factor (TGF) superfamily that is widely implicated in embryonic development (Francis-West et al., 1994; Wall and Hogan, 1995; Hogan, 1996). However, early lethality of conventional Bmp2 knockout mice (Zhang and Bradley, 1996) significantly hampered understanding of the role of BMPs in endocardial cushion epithelial–mesenchymal transition. Subsequent studies using in vitro culture assays (Sugi et al., 2004) and conditional deletion of BMP-2 indicated that myocardially-derived BMP2 is essential for atrioventricular canal endocardial transition into cushion mesenchyme (Ma et al., 2005; Rivera-Feliciano and Tabin, 2006). In these myocardial-Bmp2-deficient embryos, both TGF2 and Has2 are significantly reduced in the heart, indicating BMP may be upstream of these epithelial–mesenchymal transition regulators (Ma et al., 2005). BMPs exert their biological function by interacting with cell surface receptors similar to those utilized by TGFs as described above (Hogan, 1996: Yamashita et al., 1996; de Caestecker,

Chapter  |  6.1  Molecular Regulation of Cushion Morphogenesis

2004). While there is a specific type II receptor for BMP, BMP type I receptors include activin-receptor-like kinase1 (Alk1) (also named TTSR-I), Alk2, Alk3 (BMPR-1A) and Alk6 (BMPR-1B) (Hogan, 1996; Yamashita et al., 1996; Yi et al., 2000). Alk3 (BMPR1-A) and Alk6 (BMPR1-B) bind specifically with BMP-2, BMP-4 and with BMP-7 at a low affinity. Similar to conditional Bmp2 knockout mice, cardiac myocyte Alk3-deficient embryos have decreased expression of TGF2 (Gaussin et al., 2002). These embryos have hypoplastic cardiac cushions that may be a result of decreased TGF2 activity, or via direct regulation of molecules such as Has-2. Inhibition of Bmp receptor signaling in chick cushion explants by Noggin attenuates epithelial– mesenchymal transition (Okagawa et al., 2007). The role of Alk2 during EMT may be the missing link for some redundancy between TGF and BMP signaling in this process. As little is understood about the formation of signaling complexes between TGF and BMP receptors, apparent conflicts in the data between separate studies may be due to active and negative activities from receptors in unusual signal configurations. For example, phosphorylation of Alk5 by TBRIII causes growth arrest (Wrana et al., 1994). Alternatively, if it can be shown that BMP2 or BMP7 can interact with TBRIII receptors, this might explain some of the experimental data with regard to complexes that drive the epithelial–mesenchymal transition process. Thus, the pleiotrophic effects of TGF signaling may be a result of context-specific selective interactions between Alks and the three types of TGF receptor.

II.F.  Notch as a Regulator of Epithelial– Mesenchymal Transition The Notch pathway was first described in Drosophila (Portin, 2002). Mutations in Notch produced a characteristic notched phenotype of the wing. Analysis of the mutant revealed a receptor ligand interaction between adjacent cells using the ligand Delta and the receptor Notch. Binding of a ligand to Notch produces a response wherein a protease complex containing gamma secretase cleaves the intracellular portion of Notch. The intracellular domain of Notch (NICD) enters the nucleus and acts in a transcription complex containing the protein RBPJk (also known as CSL). Interaction between NICD and RBPJk alters the inhibitory function of RBPJk, and the complex becomes a transcriptional activator (Lai, 2004). One family of transcription factors, the hairy/enhancer of split family identified as Hey-1 or Hey-2 (alternatively Hes or Hert) are closely, but not exclusively, associated with Notch regulation. Hey-1 or Hey-2 are largely inhibitory, but can also activate transcription (Sakata et al., 2006). Notch2 expression appears to be regionally restricted within the heart endothelium and mesenchyme near the

373

time of epithelial–mesenchymal transition (Loomes et al., 2002; Varadkar et al., 2008). Notch1 is clearly expressed throughout the endothelium of the heart, but as Notch ligands, Jagged1 and Delta-like 1 and 4 are expressed in the myocardium, a low level of Notch1 expression in the myocardium seems likely (Loomes et al., 1999, 2002; Timmerman et al., 2004). Normal ventricular trabeculation requires expression of Notch1 in the ventricular endocardium (Grego-Bessa et al., 2007; Chapter 6.2). There are a series of mouse strains with perturbations of the Notch signal transduction pathway resulting in severe heart valve defects, including Notch1 and Notch2 mutants (McCright et al., 2001; Noseda et al., 2004; Timmerman et al., 2004; Watanabe et al., 2006; Chapter 6.2). A recent investigation linked bicuspid aortic valves with Notch1 mutations (Garg et al., 2005; Garg, 2006). Both Jagged1 and Notch2 are linked to Alagille Syndrome, a disorder that includes cardiovascular defects, as well as liver and pulmonary vessel defects (OMIM #118450 and #610205). While mice mutant for each of these molecules share some of the phenotypic characteristics of Alagille Syndrome, double heterozygotes appear to provide a better model for the disease (McCright et al., 2002). Timmerman et al. (2004) argued that Notch1 signaling interacted with TGF signal transduction pathways to alter invasion during epithelial–mesenchymal transition and expression of the gene for the transcription factor Snail (Snai1) as a marker of activation. Niessen et al. (2008) showed evidence to suggest that Notch1 and TGF separately regulate Snail family members, and that there is some redundancy in the activity of Snail1 and Snail2 in the mouse. Snail family members decrease the expression of cell adhesion molecules such as cadherins, thereby facilitating early migratory events and subsequent invasion into the cardiac jelly (Romano and Runyan, 2000). It is clear from the phenotype of the Notch1 mutant that Notch signaling regulates myocardial development and also specifies the endocardial cushions to execute epithelial–mesenchymal transition (Timmerman et al., 2004). Although the integration of Notch and TGF signaling pathways is not well-understood, there is evidence that the Notch intracellular domain can interact with Smads (Sun et al., 2005). Thus, evidence exists to suggest that Notch signaling may act to make endocardial cells competent to respond to a TGF signal, intersect with TGF signals intracellularly or mediate epithelial–mesenchymal transition by an independent signal transduction pathway. In addition to the potential contribution of defective epithelial–mesenchymal transition to the Alagille phenotype, the Hey2-null mouse has membranous ventricular septal defects consistent with a loss of mesenchymal cells through either epithelial–mesenchymal transition or post-epithelial–mesenchymal transition cushion remodeling (Sakata et al., 2006). Defining the dynamic associations between the Notch and TGF pathways will provide valuable insight into integrated signaling networks during endocardial cushion and heart valve formation.

374

II.G.  Receptor Tyrosine Kinase and Ras-MAPK Signaling During Epithelial– Mesenchymal Transition Several lines of evidence implicate ErbB receptor signaling during endocardial cushion morphogenesis and cardiac valve maturation. Ligand-induced homo- or heterodimerization of the receptors results in tyrosine kinase activation and transphosphorylation of tyrosine residues in the cytoplasmic domain of the dimerized ErbB partners. This leads to the recruitment of a variety of effector molecules, including Src, phosphatidylinositol 3-kinase, Shc, phospholipase C, signal transducer and activator of transcription, Grb2/SOS and others. The activation of many of these proteins results in the phosphorylation of nuclear translocating kinases, including stress-activated protein kinase/c-Jun N-terminal kinase (JNK) and the mitogen-activated protein (MAP) kinases, p38 and extracellular signal regulated kinase (ERK) 1/2. One mechanism of interest to cardiac development is the activation of MAP kinases through the recruitment of the Grb2/ SOS complex and Shp2 to the phosphorylated receptor, resulting in Ras activation and phosphorylation of Raf (a MAP3K), MAP/ERK activating kinase and ERK1/2. On activation, ERK 1/2 can translocate to the nucleus and induce transcription of a variety of genes involved in mitogenesis, differentiation, apoptosis and proliferation (Yarden and Sliwkowski, 2001). Components of this RTK pathway have been linked to cushion development both in vitro and in vivo. Heregulin (HRG) or neuregulin has restricted expression to the cardiac endocardium. ErbB2 mRNA and protein seem to be more commonly expressed in embryonic heart tissues. The prototypic receptor of this family, epidermal growth factor receptor (EGFR; ErbB1) has a more global presence in embryonic heart valve tissues. ErbB4 receptors appear to be more restricted to the myocardium. In contrast, at the time coincident with atrioventricular canal mesenchyme production, ErbB3 appears restricted to endocardium and transitioned mesenchymal cells. Thus, these restricted expression patterns during heart morphogenesis are concordant with distinct functions during muscle and valve formation. In this regard, gene-targeting studies revealed that embryos lacking ErbB3 (erbb3/) succumb to cardiovascular malformations, including those arising from endocardial cushion defects. These cardiac cushion abnormalities result in blood reflux through the defective valve tissue contributing to the lethality of erbb3/ embryos by E13.5. In contrast, ErbB4-deficient embryos do not have valve defects, suggesting that it is not essential for valvulogenesis. The hypomorphic EGFR model, waved-2, not only have wavy hair, but also exhibit hyperplastic outflow tract valves which is more penetrant on a Shp2/ background (Chen et al., 2000). These waved-2 mice only have 10–15% normal EGFR kinase activity due to a

PART  |  6  Cushions, Valves and Septa

point mutation (TA) in the Egfr gene (Luetteke et al., 1994), which suggests that the normal catalytic activity of EGFR is required for regulating epithelial–mesenchymal transition. In this regard, closer inspection of conventional Egfr/ knockout mice also shows valve defects in both the atrioventricular canal and outflow tract regions (Jackson et al., 2003). Although not as pronounced as the hypoplastic erbb3/ cushions, embryos deficient for erbB2 (erbb2/) or neuregulin/HRG (Nrg1/) also exhibit underdeveloped cushion-valve tissue, but most likely die at E10.5 due to failed myocyte differentiation and contractile deficiencies (Negro et al., 2004). An additional mouse line lacking only the kinase domain of ErbB2 supports the requirement for kinase activity of ErbB2 during cardiovascular development (Negro et al., 2004). Furthermore, inactivation of the serotonin (5-hydroxytryptamine) receptor 2B gene (Htr2B) results in midgestation lethality from heart defects in part due to downregulation of erbb2 (Nebigil and Maroteaux, 2001). More recently, gene-targeting of the gene for HBEGF (Hbegf ) demonstrated its obligate role in cushion and valve formation with homozygote animals dying prematurely, in part due to valve defects. Specifically, mice lacking HB-EGF display enlarged myxomatous cushion valve structures (Jackson et al., 2003). Hbegf / mice appear to have global cushion defects, since enlarged valves are not restricted to either the inlet or outlet valves (Iwamoto et al., 2003). Mice which are deficient for the protease (ADAM17 or TACE) which cleaves proHB-EGF into its active form phenocopy Hbegf / mice (Jackson et al., 2003). Interestingly, there appears to be elevated TGF signaling in the cardiac cushions of Hbegf / mice, indicating integration among these growth factor pathways during valvulogenesis. These observations also suggest that valve defects in Hbegf /, Adam17/ and Egfr/ embryos might be remodeling deficiencies in addition to unregulated epithelial–mesenchymal transition. This also emphasizes the need for experimental studies to assess the functional mechanisms of valve maturation following epithelial–mesenchymal transition (see section below on postepithelial–mesenchymal transition remodeling). A key effector of ErbB receptor signaling is Ras GTPase (Yarden and Sliwkowski, 2001; Camenisch et al., 2002b). Initial studies by Lakkis and Epstein (1998) showed that unregulated Ras activity in the knockout mouse for the neurofibromin gene causes hyperproliferative cardiac cushions in homozygote embryos. In addition, dominant-negative Ras in vitro drastically attenuates the elevated epithelial–mesenchymal transition caused by the absence of this Ras-specific GAP. Most striking is the ability of constitutively active Ras (caRas) to drive production of mesenchyme in endocardial explants derived from the ventricle (Camenisch et al., 2000). Remarkably, caRas is also able to rescue defective epithelial–mesenchymal transition in cushion explants derived from Has2/ embryos

Chapter  |  6.1  Molecular Regulation of Cushion Morphogenesis

(Camenisch et al., 2000). This latter observation indicates that if enough Ras effector signaling is achieved, upstream deficits in epithelial–mesenchymal transition programing such as that in Has2/ deficiency can be circumvented, restoring production of mesenchyme. The phosphatase, Shp2, is another key effector that propagates the RTK signal to MAP Kinases (Yarden and Sliwkowski, 2001). Fifty percent of Noonan syndrome cases are a result of germline mutations in the gene Ptpn11 encoding this tyrosine phosphatase, or SHP2 (Nakamura et al., 2006). A majority of the missense mutations are clustered to the interacting domain between the N-terminal SH2 domain and the phosphotyrosine phosphatase (PTP) domain. When the N-SH2 domain is interacting with the PTP, SHP2 is in a closed state or inactive. Thus, this is an autoregulatory interaction to control SHP2 activity. These defined mutations in the Ptpn11 gene result in sustained release of this inhibitory interaction, giving the PTP domain an open confirmation with elevated basal and inducible phosphatase activity. Thus, the SHP2 protein is “overactive”, resulting in heightened signaling typically downstream of receptor tyrosine kinases. The incidence of Noonan syndrome is 1:1000 to 1:2500 births (~7–10% of CHD) (Pierpont et al., 2007), and is characterized by variable penetrance (see also Chapter 6.2). Cranial–facial abnormalities, heart defects and myeloproliferative disorders are the key observed phenotypes of the syndrome. The significant heart defect most commonly associated with Noonan Syndrome is pulmonary valve stenosis. One linked mutation is a D61G change that results in a “gain-of-function” SHP2. This mutation has been successful modeled by a knockin strategy, where SHP2D61G/ mice mimic the valvular stenosis and mild myeloproliferative disorder observed in Noonan Syndrome patients (Araki et al., 2004). In this model, overactive Shp2 dramatically increases ERK activation by sevenfold, which likely contributes to the hyperproliferative valve phenotype. Additional effectors of RTK signaling have been implicated in the epithelial–mesenchymal transition process including Ras, Mek5, Map3Kinases (Raf, MEKK3, MEKK4) (Lakkis and Epstein, 1998; Camenisch et al., 2002b; Wang et al., 2005b; Stevens et al., 2006; Pandit et al., 2007; Razzaque et al., 2007 and Stevens et al., 2008) which appear to culminate in regulation of MAP kinases including ERK1/2, ERK5, p38 and JNK (Johnson et al., 2005). Importantly, germline mutations in K-RAS have been linked to Noonan Syndrome cases in humans (Schubbert et al., 2006). Collectively, these studies further substantiate this RTK-RAS-MAPK signaling pathway in executing epithelial–mesenchymal transition in cardiac cushion cells, as well as subsequent events in valvulogenesis. Additional investigations will further decipher this complex network of signaling mediators and exactly how they are all choreographed during epithelial–mesenchymal transition and remodeling of the valves.

375

II.H.  Calcium and VEGF as Regulators of Epithelial–Mesenchymal Transition Calcium (Ca2) serves as a critical second messenger in many systems, including heart development and function. In fact, disrupted heart development is observed in mouse embryos treated with calcium channel blockers (Porter et al., 2003). For example, Nifedipine exposure caused attenuated expression of GATA4, a key transcription factor related to cardiac cushion formation (Porter et al., 2003). Thus, any perturbations in calcium homeostasis in the cardiac forming field can result in changes in the programing required for morphological formation of the functioning heart. In this regard, an intriguing question 10 years ago was: “Why is endocardial epithelial–mesenchymal transition restricted to the cardiac cushion segments?” Several studies now have revealed that a Ca2 mechanism controls VEGF expression and its ability to limit the extent of epithelial–mesenchymal transition to the endocardial cushions, and not other segments in the forming heart. The interplay between inducing and inhibiting factors at the junction of the atrioventricular canal and ventricle is likely to prove interesting. Several observations emphasize the importance of the strict temporal and spatial control of VEGF to the ventricular and atrial segments, and not to the cushion regions during developmental epithelial–mesenchymal transition. Dependence on angiogenesis during heart development is demonstrated by the clinical association between congenital heart defects and gestational hypoxia (DeSesso, 1987; Lueder et al., 1995). VEGF is known to act in restoring hypoxic mature tissue to normal oxygen homeostasis by inducing the generation of new blood vessels (Dor et al., 2001). Strict control of VEGF levels is required to control the angiogenic response and prevent deleterious effects (Lee et al., 2000; Isner et al., 2001). The importance of the appropriate timing and dosage of VEGF during development is highlighted by shared cardiovascular developmental defects in several independent VEGF mouse systems. In a transgenic model, a 2–3-fold increase in endogenous VEGF production results in midgestation lethality, due to overdeveloped trebeculae and abnormalities in coronary vessels and septation (Miquerol et al., 2000). Conversely, the loss of one VEGF allele results in early embryonic lethality due to cardiovascular defects (Carmeliet et al., 1996; Ferrara et al., 1996). This haploinsufficient phenotype exhibits underdeveloped endocardial cushions and chamber malformations, in addition to impaired vascular development. In a controlled transgenic system, it was shown that production of VEGF a full day earlier than normal (E9.5 versus E10.5) results in septal and valve defects arising from malformed endocardial cushion tissues (Dor et al., 2001). This finding is consistent with other work defining the expression of VEGF in the endocardial cushions and demonstrating it as a negative regulator

376

of epithelial–mesenchymal transition (Dor et al., 2003). These findings, demonstrating a dependence on appropriate VEGF production for normal valve formation and chamber septation, appear to correlate with septal defects observed from hypoxia or exposure to air pollutants during early gestation (Miao et al., 1988; Ritz et al., 2002). These results suggest that precocious increases in VEGF may prematurely terminate epithelial-to-mesenchymal transition in the endocardial cushions, and contribute to chamber septation and valvular defects. A connection between VEGF expression and calcium is based on the regulation of the transcription factor NFATc (Johnson et al., 2003). Recently, Chang et al. (2004) showed that VEGF is repressed by a calcineurin/ NFATc mechanism to prevent VEGF production in the myocardium of the atrioventricular canal cushions until E11 or after the epithelial–mesenchymal transition process. Specifically, Ca2-dependent translocation of NFATc to the nucleus in myocytes regulates VEGF expression. Down syndrome critical region 1 (or MCIP) may compete with NFATc for binding to calcineurin which dephosphor­ ylates NFATc, allowing it to translocate into the nucleus. Thus, a complicated pathway is emerging for the regulation of VEGF during heart development whereby MCIP interacts with Ca2-dependent calcineurin attenuating the dephosphorylation of NFATc and its ability to get into the nucleus for suppressing VEGF expression. It is speculative, but the unregulated overexpression of VEGF, a potent inhibitor of epithelial–mesenchymal transition, may in part be responsible for the paucity of valve tissue in Trisomy 21 patients. This makes an attractive target for molecular medicine with regard to controlling the outgrowth of valve tissue in this select patient population.

III.  Post-epithelial–mesenchymal transition remodeling of cushions into valvuloseptal structures Mature heart valves consist of fibrous connective tissue surrounded by endocardial endothelial cells. The connective tissue cells include fibroblasts, myofibroblasts and smooth muscle cells that are collectively referred to as valve interstitial cells (Mulholland and Gotlieb, 1996; Butcher and Markwald, 2007). Constituent molecules of valve extracellular matrix include: collagens (Cole et al., 1984; Filip et al., 1986; Bashey et al., 1992; Miosge et al., 1998); elastic fibers; fibulins; proteoglycans; glycosaminoglycans (GAGs); and glycoproteins (Honda et al., 1975). The endocardial cushions become fully-defined developmentally as valves when they acquire “free” and “fixed” components. The free (flexible) portion of a mature leaflet is composed of lamellae of extracellular matrix and aligned with fibroblastic-like cells. The fixed component of

PART  |  6  Cushions, Valves and Septa

the valve is called the annulus, or cardiac skeleton, which runs circumferentially around the atrioventricular canal or outflow tract valves and anchors the free leaflet at its base (Icardo and Colvee, 1995). Atrioventricular valves also have a tendinous supporting apparatus to govern tension and movements of the leaflets during the cardiac cycle. The origin of the supportive apparatus is still debated, and the mechanism of its formation has not been fully described (Oosthoek et al., 1998; Lincoln et al., 2004). Adding complexity to valvulogenesis are reports that extra­ cardiac cells, neural crest (in outflow tract cushions) and epicardial derived cells (EPDC) (in atrioventricular cushions) invade and intermingle with cells of endocardial origin (Poelmann et al., 2002). While the role of neural crest cells in outflow tract valvulogenesis remains unclear, it is not so for EPDC. If the formation and/or migration of EPDC into the atrioventricular canal cushions are inhibited, the cushions become misshapen and hyperplastic (Poelmann et al., 2002). Additionally, the fibrous skeleton is discontinuous or absent as atrioventricular junctional myocardium abnormally persist, resulting in persistence of functional atrioventricular accessory conduction pathways (Kolditz et al., 2007). To become a mature valve, the primitive mesenchymalized cushions must be remodeled, a multi-step process that continues into postnatal life as shown in Fig. 2.

III.A.  Proliferation and Elongation The continued proliferation and expansion of the undifferentiated (post-epithelial–mesenchymal transition) mesenchyme towards the lumen is initiated by signaling from the endocardium. The rationale for this assumption is from observations with the NFATc-3 mouse. As shown by Baldwin and colleagues (Ranger et al., 1998), NFATc encodes a transcription factor expressed specifically in atrio­ ventricular canal and outflow tract endocardium, but not in the cushion mesenchyme. Knocking-out this gene did not stop endocardial transformation into new cushion mesenchyme, but it did stop cushion expansion and elongation into a normally sized leaflet, resulting in lethality at E14. On some levels, limb bud and cushion elongation appear similar. There is distal elongation of a mesenchymal core covered by an epithelium. In the limb, it is well-known that a special region of the ectodermal epithelium, the apical ectodermal ridge (AER), signals the subjacent mesoderm (the “progressive zone”) to proliferate, resulting in the distal (linear) elongation of the limb bud (Fallon et al., 1994). The signals secreted by the AER to promote proliferative expansion are fibroblast growth factors, especially FGF4 and FGF8 (Fallon et al., 1994; Lewandoski et al., 2000). In cardiac cushions, there is no AER equivalent, but there is a rim of endocardial cells enclosing cushion mesenchyme that becomes thickened and stratified over time (Markwald et al., 1998). Clusters of BrdU-proliferating

Chapter  |  6.1  Molecular Regulation of Cushion Morphogenesis

cushion mesenchymal cells can be found beneath this “thickened” endocardial rim (or “ridge”) (de la Cruz and Markwald, 1998). In a further analogy to the limb, fibroblast growth factor-4 (FGF-4) is secreted by both the AER and cushion endocardium, while FGF receptors 1, 2 and 3 are detected in the target mesenchyme. When FGF4 was added to cultures of HH 24 chick cushion cells, or delivered directly by viral vectors into cushions in vivo, there was abundant mesenchymal proliferation (Sugi et al., 2003). The response to signaling in each case appears to be proliferation (but not differentiation). That proliferation and elongation of cushion mesenchyme can occur independently of overt differentiation was shown by using an atrial ligation model (Sedmera et al., 2002). Thus, regulatory cascades are critical to the proper proliferation and elongation of the maturing valve leaflets.

III.B.  Differentiation Cushion mesenchymal cells have the potential to differentiate into multiple mesodermal derivatives. Initially, cushion cells express regulatory genes associated with cartilage and bone, but their expression eventually stops while genes associated with fibroblastic lineages persist (e.g., scleraxis, periostin, collagens 1 and 3) (Lincoln et al., 2004). In the Smad6 knockout mouse, the “switch” that normally shuts off cartilage and bone gene expression remains on and cartilage, lamellar bone, bone marrow and even blood were observed in both atrioventricular canal and outlet myxomatous valve leaflets (Galvin et al., 2000). One valvulogenic candidate proposed as a regulator of differentiation (but not proliferation) is periostin (Kruzynska-Frejtag et al., 2001; Norris et al., 2007). Periostin is an evolutionarily conserved, 90 kDa-secreted protein that is distinctly different from other extracellular matrix proteins (Lindsley et al., 2005; Litvin et al., 2005). It has a carboxy terminal that can be alternatively spliced, and four coiled (fas) domains that have closest homology to Drosophila fasciclin1. In Drosophila, the ancestral fasciclin domain functions as an adhesion molecule linked to guiding and targeting axon growth through cell surface–extracellular matrix contacts (Litvin et al., 2005). Inhibiting periostin expression in mice resulted in altered pathways of cushion differentiation from fibroblastic lineages to bone, cartilage or muscle (Oka et al., 2007; Snider et al., 2008; Norris et al., 2008). Another fasciclin protein with homology to periostin is a 67 kDa secreted protein called TGF-induced gene-Human clone 3 or igH3. Both periostin and igH3 are abundant in the forming heart valves and maintained in mature functioning valves (Kern et al., 2005; Lindsley et al., 2005; Norris et al., 2005). The known receptors for periostin and igH3 are integrins, alphaV, beta-3 and alphaV, beta-5, through which periostin can signal changes in cushion cell motility and association with collagen through Rho kinase and PI3 kinase (Butcher et al., 2007b).

377

III.C.  Attenuation and Compaction Attenuation or sculpting is a key step in valvulogenesis. It correlates with the secretion, compaction and alignment of collagen fibrils into highly-organized, stratified layers of extracellular matrix with interstitial fibroblasts (Hinton et al., 2006). Maturation of the valve extracellular matrix is a progressive process that normally proceeds to a tri­ laminar organization (i.e., z. ventricularis, z. spongiosa, z. atrialis) in most mammals (Hinton et al., 2006; Kruithof et al., 2007). In mouse heart valves, there is only a hint of trilaminar organization. Instead, they show an arborizing framework (backbone) of collagen (Norris et al., 2007). In periostin-null mice, the heart valves do not properly attenuate, but remain truncated and poorly-differentiated with no evidence of a fibrous backbone (Snider et al., 2008). Periostin directly binds to collagen and promotes its crosslinking (Norris et al., 2007), which likely contributes to compaction and attenuation of the cushions into leaflets. By promoting cross-linking and attenuation, developing valves also increase their viscoelastic properties (becoming more rigid) (Butcher et al., 2007a; Norris et al., 2007, 2008). Because periostin can function both as a signaling and scaffold-binding protein, it acts like a member of a class of proteins called “matricellular” proteins. As defined by Bornstein et al. (Bornstein, 2000), matricellular proteins are secreted extracellularly where they can bind to either scaffold proteins, like collagen, or cell surface receptors like integrins, where they are able to signal changes in adhesion, migration or differentiation. Examples of other matricellular proteins include thrombospondins, tenascinC, osteopontin, CCN1 and SPARC.

III.D.  Delamination and Formation of Supporting Valve Structures The atrioventricular canal valves form from cushion tissue that buds off the atrioventricular septum to form future septal leaflets of the mitral and tricuspid valves or from the lateral atrioventricular cushions which form the mural leaflets. The lateral cushions elongate distally on an infolding of atrioventricular junctional myocardium that acts as substratum for elongation. As development proceeds, these delaminate from this myocardial bridge, but still retain a connection to it (future papillary muscle) by tendinous cords as a supporting, tension apparatus (Markwald et al., 1998; Oosthoek et al., 1998; Kruithof et al., 2007) (Fig. 2A). Conversely, the smaller atrioventricular septal (medial) cushions grow freely into the ventricular cavity and attach directly to the myocardium of the interventricular septum at their anterior and posterior commissures. Distal outflow tract cushions also initially project freely into the lumen, but later curve back toward the myocardium forming a space that progressively deepens (excavated) to form sinuses (of Valsalva) (Fig. 2). How delamination or

378

excavation occurs and the origin of the tendinous supportive apparatus are questions for which there is only speculation. In periostin-null mice or mice with reduced periostin, delamination is variably disrupted (Snider et al., 2008). This highlights the critical importance of matricellular proteins in this maturation process.

III.E.  Abnormal Morphogenesis (Pediatric Valvular Diseases) Any “missteps” in the morphogenetic steps shown in Fig. 2 could result in a child being born with an abnormal valve structure (see also Chapter 6.2). Valvular defects are among the most common and serious of heart defects (Hoffman and Kaplan, 2002). Structural (anatomical) valvular birth defects include hypoplasia, atresia, stenosis (due to enlargement of the valve blocking the movement of blood), Ebstein’s disease (failure of the right atrioventricular mural leaflet to separate properly from the ventricular walls), a defective suspensory apparatus (parachute valves), a myxomatous valve (i.e., one retaining undifferentiated mesenchyme and excess proteoglycans) or combinations of the above and others, sometimes called polyvalvular disease (Bartram et al., 2001a). Defects in post-epithelial–mesenchymal transition valvulogenesis do not always manifest themselves in lethality. This is evident in the many variants of abnormal valves observed in human congenital heart disease. The presence of shunts (foramina) at several physiological levels in the fetus and the lack of a requirement for a pulmonary circulation combine to permit successful heart function until term for virtually all stenotic or atretic valve lesions that might be present in the human fetus. Indeed, the only valve anomalies that appear to be incompatible with term gestation are excessive regurgitation of the mitral valve or aortic valve. Thus, the consequence of a mutated extracellular or extracellular matrix receptor gene may not be manifested until postnatal or even adult life when a weakened, malformed or leaky valve may become problematic, especially if changes in hemodynamic pressure occur during the contraction cycle that lead to backflow (regurgitation) of blood back into the heart. Hence, “leaky valves” that lead to regurgitation can become a significant clinical problem when a small “dosage” of an extracellular matrix deficit at birth over time equals pathology.

IV.  Regulation of post-epithelial– mesenchymal transition cushion morphogenesis: lessons from adult valve diseases Despite their clinical relevance, surprisingly, the regulatory mechanisms required to remodel post-epithelial– mesenchymal transition cushions into mature valves are

PART  |  6  Cushions, Valves and Septa

just now beginning to be experimentally pursued. New understanding is coming from studies on embryonic postepithelial–mesenchymal transition cushions using mouse genetic approaches and adult valve diseases. This is not surprising, because adult valve diseases are stochastic and appear to have their roots in embryonic development. The recent discovery of filamin A mutations in X-linked mitral and aortic valve dystrophy (Kyndt et al., 2007) raises intriguing questions of how a cytoskeletal filament could cause defects in both inlet and outlet valves. Dietz et al. (2005) originally showed that fibrillin-1 mutations were the basis of atrioventricular mitral valve prolapse seen in Marfan’s syndrome. Fibrillin-1 is considered the principal molecule comprising extracellular matrix microfibrils, but how its misregulation causes myxomatous degeneration of mitral valves is not clear. Myxomatous valves are characterized by elongated and thickened leaflets resulting from degradation of the extracellular matrix fibrous scaffold typical of normal leaflets and increased production of extracellular matrix proteins normally seen in the z. spongiosa (e.g., proteoglycans and GAGs) (Nasuti et al., 2004; Ng et al., 2004). Recently Ng et al. (2004) established the Fbn1C1039G/ mouse that carries a C1039G mutation in the fibrillin-1 gene as a model of myxomatous changes in the mitral valve. What Ng et al. found was that myxomatous changes which followed fibrillin-1 deficiency were the direct result of enhanced TGF signaling. Thus, a root cause of myxomatous mitral valves seems to be altered growth factor signaling. Consistent with this hypothesis, Dietz et al. (2005) genetically-mutated receptors for TGF in mice, which resulted in mitral valve insufficiency similar to that found in Marfan’s syndrome. This highlights the importance of the TGF supergene family in valvulogenesis, not only in the epithelial–mesenchymal transition phases, but also in the post-epithelial–mesenchymal transition morphogenetic processes where cushions are molded into valve leaflets. How altered TGF signaling is transduced into abnormal valvulogenesis is also being actively investigated (Snider et al., 2008). Periostin, which is emerging as a major candidate for regulating post-epithelial–mesenchymal transition valvulogenesis, is a downstream target of TGF signaling (Horiuchi et al., 1999). When equivalent embryonic fibroblasts (MEFs) from E14 wild-type mice and those null for the periostin gene (Postn) were exogenously treated in culture with TGF, null-MEFs synthesized ~32% less collagen than littermate controls, indicating their response to TGF had been blunted by the loss of periostin (Snider et al., 2008). As Postn expression has been linked to TGF gene superfamily signaling and SMAD proteins are known intracellular mediators of TGF signaling, it was not surprising that targeted deletion of Smad6 resulted in abnormal atrioventricular canal and outflow tract valvulogenesis. In all cases, there was replacement of fibrous tissues with bone and/or cartilage-like tissue (Galvin et al., 2000).

Chapter  |  6.1  Molecular Regulation of Cushion Morphogenesis

Periostin expression is also downregulated in Foxc1-nulls that fail to undergo TGF-mediated mesenchymal maturation (Snider et al., 2008). Foxc1 is a TGF1-responsive gene, and null embryos exhibit valve hypoplasia due to defects in mesenchymal maturation (Winnier et al., 1999). These results indicate that a defect in the periostin-null valvular cells blunts TGF-responsiveness and alters pathways of differentiation. Thus, therapy that targets TGF ligands, e.g., Losartan or Valsartan, may ultimately prove beneficial for mitral valve defects (Habashi et al., 2006; Iekushi et al., 2007; Katsuragi et al., 2007). In the clinical setting, where mitral valve regurgitation, insufficiency or prolapse has important prognostic impacts on survival, understanding the molecular control over post-epithelial– mesenchymal transition remodeling of cushions into valves takes on added and urgent significance. Outflow tract valvulogenesis, like that of the atrioventricular inlet, is also dependent on expression of periostin, but apparently there are some differences in mechanism, perhaps reflecting quantitative or qualitative differences in the expression of transcriptional regulatory proteins (SMADs, TWIST, MSX 1 and TBX20) or growth factors (BMP 4 and FGF) associated with atrioventricular cushions (Gaussin et al., 2005; Lincoln et al., 2006). Outflow tract valves, especially aortic leaflets, showed a greater tendency than inlet atrioventricular valves towards osteogenic differentiation and calcification. In periostindeficient mice, pathological structural features are found such as dilation of the aortic root and displacement of the coronary arteries that precede or accompany severe aortic, calcific valve disease (Braverman et al., 2005). In 25% of cases, the aortic valves of periostin-null mice were severely deformed, had a bicuspid morphology and exhibited a wrinkled or “rugged” morphology and premature calcification (Tkatchenko et al., 2009). The abnormal phenotype of outflow tract-derived valves, like those of the atrioventricular myxomatous valves, has their roots in early embryonic development. Based on large-scale gene profiling using a combination of cDNA subtraction, reverse Northern and in situ hybridization, differences in expression of 69 genes were found comparing hearts isolated from periostin/-mice versus wild-type animals. Three genes were particularly upregulated in Postn-null hearts: those encoding pleiotrophin (Ptn); Delta-like 1 homolog (Dlk1); and Runx2 (Runx2). However, Notch 1 and related genes, Hes1, Hey1 and Hey2 were significantly downregulated. PTN is a differentiation factor which has been implicated in early osteoblast differentiation (Yang et al., 2003). Binding of PTN to its receptor, the receptor-type protein tyrosine phosphatase / (RPTP /), was shown to activate -catenin (by dephosphorylation) (Meng et al., 2000), which in turn promotes RUNX2-mediated transcriptional activation of the osteocalcin promoter (Kahler and Westendorf, 2003). Similarly, periostin-null mice show upregulated expression of Ptn

379

correlated with increased transcription of Runx2 and the osteocalcin gene (Bglap1) (Tkatchenko et al., 2009). Dlk1 is a negative regulator of Notch1 (Baladron et al., 2005), another signaling pathway that can affect osteogenesis and calcification. Using real time PCR and in situ hybridization, Notch, Hes1, Hey1 and Hey2 were reduced in the outflow tract of periostin-null heart, whereas Dlk1 and Ptn were upregulated. Also, the boundaries of their expression were expanded from proximal to distal outflow tract, indicating a pattern of misexpression. Taken together, these data suggest that periostin normally functions to suppress Ptn and Dlk1 expression and to enhance Notch 1 activity. By doing so, Runx2 and two main effectors of osteoblast mineralization, osteocalcin and osteopontin, are normally suppressed (Tkatchenko et al., 2009). If, however, periostin is lost, then over time changes will accumulate in postnatal life which will result in abnormal differentiation and eventually calcification of the aortic valve. This is also consistent with the upregulation of Runx2 in human patients with aortic valve calcification, and it implicates the loss of Notch1 signaling in this disease process (Garg et al., 2005).

V.  The living valve V.A.  When Does Valvulogenesis End? It is widely believed that adult valve interstitial cells have a low proliferation index (Taylor et al., 2003) unless activated by a pathogen which usually triggers abnormal remodeling (Aikawa et al., 2006). Does this mean that the allotment of valve cells we have at birth will suffice for a lifetime? Do valve fibroblasts undergo apoptosis? If so, are they replaced? The fact is we currently know very little about adult valve cell proliferation and turnover rates. What we do know is that adult valve interstitial populations, including fibroblast, myofibroblasts and smooth muscle cells are derived from developmental sources, including mesenchymal cells derived from the endocardium that populate primitive embryonic valves/endocardial cushions (de Lange et al., 2004) and cells derived from the embryonic epicardium (Gittenberger-de Groot et al., 1998), and possibly circulating progenitor cells (Zhang et al., 2006). If there is little evidence for valve cell proliferation, we are left with at least two possible alternative mechanisms for renewing valve interstitial cells that do not require cell-cycling. Both would be extensions of embryonic processes. The first option would be that the embryonic endocardial epithelial–mesenchymal transition process is never completely shut down, meaning the endocardial endothelium is able to “seed” new mesenchymal cells after birth, which presumably would differentiate into valve interstitial cells. Recently, Armstrong and Bischoff (Armstrong and Bischoff, 2004; Paruchuri et al., 2006) have found evidence for such a mechanism. Specifically, they found subendocardial cells in a human pediatric pulmonary valve

380

that expressed both endothelial and mesenchymal markers. They subsequently cloned this cell type and showed it to be a valve progenitor cell. Expression of both epithelia/ endocardial and mesenchymal markers would be consistent with an origin by an epithelial–mesenchymal transition mechanism and indicative that epithelial–mesenchymal transition extends (albeit at a low rate) into postnatal life. Progenitor cells derived in this fashion would not necessarily be expected to be detected efficiently by mitotic markers (e.g., BrdU or PCNA). A second and complementary (if not related) mechanism would be recruitment of circulating progenitor cells. Again, this could be an extension of an embryonic mechanism (Eisenberg and Markwald, 2004) in which circulating stem cells have been shown to engraft into the embryonic heart (Zhang et al., 2006). However, proving that a circulating stem cell can engraft into heart valves and differentiate into fibroblasts is no small feat. Controversy regarding the potential of any given stem cell type is due to the fact that most studies evaluating stem cell potential use mixed cell populations (even after enrichment). Definitive assignment of potential must be based on the analysis of a single cell’s potential after long-term engraftment. A single cell, in vivo, transplantation strategy utilizing EGFP hemato­ poietic stem cells derived from mice that express EGFP on a universal promoter was employed by Visconti et al. (2006) to show that bone marrow hematopoietic stem cells give origin to new valve interstitial cells. In this context, it is significant that they found “green fluorescent” cells in all four heart valves. These “transplanted” cells mimicked the unique expression patterns of valvular interstitial cells. Fig. 7 shows a mouse aortic valve with interstitial cells of hematopoietic stem cell origin.

Figure 7  Stem cell contribution to an adult aortic valve. Bone marrowderived hematopoietic stem cells were engrafted into an aortic valve following lethal radiation of the recipient and injection of a single clone of GFP-positive stem cells. Pleuripotentency of the cells was established by survival of the host. The experimental procedure and additional details are decribed in Visconti et al. (2006).

PART  |  6  Cushions, Valves and Septa

Thus, these recent studies suggest that valve development is a life-long process in which resident interstitial valve cells may turn over by apoptosis, requiring a regenerative population of cells for replacement. These new interstitial valve cells may be derived locally by epithelial– mesenchymal transition, or through recruitment of circulating progenitor cells. With the advent of regenerative medicine, a comprehensive understanding of valvular homeostasis will be paramount for treating valve diseases by harnessing the power of progenitor cells and for engineering successful long-term valve replacements.

Acknowledgments The authors acknowledge Drs Joey Barnett (Vanderbilt University) and Scott Klewer (University of Arizona) for useful comments and discussions. We thank the members of our laboratories who produced many of the findings that are summarized here. We are grateful to the National Heart, Lung and Blood Institute, NIH and the American Heart Association for their support of our work in the laboratory.

References Aikawa, E., Whittaker, P., Farber, M., Mendelson, K., Padera, R.F., Aikawa, M., Schoen, F.J., 2006. Human semilunar cardiac valve remodeling by activated cells from fetus to adult – Implications for postnatal adaptation, pathology, and tissue engineering. Circulation 113, 1344–1352. Araki, T., Mohi, M.G., Ismat, F.A., Bronson, R.T., Williams, I.R., Kutok, J.L., Yang, W.T., Pao, L.I., Gilliland, D.G., Epstein, J.A., Neel, B.G., 2004. Mouse model of Noonan syndrome reveals cell type- and gene dosage-dependent effects of Ptpn11 mutation. Nat. Med. 10, 849–857. Armstrong, E.J., Bischoff, J., 2004. Heart valve development – Endothelial cell signaling and differentiation. Circ. Res. 95, 459–470. Azhar, M., Runyan, R.B., Gard, C., Sanford, L.P., Miller, M.L., Andringa, A., Pawlowski, S., Rajan, S., Doetschman, T., 2009. Ligandspecific function of transforming growth factor beta in epithelial– mesenchymal transition in heart development. Dev. Dyn. 238, 431–442. Azhar, M., Schultz Jel, J., Grupp, I., Dorn, G.W., Meneton, P., Molin, D.G., Gittenberger-de Groot, A.C., Doetschman, T., 2003. Transforming growth factor beta in cardiovascular development and function. Cytokine Growth Factor Rev. 14, 391–407. Baladron, V., Ruiz-Hidalgo, M.J., Nueda, M.L., Diaz-Guerra, M.J., Garcia-Ramirez, J.J., Bonvini, E., Gubina, E., Laborda, J., 2005. Dlk acts as a negative regulator of Notch1 activation through interactions with specific EGF-like repeats. Exp. Cell Res. 303, 343–359. Baldwin, H.S., Solursh, M., 1990. Degradation of hyaluronic acid does not prevent looping of the mammalian heart in situ. Dev. Biol. 136 (2), 555–559. Baldwin, H.S., Shen, H.M., Yan, H.C., DeLisser, H.M., Chung, A., Mickanin, C., Trask, T., Kirschbaum, N.E., Newman, P.J., Albelda, S.M., 1994. Platelet endothelial cell adhesion molecule-1 (PECAM-1/CD31), alternatively spliced, functionally distinct isoforms expressed during mammalian cardiovascular development. Development 120 (9), 2539–2553.

Chapter  |  6.1  Molecular Regulation of Cushion Morphogenesis

Bartram, U., Bartelings, M.M., Kramer, H.H., Gittenberger-de Groot, A.C., 2001a. Congenital polyvalvular disease: A review. Pediatr. Cardiol. 22, 93–101. Bartram, U., Molin, D.G., Wisse, L.J., Mohamad, A., Sanford, L.P., Doetschman, T., Speer, C.P., Poelmann, R.E., Gittenberger-De Groot, A.C., 2001b. Double-outlet right ventricle and overriding tricuspid valve reflect disturbances of looping, myocardialization, endocardial cushion differentiation, and apoptosis in tgf-beta (2)-knockout mice. Circulation 103, 2745–2752. Bashey, R.I., Martinez-Hernandez, A., Jimenez, S.A., 1992. Isolation, characterization, and localization of cardiac collagen type VI. Associations with other extracellular matrix components. Circ. Res. 70, 1006–1017. Bernanke, D.H., Markwald, R.R., 1979. Effects of hyaluronic acid on cardiac cushion tissue cells in collagen matrix cultures. Tex. Rep. Biol. Med. 39, 271–285. Bernanke, D.H., Orkin, R.W., 1984. Hyaluronate binding and degradation by cultured embryonic chick cardiac cushion and myocardial cells. Dev. Biol. 106, 360–367. Bornstein, P., 2000. Matricellular proteins, an overview: Introduction. Matrix Biol. 19, 555–556. Bouchey, D., Argraves, W.S., Little, C.D., 1996. Fibulin-1, vitronectin, and fibronectin expression during avian cardiac valve and septa development. Anat. Rec. 244, 540–551. Boyer, A.S., Runyan, R.B., 2001. TGF beta type III and TGF beta type II receptors have distinct activities during epithelial–mesenchymal cell transformation in the embryonic heart. Dev. Dyn. 221, 454–459. Boyer, A.S., Ayerinskas, I.I., Vincent, E.B., McKinney, L.A., Weeks, D.L., Runyan, R.B., 1999. TGF beta 2 and TGF beta 3 have separate and sequential activities during epithelial–mesenchymal cell transformation in the embryonic heart. Dev. Biol. 208, 530–545. Braverman, A.C., Guven, H., Beardslee, M.A., Makan, M., Kates, A.M., Moon, M.R., 2005. The bicuspid aortic valve. Curr. Problems Cardiol. 30, 470–522. Brown, C.B., Boyer, A.S., Runyan, R.B., Barnett, J.V., 1996. Antibodies to the type II TGF beta receptor block cell activation and migration during atrioventricular cushion transformation in the heart. Dev. Biol. 174, 248–257. Brown, C.B., Boyer, A.S., Runyan, R.B., Barnett, J.V., 1999. Requirement of type III TGF-beta receptor for endocardial cell transformation in the heart. Science 283, 2080–2082. Butcher, J.T., Markwald, R.R., 2007. Valvulogenesis, the moving target. Phil. Trans. Royal Soc. B. Biol. Sci. 362, 1489–1503. Butcher, J.T., McQuinn, T.C., Sedmera, D., Turner, D., Markwald, R.R. 2007a. Transitions in early embryonic atrioventricular valvular function correspond with changes in cushion biomechanics that are predictable by tissue composition. Circ. Res. 100, 1503–1511. Butcher, J.T., Norris, R.A., Hoffman, S., Mjaatvedt, C.H., Markwald, R.R., 2007b. Periostin promotes atrioventricular mesenchyme matrix invasion and remodeling mediated by integrin signaling through Rho/PI 3-kinase. Dev. Biol. 302, 256–266. Camenisch, T.D., Spicer, A.P., Brehm-Gibson, T., Biesterfeldt, J., Augustine, M.L., Calabro, A., Kubalak, S., Klewer, S.E., McDonald, J.A., 2000. Disruption of hyaluronan synthase-2 abrogates normal cardiac morphogenesis and hyaluronan-mediated transformation of epithelium to mesenchyme. J. Clin. Invest. 106, 349–360. Camenisch, T.D., Molin, D.G., Person, A., Runyan, R.B., Gittenbergerde Groot, A.C., McDonald, J.A., Klewer, S.E., 2002a. Temporal and distinct TGFbeta ligand requirements during mouse and avian endocardial cushion morphogenesis. Dev. Biol. 248, 170–181.

381

Camenisch, T.D., Schroeder, J.A., Bradley, J., Klewer, S.E., McDonald, J.A., 2002b. Heart-valve mesenchyme formation is dependent on hyaluronan-augmented activation of ErbB2-ErbB3 receptors. Nat. Med. 8, 850–855. Carmeliet, P., Ferreira, V., Breier, G., Pollefeyt, S., Kieckens, L., Gertsenstein, M., Fahrig, M., Vandenhoeck, A., Harpal, K., Eberhardt, C., Declercq, C., Pawling, J., Moons, L., Collen, D., Risau, W., Nagy, A., 1996. Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 380, 435–439. Carvalhaes, L.S., Gervasio, O.L., Guatimosim, C., Heljasvaara, R., Sormunen, R., Pihlajaniemi, T., Kitten, G.T., 2006. Collagen XVIII/ endostatin is associated with the epithelial–mesenchymal transformation in the atrioventricular valves during cardiac development. Dev. Dyn. 235, 132–142. Chang, C.P., Neilson, J.R., Bayle, J.H., Gestwicki, J.E., Kuo, A., Stankunas, K., Graef, I.A., Crabtree, G.R., 2004. A field of myocardial– endocardial NFAT signaling underlies heart valve morphogenesis. Cell 118, 649–663. Chang, H., Huylebroech, D., Verschueren, K., Guo, Q., Matzuk, M.M., Zwijsen, A., 1999. Smad5 knockout mice die at midgestation due to multiple embryonic and extraembryonic defects. Development 126, 1631–1642. Chen, B., Bronson, R.T., Klaman, L.D., Hampton, T.G., Wang, J.F., Green, P.J., Magnuson, T., Douglas, P.S., Morgan, J.P., Neel, B.G., 2000. Mice mutant for Egfr and Shp2 have defective cardiac semi­ lunar valvulogenesis. Nat. Genet. 24, 296–299. Chen, Q., Chen, H., Zheng, D., Kuang, C., Fang, H., Zou, B., Zhu, W., Bu, G., Jin, T., Wang, Z., Zhang, X., Chen, J., Field, L.J., Rubart, M., Shou, W., Chen, Y., 2009. Smad7 is required for the development and function of the heart. J. Biol. Chem. 284, 292–300. Cole, W.G., Chan, D., Hickey, A.J., Wilcken, D.E., 1984. Collagen composition of normal and myxomatous human mitral heart-valves. Biochem. J. 219, 451–460. de Caestecker, M., 2004. The transforming growth factor-beta superfamily of receptors. Cytokine Growth Factor Rev. 15, 1–11. de la Cruz, M.V., Markwald, R.R., 1998. Embryological development of the ventricular inlets. Septation and atrioventricular valve apparatus. In: Markwald, R.R. (Ed.), Living Morphogenesis of the Heart. Birkhauser, Boston, MA, pp. 131–157. de Lange, F.J., Moorman, A.F., Anderson, R.H., Manner, J., Soufan, A.T., de Vries, C.D., Schneider, M.D., Webb, S., van den Hoff, M.J., Christoffels, V.M., 2004. Lineage and morphogenetic analysis of the cardiac valves. Circ. Res. 95, 645–654. Delot, E.C., Bahamonde, M.E., Zhao, M., Lyons, K.M., 2003. BMP signaling is required for septation of the outflow tract of the mammalian heart. Development 130, 209–220. DeSesso, J.M., 1987. Maternal factors in developmental toxicity. Teratog. Carcinog. Mutagen. 7, 225–240. Desgrosellier, J.S., Mundell, N.A., McDonnell, M.A., Moses, H.L., Barnett, J.V., 2005. Activin receptor-like kinase 2 and Smad6 regulate epithelial–mesenchymal transformation during cardiac valve formation. Dev. Biol. 280, 201–210. Dickson, M.C., Slager, H.G., Duffie, E., Mummery, C.L., Akhurst, R.J., 1993. RNA and protein localisations of TGF beta 2 in the early mouse embryo suggest an involvement in cardiac development. Development 117, 625–639. Dietz, H.C., Loeys, B., Carta, L., Ramirez, F., 2005. Recent progress towards a molecular understanding of Marfan syndrome. Am. J. Med. Genet. (Part C: Semin. Med. Genet.) 139C, 4–9.

382

Dor, Y., Klewer, S.E., McDonald, J.A., Keshet, E., Camenisch, T.D., 2003. VEGF modulates early heart valve formation. Anat. Rec. 271A, 202–208. Dunker, N., Krieglstein, K., 2000. Targeted mutations of transforming growth factor-beta genes reveal important roles in mouse development and adult homeostasis. Eur. J. Biochem. 267, 6982–6988. Eisenberg, L.M., Markwald, R.R., 2004. Cellular recruitment and the development of the myocardium. Dev. Biol. 274, 225–232. Enciso, J.M., Gratzinger, D., Camenisch, T.D., Canosa, S., Pinter, E., Madri, J.A., 2003. Elevated glucose inhibits VEGF-A-mediated endocardial cushion formation, modulation by PECAM-1 and MMP2. J. Cell Biol. 160, 605–615. Fallon, J.F., Lopez, A., Ros, M.A., Savage, M.P., Olwin, B.B., Simandl, B.K., 1994. Fgf-2-apical ectodermal ridge growth signal for chick limb development. Science 264, 104–107. Ferrara, N., Carver-Moore, K., Chen, H., Dowd, M., Lu, L., O’Shea, K.S., Powell-Braxton, L., Hillan, K.J., Moore, M.W., 1996. Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 380, 439–442. Filip, D.A., Radu, A., Simionescu, M., 1986. Interstitial-cells of the heart-valves possess characteristics similar to smooth-muscle cells. Circ. Res. 59, 310–320. Forouhar, A.S., Liebling, M., Hickerson, A., Nasiraei-Moghaddam, A., Tsai, H.J., Hove, J.R., Fraser, S.E., Dickinson, M.E., Gharib, M., 2006. The embryonic vertebrate heart tube is a dynamic suction pump. Science 312, 751–753. Francis-West, P.H., Tatla, T., Brickell, P.M., 1994. Expression patterns of the bone morphogenetic protein genes Bmp-4 and Bmp-2 in the developing chick face suggest a role in outgrowth of the primordia. Dev. Dyn. 201, 168–178. Galvin Jr., K.M., Donovan, M.J., Lynch, C.A., Meyer, R.I., Paul, R.J., Lorenz, J.N., Fairchild-Huntress, V., Dixon, K.L., Dunmore, J.H., Gimbrone, M.A., Falb, D., Huszar, D., 2000. A role for smad6 in development and homeostasis of the cardiovascular system. Nat. Genet. 24, 171–174. Garg, V., 2006. Molecular genetics of aortic valve disease. Curr. Opin. Cardiol. 21, 180–184. Garg, V., Muth, A.N., Ransom, J.F., Schluterman, M.K., Barnes, R., King, I.N., Grossfeld, P.D., Srivastava, D., 2005. Mutations in NOTCH1 cause aortic valve disease. Nature 437, 270–274. Gaussin, V., Van de Putte, T., Mishina, Y., Hanks, M.C., Zwijsen, A., Huylebroeck, D., Behringer, R.R., Schneider, M.D., 2002. Endocardial cushion and myocardial defects after cardiac myocytespecific conditional deletion of the bone morphogenetic protein receptor ALK3. Proc. Natl. Acad. Sci. U.S.A. 99, 2878–2883. Gaussin, V., Morley, G.E., Cox, L., Zwijsen, A., Vance, K.M., Emile, L., Tian, Y.M., Liu, J., Hong, C., Myers, D., Conway, S.J., Depre, C., Mishina, Y., Behringer, R.R., Hanks, M.C., Schneider, M.D., Huylebroeck, D., Fishman, G.I., Burch, J.B., Vatner, S.F., 2005. Alk3/Bmpr1a receptor is required for development of the atrioventricular canal into valves and annulus fibrosus. Circ. Res. 97, 219–226. Gittenberger-de Groot, A.C., Vrancken Peeters, M.P., Mentink, M.M., Gourdie, R.G., Poelmann, R.E., 1998. Epicardium-derived cells contribute a novel population to the myocardial wall and the atrioventricular cushions. Circ. Res. 82, 1043–1052. Goumans, M.J., Valdimarsdottir, G., Itoh, S., Rosendahl, A., Sideras, P., ten Dijke, P., 2002. Balancing the activation state of the endothelium via two distinct TGF-beta type I receptors. EMBO J. 21, 1743–1753.

PART  |  6  Cushions, Valves and Septa

Grego-Bessa, J., Luna-Zurita, L., del Monte, G., Bolós, V., Melgar, P., Arandilla, A., Garratt, A.N., Zang, H., Mukouyama, Y.S., Chen, H., Shou, W., Ballestar, E., Esteller, M., Rojas, A., Pérez-Pomares, J.M., de la Pompa, J.L., 2007. Notch signaling is essential for ventricular chamber development. Dev. Cell 12, 415–429. Habashi, J.P., Judge, D.P., Holm, T.M., Cohn, R.D., Loeys, B.L., Cooper, T.K., Myers, L., Klein, E.C., Liu, G.S., Calvi, C., Podowski, M., Neptune, E.R., Halushka, M.K., Bedja, D., Gabrielson, K., Rifkin, D.B., Carta, L., Ramirez, F., Huso, D.L., Dietz, H.C., 2006. Losartan, an AT1 antagonist, prevents aortic aneurysm in a mouse model of Marfan syndrome. Science 312, 117–121. Hall, A., 1998. Rho GTPases and the actin cytoskeleton. Science 279, 509–514. Hinton, R.B., Lincoln, J., Deutsch, G.H., Osinska, H., Manning, P.B., Benson, D.W., Yutzey, K.E., 2006. Extracellular matrix remodeling and organization in developing and diseased aortic valves. Circ. Res. 98, 1431–1438. Hoffmann, A., Preobrazhenska, O., Wodarczyk, C., Medler, Y., Winkel, A., Shahab, S., Huylebroeck, D., Gross, G., Verschueren, K., 2005. Transforming growth factor-beta-activated kinase-1 (TAK1), a MAP3K, interacts with Smad proteins and interferes with osteogenesis in murine mesenchymal progenitors. J. Biol. Chem. 280, 27271–27283. Hoffman, J.I., Kaplan, S., 2002. The incidence of congenital heart disease. J. Am. Coll. Cardiol. 39, 1890–1900. Hogan, B.L., 1996. Bone morphogenetic proteins in development. Curr. Opin. Genet. Dev. 6, 432–438. Honda, A., Kanke, Y., Mori, Y., 1975. Biochemical study of cardiac valvular tissue – isolation of glycoproteins and proteoglycans from salt-insoluble materials of bovine heart valve. J. Mol. Cell Cardiol. 7, 715–732. Horiuchi, K., Amizuka, N., Takeshita, S., Takamatsu, H., Katsuura, M., Ozawa, H., Toyama, Y., Bonewald, L.F., Kudo, A., 1999. Identification and characterization of a novel protein, periostin, with restricted expression to periosteum and periodontal ligament and increased expression by transforming growth factor beta. J. Bone Min. Res. 14, 1239–1249. Icardo, J.M., Colvee, E., 1995. Atrioventricular valves of the mouse. 3. Collagenous skeleton and myotendinous junction. Anat. Rec. 243, 367–375. Iekushi, K., Taniyama, Y., Azuma, J., Katsuragi, N., Dosaka, N., Sanada, F., Koibuchi, N., Nagao, K., Ogihara, T., Morishita, R., 2007. Novel mechanisms of valsartan on the treatment of acute myocardial infarction through inhibition of the antiadhesion molecule periostin. Hypertension 49, 1409–1414. Isner, J.M., Ropper, A., Hirst, K., 2001. VEGF gene transfer for diabetic neuropathy. Hum. Gene. Ther. 12, 1593–1594. Iwamoto, R., Yamazaki, S., Asakura, M., Takashima, S., Hasuwa, H., Miyado, K., Adachi, S., Kitakaze, M., Hashimoto, K., Raab, G., Nanba, D., Higashiyama, S., Hori, M., Klagsbrun, M., Mekada, E., 2003. Heparin-binding EGF-like growth factor and ErbB signaling is essential for heart function. Proc. Natl. Acad. Sci. U.S.A. 100, 3221–3226. Jackson, L.F., Qiu, T.H., Sunnarborg, S.W., Chang, A., Zhang, C.L., Patterson, C., Lee, D.C., 2003. Defective valvulogenesis in HB-EGF and TACE-null mice is associated with aberrant BMP signaling. EMBO J. 22, 2704–2716. Jiao, K., Langworthy, M., Batts, L., Brown, C.B., Moses, H.L., Baldwin, H.S., 2006. Tgf beta signaling is required for atrioventricular cushion mesenchyme remodeling during in vivo cardiac development. Development 133, 4585–4593.

Chapter | 6.1

Molecular Regulation of Cushion Morphogenesis

Johnson, D.W., Berg, J.N., Baldwin, M.A., Gallione, C.J., Marondel, I., Yoon, S.-J., Stenzel, T.T, Speer, M., Pericak-Vance, M.A., Diamond, A., Guttmacher, A.E., Jackson, C.E., Attisano, L., Kucherlapati, R., Porteous, M.E., Marchuk, D.A., 1996. Mutations in the activin receptor-like kinase 1 gene in hereditary haemorrhagic telangiectasia type 2. Nat. Genet. 13, 189–195. Johnson, E.N., Lee, Y.M., Sander, T.L., Rabkin, E., Schoen, F.J., Kaushal, S., Bischoff, J., 2003. NFATc1 mediates vascular endothelial growth factor-induced proliferation of human pulmonary valve endothelial cells. J. Biol. Chem. 278, 1686–1692. Johnson, G.L., Dohlman, H.G., Graves, L.M., 2005. MAPK kinase kinases (MKKKs) as a target class for small-molecule inhibition to modulate signaling networks and gene expression. Curr. Opin. Chem. Biol. 9, 325–331. Jones, C.M., Lyons, K.M., Hogan, B.L., 1991. Involvement of bone morphogenetic protein-4 (BMP-4) and Vgr-1 in morphogenesis and neurogenesis in the mouse. Development 111, 531–542. Kaartinen, V., Voncken, J.W., Shuler, C., Warburton, D., Bu, D., Heisterkamp, N., Groffen, J., 1995. Abnormal lung development and cleft palate in mice lacking TGF-beta 3 indicates defects of epithelial–mesenchymal interaction. Nat. Genet. 11, 415–421. Kaartinen, V., Haataja, L., Nagy, A., Heisterkamp, N., Groffen, J., 2002. TGF beta 3-induced activation of RhoA/Rho-kinase pathway is necessary but not sufficient for epithelio–mesenchymal transdifferentiation, Implications for palatogenesis. Int. J. Mol. Med. 9, 563–570. Kahler, R.A., Westendorf, J.J., 2003. Lymphoid enhancer factor-1 and beta-catenin inhibit Runx2-dependent transcriptional activation of the osteocalcin promoter. J. Biol. Chem. 278, 11937–11944. Kern, C.B., Hoffman, S., Moreno, R., Damon, B.J., Norris, R.A., Krug, E.L., Markwald, R.R., Mjaatvedt, C.H., 2005. Immunolocalization of chick periostin protein in the developing heart. Anat. Rec. (Part A – Discoveries in Mol. Cell. and Evol. Biol.) 284A, 415–423. Kern, C.B., Twal, W.O., Mjaatvedt, C.H., Fairey, S.E., Toole, B.P., IruelaArispe, M.L., Argraves, W.S., 2006. Proteolytic cleavage of versican during cardiac cushion morphogenesis. Dev. Dyn. 235, 2238–2247. Kinsella, M., Fitzharris, T., 1980. Origin of cushion tissue in the developing chick heart, cinematographic recordings of in situ formation. Science 207, 1359–1360. Kitten, G.T., Markwald, R.R., Bolender, D.L., 1987. Distribution of basement membrane antigens in cry preserved early embryonic hearts. Anat. Rec. 217 (4), 379–390. Klewer, S.E., Krob, S.L., Kolker, S.J., Kitten, G.T., 1998. Expression of type VI collagen in the developing mouse heart. Dev. Dyn. 211, 248–255. Klewer, S.E., Yatskievych, T., Pogreba, K., Stevens, M.V., Antin, P.B., Camenisch, T.D., 2006. Has2 expression in heart forming regions is independent of BMP signaling. Gene. Expr. Patterns 6, 462–470. Kolditz, D., Wijffels, M., Blom, N., van der Laarse, A., Markwald, R., Schalij, M., Gittenberger-de Groot, A.C., 2007. Persistence of functional atrioventricular accessory pathways in post-septated embryonic avian hearts, implications for morphogenesis and functional maturation of the cardiac conduction system. Circ. Res. 115, 17–26. Krug, E.L., Runyan, R.B., Markwald, R.R., 1985. Protein extracts from early embryonic hearts initiate cardiac endothelial cytodifferentiation. Dev. Biol. 112 (2), 414–426. Kruithof, B.P., Krawitz, S.A., Gaussin, V., 2007. Atrioventricular valve development during late embryonic and postnatal stages involves condensation and extracellular matrix remodeling. Dev. Biol. 302, 208–217.

383

Kruzynska-Frejtag, A., Machnicki, M., Rogers, R., Markwald, R.R., Conway, S.J., 2001. Periostin (an osteoblast-specific factor) is expressed within the embryonic mouse heart during valve formation. Mech. Dev. 103, 183–188. Kyndt, F., Gueffet, J.P., Probst, V., Jaafar, P., Legendre, A., Le Bouffant, F., Toquet, C., Roy, E., McGregor L., LynchS.A., Newbury-Ecob, R., Tran, V., Young, I., Trochu, J.N., Le Marec, H., Schott, J.J., 2007. Mutations in the gene encoding filamin A as a cause for familial cardiac valvular dystrophy. Circulation 115, 40–49. Lai, E.C., 2004. Notch signaling, control of cell communication and cell fate. Development 131, 965–973. Lai, Y.T., Beason, K.B., Brames, G.P., Desgrosellier, J.S., Cleggett, M.C., Shaw, M.V., Brown, C.B., Barnett, J.V., 2000. Activin receptor-like kinase 2 can mediate atrioventricular cushion transformation. Dev. Biol. 222, 1–11. Lakkis, M.M., Epstein, J.A., 1998. Neurofibromin modulation of ras activity is required for normal endocardial–mesenchymal transformation in the developing heart. Development 125, 4359–4367. Lee, R.J., Springer, M.L., Blanco-Bose, W.E., Shaw, R., Ursell, P.C., Blau, H.M., 2000. VEGF gene delivery to myocardium, deleterious effects of unregulated expression. Circulation 102, 898–901. Lewandoski, M., Sun, X., Martin, G.R., 2000. Fgf8 signaling from the AER is essential for normal limb development. Nat. Genet. 26, 460–463. Liebner, S., Cattelino, A., Gallini, R., Rudini, N., Iurlaro, M., Piccolo, S., Dejana, E., 2004. Beta-catenin is required for endothelial–mesenchymal transformation during heart cushion development in the mouse. J. Cell Biol. 166, 359–367. Lincoln, J., Alfieri, C.M., Yutzey, K.E., 2004. Development of heart valve leaflets and supporting apparatus in chicken and mouse embryos. Dev. Dyn. 230, 239–250. Lincoln, J., Alfieri, C.M., Yutzey, K.E., 2006. BMP and FGF regulatory pathways control cell lineage diversification of heart valve precursor cells. Dev. Biol. 292, 290–302. Lindsley, A., Li, W., Wang, J., Maeda, N., Rogers, R., Conway, S.J., 2005. Comparison of the four mouse fasciclin-containing genes expression patterns during valvuloseptal morphogenesis. Gene. Expr. Patterns 5, 593–600. Litvin, J., Zhu, S.M., Norris, R., Markwald, R.R., 2005. Periostin family of proteins, therapeutic targets for heart disease. Anat. Rec. (Part a – Discoveries in Mol. Cellular and Evol. Biol.) 287A, 1205–1212. Loeber, C.P., Runyan, R.B., 1990. A comparison of fibronectin, laminin, and galactosyltransferase adhesion mechanisms during embryonic cardiac mesenchymal cell migration in vitro. Dev. Biol. 140 (2), 401–412. Loomes, K.M., Underkoffler, L.A., Morabito, J., Gottlieb, S., Piccoli, D.A., Spinner, N.B., Baldwin, H.S., Oakey, R.J., 1999. The expression of Jagged1 in the developing mammalian heart correlates with cardiovascular disease in Alagille syndrome. Hum. Mol. Genet. 8, 2443–2449. Loomes, K.M., Taichman, D.B., Glover, C.L., Williams, P.T., Markowitz, J.E., Piccoli, D.A., Baldwin, H.S., Oakey, R.J., 2002. Characterization of Notch receptor expression in the developing mammalian heart and liver. Am. J. Med. Genet. 112, 181–189. Lopez-Casillas, F., Cheifetz, S., Doody, J., Andres, J.L., Lane, W.S., Massagué, J., 1991. Structure and expression of the membrane proteoglycan betaglycan, a component of the TGF-beta receptor system. Cell 67, 785–795. Lopez-Casillas, F., Payne, H.M., Andres, J.L., Massague, J., 1994. Betaglycan can act as a dual modulator of TGF-beta access to signaling

384

receptors, mapping of ligand binding and GAG attachment sites. J. Cell Biol. 124, 557–568. Lueder, F.L., Kim, S.B., Buroker, C.A., Bangalore, S.A., Ogata, E.S., 1995. Chronic maternal hypoxia retards fetal growth and increases glucose utilization of select fetal tissues in the rat. Metabolism 44, 532–537. Luetteke, N.C., Phillips, H.K., Qiu, T.H., Copeland, N.G., Earp, H.S., Jenkins, N.A., Lee, D.C., 1994. The mouse waved-2 phenotype results from a point mutation in the EGF receptor tyrosine kinase. Genes Dev. 8, 399–413. Lyons, K.M., Pelton, R.W., Hogan, B.L., 1990. Organogenesis and pattern formation in the mouse, RNA distribution patterns suggest a role for bone morphogenetic protein-2A (BMP-2A). Development 109, 833–844. Ma, L.J., Lu, M.F., Schwartz, R.J., Martin, J.F., 2005. Bmp2 is essential for cardiac cushion epithelial–mesenchymal transition and myocardial patterning. Development 132, 5601–5611. Manasek, F.J., Reid, M., Vinson, W., Seyer, J., Johnson, R., 1973. Glycosaminoglycan synthesis by the early embryonic chick heart. Dev. Biol. (Orlando) 35, 332–348. Markwald, R.R., Fitzharris, T.P., Manasek, F.J., 1977. Structural development of endocardial cushions. Am. J. Anat. 148, 85–119. Markwald, R.R., Runyan, R.B., Kitten, G.T., Funderburg, F.M., Bernanke, D.H., Brauer, P.R., 1984. Use of collagen gel cultures to study heart development, proteoglycan and glycoprotein interactions during formation of endocardial cushion tissue. In: Trelstad, R.L. (Ed.), The Role of Extracellular Matrix in Development. Alan R. Liss, Inc, New York, NY, pp. 323–350. Markwald, R., Trusk, T., Moreno-Rodriguez, R., 1998. Formation and septation of the tubular heart, integrating the dynamics of morphology with emerging molecular concepts. In: de la Cruz, M., Markwald, R.R. (Eds.). Living Morphogenesis of the Heart. Birkhauser, Boston, MA, pp. 43–84. McAllister, K.A., Grogg, K.M., Johnson, D.W., Gallione, C.J., Baldwin, M.A., Jackson, C.E., Helmbold, E.A., Markel, D.S., Mckinnon, W.C., Murrell, J., Mccormick, M.K., Pericakvance, M.A., Heutink, P., Oostra, B.A., Haitjema, T., Westerman, C.J., Porteous, M.E., Guttmacher, A.E., Letarte, M., Marchuk, D.A., 1994. Endoglin, a Tgf-Beta binding-protein of endothelial-cells, is the gene for hereditary hemorrhagic telangiectasia type-1. Nat. Genet. 8, 345–351. McCright, B., Gao, X., Shen, L.Y., Lozier, J., Lan, Y., Maguire, M., Herzlinger, D., Weinmaster, G., Jiang, R.L., Gridley, T., 2001. Defects in development of the kidney, heart and eye vasculature in mice homozygous for a hypomorphic Notch2 mutation. Development 128, 491–502. McCright, B., Lozier, J., Gridley, T., 2002. A mouse model of Alagille syndrome, Notch2 as a genetic modifier of Jag1 haploinsufficiency. Development 129, 1075–1082. McGuire, P.G., Orkin, R.W., 1992. Urokinase activity in the developing avian heart, aspatial and temporal analysis. Dev. Dyn. 193, 24–33. Meng, K., Rodriguez-Pena, A., Dimitrov, T., Chen, W., Yamin, M., Noda, M., Deuel, T.F., 2000. Pleiotrophin signals increased tyrosine phosphorylation of beta-catenin through inactivation of the intrinsic catalytic activity of the receptor-type protein tyrosine phosphatase beta/zeta. Proc. Nat. Acad. Sci. U.S.A. 97, 2603–2608. Mercado-Pimentel, M.E., Hubbard, A.D., Runyan, R.B., 2007. Endoglin and Alk5 regulate epithelial–mesenchymal transformation during cardiac valve formation. Dev. Biol. 304, 420–432.

PART | 6

Cushions, Valves and Septa

Miao, C.Y., Zuberbuhler, J.S., Zuberbuhler, J.R., 1988. Prevalence of congenital cardiac anomalies at high altitude. J. Am. Coll. Cardiol. 12, 224–228. Miosge, N., Sasaki, T., Chu, M.L., Herken, R., Timpl, R., 1998. Ultrastructural localization of microfibrillar fibulin-1 and fibulin-2 during heart development indicates a switch in molecular associations. Cell Mol. Life Sci. 54, 606–613. Miquerol, L., Langille, B.L., Nagy, A., 2000. Embryonic development is disrupted by modest increases in vascular endothelial growth factor gene expression. Development 127, 3941–3946. Mulholland, D.L., Gotlieb, A.I., 1996. Cell biology of valvular interstitial cells. Can. J. Cardiol. 12, 231–236. Nakamura, T., Colbert, M., Krenz, M., Hahn, H., Molkentin, J., Robbins, J., 2006. Rescue of the cardiac phenotype of noonan syndrome in vivo by ablation of cardiac ERK1/2 signaling. Circulation 114 7–7. Nasuti, J.F., Zhang, P.J., Feldman, M.D., Pasha, T., Khurana, J.S., Gorman, J.H., Gorman, R.C., Narula, J., Narula, N., 2004. Fibrillin and other matrix proteins in mitral valve prolapse syndrome. Ann. Thorac. Surg. 77, 532–536. Nebigil, C.G., Maroteaux, L., 2001. A novel role for serotonin in heart. Trends Cardiovasc. Med. 11, 329–335. Negro, A., Brar, B.K., Lee, K.F., 2004. Essential roles of Her2/erbB2 in cardiac development and function. Recent Prog. Horm. Res. 59, 1–12. Ng, C.M., Cheng, A., Myers, L.A., Martinez-Murillo, F., Jie, C.F., Bedja, D., Gabrielson, K.L., Hausladen, J.M., Mecham, R.P., Judge, D.P., Dietz, H.C., 2004. TGF-beta-dependent pathogenesis of mitral valve prolapse in a mouse model of Marfan syndrome. J. Clin. Invest. 114, 1586–1592. Niessen, K., Fu, Y., Chang, L., Hoodless, P.A., McFadden, D., Karsan, A., 2008. Slug is a direct Notch target required for initiation of cardiac cushion cellularization. J. Cell Biol. 182, 315–325. Nieto, M.A., Sargent, M.G., Wilkinson, D.G., Cooke, J., 1994. Control of cell behavior during vertebrate development by slug, a zinc finger gene. Science 264, 835–839. Noble, P.W., McKee, C.M., Cowman, M., Shin, H.S., 1996. Hyaluronan fragments activate an NF-kappa B/I-kappa B alpha autoregulatory loop in murine macrophages. J. Exp. Med. 183, 2373–2378. Norris, R.A., Kern, C.B., Wessels, A., Wirrig, E.E., Markwald, R.R., Mjaatvedt, C.H., 2005. Detection of beta ig-H3, a TGF beta induced gene, during cardiac development and its complementary pattern with periostin. Anat. Embryol. 210, 13–23. Norris, R.A., Damon, B., Mironov, V., Kasyanov, V., Ramamurthi, A., Moreno-Rodriguez, R., Trusk, T., Potts, J.D., Goodwin, R.L., Davis, J., Hoffman, S., Wen, X., Sugi, Y., Kern, C.B., Mjaatvedt, C.H., Turner, D.K., Oka, T., Conway, S.J., Molkentin, J.D., Forgacs, G., Markwald, R.R., 2007. Periostin regulates collagen fibrillogenesis and the biomechanical properties of connective tissues. J. Cell Biochem. 101, 695–711. Norris, R.A., Moreno-Rodriguez, R., Sugi, Y., Hoffman, S., Amos, J., Hart, M., Potts, J.D., Goodwin, R.L., Markwald, R.R., 2008. Periostin regulates atrioventricular valve maturation. Dev. Biol. 316, 200–213. Noseda, M., McLean, G., Niessen, K., Chang, L., Pollet, I., Montpetit, R., Shahidi, R., Dorovini-Zis, K., Li, L.H., Beckstead, B., Durand, R.E., Hoodless, P.A., Karsan, A., 2004. Notch activation results in phenotypic and functional changes consistent with endothelial-to-mesenchymal transformation. Circ. Res. 94, 910–917. Oka, T., Xu, J., Kaiser, R.A., Melendez, J., Hambleton, M., Sargent, M.A., Lorts, A., Brunskill, E.W., Gerald, W., Dorn, I., Conway, S.J.,

Chapter | 6.1

Molecular Regulation of Cushion Morphogenesis

Aronow, B.J., Robbins, J., Molkentin, J.D., 2007. Genetic manipulation of periostin expression reveals a role in cardiac hypertrophy and ventricular remodeling. Circ. Res. 101, 313–321. Okagawa, H., Markwald, R.R., Sugi, Y., 2007. Functional BMP receptor in endocardial cells is required in atrioventricular cushion mesenchymal cell formation in chick. Dev. Biol. 306, 179–192. Oosthoek, P.W., Wenink, A.C., Vrolijk, B.C., Wisse, L.J., DeRuiter, M.C., Poelmann, R.E., Gittenberger-de Groot, A.C., 1998. Development of the atrioventricular valve tension apparatus in the human heart. Anat. Embryol. 198, 317–329. Pandit, B., Sarkozy, A., Pennacchio, L.A., Carta, C., Oishi, K., Martinelli, S., Pogna, E.A., Schackwitz, W., Ustaszewska, A., Landstrom, A., Bos, J.M., Ommen, S.R., Esposito, G., Lepri, F., Faul, C., Mundel, P., Siguero, J.P., Tenconi, R., Selicorni, A., Rossi, C., Mazzanti, L., Torrente, I., Marino, B., Digilio, M.C., Zampino, G., Ackerman, M.J., Dallapiccola, B., Tartaglia, M., Gelb, B.D., 2007. Gain-of-function RAF1 mutations cause Noonan and LEOPARD syndromes with hypertrophic cardiomyopathy. Nat. Genet. 39, 1007–1012. Paruchuri, S., Yang, J.H., Aikawa, E., Melero-Martin, J.M., Khan, Z.A., Loukogeorgakis, S., Schoen, F.J., Bischoff, J., 2006. Human pulmonary valve progenitor cells exhibit endothelial/mesenchymal plasticity in response to vascular endothelial growth factor-A and transforming growth factor-beta(2). Circ. Res. 99, 861–869. Patten, B.M., Kramer, T.C., Barry, A., 1948. Valvular action in the embryonic chick heart by localized apposition of endocardial masses. Anat. Rec. 102, 299–311. Pierpont, M.E., Basson, C.T., Benson, W.D., Gelb, B.D., Giglia, T.M., Goldmuntz, E., McGee, G., Sable, C.A., Srivastava, D., Webb, C.L., 2007. Genetic basis for congenital heart defects: Current knowledge. Circulation 115, 3015–3038. Poelmann, R.E., Lie-Venema, H., Gittenberger-de Groot, A.C., 2002. The role of the epicardium and neural crest. Tex. Heart Inst. J. 29, 255–261. Porter Jr, G.A., Makuck, R.F, Rivkees, S.A., 2003. Intracellular calcium plays an essential role in cardiac development. Dev. Dyn. 227, 280–290. Portin, P., 2002. General outlines of the molecular genetics of the Notch signaling pathway in Drosophila melanogaster, a review. Hereditas 136, 89–96. Potts, J.D., Runyan, R.B., 1989. Epithelial–mesenchymal cell transformation in the embryonic heart can be mediated, in part, by transforming growth factor beta. Dev. Biol. 134, 392–401. Potts, J.D., Dagle, J.M., Walder, J.A., Weeks, D.L., Runyan, R.B., 1991. Epithelial-mesenchymal transformation of embryonic cardiac endothelial cells is inhibited by a modified antisense oligodeoxynucleotide to transforming growth factor beta 3. Proc. Nat. Acad. Sci. U.S.A. 88, 1516–1520. Potts, J.D., Vincent, E.B., Runyan, R.B., Weeks, D.L., 1992. Sense and antisense TGF beta 3 mRNA levels correlate with cardiac valve induction. Dev. Dyn. 193 (4), 340–345. Ranger, A.M., Grusby, M.J., Hodge, M.R., Gravallese, E.M., de la Brousse, F.C., Hoey, T., Mickanin, C., Baldwin, H.S., Glimcher, L.H., 1998. The transcription factor NF-ATc is essential for cardiac valve formation. Nature 392, 186–190. Razzaque, M.A., Nishizawa, T., Komoike, Y., Yagi, H., Furutani, M., Amo, R., Kamisago, M., Momma, K., Katayama, H., Nakagawa, M., Fujiwara, Y., Matsushima, M., Mizuno, K., Tokuyama, M., Hirota, H., Muneuchi, J., Higashinakagawa, T., Matsuoka, R., 2007. Germline gain-of-function mutations in RAF1 cause Noonan syndrome. Nat. Genet. 39, 1013–1017.

385

Ritz, B., Yu, F., Fruin, S., Chapa, G., Shaw, G.M., Harris, J.A., 2002. Ambient air pollution and risk of birth defects in Southern California. Am. J. Epidemiol. 155, 17–25. Rivera-Feliciano, J., Tabin, C.J., 2006. Bmp2 instructs cardiac progenitors to form the heart-valve-inducing field. Dev. Biol. 295, 580–588. Roberts, A.B., Tian, F., Byfield, S.D., Stuelten, C., Ooshima, A., Saika, S., Flanders, K.C., 2006. Smad3 is key to TGF-beta-mediated epithelial-to-mesenchymal transition, fibrosis, tumor suppression and metastasis. Cytokine Growth Factor Rev. 17, 19–27. Rodgers, L.S., Lalani, S., Hardy, K.M., Xiang, X.Y., Broka, D., Antin, P.B., Camenisch, T.D., 2006. Depolymerized hyaluronan induces vascular endothelial growth factor, a negative regulator of developmental epithelial-to-mesenchymal transformation. Circ. Res. 99, 583–589. Romano, L.A., Runyan, R.B., 1999. Slug is a mediator of epithelial– mesenchymal cell transformation in the developing chicken heart. Dev. Biol. 212, 243–254. Romano, L.A., Runyan, R.B., 2000. Slug is an essential target of TGF beta 2 signaling in the developing chicken heart. Dev. Biol. 223, 91–102. Runyan, R.B., Markwald, R.R., 1983. Invasion of mesenchyme into three dimensional collagen gels: A regional and temporal analysis of interaction in heart tissue. Dev. Biol. 95, 108–114. Runyan, R.B., Potts, J.D., Weeks, D.L., Sharma, R.V., Loeber, C.L., Chiang, J.J., Bhalla, R.C., 1990. Tissue interaction and signal transduction in the atrioventricular-canal of the embryonic heart. Ann. N.Y. Acad. Sci. 588, 442–443. Runyan, R.B., Potts, J.D., Weeks, D.L., 1992. TGF-beta 3-mediated tissue interaction during embryonic heart development. Mol. Repro. Dev. 32, 152–159. Sakabe, M., Ikeda, K., Nakatani, K., Kawada, N., Imanaka-Yoshida, K., Yoshida, T., Yamagishi, T., Nakajima, Y., 2006. Rho kinases regulate endothelial invasion and migration during valvuloseptal endocardial cushion tissue formation. Dev. Dyn. 235, 94–104. Sakata, Y., Koibuchi, N., Xiang, F., Youngblood, J.M., Kamei, C.N., Chin, M.T., 2006. The spectrum of cardiovascular anomalies in CHF1/Hey2 deficient mice reveals roles in endocardial cushion, myocardial and vascular maturation. J. Mol. Cell Cardiol. 40, 267–273. Sanford, L.P., Ormsby, I., Gittenberger-de Groot, A.C., Sariola, H., Friedman, R., Boivin, G.P., Cardell, E.L., Doetschman, T., 1997. TGFß2 knockout mice have multiple development defects that are non-overlapping with other TGFß knockout phenotypes. Development 124, 2659–2670. Schubbert, S., Zenker, M., Rowe, S.L., Boll, S., Klein, C., Bollag, G., van der Burgt, I., Musante, L., Kalscheuer, V., Wehner, L.E., Nguyen, H., West, B., Zhang, K.Y., Sistermans, E., Rauch, A., Niemeyer, C.M., Shannon, K., Kratz, C.P., 2006. Germline KRAS mutations cause Noonan syndrome. Nat. Genet. 38, 331–336. Sedmera, D., Hu, N., Weiss, K.M., Keller, B.B., Denslow, S., Thompson, R.P., 2002. Cellular changes in experimental left heart hypoplasia. Anat. Rec. 267, 137–145. Sheffield, V.C., Pierpont, M.E., Nishimura, D., Beck, J.S., Burns, T.L., Berg, M.A., Stone, E.M., Patil, S.R., Lauer, R.M., 1997. Identification of a complex congenital heart defect susceptibility locus by using DNA pooling and shared segment analysis. Hum. Mol. Genet. 6, 117–121. Sheng, W., Wang, G.Z., Wang, Y.L., Liang, J.Y., Wen, J.P., Zheng, P.S., Wu, Y.J., Lee, V., Slingerland, J., Dumont, D., Yang, B.B., 2005. The roles of versican V1 and V2 isoforms in cell proliferation and apoptosis. Mol. Biol. Cell 16, 1330–1340.

386

Shull, M.M., Ormsby, I., Kier, A.B., Pawlowski, S., Diebold, R.J., Yin, M., Allen, R., Sidman, C., Proetzel, G., Calvin, D., 1992. Targeted disruption of the mouse transforming growth factor-beta 1 gene results in multifocal inflammatory disease. Nature 359, 693–699. Sinning, A.R., Lepera, R.C., Markwald, R.R., 1988. Initial expression of type I procollagen in chick cardiac mesenchyme is dependent upon myocardial stimulation. Dev. Biol. 130 (1), 167–174. Sirard, C., de la Pompa, J.L., Elia, A., Itie, A., Mirtosos, C., Cheung, A., Hahn, S., Wakehem, A., Schartz, L., Kern, S.E., Rossant, J., Mak, T.W., 1998. The tumor suppressor gene Smad4/Dpc4 is required for gastrulation and later for anterior development of the mouse embryo. Genes Dev. 12, 107–119. Slevin, M., Krupinski, J., Kumar, S., Gaffney, J., 1998. Angiogenic oligosaccharides of hyaluronan induce protein tyrosine kinase activity in endothelial cells and activate a cytoplasmic signal transduction pathway resulting in proliferation. Lab. Invest. 78, 987–1003. Snider, P., Hinton, R.B., Moreno-Rodriguez, R.A., Wang, J., Rogers, R., Lindsley, A., Li, F., Ingram, D.A., Menick, D., Field, L., Firulli, A. B., Molkentin, J.D., Markwald, R., Conway, S.J., 2008. Periostin is required for maturation and extracellular matrix stabilization of noncardiomyocyte lineages of the heart. Circ. Res. 102 (7), 752–760. Somi, S., Buffing, A.A., Moorman, A.F., Van Den Hoff, M.J., 2004. Dynamic patterns of expression of BMP isoforms 2, 4, 5, 6, and 7 during chicken heart development. Anat. Rec. (Part A: Discoveries in Mol. Cellular and Evol. Biol.) 279A, 636–651. Song, W.M., Jackson, K., McGuire, P.G., 2000. Degradation of type IV collagen by matrix metalloproteinases is an important step in the epithelial–mesenchymal transformation of the endocardial cushions. Dev. Biol. 227, 606–617. Sorensen, L.K., Brooke, B.S., Li, D.Y., Urness, L.D., 2003. Loss of distinct arterial and venous boundaries in mice lacking endoglin, a vascularspecific TGF beta coreceptor. Dev. Biol. 261, 235–250. Stevens, M.V., Parker, P., Vaillancourt, R.R., Camenisch, T.D., 2006. MEKK4 regulates developmental EMT in the embryonic heart. Dev. Dyn. 235, 2761–2770. Stevens, M.V., Broka, D.M., Parker, P., Rogowitz, E., Vaillancourt, R.R., Camenisch, T.D., 2008. MEKK3 initiates transforming growth factor β2-dependent epithelial-to-mesenchymal transition during endocardial cushion morphogenesis. Circ. Res. 103, 1430–1440. Sugi, Y., Ito, N., Szebenyi, G., Myers, K., Fallon, J.F., Mikawa, T., Markwald, R.R., 2003. Fibroblast growth factor (FGF)-4 can induce proliferation of cardiac cushion mesenchymal cells during early valve leaflet formation. Dev. Biol. 258, 252–263. Sugi, Y., Yamamura, H., Okagawa, H., Markwald, R.R., 2004. Bone morphogenetic protein-2 can mediate myocardial regulation of atrioventricular cushion mesenchymal cell formation in mice. Dev. Biol. 269, 505–518. Sun, Y.P., Lowther, W., Kato, K., Bianco, C., Kenney, N., Strizzi, L., Raafat, D., Hirota, M., Khan, N.I., Bargo, S., Jones, B., Salomon, D., Callahan, R., 2005. Notch4 intracellular domain binding to Smad3 and inhibition of the TGF-beta signaling. Oncogene 24, 5365–5374. Tavares, A.L., Mercado-Pimentel, M.E., Runyan, R.B., Kitten, G.T., 2006. TGF-mediated RhoA expression is necessary for epithelial– mesenchymal transition in the embryonic chick heart. Dev. Dyn. 235, 1589–1598. Taylor, P.A., Batten, P., Brand, N.J., Thomas, P.S., Yacoub, M.H., 2003. The cardiac valve interstitial cell. Int. J. Biochem. Cell Biol. 35, 113–118.

PART | 6

Cushions, Valves and Septa

Tendijke, P., Yamashita, H., Ichijo, H., Franzen, P., Laiho, M., Miyazono, K., Heldin, C.H., 1994. Characterization of type-I receptors for transforming growth-factor-beta and activin. Science 264, 101–104. Timmerman, L.A., Grego-Bessa, J., Raya, A., Bertran, E., Perez-Pomares, J.M., Diez, J., Aranda, S., Palomo, S., McCormick, F., Izpisua-Belmonte, J.C., de la Pompa, J.L., 2004. Notch promotes epithelial-mesenchymal transition during cardiac development and oncogenic transformation. Genes Dev. 18, 99–115. Tkatchenko, A., Tkatchenko, T., Moreno-Rodriguez, R., Markwald, R., 2009. Lack of periostin leads to suppression of Notch1 signaling and calcific aortic valve disease. Physiol. Genom. 39, 160–168. Varadkar, P., Kraman, M., Despres, D., Ma, G., Lozier, J., McCright, B., 2008. Notch2 is required for the proliferation of cardiac neural crestderived smooth muscle cells. Dev. Dyn. 237, 1144–1152. Visconti, R.P., Ebihara, Y., LaRue, A.C., Fleming, P.A., McQuinn, T.C., Masuya, M., Minamiguchi, H., Markwald, R.R., Ogawa, M., Drake, C.J., 2006. An in vivo analysis of hematopoietic stem cell potential – Hematopoietic origin of cardiac valve interstitial cells. Circ. Res. 98, 690–696. Wall, N.A, Hogan, B.L., 1995. Expression of bone morphogenetic protein-4 (bmp-4), bone morphogenetic protein-7 (bmp-7), fibroblast growth factor-VIII (fgf-8) and sonic hedgehog (shh) during branchial arch development in the chick. Mech. Dev. 53, 383–392. Walsh, E.C., Stainier, D.Y., 2001. Udp-glucose dehydrogenase required for cardiac valve formation in zebrafish. Science 293, 1670–1673. Wang, J., Sridurongrit, S., Dudas, M., Thomas, P., Nagy, A., Schneider, M.D., Epstein, J.A., Kaartinen, V., 2005a. Atrioventricular cushion transformation is mediated by ALK2 in the developing mouse heart. Dev. Biol. 286, 299–310. Wang, X., Merritt, A.J., Seyfried, J., Guo, C., Papadakis, E.S., Finegan, K.G., Kayahara, M., Dixon, J., Boot-Handford, R.P., Cartwright, E.J., Mayer, U., Tournier, C., 2005b. Targeted deletion of mek5 causes early embryonic death and defects in the extracellular signal-regulated kinase 5/myocyte enhancer factor 2 cell survival pathway. Mol. Cell Biol. 25, 336–345. Watanabe, Y., Kokubo, H., Miyagawa-Tomita, S., Endo, M., Igarashi, K., Aisaki, K., Kanno, J., Saga, Y., 2006. Activation of Notch1 signaling in cardiogenic mesoderm induces abnormal heart morphogenesis in mouse. Development 133, 1625–1634. Webb, S., Anderson, R.H., Brown, N.A., 1996. Endocardial cushion development and heart loop architecture in the trisomy 16 mouse. Dev. Dyn. 206, 301–309. Webb, S., Brown, N.A., Anderson, R.H., 1998. Formation of the atrioventricular septal structures in the normal mouse. Circ. Res. 82, 645–656. Wessels, A., Markman, M.W., Vermeulen, J.L., Anderson, R.H., Moorman, A.F., Lamers, W.H., 1996. The development of the atrioventricular junction in the human heart. Circ. Res. 78, 110–117. Winnier, G.E., Blessing, M., Labosky, P.A., Hogan, B.L., 1995. Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 2105–2116. Winnier, G.E., Kume, T., Deng, K.Y., Rogers, R., Bundy, J., Raines, C., Walter, M.A., Hogan, B.L., Conway, S.J., 1999. Roles for the winged helix transcription factors MF1 and MFH1 in cardiovascular development revealed by nonallelic noncomplementation of null alleles. Dev. Biol. 213, 418–431. Wrana, J.L., Attisano, L., Wieser, R., Ventura, F., Massagué, J., 1994. Mechanism of activation of the TGF-b receptor. Nature 370, 341–347.

Chapter | 6.1

Molecular Regulation of Cushion Morphogenesis

Wunsch, A.M., Little, C.D., Markwald, R.R., 1994. Cardiac endothelial heterogeneity defines valvular development as demonstrated by the diverse expression of Jb3, an antigen of the endocardial cushion tissue. Dev. Biol. 165, 585–601. Yamada, M., Szendro, P.I., Prokscha, A., Schwartz, R.J., Eichele, G., 1999. Evidence for a role of Smad6 in chick cardiac development. Dev. Biol. 215, 48–61. Yamagishi, T., Nakajima, Y., Miyazono, K., Nakamura, H., 1999. Bone morphogenetic protein-2 acts synergistically with transforming growth factor-beta 3 during endothelial-mesenchymal transformation in the developing chick heart. J. Cell Physiol. 180, 35–45. Yamashita, H., TenDijke, P., Heldin, C.H., Miyazono, K., 1996. Bone morphogenetic protein receptors. Bone 19, 569–574. Yang, X.B., Tare, R.S., Partridge, K.A., Roach, H.I., Clarke, N.M., Howdle, S.M., Shakesheff, K.M., Oreffo, R.O., 2003. Induction of

387

human osteoprogenitor chemotaxis, proliferation, differentiation, and bone formation by osteoblast stimulating factor-1/pleiotrophin, Osteoconductive biomimetic scaffolds for tissue engineering. J. Bone Min. Res. 18, 47–57. Yarden, Y., Sliwkowski, M.X., 2001. Untangling the ErbB signaling network. Nat. Rev. Mol. Cell Biol. 2, 127–137. Yi, S.E., Daluiski, A., Pederson, R., Rosen, V., Lyons, K.M., 2000. The type IBMP receptor BMPRIB is required for chondrogenesis in the mouse limb. Development 127, 621–630. Zhang, H.B., Bradley, A., 1996. Mice deficient for BMP2 are nonviable and have defects in amnion chorion and cardiac development. Development 122, 2977–2986. Zhang, N., Mustin, D., Reardon, W., DeAlmeida, A., Mozdziak, P., Mrug, M., Eisenberg, L.M., Sedmera, D., 2006. Blood-borne stem cells differentiate into vascular and cardiac lineages during normal development. Stem Cells Dev. 15, 17–28.

Chapter 6.2

Signaling Pathways in Valve Formation: The Origin of Congenital Defects Joaquim Grego-Bessa1, José M. Pérez-Pomares2 and José Luis de la Pompa3 1 Departamento de Biología del Desarrollo Cardiovascular, Centro Nacional de Investigaciones Cardiovasculares (CNIC), Melchor Fernández Almagro 3, E-28029 Madrid, Spain 2 Departamento de Biología Animal, Facultad de Ciencias, Universidad de Málaga, Campus de Teatinos s/n, E-29071 Málaga, Spain 3 Departamento de Biología del Desarrollo Cardiovascular, Centro Nacional de Investigaciones Cardiovasculares (CNIC), Melchor Fernández Almagro 3, E-28029 Madrid, Spain

I.  Introduction Cardiac valves are crucial to adult heart function. The appearance of specialized chamber components in the vertebrate heart is linked to the development of sequential muscle contraction (coordinated by the conduction system), as well as to the formation of gating structures able to control blood flow effectively at the different pressures that arise in chambers and the vasculature at systole and diastole. This is already obvious in the hearts of fish, which have a welldeveloped valvular apparatus (Stainier et al., 2002). The progressive colonization of land by vertebrates correlated with a dramatic change in the organization of the cardiovascular system, which allowed adaptation to the new environment. The main consequence of this structural change, which coincided with the appearance of lung breathing, was a division of the bloodstream into two separate circuits, pulmonary and systemic. This physiological change led to a profound spatial reorganization of the simplest vertebrate heart design, that of the fish. The complexity of the vertebrate heart increased through time, with subdivision of the original valvular sets and the ancestral heart chambers (atrium and ventricle). Through this process, the heart was shaped into a sophisticated four-chambered organ that acts as an efficient pump, expelling blood from the main chamber reservoirs to the lungs and the rest of the body while managing large pressure differences between different cardiac regions. In parallel, developmental mechanisms responsible for the formation of the cardiac chamber, conduction system, valve and septum underwent spatio– temporal regulatory changes. These included refinements to the signaling networks that control cardiac cell determination, proliferation, differentiation and maturation. Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

In this chapter, we will focus on cardiac valve formation and the molecular signals that regulate this process (see also Chapter 6.1). To do so, we will analyze and discuss the molecular and cellular properties of the embryonic endocardium, as well as the specific molecular crosstalk between the endocardium and the myocardium. This intertissue crosstalk is an emerging and important integrative concept in cardiac development, supported by the recent description of fields of competence in the endocardium (Chang et al., 2004; Timmerman et al., 2004; Grego-Bessa et al., 2007), a finding compatible with the heterogeneity of region-specific myocardial signals and the ability of endocardial cells to respond to them (Krug et al., 1987; Mjaatvedt et al., 1987; Nakajima et al., 2000; Timmerman et al., 2004).

II.  Valve anatomy and function The four-chambered, fully-septated hearts of avians and mammals have the most sophisticated valve system of all vertebrates. Cardiac valves are crucial for coordination of an efficient blood supply to the body and lungs through two parallel circuits. Cardiac valves can be classified as atrioventricular (AV) and conoventricular (outflow tract, OFT) (Fig. 1). Heart valves are fibrous structures that originate, at least in part, from the endocardium that lines the corresponding myocardial segments; the endocardium of the atrioventricular and outflow tract regions is unique and has specific gene expression patterns (Fig. 1). The valves are not completely functional until late gestation (Larsen, 2001) and are fully mature only after birth (Kruithof et al., 2007). 389

390

PART  |  6  Cushions, Valves and Septa

Figure 1  Cardiac valve development in the mouse. (A–D, M–T) Hematoxylin/eosin (HE) staining; (E–H, U–Y) in situ hybridization; (I–L) immunohistochemistry. (A, B) General view and detail of a mouse E9.5 heart at the level of the atrioventricular canal (AVC). Arrowheads in (B) point to endocardium and the arrow to transformed mesenchymal cells in the cushion. (C, D) Outflow tract (OFT). (D) Detail showing endocardium (arrowhead) and transformed mesenchymal cells (arrow). (E–H) In situ hybridization showing NFATc1 transcription in atrioventricular canal (F, arrowheads) and outflow tract (H, arrowheads) endocardium. (I–L) Confocal images of E9.5 heart showing nuclear NFATc1 staining. (I, J) Atrioventricular canal. (J) Detail showing nuclear staining in atrioventricular canal endocardium (arrowheads). (K, L) NFATc1 staining in outflow tract endocardium (arrowheads in L). Note nuclear staining in ventricular endocardium (K, arrows). Nuclei are counterstained with DAPI. (M–P) E12.5 heart. (M, N) atrioventricular valves. (N) Detail of the developing mitral valve (MITV). Arrowheads point to endocardium and arrows to mesenchyme cells. (O, P) Aortic valve (AORV). (P) Detail of AORV with endocardium (arrowhead) and transformed cells (arrow). (Q) General view of E14.5 heart. (R) Detail of mitral valve (MITV) cusps. (S, T) Aortic valve. Arrowheads indicate endocardium and arrows, mesenchymal cells. (U–Y) In situ hybridization showing NFATc1 transcription in endocardium of MITV (U, V) and aortic valve endocardium (X, Y; arrowhead) (RV, right ventricle; LV, left ventricle).

From a structural point of view, atrioventricular and outflow tract valves differ (Figs 1; 3). Adult atrioventricular valves separate the atria from the ventricles. The mitral valve (left atrioventricular valve) controls blood flow between the left atrium and the left ventricle; the tricuspid valve (right atrioventricular valve) controls blood flow between the right atrium and the right ventricle. The adult

leaflets of the atrioventricular valves (two in the mitral valve and three in the tricuspid) are flattened laminar structures. At their base, they attach to the annulus fibrosus, a ring of fibrous tissue that insulates atrial and ventricular chambers. The annulus fibrosus appears in the ­atrioventricular region (Wessels et al., 1996), replacing the original embryonic atrioventricular myocardial sleeve (Lamers et al., 1995;

Chapter  |  6.2  Signaling Pathways in Valve Formation: The Origin of Congenital Defects

391

Figure 2  Endocardial–myocardial crosstalk in early valvular development. The atrioventricular canal region (box in (A)) is magnified in (B) to show the origin and basic signal sequence that initiates endocardial epithelial-to-mesenchymal transformation (EMT), valvuloseptal mesenchyme expansion, and other important events. Arrows indicate signal direction and numbers, the temporal hierarchy (1: myocardial activation of competent endocardium; 2: amplifying loop; 3: extracellular matrix modulation of signals). (C, D) General view of adult heart and a detail of the mitral (left) atrioventricular valve apparatus (AF: annulus fibrosus; AL: anterior leaflet of mitral valve; PL: posterior leaflet of mitral valve; PM: papillary muscles; TC: tendinous chords).

(A)

(C)

(B)

(D)

(E)

(F)

Figure 3  Cushion structure and development. (A, B) E9.5 and (C) E12.5 mouse embryos, scanning electron microscopy (SEM). The developing chambers of the embryonic heart are exposed by sectioning before SEM. The detail of (A) shown in (B) presents the complex fibrillar structure of the cardiac jelly (arrowheads), which constitutes the environment into which epithelial-to-mesenchymal-derived mesenchymal cells (arrow) immigrate. (C) Atrioventricular cushions, showing the ventral (superior) cushion (asterisk) and the dorsal (inferior) cushion (double asterisk). Note that the cushions are formed basically by a mass of mesenchymal cells covered by the atrioventricular endocardium (arrow). (D) Other mesenchymal cell types are found in the atrioventricular cushions. HH28 proepicardial quail-to-chick chimeras show epicardially-derived cells (green dots are QCPN-positive nuclei of donor epicardial derivatives) in the mesenchymal core of atrioventricular cushions (arrows). (E) At later stages, endocardial cushions (asterisks) fuse to remodel into the valvuloseptal mesenchyme. Cushion fusion is led by endocardial cell compatibility (labeled in green by an endocardially-bound FITCcoupled lectin) and the ability of vascular tissues to coalesce and fuse. (F) Transverse section through an HH36 stage quail aortic valve. The mature structure of this valve is characteristically semilunar, in contrast with the leaflet morphology of mature atrioventricular valves. Endothelial cells are stained with the QH1 antibody. Note a peritruncal coronary plexus around the valve (arrowheads). Main coronary roots penetrate artery walls and contact the vascular sinuses (arrow) (AORV: aortic valve; AV: atrioventricular valve; IVS: interventricular septum; LA: left atrium; LCS: aortic left coronary sinus; LV: left ventricle; NCS: aortic non-coronary sinus; OFT: outflow tract; RA: right atrium; RCS: aortic right coronary sinus; RV: right ventricle).

392

Wessels et al., 1996). At their tips, the valves are attached to chords of a tendinous nature (cordae tendineae or tendinous chords) that support the leaflets. These chords arise as an endocardial derivative (de Lange et al., 2004), although atrioventricular valve mural leaflets do not seem to derive completely from transformed endothelial cells (Lincoln et al., 2004). Tendinous chords are supported by thick expansions of the myocardium known as papillary muscles; together, tendinous chords and papillary muscles constitute the tension apparatus (Fig. 2D). In the adult human heart, the left ventricle has two main papillary muscles, anterior and posterior, which can be subdivided into smaller muscular strands. The right ventricle is slightly more complex, however, since an additional accessory papillary muscle (conal) attaches to the septal band and helps support the anterior and medial (septal) leaflets of the tricuspid valve. Furthermore, the right papillary muscles are inserted into the interventricular septum through the moderator band, whereas the left papillary muscles insert more caudally around the apex (Anderson and Becker, 1980). The common outflow tract embryonic valvular primordia remodel into a double set of valves (aortic and pulmonary). These valves have a characteristic semilunar design, with three swallow-nest cusps per valvular set (Figs 1; 3). The cusps are fibrous in nature and apparently differentiate from the excavation of the mesenchymal core in the conal cushions (Hurle et al., 1980).

III.  Early embryogenesis of heart valves: the origins of cardiac valve cell populations The early embryonic heart (around E8.5 in the mouse) mainly consists of two concentric tissue layers, the outer contractile myocardium and the inner endocardium. They are separated by a thick extracellular matrix (ECM) layer, known as the cardiac jelly (Eisenberg and Markwald, 1995). This cardiac jelly is secreted by the myocardium (Krug et al., 1985). It is evenly distributed in the primary embryonic heart, but soon after cardiac looping it becomes thinner in the atrial and ventricular segments, remaining as a wide extracellular matrix-filled space only in the atrioventricular and outflow tract areas. These cardiac jelly remnants mark the initiation of cardiac valve formation (Figs 1; 2). At around E9.5 in the mouse, atrioventricular and outflow tract cardiac jelly spaces begin to be populated by mesenchymal cells whose origin has long remained unknown. The most relevant finding for our understanding of cardiac valve development was the involvement of an epithelial-to-mesenchymal transformation (EMT) in valvuloseptal mesenchyme formation (Markwald et al., 1975, 1977) (Fig. 2). Epithelial-to-mesenchymal transformation is a finely-regulated cellular mechanism (Thiery and Sleeman, 2006) that results in the complete conversion

PART  |  6  Cushions, Valves and Septa

of an epithelial cell to a mesenchymal cell. Epithelialto-mesenchymal transformation requires active downregulation of cell-to-cell adhesion molecules (including cadherins), degradation of the basement membrane and extreme remodeling of the cytoskeleton, all of which correlate with the loss of the characteristic baso-apical polarity of epithelial cells (Hay, 2005). Study of the mechanism of cardiac cushion epithelial-to-mesenchymal transformation was greatly facilitated by the use of a three-dimensional type I collagen gel explant assay introduced by Markwald and colleagues (Bernanke and Markwald, 1982) (see example in Fig. 5). In this system, atrioventricular canal or outflow tract regions are isolated before epithelialto-mesenchymal transformation and explanted onto the surface of the gel, after which a subset of endocardial cells transform into mesenchyme and invade the collagen matrix. Experimental studies using this explant assay in chick and mouse embryos indicate the potential of the atrioventricular and outflow tract endocardium (but not of other endocardial populations) to undergo epithelial-tomesenchymal transformation and form mesenchymal cells (Bernanke and Markwald, 1982; Runyan and Markwald, 1983; Krug et al., 1987; Mjaatvedt and Markwald, 1989; Dor et al., 2001; Camenisch et al., 2002b; Timmerman et al., 2004; Chapter 6.1). The process of endocardial epithelial-to-mesenchymal transformation is divided in three steps, defined using the collagen gel culture system (Bernanke and Markwald, 1982). The first step is the activation of endocardial cells that swell and become hypertrophic due to enlargement of the Golgi and rough endoplasmic reticulum. The second step is transformation of endocardial cells, which lose their cell-to-cell contacts, acquire the ability to move within the plane of the endocardial layer, polarize the Golgi, and form migratory appendages such as filopodia and lamellipodia. Finally, the transforming endocardium migrates from the endocardial cell layer into the cardiac jelly. Two main cushions, ventral and dorsal, initially appear in the common atrioventricular canal (Fig. 3). Considering their relative cranio–caudal position in the human embryo, most authors refer to them as the superior and inferior cushions, respectively (Kim et al., 2001a). At later stages (mouse E12.5), two additional small cushions appear at the left and right sides of the atrioventricular canal, with an important role in the final formation of the valve leaflets. Conal (OFT) cushions also originate after a local endocardial epithelial-to-mesenchymal transformation (Krug et al., 1995); epithelial-to-mesenchymal transformation-derived mesenchyme distribution is restricted to the more proximal area of the outflow tract (Kisanuki et al., 2001; de Lange et al., 2004). Four cushions (two proximal and two distal) have been described in mammals (Ya et al., 1998). Distal cushions of the mammalian outflow tract participate in the formation of the aortic and pulmonary semilunar valves, as is also the case in avian embryos (Qayyum et al., 2001).

Chapter  |  6.2  Signaling Pathways in Valve Formation: The Origin of Congenital Defects

IV.  Late embryogenesis of valves Once the original superior and inferior atrioventricular cushions become fully cellularized, they grow until their endocardial surfaces make contact, initiating a process of fusion between the two cushions (Fig. 3E). This tissue fusion represents the joining of two independent tissue units to form a more complex anatomical element (PerezPomares and Foty, 2006). It is evident that this fusion is driven by the natural ability of endothelial cell populations to fuse with one another (Drake and Little, 1995). Cushion fusion is also critical for development of the interventricular and atrioventricular septum, since the final fusion between different areas of the cushions and the membranous portions of the septum not only determines the future shape of the adult valve leaflets, but is also crucial to the completion of cardiac septation. Formation of the mitral and tricuspid valves is a complex process. The mitral atrioventricular valve which separates the left atrium from the left ventricle has two leaflets, an aortic (anterior) leaflet and a mural or parietal (posterior) leaflet. The tricuspid atrioventricular valve has a septal (medial) leaflet and two mural leaflets, one posterior (caudal) and another anterior (cranial) (de Lange et al., 2004). Right and left mural leaflets develop in association with myocardial protrusions that express Tbx3, an atrioventricular canal myocardium marker (Hoogaars et al., 2004), and are negative for characteristic cardiac ­ chamber myocardium markers such as Cx40, Cx43 (Dupays et al., 2005) and Nppa (encoding atrial natriuretic factor) (Small and Krieg, 2003). As these protrusions increase and expand, their ventricular side (formed by an irregular trabecular layer) expresses markers of ventricular myocardium, but not of atrioventricular canal myocardium. The lateral atrioventricular cushions then become apparent through local formation of mesenchyme, which grow to fuse with the main inferior and superior cushions (de Lange et al., 2004). The septal leaflet of the tricuspid (right atrioventricular) valve and the aortic leaflet of the mitral (left atrioventricular) valve form mainly from the inferior and superior atrioventricular cushions, respectively; once these cushions have fused with the membranous tip of the interventricular septum (around E14.5 in the mouse), the tricuspid septal leaflet forms on the right side of the valve. In the mouse this leaflet, supported by the septal myocardium, lies over the dorsal side of the interventricular septum which expresses Tbx3 (Hoogaars et al., 2004). The aortic mitral leaflet is not attached to myocardium except at its cranial and caudal margins, exactly where the papillary muscles will form (de Lange et al., 2004). Valve leaflets form through different mechanisms. The septal leaflet of the tricuspid valve forms by detachment of the atrioventricular cushions from the ventricular myocardial wall. In contrast, the aortic and mural leaflets of the mitral valve and the mural leaflet of the tricuspid valve form by ­ proliferation

393

of the atrioventricular myocardium and expansion of the underlying ventricular myocardium, which gives rise to a mobile leaflet (de Lange et al., 2004). The myocardial layer of the valves disappears late in development, probably due to apoptosis (de Lange et al., 2004). The fate of all the cells that form part of the original valvulo–septal mesenchyme population is far from being completely understood. It is known that cardiac valves progressively decellularize. Many of the cells remaining in the leaflets are endocardial derivatives (de Lange et al., 2004; Lincoln et al., 2004), but it also seems clear that some of these cells contribute to the annulus fibrosus (see below) and the forming tendinous chords (de Lange et al., 2004). Apoptosis might be an important mechanism in the regulation of cushion cell numbers, but the location of endocardial-derived cells in “noncushion” areas such as the annulus fibrosus suggests a cell migration from the cushion to other heart locations (de Lange et al., 2004). The neural crest also contributes to valve development, as neural crest-derived cells are found in outflow tract cushion tissue (Jiang et al., 2000; de Lange et al., 2004; see Chapters 7.1 and 7.2) although their contribution to atrioventricular cushions or valvular primordia is debated (de Lange et al., 2004; Nakamura et al., 2006). Nevertheless, the mature leaflets of the outflow tract valves lack neural crestderived cells (Jiang et al., 2000). More recent studies indicate that outflow tract leaflets are derived exclusively from the endocardium (de Lange et al., 2004), suggesting that endocardial-derived mesenchyme replaces neural crestderived cells at late stages of valve development.

V.  Signaling pathways and effectors of endocardial epithelial-tomesenchymal transformation and valve morphogenesis Transformation of atrioventricular and outflow tract endocardium is a key step in cardiac valve development. This process is under the control of a region-specific myocardial signal produced by the atrioventricular and outflow tract myocardium (Mjaatvedt et al., 1987). In addition, only atrioventricular canal and outflow tract endocardium are competent to undergo epithelial-to-mesenchymal transformation, as shown by experiments in which ventricular endocardium incubated with atrioventricular canal myocardium did not undergo epithelial-to-mesenchymal transformation (Runyan and Markwald, 1983; Mjaatvedt et al., 1987). These studies demonstrate that only atrioventricular and outflow tract myocardium secretes specific epithelialto-mesenchymal transformation-inducing signals, and only atrioventricular canal and outflow tract endocardium is competent to transform into heart valve primordia. In this section, we will review the role of various signaling pathways and effector mechanisms in

394

e­ pithelial-to-­mesenchymal transformation and cardiac valve development (see also Chapter 6.1). Interactions between pathways are summarized in Fig. 6.

V.A.  The TGF Superfamily V.A.i.  TGF Many members of the transforming growth factor beta (TGF) family of growth factors are key components of the epithelial-to-mesenchymal transformation-inductive stimulus. Chick atrioventricular endocardium cultured with ventricular myocardium plus exogenous TGF1 and TGF2 induces endocardial epithelial-to-mesenchymal transformation in vitro (Potts and Runyan, 1989). In the mouse, TGF1 is expressed in endocardium (Akhurst et al., 1990), while TGF2 is expressed in atrioventricular and outflow tract myocardium (Dickson et al., 1993) (see Fig. 4M,N). The endocardium-derived mesenchymal cells that populate the atrioventricular endocardial cushions between E9.0 and E10.0 are also weakly TGF2-positive (Camenisch et al., 2002a). In contrast, TGF2 in the chick is expressed in both endocardium and myocardium (Boyer et al., 1999a). Inhibition of chick TGF signaling with a pan-anti-TGF antibody prevents epithelial-to-mesenchymal transformation (Potts and Runyan, 1989). These data indicate that TGF is an atrioventricular-specific epithelialto-mesenchymal transformation inducer. TGF3 expressed in endocardium has a dual function in chick, acting as an autocrine signal to sustain mesenchyme formation, and as a paracrine signal to amplify the initial myocardial inductive event (Ramsdell and Markwald, 1997; Nakajima et al., 1998). Further studies showed that TGF2 is required for endocardial cell activation and TGF3 for mesenchymal cell invasion in the chick (Boyer et al., 1999a). Only TGF2-mutants have specific defects in valves and septa (Sanford et al., 1997). These data from chick and mouse concur with studies using neutralizing antibodies in atrioventricular canal explants. In avian explants, TGF2- and TGF3-blocking antibodies inhibit endocardial cell activation and transformation, respectively. In contrast, TGF2 blockade inhibits cell transformation in the mouse, whereas TGF3 inhibition has no effect (Camenisch et al., 2002a). In addition, expression of TGF3 in E12 mouse endocardial cushions suggests that it may be involved in valve remodeling (Camenisch et al., 2002a; see Chapter 6.1). TGF2 signaling in chick endocardial cushions is mediated by the transcriptional repressor slug (snail2), which belongs to the snail/slug family. These zinc-finger transcription factors are expressed in various mesenchymal cell populations and are required for epithelial-tomesenchymal transformation processes throughout development (Nieto et al., 1992, 1994). In the chick, slug mRNA and protein are found in the outflow tract and atrio­ ventricular canal cardiac cushions from HH stages 14–20

PART  |  6  Cushions, Valves and Septa

(Romano and Runyan, 1999), suggesting its involvement in cardiac epithelial-to-mesenchymal transformation. Slug expression persists until at least HH stage 30, primarily in the transformed mesenchyme of atrioventricular and outflow tract valves (Carmona et al., 2000). Antisense oligonucleotides to slug inhibit epithelial-to-mesenchymal transformation in atrioventricular canal explant cultures by blocking endocardial cell activation (Romano and Runyan, 1999). TGF2-blocking antibodies, known to inhibit atrioventricular canal epithelial-to-mesenchymal transformation at a similar stage (Boyer et al., 1999b), reduce slug expression in atrioventricular canal explants (Romano and Runyan, 1999), suggesting that TGF2 acts upstream of slug during cardiac epithelial-to-mesenchymal transformation. Moreover, slug overexpression rescues cell transformation in cultures treated with anti-TGF2 antibodies, further indicating that slug signals downstream of TGF2 (Romano and Runyan, 1999). In the section dedicated to Notch, we will describe further evidence for slug/snail involvement in cardiac epithelial-to-mesenchymal transformation (see also Fig. 4).

V.A.ii.  Bone Morphogenetic Proteins Other TGF superfamily members such as the bone morphogenetic proteins (BMP) are expressed in developing cardiac cushions. BMP2, BMP4, BMP5, BMP6 and BMP7 are expressed in the atrioventricular canal and outflow tract during initiation of epithelial-to-mesenchymal transformation. BMP2 is restricted to the atrioventricular canal and outflow tract myocardium before and during epithelial-to-mesenchymal transformation (Lyons et al., 1990). BMP2 is sufficient to induce epithelial-to-mesenchymal transformation in mouse atrioventricular canal endocardium explants cultured in the absence of myocardium (Sugi et al., 2004). Treatment of atrioventricular canal endocardium explants with BMP2 and the BMP inhibitor Noggin blocks mesenchymal cell transformation. Noggin also blocks epithelial-to-mesenchymal transformation in atrioventricular endocardium cultured with associated atrioventricular myocardium (Sugi et al., 2004). BMP2 treatment of atrioventricular explants results in increased TGF2 protein levels, indicating that TGF2 is a downstream target of BMP2 in atrioventric­ular cushion formation (Sugi et al., 2004). Conventionally-targeted BMP2-mutant embryos die at E8.5, prior to cardiac cushion formation (Zhang and Bradley, 1996), which complicates analysis of the role of BMP2 in cushion morphogenesis. Recent studies in mice with conditional BMP2 deletion in cardiac progenitors (driven by Nkx2-5CRE) indicate that BMP2 is necessary for specification of the heart valve-inducing region, myocardial patterning and EMT (Ma et al, 2005; Rivera-Feliciano and Tabin, 2006), processes which may include cooperation with other signaling pathways such as Notch (see below). BMP4 is also expressed in cardiac cushions during the initiation of epithelial-to-mesenchymal transformation.

Chapter  |  6.2  Signaling Pathways in Valve Formation: The Origin of Congenital Defects

It does not appear to be involved in cardiac cushion epithelial-to-mesenchymal transformation, but rather is involved in growth and remodeling of cushions into mature valvular and septal structures. Conventionally-targeted BMP4-mice die very early in embryonic development (Winnier et al., 1995). Nonetheless, analysis of mice with different levels of wild-type BMP4 expression in cardiomyocytes defined a critical role for BMP4 in heart valve development (Jiao et al., 2003). A modest reduction in BMP4 activity leads to partial atrioventricular septation defects; more severe reductions in myocardial BMP4 activity result in smaller atrioventricular cushions and complete atrioventricular septation defects, with an atrioventricular valve similar to those of some Down syndrome patients. The atrioventricular septation defects in these mouse models appear to result from decreased cell proliferation within the atrioventricular cardiac cushions (Jiao et al., 2003). In mice, BMP5 is expressed in the myocardium before and during cardiac cushion epithelial-to-mesenchymal transformation (Solloway and Robertson, 1999). BMP6 expression is noted primarily in murine outflow tract myocardium at the onset of epithelial-to-mesenchymal transformation (Dudley and Robertson, 1997). Later in development, BMP6 expression persists in outflow tract myocardium, and expression becomes detectable in the atrioventricular canal but not in the outflow tract cushion mesenchyme (Kim et al., 2001b). BMP7 is expressed throughout the myocardium of the linear heart tube (Dudley and Robertson, 1997), and is later expressed weakly in atrioventricular valve mesenchyme (Kim et al., 2001b). Conventional inactivation of BMP5, BMP6, or BMP7 alone does not produce cardiac defects (Dudley and Robertson, 1997). Several groups have generated distinct BMP mutant combinations to further study its role in cardiac cushion development. BMP5;BMP7 double mutant mice do not form cardiac cushions (Solloway and Robertson, 1999). Determination of the precise role(s) of BMP5 and BMP7 in cushion morphogenesis is complicated by a severe delay and general disorganization in development (Solloway and Robertson, 1999). BMP6;BMP7 double mutant mice show a marked delay in formation of outflow tract cushions, due to reduced cell proliferation (Kim et al., 2001b).

V.A.iii.  TGF Receptors and Smad Signaling Mediators TGF signals are transduced via two serine-threonine kinase domain-containing transmembrane receptors, TRI and TRII (TGF receptors I and II (Wrana et al., 1994)). Binding of TGF ligands to TRII results in phosphorylation of TRI, which in turn activates Smad intracellular signaling mediators that activate TGF-dependent gene transcription (Shi and Massague, 2003). TRIII, also

395

known as betaglycan, can bind and present TGF ligands to the TRI and TRII complex. TRIII has a short intracellular tail and no identified intracellular signaling function (Kretzschmar and Massague, 1998). Gene inactivation of TRII in mice provided no insight into the role of this receptor in cardiac cushion development, as TRII-targeted mice die before endocardial cushion formation (Oshima et al., 1996). Functional studies in the chick showed that antibodies to TRII inhibit epithelial-to-mesenchymal transformation in atrioventricular canal explants (Brown et al., 1996). TRII is expressed throughout the vascular endothelium, as well as the endocardium (Brown et al., 1996). The localized transformation event in cardiac cushions may be due to restricted TRIII expression in the chick endocardium and transformed mesenchymal cells of developing cushions (Brown et al., 1999). Anti-TRIII function-blocking antibodies inhibit mesenchymal cell formation; TRIII thus seems to be a necessary component for cardiac cushion epithelialto-mesenchymal transformation. TRIII overexpression in ventricular explants induces enhanced epithelial-tomesenchymal transformation in the presence of TGF2 (Brown et al., 1999). Increased TRIII expression in ventricles could either permit TGF2-mediated epithelialto-mesenchymal transformation in the ventricular endocardium, or enhance invasion in atrioventricular endocardial cell explants. Either way, TRIII is functional during epithelial-to-mesenchymal transformation. Efforts are still under way to explain the mechanism of TRIII signaling during cardiac cushion epithelial-to-mesenchymal transformation, and to determine whether TRII and TRIII are also essential for epithelial-to-mesenchymal transformation in the mouse. Endoglin is another TGF superfamily receptor that binds to TGF1 and TGF3 and, similarly to TRIII, presents these ligands to TRII (Cheifetz et al., 1992). Endoglin is expressed in all endothelial cells, including endocardium, and endoglin-mutant mice show vascular defects (Bourdeau et al., 1999). In endoglin-mutant mice, the atrioventricular canal endocardial cushions do not undergo epithelial-to-mesenchymal transformation, resulting in hypocellular cushions at E9.5; these mice die of cardiac insufficiency at E10.5 (Bourdeau et al., 1999). Endoglin is necessary for epithelial-to-mesenchymal transformation in the atrioventricular canal cardiac cushions, but the ligand(s) that bind endoglin in cardiac cushions have yet to be identified. Functional studies of factors downstream of BMP provide additional support for a critical role of BMP signaling during endocardial cushion morphogenesis. Mice with a hypomorphic BMP receptor II allele (Bmpr2) die before birth as a result of cardiovascular abnormalities (Delot et al., 2003). Bmpr2 interacts with activin receptor-like kinase (ALK-3) (Shi and Massague, 2003), and can bind BMP-2, BMP-4, or BMP-7 to activate an intracellular

396

cascade involving Smad intracellular signaling mediators (Shi and Massague, 2003). Mice with hypomorphic Bmpr2 mutations show persistent truncus arteriosus and semilunar valve defects (Delot et al., 2003). Epithelial-to-mesenchymal transformation is initiated in both the atrioventricular and outflow tract endocardial cushions, which argues that Bmpr2 is not necessary for epithelial-to-mesenchymal transformation. As outflow tract cushions did not develop after the initial epithelial-to-mesenchymal transformation events, however, an obvious conclusion is that Bmpr2 is required for subsequent growth and maintenance of the conotruncal cushions (Delot et al., 2003). Persistent truncus arteriosus can be caused by inhibition of correct cardiac neural crest contribution to the distal outflow tract, resulting in absence of the aorticopulmonary septum (Kirby et al., 1983). Some cardiac neural crest cells enter the outflow tract in mice with hypomorphic Bmpr2 mutations, leading to speculation that BMP signaling may be important in regulating the interaction between migrating neural crest and cells intrinsic to the developing outflow tract (Delot et al., 2003). Myocardial-specific inactivation of the Bmpr2 signaling partner ALK3 also results in cardiac defects (Gaussin et al., 2002); interestingly, these are observed in the atrioventricular junction and not in the outflow tract cushions (Gaussin et al., 2002; Delot et al., 2003). Initial steps of cardiac cushion development are normal in both the atrio­ ventricular and outflow tract cushions, suggesting that ALK3 is dispensable for epithelial-to-mesenchymal transformation. After epithelial-to-mesenchymal transformation, however, the cushions are reduced in size, and inferior and superior cushions do not fuse (Gaussin et al., 2002). Increased cell death in ALK3-mutant cardiomyocytes may explain this abnormal atrioventricular canal phenotype. ALK3 deletion also results in diminished TGF2 expression in the atrioventricular canal, suggesting that BMP signaling in the myocardium regulates TGF2 expression in the atrioventricular canal (Gaussin et al., 2002). This is consistent with previous data from mouse atrioventricular canal explant assays (Sugi et al., 2004). Further insight into the role of ALK3 in valve development and maturation was obtained from the study of mice with ALK3 deletion in atrioventricular canal myocardium, which showed defective atrioventricular valve morphogenesis and ventricular pre-excitation reminiscent of Ebstein’s anomaly in man (Gaussin et al., 2005). Intracellular TGF signaling mediators are also necessary for cell transformation in developing endocardial cushions. Eight distinct Smad proteins are known, divided into three functional classes: (1) the receptor-activated R-Smads (Smad 1, 2, 3, 5, and 8); (2) the co-mediator Co-Smad (Smad 4); and (3) the inhibitory I-Smads (Smad6 and 7) (Feng and Derynck, 2005; Euler-Taimor and Heger, 2006). All three functional classes of Smads and nearly all isoforms have been detected in the cardiovascular system.

PART  |  6  Cushions, Valves and Septa

In nonactivated cells, R-Smads are predominantly localized in the cytoplasm, Co-Smads are distributed equally between cytoplasm and the nucleus, and I-Smads are found mostly in the nucleus. TGF superfamily receptor kinases directly phosphorylate and activate R-Smads; Smad2 and Smad3 respond to TGF subfamily signaling, whereas R-Smads Smad1, Smad5 and Smad8 respond to BMP signaling (Feng and Derynck, 2005). Activated R-Smads in both the TGF and BMP pathways form heterodimers with Smad4. These Smad4/R-Smad heterodimer complexes translocate into the nucleus to activate gene transcription (Shi and Massague, 2003). Smad6 negatively regulates BMP signaling by competing for Smad1 binding of Smad4 (Hata et al., 1998). Smad6 is expressed in the atrio­ ventricular and outflow tract regions of the heart during development (Galvin et al., 2000) and may modulate BMP signaling in these areas because it preferentially inhibits BMP signaling (Shi and Massague, 2003). Smad6 mutant mice have hypercellular atrioventricular and outflow tract cardiac cushions (Galvin et al., 2000), a phenotype consistent with BMP mediation of epithelial-to-mesenchymal transformation or subsequent mesenchymal cell proliferation within cardiac cushions (Galvin et al., 2000). It would seem that BMP-induced mesenchyme formation or proliferation is controlled by a negative feedback mechanism involving Smad6. Smad6 was proposed to block signaling by another TGF ligand expressed in the vicinity, thus limiting epithelial-to-mesenchymal transformation (Galvin et al., 2000). Further insight into cardiac epithelial-tomesenchymal transformation regulation was provided by a study in chick cardiac explants, which showed that expression of a constitutively active ALK2 receptor induces epithelial-to-mesenchymal transformation in nontransforming ventricular endocardium. In addition, Smad6 overexpression inhibits epithelial-to-mesenchymal transformation in atrioventricular canal endocardium (Desgrosellier et al., 2005), suggesting that Smad6 may act downstream of ALK2 to negatively-regulate epithelial-to-mesenchymal transformation (Desgrosellier et al., 2005).

V.B.  Vascular Endothelial Growth Factor (VEGF) Vascular endothelial growth factor (VEGF) (Carmeliet et al., 1996; Ferrara et al., 1996) and its tyrosine kinase receptors VEGF-R2/Flk1 (Shalaby et al., 1995), VEGFR1/Flt1 (Fong et al., 1995) and VEGF-R3/Flt4 (Dumont et al., 1998) are critical for early endothelial cell differentiation, vasculogenesis and angiogenesis. During embryonic development, VEGF is present in various tissues from the late blastocyst stage (Miquerol et al., 1999) and is already expressed in the heart by E8.0 (Miquerol et al., 1999). At E9.0–9.5, VEGF expression spans the entire length of the heart tube (Miquerol et al., 1999), and at E10.5 it is

Chapter  |  6.2  Signaling Pathways in Valve Formation: The Origin of Congenital Defects

restricted to the atrioventricular canal and outflow tract myocardium (Dor et al., 2001). Cardiovascular development is strongly dependent on normal levels and timely expression of VEGF, as ­demonstrated by gain- (Miquerol et al., 2000) and loss-of-function studies (Carmeliet et al., 1996; Ferrara et al., 1996). A slight increase in VEGF expression thus leads to cardiac abnormalities such as a thinner compact layer myocardium, overproduction of trabeculae, defective trabecular septation and abnormal outflow tract remodeling (Miquerol et al., 2000). Mice with inducible VEGF expression in myocardium provided insight into the function of this growth factor in valve development, as its induction in the E9.5 myocardium leads to abnormal expansion of the endocardium, which does not transform into cushion tissue (Dor et al., 2001). Addition of VEGF to the medium of E9.5 atrioventricular explants almost completely inhibits mesenchymal transformation and collagen gel invasion (Dor et al., 2001). Flk1 and Flt1 are both expressed throughout the endocardium, suggesting that VEGF secreted by the myocardium can transduce signals in the adjacent endocardium. Neither Flk1 nor Flt1 are expressed in transformed mesenchymal cells (Dor et al., 2001), indicating that after transformation cushion cells become unresponsive to VEGF which signals exclusively to endocardium. The observation that VEGF expression is induced in wild-type embryos only after the onset of transformation is consistent with a role for VEGF in the negative feedback regulation of epithelial-to-mesenchymal transformation during cardiac valve development (Dor et al., 2001). The working model for the role of VEGF in the regulation of epithelial-to-mesenchymal transformation is as follows: a myocardially-derived TGF-related signal induces epithelial-to-mesenchymal transformation. After cellularization of the cardiac jelly, myocardial-derived VEGF acts on the endocardium to promote an epithelial phenotype and attenuate endocardial competence for epithelial-tomesenchymal transformation. VEGF may thus indirectly regulate cellularity of cushion tissue; in the endocardium, it stimulates local proliferation to replenish cells lost by transformation (Dor et al., 2001). This model was recently refined, as VEGF was found to induce epithelialto-mesenchymal transformation in E10.5 atrioventricular explants (Hallaq et al., 2004). This suggests that VEGF is required in the later stages of epithelial-to-mesenchymal transformation to stimulate endocardial proliferation and replenishment of transformed endocardial cells, allowing epithelial-to-mesenchymal transformation to continue.

V.C.  Calcineurin and Nuclear Factor of Activated T-Cells (NFATc) The signaling pathway downstream of VEGF during cardiac cushion formation has been studied intensively. VEGF induces nuclear translocation of the transcription factor

397

NFAT (nuclear factor of activated T-cells) in endothelial cells (Armesilla et al., 1999), suggesting that NFAT is a key mediator of VEGF signaling in the endocardium. NFAT transcription factors are well-known ­regulators of development and adult physiology in mammals (Graef et al., 2001a). Their activity is controlled by cycles of dephosphorylation and phosphorylation that determine the amount of the transcription factor in the nucleus, and thus the final transcriptional response (Hogan et al., 2003). The phosphatase calcineurin B is responsible, following increases in intracellular Ca2, for the rapid dephosphorylation and subsequent nuclear entry of the four NFAT gene products (NFATc1–c4). Several kinases participate in the phosphorylation step that acts as a switch-off mechanism (Hogan et al., 2003). The first suggestion of NFAT transcription factor involvement in valve development came from the analysis of NFATc1-targeted mice. Early in development, NFATc1 is expressed throughout the endocardium (Fig. 1E–H), although it is progressively restricted to the endocardium of valveforming territories (de la Pompa et al., 1998; Ranger et al., 1998) (Fig. 1I–L,U–Y). Although epithelial-to-mesenchymal transformation occurs in NFATc1-mutant mice, they die at E14.5 from congestive heart failure derived from valvulo-septal defects, suggesting that NFATc1 is required for valve morphogenesis (de la Pompa et al., 1998; Ranger et al., 1998). Calcineurin B is expressed in both endocardium and myocardium, and analysis of endothelium-specific (Tie2-CRE) calcineurin B-mutant mice indicated defective valvular morphogenesis, similar to that of NFATc1 mutants; this suggests that endocardial calcineurin B is specifically required to regulate NFATc1 activity during valve elongation and remodeling (Chang et al., 2004). Transgenic rescue of the NFATc1-defective valvular phenotype with an endocardial-specific NFATc1 transgene (Tie2-NFATc1) further supports this notion (Chang et al., 2004). Calcineurin B activity is inhibited by the drug cyclosporine A (Emmel et al., 1989). Treatment of wild-type embryos with this drug between E10 and E13 showed that endocardial calcineurinB/NFATc1 activity is required at E11 (Chang et al., 2004), the time at which NFATc1 activity is restricted to the cushion endocardium where valve formation occurs (de la Pompa et al., 1998; Ranger et al., 1998). NFATc1 and endocardial-specific calcineurin B-mutants undergo normal epithelial-to-mesenchymal transformation in E9.5 atrioventricular explants (Chang et al., 2004; GregoBessa et al., unpublished data), further indicating that these molecules are not required in the endocardium for epithelial-to-mesenchymal transformation. NFATc2, NFATc3 and NFATc4 are expressed in cardiomyocytes; however, their individual inactivation does not impair cardiac development (Oukka et al., 1998; Graef et al., 2001b; Wilkins et al., 2002; Bourajjaj et al., 2008). In contrast, E10.5 NFATc2:c3:c4-triple-mutants show significantly reduced numbers of mesenchymal cells in the atrioventricular

398

canal and outflow tract, similar to conventional calcineurin B-mutants, indicating that calcineurin regulates NFATc2/c3/ c4 in the myocardium to induce epithelial-to-mesenchymal transformation (Chang et al., 2004). The observation that embryos lacking myocardial NFATc signaling did not undergo mesenchymal transformation suggested that myocardial NFATc is required to produce an epithelial-to-mesenchymal transformation inducer or to suppress an inhibitor of epithelial-to-mesenchymal transformation. Transcript array analysis indicated that VEGF was strongly upregulated in calcineurin B- and NFATc2:c3:c4mutant mice (Chang et al., 2004). As VEGF inhibits epithelial-to-mesenchymal transformation (Dor et al., 2004), it was proposed to be the myocardial NFATc-dependent inhibitor of epithelial-to-mesenchymal transformation. Indeed, EMSA showed that NFATc3 and NFATc4 directly repress the VEGF promoter (Chang et al., 2004). The current working hypothesis suggests that at E9.0 calcineurin/NFATc signaling in the atrioventricular canal myocardium represses VEGF, allowing endocardial epithelial-to-mesenchymal transformation to occur in response to TGF inductive signals (Chang et al., 2004). At E10.5, VEGF expression is upregulated in atrioventricular canal myocardium, leading to epithelial-to-mesenchymal transformation termination, endocardial cell proliferation and calcineurin/NFATc1 activation in the endocardium (Dor et al., 2001), which directs cardiac valve elongation and morphogenesis (Chang et al., 2004). It is not yet clear how VEGF is upregulated in the myocardium so that NFATc1 is activated in the endocardium and valve morphogenesis occurs.

V.D.  Neuregulin 1 (NRG1) and ErbB3 The neuregulins (NRG) are a group of secreted glycoproteins with epidermal growth factor (EGF)-like domains that signal through Erb receptors, a family of tyrosine kinase transmembrane receptors of the EGFR family (Lemke, 1996; Riese and Stern, 1998). NRG1 binds with high affinity to the ErbB3 and ErbB4 receptors (Lemke, 1996; Falls, 2003). ErbB2 itself does not bind to NRG1, but is activated through heterodimerization with other ErbB family members (Yarden and Sliwkowski, 2001). NRG1 is expressed in the endocardium from E8.5 onwards (Rentschler et al., 2002). Targeted inactivation of NRG1 disrupts ventricular trabeculation, and endocardial cushion development is also severely affected at E10.5 (Meyer and Birchmeier, 1995). The ErbB3 receptor is expressed in the endocardium and mesenchyme of cardiac cushions (Meyer and Birchmeier, 1995; Erickson et al., 1997). Targeted inactivation of ErbB3 results in a reduction of mesenchymal cells in the atrioventricular and outflow tract cardiac cushions at E9.5, and in hypoplastic, underdeveloped valves at E13.5 (Erickson et al., 1997). An insight into the mechanism of NRG1/ErbB3 signaling in cushion development came from studies of the role of the

PART  |  6  Cushions, Valves and Septa

glycosaminoglycan hyaluronan (HA) in atrioventricular canal morphogenesis. Inactivation of the hyaluronan synthase 2 (Has2) gene, which encodes an essential HA synthase, abrogates cushion formation and epithelial-tomesenchymal transformation (Camenisch et al., 2000; Chapter 7.1). This defect was reproduced by expression of a dominant-negative RAS in wild-type heart explants, and reversed in Has2 mutant explants by gene rescue, by administering exogenous HA, or by expressing activated RAS. Conversely, exogenous HA-mediated transformation in Has2 mutant explants was inhibited by dominantnegative RAS (Camenisch et al., 2000). In these explants, rescue of epithelial-to-mesenchymal transformation by HA is associated with phosphorylation of ErbB3 (Camenisch et al., 2002b). Addition of NRG1 to Has2 mutant explants also leads to ErbB2 and ErbB3 phosphorylation and rescues epithelial-to-mesenchymal transformation, and ErbB2 or ErbB3 inhibitors block NRG1 rescue of epithelial-tomesenchymal transformation in these mutants (Camenisch et al., 2002b). These data link extracellular matrix HA and NRG1/ErbB2/ErbB3 signals that converge on RAS activation during development and proliferation of cardiac valve and septal mesenchyme. Whether HA interacts directly with ErbB3 receptors or if it activates ErbB3 signaling by an indirect mechanism is still under study.

V.E.  Sox9 Sox9 encodes a high-mobility group transcription factor expressed in various tissues during development, including endocardial cushions (Ng et al., 1997; Zhao et al., 1997). In humans, heterozygous Sox9 mutations cause campomelic dysplasia, a disease characterized by hypoplasia of the endochondrial bones, XY sex reversal, kidney and pancreas defects, as well as ventricular septal defects and tetralogy of Fallot (Houston et al., 1983). In the heart, Sox9 is expressed exclusively in mesenchymal cells of atrioventricular and outflow tract endocardial cushions (Akiyama et al., 2004). Sox9-mutant mice die from congestive heart failure at E12.5 and show severely underdeveloped cardiac cushions associated with reduced cell proliferation. Accordingly, expression of ErbB3, which is involved in proliferation and differentiation of endocardial cushion mesenchyme (Erickson et al., 1997), is greatly reduced in Sox9 mutants (Akiyama et al., 2004). In addition, NFATc1 is ectopically expressed in the mesenchyme of Sox9mutant embryos, suggesting that Sox9 promotes a mature mesenchymal phenotype via NFATc1 repression (Akiyama et al., 2004). To test the involvement of defective neural crest development in outflow tract cushion defects, floxed Sox9 mice were bred onto a Wnt1-CRE background. Sox9 deletion in neural crest lead to defective distal, but not proximal, outflow tract development, demonstrating that Sox9 is required for the contribution of neural crest and endocardial cells to cushion mesenchyme (Akiyama et al., 2004).

Chapter  |  6.2  Signaling Pathways in Valve Formation: The Origin of Congenital Defects

V.F.  Gata4 The zinc-finger transcription factor Gata4 is essential for heart formation (Kuo et al., 1997; Molkentin et al., 1997). Gata4 heterozygous mutations in man are associated with defects in the ventricular or atrial septum and, to a lesser extent, with valvular pulmonary stenosis (Garg et al., 2003; Okubo et al., 2004; Hirayama-Yamada et al., 2005). In addition to the myocardium, Gata4 is also expressed in the endocardium and the endocardial cushion (Heikinheimo et al., 1994). This expression pattern and defective endocardial cushion development in mouse embryos homozygous for two different hypomorphic Gata4 alleles (Crispino et al., 2001; Pu et al., 2004) suggested that Gata4 is an important regulator of cushion development. Gata4 inactivation in endocardium using the Tie2-CRE driver line (Kisanuki et al., 2001) results in a marked decrease in the number of mesenchymal cells in atrioventricular cushions, probably caused by the reduction in expression of ErbB3, a direct Gata4 target (Rivera-Feliciano et al., 2006). Gata4 is required at two stages of atrioventricular valve formation first, to promote endocardial cell epithelial-tomesenchymal transformation to generate atrioventricular cushion mesenchyme and second, in endocardial-derived cells for growth and remodeling of the atrioventricular cushions during atrioventricular valve maturation. In contrast, outflow tract cushions are formed from two cell sources: outflow tract endocardium, which undergoes epithelial-to-mesenchymal transformation and will give rise to mesenchyme of the proximal outflow tract cushions; and neural crest, which contributes to mid- and distal-outflow tract cushions. Gata4 is expressed in both the neural crest- and endocardial-derived portions of outflow tract cushions. Gata4 is not required in the neural crest-derived portion for normal outflow tract development, but is necessary to form the proximal, endocardial-derived portion of the outflow tract cushion (Rivera-Feliciano et al., 2006). The relatively early death of Gata4floxed;Tie2CRE mice nonetheless precludes assessment of what effect the loss of this portion of outflow tract cushion has on outflow tract valve development. Congenital abnormalities of the pulmonary valve are associated with mutation of human Gata4 (Garg et al., 2003; Okubo et al., 2004; Hirayama-Yamada et al., 2005), suggesting that Gata4 activity within the proximal, endothelial-derived portion of the outflow tract cushions may be needed for outflow tract valve development (Rivera-Feliciano et al., 2006).

V.G.  Notch Notch is an ancient local cell signaling system, as both ligands (Delta and Jagged) and receptors (Notch) are membrane-bound (Artavanis-Tsakonas et al., 1999). After ligand–receptor interaction, a signal is transduced to the nucleus via three consecutive receptor cleavage steps

399

(Mumm and Kopan, 2000) to generate the Notch intracellular domain (NICD), the biologically-active molecule. In the nucleus, NICD binds to the RBPJK/CSL effector and regulates target gene expression (Artavanis-Tsakonas et al., 1999). Notch activity is dependent on cell context, and may maintain cells in a progenitor state or promote their differentiation (Lewis, 1998; Radtke and Raj, 2003; Louvi and Artavanis-Tsakonas, 2006). Specific Notch ligands and receptors are expressed in the heart from early developmental stages. Delta4 (Krebs et al., 2000), Notch1 (Del Amo et al., 1992), and Notch4 (Uyttendaele et al., 2000) are transcribed in the endocardium from gastrulation onwards, whereas other ligands and receptors show restricted myocardial expression from mid-gestation (Loomes et al., 1999; McCright et al., 2002). The Notch target genes HRT1/Hey1 and HRT2/Hey2 are expressed in the endocardium and/or myocardium at different stages of cardiogenesis (Nakagawa et al., 1999). Studies in Xenopus (Rones et al., 2000) and in mouse embryonic stem cells (Nemir et al., 2006) indicate that cardiomyogenic commitment and differentiation require Notch signaling inhibition. In vivo studies in chick (Rutenberg et al., 2006), mouse (Timmerman et al., 2004) and zebrafish (Milan et al., 2006) nevertheless indicate that abrogation of Notch signaling affects neither primary cardiac cell fate determination nor differentiation (Milan et al., 2006). In the mouse E9.5 heart, Notch1 activity delineates the atrioventricular canal and ventricular endocardium (Fig. 4A,B) (del Monte et al., 2007). In RBPJk-targeted mutants, Notch1 activity is greatly reduced (Fig. 4C,D) and valve development is severely disrupted, presumably because of defective endocardial maturation and signaling (Timmerman et al., 2004). RBPJk and Notch1 mutants have a collapsed endocardium and lack mesenchymal cushion cells, indicating that endocardial epithelial-tomesenchymal transformation is defective in these mutants. Ultrastructural analysis of E9.5 mutant atrioventricular canal endocardium reveals that cells remain in close association, abnormally maintaining adherens junctions, and do not invade the cardiac jelly, although they show features of activated premigratory endocardial cells. These observations correlate with a specific reduction in transcription of the snail repressor, which is normally expressed in the endocardium and mesenchyme of the atrioventricular canal region (Timmerman et al., 2004) (Fig. 4E–H). Concomitantly, VE-cadherin expression that is downregulated in wild-type atrioventricular canal (Fig. 4I,J) and outflow tract endocardium remains abnormally expressed at these sites in RBPJk or Notch1 mutants (Fig. 4K,L); this suggests that lack of snail expression prevents VE-cadherin downregulation in this tissue, blocking endocardial epithelial-to-mesenchymal transformation. These findings show that Notch activity is required for endocardial epithelialto-mesenchymal transformation (Timmerman et al., 2004). Interestingly, TGF2 expression in atrioventricular canal

400

PART  |  6  Cushions, Valves and Septa

snail

RBPKO

WT

N1ICD

TGFβ2

VE-Cadherin

(A)

(E)

(I)

(M)

(B)

(F)

(J)

(N)

(C)

(G)

(K)

(O)

(D)

(H)

(L)

(P)

Figure 4  Notch1 activity in AVC endocardium is associated with snail and TGF2 expression and reduced VE-cadherin expression. (A–D) Notch1 activity detected with an antibody against the activated receptor (N1ICD) shows nuclear staining in wild-type AVC�������������������������� ����������������������������� endocardium (A, B, arrowhead); the signal is greatly reduced in endocardium of RBPJk mutants (C, D, arrowhead). Nuclei are counterstained with DAPI. (E–P) In situ hybridization. (E–H) Wild-type snail1 expression in atrioventricular canal endocardium (F, arrowhead) and transformed mesenchyme cells (F, arrow), and greatly reduced signal in RBPJk mutants (G, H). (I–J) Wild-type VE-cadherin expression in endocardium of ventricles and atria (I); no expression in AVC���������������������������������������������� endocardium (J, arrowhead). (K, L) Abnormal VE-cadherin expression pattern in AVC endocardium of RBPJk mutants (K, L, arrowhead). (M–P) Wild-type TGF2 expression in atrioventricular canal myocardium (M, N, arrows) and strong reduction in RBPJk mutants (O, P, arrows).

myocardium (Fig. 4M,N) is reduced in RBPJk mutants (Fig. 4O,P), suggesting that endocardial Notch function is needed for production of the myocardial TGF2 signal. Atrioventricular canal explant assays with Notch1 or RBPJk mutants demonstrate defective endocardial epithelial-to-mesenchymal transformation (Fig. 5). This finding is supported by Notch inhibition experiments that impair valve development in zebrafish, and by gain-of-function experiments in which transient overexpression of N1ICD in heart leads to formation of hypertrophic atrioventricular valves (Timmerman et al., 2004). This result contrasts with more recent zebrafish data indicating that restricted N1ICD overexpression in the atrioventricular canal endocardium inhibits epithelial-to-mesenchymal transformation (Beis et al., 2005). This discrepancy may be due to ectopic N1ICD RNA expression in both endocardium and myocardium (Timmerman et al., 2004), which may generate additional myocardial-derived epithelial-to-mesenchymal transformation-inductive signals. Conditional activation of Notch1 in the mouse cardiac lineage using a Mesp1-CRE driver impairs atrioventricular canal myocardial differentiation and ventricular myocardium maturation, but

epithelial-to-mesenchymal transformation occurs normally (Watanabe et al., 2006). Thus, the precise role of Notch in epithelial-to-mesenchymal transformation regulation remains to be understood. HRT1 and HRT2 have also been linked to cardiac development. HRT2-targeted mutant mice have several cardiac anomalies, and show growth retardation and death within 10 days of birth. Surviving HRT2 mutants have enlarged atria and ventricles, and echocardiographic analysis revealed abnormal cardiac hemodynamics, including tricuspid valve stenosis and regurgitation, mitral valve regurgitation, VSD and secundum atrial septal defect (Kokubo et al., 2004). The phenotypic variation reported in HRT2-mutant mice probably results from the use of different genetic backgrounds and/or functional redundancy between HRT2 and other HRT family members (Fischer et al., 2004). Analysis of HRT2 mutants indicates that this gene is required for atrioventricular valve formation, suggesting that it may be a key Notch effector for valve development. Mice lacking both HRT1 and HRT2 die during embryogenesis due to severe cardiovascular malformations, including impaired development of atrioventricular

Chapter  |  6.2  Signaling Pathways in Valve Formation: The Origin of Congenital Defects

(B)

WT

(A)

(C)

Notch1KO

(D)

(F)

(G) RBPKO

(E)

(H)

(I)

Figure 5  Impaired EMT in E9.5 Notch pathway mutant AVC� ���� explants. (A–C) Wild-type explants. The cytoskeleton is phalloidinTRITC-stained; nuclei are DAPI-counterstained. (A) General view of explant. (B) Detail showing elongated transformed cells (arrows) and activated cells (arrowhead). (C) Phalloidin-TRITC and DAPI stainings. (D–F) Notch1 mutant. (E) Detail showing rare elongated transformed cells (arrow) and more abundant activated cells (arrowhead). (F) Phalloidin-TRITC and DAPI staining. (G–I) RBPJk mutant. (H) Detail showing endocardium with activated cells (arrowhead). (I) PhalloidinTRITC and DAPI staining. (m) myocardium; (e) endocardium.

cushions. Few cells undergo epithelial-to-mesenchymal transformation in the HRT1:HRT2 mice (Kokubo et al., 2005), suggesting that HRT1 and HRT2 function synergistically in this process. Ectopic HRT1 or HRT2 expression in the cardiac lineage of chick (Rutenberg et al., 2006) or mice (Kokubo et al., 2007) leads to a severe reduction of the atrioventricular canal territory. Marker analysis indicates that HRT1 and HRT2 may regulate atrioventricular boundary formation via repression of the T-box family member Tbx2. It remains to be understood how Notch and HRT genes intersect to regulate atrioventricular canal development and epithelial-to-mesenchymal transformation.

V.H.  Neurofibromatosis Type 1 (Nf1) Von Recklinghausen neurofibromatosis or neurofibromatosis type 1 (NF1) is one of the most common inherited

401

human diseases, affecting one in every 3,500 individuals. NF1 is characterized by proliferation and malignant transformation of neural crest derivatives. Affected individuals also have disorders that seem unrelated to the neural crest, including higher incidence of congenital heart disease, especially valvular pulmonic stenosis (Lin et al., 2000). The gene responsible for this disorder, NF1, encodes neurofibromin, which has RAS GTPase-activating activity and can downregulate RAS signaling. Standard inactivation of Nf1 in mice leads to midgestation lethality from cardiovascular anomalies. These defects include structural outflow tract malformations and enlarged endocardial cushions (Brannan et al., 1994; Jacks et al., 1994). Histological analysis revealed an overabundance of tissue in the outflow tract region, a thinned myocardium, and abnormal cardiac morphogenesis manifested in a double-outlet right ventricle. This defect is associated with human syndromes and animal models of abnormal cardiac neural crest function. Thus, the cardiac defect of Nf1-mutant mice was proposed to be related to neural crest dysfunction (Brannan et al., 1994; Jacks et al., 1994). In a later report, Lakkis and Epstein demonstrated that the overabundant tissue that accumulates and obstructs blood flow in the heart of Nf1-mutant embryos results from enhanced epithelial-to-mesenchymal transformation and cushion proliferation in the atrioventricular canal and outflow tract. Interestingly, abnormal epithelial-to-mesenchymal transformation was reproduced in explant cultures derived from E10.5 mutants, a time prior to neural crest migration into the cardiac region. In addition, forced RAS activation can mimic the Nf1-mutant phenotype in cardiac explants, and RAS inhibition can rescue it (Lakkis and Epstein, 1998). These findings led to a model in which growth factor activation of a tyrosine kinase receptor on endothelial cells triggers an intracellular RAS-mediated signaling cascade and contributes to epithelial-to-mesenchymal transformation, invasion and proliferation. Downregulation of the intracellular RAS-transduced signals is mediated by the GAP activity of NF1, which is expressed by transformed endothelial cells that enter the jelly and eventually undergo terminal differentiation into mature valve structures (Lakkis and Epstein, 1998). Endothelial-specific Nf1 inactivation recapitulates the cardiovascular abnormalities observed in the standard mutants, which involve deletion in both the endocardial cushions and myocardium (Gitler et al., 2003). This phenotype is associated with an elevated RAS signaling in Nf1 mutant endothelial cells and pre­ cocious nuclear accumulation of NFATc1. This transcription factor is required for cardiac valve morphogenesis (de la Pompa et al., 1998; Ranger et al., 1998) (see previous section), and RAS activity can modulate NFATc1 nuclear localization in other cell types (Woodrow et al., 1993). Nevertheless, NFATc1 involvement in the development of cell growth abnormalities in Nf1 mutants has not been clarified.

402

PART  |  6  Cushions, Valves and Septa

EMT Capacitation/induction Endoc.

Smad2,3

Smad1,5,8 TBRI TBRII BMPRII ALK5 ALK2 ALK3

NF1

Delta-Notch

Endoglin

TGFβ1

TGFβ3

EMT Progression/ mesenchyme development

EMT initiation

TBRI TBRII TBRIII

Snail1/2

RAS NFATc1-GATA4

VE-C

NRG1, Has2 Flk1/KDR

EMT

EC remodeling

ErbB2/3, HA HA NFATc1

TGFβ2,3*

Sox9 EC

TGFβ3 BMP6 EC proliferation Smad6

BMP2/4/5/6/7

Myoc.

TGFβ2*

ALK3, BMPRII

VEGF

Has2

TGFβ3 Calcineurin/NFATc2-4

Figure 6  Signaling pathways and effectors in cardiac valve development and morphogenesis. Valve formation is divided into three stages, based on functional studies in mice and chick: EMT��������������������������������������������������������������������������������������������� ������������������������������������������������������������������������������������������������ capacitation; EMT��������������������������������������������������������������������������� ������������������������������������������������������������������������������ initiation; and EMT������������������������������������������������������� ���������������������������������������������������������� progression/mesenchyme development. Various signaling pathways and transcription factors act in coordination to regulate valve development. Blunt arrows indicate inhibitory effects and discontinuous arrows indicate indirect effects. Asterisks indicate molecules expressed in different tissues in the valvular regions and with potentially different roles in mouse and chick. The colored rows refer to the endocardium (red), cardiac jelly (yellow) and the myocardium (blue). Note that colors darken from left to right, indicating temporal progression. For more details, see text.

The contribution of neural crest to defective cardiac development in Nf1 mutants was examined by breeding the Nf1 conditional with different neural crest-specific drivers. Consistent with the tissue specificity of the different Cre lines, there was hyperplasia of tissues of neural crest origin, including massively enlarged sympathetic ganglia and neural crest-derived tumors (Gitler et al., 2003). The PNS specificity of the phenotypes observed in these mice indicated that Nf1 function in the neural crest is not required for heart development (Gitler et al., 2003). Figure 6 summarizes the signaling pathways and effectors that regulate cardiac valve development and morphogenesis.

VI.  Normal and abnormal signaling in valve development: the origin of congenital defects Congenital heart defects are extremely common in humans (Hoffman, 1995), and many of these defects relate directly or indirectly to the development of cardiac valves and septa. In this section, we will attempt a systematic exploration and classification of abnormal cardiac valve phenotypes, and relate these phenotypes to reported anomalous molecular signaling in embryonic valve development.

Special emphasis will be made on translating molecular signaling into cellular dynamics, to explain the tissue mechanics that underlie normal and abnormal heart valve development. Congenital heart disease is often analyzed in the perspective of abnormal gene expression, and only early gross defects in embryonic cushions are presented and/or discussed as substrates for malformed adult valves. It follows, however, that a large percentage of studies based on manipulation of gene expression show extremely severe phenotypes whose relationship to congenital heart disease is difficult to interpret. These phenotypes are often incompatible with the progression of embryonic development, possibly representing extreme manifestations of the abnormal molecular and cellular events that underlie congenital heart disease. In other cases, subtle changes in the dose of certain molecules cause anomalies that resemble many characteristic congenital heart disease phenotypes. Some of these malformations are nonetheless perceived as cardiac defects only after birth; in these cases, dissection of the origin of embryonic defects is more difficult. Defective gene activity affects critical cell processes in valve formation. Among these, the most important are endocardial epithelial-to-mesenchymal transformation, cushion mesenchyme proliferation and apoptosis, ­cushion

Chapter  |  6.2  Signaling Pathways in Valve Formation: The Origin of Congenital Defects

fusion and tissue maturation (development of leaflet fibrous tissue and differentiation between leaflet and chord tissues). These coordinated events allow endocardial cushions to remodel into a fibrous leaflet correctly attached to its supporting apparatus. Coordination depends on the spatio–temporal patterning of many transcriptional regulators that signal through complex cascades. Altering the function of any of these components can lead to dysfunctional valves.

VI.A.  Altered Developmental Mechanisms and Their Morphological Reflection VI.A.i.  Epithelial-to-Mesenchymal Transformation The endocardial epithelial-to-mesenchymal transformation that mediates formation of the valvuloseptal mesenchyme is probably the most crucial event in valve development. Abolition of endocardial epithelial-to-mesenchymal transformation is incompatible with the progression of heart morphogenesis after cardiac looping, as evidenced by mutations in signaling pathways critical to endocardial epithelial-to-mesenchymal transformation, such as TGF/ BMP (Nakajima et al., 2000; Delot, 2003) and Notch (Timmerman et al., 2004). In most cases, abnormal cardiac performance leads to substantial alteration in blood circulation, and often to edemas and hemorrhages. Impaired epithelial-to-mesenchymal transformation, which causes a marked reduction in mesenchymal cell numbers in the cushions, also directly affects the formation of valves, which tend to be smaller.

VI.A.ii.  Cell Proliferation and Death Proliferation of the epithelial-to-mesenchymal transformation-derived mesenchyme is another important step in cardiac valve formation. The number of mesenchymal cells in cushions increases progressively during a specific period of valve development (Keyes and Sanders, 1999), such that the size of valve primordia is appropriate to the region in which they are found (atrioventricular canal and outflow tract). Cushion size depends not only on cell number, but also on production of extracellular matrix proteins and other elements. A real paradox in cardiac valve development is the relevance of cushion mesenchymal cells to cardiac valve morphogenesis, especially since adult valve leaflets have very low cell numbers. It is nonetheless evident that valvulo-septal cells are needed to complete the production of the extracellular matrix molecules that form the fibrous tissue characteristic of the mature leaflets. Cells are also an ideal substrate for modulating the acquisition of form through mechanisms such as local cell proliferation and apoptosis, which might help control

403

the size and shape of the different valve parts. A defect in valvulo-septal mesenchymal cell numbers might underlie some cases of atresia of the atrioventricular, aortic and pulmonary valves, whereas an excess of extracellular matrix secretion could explain the formation of dysplastic valves prone to regurgitation.

VI.A.iii.  Fusion of Cushions and Valve Calcification A direct consequence of cushion size is the extent of fusion of the valvular primordia, although reduced cushion fusion can cause many cardiac anomalies. The most evident result of impaired fusion is generally abnormal valve morphology or deficient ventricular chamber septation of the so-called “membranous” type, that is, an anomaly affecting the nonmuscular tissue of the septum. Hyperfusion of cushions can also lead to the formation of abnormal valvular structures, with a reduced number of leaflets or cusps compared to the normal situation. Some of these malformations are compatible with life, but abnormal structures are more likely than normal valves to act as a substrate for acquired disease (e.g., calcification or endocarditis) (Gregoratos, 2003; Garg, 2006). Calcification of the aortic valve is the third leading cause of heart disease in adults. Incidence increases with age and is associated with a bicuspid aortic valve in 1−2% of the population. The relevance of NOTCH in human cardiac valve development and homeostasis was demonstrated by the finding that altered NOTCH1 signaling leads to aortic valve disease (Garg et al., 2005). NOTCH1 mutations cause a spectrum of developmental aortic valve anomalies and severe valve calcification in nonsyndromic autosomal dominant human pedigrees (Garg et al., 2005). Studies in mice show that, similarly to Notch1, the Notch target genes HRT1 and HRT2 are expressed in aortic valve leaflets at E17.5, where these genes repress Runx2, a regulator of osteoblast fate. These results suggest that NOTCH1 mutations cause an early developmental defect in the aortic valve and later de-repression of calcium deposition, which causes progressive aortic valve disease (Garg et al., 2005). Whereas the role of NOTCH1 in preventing aortic valve calcification in adult-onset disease is incompletely understood, its essential function in normal valve development is also intriguing (Timmerman et al., 2004). Bicuspid and unicuspid aortic valves typically have a ridge at which the valve leaflets fail to separate in utero. In extreme cases, blood flow may be so restricted that the left ventricle fails to grow; this results in hypoplastic left heart syndrome, the most frequent cause of death in children with congenital heart disease. The discovery of NOTCH1 as a cause of bicuspid aortic valve and hypoplastic left ventricle in a single family suggests NOTCH1 mutation as the genetic basis for these defects in some patients. Future studies of

404

NOTCH1 mutations in this population may reveal those at risk for a subset of severe congenital heart lesions (Garg, 2006).

VI.A.iv.  Tissue Maturation Lineage analysis showed that the mature atrioventricular valves are derived mainly from the endocardial cushions originated through epithelial-to-mesenchymal transformation (de Lange et al., 2004). In mice, atrioventricular cushion remodeling results in the formation of mesenchymal leaflets at E14.5 (de Lange et al., 2004). Little is known, however, about the development of these immature leaflets into the adult, stress-resistant valves supported by tendinous chords and papillary muscles. Recent work led to a model for postnatal atrioventricular valve development whereby condensation of mesenchymal cells starts at the atrial side of the leaflet at E15.5 and expands throughout the leaflet at E18.5. Cell proliferation contributes to leaflet elongation until postnatal day 4.5. Rapid growth of the heart might then elongate leaflets by physically pulling at the papillary muscle insertion points. The valves then lose cell density, followed by massive remodeling of the leaflet extracellular matrix, including fibronectin, and collagen I and V accumulation (Kruithof et al., 2007). The cardiac valve regions also express molecules involved in chondrogenic cell differentiation, such as Sox9 (Montero et al., 2002; Akiyama et al., 2004) and periostin (Kruzynska-Frejtag et al., 2001; Kern et al., 2005). This might help explain the mesenchymal condensations, as well as the expression of cartilage-related molecules (Lincoln et al., 2004) and even the cartilage foci in several heart regions (Lopez et al., 2000), which was attributed to neural crest cells in the area (Sumida et al., 1989). Key molecules such as FGF (fibroblast growth factor) and BMP were proposed to regulate cardiac valve cell diversification by regulating chondrogenic and tendinous differentiation, respectively (Lincoln et al., 2006). Many abnormalities related to atrioventricular valve positioning and the concordance between the tendinous chords and the papillary muscles might be explained by disruption of some of these pathways.

VI.B.  Abnormal Developmental Processes in Cardiac Valve Formation Defective cushion development leads to several complex heart malformations (Sumida et al., 1989). The majority of congenital heart diseases are of a complex nature; the specific etiologies are thus unknown, and the developmental dynamics of the malformed heart is poorly-understood (Ransom and Srivastava, 2007). Congenital defects that alter valve anatomy include complete atrioventricular canal defect, tricuspid valve atresia, pulmonary atresia

PART  |  6  Cushions, Valves and Septa

(often linked to tricuspid valve atresia), pulmonary stenosis, bicuspid aortic valve, Ebstein’s anomaly, atrial and ventricular septal (membranous) defects and hypoplastic left heart syndrome. Abnormal valve morphogenesis leading to valve dysplasia can also manifest as stenotic valves or valves prone to prolapse in both syndromic and nonsyndromic dysplasia. The most patent of the atrioventricular septal defects (“endocardial cushion defects”) is the persistence of a common atrioventricular canal, which is accompanied by retention of a common atrioventricular valve with five leaflets (Lamers et al., 1995). Partial remodeling of cushion tissue can also account for mitral valve stenosis (Ruckman and Van Praagh, 1978), mitral valve arcade (Castaneda et al., 1969) and parachutelike asymmetric mitral valves (Oosthoek et al., 1997), a malformation in which the chords of the atrioventricular valves are shorter, thicker or fewer than normal. This anomaly alters atrioventricular valve form and function, as the leaflets are directly attached to the papillary muscles. Valve development is closely associated with cardiac septation (Lamers et al., 1995). Many septal defects are purely muscular, but others affect the membranous portion of the septal structures. These perimembranous defects of the interventricular septum appear after deficient alignment of the mesenchymal cap at the crest of the interventricular septum with the atrioventricular cushions and the valvular tissue of aortic and pulmonary trunks. Whereas these septal disorders are not necessarily due to anomalies in early valve development, some alignment defects may be caused by abnormalities in valvular primordia.

VI.B.i.  The Paradigmatic Example of Ebstein’s Anomaly No other valvular anomaly summarizes the developmental mechanisms involved in the formation of the delicate cardiac valves as clearly as Ebstein’s anomaly. This condition is characterized by a general downward displacement of the right atrioventricular (tricuspid) valve into the right ventricle, and defective morphogenesis of the valvular structure including an anterior fenestrated leaflet, and posterior and septal leaflets that are thick, hypoplastic, and directly adhered to the right ventricular wall (Fig. 7) (Lie et al., 1979; Attenhofer Jost et al., 2005). The molecular substrate of Ebstein’s anomaly is not known, although efforts have been made to map chromosomes in animal models for this disease (Andelfinger et al., 2003). It is suggested that Ebstein’s malformation is caused by the failure of the larger parts of the tricuspid valve leaflets to delaminate from the ventricular walls during ventricular development (Kanani et al., 2005). In a recent study (Gaussin et al., 2005) ALK3 was specifically deleted in the atrioventricular myocardium. As a result, the tricuspid mural leaflet and the mitral septal leaflet developed abnormally and were displaced with respect to

Chapter  |  6.2  Signaling Pathways in Valve Formation: The Origin of Congenital Defects

(A)

405

(B)

(C)

C'

Figure 7  Adult atrioventricular valve structure. (A, B) Adult mouse heart stained with Masson-Goldner’s trichrome. Myocardial cells are stained dark red and fibrous components, light green. The left atrioventricular mural valve region (box) is enlarged in (B). The elongated morphology of the leaflet (arrows) and the layering of the fibrous components of the valve into an elastic atrialis (white arrowheads) and fibrous ventricularis (black arrowheads) are shown. Nodular thickenings are indicated by a double black arrowhead. The annulus fibrosus is marked with an asterisk. Note the discontinuity between the atrial and ventricular myocardium. (C, C) A case of Ebstein’s anomaly. The displacement of the valve leaflets (arrows) with respect to the atrioventricular canal (white asterisks) is obvious in the human heart shown in (C). The detail in (C) illustrates the anomalous myxoid texture (arrowheads) of the leaflets, which constitutes part of this complex anomaly (AM: atrial myocardium; IVS: interventricular septum; LV: left ventricle; VM: ventricular myocardium).

c­ ontrol embryos, suggesting a possible role for ALK3 in the origin of Ebstein’s anomaly. The abnormal size and texture of the valve leaflets are difficult to explain. Maternal lithium therapy may also cause Ebstein’s anomaly (Cohen et al., 1994); lithium is known to block the Wnt-betacatenin pathway (Gould, 2006), which is crucial for valve endocardial epithelial-to-mesenchymal transformation (Hurlstone et al., 2003; Liebner et al., 2004).

VI.B.ii.  Down Syndrome Down syndrome (DS), caused by trisomy of chromosome 21, is the most common chromosomal abnormality affecting the human population. Affected individuals can show atrioventricular septal defects, atrial and ventricular septal defects, tetralogy of Fallot and hypoplastic left heart syndrome. The molecular basis of DS is unknown. The DS critical region 1 (DSCR1) gene is found in human chromosome 21 (Fuentes et al., 1995) and in the syntenic region of mouse chromosome 16 (Fuentes et al., 1997), trisomy of which is associated with congenital heart defects similar to those observed in DS (Epstein et al., 1985; Fuentes et al., 1997). DSCR1 encodes a regulatory protein in the calcineurin/NFAT signal transduction pathway that binds to calcineurin and modulates its phosphatase activity (Fuentes et al., 2000; Rothermel et al., 2000) (see prev­ious section). In the heart, DSCR1 is expressed specifically in atrioventricular and outflow tract valve endocardium, in

the muscular interventricular septum and in the ventricular myocardium (Lange et al., 2004), clearly suggesting a role for this molecule in the cardiac anomalies of DS.

VI.B.iii.  Allagile Syndrome Allagile syndrome (AGS) is an autosomal dominant disorder characterized by cardiac anomalies involving the peripheral and main pulmonary arteries, as well as the pulmonary valves and, on rare occasions, tetralogy of Fallot (Servidei et al., 1994; Eldadah et al., 2001). The NOTCH ligand JAGGED1 (JAG1) has been mapped to the AGS critical region, and distinct coding mutations in JAG1 have been identified that segregate with the disease phenotype in around 94% of patients (Li et al., 1997; Oda et al., 1997; Eldadah et al., 2001). In AGS patients without JAG1 mutations the NOTCH2 receptor is mutated (McDaniell et al., 2006). These patients also have varying degrees of renal disease, suggesting that the phenotypic profile for AGS due to NOTCH2 mutations may differ from that of AGS caused by JAG1 mutations (McDaniell et al., 2006).

VI.B.iv.  Noonan’s Syndrome Noonan’s syndrome (NS) is an autosomal dominant multiple congenital anomaly syndrome characterized by congenital heart disease including pulmonary valve stenosis, hypertrophic cardiomyopathy, atrial septal defect ostium

406

secundum-type and stenosis of the peripheral pulmonary arteries (Noonan, 1968). NS is the second most common syndromal form of congenital heart disease after DS, with an estimated frequency of 1 in 2,000 (Marino et al., 1999). Increased RAS signaling due to PTPN11 (Tartaglia et al., 2001, 2002; Fragale et al., 2004), SOS1 (Roberts et al., 2007; Tartaglia et al., 2007) and KRAS (Carta et al., 2006; Schubbert et al., 2006) mutations cause around 60% of NS cases. Gain-of-function mutations in RAF1, which encodes a downstream effector of RAS signaling, were also recently linked to NS (Pandit et al., 2007; Razzaque et al., 2007). Alterations in these genes are responsible for the different congenital heart diseases associated with NS. Thus, pulmonic stenosis, atrial septal defect ostium secundum-type and stenosis of pulmonary arteries are associated with PTPN11 (Sznajer et al., 2007) or SOS1 (Roberts et al., 2007; Tartaglia et al., 2007) mutations, and hypertrophic cardiomyopathy has been linked to RAF1 mutation (Pandit et al., 2007; Razzaque et al., 2007). Studies in mouse models (Araki et al., 2004) will help understanding of the mechanisms involved in the congenital heart diseases associated with NS.

VI.B.v.  Marfan Syndrome Marfan syndrome (MFS) is an autosomal dominant disorder of connective tissue; various abnormalities, including cardiac defects, account for reduced patient lifespan (Pyeritz and McKusick, 1979). MFS patients have dysplastic cardiac valves with signs of myxomatous degeneration, resulting in severe congestive heart failure by cardiac valve prolapse (Nasuti et al., 2004). The origin of MFS is ascribed to mutations in the fibrillin1 gene (Dietz et al., 1991). Fibrillin1 is part of a complex of at least five other proteins present in microfibrils that interact with elastin and probably other extracellular matrix components. MFS pathogenesis would thus appear to relate to disruption of extracellular matrix structure and growth factor (TGF) signaling (Robinson et al., 2006), which would also explain the tissue abnormalities found in cardiac valve leaflets (Nollen and Mulder, 2004).

VII.  Tissue engineering: in vitro generation of functional valvular tissue Dysfunctional valves are extremely common among congenital heart disease patients, as valve anomalies constitute 20–30% of all cardiac malformations (Hoffman and Kaplan, 2002). In the USA, it is estimated that more than 100,000 patients need valve replacements each year (Vesely et al., 2005); efforts have thus increased to replace damaged valvular tissue with prosthetic valves (Vesely, 2005). Transplantation of animal tissue with ­ properties

PART  |  6  Cushions, Valves and Septa

similar to that of the living human valve leaflet has become popular, and has proven more successful than classic mechanical devices which carry a high risk of thromboembolism (Jamieson, 2006). Bioprosthetic valves include porcine valves, bovine pericardial valves and allograft valves. Porcine xenografts consist of intact pig valves that have undergone low-concentration glutaraldehyde treatment (Carpentier, 1977). Bovine pericardial valves are constructed by assembling various pieces of glutaraldehydetreated calf pericardium supported by a stent. In both cases, glutaraldehyde reduces the antigenicity of the tissue and protects it from proteolytic degradation. In contrast, human allografts are cryogenically preserved. Porcine, bovine and human grafts are normally used to substitute aortic and pulmonary valves, but can also be applied to replace mitral valves (Vesely et al., 2005). Surgeons often prefer to repair and reconstruct a damaged valve using tissue pieces rather than replacing the entire structure. The use of bioprosthetic valves, nonetheless, has several well-known shortcomings. Whereas mortality during primary surgery is very low, it increases dramatically in secondary or tertiary interventions (Biglioli et al., 1994). The average lifespan of a bioprosthetic valve is 15–20 years (Rahimtoola, 2003), and although this seems quite a reasonable period, valves tend to calcify and present other complications with time. Finally, good bioprostheses (especially allografts) for children are difficult to find, because the specific needs are variable and the structures age rapidly leading to the need for further surgical intervention (Kanter et al., 2002). Valvular tissue engineering has thus emerged as a sophisticated solution to these problems, but it is far from being a reality. Porcine and bovine valve implants are acellular, and perform their functions quite well from a biomechanical point of view; endothelialization of these bioprostheses seems to be an option to bring some life to these structures (Leukauf et al., 1993). It is hoped that cellularization of the prostheses could enable the replaced valves to sustain a sort of homeostasis, which could in turn prevent complications related to valve aging and infection. This contrasts with other approaches that favor the use of acellular matrix xenografts. Despite the concerns about cellularized valve implants, several biocompatible and biodegradable matrices that act as cell scaffolds or, alternatively, collagen constructs, have been tested to replace damaged human tissues (Vesely, 2005). The issue of which cells should be used to seed these tissue-engineered prostheses remains open. Several laboratories have studied the nature of cardiac valve interstitial tissue (Della Rocca et al., 2000; Taylor et al., 2000; Cimini et al., 2002; Blevins et al., 2006). Autologous stem cells have also become a fashionable candidate to replace almost any cell type in tissue engineering applications, including valve replacement or repair (Mendelson and Schoen, 2006). Stem cell behavior is complex and the use of these cells has some risks, including potential teratoma

Chapter  |  6.2  Signaling Pathways in Valve Formation: The Origin of Congenital Defects

formation (Mitjavila-Garcia et al., 2005), but the technique will remain prominent in the concept of recapitulation in tissue engineering. In the field of tissue engineering, “to recapitulate” means to repeat in a fast, abridged manner the ontogenetic processes that take place in the development of an organic structure (Caplan, 2003; Mironov et al., 2003; Perez-Pomares et al., 2006). Hence, developmental biology acts as a guide from which we can infer the elements essential for development of functional tissue. Valve formation from cushions can be partially recapitulated in vitro in embryonic tissue maintained in collagen gel tubes (Goodwin et al., 2005), but it is not known whether these cultures can sustain the correct growth and maturation of functional cardiac valvular structures.

VIII.  Future prospects The many signaling pathways and molecules that participate in cardiac valve development, as well as the interplay among them, is still poorly-understood (reviewed in Armstrong and Bischoff, 2004; Person et al., 2005; Lincoln et al., 2006). The importance of myocardial–endocardial signaling, as well as of reciprocal endocardial–myocardial interactions for valve formation, has become apparent in the last few years (Timmerman et al., 2004). We know very little about the regulation of valve morphogenesis. Genetic studies using conditional mouse models and zebrafish will help identify additional players in valvulogenesis, and will determine the processes in which they participate. Gene transfer and biochemical assays using mouse and avian endocardial cushion explants and cell lines will help us clarify the mechanism of action of “old” and “new” players. Integration of this knowledge will be fundamental for the design of successful strategies for valve tissue engineering and their applications in biomedicine.

Acknowledgments The authors are grateful to all past and present members of their laboratories for contributions to the work presented in this chapter. Special thanks are given to Mr J. M. González-Rosa and Dr M. E. Manjón-Cabeza for their expert help in the original design of the diagrams, as well as to Drs D. Sánchez-Quintana and B. Picazo-Angelin for sharing images of human pathology specimens. This work was funded by grants SAF2007-62445 (Spanish Ministry of Science and Innovation), 200520M072 and P-2006/ BIO-194 (Regional Government of Madrid) to JLdlP and EU ID FP6 LSHM-CT-2005-018630 (HEARTREPAIR) to JLdlP and JMPP.

References Akhurst, R.J., Lehnert, S.A., Faissner, A., Duffie, E., 1990. TGF beta in murine morphogenetic processes: the early embryo and cardiogenesis. Development 108, 645–656.

407

Akiyama, H., Chaboissier, M.C., Behringer, R.R., Rowitch, D.H., Schedl, A., Epstein, J.A., de Crombrugghe, B., 2004. Essential role of Sox9 in the pathway that controls formation of cardiac valves and septa. Proc. Natl. Acad. Sci. USA 101, 6502–6507. Andelfinger, G., Wright, K.N., Lee, H.S., Siemens, L.M., Benson, D.W., 2003. Canine tricuspid valve malformation, a model of human Ebstein anomaly, maps to dog chromosome 9. J Med Genet. 40, 320–324. Anderson, R.H., Becker, A.E., 1980. Cardiac Anatomy. Gower Medical Publishing Ltd, London, UK. Araki, T., Mohi, M.G., Ismat, F.A., Bronson, R.T., Williams, I.R., Kutok, J.L., Yang, W., Pao, L.I., Gilliland, D.G., Epstein, J.A., Neel, B.G., 2004. Mouse model of Noonan syndrome reveals cell type- and gene dosage-dependent effects of Ptpn11 mutation. Nat. Med. 10, 849–857. Armesilla, A.L., Lorenzo, E., Gomez del Arco, P., Martinez-Martinez, S., Alfranca, A., Redondo, J.M., 1999. Vascular endothelial growth factor activates nuclear factor of activated T cells in human endothelial cells: a role for tissue factor gene expression. Mol. Cell Biol. 19, 2032–2043. Armstrong, E.J., Bischoff, J., 2004. Heart valve development: endothelial cell signaling and differentiation. Circ. Res. 95, 459–470. Artavanis-Tsakonas, S., Rand, M.D., Lake, R.J., 1999. Notch signaling: cell fate control and signal integration in development. Science 284, 770–776. Attenhofer Jost, C.H., Connolly, H.M., Edwards, W.D., Hayes, D., Warnes, C.A., Danielson, G.K., 2005. Ebstein’s anomaly – review of a multifaceted congenital cardiac condition. Swiss Med. Wkly. 135, 269–281. Beis, D., Bartman, T., Jin, S.W., Scott, I.C., D’Amico, L.A., Ober, E.A., Verkade, H., Frantsve, J., Field, H.A., Wehman, A., Baier, H., Tallafuss, A., Bally-Cuif, L., Chen, J.N., Stainier, D.Y., Jungblut, B., 2005. Genetic and cellular analyses of zebrafish atrioventricular cushion and valve development. Development 132, 4193–4204. Bernanke, D.H., Markwald, R.R., 1982. Migratory behavior of cardiac cushion tissue cells in a collagen-lattice culture system. Dev. Biol. 91, 235–245. Biglioli, P., Di Matteo, S., Parolari, A., Antona, C., Arena, V., Sala, A., 1994. Reoperative cardiac valve surgery: a multivariable analysis of risk factors. Cardiovasc. Surg. 2, 216–222. Blevins, T.L., Carroll, J.L., Raza, A.M., Grande-Allen, K.J., 2006. Phenotypic characterization of isolated valvular interstitial cell subpopulations. J. Heart. Valve Dis. 15, 815–822. Bourajjaj, M., Armand, A.S., da Costa Martins, P.A., Weijts, B., van der Nagel, R., Heeneman, S., Wehrens, X.H., De Windt, L.J., 2008. NFATc2 is a necessary mediator of calcineurin-dependent cardiac hypertrophy and heart failure. J. Biol. Chem. 283 (32), 22295–22303. Bourdeau, A., Dumont, D.J., Letarte, M., 1999. A murine model of hereditary hemorrhagic telangiectasia. J. Clin. Invest. 104, 1343–1351. Boyer, A.S., Erickson, C.P., Runyan, R.B., 1999b. Epithelial– mesenchymal transformation in the embryonic heart is mediated through distinct pertussis toxin-sensitive and TGFbeta signal transduction mechanisms. Dev. Dyn. 214, 81–91. Boyer, A.S., Ayerinskas, I.I., Vincent, E.B., McKinney, L.A., Weeks, D.L., Runyan, R.B., 1999a. TGFbeta2 and TGFbeta3 have separate and sequential activities during epithelial–mesenchymal cell transformation in the embryonic heart. Dev. Biol. 208, 530–545. Brannan, C.I., Perkins, A.S., Vogel, K.S., Ratner, N., Nordlund, M.L., Reid, S.W., Buchberg, A.M., Jenkins, N.A., Parada, L.F., Copeland,

408

N.G., 1994. Targeted disruption of the neurofibromatosis type-1 gene leads to developmental abnormalities in heart and various neural crest-derived tissues. Genes Dev. 8, 1019–1029. Brown, C.B., Boyer, A.S., Runyan, R.B., Barnett, J.V., 1996. Antibodies to the Type II TGFbeta receptor block cell activation and migration during atrioventricular cushion transformation in the heart. Dev. Biol. 174, 248–257. Brown, C.B., Boyer, A.S., Runyan, R.B., Barnett, J.V., 1999. Requirement of type III TGF-beta receptor for endocardial cell transformation in the heart. Science 283, 2080–2082. Camenisch Jr, T.D., Spicer, A.P., Brehm-Gibson, T., Biesterfeldt, J., Augustine, M.L., Calabro, A., Kubalak, S., Klewer, S.E., McDonald, J.A., 2000. Disruption of hyaluronan synthase-2 abrogates normal cardiac morphogenesis and hyaluronan-mediated transformation of epithelium to mesenchyme. J. Clin. Invest. 106, 349–360. Camenisch, T.D., Molin, D.G., Person, A., Runyan, R.B., Gittenberger-de Groot, A.C., McDonald, J.A., Klewer, S.E., 2002a. Temporal and distinct TGFbeta ligand requirements during mouse and avian endocardial cushion morphogenesis. Dev. Biol. 248, 170–181. Camenisch, T.D., Schroeder, J.A., Bradley, J., Klewer, S.E., McDonald, J.A., 2002b. Heart-valve mesenchyme formation is dependent on hyaluronan-augmented activation of ErbB2-ErbB3 receptors. Nat. Med. 8, 850–855. Caplan, A.I., 2003. Embryonic development and the principles of tissue engineering. Novartis Found Symp. 249, 17–25 discussion 25–33, 170–174, 239–241. Carmeliet, P., Ferreira, V., Breier, G., Pollefeyt, S., Kieckens, L., Gertsenstein, M., Fahrig, M., Vandenhoeck, A., Harpal, K., Eberhardt, C., Declercq, C., Pawling, J., Moons, L., Collen, D., Risau, W., Nagy, A., 1996. Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 380, 435–439. Carmona, R., Gonzalez-Iriarte, M., Macias, D., Perez-Pomares, J.M., Garcia-Garrido, L., Munoz-Chapuli, R., 2000. Immunolocalization of the transcription factor Slug in the developing avian heart. Anat. Embryol. (Berlin) 201, 103–109. Carpentier, A., 1977. From valvular xenograft to valvular bioprosthesis (1965–1977). Med. Instrum. 11, 98–101. Carta, C., Pantaleoni, F., Bocchinfuso, G., Stella, L., Vasta, I., Sarkozy, A., Digilio, C., Palleschi, A., Pizzuti, A., Grammatico, P., Zampino, G., Dallapiccola, B., Gelb, B.D., Tartaglia, M., 2006. Germline missense mutations affecting KRAS Isoform B are associated with a severe Noonan syndrome phenotype. Am. J. Hum. Genet. 79, 129–135. Castaneda, A.R., Anderson, R.C., Edwards, J.E., 1969. Congenital mitral stenosis resulting from anomalous arcade and obstructing papillary muscles. Report of correction by use of ball valve prosthesis. Am. J. Cardiol. 24, 237–240. Chang, C.P., Neilson, J.R., Bayle, J.H., Gestwicki, J.E., Kuo, A., Stankunas, K., Graef, I.A., Crabtree, G.R., 2004. A field of myocardial–endocardial NFAT signaling underlies heart valve morphogenesis. Cell 118, 649–663. Cheifetz, S., Bellon, T., Cales, C., Vera, S., Bernabeu, C., Massague, J., Letarte, M., 1992. Endoglin is a component of the transforming growth factor-beta receptor system in human endothelial cells. J. Biol. Chem. 267, 19027–19030. Cimini, M., Rogers, K.A., Boughner, D.R., 2002. Aortic valve interstitial cells: an evaluation of cell viability and cell phenotype over time. J. Heart. Valve Dis. 11, 881–887. Cohen, L.S., Friedman, J.M., Jefferson, J.W., Johnson, E.M., Weiner, M.L., 1994. A re-evaluation of risk of in utero exposure to lithium. JAMA 271, 146–150.

PART  |  6  Cushions, Valves and Septa

Crispino, J.D., Lodish, M.B., Thurberg, B.L., Litovsky, S.H., Collins, T., Molkentin, J.D., Orkin, S.H., 2001. Proper coronary vascular development and heart morphogenesis depend on interaction of GATA-4 with FOG cofactors. Genes Dev. 15, 839–844. de la Pompa, J.L., Timmerman, L.A., Takimoto, H., Yoshida, H., Elia, A.J., Samper, E., Potter, J., Wakeham, A., Marengere, L., Langille, B.L., Crabtree, G.R., Mak, T.W., 1998. Role of the NF-ATc transcription factor in morphogenesis of cardiac valves and septum. Nature 392, 182–186. de Lange, F.J., Moorman, A.F., Anderson, R.H., Manner, J., Soufan, A.T., de Gier-de Vries, C., Schneider, M.D., Webb, S., van den Hoff, M.J., Christoffels, V.M., 2004. Lineage and morphogenetic analysis of the cardiac valves. Circ. Res. 95, 645–654. Del Amo, F.F., Smith, D.E., Swiatek, P.J., Gendron-Maguire, M., Greenspan, R.J., McMahon, A.P., Gridley, T., 1992. Expression pattern of Motch, a mouse homolog of Drosophila Notch, suggests an important role in early postimplantation mouse development. Development 115, 737–744. Della Rocca, F., Sartore, S., Guidolin, D., Bertiplaglia, B., Gerosa, G., Casarotto, D., Pauletto, P., 2000. Cell composition of the human pulmonary valve: a comparative study with the aortic valve – the VESALIO Project (Vitalitate Exornatum Succedaneum Aorticum Labore Ingegnoso Obtinebitur). Ann. Thorac. Surg. 70, 1594–1600. del Monte, G., Grego-Bessa, J., Gonzalez-Rajal, A., Bolos, V., de La Pompa, J.L., 2007. Monitoring Notch1 activity in development: Evidence for a feedback regulatory loop. Dev. Dyn. 236, 2594–2614. Delot, E.C., 2003. Control of endocardial cushion and cardiac valve maturation by BMP signaling pathways. Mol. Genet. Metab. 80, 27–35. Delot, E.C., Bahamonde, M.E., Zhao, M., Lyons, K.M., 2003. BMP sig­ naling is required for septation of the outflow tract of the mammalian heart. Development 130, 209–220. Desgrosellier, J.S., Mundell, N.A., McDonnell, M.A., Moses, H.L., Barnett, J.V., 2005. Activin receptor-like kinase 2 and Smad6 regulate epithelial–mesenchymal transformation during cardiac valve formation. Dev. Biol. 280, 201–210. Dickson, M.C., Slager, H.G., Duffie, E., Mummery, C.L., Akhurst, R.J., 1993. RNA and protein localisations of TGF beta 2 in the early mouse embryo suggest an involvement in cardiac development. Development 117, 625–639. Dietz, H.C., Cutting, G.R., Pyeritz, R.E., Maslen, C.L., Sakai, L.Y., Corson, G.M., Puffenberger, E.G., Hamosh, A., Nanthakumar, E.J., Curristin, S.M., et al., 1991. Marfan syndrome caused by a recurrent de novo missense mutation in the fibrillin gene. Nature 352, 337–339. Dor, Y., Camenisch, T.D., Itin, A., Fishman, G.I., McDonald, J.A., Carmeliet, P., Keshet, E., 2001. A novel role for VEGF in endocardial cushion formation and its potential contribution to congenital heart defects. Development 128, 1531–1538. Drake, C.J., Little, C.D., 1995. Exogenous vascular endothelial growth factor induces malformed and hyperfused vessels during embryonic neovascularization. Proc. Natl. Acad. Sci. USA 92, 7657–7661. Dudley, A.T., Robertson, E.J., 1997. Overlapping expression domains of bone morphogenetic protein family members potentially account for limited tissue defects in BMP7 deficient embryos. Dev. Dyn. 208, 349–362. Dumont, D.J., Jussila, L., Taipale, J., Lymboussaki, A., Mustonen, T., Pajusola, K., Breitman, M., Alitalo, K., 1998. Cardiovascular failure in mouse embryos deficient in VEGF receptor-3. Science 282, 946–949.

Chapter  |  6.2  Signaling Pathways in Valve Formation: The Origin of Congenital Defects

Dupays, L., Jarry-Guichard, T., Mazurais, D., Calmels, T., Izumo, S., Gros, D., Theveniau-Ruissy, M., 2005. Dysregulation of connexins and inactivation of NFATc1 in the cardiovascular system of Nkx2-5 null mutants. J. Mol. Cell Cardiol. 38, 787–798. Eisenberg, L.M., Markwald, R.R., 1995. Molecular regulation of atrio­ ventricular valvuloseptal morphogenesis. Circ. Res. 77, 1–6. Eldadah, Z.A., Hamosh, A., Biery, N.J., Montgomery, R.A., Duke, M., Elkins, R., Dietz, H.C., 2001. Familial tetralogy of Fallot caused by mutation in the jagged1 gene. Hum. Mol. Genet. 10, 163–169. Emmel, E.A., Verweij, C.L., Durand, D.B., Higgins, K.M., Lacy, E., Crabtree, G.R., 1989. Cyclosporin A specifically inhibits function of nuclear proteins involved in T-cell activation. Science 246, 1617–1620. Epstein, C.J., Cox, D.R., Epstein, L.B., 1985. Mouse trisomy 16: an animal model of human trisomy 21 (Down syndrome). Ann. NY Acad. Sci. 450, 157–168. Erickson, S.L., O’Shea, K.S., Ghaboosi, N., Loverro, L., Frantz, G., Bauer, M., Lu, L.H., Moore, M.W., 1997. ErbB3 is required for normal cerebellar and cardiac development: a comparison with ErbB2and heregulin-deficient mice. Development 124, 4999–5011. Euler-Taimor, G., Heger, J., 2006. The complex pattern of SMAD signaling in the cardiovascular system. Cardiovasc. Res. 69, 15–25. Falls, D.L., 2003. Neuregulins: functions, forms, and signaling strategies. Exp. Cell Res. 284, 14–30. Feng, X.H., Derynck, R., 2005. Specificity and versatility in tgf-beta signaling through Smads. Annu. Rev. Cell Dev. Biol. 21, 659–693. Ferrara, N., Carver-Moore, K., Chen, H., Dowd, M., Lu, L., O’Shea, K.S., Powell-Braxton, L., Hillan, K.J., Moore, M.W., 1996. Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 380, 439–442. Fischer, A., Schumacher, N., Maier, M., Sendtner, M., Gessler, M., 2004. The Notch target genes Hey1 and Hey2 are required for embryonic vascular development. Genes Dev. 18, 901–911. Fong, G.H., Rossant, J., Gertsenstein, M., Breitman, M.L., 1995. Role of the Flt-1 receptor tyrosine kinase in regulating the assembly of vascular endothelium. Nature 376, 66–70. Fragale, A., Tartaglia, M., Wu, J., Gelb, B.D., 2004. Noonan syndromeassociated SHP2/PTPN11 mutants cause EGF-dependent prolonged GAB1 binding and sustained ERK2/MAPK1 activation. Hum. Mutat. 23, 267–277. Fuentes, J.J., Pritchard, M.A., Planas, A.M., Bosch, A., Ferrer, I., Estivill, X., 1995. A new human gene from the Down syndrome critical region encodes a proline-rich protein highly expressed in fetal brain and heart. Hum. Mol. Genet. 4, 1935–1944. Fuentes, J.J., Pritchard, M.A., Estivill, X., 1997. Genomic organization, alternative splicing, and expression patterns of the DSCR1 (Down syndrome candidate region 1) gene. Genomics 44, 358–361. Fuentes, J.J., Genesca, L., Kingsbury, T.J., Cunningham, K.W., PerezRiba, M., Estivill, X., de la Luna, S., 2000. DSCR1, overexpressed in Down syndrome, is an inhibitor of calcineurin-mediated signaling pathways. Hum. Mol. Genet. 9, 1681–1690. Galvin Jr, K.M., Donovan, M.J., Lynch, C.A., Meyer, R.I., Paul, R.J., Lorenz, J.N., Fairchild-Huntress, V., Dixon, K.L., Dunmore, J. H., Gimbrone, M.A., Falb, D., Huszar, D., 2000. A role for smad6 in development and homeostasis of the cardiovascular system. Nat. Genet. 24, 171–174. Garg, V., 2006. Molecular genetics of aortic valve disease. Curr. Opin. Cardiol. 21, 180–184. Garg, V., Kathiriya, I.S., Barnes, R., Schluterman, M.K., King, I.N., Butler, C.A., Rothrock, C.R., Eapen, R.S., Hirayama-Yamada, K.,

409

Joo, K., Matsuoka, R., Cohen, J.C., Srivastava, D., 2003. GATA4 mutations cause human congenital heart defects and reveal an interaction with TBX5. Nature 424, 443–447. Garg, V., Muth, A.N., Ransom, J.F., Schluterman, M.K., Barnes, R., King, I.N., Grossfeld, P.D., Srivastava, D., 2005. Mutations in NOTCH1 cause aortic valve disease. Nature 437, 270–274. Gaussin, V., Van De Putte, T., Mishina, Y., Hanks, M.C., Zwijsen, A., Huylebroeck, D., Behringer, R.R., Schneider, M.D., 2002. Endocardial cushion and myocardial defects after cardiac myocytespecific conditional deletion of the bone morphogenetic protein receptor ALK3. Proc. Natl. Acad. Sci. USA 99, 2878–2883. Gaussin, V., Morley, G.E., Cox, L., Zwijsen, A., Vance, K.M., Emile, L., Tian, Y., Liu, J., Hong, C., Myers, D., Conway, S.J., Depre, C., Mishina, Y., Behringer, R.R., Hanks, M.C., Schneider, M.D., Huylebroeck, D., Fishman, G.I., Burch, J.B., Vatner, S.F., 2005. Alk3/Bmpr1a receptor is required for development of the atrioventricular canal into valves and annulus fibrosus. Circ. Res. 97, 219–226. Gitler, A.D., Zhu, Y., Ismat, F.A., Lu, M.M., Yamauchi, Y., Parada, L.F., Epstein, J.A., 2003. Nf1 has an essential role in endothelial cells. Nat. Genet. 33, 75–79. Goodwin, R.L., Nesbitt, T., Price, R.L., Wells, J.C., Yost, M.J., Potts, J.D., 2005. Three-dimensional model system of valvulogenesis. Dev. Dyn. 233, 122–129. Gould, T.D., 2006. Targeting glycogen synthase kinase-3 as an approach to develop novel mood-stabilising medications. Expert. Opin. Ther. Targets 10, 377–392. Graef, I.A., Chen, F., Crabtree, G.R., 2001a. NFAT signaling in vertebrate development. Curr. Opin. Genet. Dev. 11, 505–512. Graef, I.A., Chen, F., Chen, L., Kuo, A., Crabtree, G.R., 2001b. Signals transduced by Ca(2)/calcineurin and NFATc3/c4 pattern the developing vasculature. Cell 105, 863–875. Grego-Bessa, J., Luna-Zurita, L., Del Monte, G., Bolos, V., Melgar, P., Arandilla, A., Garratt, A.N., Zang, H., Mukouyama, Y.S., Chen, H., Shou, W., Ballestar, E., Esteller, M., Rojas, A., Pérez-Pomares, J.M., de la Pompa, J.L., 2007. Notch signaling is essential for ventricular chamber development. Dev. Cell 12, 415–429. Gregoratos, G., 2003. Infective endocarditis in the elderly: diagnosis and management. Am. J. Geriatr. Cardiol. 12, 183–189. Hallaq, H., Pinter, E., Enciso, J., McGrath, J., Zeiss, C., Brueckner, M., Madri, J., Jacobs, H.C., Wilson, C.M., Vasavada, H., Jiang, X., Bogue, C.W., 2004. A null mutation of Hhex results in abnormal cardiac development, defective vasculogenesis and elevated Vegfa levels. Development 131, 5197–5209. Hata, A., Lagna, G., Massague, J., Hemmati-Brivanlou, A., 1998. Smad6 inhibits BMP/Smad1 signaling by specifically competing with the Smad4 tumor suppressor. Genes Dev. 12, 186–197. Hay, E.D., 2005. The mesenchymal cell, its role in the embryo, and the remarkable signaling mechanisms that create it. Dev. Dyn. 233, 706–720. Heikinheimo, M., Scandrett, J.M., Wilson, D.B., 1994. Localization of transcription factor GATA-4 to regions of the mouse embryo involved in cardiac development. Dev. Biol. 164, 361–373. Hirayama-Yamada, K., Kamisago, M., Akimoto, K., Aotsuka, H., Nakamura, Y., Tomita, H., Furutani, M., Imamura, S., Takao, A., Nakazawa, M., Matsuoka, R., 2005. Phenotypes with GATA4 or NKX2.5 mutations in familial atrial septal defect. Am. J. Med. Genet. A 135, 47–52. Hoffman, J.I., 1995. Incidence of congenital heart disease: II. Prenatal incidence. Pediatr. Cardiol. 16, 155–165.

410

Hoffman, J.I., Kaplan, S., 2002. The incidence of congenital heart disease. J. Am. Coll. Cardiol. 39, 1890–1900. Hogan, P.G., Chen, L., Nardone, J., Rao, A., 2003. Transcriptional regulation by calcium, calcineurin, and NFAT. Genes Dev. 17, 2205–2232. Hoogaars, W.M., Tessari, A., Moorman, A.F., de Boer, P.A., Hagoort, J., Soufan, A.T., Campione, M., Christoffels, V.M., 2004. The transcriptional repressor Tbx3 delineates the developing central conduction system of the heart. Cardiovasc. Res. 62, 489–499. Houston, C.S., Opitz, J.M., Spranger, J.W., Macpherson, R.I., Reed, M. H., Gilbert, E.F., Herrmann, J., Schinzel, A., 1983. The campomelic syndrome: review, report of 17 cases, and follow-up on the currently 17-year-old boy first reported by Maroteaux et al in 1971. Am. J. Med. Genet. 15, 3–28. Hurle, J.M., Colvee, E., Blanco, A.M., 1980. Development of mouse semilunar valves. Anat. Embryol. (Berlin) 160, 83–91. Hurlstone, A.F., Haramis, A.P., Wienholds, E., Begthel, H., Korving, J., Van Eeden, F., Cuppen, E., Zivkovic, D., Plasterk, R.H., Clevers, H., 2003. The Wnt/beta-catenin pathway regulates cardiac valve formation. Nature 425, 633–637. Jacks, T., Shih, T.S., Schmitt, E.M., Bronson, R.T., Bernards, A., Weinberg, R.A., 1994. Tumour predisposition in mice heterozygous for a targeted mutation in Nf1. Nat. Genet. 7, 353–361. Jamieson, W.R., 2006. Advanced technologies for cardiac valvular replacement, transcatheter innovations and reconstructive surgery. Surg. Technol. Int. 15, 149–187. Jiang, X., Rowitch, D.H., Soriano, P., McMahon, A.P., Sucov, H.M., 2000. Fate of the mammalian cardiac neural crest. Development 127, 1607–1616. Jiao, K., Kulessa, H., Tompkins, K., Zhou, Y., Batts, L., Baldwin, H.S., Hogan, B.L., 2003. An essential role of Bmp4 in the atrioventricular septation of the mouse heart. Genes Dev. 17, 2362–2367. Kanani, M., Moorman, A.F., Cook, A.C., Webb, S., Brown, N.A., Lamers, W.H., Anderson, R.H., 2005. Development of the atrio­ ventricular valves: clinicomorphological correlations. Ann. Thorac. Surg. 79, 1797–1804. Kanter, K.R., Budde, J.M., Parks, W.J., Tam, V.K., Sharma, S., Williams, W.H., Fyfe, D.A., 2002. One hundred pulmonary valve replacements in children after relief of right ventricular outflow tract obstruction. Ann. Thorac. Surg. 73, 1801–1806; discussion 1806–1807. Kern, C.B., Hoffman, S., Moreno, R., Damon, B.J., Norris, R.A., Krug, E.L., Markwald, R.R., Mjaatvedt, C.H., 2005. Immunolocal­ ization of chick periostin protein in the developing heart. Anat. Rec. A. Discov. Mol. Cell Evol. Biol. 284, 415–423. Keyes, W.M., Sanders, E.J., 1999. Cell death in the endocardial cushions of the developing heart. J. Mol. Cell Cardiol. 31, 1015–1023. Kim, J.S., Viragh, S., Moorman, A.F., Anderson, R.H., Lamers, W.H., 2001a. Development of the myocardium of the atrioventricular canal and the vestibular spine in the human heart. Circ. Res. 88, 395–402. Kim, R.Y., Robertson, E.J., Solloway, M.J., 2001b. Bmp6 and Bmp7 are required for cushion formation and septation in the developing mouse heart. Dev. Biol. 235, 449–466. Kirby, M.L., Gale, T.F., Stewart, D.E., 1983. Neural crest cells contribute to normal aorticopulmonary septation. Science 220, 1059–1061. Kisanuki, Y.Y., Hammer, R.E., Miyazaki, J., Williams, S.C., Richardson, J.A., Yanagisawa, M., 2001. Tie2-Cre transgenic mice: a new model for endothelial cell-lineage analysis in vivo. Dev. Biol. 230, 230–242. Kokubo, H., Miyagawa-Tomita, S., Tomimatsu, H., Nakashima, Y., Nakazawa, M., Saga, Y., Johnson, R.L., 2004. Targeted disruption of hesr2 results in atrioventricular valve anomalies that lead to heart dysfunction. Circ. Res. 95, 540–547.

PART  |  6  Cushions, Valves and Septa

Kokubo, H., Miyagawa-Tomita, S., Nakazawa, M., Saga, Y., Johnson, R.L., 2005. Mouse hesr1 and hesr2 genes are redundantly required to mediate Notch signaling in the developing cardiovascular system. Dev. Biol. 278, 301–309. Kokubo, H., Tomita-Miyagawa, S., Hamada, Y., Saga, Y., 2007. Hesr1 and Hesr2 regulate atrioventricular boundary formation in the developing heart through the repression of Tbx2. Development 134, 747–755. Krebs, L.T., Xue, Y., Norton, C.R., Shutter, J.R., Maguire, M., Sundberg, J.P., Gallahan, D., Closson, V., Kitajewski, J., Callahan, R., Smith, G.H., Stark, K.L., Gridley, T., 2000. Notch signaling is essential for vascular morphogenesis in mice. Genes Dev. 14, 1343–1352. Kretzschmar, M., Massague, J., 1998. SMADs: mediators and regulators of TGF-beta signaling. Curr. Opin. Genet. Dev. 8, 103–111. Krug, E.L., Runyan, R.B., Markwald, R.R., 1985. Protein extracts from early embryonic hearts initiate cardiac endothelial cytodifferentiation. Dev. Biol. 112, 414–426. Krug, E.L., Mjaatvedt, C.H., Markwald, R.R., 1987. Extracellular matrix from embryonic myocardium elicits an early morphogenetic event in cardiac endothelial differentiation. Dev. Biol. 120, 348–355. Krug, E.L., Rezaee, M., Isokawa, K., Turner, D.K., Litke, L.L., Wunsch, A.M., Bain, J.L., Riley, D.A., Capehart, A.A., Markwald, R.R., 1995. Transformation of cardiac endothelium into cushion mesenchyme is dependent on ES/130: temporal, spatial, and functional studies in the early chick embryo. Cell. Mol. Biol. Res. 41, 263–277. Kruithof, B.P., Krawitz, S.A., Gaussin, V., 2007. Atrioventricular valve development during late embryonic and postnatal stages involves condensation and extracellular matrix remodeling. Dev. Biol. 302, 208–217. Kruzynska-Frejtag, A., Machnicki, M., Rogers, R., Markwald, R.R., Conway, S.J., 2001. Periostin (an osteoblast-specific factor) is expressed within the embryonic mouse heart during valve formation. Mech. Dev. 103, 183–188. Kuo, C.T., Morrisey, E.E., Anandappa, R., Sigrist, K., Lu, M.M., Parmacek, M.S., Soudais, C., Leiden, J.M., 1997. GATA4 transcription factor is required for ventral morphogenesis and heart tube formation. Genes Dev. 11, 1048–1060. Lakkis, M.M., Epstein, J.A., 1998. Neurofibromin modulation of ras activity is required for normal endocardial–mesenchymal transformation in the developing heart. Development 125, 4359–4367. Lamers, W.H., Viragh, S., Wessels, A., Moorman, A.F., Anderson, R.H., 1995. Formation of the tricuspid valve in the human heart. Circulation 91, 111–121. Lange, A.W., Molkentin, J.D., Yutzey, K.E., 2004. DSCR1 gene expression is dependent on NFATc1 during cardiac valve formation and colocalizes with anomalous organ development in trisomy 16 mice. Dev. Biol. 266, 346–360. Larsen, W.J., 2001. Human Embryology, third ed. New York, Churchill Livingstone. Lemke, G., 1996. Neuregulins in development. Mol. Cell Neurosci. 7, 247–262. Leukauf, C., Szeles, C., Salaymeh, L., Grimm, M., Grabenwoger, M., Losert, U., Moritz, A., Wolner, E., 1993. In vitro and in vivo endothelialization of glutaraldehyde treated bovine pericardium. J. Heart Valve Dis. 2, 230–235. Lewis, J., 1998. Notch signalling and the control of cell fate choices in vertebrates. Semin. Cell Dev. Biol. 9, 583–589. Li, L., Krantz, I.D., Deng, Y., Genin, A., Banta, A.B., Collins, C.C., Qi, M., Trask, B.J., Kuo, W.L., Cochran, J., Costa, T., Pierpont, M.E., Rand, E.B., Piccoli, D.A., Hood, L., Spinner, N.B., 1997. Alagille

Chapter  |  6.2  Signaling Pathways in Valve Formation: The Origin of Congenital Defects

syndrome is caused by mutations in human Jagged1, which encodes a ligand for Notch1. Nat. Genet. 16, 243–251. Lie, J.T., Mann, R.J., Ludwig, J., 1979. The brothers von Winiwarter, Alexander (1848–1917) and Felix (1852–1931), and thromboangiitis obliterans. Mayo. Clin. Proc. 54, 802–807. Liebner, S., Cattelino, A., Gallini, R., Rudini, N., Iurlaro, M., Piccolo, S., Dejana, E., 2004. Beta-catenin is required for endothelial–mesenchymal transformation during heart cushion development in the mouse. J. Cell Biol. 166, 359–367. Lin, A.E., Birch, P.H., Korf, B.R., Tenconi, R., Niimura, M., Poyhonen, M., Armfield Uhas, K., Sigorini, M., Virdis, R., Romano, C., Bonioli, E., Wolkenstein, P., Pivnick, E.K., Lawrence, M., Friedman, J.M., 2000. Cardiovascular malformations and other cardiovascular abnormalities in neurofibromatosis 1. Am. J. Med. Genet. 95, 108–117. Lincoln, J., Alfieri, C.M., Yutzey, K.E., 2004. Development of heart valve leaflets and supporting apparatus in chicken and mouse embryos. Dev. Dyn. 230, 239–250. Lincoln, J., Lange, A.W., Yutzey, K.E., 2006. Hearts and bones: shared regulatory mechanisms in heart valve, cartilage, tendon, and bone development. Dev. Biol. 294, 292–302. Loomes, K.M., Underkoffler, L.A., Morabito, J., Gottlieb, S., Piccoli, D.A., Spinner, N.B., Baldwin, H.S., Oakey, R.J., 1999. The expression of Jagged1 in the developing mammalian heart correlates with cardiovascular disease in Alagille syndrome. Hum. Mol. Genet. 8, 2443–2449. Lopez, D., Duran, A.C., Sans-Coma, V., 2000. Formation of cartilage in cardiac semilunar valves of chick and quail. Ann. Anat. 182, 349–359. Louvi, A., Artavanis-Tsakonas, S., 2006. Notch signalling in vertebrate neural development. Nat. Rev. Neurosci. 7, 93–102. Lyons, K.M., Pelton, R.W., Hogan, B.L., 1990. Organogenesis and pattern formation in the mouse: RNA distribution patterns suggest a role for bone morphogenetic protein-2A (BMP-2A). Development 109, 833–844. Ma, L., Lu, M.F., Schwartz, R.J., and Martin, J.F., 2005. Bmp2 is essential for cardiac cushion epithelial-mesenchymal transition and myocardial patterning. Development 132, 5601–5611. Marino, B., Digilio, M.C., Toscano, A., Giannotti, A., Dallapiccola, B., 1999. Congenital heart diseases in children with Noonan syndrome: An expanded cardiac spectrum with high prevalence of atrioventricular canal. J. Pediatr. 135, 703–706. Markwald, R.R., Fitzharris, T.P., Smith, W.N., 1975. Structural analysis of endocardial cytodifferentiation. Dev. Biol. 42, 160–180. Markwald, R.R., Fitzharris, T.P., Manasek, F.J., 1977. Structural development of endocardial cushions. Am. J. Anat. 148, 85–119. McCright, B., Lozier, J., Gridley, T., 2002. A mouse model of Alagille syndrome: Notch2 as a genetic modifier of Jag1 haploinsufficiency. Development 129, 1075–1082. McDaniell, R., Warthen, D.M., Sanchez-Lara, P.A., Pai, A., Krantz, I.D., Piccoli, D.A., Spinner, N.B., 2006. NOTCH2 mutations cause Alagille syndrome, a heterogeneous disorder of the notch signaling pathway. Am. J. Hum. Genet. 79, 169–173. Mendelson, K., Schoen, F.J., 2006. Heart valve tissue engineering: concepts, approaches, progress, and challenges. Ann. Biomed. Eng. 34, 1799–1819. Meyer, D., Birchmeier, C., 1995. Multiple essential functions of neuregulin in development. Nature 378, 386–390. Milan, D.J., Giokas, A.C., Serluca, F.C., Peterson, R.T., MacRae, C.A., 2006. Notch1b and neuregulin are required for specification of central cardiac conduction tissue. Development 133, 1125–1132.

411

Miquerol, L., Gertsenstein, M., Harpal, K., Rossant, J., Nagy, A., 1999. Multiple developmental roles of VEGF suggested by a LacZ-tagged allele. Dev. Biol. 212, 307–322. Miquerol, L., Langille, B.L., Nagy, A., 2000. Embryonic development is disrupted by modest increases in vascular endothelial growth factor gene expression. Development 127, 3941–3946. Mironov, V., Boland, T., Trusk, T., Forgacs, G., Markwald, R.R., 2003. Organ printing: computer-aided jet-based 3D tissue engineering. Trends Biotechnol. 21, 157–161. Mitjavila-Garcia, M.T., Simonin, C., Peschanski, M., 2005. Embryonic stem cells: meeting the needs for cell therapy. Adv. Drug. Deliv. Rev. 57, 1935–1943. Mjaatvedt, C.H., Lepera, R.C., Markwald, R.R., 1987. Myocardial specificity for initiating endothelial–mesenchymal cell transition in embryonic chick heart correlates with a particulate distribution of fibronectin. Dev. Biol. 119, 59–67. Mjaatvedt, C.H., Markwald, R.R., 1989. Induction of an epithelial–mesenchymal transition by an in vivo adheron-like complex. Dev. Biol. 136, 118–128. Molkentin, J.D., Lin, Q., Duncan, S.A., Olson, E.N., 1997. Requirement of the transcription factor GATA4 for heart tube formation and ventral morphogenesis. Genes Dev. 11, 1061–1072. Montero, J.A., Giron, B., Arrechedera, H., Cheng, Y.C., Scotting, P., Chimal-Monroy, J., Garcia-Porrero, J.A., Hurle, J.M., 2002. Expression of Sox8, Sox9 and Sox10 in the developing valves and autonomic nerves of the embryonic heart. Mech. Dev. 118, 199–202. Mumm, J.S., Kopan, R., 2000. Notch signaling: from the outside in. Dev. Biol. 228, 151–165. Nakagawa, O., Nakagawa, M., Richardson, J.A., Olson, E.N., Srivastava, D., 1999. HRT1, HRT2, and HRT3: a new subclass of bHLH transcription factors marking specific cardiac, somitic, and pharyngeal arch segments. Dev. Biol. 216, 72–84. Nakajima, Y., Yamagishi, T., Nakamura, H., Markwald, R.R., Krug, E.L., 1998. An autocrine function for transforming growth factor (TGF)beta3 in the transformation of atrioventricular canal endocardium into mesenchyme during chick heart development. Dev. Biol. 194, 99–113. Nakajima, Y., Yamagishi, T., Hokari, S., Nakamura, H., 2000. Mechanisms involved in valvuloseptal endocardial cushion formation in early cardiogenesis: roles of transforming growth factor (TGF)-beta and bone morphogenetic protein (BMP). Anat. Rec. 258, 119–127. Nakamura, T., Colbert, M.C., Robbins, J., 2006. Neural crest cells retain multipotential characteristics in the developing valves and label the cardiac conduction system. Circ. Res. 98, 1547–1554. Nasuti 3rd, J.F., Zhang, P.J., Feldman, M.D., Pasha, T., Khurana, J.S., Gorman, J.H., Gorman, R.C., Narula, J., Narula, N., 2004. Fibrillin and other matrix proteins in mitral valve prolapse syndrome. Ann. Thorac. Surg. 77, 532–536. Nemir, M., Croquelois, A., Pedrazzini, T., Radtke, F., 2006. Induction of cardiogenesis in embryonic stem cells via downregulation of Notch1 signaling. Circ. Res. 98, 1471–1478. Ng, L.J., Wheatley, S., Muscat, G.E., Conway-Campbell, J., Bowles, J., Wright, E., Bell, D.M., Tam, P.P., Cheah, K.S., Koopman, P., 1997. SOX9 binds DNA, activates transcription, and coexpresses with type II collagen during chondrogenesis in the mouse. Dev. Biol. 183, 108–121. Nieto, M.A., Bennett, M.F., Sargent, M.G., Wilkinson, D.G., 1992. Cloning and developmental expression of Sna, a murine homologue of the Drosophila snail gene. Development 116, 227–237.

412

Nieto, M.A., Sargent, M.G., Wilkinson, D.G., Cooke, J., 1994. Control of cell behavior during vertebrate development by Slug, a zinc finger gene. Science 264, 835–839. Nollen, G.J., Mulder, B.J., 2004. What is new in the Marfan syndrome? Int. J. Cardiol. 97 (Suppl. 1), 103–108. Noonan, J.A., 1968. Hypertelorism with Turner phenotype. A new syndrome with associated congenital heart disease. Am. J. Dis. Child. 116, 373–380. Oda, T., Elkahloun, A.G., Pike, B.L., Okajima, K., Krantz, I.D., Genin, A., Piccoli, D.A., Meltzer, P.S., Spinner, N.B., Collins, F.S., Chandrasekharappa, S.C., 1997. Mutations in the human Jagged1 gene are responsible for Alagille syndrome. Nat. Genet. 16, 235–242. Okubo, A., Miyoshi, O., Baba, K., Takagi, M., Tsukamoto, K., Kinoshita, A., Yoshiura, K., Kishino, T., Ohta, T., Niikawa, N., Matsumoto, N., 2004. A novel GATA4 mutation completely segregated with atrial septal defect in a large Japanese family. J. Med. Genet. 41, e97. Oosthoek, P.W., Wenink, A.C., Macedo, A.J., Gittenberger-de Groot, A.C., 1997. The parachute-like asymmetric mitral valve and its two papillary muscles. J. Thorac. Cardiovasc. Surg. 114, 9–15. Oshima, M., Oshima, H., Taketo, M.M., 1996. TGF-beta receptor type II deficiency results in defects of yolk sac hematopoiesis and vasculogenesis. Dev. Biol. 179, 297–302. Oukka, M., Ho, I.C., de la Brousse, F.C., Hoey, T., Grusby, M.J., Glimcher, L.H., 1998. The transcription factor NFAT4 is involved in the generation and survival of T cells. Immunity 9, 295–304. Pandit, B., Sarkozy, A., Pennacchio, L.A., Carta, C., Oishi, K., Martinelli, S., Pogna, E.A., Schackwitz, W., Ustaszewska, A., Landstrom, A., et al., 2007. Gain-of-function RAF1 mutations cause Noonan and LEOPARD syndromes with hypertrophic cardiomyopathy. Nat. Genet. 39 (8), 1007–1012. Perez-Pomares, J.M., Foty, R.A., 2006. Tissue fusion and cell sorting in embryonic development and disease: biomedical implications. Bioessays 28, 809–821. Perez-Pomares, J.M., Mironov, V., Guadix, J.A., Macias, D., Markwald, R.R., Munoz-Chapuli, R., 2006. In vitro self-assembly of proepicardial cell aggregates: an embryonic vasculogenic model for vascular tissue engineering. Anat. Rec. A. Discov. Mol. Cell Evol. Biol. 288, 700–713. Person, A.D., Klewer, S.E., Runyan, R.B., 2005. Cell biology of cardiac cushion development. Int. Rev. Cytol. 243, 287–335. Potts, J.D., Runyan, R.B., 1989. Epithelial–mesenchymal cell transformation in the embryonic heart can be mediated, in part, by transforming growth factor beta. Dev. Biol. 134, 392–401. Pu, W.T., Ishiwata, T., Juraszek, A.L., Ma, Q., Izumo, S., 2004. GATA4 is a dosage-sensitive regulator of cardiac morphogenesis. Dev. Biol. 275, 235–244. Pyeritz, R.E., McKusick, V.A., 1979. The Marfan syndrome: diagnosis and management. N. Engl. J. Med. 300, 772–777. Qayyum, S.R., Webb, S., Anderson, R.H., Verbeek, F.J., Brown, N.A., Richardson, M.K., 2001. Septation and valvar formation in the outflow tract of the embryonic chick heart. Anat. Rec. 264, 273–283. Radtke, F., Raj, K., 2003. The role of Notch in tumorigenesis: oncogene or tumour suppressor? Nat. Rev. Cancer 3, 756–767. Rahimtoola, S.H., 2003. The next generation of prosthetic heart valves needs a proven track record of patient outcomes at  or 15 to 20 years. J. Am. Coll. Cardiol. 42, 1720–1721. Ramsdell, A.F., Markwald, R.R., 1997. Induction of endocardial cushion tissue in the avian heart is regulated, in part, by TGFbeta-3-mediated autocrine signaling. Dev. Biol. 188, 64–74. Ranger, A.M., Grusby, M.J., Hodge, M.R., Gravallese, E.M., de la Brousse, F.C., Hoey, T., Mickanin, C., Baldwin, H.S., Glimcher, L.H.,

PART  |  6  Cushions, Valves and Septa

1998. The transcription factor NF-ATc is essential for cardiac valve formation. Nature 392, 186–190. Ransom, J., Srivastava, D., 2007. The genetics of cardiac birth defects. Semin. Cell Dev. Biol. 18, 132–139. Razzaque, M.A., Nishizawa, T., Komoike, Y., Yagi, H., Furutani, M., Amo, R., Kamisago, M., Momma, K., Katayama, H., Nakagawa, M., Fujiwara, Y., Matsushima, M., Mizuno, K., Tokuyama, M., Hirota, H., Muneuchi, J., Higashinakagawa, T., Matsuoka, R., 2007. Germline gain-of-function mutations in RAF1 cause Noonan syndrome. Nat. Genet. 39 (8), 1013–1017. Rentschler, S., Zander, J., Meyers, K., France, D., Levine, R., Porter, G., Rivkees, S.A., Morley, G.E., Fishman, G.I., 2002. Neuregulin-1 promotes formation of the murine cardiac conduction system. Proc. Natl. Acad. Sci. USA 99, 10464–10469. Riese 2nd, D.J., Stern, D.F., 1998. Specificity within the EGF family/ ErbB receptor family signaling network. Bioessays 20, 41–48. Rivera-Feliciano, J., Tabin, C.J., 2006. Bmp2 instructs cardiac progenitors to form the heart-valve-inducing field. Dev. Biol. 295, 580–588. Rivera-Feliciano, J., Lee, K.H., Kong, S.W., Rajagopal, S., Ma, Q., Springer, Z., Izumo, S., Tabin, C.J., Pu, W.T., 2006. Development of heart valves requires Gata4 expression in endothelial-derived cells. Development 133, 3607–3618. Roberts, A.E., Araki, T., Swanson, K.D., Montgomery, K.T., Schiripo, T.A., Joshi, V.A., Li, L., Yassin, Y., Tamburino, A.M., Neel, B.G., Kucherlapati, R.S., 2007. Germline gain-of-function mutations in SOS1 cause Noonan syndrome. Nat. Genet. 39, 70–74. Robinson, P.N., Arteaga-Solis, E., Baldock, C., Collod-Beroud, G., Booms, P., De Paepe, A., Dietz, H.C., Guo, G., Handford, P.A., Judge, D.P., Kielty, C.M., Loeys, B., Milewicz, D.M., Ney, A., Ramirez, F., Reinhardt, D.P., Tiedemann, K., Whiteman, P., Godfrey, M., 2006. The molecular genetics of Marfan syndrome and related disorders. J. Med. Genet. 43, 769–787. Romano, L.A., Runyan, R.B., 1999. Slug is a mediator of epithelial–mesenchymal cell transformation in the developing chicken heart. Dev. Biol. 212, 243–254. Rones, M.S., McLaughlin, K.A., Raffin, M., Mercola, M., 2000. Serrate and Notch specify cell fates in the heart field by suppressing cardiomyogenesis. Development 127, 3865–3876. Rothermel, B., Vega, R.B., Yang, J., Wu, H., Bassel-Duby, R., Williams, R.S., 2000. A protein encoded within the Down syndrome critical region is enriched in striated muscles and inhibits calcineurin signaling. J. Biol. Chem. 275, 8719–8725. Ruckman, R.N., Van Praagh, R., 1978. Anatomic types of congenital mitral stenosis: report of 49 autopsy cases with consideration of diagnosis and surgical implications. Am. J. Cardiol. 42, 592–601. Runyan, R.B., Markwald, R.R., 1983. Invasion of mesenchyme into three-dimensional collagen gels: a regional and temporal analysis of interaction in embryonic heart tissue. Dev. Biol. 95, 108–114. Rutenberg, J.B., Fischer, A., Jia, H., Gessler, M., Zhong, T.P., Mercola, M., 2006. Developmental patterning of the cardiac atrioventricular canal by Notch and Hairy-related transcription factors. Development 133, 4381–4390. Sanford, L.P., Ormsby, I., Gittenberger-de Groot, A.C., Sariola, H., Friedman, R., Boivin, G.P., Cardell, E.L., Doetschman, T., 1997. TGFbeta2 knockout mice have multiple developmental defects that are non-overlapping with other TGFbeta knockout phenotypes. Development 124, 2659–2670. Schubbert, S., Zenker, M., Rowe, S.L., Boll, S., Klein, C., Bollag, G., van der Burgt, I., Musante, L., Kalscheuer, V., Wehner, L.E., Nguyen, H., West, B., Zhang, K.Y., Sistermans, E., Rauch, A., Niemeyer, C.M.,

Chapter  |  6.2  Signaling Pathways in Valve Formation: The Origin of Congenital Defects

Shannon, K., Kratz, C.P., 2006. Germline KRAS mutations cause Noonan syndrome. Nat. Genet. 38, 331–336. Servidei, S., Bertini, E., DiMauro, S., 1994. Hereditary metabolic cardiomyopathies. Adv. Pediatr. 41, 1–32. Shalaby, F., Rossant, J., Yamaguchi, T.P., Gertsenstein, M., Wu, X.F., Breitman, M.L., Schuh, A.C., 1995. Failure of blood-island formation and vasculogenesis in Flk-1-deficient mice. Nature 376, 62–66. Shi, Y., Massague, J., 2003. Mechanisms of TGF-beta signaling from cell membrane to the nucleus. Cell 113, 685–700. Small, E.M., Krieg, P.A., 2003. Transgenic analysis of the atrialnatriuretic factor (ANF) promoter: Nkx2-5 and GATA-4 binding sites are required for atrial specific expression of ANF. Dev. Biol. 261, 116–131. Solloway, M.J., Robertson, E.J., 1999. Early embryonic lethality in Bmp5;Bmp7 double mutant mice suggests functional redundancy within the 60A subgroup. Development 126, 1753–1768. Stainier, D.Y., Beis, D., Jungblut, B., Bartman, T., 2002. Endocardial cushion formation in zebrafish. Cold Spring Harb. Symp. Quant. Biol. 67, 49–56. Sugi, Y., Yamamura, H., Okagawa, H., Markwald, R.R., 2004. Bone morphogenetic protein-2 can mediate myocardial regulation of atrioventricular cushion mesenchymal cell formation in mice. Dev. Biol. 269, 505–518. Sumida, H., Akimoto, N., Nakamura, H., 1989. Distribution of the neural crest cells in the heart of birds: a three dimensional analysis. Anat. Embryol. (Berlin) 180, 29–35. Sznajer, Y., Keren, B., Baumann, C., Pereira, S., Alberti, C., Elion, J., Cave, H., Verloes, A., 2007. The spectrum of cardiac anomalies in Noonan syndrome as a result of mutations in the PTPN11 gene. Pediatrics 119, e1325–e1331. Tartaglia, M., Mehler, E.L., Goldberg, R., Zampino, G., Brunner, H.G., Kremer, H., van der Burgt, I., Crosby, A.H., Ion, A., Jeffery, S., Kalidas, K., Patton, M.A., Kucherlapati, R.S., Gelb, B.D., 2001. Mutations in PTPN11, encoding the protein tyrosine phosphatase SHP-2, cause Noonan syndrome. Nat. Genet. 29, 465–468. Tartaglia, M., Kalidas, K., Shaw, A., Song, X., Musat, D.L., van der Burgt, I., Brunner, H.G., Bertola, D.R., Crosby, A., Ion, A., Kucherlapati, R.S., Jeffery, S., Patton, M.A., Gelb, B.D., 2002. PTPN11 mutations in Noonan syndrome: molecular spectrum, genotype-phenotype correlation, and phenotypic heterogeneity. Am. J. Hum. Genet. 70, 1555–1563. Tartaglia, M., Pennacchio, L.A., Zhao, C., Yadav, K.K., Fodale, V., Sarkozy, A., Pandit, B., Oishi, K., Martinelli, S., Schackwitz, W., et al., 2007. Gain-of-function SOS1 mutations cause a distinctive form of Noonan syndrome. Nat. Genet. 39, 75–79. Taylor, P.M., Allen, S.P., Yacoub, M.H., 2000. Phenotypic and functional characterization of interstitial cells from human heart valves, pericardium and skin. J. Heart Valve Dis. 9, 150–158.

413

Thiery, J.P., Sleeman, J.P., 2006. Complex networks orchestrate epithelial–mesenchymal transitions. Nat. Rev. Mol. Cell Biol. 7, 131–142. Timmerman, L.A., Grego-Bessa, J., Raya, A., Bertran, E., Perez-Pomares, J.M., Diez, J., Aranda, S., Palomo, S., McCormick, F., Izpisua-Belmonte, J.C., de la Pompa, J.L., 2004. Notch promotes epithelial–mesenchymal transition during cardiac development and oncogenic transformation. Genes Dev. 18, 99–115. Uyttendaele, H., Closson, V., Wu, G., Roux, F., Weinmaster, G., Kitajewski, J., 2000. Notch4 and Jagged-1 induce microvessel differentiation of rat brain endothelial cells. Microvasc. Res. 60, 91–103. Vesely, I., 2005. Heart valve tissue engineering. Circ. Res. 97, 743–755. Vesely, I., Shi, Y., Dobkin, D., Iyer, R., Soundararajan, A., 2005. Progress in developing a composite tissue-engineered aortic valve. Conf. Proc. IEEE Eng. Med. Biol. Soc. 7, 7129–7130. Watanabe, Y., Kokubo, H., Miyagawa-Tomita, S., Endo, M., Igarashi, K., Aisaki, K.I., Kanno, J., Saga, Y., 2006. Activation of Notch1 signaling in cardiogenic mesoderm induces abnormal heart morphogenesis in mouse. Development 133, 1625–1634. Wessels, A., Markman, M.W., Vermeulen, J.L., Anderson, R.H., Moorman, A.F., Lamers, W.H., 1996. The development of the atrio­ ventricular junction in the human heart. Circ. Res. 78, 110–117. Wilkins, B.J., De Windt, L.J., Bueno, O.F., Braz, J.C., Glascock, B.J., Kimball, T.F., Molkentin, J.D., 2002. Targeted disruption of NFATc3, but not NFATc4, reveals an intrinsic defect in calcineurin-mediated cardiac hypertrophic growth. Mol. Cell Biol. 22, 7603–7613. Winnier, G., Blessing, M., Labosky, P.A., Hogan, B.L., 1995. Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 9, 2105–2116. Woodrow, M.A., Rayter, S., Downward, J., Cantrell, D.A., 1993. P21ras function is important for T cell antigen receptor and protein kinase C regulation of nuclear factor of activated T cells. J. Immunol. 150, 3853–3861. Wrana, J.L., Attisano, L., Wieser, R., Ventura, F., Massague, J., 1994. Mechanism of activation of the TGF-beta receptor. Nature 370, 341–347. Ya, J., van den Hoff, M.J., de Boer, P.A., Tesink-Taekema, S., Franco, D., Moorman, A.F., Lamers, W.H., 1998. Normal development of the outflow tract in the rat. Circ. Res. 82, 464–472. Yarden, Y., Sliwkowski, M.X., 2001. Untangling the ErbB signalling network. Nat. Rev. Mol. Cell Biol. 2, 127–137. Zhang, H., Bradley, A., 1996. Mice deficient for BMP2 are nonviable and have defects in amnion/chorion and cardiac development. Development 122, 2977–2986. Zhao, Q., Eberspaecher, H., Lefebvre, V., De Crombrugghe, B., 1997. Parallel expression of Sox9 and Col2a1 in cells undergoing chondrogenesis. Dev. Dyn. 209, 377–386.

Chapter 7.1

Role of Cardiac Neural Crest Cells in Morphogenesis of the Heart and Great Vessels Kimberly E. Inman1, Max Ezin3, Marianne Bronner-Fraser3 and Paul A. Trainor1,2 1

Stowers Institute for Medical Research, Kansas City, MO, USA Kansas University School of Medicine, Kansas City, KS, USA 3 California Institute for Technology, Pasadena, CA, USA 2

I.  Introduction The neural crest is a transient migratory cell population found in all vertebrates that generates an extraordinary diversity of cell and tissue derivatives, including neurons and glia of the peripheral nervous system, smooth muscle, connective tissue, melanocytes, cartilage and bone. Due to its fundamental importance in embryonic development and vertebrate evolution (Northcutt and Gans, 1983), the neural crest has been considered as the fourth germ layer (Hall, 2000). Derived from the dorsal neuroectoderm during early embryogenesis, neural crest cells are generated along almost the entire length of the neuraxis in a rostrocaudal wave. However, neural crest cells only become morphologically detectable as they emigrate from the neural tube, a delamination process that requires an epithelial-tomesenchymal conversion of neuroectoderm cells. At each axial level, neural crest cells migrate along unique pathways, contributing to specific cell and tissue types. On the basis of these unique migratory and differentiation characteristics the neural crest can be subdivided into at least four distinct axial populations: cranial; cardiac; vagal; and trunk. In this chapter we discuss the development of the cardiac neural crest and recent advances in our understanding of the patterning of this special cell population (see also Chapter 7.2).

II.  Neural crest formation Vertebrate neurogenesis commences during gastrulation with the formation of the neural plate, which is initially a columnar neuroepithelial sheet that extends the length of Heart Development and Regeneration Copyright © 2010 2010 Elsevier Inc. All rights of reproduction in any form reserved.

the body axis. Neural crest cells are born at the interface between non-neural ectoderm (presumptive epidermis) and the dorsal region of the neural plate, a territory referred to as the neural plate border. Neural crest cell induction requires contact-mediated tissue interactions between the neural plate and surface ectoderm (Rollhauser-ter Horst, 1977; Moury and Jacobson, 1990), both of which have been shown from lineage tracing to give rise to neural crest cells (Selleck and Bronner-Fraser, 1995). The generation of neural crest cells is a multistep process involving numerous morphogen signaling pathways and a network of transcriptional gene activity (Fig. 1). This genetic cascade is initiated by extracellular signaling molecules like Wnts, FGFs, BMPs and Notch that cooperate at the border between neural and non-neural ectoderm to upregulate a set of transcription factors termed “neural plate border specifiers.” These include Zic, Msx, Dlx3/5 and Pax3/7 that define the neural plate border territory from which neural crest cells will arise. Each of these “neural plate border specifiers” in turn regulate another set of transcription factors that specify neural crest migratory properties and cell fate decisions. These “neural crest specifier genes” include Snail, AP-2, FoxD3, Twist, Id, cMyc and Sox9/10. Recent evidence for involvement of a morphogen-transcription cascade has been obtained in avian embryos where Pax7 was shown to be absolutely required for specification of the neural crest genes Snail2, Sox9 and Sox10 in a Wnt/ BMP-dependent manner (Basch et al., 2006). Furthermore Pax3 was shown in Xenopus embryos to act downstream of Msx1 and to induce Slug in a WNT-dependent manner (Monsoro-Burq et al., 2003). In addition, it was determined that Wnts and FGF8 signals act in parallel on the

417

418

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

Figure 1  Molecular signaling in neural crest induction. Schematic drawing of a transverse section of a chick embryo at the time of neural crest induction. Extracellular signaling molecules, such as Wnt and BMP in the non-neural ectoderm (light blue), low levels of BMP in the neural ectoderm (dark blue) and FGF in the mesoderm (red) cooperatively upregulate transcription factors termed “neural plate border specifiers”. These specifiers, such as Zic, Msx and Pax3/7, define the territory at the boundary of neural and non-neural ectoderm from which neural crest cells (green) will arise. Neural plate border specifiers in turn regulate transcription factors such as Sox9/10, Slug, Snail and FoxD3 that specify neural crest cell migratory properties and cell fate decisions.

neural plate border to induce neural crest through either Pax3 alone (WNT) or through both Msx1 (FGF) and Pax3 transcriptional activities (Monsoro-Burq et al., 2005). The formation of neural crest cells requires substantial cytoarchitectural and cell adhesive changes, and at the heart of this process are members of the Snail zincfinger transcription factor gene family (Nieto et al., 1994; Mayor et al., 1995) that repress the cell adhesion molecule E-cadherin (Cano et al., 2000). E-cadherin is downregulated during the epithelial-to-mesenchymal transition (EMT) that converts a neuroepithelial cell to a neural crest cell. In Xenopus embryos, the inhibition of Snail2 leads to maintenance of E-cadherin activity and blocks neural crest cell formation (Carl et al., 1999; LaBonne and Bronner-Fraser, 2000). Similarly, antisense oligonucleotides directed against Snail2 in avian embryos also inhibit the epithelial-to-mesenchymal conversion of neuroepithelial cells into neural crest cells (Nieto et al., 1994). Conversely, overexpression of Snail2 in the neural tube of avian embryos enhances the production of migrating neural crest cells which highlights the importance of Snail2 in neural crest cell formation (del Barrio and Nieto, 2002). Coordinated activity of Slug, Sox9 and FoxD3 appear to be required for proper specification of the neural crest (Cheung et al., 2005). Sox9 establishes the competence to respond to signals that promote EMT, Slug is responsible for the onset of the EMT and FoxD3 regulates expression of cell adhesion molecules required for neural crest migration.

Neural crest cell delamination and migration from the neural tube is aided by major changes in cell adhesion that includes the downregulation of NCAM, N-cadherin and cad6B, together with the upregulation of cad7 and cad11 (Akitaya and Bronner-Fraser, 1992; Inoue et al., 1997; Nakagawa and Takeichi, 1998; Cano et al., 2000). Interestingly, cad6B has recently been identified as one of the first direct targets of Snail2 in avian neural crest cells (Taneyhill et al., 2007). Although initially expressed in the neuroepithelium, cad6B is rapidly downregulated as neural crest cells delaminate and emigrate from the neural tube (Nakagawa and Takeichi, 1995, 1998). Thus, Snail2 directly represses cad6B transcription in a highly dynamic manner during neural crest emigration. The cadherin switch demonstrates that a regulated balance of adhesive activity is needed for emigration, and this idea is supported by the overexpression of neuroepithelial cadherins which prevent neural crest delamination and migration (Nakagawa and Takeichi, 1998).

III.  The cardiac neural crest The cardiac neural crest was identified over 20 years ago as a specific subpopulation of cranial neural crest cells (Kirby et al., 1983, 1985). Generated from the neural plate between the axial boundaries of the otic placode and somite 3, cardiac neural crest cells migrate into the third, fourth and sixth branchial arches. Cardiac neural

Chapter  |  7.1  Role of Cardiac Neural Crest Cells in Morphogenesis of the Heart and Great Vessels

crest cells invade the heart through both the arterial pole (outflow tract) and the venous pole (inflow tract), and collectively give rise to parasympathetic cardiac ganglia, the smooth muscle layer (the tunica media) of the great vessels, conotruncal cushions and the aorticopulmonary septum component of the outflow tract of the heart (Le Lievre and Le Douarin, 1975; Kirby et al., 1983; Bockman et al., 1987; Noden, 1991; Poelmann et al., 2004). In avian embryos, ablation of the cardiac neural crest prior to migration results in defective patterning of the aortic arch arteries, cardiac outflow tract and great vessels (Kirby et al., 1985). Interestingly, cardiac neural crest is required not only for normal heart morphogenesis, but also for differentiation of the thymus, parathyroid and thyroid glands. An analogous cardiac-specific neural crest cell population has also been identified in mammals. Lineage tracing in rat and mouse embryos (Fukiishi and Morriss-Kay, 1992; Serbedzija et al., 1992; Osumi-Yamashita et al., 1996; Chan et al., 2004) revealed that between 8.5–9.5 days post coitum (dpc), a subpopulation of neural crest cells delimited rostrally by rhombomere 5 in the hindbrain and caudally by somite 3, migrated ventrolaterally and populated the third, fourth and sixth branchial arches in a pattern similar to that observed in chick embryos (Fig. 2A–C). Neural crest cells at the level of somite 2 contribute most significantly to the outflow tract (Waldo et al., 1999; Maschhoff and Baldwin, 2000; Hutson and Kirby, 2003). However, the limitation of whole embryo culture precluded an analysis of the long-term tissue contribution of cardiac neural crest cells. The first molecular evidence of long-term neural crest cell contribution to the mammalian heart came from the analysis of transgenic mice containing a lacZ reporter construct under control of the Connexin 43 (Cx43) promoter (Lo et al., 1997). In a side-by-side comparison of Cx43-lacZ expression in mouse with the migration of cardiac neural crest in quail-chick chimeras, Cx43 expression was found to have a close correlation with the spatio–temporal migration of neural crest cells into the outflow tract (Waldo et al., 1999). Several differences, however, exist, most of which can be related to the morphological differences between the hearts of mammals and birds. The most striking difference is the complete absence of labeled cells in the pulmonary artery and the limited contribution to the semilunar valves in the mouse, all of which are derived from neural crest in chick embryos. It is important to note, however, that Cx43lacZ may not label all neural crest cells, and furthermore it does label tissues such as the epicardium and pericardium which are not derived from neural crest cells. To further clarify the fate of cardiac neural crest cells in the mouse, several groups have developed P0-cre, Wnt1-cre and Pax3-cre transgenic lines for indelible long-term lin­ eage tracing (Yamauchi et al., 1999; Jiang et al., 2000; Li et al., 2000). Although P0-, Wnt1- and Pax3-expression

419

does not persist in neural crest cells as they enter the heart, these three experimental systems enabled clear detection of the cardiac neural crest-derived cells in ­embryonic ­tissues in patterns similar to Cx43lacZ (Fig. 2D–L). More importantly, labeled neural crest descendants are readily observed in the postnatal mouse heart, where they are found in a pattern that reflects their distribution in the midgestation embryo. Neural crest-derived cells contribute to the smooth muscle layers of the ascending aorta, aortic arch, proximal carotid and coronary arteries (Yamauchi et al., 1999; Jiang et al., 2000; Li et al., 2000). Cardiac neural crest cells invade the proximal and distal outflow tract cushions and contribute to semilunar valve formation (de Lange et al., 2004). Moreover, only a limited contribution is seen in the aortic and pulmonary valves, and no contribution to the postnatal pulmonary artery has been detected (Nakamura et al., 2006). The timing of emigration from the neural tube influences the derivatives to which the cardiac neural crest will contribute. The first wave of neural crest cells to emigrate forms the condensed mesenchyme of the outflow tract of the heart (the aorticopulmonary septum), as well as contributing to the vessels of the heart. Later waves of emigrating neural crest cells contribute primarily to the proximal part of the pharyngeal arteries, where they are proposed to play a role in arterial remodeling (Boot et al., 2003). Interestingly, the arrival of neural crest cells in the heart coincides temporally with electrophysiological changes that determine stroke volume and ejection fraction in the heart which are essential to circulatory development (Poelmann and Gittenberger-de Groot, 1999; Li et al., 2003; Gurjarpadhye et al., 2007). Additional evidence for nonmesoderm contributions to the heart come from fate mapping of mesoderm derivatives during development. Mesp1 is a basic helix-loop-helix transcription factor that has been detected in all mesodermal precursors of the developing mouse cardiovascular system (Saga et al., 1999, 2000). Recently, Mesp1-cre transgenic mice have been used to fate map long-term mesoderm contribution to the heart between 9.5–13.5 dpc (Kitajima et al., 2006). Although a large portion of the developing heart contained labeled cells, cardiogenic cells did not entirely derive from Mesp1-expressing cells. A population of cells in the outflow tract cushions and cells along the interventricular septum were unlabeled, indicating they did not descend from the Mesp1 population. Further analyses revealed that the Mesp1-nonexpressing cells of the outflow tract were derived from neural crest cells. These results highlight that the atrioventricular cushions are derived primarily from Mesp1-expressing endocardium, whereas the outflow tract cushions are derived mainly from neural crest. Additionally, the unlabeled cells of the interventricular septum, which contribute ­ approximately 20% of the cells comprising the ventricular cardiac conduction system, are neither of neural crest nor Mesp1 origin, ­revealing

420

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

(A)

(B)

(C)

Pax3-cre

Wnt1-cre

(D)

Cx43lacZ

(E)

(F)

PT

PT Ao

PT

Ao

Ao

(G)

(H)

(I)

(J)

(K)

(L)

Figure 2  Fatemapping studies identify the cardiac neural crest and its descendants in the mouse embryo. (A–C) Schematic depiction of DiI lineage tracing of the mouse cardiac neural crest. The fluorescent lineage tracer DiI was injected into the developing mouse hindbrain corresponding to the prorhombomeric regions (proRh) (B) and (C) prior to the onset of neural crest migration. Red circles represent neural crest cells migrating into branchial arches 2, 3, 4 and the heart, similar to the migration of the chick cardiac crest population. The otic vesicle is outlined in yellow. (D–L) Genetic fatemapping using (D, G, J) Pax3-cre, (E, H, K) Wnt1-cre and (F, I, L) Cx43lacZ mouse models revealed significant contribution of neural crest descendants to the developing heart at 12.5dpc. While there are slight differences in the number of labeled cells detected in each model, all show neural crest cell contribution surrounding the (D–F) aorta (Ao) and pulmonary trunk (PT) and (J–L) the outflow tract cushions (arrows in G–L). (1, 2, 3, 4: branchial arches 1–4; H: heart). (D–L) Reproduced with permission from the Company of Biologists from Brown, et al. (2001) Development 128, 3071–3080.

another contributor to the complex development of the heart. Populations of neural crest-derived cells within the septa and outflow tract are surrounded by mesodermderived cells, however, it remains to be seen how defects in mesoderm contribution to the heart affect neural crest cell differentiation and cardiac patterning.

IV.  Evolution of the cardiac   neural crest The most important role of the circulatory system is the continual supplementation of tissues and organs with oxygen, in concert with the removal of carbon dioxide.

Chapter  |  7.1  Role of Cardiac Neural Crest Cells in Morphogenesis of the Heart and Great Vessels

Animals have evolved a system that includes blood for carrying gas, vessels for transporting blood to points of gas exchange and a heart that provides the force to move gas-rich blood through the vessels that course the body. Although few experiments have investigated the evolutionary origins of the cardiac neural crest, it is clear that cardiac neural crest cells appeared before the development of the four-chambered heart. Furthermore, cardiac neural crest cells preceded septation of the outflow tract. Fish, with only a two-chambered heart and a nonseptated outflow tract, use their gills for gas exchange. However, fish possess a population of neural crest cells that migrate to the heart. These cardiac neural crest cells originate more cranially in fish than they do in chick and mouse, and the fate maps show that these cells develop into myocardial cells across the entirety of the heart, including the outflow tract, the atrium and ventricle (Li et al., 2003). The fish bulbus arteriosus, which is similar to the arterial trunks (aorta and pulmonary artery) in avians, remains nonseptated, and is not colonized by cardiac neural crest (Grimes et al., 2006). Interestingly, in Xiphophorus fish, HNK1 staining has revealed the presence of neural crest in the vasculature of the gill arches which form part of the respiratory organ (Sadaghiani and Vielkind, 1990). Furthermore, fate-mapping studies in salamanders, which have a three-chambered heart and a septated outflow tract, demonstrated that cardiac neural crest cells, originating from the hindbrain, populate the aortic arches, the truncus arteriosus and the outflow tract (Bashir and Armstrong, 1999). In invertebrates such as Drosophila, a cell population known as heart-anchoring cells (HANC) shares some morphogenetic and functional similarities with vertebrate cardiac neural crest cells (Zikova et al., 2003). HANCs express the homeobox gene ladybird (lb) which is orthologous to Lbx1 in mouse; a gene that demarcates a subpopulation of cardiac neural crest cells (Schafer et al., 2003). Like cardiac neural crest cells, HANCs are extracardiac in origin and migrate over substantial distances to the heart, where they form part of the cardiac outflow tissues. However, in contrast to vertebrate neural crest cells, HANCs originate in the head epidermis, not the neural tube and migrate as part of an epidermal sheet, as opposed to individual cell migration. Furthermore, HANCs function within cardiac outflow muscles, which have no homologous population in vertebrates. Thus, invertebrate HANCs and vertebrate neural crest cells are unlikely to be homologous cell populations despite the critical importance of their species-specific contributions to heart development. Nonetheless, it is clear that as the heart has evolved, neural crest cells have continued to play an increasingly important role in cardiac morphogenesis and patterning, particularly with respect to cardiovascular remodeling.

421

V.  Cardiac neural crest cells and morphogenesis of the heart and great vessels Coordinated development of the heart and blood vessels is required for the proper establishment of the circulatory system. Integral to this process is the formation and subsequent remodeling of the great arteries that arise ventrally from the aortic sac, penetrate the branchial arches and terminate in the dorsal aorta. The great arteries of the head and heart begin as six symmetrically-paired arch artery precursors that develop in a rostral-to-caudal direction during embryogenesis (Fig. 3). In the mouse, the first and second pair of embryonic arch arteries forms at approximately 8.5–9.0 dpc. However, by 11.0 dpc, remodeling occurs so that the first two arch arteries regress becoming capillaries within branchial arches 1 and 2 (Kaufman, 1992). At the same time, the third, fourth and sixth symmetrical pair of arch arteries is generated, but by 11.5 dpc remodeling results in the asymmetrical regression and persistence of specific arch arteries. The arteries in the left and right third branchial arch become the carotid arteries. The right fourth arch artery regresses, providing a connection to the vertebral and subclavian arteries. In contrast, the left fourth arch artery persists as a larger vessel that forms a portion of the mature aortic arch. The right sixth arch artery is the major capillary network associated with the developing trachea and lungs. The left sixth arch artery forms the ductus arteriosus, which forms a shunt between the lung vasculature and the dorsal aorta during gestation. This shunt collapses after cessation of placental circulation at birth and initiation of postnatal oxygenation of the blood by the lungs (Gittenberger-de Groot et al., 2005). Although the molecular control of this complex remodeling process is not well-understood, the cardiac neural crest appears to be required for the successful remodeling of the aortic arch arteries into the mature asymmetrical arterial system. When the cardiac neural crest is ablated in developing chick embryos, operated embryos lack a variable combination of the normally persisting aortic arch arteries derived from branchial arches 3, 4 and 6 (Nishibatake et al., 1987). Absence of the right fourth aortic arch artery leads to development of interrupted aortic arch, while absence of the right sixth aortic arch artery results in the absence of the ductus arteriosus. Rather than playing a role in the formation of the aortic arch system, the cardiac neural crest is required for the persistence of the caudal aortic arch arteries. Further studies of cardiac crest ablation in the chick revealed that in the absence of neural crest in the caudal pharynx, the third aortic arch artery developed from endothelial cords and formed a lumen similar to controls (Waldo et al., 1996). Normally, neural crest cells separate the developing aortic arch arteries from the pharyngeal endoderm. However, in

422

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

(A) Normal Regression of right dorsal aorta

(B)

(C) 1 2 3 4 6

(D) Double aortic arch Persistence of right dorsal aorta and right 4th arch artery

Retroesophageal right subclavian artery Persistence of right dorsal aorta with regression of right 4th arch artery

Interrupted aortic arch type B Regression of left 4th arch artery Figure 3  Normal and defective remodeling of the branchial arch arteries to form the great vessels of the heart. Schematic drawing of the symmetric aortic arch arteries (center). Branchial arch arteries 1 and 2 regress to small capillaries, while the third (green), fourth (violet) and sixth (blue) aortic arch arteries contribute to the mature arrangement of the great vessels. (A) Normal regression and remodeling results in appropriate positioning of the vessels. Abnormal persistence or regression of the fourth arch artery can lead to congenital heart defects such as: (B) double aortic arch which can compress the trachea and esophagus; (C) retroesophageal right subclavian artery; or (D) interrupted aortic arch that require surgical intervention after birth. Reprinted from Stoller, J. Z., and Epsttein, J. A. (2005) Seminars in Cell and Developmental Biology 16, 704–715. © Elsevier Ltd. 2005, with permission from Elsevier.

c­ ardiac neural crest ablated embryos, the endothelial cells remained in contact with the endoderm. Additionally, the arch artery was greatly reduced in size and had an uneven lumen, often resulting in direct contact of the artery with both endoderm and ectoderm and loss of bilateral symmetry of the arch arteries. The cardiac neural crest cells ultimately contribute to the tunica media smooth muscle of the aortic arch arteries (Waldo et al., 1999) and as such, the cells may play a role in stabilizing the developing endothelial tube. Alternatively, the neural crest cells may provide signaling molecules required for the maintenance and growth of the endothelial tube. As detailed later in this chapter, recent studies in mouse have provided some insight into the genetic and molecular control of aortic arch remodeling.

In contributing to septation of the outflow tract, cardiac neural crest cells migrate from the distal to the proximal portion of the outflow tract, located at the junction with the right ventricle, and position themselves in the center as a mass of condensed mesenchyme. This mass forms the wall that separates the aorta from the pulmonary trunk, establishing a mesenchymal septum. The majority of the neural crest in the outflow tract undergoes apoptosis and this is thought to act as a signal for myocardialization of the septum. The process of outflow tract septation demands that key vessel remodeling takes place concomitantly. The future pulmonary trunk and the future aorta start out as a single tube leading out of the right ventricle. However, the pulmonary trunk and the aorta must position themselves above the correct chambers. A significant rotation

Chapter  |  7.1  Role of Cardiac Neural Crest Cells in Morphogenesis of the Heart and Great Vessels

­ ovement occurs, whereby the aorta repositions itself as m the outflow tube above the left ventricle and the pulmonary trunk remains above the right ventricle. In addition, the anlagen of the great vessels from the aorta and the heart are housed in the pharyngeal arches and undergo remodeling to adopt their final configuration. The early wave of migrating cardiac neural crest cells participates in the cleavage of the single outflow tract into separate pulmonary and aortic trunk and populates the aortic arch arteries, while the late wave of migrating neural crest cells plays a key role in the complex remodeling of the great vessels and their branches (Boot et al., 2003). Aberrant remodel­ ing results in congenital cardiac anomalies such as interrupted aortic arch and double aortic arch, both of which are associated with ventricular septal defects (Fig. 3). Cardiac muscle cells can beat autonomously, albeit in a disorganized manner. The proper integration of the beating of the heart with the filling and emptying of blood, together with environmental cues for stress or relaxation, requires nervous input. Innervation to the heart comprises nerve bundles within the heart – called the cardiac conduction system – as well as nerves from the peripheral nervous system, including both sympathetic and parasympathetic innervation. The cardiac conduction system integrates the contraction of the atria and ventricles with blood flow out of the heart. This conduction system consists of the sinoatrial node, the atrioventricular node and the His bundle, all of which are derived from cardiomyogeneic precursors. Cardiac neural crest does not contribute to the cardiac conduction system, however the spatial proximity of neural crest derivatives suggest a possible role in its induction (Poelmann et al., 2004). In contrast, both the sympathetic nervous system and the parasympathetic nervous system derive from the neural crest (Cheng et al., 1999; Jiang et al., 2000). Neural crest cells form the nerve fibers that constitute sympathetic innervation of the heart, as well as the cardiac branches of the vagal nerve and the cardiac ganglia that contribute to the parasympathetic network (Kirby and Stewart, 1984; Verberne et al., 1998, 1999).

VI.  Interactions between cardiac neural crest and mesoderm Ablation of cardiac neural crest cells leads to a heart defect known as persistent truncus arteriosis, a condition caused by absence of truncal and aorticopulmonary septa resulting in the fusion of the pulmonary and aortic blood vessels (Kirby et al., 1985; Besson et al., 1986; Nishibatake et al., 1987; Kirby and Waldo, 1995). Interestingly, the heterotopic transplantation of mesencephalic or trunk neural crest in place of cardiac neural crest results in heart defects similar to cardiac crest ablation (Kirby, 1989). Together with models of neural crest patterning derived from classical transplantation studies in amphibian (Andres, 1946,

423

1949; Wagner, 1949, 1959) and avian embryos (Noden, 1983), this led to the idea that cranial neural crest cells, of which the cardiac neural crest is a subset, receive patterning information prior to their emigration according to their axial origin within the hindbrain. Furthermore, this patterning information was then imposed on the surrounding tissues of the branchial arches and their derivatives. However, transplantation studies in mouse and zebrafish embryos (Trainor and Krumlauf, 2000; Schilling, 2001) have revealed a high degree of neural crest cell plasticity. Heterotopic transposition of neuroepithelial cells within the developing hindbrain revealed that graftderived neural crest cells migrate according to their new location, rather than rerouting to their original axial level. Additionally, graft-derived cells adapt their genetic fates to match those appropriate for their new location (Trainor and Krumlauf, 2000; Schilling, 2001). This not only implied that neural crest cells are not prepatterned, but also suggested that patterning of the branchial arches might involve interactions between the neural crest cells and other tissue components of the arch, including the mesoderm, ectoderm and endoderm. Further experiments revealed this to be true. When isolated neural crest cells were transplanted to ectopic sites, the dispersed neural crest cells lacked signals to maintain their native identity (Trainor and Krumlauf, 2000). However, if the neural crest cells were transplanted in association with their native mesoderm, this community of mesoderm and neural crest allowed for maintenance of neural crest cell identity in the ectopic location. Multiple researchers have now demonstrated the existence of a second heart field (or anterior heart field) in both chick and mouse embryos (Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001; Cai et al., 2003; Meilhac et al., 2004; Zaffran et al., 2004). The second heart field is a region of pharyngeal mesoderm which contributes myocardial cells to the outflow tract and right ventricle of the heart. In the mouse, the second heart field has also been shown to contribute cells to the venous or inflow region of the heart (Cai et al., 2003). The second heart field is located within the developing pharyngeal region and extends into the mesoderm of the branchial arches – the same territory through which the cardiac neural crest cells migrate on their way to the aortic arch arteries and outflow tract. Clearly, investigation into the second heart field and its interactions with the neural crest are critical to understanding morphogenesis of the outflow tract and great arteries. The interplay between cardiac neural crest cells and the surrounding mesoderm is clear from examination of ablation experiments in the chick embryo. Operated embryos display abnormal cardiac looping prior to the arrival of neural crest cells in the outflow tract of the heart (Yelbuz et al., 2002). This looping defect is a result of a shortening of the outflow tract (Yelbuz et al., 2002) that occurs when the myocardial precursors located within the second

424

heart field fail to migrate into the outflow tract (Waldo et al., 2005). Interestingly, the precursors in the second heart field destined to migrate to the outflow tract and contribute to vascular smooth muscle are unaffected. This indicates that the cardiac neural crest cells are required specifically for the migration of the myocardial precursors, and not merely for the migration of all mesoderm from the second heart field. A similar extrinsic influence on cardiac neural crest cell migration and differentiation is elicited by mesodermderived retinoic acid signaling (Thorogood et al., 1982; Pratt et al., 1987; Lee et al., 1997) (see Chapter 3.3). Targeted inactivation of the mouse retinaldehyde dehydrogenase 2 (Raldh2) gene that encodes the enzyme which converts alltrans retinal into retinoic acid causes embryonic lethality due to an absence of aorticopulmonary septation (Niederreither et al., 2001). Raldh2 mutant embryos exhibit impaired third, fourth and sixth branchial arch development and post-otic neural crest cells migrate aberrantly in a caudal direction. This disruption of cardiac neural crest cell migration leads to the absence of outflow tract septation. Raldh2 is expressed in the posterior-most pharyngeal mesoderm, highlighting the potential importance of the mesoderm during neural crest cell migration. In support of this idea, treating 8.5dpc mouse embryos with a pan-RAR antagonist in vitro and in vivo result in a complete absence of the third and fourth branchial arches (Wendling et al., 2001). Consequently, neural crest cells normally destined for the third and fourth arches migrate ectopically. The requirement for retinoic acid in cardiac neural crest cell patterning is stage-dependent. Only retinoic acid treatment of early head fold stage embryos induces ectopic migration of anterior hindbrain-derived neural crest cells (Lee et al., 1995). Second arch neural crest cells migrate ectopically into the first arch (Trainor and Krumlauf, 2000) leading to fusions between the first and second branchial arch streams. Collectively, these studies reveal that neural crest cells may not intrinsically follow specific migratory pathways, but rather are directed extrinsically by a complex of signaling interactions derived from the endoderm, mesoderm and ectoderm.

VII.  Pharyngeal endoderm and the cardiac crest The branchial arches are composed of an inner mesodermal core surrounded by a largely neural crest-derived mesenchyme, all of which is lined internally by a layer of pharyngeal endoderm and covered externally by a layer of ectoderm. The juxtaposition of these different cell types allows signaling from one arch layer to another. For example, when the cardiac neural crest population is ablated in chick embryos, an increase in levels of fibroblast growth factor 8 (FGF8) in the caudal pharyngeal ­ endoderm is

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

observed (Hutson et al., 2006). Subsequent to the increase in FGF8, the outflow tract myocardial precursor cells located in the second heart field fail to migrate into the outflow tract. However, if FGF signaling is blocked by application of an FGF8 antibody or chemical inhibition of FGF receptors in the ablated embryos, the migration of second heart field cells to the outflow tract in rescued. This indicates that cardiac neural crest cells modulate levels of FGF8 in the caudal pharyngeal endoderm, which in turn alters the behavior of an FGF8-sensitive subpopulation within the second heart field mesoderm. The T-box transcription factor, Tbx1, has been identified as a contributor to the etiology of 22q11 deletion syndrome (Yagi et al., 2003) (see Chapter 9.4). This syndrome is the most common deletion syndrome in humans and is characterized by a spectrum of defects including mild craniofacial defects, aortic arch defects and defects of the outflow tract (Ryan et al., 1997; Scambler, 2000). Although these defects are similar to cardiac neural crest ablation phenotypes, Tbx1 is expressed in the pharyngeal endoderm and the mesodermal core of the branchial arches, but is not expressed by neural crest cells (Chapman et al., 1996). Homozygous inactivation of Tbx1 in the mouse recapitulates the major defects of the 22q11 deletion syndrome (Jerome and Papaioannou, 2001; Lindsay et al., 2001; Liao et al., 2004), and targeted inactivation of Tbx1 within the pharyngeal endoderm results in identical defects (Arnold et al., 2006) indicating that the primary defect is failed outgrowth of the branchial arches. In Tbx1/ embryos, only the first branchial arch forms, but migration of the cranial and cardiac neural crest initiates normally. By 9.5 dpc abnormal distribution of neural crest cells into the pharyngeal region is observed, and shortly thereafter cells that would normally migrate into the second branchial arch mix into the first arch stream (Vitelli et al., 2002; Moraes et al., 2005). This aberrant migration is a secondary consequence of defective pharyngeal endoderm, underscoring the importance of extrinsic influences over neural crest patterning. The requirement for pharyngeal endoderm Tbx1 in patterning of the aortic arteries and outflow tract has been refined using an inducible Tbx1 mouse allele (Tbx1mcm) (Xu et al., 2005). In this model, the descendants of cells in which Tbx1 had been inactivated could be traced by reporter gene expression. When loss of Tbx1 was initiated at 7.5 dpc, labeled cells were located in the endothelium of the fourth aortic arch artery, the pharyngeal endoderm and ectoderm, and the outflow tract. Defects primarily affected the pharyngeal apparatus caudal to the second branchial arch. When Tbx1 was inactivated at 8.5 dpc, labeled Tbx1mcm descendants were found in similar locations, with the exception of the pharyngeal ectoderm, and the pharyngeal region was abnormal caudal to the third branchial arch. If inactivation of Tbx1 occurred at 9.5 dpc, the pharyngeal endoderm and outflow tract did not contain labeled descendants and were normally formed. Taken together,

Chapter  |  7.1  Role of Cardiac Neural Crest Cells in Morphogenesis of the Heart and Great Vessels

these data indicate there is a cranial-to-caudal induction of, and requirement for, Tbx1 in the development of the pharyngeal apparatus for normal aortic arch artery patterning and outflow tract development. In addition to Tbx1, the hedgehog (Hh) signaling pathway has been shown to play an important role in the proper development of the outflow tract (Washington Smoak et al., 2005). This has been attributed to defects in the cardiac neural crest and second heart field. However, using tissue-specific targeted deletion of sonic hedgehog (Shh), the importance of Hh signaling in the cardiac neural crest, second heart field and pharyngeal endoderm have been further refined (Goddeeris et al., 2007). Loss of Shh function in the pharyngeal endoderm resulted in apoptosis of splanchnic mesoderm cells, cardiac neural crest cells and cells of the pharyngeal endoderm. Ultimately, these mutants recapitulated the outflow tract, arch artery and intracardiac defects observed in full Shh/ mutants. Additionally, no outflow tract defects were observed when Shh was specifically ablated from the second heart field. Comparative analyses of these tissue-specific Shh mutant alleles with both neural crest-specific and second heart field-specific knockouts of the obligate Hh receptor, smoothened (Smo) (Goddeeris et al., 2007) revealed a cell-autonomous requirement for Smo in cardiac neural crest cells (for migration into the outflow tract and later septation events) and in the second heart field myocardial precursors (for later outflow tract septation events). Collectively, this series of conditional mutants revealed that Hh signaling is required within the pharyngeal endoderm, the cardiac neural crest and the second heart field for normal outflow tract development, and within the pharyngeal endoderm for normal aortic arch artery remodeling. However, endodermally-produced SHH ligand is required for all of these processes.

VIII.  Genetic regulation of cardiac neural crest cell patterning VIII.A.  Expansion of the Cardiac Neural Crest Cell Population Over 50 years ago, Splotch (Sp) mutant mice were identified due to neural crest-associated defects, including persistent truncus arteriosus (Auerbach, 1954; Franz, 1989). Sp and other mutant alleles (Sp2H, Spd, Spr, Sp4H) have since been identified as harboring mutations in the Pax3 gene (Epstein et al., 1991, 1993; Goulding et al., 1993; Vogan et al., 1993) that is expressed in the dorsal neural tube and in migrating neural crest cells. Sp/ embryos exhibit a reduction in the migrating streams of cardiac neural crest at 10.5 dpc (Conway et al., 1997). Initially, it was suggested that Pax3 was cell-autonomously required for migration of the cardiac neural crest (Conway et al., 1997; Li et al., 1999). However, subsequent analyses have

425

shown that Pax3-expressing neural crest precursors contribute cells to the cardiac outflow tract in both wild-type and Sp/ embryos (Epstein et al., 2000), but their contribution to the heart is decreased in mutant embryos. Focal cell labeling, orthotopic grafting, and reciprocal grafts between wild-type and Sp2H/Sp2H embryos further refined the status of cardiac neural crest cell migration in Splotch (Chan et al., 2004). The mutant cardiac neural crest cells were delayed in their emigration from the neural tube, followed normal pathways of migration, but were reduced in number compared to wild-type. Reciprocal cell transplantations between wild-type and mutant embryos indicated that defects in both neural crest cells and the migratory environment are collectively responsible for the Splotch mutant phenotype. Wild-type neural crest or environment transplants rescued the retardation of cardiac neural crest migration in mutant embryos. Extensive analyses in Splotch-mutant embryos demonstrate that prior to emigration from the neural tube, neural crest stem cells within the cardiac crest domain fail to undergo proliferative expansion (Conway et al., 2000). This primary defect reduces the number of premigratory cardiac crest, and so limits the neural crest contribution to the heart; however, it is unclear how this reduction in cell number retards migration through the cranial mesoderm. Interestingly, the decreased numbers of cardiac neural crest cells in Splotch embryos is associated with defective rotation of the outflow tract (Bajolle et al., 2006), similar to ablation studies in the chick, once again highlighting the intricate relationship between neural crest cells and mesoderm. Once the cardiac crest enters the branchial arches, proliferation of the cells is required for their further contribution to the outflow tract. Dishevelled 2 (Dvl2) is a member of the Wingless/Wnt signaling pathway that regulates the Wnt signal by modulating glycogen synthase kinase 3 (GSK-3), allowing nuclear translocation of -catenin and subsequent Wnt-target gene activation (Boutros and Mlodzik, 1999). Dvl2-deficient mice exhibit defective outflow tract septation and reduced levels of cardiac neural crest markers Pitx2 in the branchial arches and outflow tract, and PlexinA2 in the heart (Hamblet et al., 2002). Further examination of Pitx2/ mice identified a genetic interaction with Dvl2, as well as the direct regulation of Pitx2 by the Wnt/Dvl/-catenin pathway (Kioussi et al., 2002). In both Pitx2/ and Wnt1-Cre/-catenin mutants, cardiac neural crest cells migrating from the fourth and sixth branchial arches to the outflow tract exhibit decreased proliferation. Further, in vitro assays provided evidence for cell type-specific regulation of proliferation. In this model, Wnt/Dvl/-catenin signaling activates Pitx2, which stimulates expression of growth control genes, such as Cyclin D2, and proliferation (Kioussi et al., 2002). In the absence of Wnt signal in Dvl2/ embryos, this Pitx2 cell typespecific proliferation signal is not activated, and so there is decreased contribution of neural crest to the outflow tract.

426

VIII.B.  Migration of the Cardiac Crest In addition to a role in neural crest cell identity, cranial mesoderm has also been shown to influence neural crest cell migration. In vertebrates, cranial and cardiac neural crest cells migrate ventrolaterally in distinct segregated streams from the neural tube into the adjacent branchial arches. Clear neural crest-free zones exist immediately adjacent to odd numbered rhombomeres 3 and 5. These neural crest cell-free zones are not attributed to axial differences in neural crest cell production (Selleck and Bronner-Fraser, 1995) or localized domains of cell death (Trainor et al., 2002). Rhombomeres 3 and 5 generate neural crest cells, but instead of migrating laterally these populations deviate rostrally and caudally to join the adjacent ventrolateral streams. Complex interactions between the hindbrain and paraxial ectoderm and mesoderm tissues govern the migratory path of cranial neural crest cells. For example, the neuregulin receptor ErbB4 is expressed in rhombomeres 3 and 5, and in ErbB4-null mutant mice the first and second branchial arch streams of neural crest cells mix resulting in ganglion fusions. Reciprocal neuroepithelial transplantations between mutant and wild-type embryos revealed ErbB4 expression in the hindbrain functions noncell-autonomously to establish the neural crest free zones within the head (Farlie et al., 1999; Golding et al., 2000). Aberrant neural crest cell migration resulting in mixing between first and second branchial arch streams is also observed in Twist mutant mice (Soo et al., 2002) (Fig. 4A–B). Twist is a basic helix-loop-helix transcription factor initially expressed throughout the mesoderm, which later becomes active in neural crest cells colonizing the branchial arches (Wolf et al., 1991). Reciprocal cell transplantations between wild-type and Twist/ mutant embryos suggested that Twist functions in the paraxial mesoderm to instruct the migration trajectories of late-emigrating neural crest cells that are responsive to Twist signaling (Soo et al., 2002). In addition, Twist is required cell-autonomously in neural crest cells for their proper migration trajectories and regionalization once neural crest cells enter the branchial arch (Soo et al., 2002) (Fig. 4C–E). Recent analysis of the branchial arch and outflow tract mesenchyme of Twist/ embryos has identified abnormal clumps or nodules of neural crest-derived cells (Soo et al., 2002; Vincentz et al., 2008). Although these neural crest nodules do arrive in the branchial arches and outflow tract, they do so in reduced numbers compared to wild-type embryos (Ota et al., 2004; Vincentz et al., 2008), perhaps due to increased cell death in the mutant branchial arches (Fig. 4 F–G). When mutant embryos were analyzed at the time late-emigrating neural crest were exiting the neural tube, the neural crest cells were impaired or delayed ­during

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

delamination and migration from the neural tube, and actually formed clumps soon after exiting the neural tube (Vincentz et al., 2008). As Twist1 is not expressed in neural crest stem cells or within the neural tube at this time, the action of Twist on neural crest cell delamination and migration appears to be noncell-autonomous. The clumped cells abnormally express high levels of Hand1 and Hand2 and show limited expression of smooth muscle actin and periostin, indicating that the neural crest-derived cells are not maturing and differentiating appropriately, despite their arrival in the outflow tract (Vincentz et al., 2008). A growing body of work indicates that neural crest cell migration is directed by the same signaling molecules that govern axon pathfinding (Krull et al., 1997; Smith et al., 1997; Wang and Anderson, 1997; Gammill et al., 2006a,b). For example, Eph receptors and their ligands, the ephrins, are membrane-bound proteins that are differentially expressed between neural crest cell populations and the associated mesoderm (Davy and Soriano, 2005). EphrinB2 is expressed in the cranial and trunk mesoderm, and its ligand EphA4 is specifically expressed in neural crest cells. Ephrin binding to the Eph receptor results in autophosphorylation of the receptor and activation of downstream signaling events (Fig. 5A). There is also evidence that interaction with the Eph receptors triggers a reverse signal through the phosphorylation of the cytoplasmic tail of ephrin itself. Both forward and reverse signaling can regulate cell motility by increasing adhesion (attraction) or decreasing adhesion (repulsion) between cells (Davy and Soriano, 2005). Transgenic mice expressing a carboxy terminal truncated ephrin B2 protein have shown that correct targeting of Eph receptor expressing cranial neural crest cells to the branchial arches occurs by forward signaling through the receptors (Adams et al., 2001) (Fig. 5B). On receptor binding, repulsive signals guide correct migration of the neural crest cells. However, the role of ephrin/Eph signaling in the proper migration of neural crest cells into the branchial arches is more complex, as recent work has shown a requirement for the cell-autonomous action of ephrinB1 acting through reverse signaling (Davy et al., 2004) (Fig. 5C). Additionally, in more caudal regions it is believed that ephrins in the developing somites act noncellautonomously on neural crest cells, expressing Eph receptors (Krull et al., 1997; Wang and Anderson, 1997). In addition to ephrin/Eph signaling, secreted class 3 semaphorin signaling molecules and their receptors have been implicated in neural crest cell migration (Brown et al., 2001; Feiner et al., 2001; Yu and Moens, 2005; Gammill et al., 2006a,b; Sato et al., 2006). Originally, semaphorins were identified as inhibitory axon guidance cues, but some semaphorins have also been shown to be attractive (Kolodkin et al., 1993; Mark et al., 1997; Raper, 2000). Recently, semaphorin (sema) 3F and its receptor neuropilin 2 (npn2) have been localized to the neural crest-free zones

Chapter  |  7.1  Role of Cardiac Neural Crest Cells in Morphogenesis of the Heart and Great Vessels

(A)

Sox10 gV

(C)

ba1 ba2

gX

ba2

*

(B)

gVll/Vll ov glX oc

ba1

* +/+

+/+

427

o

−/−

+/+ lat

−/−

+/+

+/+

−/−

(D)

(E)

(F)

+/+ (G)

−/−

ba1 ba2 sm

Figure 4  Complex interactions between mesoderm and neural crest cells govern neural crest cell migration and survival. (A, B) Whole mount in situ hybridization to detect the neural crest marker Sox10 in: (A) wild-type; and (B) Twist/ 9.5dpc mouse embryos. The Sox10-positive neural crest cells normally migrate in wellseparated streams, while in the mutant mixing (arrowhead in (B)) of the streams contributing to branchial arches 1 and 2 (ba1, ba2) is observed. (C–E) Transplantation of neural crest cells into host embryos was used to determine the requirement for Twist within the neural crest and within the mesoderm. (C) When wild-type (/) donor neural crest cells (blue label) were transplanted into / host embryos, the donor cells migrated normally and dispersed appropriately within the branchial arch (tissue section). (D) When donor cells were obtained from Twist mutants (/) and transplanted into wild-type hosts, labeled cells remained concentrated within the core of the branchial arch, indicating a cell-autonomous requirement for Twist within the neural crest population. (E) There is also a requirement for Twist function in the surrounding mesoderm, as transplantation of / neural crest cells into mutant host embryos did not restore normal migration patterns to the neural crest cells. (F–G) Loss of Twist leads to increased cell death in the developing branchial arch mesenchyme (blue stain, arrows in (G)) indicating Twist is important for survival of neural crest and cranial mesoderm cells. (A, B, F, G) Reprinted, with permission, from Ota, M. S. et al. (2004) Developmental Dynamics 230, 216–228. (C–E) Reprinted from Soo, K. et al. (2002) Developmental Biology 247, 251–270. © Elsevier Science (USA) 2002, with permission from Elsevier.

428

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

Forward signaling

Kinase

P PP EphrinB2∆C/∆C

Crabp1

EphrinB2WT/WT

Forward signaling

Forward signaling

YYKV

Eph Reverse signaling signaling mutant

SAM

Reverse Ephrin signaling signaling mutant

(B)

SAM

P P

Reverse signaling

Forward signaling

P P

Eph K/O

Kinase

P PP

Reverse signaling

Forward signaling

YYKV

Ephrin K/O

Kinase

P PP

Reverse signaling

YYKV

Wild type

Eph receptors P P

Ephrin-B

SAM

(A)

EphrinB2KO/KO

r3 r4 r5

EphrinB1 WT

Null mutant

(C) ba1 ba2 ba3

ba1 ba2 ba3

ba4

ba4

NCC mutant ba1 ba2 ba3 ba4

Figure 5  Ephrin signaling and neural crest cell migration. (A) Schematic representation of Eph receptors, transmembrane ephrins and how alteration of the proteins affects forward or reverse signaling events. PDZ-containing proteins interact with the YYKV domain located in both Ephrins and Eph receptors. Disruption of either forward or reverse signaling can alter cell migration and adhesion, thereby disrupting neural crest cell migration. (B) Comparison of expression of the neural crest marker Crabp1 in wild-type, ephrinB2C/C (ephrin signaling-mutant) and ephrinB2KO/KO-(null) embryos. Truncation of the C-terminus of ephrinB2 restores neural crest cell migration, indicating that reverse signaling is not required. The disruption of migration in the null-mutant is thus attributed to a loss of forward signaling. (C) Fatemapping of neural crest cells in wild-type, ephrinB1-null and ephrinB1 neural crest-specific mutant mice revealed alterations in the migration of cardiac neural crest cell populations (red arrows) in both mutant strains. The cell-autonomous requirement for ephrinB1 in the neural crest was shown to require reverse signaling when a third mouse line containing ephrinB1 with an altered PDZ domain (ephrin signaling mutant) was analyzed (ba: branchial arches; r: rhombomere). (A) Reprinted with permission from Davy, A., and Soriano, P. (2005) Developmental Dynamics 232, 1–10. (B) Reprinted from Adams, R. H. et al. (2001) Cell 104, 57–69. © Cell Press 2001, with permission from Elsevier. (C) Reprinted with permission from Davy, A. et al. (2004) Genes & Development 18, 572–583.

of the cranial mesenchyme and migrating cranial neural crest cells, respectively (Gammill et al., 2006a). Targeted deletion of either sema3F or npn2 resulted in aberrant

migration of neural crest cells entering the branchial arches, which indicates that intact sema3F/npn2 signaling is required for proper cranial neural crest migration.

Chapter  |  7.1  Role of Cardiac Neural Crest Cells in Morphogenesis of the Heart and Great Vessels

Although a role for npn2/sema3F signaling in cardiac neural crest migration has not been unequivocally demonstrated, expression analyses of Sema3C and PlexinA2, which is a co-receptor of npn2, suggest that these molecules may play a similar role in cardiac neural crest migration (Brown et al., 2001; Feiner et al., 2001). Presumptive cardiac neural crest cells are positive for PlexinA2 (Brown et al., 2001), and in the midgestation mouse heart, mesenchymal cells surrounding regions of cardiac neural crest colonization express Sema3C (Feiner et al., 2001). Null mutation of Sema3C results in misexpression of PlexinA2 and aortic arch defects (Feiner et al., 2001), similar to ablation of the cardiac neural crest in chick. Interpretation of the precise role of Sema/Npn signaling is complicated by evidence that Npn receptors also bind to the VEGF-165 isoform, the disruption of which leads to malformations of the outflow tract (Stalmans et al., 2003). Hence, disruption of Sema/Npn signaling may shift the balance of VEGF/ Npn signaling, leading to defective migration. Future studies tracing neural crest migration in these mouse models will clarify the role of Sema/npn signaling in the cardiac neural crest. Neural crest cells have been observed to migrate in groups or sheets, indicating that cell–cell interactions are important in their organized migration (Nakagawa and Takeichi, 1998). As previously discussed, the Connexin43lacZ (Cx43lacZ) reporter mouse line provided the first molecular evidence of the mammalian cardiac neural crest (Lo et al., 1997). Cx43 is a member of the multigene connexin family of proteins that comprise gap junction membrane channels that allow the passage of ions and small molecules between cells. Cx43 knockout mice exhibit conotruncal heart malformations and obstruction of the pulmonary outflow tract which result in death soon after birth (Reaume et al., 1995). Similar defects were also observed when Cx43 was overexpressed in subpopulations of neural crest cells (Ewart et al., 1997; Huang et al., 1998). These defects do not resemble the defects generally associated with surgical ablation of the cardiac neural crest, and therefore the function of Cx43 in cardiac neural crest is not to simply maintain the cell population. Neural crest-specific deletion of Cx43 has now shown that Cx43 expression within the cardiac neural crest is not critical for normal heart morphogenesis (Kretz et al., 2006; Liu et al., 2006). Rather, studies using explants of postotic hindbrain neural tube to study the migration distance, direction and speed of cardiac neural crest cells (Huang et al., 1998; Xu et al., 2001, 2006) have suggested that Cx43 gap junctions mediate the cross-talk between signaling pathways that regulate directional migration. Overexpression of Cx43 results in increased speed and directionality of migration of cardiac neural crest, while loss of Cx43 or expression of dominant negative constructs leads to reduced directionality and speed (Huang et al., 1998; Xu et al., 2006). Perturbation of Cx43 expression also altered the

429

association of 1 integrin with vinculin focal adhesions in cardiac neural crest cells. These focal contacts, which are reduced in Cx43/ cardiac neural crest cells, provide a connection to the actin cytoskeleton which is of fundamental importance for cell motility (Xu et al., 2006). Interestingly, Cx43 and N-cadherin co-localize along cell processes of wild-type migrating neural crest cells. Inhibition or overexpression of N-cadherin has been shown to disrupt the normal emigration of neural crest cells from the neural tube (Bronner-Fraser et al., 1992; Nakagawa and Takeichi, 1998). In hindbrain neural tube explants, the migration paths of N-cadherin-deficient neural crest cells are tortuous and meandering compared to wild-type controls, indicating that cadherin-mediated cell–cell adhesion mediates motility (Xu et al., 2001). Both Cx43 and N-cadherin molecules extensively co-localize at the cell surface with p120 catenin, which modulates cell motility through Rho GTPases. In the absence of either Cx43 or N-cadherin, p120 catenin was primarily found localized to the nucleus of cells. Thus the coordinated interaction of Cx43, N-cadherin and p120 catenin serve to facilitate interaction with and reorganization of the actin ­ cytoskeleton needed for cell movement.

IX.  The post-migratory cardiac crest The cardiac neural crest cells that remain associated with the developing aortic arch arteries play a little-understood role in the proper regression or persistence of these blood vessels (Fig. 3). Ultimately, these cells differentiate into the smooth muscle layer of the mature vessels. The subset of cardiac neural crest cells that further migrate into the outflow tract are necessary for proper division of the aorta and pulmonary artery. Although relatively little is known concerning the molecular and genetic control of these remodeling and septation events, several mutant mouse strains have provided important insights into these processes.

IX.A.  Aortic Arch Remodeling Genetic disruption of platelet-derived growth factor receptor  (PDGFR) results in a wide spectrum of defects including kidney, cardiac, facial and axial skeleton defects (Tallquist and Soriano, 2003). In chimeras of PDGFRnull embryonic stem cells and wild-type blastocysts the mutant cells were excluded from several regions of the embryo, including the branchial arches. Conditional knockout of PDGFR in neural crest (PDGFR-NCC) showed normal induction and migration of cardiac neural crest cells, but the third, fourth, and sixth arch arteries were either dilated or reduced in diameter compared with controls. Additionally, the mutant cells differentiated ­normally into smooth muscle of the arteries. The aortic arch defects

430

in PDGFR-NCC occurred with incomplete penetration, possibly due to compensation by PDGFR which is also expressed in cardiac neural crest cells. To investigate this possibility, both PDGFR and PDGFR were conditionally ablated in neural crest cells (PDGFR/-NCC) (Richarte et al., 2007). In these mutants, aortic arch defects occurred with complete penetrance, and migration of the cardiac neural crest into the proximal outflow tract and aortic sac was reduced. These studies reveal that signaling through PDGFRs is required cell-autonomously within the cardiac neural crest for normal aortic arch development. Endothelin-1/Endothelin-A Receptor (ET-1/ETA)­mediated signaling has been implicated in aortic arch remodeling. In cranial regions, neural crest cells express ETA as they exit the hindbrain into cranial mesenchyme. The ET-1 ligand is found in the pharyngeal pouch endoderm, pharyngeal ectoderm and the mesodermal core of the branchial arches (Clouthier et al., 2003). Neural crest cell migration appears normal; however, mice deficient in either ETA or the endothelin-converting enzyme (ECE-1) exhibit interruption of the aortic arch and absence of the right subclavian artery (Clouthier et al., 1998; Yanagisawa et al., 1998b). This indicates that the role of ET-1/ETA signaling plays an essential role in the post-migratory neural crest cells. Introduction of the smooth muscle 22lacZ reporter transgene into ET-1- or ETA-mutant backgrounds enabled visualization of aortic arch artery development (Yanagisawa et al., 1998a). Initial formation of the arch arteries is normal, but arch arteries 3, 4 and 6 were most severely affected with respect to abnormal persistence or regression. Chimera analysis has shown that ETA/ cells are excluded from the most ventral aspects of the branchial arches, the walls of the developing arch arteries and outflow tract, indicating a cell-autonomous requirement for ETA signaling in these tissues (Clouthier et al., 2003). Although ET-1/ETA/ ECE-1 are not expressed with any left–right asymmetry in the developing branchial arches, intact ET-1/ETA signaling is essential to correct asymmetric maintenance of mesenchyme surrounding specific arch arteries (Yanagisawa et al., 1998a). Interestingly, if pre-proendothelin 1 is constitutively expressed in the cardiac neural crest, the diameter of the great vessels is increased, due to an increase in the size of the adventitial layer (Ballard and Mikawa, 2002). This could indicate a role for ET-1/ETA signaling maintaining mesenchymal support of the great arteries and proper hemodynamic flow through the vessels. Hemodynamic flow has been implicated as an important player in remodeling of at least the sixth aortic arch artery (Yashiro et al., 2007), as revealed by analysis of Pitx2-mutant mouse embryos. Nodal signaling in the lateral plate mesoderm of the early embryo establishes the left–right asymmetry of the internal organs of vertebrates (Shiratori and Hamada, 2006), and Pitx2 is a transcription factor induced by the nodal signal. Pitx2c is expressed

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

asymmetrically in the branchial arch mesoderm and second heart field, and Pitx2c-mutants display aortic arch remodeling defects (Liu et al., 2002). However, Pitx2 expressing cells are not detected around the fourth and sixth aortic arch arteries that regress or persist in an asymmetric fashion (Yashiro et al., 2007). Mutant mice lacking the asymmetric enhancer of Pitx2 (Pitx2ASE/ASE) express Pitx2 bilaterally and have a randomization of the sidedness of the patent sixth aortic arch artery, as well as the patent dorsal aorta (Yashiro et al., 2007). In addition, it was noted that at 11.5 dpc, immediately prior to aortic arch remodeling, the outflow tract morphology in these mutants was abnormal. Rather than taking on a spiral configuration, the outflow tract remained straight. At 12.0 dpc, the outflow tract should rotate again to shift the right sixth aortic arch artery to the left. This also fails to occur when Pitx2 is expressed bilaterally. Normally, this rotation of the outflow tract creates a right sixth aortic arch artery that is longer and thinner than its left-side counterpart. This would potentially reduce the hemodynamic flow to the right sixth artery. To test the idea that asymmetric flow could determine which artery persists and which regresses, the left sixth aortic arch artery was ligated to decrease the flow abnormally in wild-type embryos (Yashiro et al., 2007). After culture of the manipulated embryos the right sixth artery persisted, and the left sixth artery with reduced flow had regressed. The ligation also induced cell death surrounding the ligated side, and the right side of the dorsal aorta was abnormally enlarged over the left side. The change in left–right persistence of the sixth aortic arch artery was also associated with alterations in the expression of PDGFA and VEGFR2. Both molecules are normally downregulated on the regressing right side and maintained on the left. This pattern was reversed in both ligated wild-type and Pitx2ASE/ASE mutant embryos. Taken together, these results suggest a model of aortic arch remodeling (Yashiro et al., 2007) in which asymmetric Pitx2 expression in the second heart field is required for proper morphogenesis of the outflow tract. In turn, this rearrangement of the outflow tract leads to differential blood flow between the left and right sides of the aortic arch system. Differential blood flow then stimulates VEGFR2 and PDGFA expression around the left sixth artery, while they are downregulated on the right side. In the absence of VEGFR2 and PDGFA, the right sixth arch artery regresses. Exactly how Pitx2 regulates spiral rotation of the outflow tract, and why VEGFR2 and PDGFA signal specifically to the sixth arch artery remains unknown.

IX.B.  Outflow Tract Septation Normally, outflow tract endocardial cushions are derived from: (1) endocardial endothelial cells that undergo epithelial-to-mesenchymal transformation (EMT); and

Chapter  |  7.1  Role of Cardiac Neural Crest Cells in Morphogenesis of the Heart and Great Vessels

(2) cardiac neural crest cells. Once epithelial-to-mesenchymal transformation has occurred, both types of cells proliferate and invade the cardiac jelly, a thick layer of acellular extracelluar matrix, to form the cushions. Sox9 is a member of a large family of transcription factors and is expressed in the mesenchyme of endocardial cushions (Akiyama et al., 2004). Sox9 plays important roles in epithelial-to-mesenchymal transformation (Akiyama et al., 2004), and is a direct regulator of type II collagen (Bell et al., 1997), which is transiently expressed from 10.5–14.5 dpc in regions of the developing heart undergoing epithelial-to-mesenchymal transformation (Rahkonen et al., 2003). Targeted deletion of Sox9 in neural crest cells indicated that Sox9 is required in both the endothelial and neural crest cells contributing to cushion formation (Akiyama et al., 2004). Although epithelial-to-mesenchymal transformation was initiated, proliferation of the mesenchymal cells invading the cardiac jelly was diminished due to the inactivation of the epidermal growth factor receptor tyrosine kinase, ErbB3. As the cardiac neural crest cells migrating to the outflow tract are already mesenchymal in nature, it is unclear if abnormal septation in Sox9/ cardiac neural crest is secondary to deficiencies in the invading endothelial cells with which the crest cells might interact, or if defective type II collagen regulation might play a role. BMP growth factor signaling, which has been implicated in induction of neural crest cell formation, is also important for outflow tract septation. BMP ligands bind to a two-part receptor complex composed of type I and type II receptors (Kishigami and Mishina, 2005). In mouse there are three known type I receptors (BMPRIA, BMPRIB and ALK2) and two type II receptors (ActRII and BMPR2). Based on their spatio–temporal expression during embryonic development and results of targeted deletion of the receptors (Matzuk et al., 1995; Gu et al., 1999; Beppu et al., 2000; Panchision et al., 2001), BMPRIA, ALK2 and BMPR2 are the most likely to play a role in proper migration, patterning, or differentiation of neural crest cells. Once bound by ligand, phosphorylation of the receptorregulated Smad (R-Smads) transcription factors occurs. Phosphorylated R-Smads then form a complex with the common Smad, Smad4, which facilitates traslocation to the nucleus and activation of BMP target genes (Massague et al., 2005). Neural crest-specific deletion of BMPRIA revealed no deficiencies in induction or migration of the neural crest; however, the outflow tracts of mutant embryos were shortened, showed decreased contribution of cardiac neural crest cells and had underdeveloped endocardial cushions (Stottmann et al., 2004). Similar outflow tract defects were observed in a hypomorphic BMPR2-mutant mouse in which signaling through this type II receptor is diminished (Delot et al., 2003). The defective outflow tract cushions appeared to initiate epithelial-to-mesenchymal

431

t­ransformation and contained evidence of neural crest cell derivatives, but the cushions failed to expand, leading to septation failure. Similarly, when Smad4 is specifically deleted in neural crest cells, cardiac neural crest cells undergo normal migration to the distal outflow tract, but do not complete migration into the proximal outflow tract (Jia et al., 2007). Ultimately, the neural crest cells of the outflow tract exhibit decreased levels of BMP targets Msx1 and Msx2 and undergo apoptosis. This is not surprising, as one functional copy of either Msx1 or Msx2 is required to maintain appropriate levels of post-migratory neural crest cells in the outflow tract (Chen et al., 2007). Additionally, in the neural crest-specific Smad4 mutants, decreased expression of both Sox9 and hyaluronan synthase 2 (Has2) were reported (Jia et al., 2007). Both Sox9 and Has2 are required in the developing cardiac cushions for proper EMT. Loss of expression was specific to the outflow tract cushions, whereas the atrioventricular cushions were normal, indicating that the disruption of epithelial-to-­mesenchymal transformation was associated with the loss of BMP sig­ naling in the cardiac neural crest. Finally, neural crest-specific ablation of ALK2 not only results in abnormal aortic arch remodeling, but also in defective endocardial cushion formation (Kaartinen et al., 2004). Again, mutant neural crest derivatives arrived in the pharyngeal arches and outflow tract, but the cushions failed to expand properly. Interestingly, expression of PlexinA2, which is involved in Sema/npn signaling, and Msx1, a major effector of BMP signaling, were diminished in the mutant outflow tract. Taken together, these studies implicate BMP signaling in the cardiac neural crest as critical to normal development of the outflow tract, malformations in which are a major cause of neonatal mortality in humans (Creazzo et al., 1998).

IX.C.  Differentiation of Cardiac Neural Crest Cardiac neural crest cells within the aortic arch arteries and in the outflow tract ultimately differentiate into smooth muscle. Despite a number of models exhibiting defects in some aspect of cardiac neural crest development, there are few published reports that have identified a defect in differentiation. Myocardin-related transcription factor B (MRTF-B) was detected in rhombomeres 3 and 5 and the neural crest derived mesenchyme around the aortic arch arteries, and MRTF-B gene trap mice produced MRTF-B/ embryos that died neonatally due to cardiac outflow tract defects (Li et al., 2005). PlexinA2 was detected in the aortic arch arteries and in the defective endocardial cushions, indicating that cardiac neural crest cells migrated properly and contributed to the heart. However, smooth muscle alpha actin (Acta2a/SMA) was diminished or absent from aortic arch arteries 3, 4 and 6,

432

and absent from the outflow tract cushions. As Acta2a was readily detected in other regions of the heart, this indicated that the cardiac crest failed to differentiate into smooth muscle. When MRTF-B activity was specifically restored in the neural crest cell population, all aortic arch artery defects, outflow tract defects and neonatal lethality were rescued. Although little is known about smooth muscle cell differentiation from cardiac neural crest cells, it is known that SMCs that arise from lateral plate mesoderm are regulated by signaling from the serum response factor (Owens et al., 2004). Serum response factor activates smooth muscle cell differentiation by interaction with the transcriptional coactivator myocardin (Wang et al., 2001; Du et al., 2003; Wang and Olson, 2004), and based on the MRTF-B/ phenotype serum response factor-mediated signaling is likely involved in smooth muscle cell differentiation from the cardiac neural crest. The Notch signaling pathway has also been shown to play a role in the differentiation of cardiac neural crest cells into smooth muscle (High et al., 2007). When a dominant negative Notch inhibitor (DN-MAML) was activated specifically within neural crest cells, a wide range of outflow tract, cardiac and aortic arch defects were observed. Neural crest cell migration, survival and contribution to the branchial arches and outflow tract were normal in these embryos. However, a decrease in the Notch target genes (HRT1, HRT2, and HRT3) and smooth muscle markers (Acta2a/SMA and SM22) was detected surrounding the aortic arch arteries. Additionally, when mutant neural tubes were explanted or when Notch signaling was inhibited in wild-type explants, neural crest cells failed to differentiate into smooth muscle cells. Thus, Notch signaling plays a cell-autonomous role in post-migratory differentiation of the cardiac neural crest into smooth muscle.

X.  Persistence of cardiac neural crest cells in the heart The neural crest is a multipotent population of cells that gives rise to a large number of differentiated cell types including bone, neurons, glia, melanocytes and those of connective, endocrine and adipose tissue. Such diversity suggests that the neural crest population might represent a stem cell population within the developing embryo. There is still debate among researchers as to whether the neural crest cells are truly pluripotent, or whether they are faterestricted shortly after their emergence from the neural tube. Pluripotency of pre-migratory neural crest cells has been demonstrated in vivo through the labeling and transplantation of single cells into avian embryos. The labeled cells generated descendants that were of multiple distinct cell types (i.e., melanocytes, glia and neurons) (BronnerFraser and Fraser, 1988, 1989). Further studies of the cranial neural crest revealed that once migratory the majority

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

of neural crest cells are multipotent, biotent and unipotent progenitor cells with limited differentiation capabilities (Baroffio et al., 1991). Only a small percentage of cells within the migrating neural crest are believed to be truly pluripotent and have been deemed the neural crest stem cell population. This is similar to that which has been determined for the cardiac neural crest in the mouse embryo. Targeted disruption of the tyrosine kinase receptor type 3 (TrkC), which is involved in signaling by the neurotrophic factor NT-3, surprisingly resulted in defects of the outflow tract associated with cardiac neural crest deficiencies (Tessarollo et al., 1997). Migrating neural crest cells in the domain of the cardiac neural crest express TrkC and appear to be activated by paracrine-NT-3 from endothelial cells scattered throughout the cranial mesoderm (Youn et al., 2003). Explants of the neural tube from the midotic region to somite 3 were used in colony assays to identify three types of cell colonies: (1) cardiac neural crest stem cell (NCS-SC) colonies capable of self-renewal and production of all neural crest derivative cell types; (2) fate-restricted colonies (CNC-RC) which do not contain neurons or melanocytes; and (3) smooth muscle lineage committed colonies (CNC-smC) which only contain smooth muscle cells (Youn et al., 2003) (Fig. 6A–B). In the absence of TrkC, the percentage of CNC-SC in cultures decreased by greater than 50% compared to wild-type cultures, whereas an equivalent increase in CNC-RC was observed. It is believed that NT-3 signaling through TrkC retains CNCSC in the stem cell compartment and loss of this signal results in their premature fate restriction. Precocious commitment of CNC-SC would decrease the stem cell population giving rise to the cardiac crest; however, the effect of this on cardiac neural crest migration and differentiation are not known. Clonal analysis of cells migrating in the cardiac migration pathway revealed cardiac neural crest stem cells that generated many cell types but comprised only a small percentage of the clone colonies generated (Youn et al., 2003). These cells could be serially cloned and so were capable of self-renewal; however, it was unclear from these studies how long the cardiac neural crest stem cells persisted in the embryonic or adult heart. Recently, cardiac neural crest stem cells were isolated from both the postnatal and adult mouse heart (Tomita et al., 2005). Primary cultures of cardiomyocytes and nonmyocytes were isolated and used in a cardiosphere formation assay. Similar to the generation of neurospheres from neural stem cells, this resulted in free floating spheres that were characterized with respect to their gene expression and differentiation capabilities both in vitro and in vivo. The cardiospheres expressed classic stem cell markers Nestin and Musashi-1 at their isolation and prior to the onset of differentiation. When permitted to differentiate, the cardiospheres differentiated into multiple cell types of the neural crest lineage. Additionally,

Chapter  |  7.1  Role of Cardiac Neural Crest Cells in Morphogenesis of the Heart and Great Vessels

433

(A)

(B)

(C)

(D)

(E)

Figure 6  Cardiac neural crest stem cells and their persistence in the adult heart. (A) Bright field image of colony type formed by cardiac neural crest stem cells. Large flattened cells (red arrow), small stellate cells (white arrow) and neurites (green arrow) are visibile. (B) Neural crest cell types identified within the stem cell colonies: (1) mature and immature smooth muscle cells (SMA, red), Schwann cells (S100, green); (2) neurons (-III tubulin, green); (3) pigment cells containing melanin; (4) chondrocytes (Coll type II, green); (5) progenitor cells (4E9R, green). (C–E) A transgenic mouse line expressing GFP in the neural crest was used to trace neural crest and their descendants in the heart. (C) GFP-positive cells were detected in the 17.5 dpc mouse heart. (D) These GFP-positive cells co-expressed the stem cell marker Nestin indicating that they retained multipotency. (E) Cells were isolated from 10-week-old mouse hearts and used to generate cardiospheres. These cardiospheres were GFP-positive, indicating their neural crest origin. (A, B) Reprinted from Youn, Y. H. et al. (2003) Molecular and Cellular Neuroscience 24, 160–170. © Elsevier Science (USA) 2003, with permission from Elsevier. (C–E) Reproduced from Journal of Cell Biology (2005) 170, 1135–1146. © 2005 Rockefeller University Press.

434

labeled cardiospheres were transplanted into two locations in chick embryos to test their ability to contribute to neural crest populations. When transplanted into either the endogenous cardiac neural crest region or into the migration pathway normally taken by melanocytes, labeled cells migrated along with the endogenous neural crest in the region. Strikingly, labeled cells were detected in the outflow tract of the heart, within the dorsal root ganglia and peripheral nerves. As a final assay for these cardiac neural crest stem cells within the adult heart, a transgenic mouse line expressing green fluorescent protein (GFP) specifically within the neural crest population was used to track the distribution of GFP-positive cells in the heart (Tomita et al., 2005). Although many GFP-positive cells in the adult heart had differentiated and co-expressed cardiac specific markers, a small population of GFP-positive cells retained Nestin expression in the absence of cardiac specific markers (Fig. 6C–D). Finally, hearts from 10-week-old GFP reporter mice were used to isolate cardiospheres that were clearly GFP-positive, verifying the neural crest origin of the cardiospheres (Fig. 6E). Taken together, these studies indicate the presence of pluripotent cardiac neural crest stem cells that are capable of self-renewal and generation of a wide range of neural crest cells types. It appears that in the adult these stem cells are a small, dormant population, but which if called upon can differentiate into various cell types. Given the potential for clinical treatment of various types of heart disease and congenital heart malformations, further study of these important cells is critical to identify the molecular control of stem cell maintenance, self-renewal and the cues that trigger their differentiation.

XI.  Summary Cardiac conotruncal malformations affect four in 10,000 live births, while valve and septal defects afflict about four in every 1,000 newborn children. The congenital cardiovascular anomalies encompass failure of outflow tract septation (persistent truncus arteriosis) and abnormal patterning of the aortic arch arteries (Gittenberger-de Groot et al., 2005). Such conotruncal, valve and septal defects comprise pathophysiological components of DiGeorge, tetralogy of Fallot and velocardiofacial syndromes, each of which is caused at least in part by aberrant cardiac neural crest cell patterning. Neural crest cells are generated transiently in a discrete location within the embryo, but through their directed migration over relatively long distances they contribute extensively to a wide array of embryonic and adult tissues. Neural crest cells are generated at all levels of the embryonic anteroposterior axis and exhibit unique axial properties with respect to their migration paths and the cell types into which they differentiate. At the level of the posterior

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

hindbrain, neural crest cells contribute to the glands of the neck, the tunica media of the aortic arch arteries and great vessels, and the outflow tract of the heart. These cells function in the remodeling of the aortic arch arteries into their mature vessel derivatives, and in the proper septation of the aorta and pulmonary artery. Recent advances in transgenic and knockout mouse technology have verified the existence of the mammalian cardiac neural crest, identified a series of molecules and signaling pathways involved in the formation, migration and function of the cardiac neural crest, and isolated cardiac neural crest stem cells from both neonatal and adult hearts. Despite all of these advances, many questions remain regarding the signal(s) involved in the induction of neural crest cell formation in the mouse, the signals guiding migration of the neural crest and the specificity with which this may occur. For example, do the cells of the cardiac region respond to different signals than those in the trunk region? Is specificity conferred by differences in timing or combinations of signals? How are the cardiac neural crest cells that migrate from the pharyngeal arches to the heart segregated from those that remain and contribute to glands of the throat? In human 22q11 deletion syndrome, the primary cardiac defect is a partial or complete lack of division of the single great vessel of the embryonic heart (the truncus arteriosus) into the pulmonary artery and the aorta (Scambler, 2000). A similar spectrum of branchial arch defects have been reported in mutant and teratogenic mouse models (Wilson and Warkany, 1949; Auerbach, 1954; Franz, 1989; Kurihara et al., 1995; Manley and Capecchi, 1995; Lee et al., 1997). It is interesting to note that in humans outflow tract and aortic arch remodeling defects, such as those seen in 22q11 deletion syndrome and in persistant truncus arteriosus, occur in the absence of defects in other heart structures. These are some of the most common defects in humans, and based on studies in mouse and chick, they arise from defects within the cardiac neural crest. Our understanding of the etiology of these disorders will increase with a deeper appreciation of the formation and patterning of the cardiac neural crest. With the identification of cardiac neural crest stem cells, advances in isolation, transplantation and understanding of their differentiation will provide for their use in repair of injured tissue and other stem cell therapies.

References Adams, R.H., Diella, F., Hennig, S., Helmbacher, F., Deutsch, U., Klein, R., 2001. The cytoplasmic domain of the ligand ephrinB2 is required for vascular morphogenesis but not cranial neural crest migration. Cell 104, 57–69. Akitaya, T., Bronner-Fraser, M., 1992. Expression of cell adhesion molecules during initiation and cessation of neural crest cell migration. Dev. Dyn. 194, 12–20.

Chapter  |  7.1  Role of Cardiac Neural Crest Cells in Morphogenesis of the Heart and Great Vessels

Akiyama, H., Chaboissier, M.C., Behringer, R.R., Rowitch, D.H., Schedl, A., Epstein, J.A., de Crombrugghe, B., 2004. Essential role of Sox9 in the pathway that controls formation of cardiac valves and septa. Proc. Natl. Acad. Sci. USA 101, 6502–6507. Andres, G., 1946. Uber Induction und Entwicklung von Kopforganen aus Unkenektoderm im Molch (Epidermis, Plakoden and Derivate der Nemalleiste). Rev. Suisse. Zool. 53, 502–510. Andres, G., 1949. Untersuchungen an chimaren von Triton und Bombinator. Genetics 24, 387–534. Arnold, J.S., Werling, U., Braunstein, E.M., Liao, J., Nowotschin, S., Edelmann, W., Hebert, J.M., Morrow, B.E., 2006. Inactivation of Tbx1 in the pharyngeal endoderm results in 22q11DS malformations. Development 133, 977–987. Auerbach, R., 1954. Analysis of the developmental effects of a lethal mutation in the house mouse. J. Exp. Zool. 127, 305–329. Bajolle, F., Zaffran, S., Kelly, R.G., Hadchouel, J., Bonnet, D., Brown, N.A., Buckingham, M.E., 2006. Rotation of the myocardial wall of the outflow tract is implicated in the normal positioning of the great arteries. Circ. Res. 98, 421–428. Ballard, V.L., Mikawa, T., 2002. Constitutive expression of preproendothelin in the cardiac neural crest selectively promotes expansion of the adventitia of the great vessels in vivo. Dev. Biol. 251, 167–177. Baroffio, A., Dupin, E., Le Douarin, N., 1991. Common precursors for neural and mesenchymal derivatives in the cephalic neural crest. Development 112, 301–305. Basch, M.L., Bronner-Fraser, M., Garcia-Castro, M.I., 2006. Specification of the neural crest occurs during gastrulation and requires Pax7. Nature 441, 218–222. Bashir, N.S., Armstrong, J.B., 1999. The cardiac neural crest in Ambystoma mexicanum. Int. J. Dev. Biol. 43, 269–274. Bell, D.M., Leung, K.K., Wheatley, S.C., Ng, L.J., Zhou, S., Ling, K.W., Sham, M.H., Koopman, P., Tam, P.P., Cheah, K.S., 1997. SOX9 directly regulates the type-II collagen gene. Nat. Genet. 16, 174–178. Beppu, H., Kawabata, M., Hamamoto, T., Chytil, A., Minowa, O., Noda, T., Miyazono, K., 2000. BMP type II receptor is required for gastrula­ tion and early development of mouse embryos. Dev. Biol. 221, 249–258. Besson, W.T., Kirby, M.L., Van Mierop, L.H., Teabeaut, J.R., 1986. Effects of the size of lesions of the cardiac neural crest at various embryonic ages of incidence and type of cardiac defects. Circulation 73, 360–364. Bockman, D.E., Redmond, M.E., Waldo, K., Davis, H., Kirby, M.L., 1987. Effect of neural crest ablation on development of the heart and arch arteries in the chick. Am. J. Anat. 180, 332–341. Boot, M.J., Gittenberger-de Groot, A.C., Van Iperen, L., Hierck, B.P., Poelmann, R.E., 2003. Spatiotemporally separated cardiac neural crest subpopulations that target the outflow tract septum and pharyngeal arch arteries. Anat. Rec. A. Discov. Mol. Cell. Evol. Biol. 275, 1009–1018. Boutros, M., Mlodzik, M., 1999. Dishevelled, at the crossroads of divergent intracellular signaling pathways. Mech. Dev. 83, 27–37. Bronner-Fraser, M., Fraser, S., 1988. Cell lineage analysis reveals multipotency of some avian neural crest cells. Nature 335, 161–164. Bronner-Fraser, M., Fraser, S., 1989. Developmental potential of avian trunk neural crest cells in situ. Neuron 3, 755–766. Bronner-Fraser, M., Wolf, J.J., Murray, B.A., 1992. Effects of antibodies against N-cadherin and N-CAM on the cranial neural crest and neural tube. Dev. Biol. 153, 291–301. Brown, C.B., Feiner, L., Lu, M.M., Li, J., Ma, X., Webber, A.L., Jia, L., Raper, J.A., Epstein, J.A., 2001. PlexinA2 and semaphorin signaling

435

during cardiac neural crest development. Development 128, 3071–3080. Cai, C.L., Liang, X., Shi, Y., Chu, P.H., Pfaff, S.L., Chen, J., Evans, S., 2003. Isl1 identifies a cardiac progenitor population that proliferates prior to differentiation and contributes a majority of cells to the heart. Dev. Cell 5, 877–889. Cano, A., Perez-Moreno, M.A., Rodrigo, I., Locascio, A., Blanco, M.J., del Barrio, M.G., Portillo, F., Nieto, M.A., 2000. The transcription factor snail controls epithelial-mesenchymal transitions by repressing E-cadherin expression. Nat. Cell Biol. 2, 76–83. Carl, T.F., Dufton, C., Hanken, J., Klymkowsky, M.W., 1999. Inhibition of neural crest migration in Xenopus using antisense slug RNA. Dev. Biol. 213, 101–115. Chan, W.Y., Cheung, C.S., Yung, K.M., Copp, A.J., 2004. Cardiac neural crest of the mouse embryo, axial level of origin, migratory pathway and cell autonomy of the splotch (Sp2H) mutant effect. Development 131, 3367–3379. Chapman, D.L., Garvey, N., Hancock, S., Alexiou, M., Agulnik, S.I., Gibson-Brown, J.J., Cebra-Thomas, J., Bollag, R.J., Silver, L.M., Papaioannou, V.E., 1996. Expression of the T-box family genes, Tbx1-Tbx5, during early mouse development. Dev. Dyn. 206, 379–390. Chen Jr., Y.H., Ishii, M., Sun, J., Sucov, H.M., Maxson, R.E., 2007. Msx1 and Msx2 regulate survival of secondary heart field precursors and post-migratory proliferation of cardiac neural crest in the outflow tract. Dev. Biol. 308, 421–437. Cheng, G., Litchenberg, W.H., Cole, G.J., Mikawa, T., Thompson, R.P., Gourdie, R.G., 1999. Development of the cardiac conduction system involves recruitment within a multipotent cardiomyogenic lineage. Development 126, 5041–5049. Cheung, M., Chaboissier, M.-C., Mynett, A., Hirst, E., Schedl, A., Briscoe, J., 2005. The transcriptional control of trunk neural crest induction, survival, and delamination. Dev. Cell 8, 179–192. Clouthier, D.E., Hosoda, K., Richardson, J.A., Williams, S.C., Yanagisawa, H., Kuwaki, T., Kumada, M., Hammer, R.E., Yanagisawa, M., 1998. Cranial and cardiac neural crest defects in endothelin-A receptor-deficient mice. Development 125, 813–824. Clouthier, D.E., Williams, S.C., Hammer, R.E., Richardson, J.A., Yanagisawa, M., 2003. Cell-autonomous and nonautonomous actions of endothelin-A receptor signaling in craniofacial and cardiovascular development. Dev. Biol. 261, 506–519. Conway, S.J., Henderson, D.J., Copp, A.J., 1997. Pax3 is required for cardiac neural crest migration in the mouse, evidence from the splotch (Sp2H) mutant. Development 124, 505–514. Conway, S.J., Bundy, J., Chen, J., Dickman, E., Rogers, R., Will, B.M., 2000. Decreased neural crest stem cell expansion is responsible for the conotruncal heart defects within the splotch (Sp(2H))/Pax3 mouse mutant. Cardiovasc. Res. 47, 314–328. Creazzo, T.L., Godt, R.E., Leatherbury, L., Conway, S.J., Kirby, M.L., 1998. Role of cardiac neural crest cells in cardiovascular development. Annu. Rev. Physiol. 60, 267–286. Davy, A., Aubin, J., Soriano, P., 2004. Ephrin-B1 forward and reverse signaling are required during mouse development. Genes. Dev. 18, 572–583. Davy, A., Soriano, P., 2005. Ephrin signaling in vivo, look both ways. Dev. Dyn. 232, 1–10. de Lange, F.J., Moorman, A.F., Anderson, R.H., Manner, J., Soufan, A.T., de Gier-de Vries, C., Schneider, M.D., Webb, S., van den Hoff, M.J., Christoffels, V.M., 2004. Lineage and morphogenetic analysis of the cardiac valves. Circ. Res. 95, 645–654.

436

del Barrio, M.G., Nieto, M.A., 2002. Overexpression of Snail family members highlights their ability to promote chick neural crest formation. Development 129, 1583–1593. Delot, E.C., Bahamonde, M.E., Zhao, M., Lyons, K.M., 2003. BMP sig­ naling is required for septation of the outflow tract of the mammalian heart. Development 130, 209–220. Du, K.L., Ip, H.S., Li, J., Chen, M., Dandre, F., Yu, W., Lu, M.M., Owens, G.K., Parmacek, M.S., 2003. Myocardin is a critical serum response factor cofactor in the transcriptional program regulating smooth muscle cell differentiation. Mol. Cell Biol. 23, 2425–2437. Epstein, D.J., Vekemans, M., Gros, P., 1991. Splotch (Sp2H), a mutation affecting development of the mouse neural tube, shows a deletion within the paired homeodomain of Pax3. Cell 67, 767–774. Epstein, D.J., Vogan, K.J., Trasler, D.G., Gros, P., 1993. A mutation within intron 3 of the Pax-3 gene produces aberrantly spliced mRNA transcripts in the splotch (Sp) mouse mutant. Proc. Natl. Acad. Sci. USA 90, 532–536. Epstein, J.A., Li, J., Lang, D., Chen, F., Brown, C.B., Jin, F., Lu, M.M., Thomas, M., Liu, E., Wessels, A., Lo, C.W., 2000. Migration of cardiac neural crest cells in Splotch embryos. Development 127, 1869–1878. Ewart, J.L., Cohen, M.F., Meyer, R.A., Huang, G.Y., Wessels, A., Gourdie, R.G., Chin, A.J., Park, S.M., Lazatin, B.O., Villabon, S., Lo, C.W., 1997. Heart and neural tube defects in transgenic mice overexpressing the Cx43 gap junction gene. Development 124, 1281–1292. Farlie, P.G., Kerr, R., Thomas, P., Symes, T., Minichiello, J., Hearn, C.J., Newgreen, D., 1999. A paraxial exclusion zone creates patterned cranial neural crest cell outgrowth adjacent to rhombomeres 3 and 5. Dev. Biol. 213, 70–84. Feiner, L., Webber, A.L., Brown, C.B., Lu, M.M., Jia, L., Feinstein, P., Mombaerts, P., Epstein, J.A., Raper, J.A., 2001. Targeted disruption of semaphorin 3C leads to persistent truncus arteriosus and aortic arch interruption. Development 128, 3061–3070. Franz, T., 1989. Persistent truncus arteriosus in the Splotch mutant mouse. Anat. Embryol. (Berlin). 180, 457–464. Fukiishi, Y., Morriss-Kay, G.M., 1992. Migration of cranial neural crest cells to the pharyngeal arches and heart in rat embryos. Cell Tissue. Res. 268, 1–8. Gammill, L.S., Gonzalez, C., Bronner-Fraser, M., 2006a. Neuropilin 2/ semaphorin 3F signaling is essential for cranial neural crest migration and trigeminal ganglion condensation. J. Neurobiol. 67 (1), 47–56. Gammill, L.S., Gonzalez, C., Gu, C., Bronner-Fraser, M., 2006b. Guidance of trunk neural crest migration requires neuropilin 2/semaphorin 3F signaling. Development 133, 99–106. Gittenberger-de Groot, A.C., Bartelings, M.M., Deruiter, M.C., Poelmann, R.E., 2005. Basics of cardiac development for the understanding of congenital heart malformations. Pediatr. Res. 57, 169–176. Goddeeris, M.M., Schwartz, R., Klingensmith, J., Meyers, E.N., 2007. Independent requirements for Hedgehog signaling by both the anter­ ior heart field and neural crest cells for outflow tract development. Development 134, 1593–1604. Golding, J.P., Trainor, P., Krumlauf, R., Gassman, M., 2000. Defects in pathfinding by cranial neural crest cells in mice lacking the Neuregulin receptor ErbB4. Nat. Cell Biol. 2, 103–109. Goulding, M., Sterrer, S., Fleming, J., Balling, R., Nadeau, J., Moore, K.J., Brown, A.D.M., Steel, K.P., Gruss, P., 1993. Analysis of the Pax3 mutant Splotch. Genomics 17, 355–363. Grimes, A.C., Stadt, H.A., Shepherd, I.T., Kirby, M.L., 2006. Solving an enigma, Arterial pole development in the zebrafish heart. Dev. Biol. 290, 265–276.

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

Gu, Z., Reynolds, E.M., Song, J., Lei, H., Feijen, A., Yu, L., He, W., MacLaughlin, D.T., van den Eijnden-van Raaij, J., Donahoe, P.K., Li, E., 1999. The type I serine/threonine kinase receptor ActRIA (ALK2) is required for gastrulation of the mouse embryo. Development 126, 2551–2561. Gurjarpadhye, A., Hewett, K.W., Justus, C., Wen, X., Stadt, H., Kirby, M.L., Sedmera, D., Gourdie, R.G., 2007. Cardiac neural crest ablation inhibits compaction and electrical function of conduction system bundles. Am. J. Physiol. Heart Circ. Physiol. 292, H1291–H1300. Hall, B.K., 2000. The neural crest as a fourth germ layer and vertebrates as quadroblastic not triploblastic. Evol. Dev. 2, 3–5. Hamblet, N.S., Lijam, N., Ruiz-Lozano, P., Wang, J., Yang, Y., Luo, Z., Mei, L., Chien, K.R., Sussman, D.J., Wynshaw-Boris, A., 2002. Dishevelled 2 is essential for cardiac outflow tract development, somite segmentation and neural tube closure. Development 129, 5827–5838. High, F.A., Zhang, M., Proweller, A., Tu, L., Parmacek, M.S., Pear, W.S., Epstein, J.A., 2007. An essential role for Notch in neural crest during cardiovascular development and smooth muscle differentiation. J. Clin. Invest. 117, 353–363. Huang, G.Y., Cooper, E.S., Waldo, K., Kirby, M.L., Gilula, N.B., Lo, C.W., 1998. Gap junction-mediated cell–cell communication modulates mouse neural crest migration. J. Cell Biol. 143, 1725–1734. Hutson, M.R., Kirby, M.L., 2003. Neural crest and cardiovascular development, a 20-year perspective. Birth Defects Res. C. Embryo. Today 69, 2–13. Hutson, M.R., Zhang, P., Stadt, H.A., Sato, A.K., Li, Y.X., Burch, J., Creazzo, T.L., Kirby, M.L., 2006. Cardiac arterial pole alignment is sensitive to FGF8 signaling in the pharynx. Dev. Biol. 295, 486–497. Inoue, T., Chisaka, O., Matsunami, H., Takeichi, M., 1997. Cadherin-6 expression transiently delineates specific rhombomeres, other neural tube subdivisions, and neural crest subpopulations in mouse embryos. Dev. Biol. 183, 183–194. Jerome, L.A., Papaioannou, V.E., 2001. DiGeorge syndrome phenotype in mice mutant for the T-box gene, Tbx1. Nat. Genet. 27, 286–291. Jia, Q., McDill, B.W., Li, S.Z., Deng, C., Chang, C.P., Chen, F., 2007. Smad signaling in the neural crest regulates cardiac outflow tract remodeling through cell autonomous and non-cell autonomous effects. Dev. Biol. 311, 172–184. Jiang, X., Rowitch, D.H., Soriano, P., McMahon, A.P., Sucov, H.M., 2000. Fate of the mammalian cardiac neural crest. Development 127, 1607–1616. Kaartinen, V., Dudas, M., Nagy, A., Sridurongrit, S., Lu, M.M., Epstein, J.A., 2004. Cardiac outflow tract defects in mice lacking ALK2 in neural crest cells. Development 131, 3481–3490. Kaufman, M.H., 1992. The Atlas of Mouse Development. Academic Press, London, UK. Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1, 435–440. Kioussi, C., Briata, P., Baek, S.H., Rose, D.W., Hamblet, N.S., Herman, T., Ohgi, K.A., Lin, C., Gleiberman, A., Wang, J., Brault, V., Ruiz-Lozano, P., Nguyen, H.D., Kemler, R., Glass, C.K., Wynshaw-Boris, A., Rosenfeld, M.G., 2002. Identification of a Wnt/Dvl/beta-catenin -- Pitx2 pathway mediating cell-type-specific proliferation during development. Cell 111, 673–685. Kirby, M., 1989. Plasticity and predetermination of the mesencephalic and trunk neural crest transplanted into the region of cardiac neural crest. Dev. Biol. 134, 402–412. Kirby, M.L., Gale, T.F., Stewart, D.E., 1983. Neural crest cells contribute to normal aorticopulmonary septation. Science 220, 1059–1061.

Chapter  |  7.1  Role of Cardiac Neural Crest Cells in Morphogenesis of the Heart and Great Vessels

Kirby, M., Stewart, D., 1984. Adrenergic innervation of the developing chick heart, neural crest ablations to produce sympathetically anueral hearts. Am. J. Anat. 171, 295–305. Kirby, M.L., Turnage, K.L., 3rd, Hays, B.M., 1985. Characterization of conotruncal malformations following ablation of “cardiac” neural crest. Anat. Rec. 213, 87–93. Kirby, M.L., Waldo, K.L., 1995. Neural crest and cardiovascular patterning. Circ. Res. 77, 211–215. Kishigami, S., Mishina, Y., 2005. BMP signaling and early embryonic patterning. Cytokine. Growth Factor Rev. 16, 265–278. Kitajima, S., Miyagawa-Tomita, S., Inoue, T., Kanno, J., Saga, Y., 2006. Mesp1-nonexpressing cells contribute to the ventricular cardiac conduction system. Dev. Dyn. 235, 395–402. Kolodkin, A.L., Matthes, D.J., Goodman, C.S., 1993. The semaphorin genes encode a family of transmembrane and secreted growth cone guidance molecules. Cell 75, 1389–1399. Kretz, M., Eckardt, D., Kruger, O., Kim, J.S., Maurer, J., Theis, M., van Rijen, H.V., Schorle, H., Willecke, K., 2006. Normal embryonic development and cardiac morphogenesis in mice with Wnt1-Cremediated deletion of connexin43. Genesis 44, 269–276. Krull, C.E., Lansford, R., Gale, N.W., Marcelle, C., Collazo, A., Yancopoulos, G.D., Fraser, S.E., Bronner-Fraser, M., 1997. Interactions of Eph-related receptors and ligands confer rostrocaudal pattern to trunk neural crest migration. Curr. Biol. 7, 571–580. Kurihara, Y., Kurihara, H., Oda, H., Maemura, K., Nagai, R., Ishikawa, T., Yazaki, Y., 1995. Aortic arch malformations and ventricular septal defect in mice deficient in endothelin-1. J. Clin. Invest. 96, 293–300. LaBonne, C., Bronner-Fraser, M., 2000. Snail-related transcriptional repressors are required in Xenopus for both the induction of the neural crest and its subsequent migration. Dev. Biol. 221, 195–205. Le Lievre, C.S, Le Douarin, N.M., 1975. Mesenchymal derivatives of the neural crest: Analysis of chimaeric quail and chick embryos. J. Embryol. Exp. Morphol. 34, 125–154. Lee, R.Y., Luo, J., Evans, R.M., Giguere, V., Sucov, H.M., 1997. Compartment-selective sensitivity of cardiovascular morphogenesis to combinations of retinoic acid receptor gene mutations. Circ. Res. 80, 757–764. Lee, Y.M., Osumi-Yamashita, N., Ninomiya, Y., Moon, C.K., Eriksson, U., Eto, K., 1995. Retinoic acid stage-dependently alters the migration pattern and identity of hindbrain neural crest cells. Development 121, 825–837. Li, J., Liu, K.C., Jin, F., Lu, M.M., Epstein, J.A., 1999. Transgenic rescue of congenital heart disease and spina bifida in Splotch mice. Development 126, 2495–2503. Li, J., Chen, F., Epstein, J.A., 2000. Neural crest expression of Cre recombinase directed by the proximal Pax3 promoter in transgenic mice. Genesis 26, 162–164. Li, J., Zhu, X., Chen, M., Cheng, L., Zhou, D., Lu, M.M., Du, K., Epstein, J.A., Parmacek, M.S., 2005. Myocardin-related transcription factor B is required in cardiac neural crest for smooth muscle differentiation and cardiovascular development. Proc. Natl. Acad. Sci. USA 102, 8916–8921. Li, Y.X., Zdanowicz, M., Young, L., Kumiski, D., Leatherbury, L., Kirby, M.L., 2003. Cardiac neural crest in zebrafish embryos contributes to myocardial cell lineage and early heart function. Dev. Dyn. 226, 540–550. Liao, J., Kochilas, L., Nowotschin, S., Arnold, J.S., Aggarwal, V.S., Epstein, J.A., Brown, M.C., Adams, J., Morrow, B.E., 2004. Full spectrum of malformations in velo-cardio-facial syndrome/DiGeorge syndrome mouse models by altering Tbx1 dosage. Hum. Mol. Genet. 13, 1577–1585.

437

Lindsay, E.A., Vitelli, F., Su, H., Morishima, M., Huynh, T., Pramparo, T., Jurecic, V., Ogunrinu, G., Sutherland, H.F., Scambler, P.J., Bradley, A., Baldini, A., 2001. Tbx1 haploinsufficieny in the DiGeorge syndrome region causes aortic arch defects in mice. Nature 410, 97–101. Liu, C., Liu, W., Palie, J., Lu, M.F., Brown, N.A., Martin, J.F., 2002. Pitx2c patterns anterior myocardium and aortic arch vessels and is required for local cell movement into atrioventricular cushions. Development 129, 5081–5091. Liu, S., Liu, F., St Schneider, A.E., Amand, T., Epstein, J.A., Gutstein, D.E., 2006. Distinct cardiac malformations caused by absence of connexin 43 in the neural crest and in the non-crest neural tube. Development 133, 2063–2073. Lo, C.W., Cohen, M.F., Huang, G.Y., Lazatin, B.O., Patel, N., Sullivan, R., Pauken, C., Park, S.M., 1997. Cx43 gap junction gene expression and gap junctional communication in mouse neural crest cells. Dev. Genet. 20, 119–132. Manley, N., Capecchi, M., 1995. The role of Hoxa-3 in mouse thymus and thyroid development. Development 121, 1989–2003. Mark, M.D., Lohrum, M., Puschel, A.W., 1997. Patterning neuronal connections by chemorepulsion, the semaphorins. Cell Tissue Res. 290, 299–306. Maschhoff, K.L., Baldwin, H.S., 2000. Molecular determinants of neural crest migration. Am. J. Med. Genet. 97, 280–288. Massague, J., Seoane, J., Wotton, D., 2005. Smad transcription factors. Genes. Dev. 19, 2783–2810. Matzuk, M.M., Kumar, T.R., Bradley, A., 1995. Different phenotypes for mice deficient in either activins or activin receptor type II. Nature 374, 356–360. Mayor, R., Morgan, R., Sargent, M.G., 1995. Induction of the prospective neural crest of Xenopus. Development 121, 767–777. Meilhac, S.M., Esner, M., Kelly, R.G., Nicolas, J.F., Buckingham, M.E., 2004. The clonal origin of myocardial cells in different regions of the embryonic mouse heart. Dev. Cell 6, 685–698. Mjaatvedt, C.H., Nakaoka, T., Moreno-Rodriguez, R., Norris, R.A., Kern, M.J., Eisenberg, C.A., Turner, D., Markwald, R.R., 2001. The outflow tract of the heart is recruited from a novel heart-forming field. Dev. Biol. 238, 97–109. Monsoro-Burq, A.H., Fletcher, R.B., Harland, R.M., 2003. Neural crest induction by paraxial mesoderm in Xenopus embryos requires FGF signals. Development 130, 3111–3124. Monsoro-Burq, A.H., Wang, E., Harland, R., 2005. Msx1 and Pax3 cooperate to mediate FGF8 and WNT signals during Xenopus neural crest induction. Dev. Cell 8, 167–178. Moraes, F., Novoa, A., Jerome-Majewska, L.A., Papaioannou, V.E., Mallo, M., 2005. Tbx1 is required for proper neural crest migration and to stabilize spatial patterns during middle and inner ear development. Mech. Dev. 122, 199–212. Moury, J.D., Jacobson, A.G., 1990. The origins of neural crest cells in the axolotl. Dev. Biol. 141, 243–253. Nakagawa, S., Takeichi, M., 1995. Neural crest cell–cell adhesion controlled by sequential and subpopulation-specific expression of novel cadherins. Development 121, 1321–1332. Nakagawa, S., Takeichi, M., 1998. Neural crest emigration from the neural tube depends on regulated cadherin expression. Development 125, 2963–2971. Nakamura, T., Colbert, M.C., Robbins, J., 2006. Neural crest cells retain multipotential characteristics in the developing valves and label the cardiac conduction system. Circ. Res. 98, 1547–1554. Niederreither, K., Vermot, J., Messaddeq, N., Schuhbaur, B., Chambon, P., Dolle, P., 2001. Embryonic retinoic acid synthesis is essential for heart morphogenesis in the mouse. Development 128, 1019–1031.

438

Nieto, M.A., Sargent, M.G., Wilkinson, D.G., Cooke, J., 1994. Control of cell behavior during vertebrate development by slug, a zinc finger gene. Science 264, 835–839. Nishibatake, M., Kirby, M.L., Van Mierop, L.H., 1987. Pathogenesis of persistent truncus arteriosus and dextroposed aorta in the chick embryo after neural crest ablation. Circulation 75, 255–264. Noden, D., 1983. The role of the neural crest in patterning of avian cranial skeletal, connective, and muscle tissues. Dev. Biol. 96, 144–165. Noden, D.M., 1991. Origins and patterning of avian outflow tract endocardium. Development 111, 867–876. Northcutt, R.G., Gans, C., 1983. The genesis of neural crest and epidermal placodes, a reinterpretation of vertebrate origins. Q Rev. Biol. 58, 1–28. Osumi-Yamashita, N., Ninomiya, Y., Doi, H., Eto, K., 1996. Rhombomere formation and hind-brain crest cell migration from prorhombomeric origins in mouse embryos. Develop. Growth Differ. 38, 107–118. Ota, M.S., Loebel, D.A., O’Rourke, M.P., Wong, N., Tsoi, B., Tam, P.P., 2004. Twist is required for patterning the cranial nerves and maintaining the viability of mesodermal cells. Dev. Dyn. 230, 216–228. Owens, G.K., Kumar, M.S., Wamhoff, B.R., 2004. Molecular regulation of vascular smooth muscle cell differentiation in development and disease. Physiol Rev. 84, 767–801. Panchision, D.M., Pickel, J.M., Studer, L., Lee, S.H., Turner, P.A., Hazel, T.G., McKay, R.D., 2001. Sequential actions of BMP receptors control neural precursor cell production and fate. Genes. Dev. 15, 2094–2110. Poelmann, R.E., Gittenberger-de Groot, A.C, 1999. A subpopulation of apoptosis-prone cardiac neural crest cells targets to the venous pole, multiple functions in heart development? Dev. Biol. 207, 271–286. Poelmann, R.E., Jongbloed, M.R., Molin, D.G., Fekkes, M.L., Wang, Z., Fishman, G.I., Doetschman, T., Azhar, M., Gittenberger-de Groot, A. C., 2004. The neural crest is contiguous with the cardiac conduction system in the mouse embryo, a role in induction? Anat. Embryol. (Berlin) 208, 389–393. Pratt, R.M., Goulding, E.H., Abbott, B.D., 1987. Retinoic acid inhibits migration of cranial neural crest cells in the cultured mouse embryo. J. Craniofac. Genet. Dev. Biol. 7, 205–217. Rahkonen, O., Savontaus, M., Abdelwahid, E., Vuorio, E., Jokinen, E., 2003. Expression patterns of cartilage collagens and Sox9 during mouse heart development. Histochem. Cell Biol. 120, 103–110. Raper, J.A., 2000. Semaphorins and their receptors in vertebrates and invertebrates. Curr. Opin. Neurobiol. 10, 88–94. Reaume, A.G., de Sousa, P.A., Kulkarni, S., Langille, B.L., Zhu, D., Davies, T.C., Juneja, S.C., Kidder, G.M., Rossant, J., 1995. Cardiac malformation in neonatal mice lacking connexin43. Science 267, 1831–1834. Richarte, A.M., Mead, H.B., Tallquist, M.D., 2007. Cooperation between the PDGF receptors in cardiac neural crest cell migration. Dev. Biol. 306, 785–796. Rollhauser-ter Horst, J., 1977. Artificial neural induction in amphibia. I. Sandwich explants. Anat. Embryol. (Berlin) 151, 309–316. Ryan, A.K., Goodship, J.A., Wilson, D.I., Philip, N., Levy, A., Seidel, H., Schuffenhauer, S., Oechsler, H., Belohradsky, B., Prieur, M., Aurias, A., Raymond, F.L., Clayton-Smith, J., Hatchwell, E., McKeown, C., Beemer, F.A., Dallapiccola, B., Novelli, G., Hurst, J.A., Ignatius, J., Green, A.J., Winter, R.M., Brueton, L., Brondum-Nielsen, K., Scambler, P.J., et al., 1997. Spectrum of clinical features associated with interstitial chromosome 22q11 deletions, a European collaborative study. J. Med. Genet. 34, 798–804. Sadaghiani, B., Vielkind, J.R., 1990. Distribution and migration pathways of HNK-1-immunoreactive neural crest cells in teleost fish embryos. Development 110, 197–209.

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

Saga, Y., Miyagawa-Tomita, S., Takagi, A., Kitajima, S., Miyazaki, J., Inoue, T., 1999. MesP1 is expressed in the heart precursor cells and required for the formation of a single heart tube. Development 126, 3437–3447. Saga, Y., Kitajima, S., Miyagawa-Tomita, S., 2000. Mesp1 expression is the earliest sign of cardiovascular development. Trends. Cardiovasc. Med. 10, 345–352. Sato, M., Tsai, H.J., Yost, H.J., 2006. Semaphorin3D regulates invasion of cardiac neural crest cells into the primary heart field. Dev. Biol. 298, 12–21. Scambler, P.J., 2000. The 22q11 deletion syndromes. Hum. Mol. Genet. 9, 2421–2426. Schafer, K., Neuhaus, P., Kruse, J., Braun, T., 2003. The homeobox gene Lbx1 specifies a subpopulation of cardiac neural crest necessary for normal heart development. Circ. Res. 92, 73–80. Schilling, T., 2001. Plasticity of zebrafish Hox expression in the hindbrain and cranial neural crest hindbrain. Dev. Biol. 231, 201–216. Selleck, M.A., Bronner-Fraser, M., 1995. Origins of the avian neural crest, the role of neural plate-epidermal interactions. Development 121, 525–538. Serbedzija, G.N., Bronner-Fraser, M., Fraser, S.E., 1992. Vital dye analysis of cranial neural crest cell migration in the mouse embryo. Development 116, 297–307. Shiratori, H., Hamada, H., 2006. The left–right axis in the mouse, from origin to morphology. Development 133, 2095–2104. Smith, A., Robinson, V., Patel, K., Wilkinson, D.G., 1997. The EphA4 and EphB1 receptor tyrosine kinases and ephrin-B2 ligand regulate targeted migration of branchial neural crest cells. Curr. Biol. 7, 561–570. Soo, K., O’Rourke, M.P., Khoo, P.L., Steiner, K.A., Wong, N., Behringer, R.R., Tam, P.P., 2002. Twist function is required for the morphogenesis of the cephalic neural tube and the differentiation of the cranial neural crest cells in the mouse embryo. Dev. Biol. 247, 251–270. Stalmans, I., Lambrechts, D., De smet, F., Jansen, S., Wang, J., Maity, S., Kneer, P., von der Ohe, M., Swillen, A., Maes, C., Gewillig, M., Molin, D.G., Hellings, P., Boetel, T., Haardt, M., Compernolle, V., Dewerchin, M., Plaisance, S., Vlietinck, R., Emanuel, B., Gittenberger-de Groot, A.C., Scambler, P., Morrow, B., Driscol, D.A., Moons, L., Esguerra, C.V., Carmeliet, G., Behn-Krappa, A., Devriendt, K., Collen, D., Conway, S.J., Carmeliet, P., 2003. VEGF, A modifier of the del22q11 (DiGeorge) syndrome? Nat. Med. 9, 173–182. Stottmann, R.W., Choi, M., Mishina, Y., Meyers, E.N., Klingensmith, J., 2004. BMP receptor IA is required in mammalian neural crest cells for development of the cardiac outflow tract and ventricular myocardium. Development 131, 2205–2218. Tallquist, M.D., Soriano, P., 2003. Cell autonomous requirement for PDGFRalpha in populations of cranial and cardiac neural crest cells. Development 130, 507–518. Taneyhill, L.A., Coles, E.G., Bronner-Fraser, M., 2007. Snail2 directly represses cadherin6B during epithelial-to-mesenchymal transitions of the neural crest. Development 134, 1481–1490. Tessarollo, L., Tsoulfas, P., Donovan, M.J., Palko, M.E., Blair-Flynn, J., Hempstead, B.L., Parada, L.F., 1997. Targeted deletion of all isoforms of the trkC gene suggests the use of alternate receptors by its ligand neurotrophin-3 in neuronal development and implicates trkC in normal cardiogenesis. Proc. Natl. Acad. Sci. USA 94, 14776–14781. Thorogood, P., Smith, L., Nicol, A., McGinty, R., Garrod, D., 1982. Effects of vitamin A on the behaviour of migratory neural crest cells in vitro. J. Cell Sci. 57, 331–350.

Chapter  |  7.1  Role of Cardiac Neural Crest Cells in Morphogenesis of the Heart and Great Vessels

Tomita, Y., Matsumura, K., Wakamatsu, Y., Matsuzaki, Y., Shibuya, I., Kawaguchi, H., Ieda, M., Kanakubo, S., Shimazaki, T., Ogawa, S., Osumi, N., Okano, H., Fukuda, K., 2005. Cardiac neural crest cells contribute to the dormant multipotent stem cell in the mammalian heart. J. Cell Biol. 170, 1135–1146. Trainor, P.A., Krumlauf, R., 2000. Plasticity in mouse neural crest cells reveals a new patterning role for cranial mesoderm. Nat. Cell Biol. 2, 96–102. Trainor, P.A., Sobieszczuk, D., Wilkinson, D., Krumlauf, R., 2002. Signalling between the hindbrain and paraxial tissues dictates neural crest migration pathways. Development 129, 433–442. Verberne, M.E., Gittenberger-de Groot, A.C., Poelmann, R.E., 1998. Lineage and development of the parasympathetic nervous system of the embryonic chick heart. Anat. Embryol. (Berlin), 198, 171–184. Verberne, M.E., Gittenberger-de Groot, A.C., Van Iperen, L., Poelmann, R.E., 1999. Contribution of the cervical sympathetic ganglia to the innervation of the pharyngeal arch arteries and the heart in the chick embryo. Anat. Rec. 255, 407–419. Vincentz, J.W., Barnes, R.M., Rodgers, R., Firulli, B.A., Conway, S.J., Firulli, A.B., 2008. An absence of Twist1 results in aberrant cardiac neural crest morphogenesis. Dev. Biol. 320, 131–139. Vitelli, F., Morishima, M., Taddei, I., Lindsay, E.A., Baldini, A., 2002. Tbx1 mutation causes multiple cardiovascular defects and disrupts neural crest and cranial nerve migratory pathways. Hum. Mol. Genet. 11, 915–922. Vogan, K.J., Epstein, D.J., Trasler, D.G., Gros, P., 1993. The splotchdelayed (Spd) mouse mutant carries a point mutation within the paired box of the Pax-3 gene. Genomics 17, 364–369. Wagner, G., 1949. Die Bedeutung der Neualleiste fur die Kopfgestaltung der Amphibienlarven. Rev. Suisse. Zool. 56, 519–620. Wagner, G., 1959. Untersuchungen an Bombinator-Triton-Chimareren. Roux. Arch. Entwicklungsmech. 151, 136–158. Waldo, K.L., Kumiski, D., Kirby, M.L., 1996. Cardiac neural crest is essential for the persistence rather than the formation of an arch artery. Dev. Dyn. 205, 281–292. Waldo, K.L., Lo, C.W., Kirby, M.L., 1999. Connexin 43 expression reflects neural crest patterns during cardiovascular development. Dev. Biol. 208, 307–323. Waldo, K.L., Kumiski, D.H., Wallis, K.T., Stadt, H.A., Hutson, M.R., Platt, D.H., Kirby, M.L., 2001. Conotruncal myocardium arises from a secondary heart field. Development 128, 3179–3188. Waldo, K.L., Hutson, M.R., Stadt, H.A., Zdanowicz, M., Zdanowicz, J., Kirby, M.L., 2005. Cardiac neural crest is necessary for normal addition of the myocardium to the arterial pole from the secondary heart field. Dev. Biol. 281, 66–77. Wang, D.-Z., Chang, P.S., Wang, Z., Sutherland, L., Richardson, J.A., Small, E., Krieg, P.A., Olson, E.N., 2001. Activation of cardiac gene expression by myocardin, a transcriptional cofactor for serum response factor. Cell 105, 851–862. Wang, D.-Z., Olson, E.N., 2004. Control of smooth muscle development by the myocardin family of transcriptional coactivators. Curr. Opin. Genet. Dev. 14, 558–566. Wang, H.U., Anderson, D.J., 1997. Eph family transmembrane ligands can mediate repulsive guidance of trunk neural crest migration and motor axon outgrowth. Neuron 18, 383–396. Washington Smoak, I., Byrd, N.A., Abu-Issa, R., Goddeeris, M.M., Anderson, R., Morris, J., Yamamura, K., Klingensmith, J., Meyers, E.N.,

439

2005. Sonic hedgehog is required for cardiac outflow tract and neural crest cell development. Dev. Biol. 283, 357–372. Wendling, O., Ghyselinck, N.B., Chambon, P., Mark, M., 2001. Roles of retinoic acid receptors in early embryonic morphogenesis and hindbrain patterning. Development 128, 2031–2038. Wilson, J.G., Warkany, J., 1949. Aortic arch and cardiac anomalies in the offspring of vitamin A deficient rats. Am. J. Anat. 85, 113–155. Wolf, C., Thisse, C., Stoetzel, C., Thisse, B., Gerlinger, P., PerrinSchmitt, F., 1991. The M-twist gene of Mus is expressed in subsets of mesodermal cells and is closely related to the Xenopus X-twi and the Drosophila twist genes. Dev. Biol. 143, 363–373. Xu, H., Cerrato, F., Baldini, A., 2005. Timed mutation and cell-fate mapping reveal reiterated roles of Tbx1 during embryogenesis, and a crucial function during segmentation of the pharyngeal system via regulation of endoderm expansion. Development 132, 4387–4395. Xu, X., Li, W.E., Huang, G.Y., Meyer, R., Chen, T., Luo, Y., Thomas, M.P., Radice, G.L., Lo, C.W., 2001. Modulation of mouse neural crest cell motility by N-cadherin and connexin 43 gap junctions. J. Cell Biol. 154, 217–230. Xu, X., Francis, R., Wei, C.J., Linask, K.L., Lo, C.W., 2006. Connexin 43mediated modulation of polarized cell movement and the directional migration of cardiac neural crest cells. Development 133, 3629–3639. Yagi, H., Furutani, Y., Hamada, H., Sasaki, T., Asakawa, S., Minoshima, S., Ichida, F., Joo, K., Kimura, M., Imamura, S., Kamatani, N., Momma, K., Takao, A., Nakazawa, M., Shimizu, N., Matsuoka, R., 2003. Role of TBX1 in human del22q11.2 syndrome. Lancet 362, 1366–1373. Yamauchi, Y., Abe, K., Mantani, A., Hitoshi, Y., Suzuki, M., Osuzu, F., Kuratani, S., Yamamura, K., 1999. A novel transgenic technique that allows specific marking of the neural crest cell lineage in mice. Dev. Biol. 212, 191–203. Yanagisawa, H., Hammer, R.E., Richardson, J.A., Williams, S.C., Clouthier, D.E., Yanagisawa, M., 1998a. Role of Endothelin-1/ Endothelin-A receptor-mediated signaling pathway in the aortic arch patterning in mice. J. Clin. Invest. 102, 22–33. Yanagisawa, H., Yanagisawa, M., Kapur, R.P., Richardson, J.A., Williams, S.C., Clouthier, D.E., de Wit, D., Emoto, N., Hammer, R. E., 1998b. Dual genetic pathways of endothelin-mediated intercellular signaling revealed by targeted disruption of endothelin converting enzyme-1 gene. Development 125, 825–836. Yashiro, K., Shiratori, H., Hamada, H., 2007. Haemodynamics determined by a genetic programme govern asymmetric development of the aortic arch. Nature 450, 285–288. Yelbuz, T.M., Waldo, K.L., Kumiski, D.H., Stadt, H.A., Wolfe, R.R., Leatherbury, L., Kirby, M.L., 2002. Shortened outflow tract leads to altered cardiac looping after neural crest ablation. Circulation 106, 504–510. Youn, Y.H., Feng, J., Tessarollo, L., Ito, K., Sieber-Blum, M., 2003. Neural crest stem cell and cardiac endothelium defects in the TrkC null mouse. Mol. Cell Neurosci. 24, 160–170. Yu, H.H., Moens, C.B., 2005. Semaphorin signaling guides cranial neural crest cell migration in zebrafish. Dev. Biol. 280, 373–385. Zaffran, S., Kelly, R.G., Meilhac, S.M., Buckingham, M.E., Brown, N.A., 2004. Right ventricular myocardium derives from the anterior heart field. Circ. Res. 95, 261–268. Zikova, M., Da Ponte, J.-P., Dastugue, B., Jagla, K., 2003. Patterning of the cardiac outflow region in Drosophila. Proc. Natl. Acad. Sci. USA 100, 12189–12194.

Chapter 7.2

Role of Cardiac Neural Crest in the Development of the Caudal Pharyngeal Arches, the Cardiac Outflow and Disease Mary R. Hutson and Margaret L. Kirby Department of Pediatrics (Neonatology), Neonatal-Perinatal Research Institute, Duke University Medical Center, Durham, NC, USA

I.  Introduction

I.A.  Origin of the Neural Crest

Neural crest cells are multipotential cells that delaminate from neuroepithelium and migrate widely throughout the body. The neural crest is broadly subdivided into cranial and trunk regions based on axial origin. A subregion of the cranial neural crest has been called “cardiac neural crest” because of the importance of these cells in heart development. This chapter will discuss the requirement of the cardiac neural crest cells in the formation of the septum that divides the cardiac arterial pole into systemic and pulmonary circulations (see also Chapter 7.1, Vol. I). Further, cardiac neural crest cells directly support the normal development and patterning of derivatives of the caudal pharyngeal arches, including the great arteries and the thymus, thyroid and parathyroids. Recently, cardiac neural crest cells have also been shown to indirectly influence the development of the secondary heart field, another derivative of the caudal pharynx, by modulating signaling in the pharynx. While most of what we know about the contribution and function of the cardiac neural crest has been learned in avian models, most of the genes associated with cardiac neural crest function have been identified using mouse models. Together these studies show that the neural crest cells may not always directly cause abnormal cardiovascular development, but may be involved secondarily because they represent a major component in the complex tissue interactions in the caudal pharynx and outflow tract. Cardiac neural crest cells span from the caudal pharynx into the outflow tract, and therefore may be susceptible to any perturbation in or by other cells in these regions. Thus, understanding congenital cardiac outflow malformations in human sequences of malformations resulting from genetic and/or environmental insults necessarily requires understanding development of the cardiac neural crest.

The neural crest generates a pluripotent cell population originating from most of the craniocaudal length of the dorsal neural tube. The cells of the neural crest delaminate and migrate widely throughout the embryo during development (Horstadius, 1950; Le Douarin, 1982; Bronner-Fraser and Fraser, 1989). Neural crest cells require a variety of environmental signals in order to be specified and then to delaminate from the neural tube. Growth factors and morphogens such as Wnts, fibroblast growth factors (FGFs), bone morphogenetic proteins (BMPs) and retinoic acid (RA) are required for neural crest induction (Steventon et al., 2005). Delamination from the neural tube is regulated via BMP-dependent Wnt1 activity (Burstyn-Cohen and Katcheim, 2002), with Wnt1 expression turning off soon after the cells leave the neural tube. Many transcription factors and signaling molecules have been implicated in migration, proliferation, survival and differentiation of the cardiac neural crest after their delamination, some of which are discussed below. The neural crest cells give rise to a wide variety of cell types including melanocytes, neurons and glia of the peripheral nervous system, cartilage, bone, smooth muscle and endocrine cells. Neural crest cell lineage is determined in part by the original axial location of the cell within the neural tube, as well as by the environmental influences encountered during migration (Fraser and Bronner-Fraser, 1991; Stern et al., 1991). Cranial neural crest extends rostrally from somite 5 and trunk neural crest extends caudally from somite 5. The cranial neural crest cells have a greater potential for generating mesenchymal cells than cells originating from the trunk neural crest (Horstadius, 1950; Le Douarin, 1982; Bronner-Fraser and

Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

441

442

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

Fraser, 1989). A subregion of the cranial neural crest extending from the level of the mid-otic placode to somite 3 has been named the cardiac neural crest because of its role in cardiovascular development (Kirby et al., 1985). Neurons, glia (Schwann and support cells), and melanocytes are generated from the entire length of the neural crest (Horstadius, 1950; Le Douarin, 1982; Bronner-Fraser and Fraser, 1989).

I.B.  Neural Crest Cell Lineages Numerous in vivo and in vitro studies have assessed whether certain neural crest cell lineages are determined before migration (Horstadius, 1950; Noden, 1978; Noden, 1980; Nakamura and Ayer-Le Lièvre, 1982; Weston, 1982; Wachtler, 1984; Le Douarin, 1990; Fraser and BronnerFraser, 1991; Ito and Sieber-Blum, 1991; Riccardi, 1991; Artinger and Bronner-Fraser, 1992; Erickson et al., 1992; Steller, 1995). Some cells are certainly multipotent cells, in that a single cell can give rise to both neuronal and nonneuronal cell types. In general, certain individual cells are determined for their lineage, i.e., melanocytes, neural, or mesenchymal before migrating, whereas others are established during migration or after they reach their terminal destination. While this can be said for individual cells, it appears that neither entire lineages nor cell populations are fully committed prior to migration. Neural crest cells from various axial positions can be grafted into new regions prior to migration and these adopt the cell fates of the new region, indicating a proportion of cells leaving the neural folds can be influenced by factors in the migration pathway before establishing their identity. It is not known whether a certain number of cells leaving the neural folds are determined to follow a neuronal or smooth muscle lineage. Certain cellular characteristics within a particular lineage, i.e., neurotransmitter identity, are not decided until neuronal precursors reach their terminal destination (Le Douarin and Kalcheim, 1999). However, as mentioned previously, the mesenchymal potential is limited to the cranial neural crest, and trunk neural crest transplanted into the cranial crest region cannot give rise to the skeletal structures of the head or support cardiovascular development (Kirby, 1989).

II.  Cardiac neural crest cells in   the caudal pharynx Much of the information regarding derivatives of the cardiac neural crest is derived from chick embryos using quailchick chimeras, in situ marking and ablation of the neural folds (Le Douarin and Jotereau, 1975; d’Amico-Martel and Noden, 1983; Kirby et al., 1983; Noden, 1983a,b; Nishibatake et al., 1987; Lumsden et al., 1991; MiyagawaTomita et al., 1991; Serbedzija et al., 1991; Couly et al., 1992; Le Douarin et al., 1993; Waldo et al., 1996), although

Figure 1  Cardiac neural crest cells migrate from the caudal rhombomeres into the caudal pharyngeal arches (3, 4 and 6) and from there into the outflow tract of the heart.

excellent neural crest lineage marked mouse models are now available and several laboratories have performed lineage tracing in mice and rats with vital dyes (Jaenisch, 1985; Smits-van Prooije et al., 1986; Tan and MorrissKay, 1986; Fukiishi and Morriss-Kay, 1992; Serbedzija et al., 1990, 1991, 1992; Lee et al., 1997; Jiang et al., 2000, 2002; Pietri et al., 2003; Nakamura et al., 2006). Results from all these experiments indicate relatively good agreement between rodent and avian models. The cardiac crest cells in avian and mammalian embryos originate from the transition region between cranial and trunk crest. Studies of zebrafish cardiac crest indicate that the axial origin is somewhat different and may be broader than in avian and rodents (Li et al., 2003; Sato and Yost, 2003).

II.A.  Pharyngeal Arch Arteries In avians and rodents, the cardiac neural crest cells originating from the caudal rhombencephalon migrate to pharyngeal arches 3, 4 and 6 (Fig. 1), and from there a subpopulation of cells continues to migrate into the heart, where they participate in outflow septation (see below) or form the cardiac parasympathetic ganglia of the heart (Kirby et al., 1983b). Cells from the cardiac neural crest also populate the esophagus and stomach en route to forming the parasympathetic enteric plexus of the midgut and hindgut (Peters-Van der Sanden et al., 1993; Kirby and Waldo, 1995). The neural crest-derived cells remaining in the caudal pharyngeal arches 3, 4 and 6 support development of the pharyngeal (aortic) arch arteries (Fig. 2) and the glands derived from this region (Fig. 2). Neural crestderived cells in the caudal pharyngeal arches, in contrast to those populating the cranial pharyngeal arches (1 and 2), have few skeletal derivatives but instead support vascular development of the great arteries that will become the major arterial conduits of the definitive, adult cardiovascular system (aorta, carotid and subclavian arteries, as well as

Chapter  |  7.2  Role of Cardiac Neural Crest in the Development of the Caudal Pharyngeal Arches

443

Figure 2  Cardiac neural crest cells in the caudal pharynx support development of the persisting aortic arch arteries into the great arteries of the thorax and form their smooth muscle tunics. In addition, the cells support development of the thymus, parathyroids and thyroid glands, and provide stromal cells after gland development.

the ductus arteriosus, a vessel that closes shortly after birth). The mesenchyme in the pharyngeal arches comes from two sources, the mesodermal cores and the neural crest-derived cells. In pharyngeal arches 3–6, neural crest-derived cells provide the majority of the mesenchyme with only a small portion contributed by mesodermal cores. This is in contrast to arches 1 and 2 which have substantial mesodermal cores (Kelly et al., 2001). There are 5 pharyngeal arches (1–4 and 6) and each arch artery develops in each pharyngeal arch, but only the arch arteries in arches 3–6 persist. The persistence of the aortic arch artery as a great artery is dependent on the neural crest-derived cells. The neural crest-derived cells in arches 3–6 invest strands of endothelial cells that will form the arch arteries (Waldo et al., 1996), while the arteries in arches 1 and 2 are not invested in neural crest-derived cells and they regress. It is possible that artery regression occurs because the mesenchyme of arches 1 and 2 is largely diverted to the formation of craniofacial skeleton and accessory structures. Further, the substantial mesodermal cores in arches 1 and 2 surround the arch arteries rather than neural crest cells. This suggests that patterning genes expressed by the neural crest cells are not available to these arch arteries to support their maintenance, whereas this is a primary function of the cells populating arches 3, 4 and 6 (Kirby et al., 1997; Qiu et al., 1997).

II.B.  Pharyngeal Glands The thymus and parathyroid glands form from endodermal pouches in the caudal pharynx via the interaction of

endoderm and neural crest-derived mesenchyme in the arches (Le Douarin and Jotereau, 1975; Bockman and Kirby, 1984). While the thyroid anlage is initially formed in the cranial pharynx, it migrates into the caudal pharynx and is dependent on mesenchyme derived from the caudal arches for normal development (Fig. 2) (Bockman and Kirby, 1984; Kuratani and Bockman, 1990).

III.  Cardiac neural crest and the formation of the arterial pole III.A.  Septation of the Arterial Pole After the cardiac neural crest cells migrate into arches 3, 4 and 6, they proliferate. In the chick these neural crest cells have been shown to form a shelf of condensed mesenchyme in the back of the aortic sac (distal outflow tract) between the origins of the fourth and sixth pairs of arch arteries to separate the systemic (third and fourth arch arteries) from the pulmonary outflow (sixth arch artery). A subset of the cells continues on into the preformed cardiac cushions of the middle outflow tract, where they collect as condensed cells in two columns or prongs centered in the outflow cushions (Fig. 3) (Waldo et al., 1998, 1999). The prongs are connected distally by the shelf of condensed neural crest cells that protrudes into the dorsal wall of the aortic sac. This complex of condensed mesenchyme located in the middle and distal outflow tract is called the aorticopulmonary septation complex (Waldo et al., 1998). The prongs end abruptly when they reach the proximal

444

(A)

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

(B)

(C)

Figure 3  Various stages in development of the arterial pole septum seen in quail-chick chimeras. (A) Condensed mesenchyme of the aorticopulmonary septation complex. The cells are continuous with the smooth muscle tunics of the great arteries. (B) Septation of the distal and middle portions of the arterial pole by the aorticopulmonary septation complex. The septum forms at the expense of the prongs. (C) Septation of the conus leaves a seam of cardiac neural crest cells.

outflow tract where neural crest cells are dispersed subendocardially in no set pattern (Fig. 3). Outflow septation occurs by three different mechanisms based on this initial configuration of neural crest cells (Fig. 3) (Waldo et al., 1998). First, the shelf protruding into the distal outflow tract elongates into the middle outflow tract at the expense of the prongs. Because the prongs spiral, the elongating septum spirals. The proximal outflow septum closes zipperlike from distal to proximal toward the ventricles. The proximal cushions are closed by the process of myocardialization in which invading myocardial cells cause the cushions to bulge into the lumen. The endocardium covering the cushions breaks down, allowing mixing of the underlying mesenchyme and myocardium, and this brings about the fusion of the opposing cushions to form a septum. Because the neural crest cells were subendocardial prior to the fusion, they now appear as a seam where the two cushions fused (Waldo et al., 1998, 1999). After division of the outflow tract, the mesenchyme that formed the outflow cushions is remodeled into aortic and pulmonary semilunar valves (see Chapters 6.1 and 6.2, Vol. I). Each valve has three cusps or leaflets. Some neural crest cells have been shown to be localized on the tips of the leaflets of the valves in chick and mouse (Waldo et al., 1998; Nakamura et al., 2006). The neural crest cells never represent a large proportion of the cell in the valves, and their role in valve formation is not known.

III.B.  Formation of the Arterial Pole The cardiac neural crest cells are only one cell population involved in the complex morphogenetic process to form the arterial pole of the heart. The arterial pole includes a distal

portion, the aortic sac, which is not invested in myocardium and which develops into the base of the aorta and pulmonary trunk; a myocardial covered middle portion, roughly comparable to the truncus arteriosus, which gives rise to the semilunar valve region; and a myocardial covered proximal portion, roughly comparable to the conus arteriosus which develops as the subvalvar region of both ventricles (Fig. 4). In the mature heart, the arterial pole is represented by the junction of the myocardium and smooth muscle at the base of the aorta and pulmonary trunk (Fig. 4). Normal development of the arterial pole is critical for normal connection of the right and left ventricles with the pulmonary and aortic trunks, respectively. A major source of cells that form the arterial pole is the splanchnic mesoderm located in the ventral caudal pharynx behind the outflow attachment to the pharynx. This mesoderm contributes the myocardium (Kelly et al., 2001; Mjaatvedt et al., 2001; Waldo et al., 2001) and smooth muscle to the arterial pole (Waldo et al., 2005a). This region of splanchnic mesoderm, called the secondary heart field, has been shown in tracing studies to be a subdivision of the cardiogenic field, but the cells do not differentiate as myocardium until the looping stage of heart development (Fig. 5) (Waldo et al., 2005; Abu-Issa and Kirby, 2008). In the chick, the presumptive myocardial cells from the secondary heart field move into the arterial pole in a spiraling pattern which may set up the spiraling cushions which the cardiac neural crest cells will migrate into later to effect aorticopulmonary septation (Waldo et al., 2005; Ward et al., 2005). The presumptive smooth muscle from the secondary heart field, which is added after the myocardium, does not move into the arterial pole in a spiraling pattern, but rather “drops” into place as the outflow

445

Chapter  |  7.2  Role of Cardiac Neural Crest in the Development of the Caudal Pharyngeal Arches

(A)

(B)

AS

RA

LA

T RPA

PT

AVC

C

LPA

Ao

CNC-derived vascular smooth muscle

Second seam

SHF-derived vascular smooth muscle

Primitive ventricle

SHF-derived myocardium

First seam

Figure 4  (A) The components of the arterial pole: conus (C), truncus (T) and aortic sac (AS) defined as proximal, middle and distal portions that undergo septation. (B) The transitions from myocardium (green) to smooth muscle (red) at the semilunar valves and more distally from smooth muscle derived from splanchnic mesoderm to smooth muscle derived from cardiac neural crest.

Heart fields

AS C AP

RV

RA

LV

C

A

LA RV

Stage 5

AP

T

Stage 14

Stages 14–18

LV

Stage 26

Figure 5  The myocardium and smooth muscle of the arterial pole arise from splanchnic mesoderm in the ventral pharynx that originates in the medial heart field.

tract moves caudally along the pharynx (Waldo et al., 2005). Therefore, the right side of the secondary heart field provides myocardium to the base of the pulmonary side of the outflow tract, but smooth muscle to the base of the aorta. Ablation of the right side of the secondary heart field results in overriding aorta with pulmonary stenosis or atresia with abnormal patterning of the main coronary arteries (Ward et al., 2005). Smooth muscle cells derived from the neural crest continue the smooth muscle tunics of the aorta and pulmonary trunk distal to the smooth muscle derived from the secondary heart field. Thus there are two “seams” at the arterial pole, i.e., the myocardial-tosmooth muscle junction derived from the secondary heart field at the base of the arterial trunks, and the junction of

the secondary heart field-derived smooth muscle with cardiac neural crest-derived smooth muscle (Fig. 4). A similar contribution of non-neural crest-derived smooth muscle cells can be seen in mouse with lineage marked neural crest cells indicating that the same junctions form in the mouse (Fig. 5) (Pietri et al., 2003).

IV.  Cardiac neural crest ablation model Much that we know about the roles of cardiac neural crest in the development of the caudal pharynx and heart came from studying the chick neural crest ablation model

446

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

Heart

Cardiac neural crest ablation

Great arteries

Persistent truncus arteriosus

Interrupted aorta

Double aortic arch

Variable absence of the carotid arteries

Left aortic arch

Alignment

Veins

Double-outlet right ventricle

Normal

Myocardial function

EC coupling Contractility L-type Ca++ Current

Non-Cardiovascular

Thyroid, Parathyroids

Thymus

Figure 6  Components of the cardiac neural crest ablation phenotype. Abnormal myocardial function and abnormal patterning of the great arteries occur in 100% of embryos after cardiac neural crest ablation. Ninety percent of embryos have persistent truncus arteriosus, with 10% showing arterial pole misalignment defects such as double-outlet right ventricle.

(Fig. 6). Ablation of the cardiac neural crest (midotic placode to somite 3) prior to migration has been variously performed using vibrating needle, tungsten needle and laser, and the phenotypes produced are identical. The morphological and functional phenotype of the cardiac neural crest ablation model has been characterized extensively since the first ablation studies published in 1983 (Kirby et al., 1983). The cardiac neural crest ablation phenotype includes defective outflow septation, abnormal patterning of the aortic arch arteries and great arteries, hypoplasia/aplasia of the thymus, thyroid and parathyroids, abnormal heart tube looping leading to arterial pole malalignment defects, and abnormal myocardial function resulting in decreased excitation-contraction (EC) coupling, contractility and

L-type Ca2 current. These defects can be broadly divided into defects which are directly related to the absence of structural contribution of the cardiac neural crest, and those which are indirectly caused by absence of the cardiac neural crest cells in the caudal pharynx resulting in abnormal signaling interactions in the pharynx.

IV.A.  Direct Defects: Outflow Septation, Aortic Arch Arteries and Pharyngeal Glands Nearly all cardiac neural crest-ablated embryos show a failure of cardiac outflow septation resulting in persistent truncus arteriosus (PTA), abnormal patterning of the great

Chapter  |  7.2  Role of Cardiac Neural Crest in the Development of the Caudal Pharyngeal Arches

arteries and variable atresia or hypoplasia of the pharyngeal glands (thyroid, thymus and parathyroid glands) (Fig. 6) (Kirby and Bockman, 1984; Kirby, 1987; Nishibatake et al., 1987; Kirby, 1988; Kirby and Waldo, 1990, 1995; Kirby and Creazzo, 1995). PTA and thymic hypoplasia are part of the spectrum in several clinical phenotypes (Kirby and Bockman, 1984; Van Mierop and Kutsche, 1986). The neural crest ablation phenotype is partially mitigated by cells from the nodose placodes that appear to follow the cardiac neural crest pathways and populate their final destinations, but are not able to perform the same functions as cardiac neural crest cells (Kirby, 1988a, 1988b, 1989). The phenotype is not mitigated by reconstitution from any other population within the post-otic neural tube, even though some of the preotic neural tube can reconstitute itself from the remaining neural tube (Scherson et al., 1993; Sechrist et al., 1995; Suzuki and Kirby, 1997). The phenotypic defects observed after neural crest ablation are remarkably consistent and reproducible. Heart defects are observed in 90% of neural crest-ablated embryos surviving to days 8–11 (Nishibatake et al., 1987). The most prevalent defects include PTA, double-outlet right ventricle (DORV) and ventricular septal defect (VSD). In PTA, the outflow septum does not form, resulting in a persisting single common outflow with a concomitant VSD. Until recently, double-outlet right ventricle, an outflow alignment defect, was less well-understood and is now considered to be indirectly related to the absence of the cardiac neural crest (see below). Virtually every cardiac neural crest-ablated embryo has abnormal patterning of the great arteries derived from the aortic arch arteries after neural crest ablation (Bockman et al., 1987, 1989; Tomita et al., 1991; Manner et al., 1996). Without the presence of neural crest-derived mesenchyme in the pharyngeal arches to support the development of the aortic arch arteries, the persistence of these vessels is unpredictable and variable, with the patterning in each embryo being unique. Most of the patterns include regression of an artery that should persist, although in some abnormal persistence of arch 2 arteries appears to have substituted for one of the more caudal arteries that are missing (Bockman et al., 1987; Rosenquist et al., 1989). The persistence of these vessels may be hemodynamically favored. Since every embryo is unique, it is not possible to provide quantitative information regarding the patterns other than that patterning anomalies are present in 100% of the embryos following cardiac neural crest ablation. In this regard, aortic arch anomalies are more characteristic of neural crest ablation than any other portion of the phenotype. Pharyngeal gland defects are also observed in nearly all of the cardiac neural crest-ablated embryos. It is thought that a threshold amount of neural crest cells with mesenchymal potential must reach the caudal arches for normal induction of the pharyngeal endoderm for gland development in the case of the thymus and parathyroid glands,

447

and for support of stromal development in the case of the thyroid gland (Kirby and Bockman, 1984; Kuratani and Bockman, 1990).

IV.B.  Indirect Defects: Role for Cardiac Neural Crest Cells to Modulate Signaling in the Caudal Pharynx One of the earliest defects observed after neural crest ablation is abnormal cardiac looping which can be seen several days before the cardiac neural crest cells should, under normal circumstances, migrate into the outflow. Defective looping is caused by failure of addition of myocardium from the secondary heart field mesoderm (Yelbuz et al., 2002). In neural crest-ablated embryos secondary heart field cell proliferation is elevated, suggesting that the presence of the cardiac neural crest cells is required for normal secondary heart development (Waldo et al., 2005b). Interestingly, even though this splanchnic mesoderm generates both myocardium and smooth muscle to the arterial pole, only the myocardium appears to be affected by neural crest ablation. The abnormal looping seen after neural crest ablation has been associated with arterial pole malalignment defects such as overriding aorta and doubleoutlet right ventricle (Yelbuz et al., 2002; Ward et al., 2005; Hutson et al., 2006). It was originally reported that neural crest ablation led to a phenotype resembling tetralogy of Fallot; however, pulmonary stenosis and atresia have never been seen after neural crest ablation, although dextroposed aorta and double-outlet right ventricle are seen on occasion (Tomita et al., 1991). It appears that pulmonary stenosis and atresia are caused by direct injury to the secondary heart field, but not the indirect injury induced by cardiac neural crest ablation (Ward et al., 2005). Because abnormal proliferation in the secondary heart field associated with abnormal heart looping occur prior to the time when neural crest cells should arrive at the outflow tract, it follows that there must be a factor normally regulated by neural crest cells in the pharynx that is dysregulated after neural crest ablation. FGF8 may be that factor, as FGF8 signaling is elevated in the pharynx of neural crest-ablated embryos during the time when the myocardium should be added from the secondary heart field (Hutson et al., 2006). If the excess FGF8 in the pharynx is neutralized in neural crest-ablated embryos, normal development of the secondary heart field is restored, leading to normal looping and normal outflow alignment. Since the neural crest cells that should populate the outflow tract are not restored, outflow septation still does not occur. Myocardial function is also compromised after cardiac neural crest ablation. Myocardial dysfunction is first observed in neural crest-ablated embryos about the time neural crest cells should migrate into the caudal pharyngeal arches, several days prior to the normal arrival of the

448

neural crest cells into the outflow tract (Leatherbury et al., 1990). The myocardium shows depressed ejection fraction with decreased calcium transient, L-type calcium current, excitation-contraction coupling and calcium sensitivity of the contractile apparatus. The embryos that survive do so by ventricular dilation to maintain cardiac output in the normal range (Leatherbury et al., 1991; Creazzo et al., 1998). Interestingly, abnormally high FGF signaling also appears to cause myocardial dysfunction because the myocardial calcium transient can be normalized when FGF signaling is reduced in hearts from neural crest-ablated embryos (Farrell et al., 2001; Hutson et al., 2006).

IV.C.  Arterial Pole Alignment versus Septation Alignment and septation of the arterial pole were thought for many years to be interdependent processes. However, it is apparent from the foregoing discussion that outflow alignment and outflow septation are related to separate cell populations and events during development. Outflow alignment is dependent on lengthening of the cardiac tube allowing the inflow and outflow poles to converge with enough length in the loop to allow the outflow tract to rotate and settle correctly. Further, the addition of the spiraling myocardium is also required for proper alignment. Other factors are important in this alignment, such as the foci of cell death that are involved in remodeling the shortening outflow myocardium (Schaefer et al., 2004). On the other hand, outflow septation is dependent on ingression of cardiac neural crest cells that form the aorticopulmonary septation complex and divide the arterial trunks down to the semilunar valve level of the outflow tract. Several studies have shown that septation and alignment of arterial pole can be uncoupled. Knockdown of HIRA, a gene thought to be involved in chromatin remodeling, leads to PTA but with correct alignment of the arterial pole (Farrell et al., 1999). As discussed above, the PTA in cardiac neural crest-ablated embryos also overrides the right ventricle. Reducing FGF signaling in the embryos results in a persistent truncus arteriosus with correct alignment; thus, the secondary heart field and cardiac neural crest are discrete cell populations with distinctly different roles in arterial pole formation and both are integral components of the pharynx. Any perturbation of development in the pharynx has the potential to affect both populations.

V.  Cardiac neural crest cells at the venous pole are associated with the conduction system Most of our knowledge of the distribution of cardiac neural crest cells in the heart comes from quail-chick chimeras.

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

More recently, adenoviral and retroviral expression vectors have been used to trace the cells, but because of the limited ability to confine these vectors to precise locations and the abnormal cell death associated with them, the most reliable information is still from chimeras. Neural crest cells in chimeras rarely undergo cell death and are seen in the mature heart (Waldo et al., 1994). Retrovirally marked neural crest cells mostly die before the cardiovascular system is mature. The reason for this death is not known, but during the integration process, retroviruses can cause DNA damage that might ultimately result in cell death (Daniel et al., 2003). In quail-chick chimeras, the cardiac neural crest cells have been traced to both the outflow and inflow poles. Retroviral studies have shown a late population of emigrating neural crest cells migrate to the inflow pole (Poelmann et al., 1998). These cells also die and their role in cardiovascular development remains to be elucidated. Retrovirallylabeled cells are found around the posterior walls of the atria, in the atrioventricular valve leaflets and near sites where the conduction system will form (see Chapter 2.3, Vol. I). The death of these cells was thought to be important for induction of the conduction system myocardium (Poelmann et al., 2004). In quail-chick chimeras, the cardiac neural crest cells reach the inflow pole and differentiate into cardiac ganglia. These cells do not enter into the inflow pole and do not die. Interestingly, after neural crest ablation the conduction myocardium differentiates normally, but does not function normally (Gurjarpadhye et al., 2007). The atrioventricular conduction is abnormal, not because of a failure of conduction tissue differentiation, but because of a failure in connective tissue insulation of these tissues after cardiac neural crest ablation. Thus, cardiac crest may have a role in insulation of the tissues or induction of the cells that insulate the central conduction system.

VI.  Factors important in cardiac neural crest induction and function The neural crest ablation phenotype has a number of genetically-based mimics in mice that have been useful in beginning to understand the molecular nature of the various disturbances that can lead to this phenotype. Understanding neural crest cell behavior entails understanding transcriptional regulation in the cells themselves and of signals in their environment. Neural crest cells respond to a variety of factors during their induction, migration and population of various target sites. Changes in the response of neural crest cells to the same signal over time, or their sensitivity to different signals may be mediated, in part, by an everchanging combination of transcription factors expressed within these cells. In addition, changes in transcription factor expression in the cells within the neural crest cell environment also determine which signals are present.

Chapter  |  7.2  Role of Cardiac Neural Crest in the Development of the Caudal Pharyngeal Arches

VI.A.  Transcription Factors VI.A.i.  Tbx1 Tbx1 is a member of the T-box family of transcription factors (see Chapter 1.4, Vol. II) that lies within the deleted region of chromosome 22q11, which is most commonly deleted in patients with the DiGeorge phenotype (Jerome and Papaioannou, 2001). Tbx1 is not expressed by cardiac neural crest cells as they migrate through the pharyngeal arches, thus any effect of Tbx1 on neural crest development is most likely indirect. Tbx1 has been shown to be haploinsufficient in several patients with conotruncal and arch artery defects (Yagi et al., 2003). In mouse, Tbx1 homozygous mutation recapitulates the cardiovascular and glandular defects common in DiGeorge syndrome, but mice haploinsufficient for Tbx1 have abnormal aortic arch patterning with normal outflow septation. One of the ways Tbx1 appears to affect conotruncal development is by supporting proliferation of cells in the secondary heart field (Xu et al., 2004).

VI.A.ii.  Pax3 Pax3 is a member of the paired family of transcription factors that contain two DNA-binding domains referred to as the paired box and the homeobox. Pax3 is expressed by all of the cells in the dorsal neural tube, including neural crest cells. In its heterozygous state the gene has been associated with Waardenburg’s syndrome in humans, which has a mild phenotype consisting of deafness and pigmentary deficiencies (Chalepakis et al., 1994; Conway et al., 1997). The absence of one allele of Pax3 is not associated with cardiovascular defects in either humans or mice (Tassabehji et al., 1992; Chalepakis et al., 1994; Tassabehji et al., 1994; Conway et al., 1997). However, in the homozygous state in mice, this mutation results in embryonic lethality at around 14 days of gestation (Chalepakis et al., 1994; Conway et al., 1997). The Splotch mutant has an ENU-induced functionally null allele of Pax3, and has been studied extensively. Homozygous mice have a complete cardiac neural crest ablation phenotype: persistent truncus arteriosus; aortic arch anomalies; and hypoplasia or aplasia of the thymus, parathyroid and thyroid glands (Franz, 1989; Chalepakis et al., 1994; Epstein, 1996; Conway et al., 1997). Pax3 is expressed in the dorsal neural tube in the region that gives rise to the emigrating neural crest cells, and Cre-lox technology has been used to fate map Pax3-expressing neural crest cells. The embryos have more global neural tube defects in addition to the cardiac neural crest ablation phenotype which makes this mutant an imperfect mimic of cardiac neural crest ablation. The cardiovascular phenotype has been associated with a deficiency of neural crest-derived cells traversing the pharyngeal arches and migrating into the cardiac outflow tract. This suggests that Pax3 is not necessary for migration of the cardiac neural crest, but may play a role in expansion of the neural crest

449

cell population (Chalepakis et al., 1994; Conway et al., 1997; Epstein et al., 2000; Chan et al., 2004).

VI.A.iii.  Hoxa3 Hoxa3, a homeobox-containing transcription factor, is expressed by cardiac neural crest cells, but also by pharyngeal endoderm. Null mutation of the Hoxa3 gene results in death during the first few hours after birth. These mice have hypoplastic and/or aplastic thymus, parathyroid and thyroid development, and reduced submaxillary glandular tissue even though normal numbers of neural crest cells arrive in the pharyngeal arches (Manley and Capecchi, 1995). Hoxa3 is expressed in the pharyngeal endoderm, as well as in the neural crest, although it is not known if expression is necessary in one or both cell types for glandular development to occur. The Hoxa3 mutant is interesting, in that gland abnormalities partially mimic the neural crest ablation phenotype but no heart malformations have been found. These animals are unable to inflate their lungs, which results in persistence of the fetal pulmonary-to-systemic vascular shunts. Without respiration the ductus arteriosus remains as a large patent channel connecting the left pulmonary artery to the thoracic aorta (Ruano and Kidd, 1991). The third pharyngeal arch and artery do not develop normally in the Hoxa3 mutant. Treatment of chick premigratory cardiac neural crest with antisense oligonucleotides to the Hoxa3 paralogous group followed by reimplantation leads to regression of the third arch artery. In this experimental paradigm, neither the treatment nor the patterning abnormality is associated with cardiac outflow defects. This, and the lack of predictable arch artery patterning in the ablation phenotype, suggests that neural crest cells carry instructive information regarding aortic arch remodeling to form the great arteries, but that the same patterning instructions are not necessary for cardiac outflow septation (Kirby et al., 1997).

VI.A.iv.  AP2 The transcription factor AP2 (Tfap2) has an important role in regulating gene expression in both epidermis and neural crest cells (Luo et al., 2005). Tfap2 is positively regulated by bone morphogenetic protein (BMP) and Wnt signaling, both of which are essential for neural crest specification and migration. Ectopic expression of Tfap2 activates expression of Slug and Sox9, which are also neural crest-specific transcription factors. Activation of Tfap2 requires some attenuation of endogenous BMP signaling. Loss-of-function of Tfap results in severe reduction in the neural crest (Luo et al., 2003). Several of the defects associated with the Tfap2-null mutation affect neural crest cell derivatives, including the craniofacial skeleton, cranial ganglia and outflow tract. Tfap expression is utilized as a marker for premigratory and migratory neural crest cells in

450

many vertebrate species. Neural crest-specific disruption of Tfap2 results in perinatal lethality associated with neural tube closure defects and cleft secondary palate. Some of the mutant mice can survive into adulthood, but have retarded craniofacial growth, abnormal middle ear development and defects in pigmentation. However, absence of Tfap2 expression does not result in a cardiac neural crest ablation phenotype (Brewer et al., 2004).

VI.A.v.  Fox Genes Foxc genes are expressed in the secondary heart field and in cardiac neural crest cells, endocardium and proepicardium. Foxc1 and Foxc2 are required for head mesenchyme and aortic arch formation (Iida et al., 1997; Winnier et al., 1999; Kume et al., 2001). Foxc1:Foxc2 compound heterozygotes have a wide spectrum of cardiac abnormalities, including hypoplasia or lack of the outflow tract and right ventricle, inflow tract, dysplasia of the outflow and atrioventricular cushions, and abnormal formation of the epicardium (Kume et al., 2001). Compound mutants show significant downregulation of Tbx1 and Fgf8 and 10, and a reduction in cell proliferation. Neural crest cells in compound mutants have extensive cell death during migration, which is thought to be the cause of persistent truncus arteriosus (Seo and Kume, 2006). Foxd3 is specifically required for neural crest specification, migration and survival (Stewart et al., 2006). However, because Foxd3 is also required in trophoblast development and tissue-specific deletions have not been made, it is not known what effect the gene may have on heart development. Not all genes important for outflow septation are expressed by the cardiac neural crest cells. Foxp1 is expressed in the myocardium and endocardium, and in Foxp1-mutant embryos the outflow cushions fail to fuse (Wang et al., 2004).

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

tract (Kwang et al., 2002; Ishii et al., 2005). However, Msx1;Msx2 double homozygous null mice have conotruncal abnormalities, including double-outlet right ventricle, tetralogy of Fallot and PTA, due to increased neural crest cell death in the pharyngeal arches (Ishii et al., 2005).

VI.A.vii.  Slug/Snail Slug and Snail belong to a family of transcriptional repressors. Both are expressed in neural crest-forming regions of the neural plate and are dependent on Wnt signaling. These genes are thought to be involved in the epithelial– mesenchymal transformation that occurs at the onset of neural crest migration (Nieto et al., 1994; Carl et al., 1999; LaBonne and Bronner-Fraser, 2000). Overexpression of Slug or Snail leads to expanded neural crest cells (LaBonne and Bronner-Fraser, 2000; Del Barrio and Nieto, 2002; Aybar et al., 2003). Homozygous deletions in Slug have been identified in human patients with Waardenburg syndrome (Sanchez-Martin et al., 2002).

VI.A.viii.  Sox 4 Sox transcription factors contain an Sry DNA-binding domain. Several members of the Sox family of transcription factors are expressed by neural crest cells and appear to be needed for their induction (Bell et al., 2000; Sock et al., 2001; Spokony et al., 2002; Aoki et al., 2003; Honore et al., 2003). However, cardiovascular defects have not been reported in embryos with altered expression of these factors. On the other hand, Sox4-null mutation is associated with PTA (Schilham et al., 1996). Interestingly, Sox4 is not expressed by neural crest cells, but in the non-neural crestderived mesenchyme of the outflow cushions, and it thus serves as another example of a gene that orchestrates the neural crest environment needed for outflow septation.

VI.A.vi.  Msx Genes Msx1 and Msx2 transcription factors are expressed in premigratory and migratory neural crest, as well as in the neural crest-derived mesenchyme of the pharyngeal arches (Davidson, 1995). Msx1 is highly-expressed at the border of the neural plate and may be important for induction of downstream genes that induce neural crest formation (Tribulo et al., 2003). It is expressed in response to BMP and Wnt signaling, and allows induction of neural crest markers (Bang et al., 1999). Msx2 has been shown to be a downstream effector of Pax3. Msx2 is upregulated in the Splotch mutant, and a loss-of-function Msx2 mutation rescues the cardiac defects of the Splotch mutant embryos, as well as defects in the dorsal root ganglia, thymus and thyroid (Kwang et al., 2002). Misexpression of Msx1 increases expression of Slug and FoxD3 (Tribulo et al., 2003). Msx1 or Msx2 individual mutant mice do not exhibit defects in the development of the cardiac outflow

VI.B.  Signaling Factors VI.B.i.  Platelet-Derived Growth Factor (PDGF) PDGF belongs to a family of growth factors that is important for proliferation, migration and angiogenesis. The Patch-mutant mouse has a phenotype that in some respects mimics the neural crest ablation model. The Patch mutation is a deletion that encompasses part of the PDGF receptor gene, as well as the locus control region for the c-kit gene which is also in the PDGF family and encodes the receptor for Steel factor (Orr-Urtreger et al., 1992; Wehrle-Haller et al., 1996). Because PDGF receptor and c-kit are expressed by neural crest cells, the Patch phenotype cannot be ascribed to a single gene. However, the Patch phenotype is largely replicated in the PDGF receptornull mouse. Mice homozygous for null mutation of the PDGF receptor show increased cell death along the

Chapter  |  7.2  Role of Cardiac Neural Crest in the Development of the Caudal Pharyngeal Arches

neural crest migration pathways (Orr-Urtreger et al., 1992; Wehrle-Haller et al., 1996; Soriano, 1997). Steel factor is a known survival and proliferative factor for melanocytes (Raid et al., 1996). PDGF and c-kit are expressed in neural crest cells from the onset of migration, with expression continuing during migration (Besmer et al., 1993). In Xenopus, PDGF receptor is also expressed by cranial neural crest cells prior to migration and during their migration into the pharyngeal arches. The Patch mutant and PDGFnull homozygous mice have a number of cardiovascular anomalies that include PTA and abnormal patterning of the persisting aortic arch arteries.

VI.B.ii.  Fibroblast Growth Factor (FGF) FGF signaling is important from the earliest stages of cardiac neural crest development in the neural tube and later as the cells migrate through the pharynx. Mice hypomorphic for Fgf8 display abnormal apoptosis of cardiac neural crest cells, the typical neural crest ablation cardiovascular phenotype including PTA, arterial pole misalignment, and arch artery patterning defects, suggesting that FGF8 is an important survival factor for the neural crest (AbuIssa et al., 2002; Frank et al., 2002; Brown et al., 2004). Conversely, neural crest-ablated embryos have excessive FGF8 signaling in the caudal pharynx, which is detrimental to normal secondary heart field development (Hutson et al., 2006). Arterial pole alignment defects, but not PTA, are observed in chick embryos when FGF8 levels are reduced to below-normal levels during secondary heart field migration (Hutson et al., 2006). These studies suggest that the secondary heart field is particularly sensitive to FGF8 signaling levels. In addition to levels of FGF8, the source of the FGF8 signal may also be important. It should be noted that FGF8 is produced by the pharyngeal ectoderm and the endoderm, and not by the neural crest cells. Tissue-specific ablation of Fgf8 in the pharyngeal arch ectoderm results in arch artery patterning defects which may involve abnormal neural crest-related arch patterning, but no arterial pole defects (Macatee et al., 2003).

VI.B.iii.  Bone Morphogenetic Protein (BMP) BMP2 and BMP4 are thought to regulate neural crest cell induction, maintenance, migration, differentiation and survival. The BMP receptor Bmpr1a is expressed in the neural tube sufficiently early to be involved in neural crest specification and/or migration. Mice in which Bmpr1a has been ablated from the neural crest display a shortened cardiac outflow tract and defective septation. These mice also show reduced myocardial proliferation and die at midgestation due to heart failure. Surprisingly, the mice do not show abnormal induction, delamination or initial migration of the neural crest cells (Stottmann et al., 2004). Deletion of Alk4, another BMP receptor, in the cardiac neural crest also

451

results in arch artery patterning defects and PTA (Kaartinen et al., 2004). In these mice not enough neural crest cells reach the outflow tract to effect normal septation.

VI.B.iv.  Semaphorin3C Semaphorin (Sema) 3C is a member of a family of secreted ligands used in axon guidance. It is also important in the migration and targeting of cardiac neural crest cells to the outflow tract. Sema3C-null mice have interrupted aortic arch and PTA (Feiner et al., 2001). Because other neural crest derivatives are normal in Sema3C-null embryos, Sema3C signaling appears to be particularly important for cardiac neural crest. However, Sema3C is not expressed by cardiac crest cells, but is expressed in outflow myocardium. The neural crest cells express the receptors for semaphorin ligands which are multimeric complexes of plexins and neuropilin (Np)1 and/or Np2 (Kawasaki et al., 1999). A complex of Np1 and plexinA2 may be the functional receptor that signals the cardiac neural crest, as either plexinA2 or Np1-null mice have PTA and interrupted aortic arch (Kawasaki et al., 1999; Brown et al., 2001). This suggests that cardiac neural crest cells use guidance cues that target them to the cardiac outflow tract. Interestingly, GATA6, a transcription factor expressed by vascular smooth muscle, has been shown to regulate Sema3C expression in the outflow tract and vascular smooth muscle. Mice with targeted deletion of GATA6 in the vascular smooth muscle have interrupted aortic arch and PTA (Lepore et al., 2006).

VI.B.v.  Endothelin (ET) The endothelin signaling family consists of three known ligands and two receptors. The endothelin (ET) ligands 1–3 (ET1, ET2 and ET3) effect cellular responses via the endothelin-A (ETA) or endothelin-B (ETB) receptors. ET1 has highest affinity for the ETA and ET3 response is mediated by ETB. ETB is expressed by neural crest cells before and during migration at all levels of the neural axis (Nataf et al., 1996). In culture, ET3 stimulation causes proliferation in the pleuripotent neural crest cells and melanocyte progenitors. Steel factor synergizes with ET3 to promote both survival and proliferation of the melanocyte progenitors. ET3 also induces differentiation of melanoblasts into mature melanocytes (Raid et al., 1996). Mutations in either ET3 or ETB result in pigment defects (Nataf et al., 1996; Pasini et al., 1996). No cardiac defects are observed in the ET3/ETB-null mice and the phenotype appears to be limited to neural crest cells that form the distal-most enteric plexus and melanocytes. In contrast to ET3, ET1 is expressed in the endothelium of the arch arteries, the endocardial cushions, as well as the pharyngeal arch epithelium, but not by the neural crest. ET1 induces vasoconstriction and increases cardiac contractility, as well as cell proliferation. ET1 is processed from a

452

38-amino acid inactive propeptide called “big endothelin-1” by endothelin-converting enzyme-1 (ECE1) (Xu et al., 1994; Torres et al., 1997). ECE-1 is a membrane-bound metalloprotease that activates ET1 and ET3 (Xu et al., 1994; Torres et al., 1997). When ET1 is disrupted, the homozygous mutants are smaller than normal and have hypoplastic thymus and thyroid glands that are not fused in the midline (Kurihara et al., 1995). In addition, the thymus does not descend to its appropriate position (Kurihara et al., 1995a, 1995b). A small proportion of the homozygotes have aortic arch anomalies, whereas a larger percentage (50%) have ventricular septal defects and outflow tract defects that include double-outlet right ventricle and PTA, suggesting that the cardiac neural crest requires ET1 signaling to effect normal outflow septation. The percentage of homozygotes with cardiovascular anomalies can be enhanced by administering a blocking antibody or by selective ETA antagonists during embryogenesis. Thus, it is thought that circulating maternal ET1 or other endothelin isoforms may provide functional redundancy in the endothelin system (Kurihara et al., 1995). This assumption is supported by the fact that disruption of ECE1 causes a much more severe phenotype in a larger number of mutant embryos. The endothelin receptors are G-protein coupled receptors. The G-proteins G-12 and G-13, have been shown to be important in mediating endothelin signaling in the neural crest. Mice in which G12 and G13 were specifically knocked-out in the neural crest using P0 cre had double-outlet right ventricle, VSD and some coronary artery defects. Gq/G11 have been shown by another group (Ivey, 2003) to mediate ET1 signaling in the pharynx, but Dettlaff-Swiercz and colleagues (2005) did double knockouts and these mice had no heart defects. These experiments lend further evidence to the importance of neural crest in modulating signaling in the pharynx coincident with secondary heart field development.

VI.B.vi.  Retinoic Acid (RA) Retinoic acid is an endogenous signal as well as an exogenous teratogen that has been associated with defective neural crest development. Disruption of endogenous retinoic acid signaling by mutation of the Retinoid receptors or synthetic enzymes, or administration of exogenous retinoic acid, causes disrupted development resulting in phenotypes similar to neural crest ablation. Retinoic acid signaling is transduced via two families of receptors composed of nuclear proteins that act as transcriptional regulators when complexed with the retinoic acid ligand (Chen et al., 1996a). The retinoid acid receptor (RAR) family is activated by all-trans-retinoic acid and 9-cis-retinoic acid, whereas the retinoid X receptor (RXR) family is activated by 9-cis retinoic acid, as well as a variety of other nonretinoid ligands including vitamin D and thyroid hormone. There are three receptor isoforms in each family which are designated ,  and . Further variants are produced within

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

these groups of RAR receptors by differential splicing, and there is a specific spatio–temporal distribution of each isoform during embryogenesis. It has been shown in cultured cells that retinoic acid-responsive transcription is controlled by RAR-RXR heterodimers (Chen et al., 1996a; Lu et al., 1997). Null mutations that affect all the subtypes of a single isoform (RAR or RAR) show postnatal growth deficiency and mortality, but otherwise have no apparent morphological defects. However, compound mutations in RAR isoforms mimic all the abnormalities seen in vitamin A deficiency, in addition to several that are not seen. Phenotype synergy is observed when RXR mutation is introduced into the RAR or RAR mutant background. RAR;RAR double mutants also have several malformations not seen in single mutants. This and other evidence from null mutations suggests that RXR-RAR heterodimers mediate retinoid signaling (Kastner et al., 1994; Krezel et al., 1996). RXRs serve as markers for undifferentiated neural derivatives of the trunk neural crest (Rowe et al., 1991), and in the chick RXR receptor transcripts are a good marker for migrating neural crest cells. Transcripts are gradually restricted to the differentiating neural derivatives, whereas expression is lost in the ectomesenchymal derivatives by stage 15 (Rowe and Brickell, 1995). In mice, RXRnull mutations result in ocular and cardiac malformations and in utero death (Kastner et al., 1994; Krezel et al., 1996). When RAR or RAR mutations are combined, more severe ocular defects occur, and PTA and aortic defects appear (Kastner et al., 1994; Krezel et al., 1996). RAR;RAR and RAR;RAR double mutants have PTA, double-outlet right ventricle and aortic arch anomalies (Sucov et al., 1995). In addition, RXR-null embryos have thin myocardium, perhaps caused by premature differentiation of the ventricular myocardium, indicating that the RXR receptors are also important in development of the myocardium (Gruber et al., 1996). When RAR1;RAR mutant embryos were crossed with the Wnt1-cre to specifically knockout the retinoic acid receptors in neural crest cells, the number, migration and terminal fate of the cardiac neural crest was normal; however, the specific function of these cells in forming the aorticopulmonary septum was impaired (Jiang et al., 2002). Targeted inactivation of the mouse retinaldehyde dehydrogenase 2 (RALDH2/ALDH1a2), which codes for the enzyme responsible for retinoic acid synthesis, is embryonic lethal at E9.5–10 due to severe heart defects. Transient retinoic acid supplementation to the dams between E7.5 to E8.5 rescues most of these defects, but the retinoic acid-supplemented null mice have still have PTA (Niederreither et al., 2001). Furthermore, all derivatives of the caudal pharyngeal arches including the aortic arch arteries, thymus, parathyroid gland and post-otic neural crest cells are abnormal (Niederreither et al., 2003). Additionally, reduced and/or altered expression of Cyp26 enzyme function, needed for retinoic acid inactivation, also causes a dose-dependent loss of the caudal pharyngeal arches and persistent

Chapter  |  7.2  Role of Cardiac Neural Crest in the Development of the Caudal Pharyngeal Arches

truncus arteriosus (Roberts et al., 2006). This suggests that regulation of retinoic acid signaling must be tightly controlled, because too little or too much retinoic acid is detrimental to the derivatives of the neural crest and the caudal pharynx.

VI.B.vii.  Wnt1 Wnts are extracellular ligands that signal through frizzled receptors. Wnt1 expression is limited to the dorsal margins of the myelencephalon and the mesencephalon, and subsequently extends to the entire lateral margins of the neural plate where neural crest cells form (McMahon et al., 1992). Wnt1-null embryos show severe abnormalities in the development of the mesencephalon and metencephalon, but display no components of the cardiac neural crest ablation phenotype (Thomas and Capecchi, 1990). Wnt-1 and a related gene, Wnt-3a, are co-expressed from early somite stages in dorsal aspects of the myelencephalon and spinal cord, and functional redundancy between these two genes may account for the lack of a neural crest ablation phenotype (McMahon et al., 1992). Attenuation of Wnt-1 expression using antisense oligonucleotide inhibition in mouse embryos grown in culture induces mid- and hindbrain abnormalities as those seen in the Wnt-1-null mutant mice, but in addition it also is associated with cardiomegaly resulting in hemostasis consistent with the possibility that neural crest might be affected (Augustine et al., 1993). Even though its role in neural crest development is still unclear, Wnt1 is a good marker of the neural crest cell lineage, and Wnt1-cre has been one of the most popular of the neural crest-specific gene activators and inactivators.

VI.B.viii.  Transforming Growth Factor (TGF)  Tissue-specific deletion of the TGF receptor in neural crest cells shows a completely penetrant phenotype of PTA and interrupted aortic arch type B. However, although it was speculated that this null mutation would interfere with smooth muscle differentiation by the neural crest cells, it does not alter neural crest cell specification to a smooth muscle fate in the cranial or cardiac domains. Pharyngeal organ defects such as those seen in the ablation phenotype are not observed, arguing against an early perturbation of the cardiac neural crest cell lineage (Choudhary et al., 2006). Neural crest-specific deletion of the TGF-beta type I receptor, Alk5, leads to severely abnormal remodeling of pharyngeal arch arteries, abnormal aortic sac development, failure in pharyngeal organ migration and PTA. ALK5 is not required for neural crest cell migration, but appears to be important in the survival of post-migratory cardiac neural crest cells (Wang et al., 2006). Thus, TGF is an essential morphogenic signal for the neural crest cell lineage involved in outflow septation and pharyngeal arch patterning.

453

VI.C.  Gap Junctions VI.C.i.  Cx43 Cx43 is a gap-junction protein expressed during neural crest cell migration (Reaume et al., 1995; Lo et al., 1999). Gap junctions are intracellular channels that allow the passage of small signaling molecules, low molecular weight metabolites and ions between cells. Cx43-null mice die soon after birth because of pouches in the outflow of the right ventricle that obstruct blood leaving the heart. This is a defect that is not seen in the neural crest ablation phenotype. Cardiac neural crest cell migration in Cx43 knockout mice show reduced directionality and speed, while cardiac neural crest cells overexpressing Cx43 in transgenic mice show increased directionality and speed (Huang et al., 1998). Interestingly, two studies have recently shown that conditional knockout of Cx43 targeted to neural crest cells by Wnt1-cre failed to reproduce the outflow obstruction phenotype found in the germline knockout of Cx43 (Kretz et al., 2006; Liu et al., 2006). However, a broader conditional knockout mediated by Pax3-cre knocked out Cx43 expression in a larger dorsoventral extent of the neural tube and neural crest (Liu et al., 2006). This resulted in outflow obstruction defects similar to those seen in germline Cx43null embryos. The Pax3-cre mediated knockout resulted in an increased number of neuroepithelial cells leaving the dorsal and lateral neural tube and an excess of cells migrating to the outflow tract.

VII.  Human syndromes that   are likely to involve cardiac neural crest VII.A.  DiGeorge and Velocardiofacial Syndromes Van Mierop and Kusche, and Bockman and Kirby first suggested the association of DiGeorge syndrome with neural crest-derived mesenchyme in pharyngeal arches 3 and 4 (Bockman and Kirby, 1984; Van Mierop and Kutsche, 1986). Both DiGeorge syndrome and velocardiofacial syndrome (VCFS) provide phenocopies of cardiac neural crest ablation. The most distinct features of these syndromes are interrupted aortic arch type B, outflow tract malformation (PTA, tetralogy of Fallot or double-outlet right ventricle, hypoplastic thymus with some degree of immunocompromise and hypoparathyroidism). Additional features include mild craniofacial irregularities, hypothyroidism, psychosis and, in the case of VCFS, facial clefting, i.e., cleft lip and palate (Karayiorgou et al., 1995; Papolos et al., 1996). These syndromes have been linked to a microdeletion in chromosome 22q11. Other forms of these syndromes that have been reported and associated with similar microdeletions are conotruncal anomaly face syndrome (Matsuoka

454

et al., 1994) and right-sided aortic arch (Strong, 1968). The name CATCH (Cardiac defects, Abnormal facies, Thymic hypoplasia, Cleft palate and Hypocalcemia)-22 has been proposed for syndromes with these manifestations that have the 22q11 microdeletion (Wilson et al., 1993). The search for the gene or genes underlying these physical traits has been intense, and several candidates have emerged. Identification of the chromosome 22 microdeletion led to the identification of Tbx1 as the gene that is most likely to underlie the phenotype, although it is generally recognized that the multiple forms of the phenotype are likely to involve other genetic loci and modifier genes (see below). Tbx1 is not expressed by cardiac neural crest cells while they are migrating, or when they populate the pharyngeal arches, but instead it is expressed in the pharyngeal ectoderm, endoderm and in the secondary heart field mesenchyme (Xu et al., 2004). Tbx1 expression in the endoderm may control Fgf8 expression, which affects some aspect of neural crest development, perhaps proliferation (Garg et al., 2001). One of the more remarkable properties of Tbx1 is that different structures have different sensitivity to Tbx1 dose. Recently, Baldini and colleagues showed mesoderm-specific deletion of Tbx1 causes severe pharyngeal patterning and cardiovascular defects (Zhang et al., 2006). When mesoderm-specific Tbx1 expression was restored in a mutant background, the outflow morphology was rescued but not the arch artery patterning. While the 22q11 microdeletion is detected in many patients with diagnosed DiGeorge syndrome, not all patients with the DiGeorge phenotype have the microdeletion and/or a mutation in Tbx1. Furthermore, not all DiGeorge patients with the deletion have the same severity of malformations, again suggesting that there are other genetic or epigenetic modifiers. Many studies have revealed interactions between Tbx1 and FGF (Vitelli et al., 2002; Hu et al., 2004; Xu et al., 2004), Pitx2 (Nowotschin et al., 2006), hedgehog (Yamagishi et al., 2003), retinoic acid (Roberts et al., 2005; Guris et al., 2006) and vascular endothelial growth factor (Stalmans et al., 2003) signaling. Crkl, an adaptor protein important for intercellular signaling, is also located within the 22 microdeleted region and homozygous mice null for Crkl have some but not all of the DiGeorge features (Guris et al., 2002). However, Crkl;FGF8 compound heterozygous mice have increased incidence and severity of cardiovascular defects (Moon et al., 2006). Interestingly, Cyp26a1, a gene required for retinoic acid-inactivation, is a target of Tbx1 (Roberts et al., 2006). Cyp26 is reduced in Tbx1-null embryos raising the level of retinoic acid signaling. HIRA (originally called Tuple1) is another gene located in the 22q11 deleted region that has an embryonic expression pattern consistent with the spatial and temporal genesis of the defects. Tuple1 was renamed Hira because of its similarity with Hir1 and Hir2, yeast genes that encode repressors of histone gene transcription (Lamour et al., 1995; Scamps et al., 1996). Hira is highly conserved

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

in mouse, chick and human. In both chick and mouse embryos, it is expressed in the neural plate, neural tube, neural crest and mesenchyme of the head and pharyngeal arches (Halford et al., 1993; Roberts et al., 1997). Hira on the nondeleted chromosome 22 in DiGeorge patients shows delayed replication timing, with an increased ratio of cells with the Hira locus placed toward the nuclear periphery. Replication timing and nuclear location are generally correlated with transcription activity of the relative DNA region. Thus the nondeleted Hira gene may have altered transcription in DiGeorge patients which could alter timing and expression of critical genes (D’Antoni et al., 2004). While the advances in understanding this neural crest ablation-type phenotype linked to a microdeletion of chromosome 22q11 are very exciting, it is important to keep in mind that phenocopies of these syndromes are linked to microdeletions on two other chromosomes. A fetus with a 17q13 chromosomal deletion was found to have multiple DiGeorge syndrome-like anomalies, including thymic hypoplasia and double-outlet right ventricle. In addition, the fetus showed polyhydramnios and intrauterine growth retardation (Greenberg et al., 1988). Terminal deletions of chromosome 10 resulting in the loss of p13; p14 are associated with hypoparathyroidism and other manifestations of DiGeorge syndrome and VCFS (Daw et al., 1996).

VII.B.  CHARGE Syndrome The CHARGE phenotype includes coloboma, heart disease, atresia of choanae, retardation of physical and mental development, genital hypoplasia and ear anomalies and/or deafness (Siebert et al., 1985). This constellation of defects appears to have an identifiable DiGeorge/VCFS basis, with multiple other anomalies. Thyroid and parathyroid glands are frequently absent and accompany outflow anomalies and aortic arch artery malformations. Malformation of the foregut, reproductive organs, kidneys, limbs and digits, and brain, including pituitary gland, with lung abscesses and focal hepatic necrosis suggests an etiology from an earlier developmental stage or involving a gene that has a broader expression pattern than those involved in DiGeorge/VCFS. Recently, microdeletions on chromosome 8q12 were linked to two individuals with CHARGE syndrome. Sequence analysis of individuals with CHARGE syndrome, without microdeletions in this chromosomal region, detected mutations in the gene CHD7, a member of the chromodomain helicase DNA-binding gene family accounting for the disease in 70% of affected individuals (Aramaki et al., 2006; Vissers et al., 2006). CHD proteins affect chromatin structure and gene expression and are ubiquitously expressed during early embryonic development, helping to explain the wide range of congenital defects observed in CHARGE syndrome.

Chapter  |  7.2  Role of Cardiac Neural Crest in the Development of the Caudal Pharyngeal Arches

VII.C.  Fetal Alcohol Syndrome (FAS) Alcohol exposure during the time when neural crest cells are populating the frontonasal process and caudal pharyngeal arches can cause a phenotype similar to that seen in DiGeorge syndrome, and has been proposed as an environmental causative agent in DiGeorge syndrome (Sulik et al., 1986). Epidemiological studies suggest that women consuming alcoholic beverages more than once weekly during the periconceptional period have a 2- to 2.5-fold increase in the risk of having offspring with conotruncal heart defects (Carmichael et al., 2003). Ethanol can exert a teratogenic effect by disrupting microtubules and microfilaments that would interfere with cell migration (Hassler and Moran, 1986), decreasing mitochondrial respiration (Nyquist-Battie and Freter, 1988), or causing excessive cell death in selected cell populations, including craniofacial neural crest, which might be caused by heightened membrane fluidity (Chen et al., 1996b). Interestingly, there is growing evidence that ethanol exposure is antagonistic to retinol (Vitamin A) and retinal conversion to retinoic acid, suggesting that FAS is induced in part by an ethanoldependent reduction in retinoic acid levels that are necessary for normal development (Yelin et al., 2005). Further, ethanol-induced defects in cultured chick embryos can be ameliorated by treatment with all-trans-retinoic acid, indicating retinoic acid signaling is reduced in FAS (SatirogluTufan and Tufan, 2004).

VII.D.  Retinoic Acid (RA) Embryopathy Interestingly, the retinoic acid system (see Chapter 3.3, Vol. I) is one of the few in which either absent or excess signaling is associated with similar phenotypes in humans and other animal models. While the defects that occur in retinoid embryopathy are not confined to the cardiovascular system, heart and pharyngeal arch development are widely studied in animal models that employ retinoic acid. Cardiac and aortic arch anomalies in offspring of rats with retinoic acid deficiency or excess are directly correlated with the anomalies in humans (Wilson and Warkany, 1950; Rothman et al., 1995). The lower jaw, palate, limbs, vertebrae and tail are consistently abnormal after retinoic acid treatment. In cultured mouse embryos, retinoic acid causes a reduction in the size of arches 1 and 2 (Goulding and Pratt, 1986). Retinoic acid exposure of mice causes complete transposition of the great arteries, which appears to be associated with hypoplasia of the conal but not the atrioventricular cushions (Yasui et al., 1995; Nakajima et al., 1996). In the chick, retinoic acid exposure causes a range of cardiac outflow defects from overriding aorta to double-outlet right ventricle (Broekhuizen et al., 1992; Bouman et al., 1995). However, transposition is not seen in the chick. While the mechanism of retinoic acid teratogenicity is not known, several clues have been identified. The head

455

and hindbrain are especially sensitive to retinoid exposure, and some members of the hox gene family (i.e., Hoxa1 and -b1 genes) contain retinoid-responsive elements (Langston et al., 1997). Thus, retinoic acid treatment appears to alter the segmental expression of the hox gene code which in turn causes transformation of pharyngeal segmental identity, the hindbrain and otocyst are shifted cranially relative to foregut, and the preotic neural crest is retarded in migration (Marshall et al., 1992). Retinoic acid also retards neural crest cell migration in vitro, but does not affect viability or DNA synthesis (Maxwell et al., 1982). It is likely that retinoic acid alters the regional identity of cranial crest cells (Lee et al., 1995). Unresolved data show that the effect may be on neural crest cells committed to somatic mesoderm (Yasuda et al., 1986), or it may be selective to cells undergoing migration rather than affecting a particular cell lineage (Thorogood et al., 1982). There is a large accumulation of labeled retinoic acid in neural crest derivatives in the pharyngeal arches (Dencker et al., 1990) which also express cellular retinoic acid-binding protein (CRABP) (Maden et al., 1990; Vaessen et al., 1990). HNK1 expression, which is characteristic of migrating neural crest cells in avians, disappears as CRABP expression appears in the same cell population (Maden et al., 1991). Even so, the regions affected in retinoid embryopathy are not correlated with CRABPI or CRABPII, so other factors must be responsible for the teratogenic effect (Horton and Maden, 1995).

VIII.  Conclusions and future perspectives The neural crest ablation model provides a valuable framework for beginning to understand the actions of various teratogens and gene mutations in cardiac and great vessel development. The primary signs of neural crest malfunction are PTA and aortic arch anomalies, with or without anomalous development of the glands derived from the caudal pharynx. The best phenocopy of cardiac neural crest ablation is in human DiGeorge syndrome. Interestingly, the genes that have been found to underlie the DiGeorge phenotype are generally not expressed in cardiac neural crest, and so the role of neural crest in generating the phenotype is mostly secondary to the primary insult. In the same line of thinking, double-outlet right ventricle is only indirectly a component of the neural crest ablation phenotype. We have only realized recently the importance of the neural crest to modulate signaling in the caudal pharynx, and this signaling is critical for development of secondary heart field-derived outflow myocardium during cardiac looping. Failure of the addition of secondary heart field myocardium to the lengthening tube leads to overriding aorta/double-outlet right ventricle-type misalignment defects.

456

Overriding aorta is a component of tetralogy of Fallot, but this phenotype also has pulmonary stenosis or atresia as a major component. Pulmonary stenosis and atresia are caused by abnormal development of the secondary heart field and have not been seen after neural crest ablation, even though we reported it as part of the neural crest ablation phenotype thinking at the time that pulmonary stenosis must be something specific to human development. Now that we are able to ascribe particular developmental processes to specific outflow malformations, the next task will be to understand the mechanisms by which neural crest cells modulate signaling in the caudal pharynx, and the tissue responses of the cells in the secondary heart field to the signals that control their translocation into the outflow tract and differentiation as myocardium.

References Abu-Issa, R., Smyth, G., Smoak, I., Yamamura, K-I., Meyers, E., 2002. Fgf8 is required for pharyngeal arch and cardiovascular development in the mouse. Development 129, 4613–4625. Aramaki, M., Udaka, T., Kosaki, R., Makita, Y., Okamoto, N., Yoshihashi, H., Oki, H., Nanao, K., Moriyama, N., Oku, S., Hasegawa, T., Takahashi, T., Fukushima, Y., Kawame, H., Kosaki, K., 2006. Phenotypic spectrum of CHARGE syndrome with CHD7 mutations. J. Pediatr. 148, 410–414. Artinger, K.B., Bronner-Fraser, M., 1992. Partial restriction in the developmental potential of late emigrating avian neural crest cells. Dev. Biol. 149, 149–157. Augustine, K., Liu, E.T., Sadler, T.W., 1993. Antisense attenuation of Wnt-1 and Wnt-3a expression in whole embryo culture reveals roles for these genes in craniofacial, spinal cord, and cardiac morphogenesis. Dev. Genet. 14, 500–520. Aybar, M.J., Nieto, M.A., Mayor, R., 2003. Snail precedes Slug in the genetic cascade required for the specification and migration of the Xenopus neural crest. Development 130, 483–494. Bang, A.G., Papalopulu, N., Goulding, M.D., Kintner, C., 1999. Expression of Pax-3 in the lateral neural plate is dependent on a Wnt-mediated signal from posterior nonaxial mesoderm. Dev. Biol. 212, 366–380. Bell, K.M., Western, P.S., Sinclair, A.H., 2000. SOX8 expression during chick embryogenesis. Mech. Dev. 94, 257–260. Besmer, P., Manova, K., Duttlinger, R., Huang, R.J., Packer, A., Gyssler, C., Bachvarova, R.F., 1993. The kit-ligand (steel factor) and its receptor c-kit/W: pleiotropic roles in gametogenesis and melanogenesis. Development (Review), 125–137. Bockman, D.E., Kirby, M.L., 1984. Dependence of thymus development on derivatives of the neural crest. Science 223, 498–500. Bockman, D.E., Redmond, M.E., Waldo, K., Davis, H., Kirby, M.L., 1987. Effect of neural crest ablation on development of the heart and arch arteries in the chick. Am. J. Anat. 180, 332–341. Bockman, D.E., Redmond, M.E., Kirby, M.L., 1989. Alteration of early vascular development after ablation of cranial neural crest. Anat. Rec. 225, 209–217. Bouman, H.G., Broekhuizen, M.L., Baasten, A.M., Gittenberger-de Groot, A.C., Wenink, A.C., 1995. Spectrum of looping disturbances in stage 34 chicken hearts after retinoic acid treatment. Anat. Rec. 243, 101–108.

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

Brewer, S., Feng, W., Huang, J., Sullivan, S., Williams, T., 2004. Wnt1Cre-mediated deletion of AP-2alpha causes multiple neural crestrelated defects. Dev. Biol. 267, 135–152. Broekhuizen, M.L., Wladimiroff, J.W., Tibboel, D., Poelmann, R.E., Wenink, A.C., Gittenberger-de Groot, A.C., 1992. Induction of cardiac anomalies with all trans retinoic acid in the chick embryo. Cardiol. Young 2, 311–317. Bronner-Fraser, M., Fraser, S., 1989. Developmental potential of avian trunk neural crest cells in situ. Neuron 3, 755–766. Bronner-Fraser, M., Stern, C., 1991. Effects of mesodermal tissues on avian neural crest cell migration. Dev. Biol. 143, 213–217. Brown, C.B., Feiner, L., Lu, M.M., Li, J., Ma, X.K., Webber, A.L., Jia, L., Raper, J.A., Epstein, J.A., 2001. PlexinA2 and semaphorin signaling during cardiac neural crest development. Development 128, 3071–3080. Brown, C.B., Wenning, J.M., Lu, M.M., Epstein, D.J., Meyers, E.N., Epstein, J.A., 2004. Cre-mediated excision of Fgf8 in the Tbx1 expression domain reveals a critical role for Fgf8 in cardiovascular development in the mouse. Dev. Biol. 267, 190–202. Burstyn-Cohen, T., Katcheim, C., 2002. Association between the cell cycle and neural crest delamination through specific regulation of G1/S transition. Dev. Cell 3, 383–395. Carl, T.F., Dufton, C., Hanken, J., Klymkowsky, M.W., 1999. Inhibition of neural crest migration in Xenopus using antisense slug RNA. Dev. Biol. 213, 101–115. Carmichael, S.L., Shaw, G.M., Yang, W., Lammer, E.J., 2003. Maternal periconceptional alcohol consumption and risk for conotruncal heart defects. Birth Defects Res. A Clin. Mol. Teratol. 67, 875–878. Chalepakis, G., Goulding, M., Read, A., Strachan, T., Gruss, P., 1994. Molecular basis of splotch and Waardenburg Pax-3 mutations. Proc. Natl. Acad. Sci. USA 91, 3685–3689. Chan, W.Y., Cheung, C.S., Yung, K.M., Copp, A.J., 2004. Cardiac neural crest of the mouse embryo: axial level of origin, migratory pathway and cell autonomy of the splotch (Sp2H) mutant effect. Development 131, 3367–3379. Chen, J.-Y., Clifford, J., Zusi, C., Starrett, J., Tortolani, D., Ostrowski, J., Reczek, P.R., Chambon, P., Gronemeyer, H., 1996a. Two distinct actions of retinoid-receptor ligands. Nature 382, 819–822. Chen, S.Y., Yang, B., Jacobson, K., Sulik, K.K., 1996b. The membrane disordering effect of ethanol on neural crest cells in vitro and the protective role of GM1 ganglioside. Alcohol 13, 589–595. Choudhary, B., Ito, Y., Makita, T., Sasaki, T., Chai, Y., Sucov, H.M., 2006. Cardiovascular malformations with normal smooth muscle differentiation in neural crest-specific type II TGFbeta receptor (Tgfbr2) mutant mice. Dev. Biol. 289, 420–429. Conway, S.J., Henderson, D.J., Copp, A.J., 1997. Pax3 is required for cardiac neural crest migration in the mouse: evidence from the splotch (Sp2H) mutant. Development 124, 505–514. Couly, G.F., Coltey, P.M., Le Douarin, N.M., 1992. The developmental fate of the cephalic mesoderm in quail-chick chimeras. Development 114, 1–15. Creazzo, T.L., Godt, R.E., Leatherbury, L., Conway, S.J., Kirby, M.L., 1998. Role of cardiac neural crest cells in cardiovascular development. Ann. Rev. Physiol. 60, 267–286. d’Amico-Martel, A., Noden, D.M., 1983. Contribution of placodal and neural crest cells to avian cranial peripheral ganglia. Am. J. Anat. 166, 445–468. D’Antoni, S., Mattina, T., Di Mare, P., Federico, C., Motta, S., Saccone, S., 2004. Altered replication timing of the HIRA/Tuple1 locus in the DiGeorge and Velocardiofacial syndromes. Gene 333, 111–119.

Chapter  |  7.2  Role of Cardiac Neural Crest in the Development of the Caudal Pharyngeal Arches

Davidson, D., 1995. The function and evolution of Msx genes: pointers and paradoxes. TIG 11, 405–411. Daw, S.C., Taylor, C., Kraman, M., Call, K., Mao, J.I., Schuffenhauer, S., Meitinger, T., Lipson, T., Goodship, J., Scambler, P., 1996. A common region of 10p deleted in DiGeorge and velocardiofacial syndromes. Nat. Genet. 13, 458–460. Del Barrio, M.G., Nieto, M.A., 2002. Overexpression of Snail family members highlights their ability to promote chick neural crest formation. Development 129, 1583–1593. Dencker, L., Annerwall, E., Busch, C., Eriksson, U., 1990. Localization of specific retinoid-binding sites and expression of cellular retinoic-acidbinding protein (CRABP) in the early mouse embryo. Development 110, 343–352. Dettlaff-Swiercz, D.A., Wettschureck, N., Moers, A., Huber, K., Offermanns, S., 2005. Characteristic defects in neural crest cellspecific Galphaq/Galpha11- and Galpha12/Galpha13-deficient mice. Devel. Biol. 282, 174–182. Epstein, J.A., 1996. Pax3, neural crest and cardiovascular development. Trends Cardiovasc. Med. 6, 255–261. Epstein, J.A., Li, J., Lang, D., Chen, F., Brown, C.B., Jin, F.Z., Li, M.M., Thomas, M., Liu, E.C.J., Wessels, A., Lo, C.W., 2000. Migration of cardiac neural crest cells in Splotch embryos. Development 127, 1869–1878. Erickson, C.A., Duong, T.D., Tosney, K.W., 1992. Descriptive and experimental analysis of the dispersion of neural crest cells along the dorsolateral path and their entry into ectoderm in the chick embryo. Dev. Biol. 151, 251–272. Farrell, M.J., Burch, J.L., Wallis, K., Rowley, L., Kumiski, D., Stadt, H., Godt, R.E., Creazzo, T.L., Kirby, M.L., 2001. FGF-8 in the ventral pharynx alters development of myocardial calcium transients after neural crest ablation. J. Clin. Invest. 107, 1509–1517. Farrell, M.J., Stadt, H., Wallis, K.T., Scambler, P., Hixon, R.L., Wolfe, R., Leatherbury, L., Kirby, M.L., 1999. HIRA, a DiGeorge syndrome candidate gene, is required for cardiac outflow tract septation. Circ. Res. 84, 127–135. Feiner, L., Webber, A.L., Brown, C.B., Lu, M.M., Jia, L., Feinstein, P., Mombaerts, P., Epstein, J.A., Raper, J.A., 2001. Targeted disruption of semaphorin 3C leads to persistent truncus arteriosus and aortic arch interruption. Development 128, 3061–3070. Frank, D., Fotheringham, L., Brewer, J., Muglia, L., Tristani-Firouzi, M., Capecchi, M., Moon, A., 2002. An Fgf8 mouse mutant phenocopies human 22q11 deletion syndrome. Development 129, 4591–4603. Franz, T., 1989. Persistent truncus arteriosus in the Splotch mutant mouse. Anat. Embryol. 180, 457–464. Fraser, S.E., Bronner-Fraser, M., 1991. Migrating neural crest cells in the trunk of the avian embryo are multipotent. Development 112, 913–920. Fukiishi, Y., Morriss-Kay, G.M., 1992. Migration of cranial neural crest cells to the pharyngeal arches and heart in rat embryos. Cell Tiss. Res. 268, 1–8. Garg, V., Yamagishi, C., Hu, T.H., Kathiriya, I.S., Yamagishi, H., Srivastava, D., 2001. Tbx1, a DiGeorge syndrome candidate gene, is regulated by Sonic hedgehog during pharyngeal arch development. Dev. Biol. 235, 62–73. Goulding, E.H., Pratt, R.M., 1986. Isotretinoin teratogenicity in mouse whole embryo culture. J. Craniofac. Genet. Dev. Biol. 6, 99–112. Greenberg, F., Courtney, K.B., Wessels, R.A., Huhta, J., Carpenter, R.J., Rich, D.C., Ledbetter, D.H., 1988. Prenatal diagnosis of deletion 17p13 associated with DiGeorge anomaly. Am. J. Med. Genet. 31, 1–4. Gruber, P., Kubalak, S.W., Pexieder, T., Sucov, H.M., Evans, R.M., Chien, K.R., 1996. RXRalpha deficiency confers genetic susceptibility

457

to aortic sac, conotruncal, atrioventricular cushion, and ventricular muscle defects. J. Clin. Invest. 98, 1332–1343. Guris, D.L., Fantes, J., Tara, D., Druker, B.J., Imamoto, A., 2002. Mice lacking the homologue of the human 22q11.2 gene CRKL phenocopy neurocristopathies of DiGeorge syndrome. Nat. Genet. 27, 293–298. Guris, D.L., Duester, G., Papaioannou, V.E., Imamoto, A., 2006. Dosedependent interaction of Tbx1 and Crkl and locally aberrant RA signaling in a model of del22q11 syndrome. Dev. Cell 10, 81–92. Gurjarpadhye, A., Hewett, K.W., Justus, C., Wen, X., Stadt, H., Kirby, M.L., Sedmera, D., Gourdie, R.G., 2007. Cardiac neural crest ablation inhibits compaction and electrical function of conduction system bundles. Am. J. Physiol. Heart Circ. Physiol. 292 (3), H1291–H1300. Halford, S., Wilson, D.I., Daw, S.C., Roberts, C., Wadey, R., Kamath, S., Wickremasinghe, A., Burn, J., Goodship, J., Mattei, M.G., 1993. Isolation of a gene expressed during early embryogenesis from the region of 22q11 commonly deleted in DiGeorge syndrome. Hum. Mol. Genet. 2, 1577–1582. Hassler, J.A., Moran, D.J., 1986. Effectives of ethanol on the cytoskeleton of migrating and differentiating neural crest cells: Possible role in teratogenesis. J. Craniofac. Genet. Dev. Biol. (Suppl. 2), 129–136. Honore, S.M., Aybar, M.J., Mayor, R., 2003. Sox10 is required for the early development of the prospective neural crest in Xenopus embryos. Dev. Biol. 260, 79–96. Horstadius, S., 1950. The Neural Crest. Its Properties and Derivatives in the Light of Experimental Research. Oxford University Press, London, UK. Horton, C., Maden, M., 1995. Endogenous distribution of retinoids during normal development and teratogenesis in the mouse embryo. Dev. Dyn. 202, 312–323. Hu, T., Yamagishi, H., Maeda, J., McAnally, J., Yamagishi, C., Srivastava, D., 2004. Tbx1 regulates fibroblast growth factors in the anterior heart field through a reinforcing autoregulatory loop involving forkhead transcription factors. Development 131, 5491–5502. Huang, G.Y., Cooper, E.S., Waldo, K., Kirby, M.L., Gilula, N.B., Lo, C.W., 1998. Gap junction-mediated cell–cell communication modulates mouse neural crest migration. J. Cell Biol. 143, 1725–1734. Hutson, M.R., Zhang, P., Stadt, H.A., Sato, A., Li, Y.-X., Burch, J., Creazzo, T.L., Kirby, M.L., 2006. Cardiac outflow alignment is sensitive to levels of FGF8 signaling in the pharynx. Dev. Biol. 295, 486–497. Iida, K., Koseki, H., Kakinuma, H., Kato, N., Mizutani-Koseki, Y., Ohuchi, H., Yoshioka, H., Noji, S., Kawamura, K., Kataoka, Y., Ueno, F., Taniguchi, M., Yoshida, N., Sugiyama, T., Miura, N., 1997. Essential roles of the winged helix transcription factor MFH-1 in aortic arch patterning and skeletogenesis. Development 124, 4627–4638. Ishii, M., Han, J., Yen, H.Y., Sucov, H.M., Chai, Y., Maxson, R.E., 2005. Combined deficiencies of Msx1 and Msx2 cause impaired patterning and survival of the cranial neural crest. Development 132, 4937–4950. Ito, K., Sieber-Blum, M., 1991. In vitro clonal analysis of quail cardiac neural crest development. Dev. Biol. 148, 95–106. Jaenisch, R., 1985. Mammalian neural crest cells participate in normal embryonic development on microinjection into post-implantation mouse embryos. Nature 318, 181–183. Jerome, L.A., Papaioannou, V.E., 2001. DiGeorge syndrome phenotype in mice mutant for the T-box gene. Tbx1. Nat. Genet. 27, 286–291. Jiang, X., Rowitch, D.H., Soriano, P., McMahon, A.P., Sucov, H.M., 2000. Fate of the mammalian cardiac neural crest. Development 127, 1607–1616. Jiang, X.B., Choudhary, B., Merki, E., Chien, K., Maxson, R., Sucov, H. M., 2002. Normal fate and altered function of the cardiac neural crest

458

cell lineage in retinoic acid receptor mutant embryos. Mech. Dev. 117, 115–122. Kaartinen, V., Dudas, M., Nagy, A., Sridurongrit, S., Lu, M.M., Epstein, J.A., 2004. Cardiac outflow tract defects in mice lacking ALK2 in neural crest cells. Development 131, 3481–3490. Karayiorgou, M., Morris, M.A., Morrow, B., Shprintzen, R.J., Goldberg, R., Borrow, J., Gos, A., Nestadt, G., Wolyniec, P.S., Lasseter, V.K., Eisen, H., Childs, B., Kazazian, H.H., Kucherlapati, R., Antonarakis, S.E., Pulver, A.E., Housman, D.E., 1995. Schizophrenia susceptibility associated with interstitial deletions of chromosome 22q11. Proc. Natl. Acad. Sci. USA 92, 7612–7616. Kastner, P., Grondona, J.M., Mark, M., Gansmuller, A., LeMeur, M., Decimo, D., Vondsch, J.-L., Doll, P., Chambon, P., 1994. Genetic analysis of RXRalpha developmental function: convergence of RXR and RAR signaling pathways in heart and eye morphogenesis. Cell 78, 987–1003. Kawasaki, T., Kitsukawa, T., Bekku, Y., Matsuda, Y., Sanbo, M., Yagi, T., Fujisawa, H., 1999. A requirement for neuropilin-1 in embryonic vessel formation. Development 126, 4895–4902. Kelly, R.G., Brown, N.A., Buckingham, M.E., 2001. The arterial pole of the mouse heart forms from Fgf10-expressing cells in pharyngeal mesoderm. Dev. Cell 1, 435–440. Kirby, M., 1989. Plasticity and predetermination of the mesencephalic and trunk neural crest transplanted into the region of cardiac neural crest. Dev. Biol. 134, 402–412. Kirby, M.L., 1987. Cardiac morphogenesis: recent research advances. Ped. Res. 21, 219–224. Kirby, M.L., 1988. Nodose placode contributes autonomic neurons to the heart in the absence of cardiac neural crest. J. Neurosci. 8, 1089–1095. Kirby, M.L., 1988. Nodose placode provides ectomesenchyme to the developing heart in the absence of cardiac neural crest. Cell Tiss. Res. 252, 17–22. Kirby, M.L., 1988. Role of extracardiac factors in heart development. Experientia 44, 944–950. Kirby, M.L., Gale, T.F., Stewart, D.E., 1983. Neural crest cells contribute to aorticopulmonary septation. Science 220, 1059–1061. Kirby, M.L., Bockman, D.E., 1984. Neural crest and normal development: a new perspective. Anat. Rec. 209, 1–6. Kirby, M.L., Aronstam, R.S., Buccafusco, J.J., 1985. Changes in cholinergic parameters associated with conotruncal malformations in embryonic chick hearts. Circ. Res. 56, 392–401. Kirby, M.L., Turnage, K.L., Hays, B.M., 1985. Characterization of conotruncal malformations following ablation of “cardiac” neural crest. Anat. Rec. 213, 87–93. Kirby, M.L., Waldo, K.L., 1990. Role of the neural crest in congenital heart disease. Circulation 82, 332–340. Kirby, M.L., Creazzo, T.L., 1995. Cardiovascular development. Neural crest and new perspectives. Cardiol. Rev. 3, 226–235. Kirby, M.L., Waldo, K.L., 1995. Neural crest and cardiovascular patterning. Circ. Res. 77, 211–215. Kirby, M.L., Hunt, P., Wallis, K.T., Thorogood, P., 1997. Abnormal patterning of the aortic arch arteries does not evoke cardiac malformations. Dev. Dyn. 208, 34–47. Kretz, M., Eckardt, D., Kruger, O., Kim, J.S., Maurer, J., Theis, M., van Rijen, H.V., Schorle, H., Willecke, K., 2006. Normal embryonic development and cardiac morphogenesis in mice with Wnt1Cre-mediated deletion of connexin43. Genesis 44, 269–276. Krezel, W., Dup, V., Mark, M., Dierich, A., Kastner, P., Chambon, P., 1996. RXRgamma null mice are apparently normal and compound

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

RXRalpha//RXRalpha-//RXRgamma/ mutant mice are viable. Proc. Natl. Acad. Sci. USA 93, 9010–9014. Kume, T., Jiang, H.Y., Topczewska, J.M., Hogan, B.L., 2001. The murine winged helix transcription factors, Foxc1 and Foxc2, are both required for cardiovascular development and somitogenesis. Genes Dev. 15, 2470–2482. Kuratani, S., Bockman, D.E., 1990. Impaired development of the thymic primordium after neural crest ablation. Anat. Rec. 228, 185–190. Kurihara, Y., Kurihara, H., Maemura, K., Kuwaki, T., Kumada, M., Yazaki, Y., 1995a. Impaired development of the thyroid and thymus in endothelin-1 knockout mice. J. Cardiovasc. Pharmacol. 26 (Suppl. 3), S13–S16. Kurihara, Y., Kurihara, H., Oda, H., Maemura, K., Nagai, R., Ishikawa, T., Yazaki, Y., 1995b. Aortic arch malformations and ventricular septal defect in mice deficient in endothelin-1. J. Clin. Invest. 96, 293–300. Kwang S.J., Brugger, S.M., Lazik, A., Merrill, A.E., Wu, L.Y., Liu, Y. H., Ishii, M., Sangiorgi, F.O., Rauchman, M., Sucov, H.M., Maas, R. L., Maxson, R.E., Jr.,2002. Msx2 is an immediate downstream effector of Pax3 in the development of the murine cardiac neural crest. Development 129, 527–538. LaBonne, C., Bronner-Fraser, M., 2000. Snail-related transcriptional repressors are required in Xenopus for both the induction of the neural crest and its subsequent migration. Dev. Biol. 221, 195–205. Lamour, V., Lecluse, Y., Desmaze, C., Spector, M., Bodescot, M., Aurias, A., Osley, M.A., Lipinski, M., 1995. A human homolog of the S. cere­ visiae HIR1 and HIR2 transcriptional repressors cloned from the DiGeorge syndrome critical region. Hum. Mol. Genet. 4, 791–799. Langston, A.W., Thompson, J.R., Gudas, L.J., 1997. Retinoic acidresponsive enhancers located 3 of the Hox A and Hox B homeobox gene clusters – Functional analysis. J. Biol. Chem. 272, 2167–2175. Le Douarin, N., 1982. The Neural Crest. Cambridge University Press, Cambridge, UK. Le Douarin, N.M., 1990. Cell lineage segregation during neural crest ontogeny. Ann. NY Acad. Sci. 599, 131–140. Le Douarin, N.M., Jotereau, F.V., 1975. Tracing of cells of the avian thymus through embryonic life in interspecific chimeras. J. Exp. Med. 142, 17–40. Le Douarin, N.M., Ziller, C., Couly, G.F., 1993. Patterning of neural crest derivatives in the avian embryo: In vivo and in vitro studies. Dev. Biol. 159, 24–49. Le Douarin, N.M., Kalcheim, C., 1999. The Neural Crest. Cambridge University Press, Cambridge, UK. Leatherbury, L., Gauldin, H.E., Waldo, K.L., Kirby, M.L., 1990. Microcinephotography of the developing heart in neural crest-ablated chick embryos. Circulation 81, 1047–1057. Leatherbury, L., Connuck, D.M., Gauldin, H.E., Kirby, M.L., 1991. Hemodynamic changes and compensatory mechanisms during early cardiogenesis after neural crest ablation in chick embryos. Ped. Res. 30, 509–512. Lee, J.W., Ryan, F., Swaffield, J.C., Johnston, S.A., Moore, D.D., 1995. Interaction of thyroid-hormone receptor with a conserved transcriptional mediator. Nature 374, 91–94. Lee, M.J., Brennan, A., Blanchard, A., Zoidl, G., Dong, Z., Tabernero, A., Zoidl, C., Dent, M.A.R., Jessen, K.R., Mirsky, R., 1997. P0 is constitutively expressed in the rat neural crest and embryonic nerves and is negatively and positively regulated by axons to generate non-myelinforming and myelin-forming Schwann cells, respectively. Mol. Cell Neurosci. 8, 336–350. Lepore, J.J., Mericko, P.A., Cheng, L., Lu, M.M., Morrisey, E.E., Parmacek, M.S., 2006. GATA-6 regulates semaphorin 3C and is

Chapter  |  7.2  Role of Cardiac Neural Crest in the Development of the Caudal Pharyngeal Arches

required in cardiac neural crest for cardiovascular morphogenesis. J. Clin. Invest. 116, 929–939. Li, Y.X., Zdanowicz, M., Young, L., Kumiski, D., Leatherbury, L., Kirby, M.L., 2003. Cardiac neural crest in zebrafish embryos contributes to myocardial cell lineage and early heart function. Dev. Dyn. 226, 540–550. Liu, S., Liu, F., Schneider, A.E., St. Amand, T., Epstein, J.A., Gutstein, D.E., 2006. Distinct cardiac malformations caused by absence of connexin 43 in the neural crest and in the non-crest neural tube. Development 133, 2063–2073. Lo, C.W., Waldo, K.L., Kirby, M.L., 1999. Gap junction communication and the modulation of cardiac neural crest cells. Trends Cardiovasc. Med. 9, 63–69. Lu, H.-C., Eichele, G., Thaller, C., 1997. Ligand-bound RXR can mediate retinoid signal transduction during embryogenesis. Development 124, 195–203. Lumsden, A., Sprawson, N., Graham, A., 1991. Segmental origin and migration of neural crest cells in the hindbrain region of the chick embryo. Development 113, 1281–1291. Luo, T., Lee, Y.H., Saint-Jeannet, J.P., Sargent, T.D., 2003. Induction of neural crest in Xenopus by transcription factor AP2alpha. Proc. Natl. Acad. Sci. USA 100, 532–537. Luo, T., Zhang, Y., Khadka, D., Rangarajan, J., Cho, K.W., Sargent, T. D., 2005. Regulatory targets for transcription factor AP2 in Xenopus embryos. Dev. Growth Differ. 47, 403–413. Macatee, T.L., Hammond, B.P., Benjamin, R., Arenkiel, B.R., Francis, L., Frank, D.U., Moon, A.M., 2003. Ablation of specific expression domains reveals discrete functions of ectoderm- and endodermderived FGF8 during cardiovascular and pharyngeal development. Development 130, 6361–6374. Maden, M., Ong, D.E., Chytil, F., 1990. Retinoid-binding protein distribution in the developing mammalian nervous system. Development 109, 75–80. Maden, M., Hunt, P., Eriksson, U., Kuroiwa, A., Krumlauf, R., Summerbell, D., 1991. Retinoic acid-binding protein, rhombomeres and the neural crest. Development 111, 35–44. Manley, N.R., Capecchi, M.R., 1995. The role of Hoxa-3 in mouse thymus and thyroid development. Development 121, 1989–2003. Manner, J., Seidl, W., Steding, G., 1996. Experimental study on the significance of abnormal cardiac looping for the development of cardiovascular anomalies in neural crest-ablated chick embryos. Anat. Embryol. 194, 289–300. Marshall, H., Nonchev, S., Sham, M.H., Muchamore, I., Lumsden, A., Krumlauf, R., 1992. Retinoic acid alters hindbrain Hox code and induces transformation of rhombomeres 2/3 into a 4/5 identity. Nature 360, 737–741. Matsuoka, R., Takao, A., Kimura, M., Imamura, S., Kondo, C., Joh-o, K., Ikeda, K., Nishibatake, M., Ando, M., Momma, K., 1994. Confirmation that the conotruncal anomaly face syndrome is associated with a deletion within 22q11.2. Am. J. Med. Genet. 53, 285–289. Maxwell, G.D., Sietz, P.D., Rafford, C.E., 1982. Synthesis and accumulation of putative neurotransmitters by cultured neural crest cells. J. Neurosci. 2, 879–888. McMahon, A.P., Joyner, A.L., Bradley, A., McMahon, J.A., 1992. The midbrain-hindbrain phenotype of Wnt-1-/Wnt-1- mice results from stepwise deletion of engrailed-expressing cells by 9.5 days post­ coitum. Cell 69, 581–595. Miyagawa-Tomita, S., Waldo, K., Tomita, H., Kirby, M.L., 1991. Temporospatial study of the migration and distribution of cardiac neural crest in quail-chick chimeras. Am. J. Anat. 192, 79–88.

459

Mjaatvedt, C.H., Nakaoka, T., Moreno-Rodriguez, R., Norris, R.A., Kern, M.J., Eisenberg, C.A., Turner, D., Markwald, R.R., 2001. The outflow tract of the heart is recruited from a novel heart-forming field. Dev. Biol. 238, 97–109. Moon, A.M., Guris, D.L., Seo, J.H., Li, L., Hammond, J., Talbot, A., Imamoto, A., 2006. Crkl deficiency disrupts Fgf8 signaling in a mouse model of 22q11 deletion syndromes. Dev. Cell 10, 71–80. Nakajima, Y., Morishima, M., Nakazawa, M., Momma, K., 1996. Inhibition of outflow cushion mesenchyme formation in retinoic acid-induced complete transposition of the great arteries. Cardiovasc. Res. 31, E77–E85. Nakamura, H., Ayer-Le Lièvre, C.S., 1982. Mesectodermal capabilities of the trunk neural crest of birds. J. Embryol. Exp. Morphol. 70, 1–18. Nakamura, T., Colbert, M.C., Robbins, J., 2006. Neural crest cells retain multipotential characteristics in the developing valves and label the cardiac conduction system. Circ. Res. 98, 1547–1554. Nataf, V., Lecoin, L., Eichmann, A., Le Douarin, N.M., 1996. EndothelinB receptor is expressed by neural crest cells in the avian embryo. Proc. Natl. Acad. Sci. USA 93, 9645–9650. Niederreither, K., Vermot, J., Messaddeq, N., Schuhbaur, B., Chambon, P., Dolle, P., 2001. Embryonic retinoic acid synthesis is essential for heart morphogenesis in the mouse. Development 128, 1019–1031. Niederreither, K., Vermot, J., Le Roux, I., Schuhbaur, B., Chambon, P., Dolle, P., 2003. The regional pattern of retinoic acid synthesis by RALDH2 is essential for the development of posterior pharyngeal arches and the enteric nervous system. Development 130, 2525–2534. Nieto, M.A., Sargent, M.G., Wilkinson, D., Cooke, J., 1994. Control of cell behavior during vertebrate development by Slug, a zinc finger gene. Science 264, 835–839. Nishibatake, M., Kirby, M.L., van Mierop, L.H., 1987. Pathogenesis of persistent truncus arteriosus and dextroposed aorta in the chick embryo after neural crest ablation. Circ. Res. 75, 255–264. Noden, D.M., 1978. The control of avian cephalic neural crest cytodifferentiation. I. Skeletal and connective tissues. Dev. Biol. 67, 296–312. Noden, D.M., 1980. The migration and cytodifferentiation of cranial neural crest cells. In: Pratt, R., Christiansen, R.L. (Eds.), Current Research Trends in Prenatal Craniofacial Development. Elsevier North Holland, Inc., Amsterdam, The Netherlands, pp. 3–25. Noden, D.M., 1983a. The embryonic origins of avian cephalic and cervical muscles and associated connective tissues. Am. J. Anat. 168, 257–276. Noden, D.M., 1983b. The role of neural crest in patterning of avian cranial skeletal, connective, and muscle tissues. Dev. Biol. 96, 144–165. Nowotschin, S., Liao, J., Gage, P.J., Epstein, J.A., Campione, M., Morrow, B.E., 2006. Tbx1 affects asymmetric cardiac morphogenesis by regulating Pitx2 in the secondary heart field. Development 133, 1565–1573. Nyquist-Battie, C., Freter, M., 1988. Cardiac mitochondrial abnormalities in a mouse model of the fetal alcohol syndrome. Alcoholism: Clin. Exp. Res. 12, 264–267. Orr-Urtreger, A., Bedford, M.T., Do, M.-S., Eisenbach, L., Lonai, P., 1992. Developmental expression of the alpha receptor for plateletderived growth factor, which is deleted in the embryonic lethal Patch mutation. Development 115, 289–303. Papolos, D.F., Faedda, G.L., Veit, S., Goldberg, R., Morrow, B., Kucherlapati, R., Shprintzen, R.J., 1996. Bipolar spectrum disorders in patients diagnosed with velo- cardio-facial syndrome: Does a hemizygous deletion of chromosome 22q11 result in bipolar affective disorder? Am. J. Psychiat. 153, 1541–1547.

460

Pasini, B., Ceccherini, I., Romeo, G., 1996. RET mutations in human disease. TIG 12, 138–144. Peters-Van der Sanden, M.J., Kirby, M.L., Gittenberger-de Groot, A.C., Tibboel, D., Mulder, M.P., Meijers, C., 1993. Ablation of various regions within the avian vagal neural crest has differential effects on ganglion formation in the fore-, mid- and hindgut. Dev. Dyn. 196, 183–194. Pietri, T., Eder, O., Blanche, M., Thiery, J.P., Dufour, S., 2003. The human tissue plasminogen activator-Cre mouse: a new tool for targeting specifically neural crest cells and their derivatives in vivo. Dev. Biol. 259, 176–187. Poelmann, R.E., Jongbloed, M.R., Molin, D.G., Fekkes, M.L., Wang, Z., Fishman, G.I., Doetschman, T., Azhar, M., Gittenberger-de Groot, A.C., 2004. The neural crest is contiguous with the cardiac conduction system in the mouse embryo: a role in induction?. Anat. Embryol. (Berlin) 208, 389–393. Qiu, M.S., Bulfone, A., Ghattas, I., Meneses, J.J., Christensen, L., Sharpe, P.T., Presley, R., Pedersen, R.A., Rubenstein, J.L., 1997. Role of the Dlx homeobox genes in proximodistal patterning of the branchial arches: Mutations of Dlx-1, Dlx-2, and Dlx-1 and -2 alter morphogenesis of proximal skeletal and soft tissue structures derived from the first and second arches. Dev. Biol. 185, 165–184. Raid, K., Turnley, A.M., Maxwell, G.D., Kurihara, Y., Kurihara, H., Bartlett, P.F., Murphy, M., 1996. Multiple roles for endothelin in melanocyte development: Regulation of progenitor number and stimulation of differentiation. Development 122, 3911–3919. Reaume, A.G., de Sousa, P.A., Kulkarni, S., Langille, B.L., Zhu, D., Davies, T.C., Juneja, S.C., Kidder, G.M., Rossant, J., 1995. Cardiac malformation in neonatal mice lacking connexin43. Science 267, 1831–1834. Reifers, F., Walsh, E.C., Léger, S., Stainier, D.Y., Brand, M., 2000. Induction and differentiation of the zebrafish heart requires fibroblast growth factor 8 (fgf8/acerebellar). Development 127, 225–235. Riccardi, V.M., 1991. Neurofibromatosis: Past, present and future. New England J. Med. 324, 1283–1285. Roberts, C., Daw, S.C., Halford, S., Scambler, P.J., 1997. Cloning and developmental expression analysis of chick Hira (Chira), a candidate gene for DiGeorge syndrome. Hum. Mol. Genet. 6, 237–245. Roberts, C., Ivins, S.M., James, C.T., Scambler, P.J., 2005. Retinoic acid down-regulates Tbx1 expression in vivo and in vitro. Dev. Dyn. 232, 928–938. Roberts, C., Ivins, S., Cook, A.C., Baldini, A., Scambler, P.J., 2006. Cyp26 genes a1, b1 and c1 are down-regulated in Tbx1 null mice and inhibition of Cyp26 enzyme function produces a phenocopy of DiGeorge Syndrome in the chick. Hum. Mol. Genet. 15, 3394–3410. Rosenquist, T.H., Kirby, M.L., Van Mierop, L.H., 1989. Solitary aortic arch artery. A result of simultaneous ablation of cardiac neural crest and nodose placode in the avian embryo. Circulation 80, 1469–1475. Rothman, K.J., Moore, L.L., Singer, M.R., Nguyen, U.S., Mannino, S., Milunsky, A., 1995. Teratogenicity of high vitamin A intake. New Engl. J. Med. 333, 1369–1373. Rowe, A., Eager, N.S., Brickell, P.M., 1991. A member of the RXR nuclear receptor family is expressed in neural-crest-derived cells of the developing chick peripheral nervous system. Development 111, 771–778. Rowe, A., Brickell, P.M., 1995. Expression of the chicken retinoid X receptor-gamma gene in migrating cranial neural crest cells. Anat. Embryol. 192, 1–8. Sanchez-Martin, M., Rodriguez-Garcia, A., Perez-Losada, J., Sagrera, A., Read, A.P., Sanchez-Garcia, I., 2002. SLUG (SNAI2) deletions in patients with Waardenburg disease. Hum. Mol. Genet. 11, 3231–3236.

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

Satiroglu-Tufan, N.L., Tufan, A.C., 2004. Amelioration of ethanolinduced growth retardation by all-trans-retinoic acid and alpha-tocopherol in shell-less culture of the chick embryo. Reprod. Toxicol. 18, 407–412. Sato, M., Yost, H.J., 2003. Cardiac neural crest contributes to cardiomyogenesis in zebrafish. Dev. Biol. 257, 127–139. Scamps, C., Lorain, S., Lamour, V., Lipinski, M., 1996. The HIR protein family: isolation and characterization of a complete murine cDNA. Biochem. Biophys. Acta. 1306, 5–8. Schaefer, K.S., Doughman, Y.Q., Fisher, S.A., Watanabe, M., 2004. Dynamic patterns of apoptosis in the developing chicken heart. Dev. Dyn. 229, 489–499. Scherson, T., Serbedzija, G., Fraser, S., Bronner-Fraser, M., 1993. Regulative capacity of the cranial neural tube to form neural crest. Development 118, 1049–1062. Schilham, M.W., Oosterwegel, M.A., Moerer, P., Ya, J., de Boer, P.A., van de Wetering, M., Verbeek, S., Lamers, W.H., Kruisbeek, A.M., Cumano, A., Clevers, H., 1996. Defects in cardiac outflow tract formation and pro-B-lymphocyte expansion in mice lacking Sox-4. Nature 380, 711–714. Sechrist, J., Nieto, M.A., Zamanian, R.T., Bronner-Fraser, M., 1995. Regulative response of the cranial neural tube after neural fold ablation: spatiotemporal nature of neural crest regeneration and up-regulation of Slug. Development 121, 4103–4115. Seo, S., Kume, T., 2006. Forkhead transcription factors, Foxc1 and Foxc2, are required for the morphogenesis of the cardiac outflow tract. Dev. Biol. 296, 421–436. Serbedzija, G.N., Fraser, S.E., Bronner-Fraser, M., 1990. Pathways of trunk neural crest cell migration in the mouse embryo as revealed by vital dye labelling. Development 108, 605–612. Serbedzija, G.N., Burgan, S., Fraser, S.E., Bronner-Fraser, M., 1991. Vital dye labelling demonstrates a sacral neural crest contribution to the enteric nervous system of chick and mouse embryos. Development 111, 857–866. Serbedzija, G.N., Bronner-Fraser, M., Fraser, S.E., 1992. Vital dye analysis of cranial neural crest cell migration in the mouse embryo. Development 116, 297–307. Siebert, J.R., Graham, J.M., MacDonald, C., 1985. Pathologic features of the CHARGE association: support for involvement of the neural crest. Teratology 31, 331–336. Smits-van Prooije, A.E., Poelmann, R.E., Dubbeldam, J.A., Mentink, M.M., Vermeij-Keers, C., 1986. Wheat germ agglutinin-gold as a novel marker for mesectoderm formation in mouse embryos cultured in vitro. Stain Technol. 61, 97–106. Soriano, P., 1997. The PDGFalpha receptor is required for neural crest cell development and for normal patterning of the somites. Development 124, 2691–2700. Spokony, R.F., Aoki, Y., Saint-Germain, N., Magner-Fink, E., SaintJeannet, J.P., 2002. The transcription factor Sox9 is required for cranial neural crest development in Xenopus. Development 129, 421–432. Stalmans, I., Lambrechts, D., De Smet, F., Janssen, S., Wang, J., Maity, S., Kneer, P., Von der Ohe, M., Swillen, A., Maes, C., Gewillig, M., Molin, D.G., Hellings, P., Boetel, T., Haardt, M., Compernolle, V., Dewerchin, M., Plaisance, S., Vlietinck, R., Emanuel, B., Gittenberger-de Groot, A.C., Esguerra, C., Scambler, P., Morrow, B., Driscol, D., Moons, L., Carmeliet, G., Behn-Krappa, A., Devriendt, K., Collen, D., Conway, S., Carmeliet, P., 2003. VEGF: A modifier of the del22q11 (DiGeorge) syndrome? Nature Med. 9, 1–10. Steller, H., 1995. Mechanisms and genes of cellular suicide. Science 267, 1445–1449.

Chapter  |  7.2  Role of Cardiac Neural Crest in the Development of the Caudal Pharyngeal Arches

Stern, C.D., Artinger, K.B., Bronner-Fraser, M., 1991. Tissue interactions affecting the migration and differentiation of neural crest cells in the chick embryo. Development 113, 207–216. Steventon, B., Carmona-Fontaine, C., Mayor, R., 2005. Genetic network during neural crest induction: from cell specification to cell survival. Semin. Cell Dev. Biol. 16, 647–654. Stewart, R.A., Arduini, B.L., Berghmans, S., George, R.E., Kanki, J.P., Henion, P.D., Look, A.T., 2006. Zebrafish foxd3 is selectively required for neural crest specification, migration and survival. Dev. Biol. 292, 174–188. Stottmann, R.W., Choi, M., Mishina, Y., Meyers, E.N., Klingensmith, J., 2004. BMP receptor IA is required in mammalian neural crest cells for development of the cardiac outflow tract and ventricular myocardium. Development 131, 2205–2218. Strong, W.B., 1968. Familial syndrome of right-sided aortic arch, mental deficiency, and facial dysmorphism. J. Pediatr. 73, 882–888. Sulik, K.K., Johnston, M.C., Daft, P.A., Russell, W.E., Dehart, D.B., 1986. Fetal alcohol syndrome and DiGeorge anomaly: critical ethanol exposure periods for craniofacial malformations as illustrated in an animal model. Am. J. Med. Genet. (Suppl. 2), 97–112. Suzuki, H.R., Kirby, M.L., 1997. Absence of neural crest cell regeneration from the postotic neural tube. Dev. Biol. 184, 222–233. Tan, S.S., Morriss-Kay, G.M., 1986. Analysis of cranial neural crest cell migration and early fates in postimplantation rat chimeras. J. Embryo. Exp. Morphol. 98, 21–58. Tassabehji, M., Newton, V.E., Leverton, K., Turnbull, K., Seemanova, E., Kunze, J., Sperling, K., Strachan, T., Read, A.P., 1994. PAX3 gene structure and mutations: Close analogies between Waardenburg syndrome and the Splotch mouse. Hum. Mol. Genet. 3, 1069–1074. Tassabehji, M., Read, A.P., Newton, V.E., Harris, R., Balling, R., Gruss, P., Strachan, T., 1992. Waardenburg’s syndrome patients have mutations in the human homologue of the Pax-3 paired box gene. Nature 355, 635–636. Thomas, K.R., Capecchi, M.R., 1990. Targeted disruption of the murine int-1 proto-oncogene resulting in severe abnormalities in midbrain and cerebellar development. Nature 346, 847–850. Thorogood, P., Smith, L., Nicol, A., McGinty, R., Garrod, D., 1982. Effects of vitamin A on the behavior of migratory neural crest cells in vitro. J. Cell Sci. 57, 331–350. Tomita, H., Connuck, D.M., Leatherbury, L., Kirby, M.L., 1991. Relation of early hemodynamic changes to final cardiac phenotype and survival after neural crest ablation in chick embryos. Circulation 84, 1289–1295. Torres, M., Stoykova, A., Huber, O., Chowdhury, K., Bonaldo, P., Mansouri, A., Butz, S., Kemler, R., Gruss, P., 1997. An alpah-E-catenin gene trap mutation defines its function in preimplantation development. Proc. Natl. Acad. Sci. USA 94, 901–906. Tribulo, C., Aybar, M.J., Nguyen, V.H., Mullins, M.C., Mayor, R., 2003. Regulation of Msx genes by a Bmp gradient is essential for neural crest specification. Development 130, 6441–6452. Vaessen, M.-J., Meijers, J.H., Bootsma, D., van Kessel, A.D., 1990. The cellular retinoic-acid-binding protein is expressed in tissues associated with retinoic-acid-induced malformations. Development 110, 371–378. Van Mierop, L.H., Kutsche, L.M., 1986. Cardiovascular anomalies in DiGeorge syndrome and importance of neural crest as a possible pathogenetic factor. Am. J. Cardiol. 58, 133–137. Vissers, L.E., van Ravenswaaij, C.M., Admiraal, R., Hurst, J.A., de Vries, B.B., Janssen, I.M., van der Vliet, W.A., Huys, E.H., de Jong, P.J., Hamel, B.C., Schoenmakers, E.F., Brunner, H.G., Veltman, J.A., van Kessel, A.G., 2006. Mutations in a new member of the chromodomain gene family cause CHARGE syndrome. Nat. Genet. 36, 955–957.

461

Vitelli, F., Taddei, I., Morishima, M., Meyers, E., Lindsay, E., Baldini, A., 2002. A genetic link between Tbx1 and fibroblast growth factor signaling. Development 129, 4605–4611. Wachtler, F., 1984. On the differentiation and migration of some nonneuronal neural crest derived cell types. Anat. Embryol. 170, 161–168. Waldo, K.L., Kumiski, D., Kirby, M.L., 1996. Cardiac neural crest is essential for the persistence rather than the formation of an arch artery. Dev. Dyn. 205, 281–292. Waldo, K., Miyagawa-Tomita, S., Kumiski, D., Kirby, M.L., 1998. Cardiac neural crest cells provide new insight into septation of the outflow tract: Aortic sac to ventricular septal closure. Dev. Biol. 196, 129–144. Waldo, K.L., Lo, C.W., Kirby, M.L., 1999. Connexin 43 expression reflects neural crest patterns during cardiovascular development. Dev. Biol. 208, 307–323. Waldo, K.L., Kumiski, D.H., Wallis, K.T., Stadt, H.A., Hutson, M.R., Platt, D.H., Kirby, M.L., 2001. Conotruncal myocardium arises from a secondary heart field. Development 128, 3179–3188. Waldo, K.L., Hutson, M.R., Stadt, H.A., Zdanowicz, M., Zdanowicz, J., Kirby, M.L., 2005a. Cardiac neural crest is necessary for normal addition of the myocardium to the arterial pole from the secondary heart field. Dev. Biol. 281(1), 66–77. Waldo, K.L., Hutson, M.R., Ward, C.C., Zdanowicz, M., Stadt, H.A., Kumiski, D., Abu-Issa, R., Kirby, M.L., 2005b. Secondary heart field contributes myocardium and smooth muscle to the arterial pole of the developing heart. Dev. Biol. 281, 78–90. Wang, B., Weidenfeld, J., Lu, M.M., Maika, S., Kuziel, W.A., Morrisey, E.E., Tucker, P.W., 2004. Foxp1 regulates cardiac outflow tract, endocardial cushion morphogenesis and myocyte proliferation and maturation. Development 131, 4477–4487. Wang, H., Charles, P.C., Wu, Y., Ren, R., Pi, X., Moser, M., BarshishatKupper, M., Rubin, J.S., Perou, C., Bautch, V., Patterson, C., 2006. Gene expression profile signatures indicate a role for Wnt signaling in endothelial commitment from embryonic stem cells. Circ. Res. 98, 1331–1339. Ward, C., Stadt, H.A., Hutson, M.R., Kirby, M.L., 2005. Ablation of the secondary heart field leads to tetralogy of Fallot and pulmonary atresia. Dev. Biol. 284, 72–83. Wehrle-Haller, B., Morrison-Graham, K., Weston, J.A., 1996. Ectopic c-kit expression affects the fate of melanocyte precursors in patch mutant embryos. Dev. Biol. 177, 463–474. Wendling, O., Ghyselinck, N.B., Chambon, P., Mark, M., 2001. Roles of retinoic acid receptors in early embryonic morphogenesis and hindbrain patterning. Development 128, 2031–2038. Weston, J.A., 1982. Neural crest cell development. Prog. Clin. Biol. Res. 85, 359–379. Wilson, D.I., Burn, J., Scambler, P., Goodship, J., 1993. DiGeorge syndrome: part of CATCH 22. J. Med. Genet. 30, 852–856. Wilson, J.G., Warkany, J., 1950. Cardiac and aortic arch anomalies in offspring of vitamin A deficient rats correlated with similar human anomalies. Pediatrics 5, 708–725. Winnier, G.E., Kume, T., Deng, K.Y., Rogers, R., Bundy, J., Raines, C., Walter, M.A., Hogan, B.L., Conway, S.J., 1999. Roles for the winged helix transcription factors MF1 and MFH1 in cardiovascular develop­ ment revealed by nonallelic noncomplementation of null alleles. Dev. Biol. 213, 418–431. Xu, D., Emoto, N., Giaid, A., Slaughter, C., Kaw, S., DeWit, D., Yanagisawa, M., 1994. ECE-1: a membrane-bound metalloprotease that catalyzes the proteolytic activation of big endothelin-1. Cell 78, 473–785.

462

Xu, H., Morishima, M., Wylie, J.N., Schwartz, R.J., Bruneau, B.G., Lindsay, E.A., Baldini, A., 2004. Tbx1 has a dual role in the morphogenesis of the cardiac outflow tract. Development 131, 3217–3227. Yagi, H., Furutani, Y., Hamada, H., Sasaki, T., Asakawa, S., Minoshima, S., Ichida, F., Joo, K., Kimura, M., Imamura, S., Kamatani, N., Momma, K., Takao, A., Nakazawa, M., Shimizu, N., Matsuoka, R., 2003. Role of TBX1 in human del22q11.2 syndrome. Lancet 362, 1366–1373. Yamagishi, H., Maeda, J., Hu, T.H., McAnally, J., Conway, S.J., Kume, T., Meyers, E.N., Yamagishi, C., Srivastava, D., 2003. Tbx1 is regulated by tissue-specific forkhead proteins through a common Sonic hedgehog-responsive enhancer. Genes. Dev. 17, 269–281. Yasuda, Y., Okamoto, M., Konishi, H., Matsuo, T., Kihara, T., Tanimura, T., 1986. Developmental anomalies induced by all-trans retinoic acid in fetal mice: I. Macroscopic findings. Teratology 34, 37–49.

PART  |  7  Cardiac Neural Crest and Pharyngeal Patterning

Yasui, H., Nakazawa, M., Morishima, M., Miyagawa-Tomita, S., Momma, K., 1995. Morphological observations on the pathogenetic process of transposition of the great arteries induced by retinoic acid in mice. Circulation 91, 2478–2486. Yelbuz, T.M., Waldo, K.L., Kumiski, D.H., Stadt, H.A., Wolfe, R.R., Leatherbury, L., Kirby, M.L., 2002. Shortened outflow tract leads to altered cardiac looping after neural crest ablation. Circulation 106, 504–510. Yelin, R., Schyr, R.B., Kot, H., Zins, S., Frumkin, A., Pillemer, G., Fainsod, A., 2005. Ethanol exposure affects gene expression in the embryonic organizer and reduces retinoic acid levels. Dev. Biol. 279, 193–204. Zhang, Z., Huynh, T., Baldini, A., 2006. Mesodermal expression of Tbx1 is necessary and sufficient for pharyngeal arch and cardiac outflow tract development. Development 133, 3587–3595.

Chapter 8.1

Origin of the Vertebrate Endothelial Cell Lineage: Ontogeny and Phylogeny Ramón Muñoz-Chápuli and José M. Pérez-Pomares Department of Animal Biology, Faculty of Science, University of Málaga, Spain

I.  Introduction Circulatory systems are present in most metazoans, and are in charge of sustaining animal metabolism and homeo­ stasis. In small-sized animals (around 1 mm in thickness or diameter) the major portion of metabolic substrates are obtained and exchanged by diffusion through the body walls (Brusca and Brusca, 2003). This mechanism, which is slow and not especially efficient, becomes useless in bigger animals. The reason is evident; the surface:volume ratio decreases and diffusion is no longer a reasonable physical mechanism to capture and distribute oxygen and nutrients, and to eliminate metabolic waste. In this sce­ nario, the inner cavities of the animal constitute an ideal support to outline inner systems of transport. Two main types of circulatory systems exist in animals, based on two kinds of cavities, coelomic and hemal. The term “circulatory system” is frequently used only for hemal systems, but this restriction ignores how important coe­ lomic cavities are for the performance of circulatory func­ tions in some animal phyla, such as echinoderms. Hemal systems are constituted of vessels which, in invertebrates, are always lined by extracellular matrix, normally the basal lamina of adjacent epithelia (Fig. 1). This feature, known as the R&C model (Ruppert and Carle, 1983), reveals that the origin of hemal cavities resides in the opening of spaces between the endodermal and coelomic epithelia. In vertebrates, however, a complete cellular lining is always present in the inner surface of vessels (Fig. 1). This cell lining constitutes a true, continuous and coherent layer of epithelial cells (polarized and tightly adhered between them) called endothelium. Although the endothelium, as defined above, is only present in vertebrates, no theory has hitherto been formulated about the origin of this cell type. Therefore, in order to formulate a working model about the evolutionary origin of endothelial cells from precursor cells Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

already present in invertebrates, we have gathered com­ parative molecular and cellular information on the embryo­ logical and adult features of metazoan circulatory systems. The aim of this chapter is to introduce this evolutionary model about the origin of the endothelium and, in general, about the transition between the invertebrate and vertebrate circulatory systems. We think that this model can explain

Figure 1  Invertebrate and vertebrate vessels. Invertebrate vessels (left) are constituted of hemal spaces (H) open between the basal laminae of endoderm (EN) and mesothelium (ME) or between adjacents mesothe­ lia in the mesenteries (M). In the left bottom photograph of an annelid, all these elements can be seen. Note the thickened basal lamina (BL) of the mesenteric vessel lining the hemal space. In vertebrates (right), however, endothelial cells (EC) always line the inner surface of the basal lamina, and the vessels are surrounded by perivascular cells, pericytes or smooth muscle cells (SMC) and fibroblasts (F). The right bottom photograph of a mouse coronary artery show immunolocalization of SMC specific -actin (green) and the endothelial adhesion molecule PECAM-1 (red). The space between endothelium and perivascular cells is an artifact.

465

466

many of the peculiar features of the endothelium, and also of its origin and differentiation during normal embryonic development. We are aware that “invertebrates” cannot be considered a monophyletic group, and for this reason we will not use the term with a taxonomic meaning. However, the fact that vessels lined with a true endothelium are only found in vertebrates, lead us to use “invertebrates”, for the sake of simplicity, to refer to the set of animal taxa bearing ves­ sels devoid of endothelium. Before starting to expose the comparative data, it is important to define precisely the meaning of a number of terms that we will use along this chapter. These terms are: Circulatory system A system of cavities that allow for the active movement of fluids through the body, pro­ pelled by some kind of mechanism. There are two types of circulatory systems, coelomic and hemal, and usually only one of them assumes the main circula­ tory functions in each animal phyla. Thus, nemerteans, arhynchobdellid hirudineans and echinoderms show well-developed coelomic circulatory systems, while the circulation in cephalopods, cephalochordates or verte­ brates occurs exclusively through hemal systems. Coelomic circulatory system A network of cavities derived from the embryonic coelom (the cavities of mesodermal origin present in many animals) through which the coelomic fluid circulates. These cavities are lined by a mesothelium (i.e., an epithelium of meso­ dermal origin) commonly referred to as the coelomic epithelium. This mesothelium, which is frequently cili­ ated, can be a peritoneum or a myoepithelium, as we will see below. The movement of the coelomic fluid usually occurs by ciliary beating, but it can also be pro­ duced by the body movements or by contractions of the coelomic myoepithelium. Hemal circulatory system A network of cavities originally limited by the basal lamina of two adjacent epithelia, both of which can be of mesodermal (i.e., coelomic) origin, as it occurs in the mesenteries or in the gonads, or both can have different blastodermic origins (meso­ dermal or endodermal). Although coelomic cells invest hemal circulatory systems, the difference with respect to the coelomic circulatory systems is clear-cut. The lumi­ nal surface of the coelomic cells in hemal systems is always basal, whereas in coelomic circulatory systems, the apical surface is directed to the lumen. It follows that ciliary movements cannot be involved in propul­ sion of the hemal fluid, which either occurs by contrac­ tion of the vessel walls or by the pumping of a heart. From this definition, it is obvious that hemal systems of invertebrates are originally related to the digestive tract, since they originally form between endodermal and coelomic epithelia. Thus, the extension of the hemal circulatory systems to other body regions occurs either

PART  |  8  Making Vessels

by connection with nonhemal cavities (see “open hemal circulatory systems” below) or, in a few cases, through microvessels which are able to invade and supply non­ visceral territories (for example somatic muscle or neu­ ral organs). Except for vertebrates, the microvessels are always formed by myoepithelial cells of coelomic ori­ gin which constitute tubes keeping their basal surfaces limiting the vascular lumen. The mechanisms by which these cells are able to invade the tissues and organize themselves in microvessels (i.e., the invertebrate “ang­ iogenesis”) are unknown. Open and closed hemal circulatory systems We will talk about closed hemal systems when they are composed by a continuous network of arteries, veins and capil­ laries interconnected between them. The hemal fluid is always in contact with the basal lamina of the vascular walls or, in vertebrates, with the endothelium. In open circulatory systems the vessels open into coelomic or coelom-derived cavities (hemocoels), or into interstitial spaces in the connective tissue. Mesothelium We think that the term “mesothelium” can be misleading, and we have restricted the use of this term in this chapter. Mesothelium is a rather general term which is normally used to refer to the cell layer which invests the coelomic cavities. However, this use of the term refers to many different cell types, includ­ ing at least the multipotential embryonic coelomic cells and the adult myoepithelial and peritoneal cells, which also show heterogeneity depending on their location in the body. In a broader sense, mesothelium has even been used for any mesodermal epithelium, including endothelial and renal epithelia. Thus, we have used “mesothelium” as a general term, using more specific terms when necessary, such as “embryonic coelomic epithelium”, “myoepithelium” or “peritoneum”. Myoepithelium A set of polarized cells originally forming part of the coelomic lining, and containing myofila­ ments which render them contractile. They are often ciliated. Myoepithelial cells also constitute the micro­ vessels of invertebrates. In this case, the apical sur­ face is oriented towards the embedding tissue and the cilia are absent, but the remaining features are similar. Probably myoepithelial cells were the original lining of the coelomic cavities, being the progenitor cells of both the peritoneal and visceral smooth muscle cells, as dis­ cussed below. Endothelium A set of polarized cells joined by intercellu­ lar junctions, anchored to the inner basal lamina of the vessels and thus lining the inner surface of the hemal system. According to this definition, only the verte­ brates have a true endothelium, since other “endothe­ lia” described in invertebrate representatives are either not joined by intercellular junctions (e.g., cephalopods or cephalochordates) or are not part of a hemal system (e.g., nemerteans).

Chapter  |  8.1  Origin of the Vertebrate Endothelial Cell Lineage: Ontogeny and Phylogeny

II.  A Brief Synopsis of the Circulatory Systems In Metazoans II.A.  Coelomic Circulatory Systems In the earliest animal phyla, which can be currently rep­ resented by the cnidarian or the plathyhelminth bauplan, circulation of nutrients, oxygen and catabolites occurs through free diffusion across the intercellular medium (Fig. 2). In this way, substances absorbed through the digestive epithelium freely diffuse by the extracellular matrix of the compact body walls and reach the mesen­ chymal and ectodermal cells. Oxygen, carbon dioxyde and nitrogenated products from catabolism are also dissolved in the medium which surrounds all these cells. This sim­ ple diffusion mechanism is slow and unefficient, and this limits the volume and/or the activity of these animals. In the cnidarians, the semifluid nature of the mesoglea, i.e., the extracellular matrix located between the epidermis (of ectodermal origin) and the gastrodermis (derived from endoderm), facilitates diffusion and allows for large sizes in some jellyfish. In platyhelminths, however, there is a cell layer located between the epidermis and gastrodermis. This layer is the mesenchyme, and it arises from the embryonic mesoderm, cells which migrate from the ectodermal or endodermal layers and characterize all bilateral animals. Coeloms provided a major improvement for the deliv­ ery of nutrients and removal of metabolic waste in animals.

467

The existence of fluid-filled cavities in the body, coupled with a mechanism for propulsion of the fluid, improved the diffusion of substances, and in fact most excretory systems of invertebrates filter the coelomic fluid remov­ ing all waste emanating from the body. Other functional advantages of the coelom are to provide the animal with a hydrostatic skeleton, or to facilitate the peristaltic contrac­ tions of the gut and heart (Brusca and Brusca, 2003). Coelomic cavities appeared in mesenchymal tissues of mesodermal origin and are lined by a mesodermal epithe­ lium (the coelomic epithelium or mesothelium) (Ruppert, 1991). A number of animal phyla are characterized by cavities in the mesenchymal tissues that are not lined by epithelium (Fig. 2). These animals are known as pseu­ docoelomates (or blastocoelomates; Brusca and Brusca, 2003), and they were first considered as intermediate between the acoelomate condition (that of platyhelminths) and the true coelomates. However, it is possible that pseu­ docoelomates have derived from coelomates which have secondarily lost the coelomic lining, leaving a set of cavi­ ties between the digestive tube and the ectodermal deriva­ tives (reviewed in Halanych, 2004; Jenner, 2004). It is important to distinguish two main types of coe­ lomic linings (Ruppert, 1991). Peritoneum is a simple, noncontractile mesothelium which invests the adult coe­ lomic cavities of vertebrates and members from other phyla (chaetognaths or sipunculans), and lies behind the visceral or somatic muscles. However, in hemichordates,

Figure 2  Phylogeny of the coelomic and hemal circulatory systems. The central nervous system of vertebrates is shown in dark gray. Only annelids, cephalopods and vertebrates show extensive vascularization of the somatic tissues, consisting of coelomic-derived myoepithelial cells in annelids and cephalopods, and endothelial/perivascular cells in vertebrates (color key: blue: ectoderm; gray: connective tissue; yellow: endoderm; pink: muscle; green: coelomic epithelium; white: inner cavities; orange: hemal spaces; red: endothelium).

468

some echinoderms, lophophorates and many annelids the coelomic lining consists of a myoepithelium, endowed with actin/myosin filaments, often ciliated and with con­ tractile ability. It is thought that this myoepithelium could be the most primitive type of coelomic lining in bilat­ erians, being a peritoneum-derived condition (Ruppert, 1991). The origin of the noncontractile peritoneum from the original contractile myoepithelium has been explained by the progressive muscular specialization of a subset of myoepithelial cells, finally delaminating from the epi­ thelium and gathering into a submesothelial layer of vis­ ceral muscle. Other subsets of myoepithelial cells would retain the ciliated condition, becoming peritoneal cells. A pseudostratified coelomic epithelium, formed by two types of cells, could have been an intermediate condi­ tion before separation of the peritoneal and the muscular cell lineages (Rieger and Lombardi, 1987; Ruppert, 1991). The close relationship between coelomic cells and the visceral smooth muscle is illustrated by the reported mus­ cular regeneration in adult holothurians through transdif­ ferentiation of the coelomic epithelium (Dolmatov and Ginanova, 2001; Murray and García-Arrarás, 2004). In some cases, the ability of embryonic coelomic cells to dif­ ferentiate into muscular cells leads to a massive increase in muscular volume and obliteration of the coelomic cavity. This happens during somitogenesis in cephalo­ chordates and vertebrates, but a similar phenomenon has also been described in phoronids and hemichordates (Ruppert, 1991). Coelomic circulatory systems are well-developed in echinoderms and arhynchobdellid hirudineans (Fig. 3). In other cases, such as in arthropods and brachiopods, coelomic systems fuse with hemal systems. A special case among the coelomic circulatory systems is that of nemerteans, considered to constitute an acoelomate phy­ lum for a long period of time, although their circulatory system was considered to be strikingly advanced (Fig. 3A,B). The nemertean circulatory system is closed and shows a complete inner cellular lining. In these marine worms, two main lateral “vessels” are joined by anter­ior and posterior transverse commissures. The vessels are composed of an inner continuous epithelium, a subepithe­ lial extracellular matrix layer and a muscle layer of circu­ lar fibers. Cells of the lining epithelium (originally named “endothelium”) are squamous, and they are joined by adherens junctions or septate desmosomes. This unique endothelium shows rudimentary cilia and myofilaments. The embryonic development of these exceptional ves­ sels was investigated ultrastructurally in Prosorhochmus (Turbeville, 1986). According to this author, vessels are formed by hollowing of a solid mesodermal band (schizo­ coely) in the same way as the coelom develops in spiralian representatives such as annelids. This, together with the lateral position of the main channels, adherens junctions,

PART  |  8  Making Vessels

Figure 3  Examples of coelomic and hemal circulatory systems in invertebrates. (A, B) Coelomic channels (COE) in the nemertean Cerebratulus. Transverse section. These channels are mainly located at the sides of the intestine (I), and they branch among the muscles (M). Circulating cells (coelomocytes) can be seen inside the channels (arrow in (B)). The epithelial cells lining these channels are shown by an arrow­ head in (B). (C) Contractile dorsal vessel (DV) in an oligochaete annelid. The myoepithelial cell layer (ME), continuous with the dorsal mesentery (DM), surrounds a wide hemal space. Amoebocytes (arrows) can be seen adhered to the basal lamina of the myoepithelial cells (COE: coelomic cavity). (D) Hemal spaces (HE) around the intestine (I) of the cephalo­ chordate Branchiostoma are much reduced. Note the thin myoepithelial cell layer (ME) and the wall (AW) of the atrium (AT), a specialized coe­ lomic cavity. Adherent amoebocytes (arrows) are the only blood cells present in the hemal cavity.

cilia and myofilaments, clearly demonstrated that these “vessels” are coelomic channels, and the inner lining, formerly considered as an endothelium, is a typical coe­ lomic epithelium (Turbeville, 1991). This finding repo­ sitioned nemerteans among coelomate Protostomozoa. Therefore, nemerteans constitute a good example of a coelomic circulatory system.

II.B.  Hemal Circulatory Systems We have already indicated that the perivisceral coelomic cavities improved the circulation of nutrients. Substances absorbed through the gut epithelium reach the coelom, and the coelomic fluid distributes them as far as these cavities can reach in the body. Loss of contact between the basal surfaces of the digestive and the coelomic epithelia was the probable origin of the hemal circulatory systems (Figs 2; 3). Hemal systems of invertebrates are originally constituted by a network of rudimentary cavities located between the basal laminae of the digestive and the coe­ lomic epithelia. Some factors probably contributed to the evolutionary success of these cavities which constitute the

Chapter  |  8.1  Origin of the Vertebrate Endothelial Cell Lineage: Ontogeny and Phylogeny

main circulatory systems in a number of phyla, depriving the coelom of this function. Among these factors we can consider the proximity of the hemal cavities to the endo­ dermal cells (it is the hemal fluid, and not the coelomic fluid, that receives the substances absorbed through the gut directly) and, especially, the ability of the primitive coelomic myoepithelium to contract, which allows pump­ ing of the hemal fluid and improved circulation of nutri­ ents. As a matter of fact, the myoepithelial nature of the invertebrate large vessels accounts for their contractile ability, constituting “secondary hearts”. This is the case, for example, of the endostylar artery, the hepatic vein or the contractile bulbs of the pharyngeal arteries of the cephalochordate Branchiostoma (Moller and Philpott, 1973). Although we have defined the hemal systems as the network of cavities originally developed between the basal laminae of two adjacent epithelia, at least one of these epithelia being coelomic in nature, hemal or hemal-like cavities have been described in acoelomate and pseudocoe­ lomate invertebrates (that formally lack of coelom). Thus, a so-called “lymphatic system” has been described in platy­ helminthes (Fried and Haseeb, 1991) composed of intercon­ nected, fluid-filled blind sinuses and channels. In the worm Megalodiscus these sinuses are lined by a syncytial epithe­ lium (Strong and Bogitsh, 1973). These cavities are simi­ lar to those reported in pseudocoelomates as we described above. However, in all these cases, cavities devoid of cell lining cannot be considered as true circulatory systems, since they lack an active mechanism for fluid propulsion.

II.C.  Anatomy of Coelomic and Hemal Invertebrate Circulatory Systems II.C.i.  Prostostomozoa Lophophorates are thought to be phylogenetically close to the Protostome–Deuterostome ancestor. In this group, two phyla with very different types of circulatory systems are included. Phoronids show a closed circulatory system com­ posed of vessels lacking of endothelium. Therefore, blood contacts the basal lamina of the coelomic epithelium. Blood is pumped by rhythmic waves of contraction of myoepithe­ lial cells of a median vessel wall, and contains red cells carrying hemoglobin (Herrmann, 1997). Circulation in brachiopods, instead, proceeds through a system of ciliated coelomic channels which connect with an open hemal sys­ tem (James, 1997). Extensions of the coelom form a series in the mantle, lophophore and even the pedicle. Sporadic muscular contractions originating from ciliated coelomic myoepithelia move the coelomic fluid (Chuang, 1964). The hemal system includes contractile vessels (referred to as “hearts”) located in the dorsal mesentery of the stom­ ach region that branch and flow into sinuses. Capillary-like

469

vessels have also been described, composed of a single layer of squamous myoepithelial cells. Arthropods also have a variety of circulatory systems. In crustaceans there are large and small arteries, as well as capillaries (Martin and Hose, 1992; Chan et al., 2006). According to these authors, these vessels are formed from the lumen to the exterior, by a thick basal lamina or inner intima, a so-called “endothelial” layer, a loose connective tissue layer and an outer limiting intima. The endothelial layer is described as a single layer of cells with highlyconvoluted basal and lateral plasma membranes, suggest­ ing that they stretch and shorten with changes in diameter of the vessel lumen. These authors state that the “endo­ t­helial cells” probably contract. The localization of these cells (abluminal with respect to the vascular basal lamina) and their contractility strongly suggest that they are typi­ cal myoepithelial cells. In some vessels a single layer of striated muscle adjacent to the inner intima appears (e.g., in the dorsal abdominal artery), and in some species this artery contracts. In insects, the hemal system is much reduced. In fact, it is restricted to the pulsatile dorsal ves­ sel or “heart”, which pumps hemolymph into the hemo­ coel, i.e., into the remnants of the coelomic cavity. The insect dorsal vessel is composed of two types of cells, the inner contractile cells, aligned in two longitudinal rows, and the pericardial cells whose function is poorlyunderstood (Bodmer and Frasch, 1999). The lack of a developed hemal circulatory system in insects is compen­ sated for by an extensive network of intercellular lymph spaces surrounding all the organs, together with a set of pulsatile structures, probably derived from local muscles, which move the hemolymph throughout the body. This is a unique and probably derived condition of insects (Locke, 1998; Pass, 1998). Among molluscs, cephalopods have a virtually closed hemal circulatory system, composed of arteries, veins and capillaries (Schipp, 1987). This hemal system, probably the best-developed of all the invertebrates, is definitely related to the large size attained by some cephalopods such as the giant squids, which are the biggest invertebrates known. All the vessels show a continuous basal lamina and a discontinuous layer of endothelial cells (almost con­ tinuous in the aorta). These cells never show intercellular junctions, and they seem to migrate freely over the luminal surface of the vessels. Apparently, these endothelial cells discharge dense bodies in the abluminal surface by exocy­ tosis to form the matrix of the basal lamina (Schipp, 1987). The main arteries show a thick medial layer composed of muscle cells. Small vessels are formed by myoepithelial cells (with myofilaments), pericytes (without myofila­ ments) and sometimes by a single pericyte, always with a continuous, thick basal lamina in the luminal surface. Nearly all the veins are contractile, ensuring blood back­ flow. Branchial hearts derive from “caval” veins, receive

470

deoxygenated blood and pump it into the gills. They are also excretory and show a discontinuous layer of endothe­ lial cells. The hemal system from oligochaete annelids fits well with the classical R&C model, since they are lined by the basal lamina of myoepithelial cells of coelomic origin (Jamieson, 1992). Adherent cells are occasionally present, partially covering their inner surface. Hirudineans show modifications of this basic model, since the hemal system coexists, in acanthobdellids and rhynchobdellids, with a hemocoelomic system, a network of closed, intercon­ nected cavities of coelomic origin (Fernández et al., 1992) (Fig. 3C,D). Connections between both systems have not been found. The hemocoelomic system consists of four major longitudinal sinuses whose coelomic cell lining is interrupted by chloragogen cells, a type of immunocom­ petent, nonphagocytic coelomocyte (Cholewa et al., 2006). Circulation of the coelomic fluid occurs by contraction of the body walls. The vessels are dorsal and ventral, with a contractile area in the dorsal vessels (heart). The vessel wall is composed of an inner muscle layer (whose basal lamina is lining the vascular lumen), a middle connective tissue layer and an outer layer formed by the coelomic epi­ thelium. Arhynchobdellid hirudineans lack blood vessels, and the circulatory system is limited to the hemocoelomic network of channels. In some channels the epithelium is replaced by large cells containing brown pigment (botry­ oidal tissue) that are thought to be similar to chloragogen cells. We will deal with the importance of this tissue in vascular growth below.

II.C.ii.  Deuterostomozoa In echinoderms the open hemal circulatory system is poorly-developed, as we would expect from an ani­ mal group in which the coelom is complex and highlydeveloped, forming a phylum-specific, fluid-filled sytem (the ambulacral system). The hemal system consists of channels that unite the perivisceral sinuses. The axial organ, which is an enlarged blood vessel connecting the portions of the hemal system, is composed of myoepithe­ lial cells. The function of the axial organ seems to be to pump the hemal fluid, but its performance is poor. When there are well-defined vessels (e.g., in holoturians) they are always lined by a coelomic myoepithelium which is contractile in some places. In these echinoderms, cells adhered to the inner basal lamina of the vessels have been described and called “fibroblasts” (Doyle, 1964; Smiley, 1994), a type of cell that has been described in other taxa as “amoebocytes”, i.e., circulating blood cells with the ability to adhere to the inner basal lamina of the vessels. The hemal circulatory system of hemichordates is more developed than that in echinoderms. Two contractile longitudinal blood vessels, dorsal and ventral, are formed by the coelomic myoepithelial cells from the mesenteries.

PART  |  8  Making Vessels

There is a heart vesicle in the head, formed by a sac of the coelomic epithelium which contracts and presses a hemal sinus against the stomochord (a small compartment of the anterior intestine). The hemal circulatory system of urochordates is restricted to a tubular heart (described below) which pumps the blood fluid into a perivisceral, noncoelomic cavity (Burighel and Cloney, 1997). In contrast, the hemal circu­ latory system of cephalochordates is well-developed and similar to that of vertebrates in their “closed” nature, and also in its anatomical pattern (Ruppert, 1997). Cells called “endothelial” have been reported adhered to the inner basal lamina of the vessels, as described below (Moller and Philpott, 1973). These cells are more or less abun­ dant, but always lack cell junctions, even in the rare cases when the cells are in contact. Vessels are located through­ out the body, and they are formed by spaces between the basal lamina of adjacent epithelia (usually endodermal and coelomic, as in the gills or viscerae) (Fig. 3). Some ves­ sels are limited by a fibrilar layer of collagen produced by the connective tissue. This is the case in the segmen­ tal vessels of the myotomes, which are comprised between the atrial (coelomic) epithelium and the dense collagenic layer of the myosepta. These segmental vessels branch in microvessels which penetrate within the myosepta. Thus, the vascular lumen is always limited by the extracellular matrix, either the basal lamina or a collagenic mesh, and the hemal circulatory system is basically closed.

II.D.  Blood Cells in Invertebrates Blood cells should be considered as an important part of the cardiovascular system. In vertebrates the ontogenetic relationship existing between blood and endothelial cells is well-known (see below). This parallel cannot be estab­ lished in invertebrate animals, because these lack a true endothelium but, as will be discussed, blood cells of non­ vertebrate metazoans have a tight relationship to coelomic tissues. Two types of circulating cells have been described in invertebrates, i.e., coelomocytes and hemocytes, which circulate through the coelomic and hemal spaces, respec­ tively (Fig. 3). However, no clear distinction has been found between these cell types, and it is generally agreed that circulating cell can actually move between the coe­ lomic and the hemal compartments when these compart­ ments are connected (Hartenstein, 2006). Thus, the terms coelomocyte or hemocyte reflect the location of the cell, rather than their origin or function. The diversity of hemocytes and coelomocytes in inver­ tebrates is astonishing, and virtually all the functions acco­m­ plished by vertebrate blood cells have been found in one or other type of invertebrate blood cells, including blood coag­ ulation, phagocytosis, antigen recognition, immunogl­obulin production or transport of respiratory pigments (reviewed in

Chapter  |  8.1  Origin of the Vertebrate Endothelial Cell Lineage: Ontogeny and Phylogeny

Hartenstein, 2006). This author proposed a classification in four main types: prohemocytes (progenitor cells); hyaline hemocytes (plasmatocytes or monocytes); granulocytes; and eleocytes (chloragogenocytes). We will use these names in this chapter. Hemoglobin-containing blood cells have been found in nemerteans, bivalves, annelids, echiurans, phoronids, echinoderms and vertebrates (Vinogradov et al., 1993). The vertebrate erythrocyte seems to be exceptional, since in the remaining phyla hemoglobin is packed in hyaline hemocytes (Hartenstein, 2006). In other phyla such as bra­ chiopods or arthropods, blood cells carry other respiratory pigments, such as hemerythrin or hemocyanin. However, in polichaete annelids hemoglobin is synthesized in coe­ lomic cells such as the heart wall and chloragogen tissue (Friedman and Weiss, 1980). In hemichordates and some echinoderms, hemoglobin is dissolved in the extracellular medium, and it is also synthesized and secreted by coe­ lomic epithelial cells (Smiley, 1994; Benito and Pardos, 1997). On the other hand, free hemoglobin is contained in the interstitial fluid of the pseudocoelomate nematode Nippostrongylus (Sharpe and Lee, 1981) and even in acoe­ lomate trematodes (Haque et al., 1992), raising questions about the location where this pigment is synthesized. A special type of hyaline hemocyte has been described as the “amoebocyte”, a cell which is able to adhere to the inner basal lamina of the vessels. It has been described under different names in annelids, phoronids, holoturians (as “fibroblasts”, Smiley, 1994), brachiopods (phagocytic and carrying granules, Pan and Watabe, 1989), hemichor­ dates or cephalochordates (as “endothelial cells”, Ruppert, 1997). Amoebocytes in hemichordates show sometimes one or two centrioles associated with striated rootlets, sug­ gesting a ciliated, coelomic origin (Benito and Pardos, 1997). In cephalochordates, the amoebocyte or endothe­ lial cell (Fig. 3D) is the only circulating cell type which has been described in the hemal spaces, although cilated coelomocytes have been reported to exist in the coelom of cephalochordates (Rhodes et al., 1982; Zhang, 1992). In large vessels of cephalochordates, amoebocytes form a continuous, squamous lining, but lack of intercellular junc­ tions. Therefore, they do not form a real epithelium. They are endocytic, and able to remove exogenous protein from blood (Moller and Philpott, 1973). According to these authors, another function of these cells is policing the luminal surface of the vessels. In fact, their ultrastructural features (cell processes with microtubules oriented paral­ lel to the long axis of the cell) suggest that these cells are free-floating (circulating) ones.

II.E.  The Invertebrate Heart The invertebrate heart must be interpreted in the context of the basic plan of hemal systems, i.e., composed of vascular walls limited by a myoepithelium with contractile abilities.

471

Thus, the heart is established in a precise area of the myoepithelium where contractions become autonomous, stronger and organized, pumping the hemal fluid in the preferred direction. This special area of the vascular sys­ tem could, in theory, be placed in any part of the hemal circulatory system, as illustrated by the dorsal position of the heart in protostomes, or its ventral location in deuter­ ostomes. However, this location is probably due to the for­ mation of a new, secondary mouth in deuterostomes. This major evolutionary event reversed the primitive dorsal and ventral faces of the deuterostomozoans. In conclusion, although the heart could develop in different places in the body, its location has been well-conserved throughout evo­ lution, revealing a relationship between cardiogenesis and the establishment of the dorsoventral axis of metazoa. Besides its location in the body, the metazoan heart organs show a structural bauplan common to most inverte­ brates (for more detail see Chapter 1.1). The cardiac mus­ cle can retain the original myoepithelial morphology (as in urochordates), or it can become a layer of striated muscle cells as happens in crustaceans (Martin and Hose, 1992) or in cephalopods (Budelmann et al., 1997). The heart can be tubular, as in insects (Bodmer and Frasch, 1999), or multi­ chambered as in cephalopods (Budelmann et al., 1997), but in all cases the cardiac wall is derived from the coelomic wall, and a basal lamina separates this wall from the blood. Frequently, the heart is just a widened myoepithelial tube continuous with the peritoneum and showing rhythmic contractions, as has been described in annelids and brachio­ pods (Martynova and Chaga, 1997, 2002) (Fig. 3E). Among the deuterostomozoans, the echinoderms lack a heart, although the axial organ has been considered as a homologous structure of the chordate heart by Nielsen (1995). However, it is difficult to consider the axial organ as a heart, since it has no pumping power to circulate the blood. As stated above, the so-called “heart” of hemichordates is located in the head, and it consists of a vesicle of coelomic myoepithelium which contracts and presses a hemal sinus against the rigid stomochord. Its position in the body sug­ gests that it is not a true heart. The localization of the heart of urochordates is more similar to that of vertebrates, ventral and posterior to the pharynx (see Chapters 7.1 and 7.2). This heart is formed from a compact mass of cells which enlarge to form a hollow cylindrical vesicle that later invaginates, giving rise to a tube within a tube. The inner tube (called “myocardium”) is a simple, contractile myoepithelium. The outer tube is the “pericardium” and it is also a simple, noncontractile coelomic epithelium (see Chapter 5.1). The luminal surface of the myocardium is lined by a basal lam­ ina (Burighel and Cloney, 1997). In cephalochordates, how­ ever, a distinct heart is absent, but many large vessels are contractile, as happens in other groups. The contractility and the larger diameter of the posterior area of the subendostilar vessel has led to the proposition that this area is homologous to the vertebrate heart, a hypothesis that is consistent with

472

the expression of AmphiNk2-tin, a gene similar to the verte­ brate NK2 cardiac genes (Holland et al., 2003).

II.F.  Development of the Invertebrate   Hemal Circulatory System The main elements of the invertebrate circulatory system, myoepithelial cells and hemocytes, originate from the coe­ lomic epithelium. The myoepithelial cells which form the large vessels, as well as the microvessels of invertebrates, derive from the embryonic coelomic lining and their ontogenetic origin is related to the process of coelomoge­ nesis (Ruppert, 1991). On the other hand, the embryonic origin of blood cells in invertebrates has recently been reviewed by Hartenstein (2006). This author concludes that the ontogeny and phylogeny of blood cell progeni­ tors is also related to the coelomic epithelium. In policha­ ete annelids, for example, the adult coelomic epithelium proliferates in specialized domains and buds off hemo­ cytes into the coelomic lumen (Dales and Dixon, 1981). A delamination of phagocytic cells from the coelomic lin­ ing was also described in adult echinoderms (Bossche and Jangoux, 1976). This coelomic origin explains the frequent presence of cilia or centrioles in echinoderm coelomocytes (Holland, 1965; Mae and Jangoux, 1983). In other cases, the presence of specialized hemopoietic organs makes it more difficult to know where the hemocyte progenitors arise from. This is the case of the white bodies in cephalo­ pods and the lymph glands of arthropods, oligochaetes or ascidians (Martin and Hose, 1992; Claes, 1996). In these cases it is conceivable that hemopoietic stem cells have migrated out of the coelomic epithelium and coalesced in specialized hemopoietic organs (Hartenstein, 2006).

II.G.  Vertebrate Endothelium is an Exception in the Animal Kingdom As we have shown in the preceding sections, circulatory systems lined with a true endothelium are not known in invertebrates, although in three different cases (nemerteans, cephalopods and cephalochordates) “endothelial-like” cells have been described. None of these cases can fit the defi­ nition of an endothelium, i.e., a layer of mesodermal, epi­ thelial cells joined by intercellular junctions and anchored to a vascular basal lamina through hemidesmosomes. Not surprisingly, many of the classic molecular biology charac­ teristics used to define endothelial cells are directly related to their functional properties. Endothelial cells secrete von Willebrand factor (vWF), a glycoprotein that acts a carrier of factor VIII. vWF can accumulate in the form of WeibelPalade bodies (WPb), a relatively large rod-shaped corpus­ cle when seen by transmission electron microscopy (TEM). TEM studies reveal that the presence of WPb, caveolae, reduced cytoplasm and a nucleus with an elliptical section

PART  |  8  Making Vessels

are the most typical ultrastructural features of endothelium (Bouïs et al., 2001). Endothelial cells have angiotensinconverting enzyme activity, secrete prostaglandin I2, syn­ thesize a whole series of molecules with antithrombotic (2-macroglobulin) and thrombotic (plasminogen acti­ vator, tissue factor) properties, and are capable of 5-HT (serotonin), catecholamine, acetylcholine, adenosine and acetylated low density lipoprotein (acLDL) uptake. Finally, this cell type has a specific cell membrane profile, character­ ized by the presence of vascular endothelial (VE)-cadherin and CD31/platelet endothelial cell adhesion molecule (PECAM; that participate in endothelial cell-to-cell attach­ ment, together with other molecules such as ICAM, VCAM and E-selectin), as well as expression of specific tyrosine kinase receptors like VEGFR-2/Flk-1 (vascular endothe­ lial growth factor receptor type-2) or Tie-2 (angiopoietin-1 receptor) (Johnson et al., 1977; Holthöfer et al., 1982; Edgell et al., 1983; Voyta et al., 1984; Pober et al., 1986; Dejana et al., 1996; Bouïs et al., 2001). The close relationship that exists between the specific activities of endothelial cells and both their morphologi­ cal and molecular phenotype is evident when comparing endothelial cells from big vessels (macroendothelium) to those of small (capillary) ones (microendothelium). From a structural point of view, blood–brain barrier endothe­ lial cells display tight cell-to-cell junctions that transform this endothelium into a highly selective one, while liver sinusoidal and renal glomerular endothelium have fenes­ trations that are quite permeable. At the molecular level, microendothelial cells tend to show a diffuse cytoplasmic distribution of vWF with no WPb as happens in embry­ onic endothelium. They can synthesize collagen IV, but not other extracellular matrix glycoproteins like elastin and enactin. Microvascular and macrovascular matrix met­ alloproteinase (MMP)-expression and lectin binding (Del Vecchio et al., 1992; Magee et al., 1994; Abdi et al., 1995; King et al., 2004) are also different (Jackson and Nguyen, 1997). Finally, type III transforming growth factor  (TGF) receptor is preferentially expressed by microvas­ cular endothelial cells (Morello et al., 1995). The vertebrate endothelium is exceptional, and the knowledge of its embryonic development can supply us with clues about its evolutionary origin, in an Evo–Devo approach. We will now briefly review the current knowledge about the differentiation of endothelial cells in vertebrates.

III.  The ontogeny of the endothelium III.A.  The Basic Anatomy of Vertebrate Vasculature Vertebrates are the only animals that have an endothelialbased circulatory system formed by a complex network

Chapter  |  8.1  Origin of the Vertebrate Endothelial Cell Lineage: Ontogeny and Phylogeny

of arteries, veins and capillaries that support a continuous stream of blood propelled by a chambered heart. As indi­ cated above, all these vessel types are basically formed by a continuous, flattened, nonstratified epithelium of meso­d­ermal origin called endothelium. Endothelial cells “define” the vertebrate vascular system, so that a vertebrate vessel can lack smooth muscle or fibrous adventitial com­ ponent, but not endothelium. Therefore, vascular growth (angiogenesis or vasculogenesis, see below) is always, and first, defined by endothelial activity. Vertebrate arteries have thick, multilayered walls that are characteristically elastic. In the adult, three classic anatomical regions are found in an artery: an outer tunica adventitia, mainly constituted of fibrous cells; the tunica media, that includes circular and longitudinal smooth mus­ cle fibers; and the tunica intima, formed by the innermost endothelium and the underlying elastic fibers, which are normally absent early in development. Arteries are capa­ ble of peristaltic blood pumping, and some of them even have their own blood supply (vasa vasorum). Arteries con­ nect the heart to the capillary bed of tissues through arteri­ oles and act as a pressure reservoir, as well as an absorber system for the oscillations of the circulatory flow (Eckert et al., 1988). Capillaries are the smallest vessels of the body and are responsible for the cell-by-cell delivery of oxygen and nutrients. Capillaries consist of endothelial cells discon­ tinuously covered by supporting cells called pericytes. These pericytes are embedded within the endothelial basal lamina, and make focal contacts with endothelial cells (reviewed in Armulik et al., 2005). Pericytes not only appear in capillaries, but also in pre-capillary metarterioles and post-capillary venules. They are modulators of blood vessel development, maturation and remodeling, and are especially abundant in the microvessels of the central nerv­ ous system. Pericytes are closely related to smooth muscle cells (SMCs), and many authors favor the idea that a molec­ ular and phenotypical continuum ranging from pericytes to differentiated and functional smooth muscle cells exists, suggesting a complex reality that is even more obscure in the embryo (Hungerford and Little, 1999). Markers used to identify pericytes include -smooth muscle actin (-SMA), desmin, platelet-derived growth factor recep­ tor beta (PDGFR-), aminopeptidases A and N, and RGS5 (a protein involved in tuning heterotrimeric G-protein signaling by acting as GTPase-activating proteins for G subunits), but none of these markers is absolutely specific for pericytes, and many of them are also expressed by smooth muscle cells (Armulik et al., 2005). Very interest­ ingly, pericytes also have some similarities with endothelial cells. Studies in adult cancer malignancy have shown that both pericytes and endothelial precursor cells (EPCs) are capable of tube and network formation, as well as response to kinase inhibitors selective for angiogenic pathways. The expression of cell surface proteins, including PDGFR

473

VCAM, intercellular adhesion molecule, CD105, desmin and neural growth proteoglycan 2, was found to be very similar between pericytes and EPC (Bagley et al., 2005). Normally, any given cell is at a distance of no more than three cell bodies from a capillary, and in some cases capillaries can irrigate, one by one, all the cells of a tissue (e.g., in the heart). Arterioles connect to capillaries through metarterioles that normally end in a pre-capillary sphincter (a ring formed by a few smooth muscle cells) that controls the blood flow in a capillary bed. For this reason, under normal conditions, the blood content of all the capillar­ ies of an individual is only half (around 7%) of its total potential volume (around a 15% of all the blood of the body). Veins act as a return pathway for the blood circu­ lating from the capillaries to the heart. Veins have a great inner diameter, and normally present a less complex wall than arteries (some veins lack a well-developed muscular layer). It is known that at least 50% of the total volume of the blood of vertebrates is found in veins which can act as a reservoir of blood to sustain arterial pressure in case of hemorrhage (McDonald, 1960; Eckert et al., 1988). What advantage can the vertebrate endothelial cell type represent with respect to the nude vessels of invertebrates? Endothelial cells do not form a passive barrier, but can actively transport micro- and macromolecules. However, this control of molecular transportation might be a minor advantage when compared with other endothelial func­ tions. Endothelial cells play a major direct role in the regu­ lation of blood vessel contractility, blood coagulation and inflammatory responses. Finally, endothelial cells allow for guided blood vessel growth when needed.

III.B.  Mechanisms of Vascular Formation and Growth: Angiogenesis versus Vasculogenesis Vertebrate embryonic blood vessels develop through a mor­ phogenetic process termed vasculogenesis, a cellular mech­ anism that involves the coalescence of isolated endothelial cell progenitors of mesodermal origin (referred to as angio­ blasts) to form tubular endothelial structures of different caliber (González-Crussi, 1971; Risau and Lemmon, 1988; Poole and Coffin, 1989; Risau and Flamme, 1995; Risau, 1997). With time, these primitive vascular networks con­ tinue growing by angiogenesis, an alternative mechanism of vessel growth based on the invasive budding or division of pre-existing blood vessels (Hertig, 1935; Folkman and Klagsbrun, 1987; Risau and Lemmon, 1988). Therefore, angiogenesis, which will become the main normal and path­ ological mechanism of blood vessel growth and remodeling throughout adulthood, depends on embryonic vasculogen­ esis to take place. The endothelium is the primary cell type forming blood vessels in vertebrates. It has been proposed that an initial trigger from endodermal (or other) cell types is required in

474

order to initiate and then sustain effective vasculogenesis from angioblasts, but not for their differentiation from mes­ odermal progenitors (Mato et al., 1964; Wilt, 1965; Kessel and Fabian, 1986; Sugi and Lough, 1994, 1995; Lough and Sugi, 2000; Vokes and Krieg, 2000; Vokes et al., 2004). However, it is also true that the endothelium itself can lead the recruitment of endothelial progenitors from different origins if necessary. In later stage of vasculogen­ esis, the endothelium produces a diversity of molecules, including trophic factors and receptors that are involved, directly or indirectly, in the attraction of pericytes/smooth muscle progenitors toward the forming endothelial tubes, therefore initiating the formation of the vessel wall (Sato et al., 1995; Davis et al., 1996; Suri et al., 1996; Mais­onpierre et al., 1997; Hellstrom et al., 1999; Hirschi et al., 1999; Zhang et al., 2001; Oike et al., 2002). All these characteristic features of the endothelium are crucial to understanding its patterning potential. Endothelial cells follow a strict outline to create an optimally-efficient vas­ cular system with a minimal investment of energy. This, together with the still poorly-known molecular switch and signal transduction mechanisms that cause angiob­ last determination, can be in the origin of the geometri­ cal distribution of differentiating angioblasts. These are known to occur in the space forming planar polygones of pentagonal profile even before expressing molecules that reveal their full commitment to the endothelial line­ age (Drake et al., 1997; Drake and Little, 1999). However, the molecular and cellular laws that govern this pattern are not known. Seminal studies by Hans Meinhardt used reac­ tion–diffusion models to explain the formation of net-like structures (Gierer and Meinhardt, 1972; Meinhardt, 1982); these mathematical models, which can easily be simulated by computers, are quite accurate in the design of networks which are very similar to developing blood vessels. Further modification of these models has been able to formalize the recruitment or incorporation of new elements (cells) to the forming network (Andreucci et al., 2004). However, it is still difficult to explain which molecules are the longrange effect, highly-diffusible inhibitor and the local acti­ vator with low diffusion properties that the model requires.

III.C.  Fine Features of Embryonic Vasculogenesis The vasculogenic formation of any primary endothelial net­ work strongly relies on the properties of endothelial cells. Although certain molecular differences can be recorded among endothelial cells (mostly related to their arterial or venous nature), early angioblasts and endothelial cells are known to be plastic (Moyon et al., 2001). They interact between themselves, and their mutual molecular compati­ bility (e.g., expression profiles of adhesion molecules such as VE-cadherin) makes their coalescence straightforward.

PART  |  8  Making Vessels

This is extremely important, as the mechanism of vascu­ logenesis, as indicated above, depends strongly on the so-called “fusion” of groups of angioblastic cells that have previously formed endothelial colonies (blood islands, vascular chords), so that a cohesive network of vessels of different calibers can form (Drake and Little, 1995, 1999). The self-organizing properties of the endothelial tis­ sue rely on the ability of the primary unicellular vascular units (angioblasts) to associate into secondary multicel­ lular vascular units (blood islands, vascular chords) that will fuse, giving rise to vascular beds (Risau and Flamme, 1995; Risau, 1997). This tissue fusion depends on the cell proliferation rates of differentiating endothelial cells, cell-to-cell contact, extracellular matrix (ECM) production and cell-to-extracellular matrix interactions, and its final outcome is the appearance of a new, more complex tissue unit. Most importantly, it has been indicated that tissue fusion is an event that involves the union, but not neces­ sarily the mixing, of the individual cells that comprise the tissue (Pérez-Pomares and Foty, 2006). An important controversy directly related to the origin of the angioblastic cell population is that of the relation­ ship which exists between the origin of angioblasts and the locations where vascular formation takes place. Most evi­ dence indicates that the major part of embryonic vascular structures that form through vasculogenesis (but not all of them) do so from angioblasts differentiated in situ without a significant recruitment of vascular progenitors from adja­ cent locations, suggesting a reduced migration of angio­ blasts. This seems to be the case for many of the organs whose vascularization takes place through vasculogenesis, such as the aortic vessels and main veins (cardinal and vitelline veins), as well as the primary vascular networks of visceral organs like the lungs, kidneys, spleen, heart (endo­ cardium and coronary vessels), liver, pancreas and other elements of the digestive system (Poole and Coffin, 1991; Sherer, 1991; Tufro et al., 1999; Pérez-Pomares et al., 2002a,b; Kattan et al., 2004; Anderson-Berry et al., 2005). An evident exception is that of the nervous system, in which highly-migratory populations of angioblasts first pre-organize themselves through vasculogenesis in a perineural vascular plexus that then undergoes an extensive angiogenesis process to fully-vascularize all the neural tis­ sues (Noden, 1990, 1991). This may reflect the close rela­ tionship that exists between the endoderm and the origin of angioblasts, thus pointing to the division of the embry­ onic territories into two main regions, a ventral splanch­ nic one which produces high numbers of angioblasts that tend to form blood vessels in situ and a dorsal somatic one, where angioblast differentiation is reduced, that thus has to recruit cells from more ventral locations. Some ad hoc explanations to this situation have appeared in the lit­ erature like the classification of vasculogenesis into types I (in situ vasculogenesis) and type II (vasculogenesis from recruited angioblasts that migrate to the vascularizing

Chapter  |  8.1  Origin of the Vertebrate Endothelial Cell Lineage: Ontogeny and Phylogeny

region from distant locations) (Poole and Coffin, 1991). We believe these differences clearly reveal the phyloge­ netic origin of the circulatory hemal system closely related to the visceral areas, and its subsequent extension to the somatic domains, as we will discuss below.

III.D.  Origin and Differentiation   of Vertebrate Angioblasts The angioblasts are the vascular progenitors of the endothelial cells. The mechanisms that control the deter­ mination and differentiation of angioblasts are still subject to controversy. It is widely-accepted that the origin of these cells is mesodermal, but it still remains unclear whether all embryonic endothelial cells originate from a common original pool of angioblasts differentiated soon after gas­ trulation, or if different waves of angioblastic differentia­ tion take place in different locations as the embryo forms. The main hypothesis on the origin of angioblasts states that all the endothelial progenitors of the developing embryo come from a population of early primary mesodermal cells that acquire an angioblastic fate soon after, but independ­ ently from, gastrulation (Azar and Eyal-Giladi, 1979). Alternatively, recent findings suggest that the embryo may locally generate several generations of angioblasts from mesodermal undifferentiated cells at different moments of its development (Sherer, 1991; Pérez-Pomares and Muñoz-Chápuli, 2002). Whereas in the first case a sus­ tained proliferation of angioblasts would be required, the second hypothesis would suggest that high proliferation of angioblasts is not necessary for vasculogenesis to be initi­ ated. In our opinion, the literature published on this issue somehow overrates the significance of cell proliferation in early vascular development. It is commonly accepted that endothelial cells have a low proliferation rate in the adult, and a high rate in the embryo (Risau, 1995). This general concept seems to be supported by the known specific and mytogenic effect of VEGF, a crucial molecule in vascular formation (Ferrara et al., 1995). However, the reported pro­ liferation rates of in vivo angioblasts (Breier et al., 1992; Millauer et al., 1993) are not necessarily high, and as a matter of fact the promitotic effect of VEGF is characteris­ tic of in vitro-cultured endothelial cells (Keck et al., 1989; Alon et al., 1995), but has not specifically been reported for angioblasts. Then, addition of exogenous VEGF to developing embryos seems to produce vascular dysmor­ phogenesis by endothelial hyperfusion, rather than by increasing cell numbers (Drake and Little, 1995). Finally, several studies have demonstrated that isolated embry­ onic angioblasts display low cell proliferation (BrandSaberi et al., 1995; Guadix et al., 2006), and that tight control of endothelial progenitor and immature endothelial cell proliferation is essential to proper vascular develop­ ment and remodeling (Lai et al., 2003). This angioblastic

475

low-proliferation scenario could explain the complex and progressive vascularization of the embryo as development proceeds, without having to assume that some angioblasts, completely surrounded by provascular signals, have to remain isolated and do not undergo vascular morphogen­ esis until many hours and even days after the completion of gastrulation. Yet another crucial question remains open: how are the different blood vessel cell lineages related? Several old and modern theories try to explain the ontogenetic relation­ ship that exists between endothelial, smooth muscle and blood cells. The classical hypothesis suggests that both endothelial and blood cells derive from a common cell progenitor, the hemangioblast (His, 1900). Morphological data supporting this hypothesis include the formation in the extra-embryonic membranes of isolated small blood cell colonies completely covered by endothelial cells, the so-called blood islands (González-Crussi et al., 1971; Risau and Flamme, 1995) and the formation of “blast colonies” derived from embryonic stem (ES) cell cultures that are capable (though transiently) of giving rise to both blood and endothelial cells (Choi et al., 1998). Among the molecular evidence that credits the hemangioblast hypoth­ esis, the common expression by both early endothelial and blood cells of different molecules like Flk1 is the most important one (Millauer et al., 1993; Yamaguchi et al., 1993; Eichmann et al., 1997; Kabrun et al., 1997). Other molecules that are expressed by both endothelial and blood progenitors are CD34 (Young et al., 1995), SCL/Tal-1 (Kallianpur et al., 1994), Flt1 (Fong et al., 1996), GATA-2 (Orkin, 1992), Cbfa2/Runx1/AML1 (North et al., 1999) and PECAM-1 (Watt et al., 1995). An important criti­ cism of the hemangioblast hypothesis is that this cell type might be an in vitro artifact, rather than an in vivo reality. Differences recorded between the vascularization of the extraembryonic territories, which normally correlate with blood differentiation, and the vascularization of the intraembryonic domain, where blood vessels develop without significant associated hemopoiesis, really confirms the complexity of the embryonic vascular development. The hemangioblast hypothesis has somehow acquired a new dimension after the discovery of intraembry­ onic hemopoietic stem cells (HSCs) which seem to be released to the bloodstream from the aortic endothe­ lium (Medvinsky et al., 1993; Medvinsky and Dzierzak, 1996; Pardanaud et al., 1996; Jaffredo et al., 1998, 2000; reviewed in Dieterlen-Lièvre et al., 2006). The pres­ ence of such hemogenic endothelium (that in the mouse embryo can also be found in large veins like the umbili­ cal ones, Wood et al., 1997) again illustrates the close ontogenetic relationship which exists between endothe­ lial and blood cells. However, recent discoveries discuss the endothelial origin of intraembryonic HSCs, and there­ fore challenge the lineage relationship between these cell types.

476

A different series of findings support the so-called vas­ cular bipotential progenitor hypothesis that states that both endothelial and vascular smooth muscle cells might have a common cell precursor (Yamashita et al., 2000). The idea is exciting and is clearly consistent with some findings in the developing embryo. A good example is provided by the application of CRE-Lox technology to drive permanent LacZ expression under VEGFR2/Flk-1 promoter activ­ ity, that revealed that both endothelial and smooth muscle cells express -galactosidase in these embryos (Motoike et al., 2003), thus suggesting a lineage relationship between them. This finding was challenged by other groups that, using the same approach, reported transgene expression only in endothelial cells (Licht et al., 2004). Additional support for the bipotential progenitor hypothesis comes from the finding that epicardial cells are able to give rise to both coronary endothelial and smooth muscle cells (PérezPomares et al., 1997, 1998a,b, 2002a,b; Dettman et al., 1998; Männer et al., 1998; Vrancken Peeters et al., 1998). Finally, neointimal differentiation of smooth muscle cells from adult endothelial cells (De Meyer and Bult, 1997) or late embryonic endothelial cells (Arciniegas et al., 2000) also fits this hypothesis. Nevertheless, in the case of the embryo, the smooth muscle cell population is known to be highly heterogeneous in terms of its origin, and it is rea­ sonable to accept that only some portions of the vascula­ ture would derive from such bipotential precursor. Are all these hypotheses mutually exclusive? We believe this is not the case, since, as we will discuss below, all these vascular cell lineages seem to share a phylogenetic origin, the coelomic epithelial cells. This common ancestry could explain the different and sometimes contradictory models about the developmental relationships between endothelium, blood cells and vascular smooth muscle.

III.E.  Endothelial Cell Types: The Paradigmatic Case of Cardiac Endothelial Lineages The heart includes at least two different populations of endothelial cells, the endocardial cell type that line the car­ diac lumen, and the coronary endothelium, which appears later in development. Both of them are unique, and this section will be dedicated to them. Endocardial cells originate from the precardiac epithelium of the primary heart fields, a subpopulation of the splanch­ nopleural component of the lateral plate mesoderm in close relationship with the pharyngeal endoderm, a critical tissue in early cardiac development (Lough and Sugi, 1995, 2000; see Chapters in part 1). Endocardial progenitors are thought to segregate from this epithelial layer by an apparent dela­ mination, mediated by an epithelial-to-mesenchymal (EMT) transformation. Endocardial progenitor cells that downregu­ late N-cadherin expression (Manasek, 1968; Linask and Lash, 1993), migrate away from the rest of the cells in the

PART  |  8  Making Vessels

precardiac epithelia (that will become N-cadherinexpressing myocardial cells) and arrange internally to the muscular component of the paired cardiac primor­ dium (Linask and Lash, 1993; Sugi and Markwald, 1996). A great deal of controversy exists on the lineage relation­ ship of these two cell types. While some authors have sug­ gested that precardiac cells represent only one cell type with the potential to give rise to both myocardial and endo­ cardial progenitors (Linask and Lash, 1993; Eisenberg and Bader, 1995), data from retroviral lineage have shown clonal differentiation from these cells without significant intermixing of cells, therefore indicating that myocardial and endocardial progenitors are very likely to derive from two different subpopulations of cells already present in the primary heart field (Cohen-Gould and Mikawa, 1996). Independent of this, it seems clear that the endocardium includes at least two distinct populations of cells, of which only one is able to undergo an EMT and transform into, at least part of, the valvuloseptal mesenchyme of the atrio­ ventricular and conoventricular valves. This subpopula­ tion of transformation-competent endocardial cells was first identified in the avian embryo by its fibrillin-2 immu­ noreactivity (evidenced by the JB3 antibody) (Wunsch et al., 1994). Further research on mouse embryos has shown that the endocardium of this cushion segment expresses markers like the NF-ATc and Snail transcription factors that are not expressed by other endocardial cells (de la Pompa et al., 1998; Ranger et al., 1998; Timermann et al., 2004). The final fate of the endocardially-derived mes­ enchyme also seems to be different. Experiments using the CRE/Lox technology to drive permanent LacZ expres­ sion after Tie2 activation revealed that atrioventricular endocardially-derived mesenchymal cells remain in the atri­ oventricular cushions and contribute to the formation of the valvular apparatus (De Lange et al., 2004), whereas cono­ ventricular endocardially-derived mesenchyme is only found in the more proximal (conal) area of the outflow tract, but not in the more distal one (Kisanuki et al., 2001; De Lange et al., 2004). The other heart endothelial lineage is that of the coro­ nary vessels. The coronary system nourishes the compact myocardial layer that appears in the ventricles of some ver­ tebrates (it is the presence of a trabecular or spongy layer, and not the presence of a compact one, that characterizes the original vertebrate ventricle). It has been assumed that the reason for the coronaries to develop was the existence of this compact myocardial layer (Ostadal et al., 1975; Ostadal, 1979), but we now know that coronary vessels and their precursors are critical to the proliferation of the compact myocardial layer (Moore et al., 1999; Tevosian et al., 2000; Chen et al., 2001; Stuckmann et al., 2001; Pérez-Pomares et al., 2002b; Pennisi et al., 2003). For many years it was thought that coronary endothe­ lium was a derivative of the aortic endothelium (Licata, 1954, 1955; Vernall, 1962; Dbaly et al., 1968; Rychter and

Chapter  |  8.1  Origin of the Vertebrate Endothelial Cell Lineage: Ontogeny and Phylogeny

Ostadal, 1971; Hirakow, 1983; Conte and Pellegrini, 1984; Corone et al., 1984; Hutchins et al., 1988; Ogden, 1988) or the sinus venosus endocardium (Aikawa and Kawano, 1982; Viragh and Challice, 1981). This controversy was somehow linked to the discussion on the vascular mor­ phogenetic mechanisms that guided coronary vascular development. Angiogenesis was the preferred candidate for decades (Aikawa and Kawano, 1982; Hirakow, 1983; Viragh et al., 1993; Vrancken Peeters et al., 1997a,b) (among others), but retroviral studies from the 1990s (Mikawa and Fishman, 1992; Mikawa and Gourdie, 1996) demonstrated that coronary cells are a proepicardial/epi­ cardial derivative that originally form through vasculogen­ esis. These researchers carried out both direct retroviral tagging of proepicardial cells and labeling of the early endocardium. The first set of experiments showed that proepicardial cells differentiate into coronary endothelial and smooth muscle cells and the second that there does not seem to be a clear contribution of the endocardium to coro­ nary development. These results have received further sup­ port from descriptive as well as experimental studies, based on proepicardial quail-to-chick transplantations (quail to chick chimeras) or diverse cell labeling procedures (PérezPomares et al., 1997, 1998a,b, 2002a,b; Dettman et al., 1998; Männer et al., 1998; Vrancken Peeters et al., 1998; Kattan et al., 2004). An important finding was that the dif­ ferentiation of the coronary progenitors also takes place after the EMT of the primitive epicardium (Pérez-Pomares et al., 1997, 1998a; Dettman et al., 1998; Gittenbergerde Groot et al., 1998), so that the mesenchymal popula­ tion that invades the subepicardial space and contains the coronary progenitors is a main proepicardial/epicardial derivative (these cells are often referred to as epicardiallyderived cells or EPDCs, Gittenberger-de Groot et al., 1998). Recent in vivo and in vitro studies have sug­ gested that some EPDCs might have the potential to differentiate into both endothelial and smooth muscle cells (Guadix et al., 2006), an idea that fits the hypoth­ esis that some of the molecular mechanisms that regu­ late epicardial epithelial-to-mesenchymal transition are directly involved in the activation of the multipotential properties of EPDCs (the “coronary bipotential progeni­ tor model”) (Muñoz-Chápuli et al., 2002; Pérez-Pomares et al., 2002a; Wessels and Pérez-Pomares, 2004; Smart et al., 2007). This hypothesis is compatible with the reported formation of coronary vessels from discontinuous cell colonies (derived from the epicardium) containing only one cell type (either endothelial or smooth muscle cells, but not both) (Mikawa and Fischman, 1992), because the “coronary bipotential progenitor model” states that: (1) some coronary progenitors have the potential to differ­ entiate into endothelial and smooth muscle cells (not that they actually do so); and that (2) coronary endothelial cells differentiate before smooth muscle cells, following envi­ ronmental cues. Therefore, a tagged bipotential precursor,

477

at a certain time point and in a given anatomical location, would only give rise to a clone of one cell type. A final aspect of coronary development that deserves some specific attention is that of the differences between arteries and veins. The coronary vascular system is known to be composed of both arterial and venous blood vessels that develop in parallel but are anatomically connected to different heart segments. Coronary veins, which are mainly dorsal, connect to the venous pole of the heart (sinus venosus, future coronary sinus of mammals) whereas coronary arteries connect to the aortic root. Many studies on coronary vessel development ambiguously try to avoid discussions of these two vessel types, in part because during the first stages of coronary development it is not possible to distin­ guish between arteries and veins in many cardiac regions. Only a couple of studies (Vrancken Peeters et al., 1997a,b) pointed to this conflicting aspect of coronary develop­ ment, but always regarding arteries and veins as deriva­ tives of the sinus venosus endocardium and neglecting the data on (pro)epicardial involvement in coronary formation (Chapter 5.1). Published and unpublished results from our laboratory suggest that the arterial and venous components of the coronary tree have different origins. This conclusion is first drawn from the possible differential contribution of proepicardial derivatives to arteries with respect to veins (Pérez-Pomares et al., 2002a). Coronary arteries are known to join to the aortic root relatively late in development (at least three days after coronary formation is initiated), while veins seem to connect to the sinus venosus endocardium quite early in development (Vrancken Peeters et al., 1997a,b; Muñoz-Chápuli et al., unpublished results). This allows early coronary arteries to be distinguished from veins on the basis of their connection to the systemic flow. This approach, combined with quail-chick chimeras, yielded results sug­ gesting that the main derivatives of (pro)epicardial tissues are coronary arteries, but not necessarily the cardiac veins (Pérez-Pomares et al., in preparation). As a summary of this section we would like to empha­ size that cardiac endothelial tissues are an excellent example of the close ontogenetic relationship that exists between the endothelial and smooth muscle cells and the coelomic tissues. The results discussed are in accordance with the hypothesis that embryonic coelomic tissues are a source of hemangioblasts (Muñoz-Chápuli et al., 1999), as well as with the suggested evolutionary origin of the verte­ brate vascular system from coelomic progenitors (MuñozChápuli et al., 2005), an idea we will further develop in the following sections of this chapter.

III.F.  Molecular Regulation of Embryonic Vasculogenesis: Essential Elements Chapter 8.2 focuses on the molecular regulation of vas­ cular development. This revision summarizes most of

478

our current knowledge on the molecular network con­ trolling vascular morphogenesis, and we have therefore tried to avoid conceptual overlap in this chapter. We would, however, like to emphasize some specific fea­ tures of embryonic endothelium and their progenitors, and consider them in a context that attempts a systemic analysis of the developing vasculature. A recent study (Argraves and Drake, 2005) has shown that, among the many molecules that have been reported to be involved in vasculogenic events, only some of them are really criti­ cal to the process. A quick overview of the list of these essential molecules immediately allows realization that they fall into one of these three groups: (1) VEGF/angio­ poietin signaling and transduction (VGFA, VEGFR1Flt-1, VEGFR2-Flk-1, Tie2/Tek, CD148/DEP-1, ShcA – part of the MAP Kinase pathway); (2) Retinoic acid (RA) signaling (cytochrome P450 reductase); (3) Cell-to-cell or cell-to-matrix contact mediators (fibronectin, 5 integrin, PECAM/CD31, VE-cadherin, connexin 45, ephrinB2). Why are these three aspects of cell signaling/behavior so important for endothelial vasculogenesis? The relevance of the VEGF signaling pathway in vascular morphogen­ esis is clear, the importance of retinoic acid in the con­ trol of endothelial proliferation during the formation of embryonic vascular beds has already been discussed (Lai et al., 2003), and the significance of adhesion mol­ ecules for cells that basically are epithelial in nature seems obvious. However, the results presented and dis­ cussed by Argraves and Drake may be considered from the point of view of the ancestral roles of these molecules and the biological processes they represent in the ances­ tors of vertebrates. The importance of VEGF signaling in invertebrates, as well as its linkage to hemocytic cell lin­ eages, have been reported (Cho et al., 2002; Munier et al., 2002; Bruckner et al., 2005); how these findings relate to vertebrate endothelium is discussed below. RA-related machinery has been found to be present in invertebrates (Chapter 3.3), and has been suggested to be pivotal to the morphological radiation of deuterostomes (Bouton et al., 2005; Canestro et al., 2006). Although in these and other studies (Marletaz et al., 2006) an emphasis in the role of RA in the establishment of a vertebrate A–P axis is made (mostly focused in central nervous system patterning), it is also suggested that RA is critical to the differentiation of pharyngeal endoderm, a tissue crucial to cardiovas­ cular development in the vertebrate embryo (Lough and Sugi, 2000; Chapter 7.2). Finally, adhesion properties of endothelial cells might not only be a requirement of their epithelial nature, but might also reflect their possible origin from circulating cell precursors that would have acquired adhesive characteristics in order to seed into the blood vessel intimal lining. This possibility is accepted for the case of adult vertebrate circulating progenitors, including human ones (Asahara et al., 1999; Kalka et al., 2000). This evidence, taken together, offers a new

PART  |  8  Making Vessels

evolutionary perspective on the origin of the vertebrate cardiovascular system which will be the focus of the rest of this chapter.

IV.  Phylogeny of the endothelium IV.A.  A Hypothesis on the Origin   of Endothelial Cells After this comparative analysis of the circulatory system in metazoans, as well as the review of the emergence of endothelial progenitors in vertebrate development, we can propose a model of the evolutionary origin of the endothelium in vertebrates, together with a hypothetical reconstruction of the invertebrate–vertebrate transition of the circulatory system. This model has been put forward in Muñoz-Chápuli et al. (2005). As far as we know, the first speculation about the origin of the endothelium from freely-migrating cells was made by Kuprijanov (1990), although this author admitted the presence of endothelium in some groups of invertebrates. Recently, Hartenstein and Mandal (2006) have reviewed the phylogeny of the blood/vascular system, stressing the relationships between coelom, blood cells and vasculature, and looking for the ancestral condition of the vascular system. We think that endothelial cells are evolutionary deriva­ tives from amoebocytes, a type of hyaline hemocyte with the ability to adhere to and migrate over the basal laminal of the vessels. In the origin of vertebrates these circulating cells acquired a function of lining all the luminal surface of the vascular system (Fig. 4). This proposal is supported by the absence of a true endothelium in invertebrates. Given that vascular myoepithelial cells of invertebrates seem to be a clear candidate for evolutionary progenitors of ver­ tebrate vascular smooth muscle cells, as we will discuss below, the only possibility for an ancestry of the endothe­ lium must be in the blood cells, among which the amoebo­ cytes are the most suitable candidates. It is interesting to note that amoebocytes are apparently the only blood cells present in the hemal system of Branchiostoma, an inverte­ brate taxon closely related to vertebrates (Ruppert, 1997). Endothelialization required development of intercel­ lular junctions between amoebocyte-like cells which had already developed molecular systems for anchoring to the basal lamina of the vessel walls. It is interesting to note that two main molecules participating in these intercellular junctions, PECAM-1 and VE-cadherin, are also expressed by blood cells or their progenitors (Fraser et al., 2003). This suggests that protoendothelial cells did not need to “invent” anything new, but resorted to molecules which were already present in blood cells. The drastic change from the invertebrate to the verte­ brate type of vessels leads to the question of what advan­ tage was obtained from lining the whole vascular surface

Chapter  |  8.1  Origin of the Vertebrate Endothelial Cell Lineage: Ontogeny and Phylogeny

1

2

479

3

Figure 4  A model for the invertebrate–vertebrate transition of the circulatory system. The whole cartoon should be interpreted from left to right. The 1–3 sequence summarizes the hypothesis of the origin of true endothelial cells from circulating (blood-like) cell progenitors. Hemal cavities, which are between the basal laminae of the purple endoderm and the orange coelomic epithelium (see also Fig. 1), become progressively lined by adher­ ent blood cells (in red) that acquired epithelial features. At the same time, myoepithelial cells (in blue) gave rise to perivascular cells (green) that can build a multilayered vessel wall. The process is completed in vertebrates (3). Sections 2 and 3 can also be analyzed from left to right, so that both the hierarchy of circulatory vessels based on the proportion of basal lamina-adhered circulating cells (2) or the developmental stage of maturation of the vertebrate blood vessels (3) are depicted.

with blood cells. We think that endothelium confers three advantages which might have been decisive for vertebrate evolution: 1. Immunological defense. The progenitors of the endothelial cells were probably compromised in defensive functions. Cooperation and communication between different types of blood cells is a well-known phenomenon in both vertebrates and invertebrates. The immobilization of endothelial cells at the vascular wall throughout the body allows rapid movement of cells from the immune system to areas of infection. The expression of specific molecules, such as selectins, in the endothelial surface mediates the docking of these circulatory cells, which can then extravasate and invade the tissues. On the other hand, endothelial cells care­ fully regulate plasma extravasation through response to VEGF, and participate in coagulation in cooperation with blood cells. 2. Fine regulation of blood flow. In the invertebrate microvessels, blood flow is regulated through contrac­ tion or relaxation of the myoepithelial cells under neu­ rogenic stimuli (Shigei et al., 2001). Vasoconstriction and vasorelaxation is also performed in vertebrates by smooth muscle cell contraction, but these processes are locally controlled by endothelial cells through specific molecules, such as nitric oxide. 3. Angiogenesis. We think that the key achievement of the endothelialization of the whole hemal circulatory system is the angiogenic ability (i.e., the possibility of rapid vascular growth) provided by endothelial cells. In fact, we can now redefine angiogenesis as recupera­ tion by the endothelium of the ancestral features of circulating blood cells able to migrate and invade the tissues. The possibility of shifting between an epi­ thelial, quiescent and an invasive phenotype allowed

the primitive hemal system to extend to body areas located away from the coelom. This extension of the hemal system is restricted in invertebrates, because the ability of myoepithelial cells to migrate seems to be more limited. We think that endothelial-dependent ang­ iogenic growth of vessels, a phenomenon exclusive of vertebrates, explains why the “somatic” compartment of the vertebrate body (i.e., limbs, myotomes, head, central nervous system and sense organs) was able to develop to an extent never reached by invertebrates. The only exception, maybe, are the cephalopods, whose potent musculature is supplied by microvessels consisting of myoepithelial cells. It will be interesting to study the cellular and molecular regulation of the exceptional growth of these vessels in cephalopods, no doubt related to the large size attained by these ani­ mals, and to compare it with the angiogenic process in vertebrates.

IV.B.  The Origin of Pericytes and Vascular Smooth Muscle According to our comparative analysis, it seems clear that vascular smooth muscle (and most probably pericytes) evolutionarily originated from the vascular myoepithelial cells, which in turn are derived from the coelomic epithe­ lium, as stated above. It is important to remark that, in this way, vascular smooth muscle constitutes the original ele­ ment of the vascular system in a phylogenetic perspective, being the endothelial cell type “newcomer” in the verte­ brate vessels. This is probably anti-intuitive and a viola­ tion of the “principle of recapitulation”, due to the fact that pericytes and smooth muscle are recruited during vascular development by a primary plexus of capillaries composed exclusively of endothelium. This is true, but it also reflects

480

the new properties acquired by the endothelial cells, which are able to form lumenized structures before the avail­ ability of mural progenitors. On the other hand, the close relationship between coelomic-derived cells and vascular mural cells is appealing. Coelomic origin has been shown for smooth muscle of the coronary arteries (Mikawa and Gourdie, 1996; Vrancken Peeters et al., 1999), gut vascu­ lature (Wilm et al., 2005), part of the aorta (Pérez-Pomares et al., 1999) and also for the perisinusoidal or stellate cells of the liver (Pérez-Pomares et al., 2004). On the other hand the stabilizing role of the mural cells for the endothelium, and the control that mural cells exert on the endothelial cell phenotype through specific molecules (angiopoie­ tins), show that these cells play a main role in the vascular homeostasis.

IV.C.  Supporting Evidence The explanatory potential is a good test for providing support to a new hypothesis. We think that considering endothelial cells as a specialized type of blood cells allow the explanation of a number of phenomena which are more difficult to explain with alternative assumptions. The first example is the ontogenetic origin of endothelial and blood cells from a common progenitor, the hemangioblast. The

PART  |  8  Making Vessels

origin of endothelial and vascular smooth muscle cells from a common vascular progenitor, as described by some authors, also makes sense when we consider that the origin of both myoepithelial-derived vascular smooth muscle and blood cells can be traced back to the coelomic epithelium, i.e., both are coelomic derivatives (Hartenstein, 2006). Another observation which can be easily explained by our model is the presence of VEGFR-like molecules in invertebrates. In fact, in Drosophila, VEGFR ortholog is expressed by blood cells, and the VEGF signaling system is involved in control of blood cell migration (Cho et al., 2002). In other words, the same signaling system which stimulates angiogenesis in vertebrates is regulating blood cell migration in insects. We think that this is an example of conserved function in endothelial cells of a mechanism which was already operative in the blood cells from which endothelial cells derived. The striking observations made on the botryoidal tissue in hirudineans can also be regarded from the point of view of our model. Botryoidal tissue can be stimulated by tissue injury or by injection of VEGF, giving rise to hollow chan­ nels, lined by endothelial-like cells, which allow cells from the immune system to arrive at the site of the stimulus (Tettamanti et al., 2003). Botryoidal tissue, a subtype of chloragogen cells, has a coelomic origin (Hartenstein,

Figure 5  Parallels between ontogeny and phylogeny in the origin of the vertebrate vascular system. The observed ontogenetic relationship between the vascular cell lineages can be explained by a multipotential hypothetical vascular progenitor related with the embryonic mesoderm. In our model, the coelomic epithelium would be the immediate ancestor of these vascular progenitors. In this way, ontogeny would be a recapitulation of the phylogenetic origin of all the elements of the vascular system from coelomic-derived cells. Note that the hypothetical “hemangioblast” would become an ontogenetic recapitulation of the common ancestor of invertebrate blood cells, probably a macrophage-like type of coelomocyte. The existence of a common vascu­ lar progenitor cell for endothelium and vascular smooth muscle (Yamashita et al., 2000) can also be explained by this model.

Chapter  |  8.1  Origin of the Vertebrate Endothelial Cell Lineage: Ontogeny and Phylogeny

2006). This is an interesting example of cooperation between blood cells with the capacity to make tunnels in the tissues and defensive cells. This observation can also be related to the striking observation of the “drilling mac­ rophages” in ischemic tissues, a type of poorly-known, angiogenic-like process (Moldovan et al., 2000). Finally, an observation which in our opinion can be well-explained by our model, is the presence of endothe­ lial progenitor cells in the bone marrow which can be mobilized to the circulation and can participate in vascu­ lar repair (reviewed by Murasawa and Asahara, 2005). We think that this observation reveals the close relationships between endothelial and blood cells which, according to our model, are two subtypes from the same cell lineage, from the ontogenetic as well as from the phylogenetic point of view.

IV.D.  An Evo–Devo Conclusion A clear parallel can be established between evolution and development of the components of the circulatory sys­ tem (Fig. 5). The embryonic origin of the early progeni­ tors of the vascular system in the endoderm–mesoderm interphase can be related to the anatomical pattern of the primitive hemal system, always located around the diges­ tive tract. On the other hand, the origin of endothelial and perivascular cells from the embryonic coelomic epithelium corresponds with the evolutionary origin of blood cells and myoepithelial cells from coelomic-derived cells. An interesting prediction which derives from this Evo–Devo approach is that it should be possible to find a relationship in vertebrate embryos between the emergence of hemo­ poietic stem cells and the coelomic epithelium. In fact, such a relationship has already been suggested by Olah and collaborators (1988), and we have also described an important contribution of coelomic-derived cells to the aorta-gonad-mesonephros region of the embryo where the definitive hemopoietic stem cells differentiate (PérezPomares et al., 1999). We think that models and work­ ing hypotheses derived from this Evo–Devo analysis can orientate research towards new, unexpected and excit­ ing possibilities in our knowledge on endothelial, smooth muscle and blood cell differentiation.

References Abdi, K., Kobzik, L., Li, X., Mentzer, S.J., 1995. Expression of mem­ brane glycoconjugates on sheep lung endothelium. Lab. Invest. 72, 445–452. Aikawa, E., Kawano, J., 1982. Formation of the coronary arteries sprouting from the primitive aortic sinus wall of the chick embryo. Experientia 38, 816–818. Alon, R., Kassner, P.D., Carr, M.W., Finger, E.B., Hemler, M.E., Springer, T.A., 1995. The integrin VLA-4 supports tethering and rolling in flow on VCAM-1. J. Cell Biol. 128, 1243–1253.

481

Anderson-Berry, A., O’Brien, E.A., Bleyl, S.B., Lawson, A., Gundersen, N., Ryssman, D., Sweeley, J., Dahl, M.J., Drake, C.J., Schoenwolf, G.C., Albertine, K.H., 2005. Vasculogenesis drives pulmonary vascular growth in the developing chick embryo. Dev. Dyn. 233, 145–153. Andreucci, D., Herrero, M.A., Velázquez, J.J., 2004. On the growth of filamentary structures in planar media. Math. Meth. Appl. Sci. 27, 1935–1968. Arciniegas, E., Ponce, L., Hartt, Y., Graterol, A., Carlini, R.G., 2000. Intimal thickening involves transdifferentiation of embryonic endo­ thelial cells. Anat. Rec. 258, 47–57. Argraves, W.S., Drake, C.J., 2005. Genes critical to vasculogenesis as defined by systematic analysis of vascular defects in knockout mice. Anat. Rec. 286, 875–884. Armulik, A., Abramsson, A., Betsholtz, C., 2005. Endothelial/pericyte interactions. Circ. Res. 97, 512–523. Asahara, T., Masuda, H., Takahashi, T., Kalka, C., Pastore, C., Silver, M., Kearne, M., Magner, M., Isner, J.M., 1999. Bone marrow origin of endothelial progenitor cells responsible for postnatal vasculogenesis in physiological and pathological neovascularization. Circ. Res. 85, 221–228. Azar, Y., Eyal-Giladi, H., 1979. Marginal zone cells – the primitive streak-inducing component of the primary hypoblast in the chick. J. Embryol. Exp. Morphol. 52, 79–88. Bagley, R.G., Weber, W., Rouleau, C., Teicher, B.A., 2005. Pericytes and endothelial precursor cells: cellular interactions and contributions to malignancy. Cancer Res. 65, 9741–9750. Benito, J., Pardos, F., 1997. Hemichordata. In: Harrison, F.W., Ruppert, E.E., (Eds.), Microscopic Anatomy of Invertebrates. vol. 15, Hemi­ chordata, Chaethognatha and the Invertebrate Chordates. Wiley-Liss, New York, NY, pp.15–102. Bodmer, R., Frasch, M., 1999. Genetic determination of Drosophila heart development. In: Harvey, R.P., Rosenthal, N. (Eds.), Heart Development. Academic Press, San Diego, CA, pp. 65–90. Bossche, J.P., Jangoux, M., 1976. Epithelial origin of starfish coelomo­ cytes. Nature 261, 227–228. Bouïs, D., Hospers, G.A., Meijer, C., Molema, G., Mulder, N.H., 2001. Endothelium in vitro: A review of human vascular endothelial cell lines for blood-vessel-related research. Angiogenesis 4, 91–102. Bouton, D., Escriva, H., de Mendonça, R.L., Glineur, C., Bertin, B., Noël, C., Robinson-Rechavi, M., de Groot, A., Cornette, J., Laudet, V., Pierce, R.J., 2005. A conserved retinoid X receptor (RXR) from the mollusk Biomphalaria glabrata transactivates transcription in the presence of retinoids. J. Mol. Endocrinol. 34, 567–582. Brand-Saberi, B., Seifert, R., Grim, M., Wilting, J., Kuhlewein, M., Christ, B., 1995. Blood vessel formation in the avian limb bud involves angioblastic and angiotrophic growth. Dev. Dyn. 202, 181–194. Breier, G., Albrecht, U., Sterrer, S., Risau, W., 1992. Expression of vas­ cular endothelial growth factor during embryonic angiogenesis and endothelial cell differentiation. Development 114, 521–532. Bruckner, K., Kockel, L., Duchek, P., Luque, C.M., Rorth, P., Perrimon, N., 2004. The PDGF/VEGF receptor controls blood cell survival in Drosophila. Dev. Cell 7, 73–84. Brusca, R., Brusca, G., 2003. Invertebrates. Sinauer Associates, Sunderland, MA. Budelmann, B.U., Schipp, R., von Boletzky, S., 1997. Cephalopoda. In: Harrison, F.W., Kohn, A.J. (Eds.), Microscopic Anatomy of Invertebrates. vol. 6A, Molluscs. Wiley-Liss, New York, NY, pp. 119–414.

482

Burighel, P., Cloney, R.A., 1997. Urochordata: Ascidiacea. In: Harrison, F.W., Ruppert, E.E. (Eds.), Microscopic Anatomy of Invertebrates. vol. 15, Hemichordata, Chaethognatha and the Invertebrate Chordates. Wiley-Liss, New York, NY, pp. 221–348. Canestro, C., Postlethwait, J.H., Gonzalez-Duarte, R., Albalat, R., 2006. Is retinoic acid genetic machinery a chordate innovation? Evol. Dev. 8, 394–406. Chan, K.S., Cavey, M.J., Wilkens, J.L., 2006. Microscopic anatomy of the thin-walled vessels leaving the heart of the lobster Homarus americanus: anterior lateral arteries. Invert. Biol. 125, 70–82. Chen, T.H., Chang, T.C., Kang, J.O., Choudhary, B., Makita, T., Tran, C.M., Burch, J.B., Eid, H., Sucov, H.M., 2002. Epicardial induction of fetal cardiomyocyte proliferation via a retinoic acid-inducible trophic fac­ tor. Dev. Biol. 250, 198–207. Cho, N.K., Keyes, L., Johnson, E., Heller, J., Ryner, L., Karim, F., Krasnow, M.A., 2002. Developmental control of blood cell migration by the Drosophila VEGF pathway. Cell 108, 865–876. Choi, K., Kennedy, M., Kazarov, A., Papadimitriou, J.C., Keller, G., 1998. A common precursor for hemopoietic and endothelial cells. Development 125, 725–732. Cholewa, J., Feeney, G.P., O’Reilly, M., Sturzenbaum, S.R., Morgan, A.J., Plytycz, B., 2006. Autofluorescence in eleocytes of some earthworm species. Folia. Histochem. Cytobiol. 44, 65–71. Claes, M.F., 1996. Functional morphology of the white bodies of the cephalopod mollusc Sepia officinalis. Acta. Zool. 77, 173–190. Coffin, J.D., Poole, T.J., 1991. Endothelial cell origin and migration in embryonic heart and cranial blood vessel development. Anat. Rec. 231, 383–395. Cohen-Gould, L., Mikawa, T., 1996. The fate diversity of mesodermal cells within the heart field during chicken early embryogenesis. Dev. Biol. 177, 265–273. Conte, G., Pellegrini, A., 1984. On the development of the coronary arteries in the human embryos, stages 14–19. Anat. Embryol. 169, 209–218. Corone, P., Corone, A., Dor, X., Binet, J.P., 1984. Les artères coronaires et leurs variations, une explication embryologique. C. R. Acad. Sci. (Paris) 299, 451–458. Davis, S., Aldrich, T.H., Jones, P.F., Acheson, A., Compton, D.L., Jain, V., Ryan, T.E., Bruno, J., Radziejewski, C., Maisonpierre, P.C., Yancopoulos, G.D., 1996. Isolation of angiopoietin-1, a ligand for the TIE2 receptor, by secretion-trap expression cloning. Cell 87, 1161–1169. Dbaly, J., Ostadal, B., Rychter, Z., 1968. Development of the coronary arteries in the rat embryos. Acta. Anat. 71, 209–222. De la Pompa, J.L., Timmerman, L.A., Takimoto, H., Yoshida, H., Elia, A.J., Samper, E., Potter, J., Wakeham, A., Marengere, L., Langille, B.L., Crabtree, G.R., Mak, T.W., 1998. Role of the NF-ATc transcription factor in morphogenesis of cardiac valves and septum. Nature 392, 182–186. de Lange, F.J., Moorman, A.F., Anderson, R.H., Manner, J., Soufan, A.T., de Gier-de Vries, C., Schneider, M.D., Webb, S., van den Hoff, M.J., Christoffels, V.M., 2004. Lineage and morphogenetic analysis of the cardiac valves. Circ. Res. 95, 645–654. De Meyer, G.R., Bult, H., 1997. Mechanisms of neointima formationlessons from experimental models. Vasc. Med. 2, 179–189. Dejana, E., 1996. Endothelial adherens junctions: Implications in the control of vascular permeability and angiogenesis. J. Clin. Invest. 98, 1949–1953. Del Vecchio, P.J., Siflinger-Birnboim, A., Belloni, P.N., Holleran, L.A., Lum, H., Malik, A.B., 1992. Culture and characterization of

PART  |  8  Making Vessels

pulmonary microvascular endothelial cells. In Vitro Cell Dev. Biol. 28A, 711–715. Dettman, R.W., Denetclaw, W., Ordahl, C.P., Bristow, J., 1998. Common epicardial origin of coronary vascular smooth muscle, perivascular fibroblasts, and intermyocardial fibroblasts in the avian heart. Dev. Biol. 193, 169–181. Dieterlen-Lievre, F., Pouget, C., Bollerot, K., Jaffredo, T., 2006. Are intra-aortic hemopoietic cells derived from endothelial cells during ontogeny? Trends Cardiovasc. Med. 16, 128–139. Dolmatov, I.Y., Ginanova, T.T., 2001. Muscle regeneration in holothuri­ ans. Microsc. Res. Tech. 55, 452–463. Doyle, W.L., 1964. Basement membrane and its precursor in haemal vessels of Holothurian. J. Cell Biol. 23, A26. Drake, C.J., Little, C.D., 1995. Exogenous vascular endothelial growth factor induces malformed and hyperfused vessels during embryonic neovascularization. Proc. Natl. Acad. Sci. USA 92, 7657–7661. Drake, C.J., Little, C.D., 1999. VEGF and vascular fusion: implications for normal and pathological vessels. J. Histochem. Cytochem. 47, 1351–1356. Drake, C.J., Brandt, S.J., Trusk, T.C., Little, C.D., 1997. TAL1/SCL is expressed in endothelial progenitor cells/angioblasts and defines a dorsal-to-ventral gradient of vasculogenesis. Dev. Biol. 192, 17–30. Eckert, R., Randall, D., 1988. Animal Physiology: Mechanisms and Adaptations. W. H. Freeman and Co, New York, NY. Edgell, C.J., McDonald, C.C., Graham, J.B., 1983. Permanent cell line expressing human factor VIII-related antigen established by hybridi­ zation. Proc. Natl. Acad. Sci. USA 80, 3734–3737. Eisenberg, C.A., Bader, D., 1995. QCE/6: a clonal cell line with cardiac myogenic and endothelial cell potentials. Dev. Biol. 167, 525–536. Fernández, J., Téllez, V., Olea, N., 1992. Hirudinea. In: Harrison, F.W., Gardiner, S.L. (Eds.), Microscopic Anatomy of Invertebrates. vol. 7, Annelida. Wiley-Liss, New York, NY, pp. 323–3394. Ferrara, N., Heinsohn, H., Walder, C.E., Bunting, S., Thomas, G.R., 1995. The regulation of blood vessel growth by vascular endothelial growth factor. Ann. NY Acad. Sci. 752, 246–256. Folkman, J., Klagsbrun, M., 1987. Angiogenic factors. Science 235, 442–447. Fong, G.H., Rossant, J., Gertsenstein, M., Breitman, M.L., 1995. Role of the Flt-1 receptor tyrosine kinase in regulating the assembly of vascular endothelium. Nature 376, 66–70. Fraser, S.T., Ogawa, M., Yokomizo, T., Ito, Y., Nishikawa, S., Nishikawa, S., 2003. Putative intermediate precursor between hematogenic endothelial cells and blood cells in the developing embryo. Dev. Growth Differ. 45, 63–75. Fried, B., Haseeb, M.A., 1991. Platyhelminthes: Aspidogastrea, Monogenea and Digenea. In: Harrison, F.W., Bogitsch, B.J. (Eds.), Microscopic Anatomy of Invertebrates. vol. 3, Platyhelminthes and Nemertinea. Wiley-Liss, New York, NY, pp. 141–210. Friedman, M.M., Weiss, L., 1979. Fine-structure of blood follicles in the earthworm genera Amynthas and Lumbricus (Annelida, Oligochaeta). J. Morph. 161, 123–144. Gierer, A., Meinhardt, H., 1972. A theory of biological pattern formation. Kybernetik 12, 30–39. Gittenberger-de Groot, A.C., Vrancken Peeters, M.P., Mentink, M.M., Gourdie, R.G., Poelmann, R.E., 1998. Epicardium-derived cells con­ tribute a novel population to the myocardial wall and the atrioven­ tricular cushions. Circ. Res. 82, 1043–1052. González-Crussi, F., 1971. Vasculogenesis in the chick embryo. An ultrastructural study. Am. J. Anat. 130, 441–460.

Chapter  |  8.1  Origin of the Vertebrate Endothelial Cell Lineage: Ontogeny and Phylogeny

Guadix, J.A., Carmona, R., Muñoz-Chapuli, R., Pérez-Pomares, J.M., 2006. In vivo and in vitro analysis of the vasculogenic potential of avian proepicardial and epicardial cells. Dev. Dyn. 235, 1014–1026. Haque, M., Rashid, K.A., Stern, M.S., Sharma, P.K., Siddiqi, A.H., Vinogradov, S.N., Walz, D.A., 1992. Comparison of the hemoglobins of the platyhelminths Gastrothylax crumenifer and Paramphistomum epiclitum (Trematoda, Paramphistomatidae). Comp. Biochem. Physiol. B101, 673–676. Hartenstein, V., 2006. Blood cells and blood cell development in the animal kingdom. Annu. Rev. Cell Dev. Biol. 22, 677–712. Hartenstein, V., Mandal, L., 2006. The blood/vascular system in a phylo­ genetic perspective. BioEssays 28, 1203–1210. Hellstrom, M., Kalen, M., Lindahl, P., Abramsson, A., Betsholtz, C., 1997. Role of PDGF-B and PDGFR-beta in recruitment of vascular smooth muscle cells and pericytes during embryonic blood vessel formation in the mouse. Development 126, 3047–3055. Hermann, K., 1997. Phoronida. In: Harrison, F.W., Woollacott, R.M. (Eds.), Microscopic Anatomy of Invertebrates. vol. 13, Lophophorates, Entoprocta and Cycliophora. Wiley-Liss, New York, NY, pp. 207–236. Hertig, A.T., 1935. Angiogenesis in the early human chorion and in the primary placenta of the macaque monkey. Contrib. Embryol. Carnegie. Inst. 25, 37–81. Hirakow, R., 1983. Development of the cardiac blood vessels in staged human embryos. Acta. Anat. 115, 220–230. Hirschi, K.K., Rohovsky, S.A., Beck, L.H., Smith, S.R., D’Amore, P.A., 1999. Endothelial cells modulate the proliferation of mural cell pre­ cursors via platelet-derived growth factor-BB and heterotypic cell contact. Circ. Res. 84, 298–305. His, W., 1900. Lecitoblast und Angioblast der Wirbeltiere. Abhandl. Math-Phys. Ges. Wiss. 26, 171–328. Holland, N.D., Venkatesh, T.V., Holland, L.Z., Jacobs, D.K., Bodmer, R., 2003. AmphiNk2-tin, an amphioxus homeobox gene expressed in myocardial progenitors: insights into evolution of the vertebrate heart. Dev. Biol. 255, 128–137. Holthofer, H., Virtanen, I., Kariniemi, A.L., Hormia, M., Linder, E., Miettinen, A., 1982. Ulex europaeus I lectin as a marker for vascular endothelium in human tissues. Lab. Invest. 47, 60–66. Hungerford, J.E., Little, C.D., 1999. Developmental biology of the vascular smooth muscle cell: building a multilayered vessel wall. J. Vasc. Res. 36, 2–27. Hutchins, G.M., Kessler-Hanna, A., Moore, G.W., 1988. Development of the coronary arteries in the embryonic human heart. Circulation 77, 1250–1258. Jackson, C.J., Nguyen, M., 1997. Human microvascular endothelial cells differ from macrovascular endothelial cells in their expres­ sion of matrix metalloproteinases. Int. J. Biochem. Cell Biol. 29, 1167–1177. Jaffredo, T., Gautier, R., Eichmann, A., Dieterlen-Lievre, F., 1998. Intraaortic hemopoietic cells are derived from endothelial cells dur­ ing ontogeny. Development 125, 4575–4583. Jaffredo, T., Gautier, R., Brajeul, V., Dieterlen-Lievre, F., 2000. Tracing the progeny of the aortic hemangioblast in the avian embryo. Dev. Biol. 224, 204–214. James, M.A., 1997. Brachiopoda. Internal anatomy, embryology and develop­ ment. In: Harrison, F.W., Woollacott, R.M. (Eds.), Microscopic Anatomy of Invertebrates. vol. 13, Lophophorates, Entoprocta and Cycliophora. WileyLiss, New York, NY, pp. 297–408. Jamieson, B.G., 1992. Oligochaeta. In: Harrison, F.W., Woollacott, R.M. (Eds.), Microscopic Anatomy of Invertebrates. vol. 7, Annelida. Wiley-Liss, New York, NY, pp. 217–322.

483

Jenner, R.A., 2004. Towards a phylogeny of the Metazoa: evaluating alternative phylogenetic positions of Platyhelminthes, Nemertea, and Gnathostomulida, with a critical reappraisal of cladistic characters. Contrib. Zool. 73, 3–163. Johnson, A., Erdös, E.G., 1977. Metabolism of vasoactive peptides by human endothelial cells in culture: Angiotensin I converting enzyme (kininase II) and angiotensinase. J. Clin. Invest. 59, 684–695. Kabrun, N., Buhring, H.J., Choi, K., Ullrich, A., Risau, W., Keller, G., 1997. Flk-1 expression defines a population of early embryonic hemopoietic precursors. Development 124, 2039–2048. Kalka, C., Masuda, H., Takahashi, T., Gordon, R., Tepper, O., Gravereaux, E., Pieczek, A., Iwaguro, H., Hayashi, S.I., Isner, J.M., Asahara, T., 2000. Vascular endothelial growth factor (165) gene transfer augments circulating endothelial progenitor cells in human subjects. Circ. Res. 86, 1198–1202. Kallianpur, A.R., Jordan, J.E., Brandt, S.J., 1994. The SCL/TAL-1 gene is expressed in progenitors of both the hemopoietic and vascular systems during embryogenesis. Blood 83, 1200–1208. Keck, P.J., Hauser, S.D., Krivi, G., Sanzo, K., Warren, T., Feder, J., Connolly, D.T., 1989. Vascular permeability factor, an endothelial cell mitogen related to PDGF. Science 246, 1309–1312. Kessel, J., Fabian, B.C., 1986. The pluripotency of the extraembryonic mesodermal cells of the early chick blastoderm: effects of the AP and AOV environments. Dev. Biol. 116, 319–327. King, J., Hamil, T., Creighton, J., Wu, S., Bhat, P., McDonald, F., Stevens, T., 2004. Structural and functional characteristics of lung macro- and microvascular endothelial cell phenotypes. Microvasc. Res. 67, 139–151. Kisanuki, Y.Y., Hammer, R.E., Miyazaki, J., Williams, S.C., Richardson, J.A., Yanagisawa, M., 2001. Tie2-Cre transgenic mice: a new model for endothelial cell-lineage analysis in vivo. Dev. Biol. 230, 230–242. Kuprijanov, V.V., 1990. [Vascular endothelium (review). I. General mor­ phology. 2B: Phylogenesis of the vascular endothelium] (article in german). Gegenbaurs. Morphol. Jahrb. 136, 201–217. Lai, L., Bohnsack, B.L., Niederreither, K., Hirschi, K.K., 2003. Retinoic acid regulates endothelial cell proliferation during vasculogenesis. Development 130, 6465–6474. Licata, R.H., 1954. The human embryonic heart in the ninth week. Am. J. Anat. 94, 73–125. Licata, R.H., 1955. The development basis of the blood supply to the human heart. Anat. Rec. 121, 330–331. Licht, A.H., Raab, S., Hofmann, U., Breier, G., 2004. Endotheliumspecific Cre recombinase activity in flk-1-Cre transgenic mice. Dev. Dyn. 229, 312–318. Linask, K.K., Lash, J.W., 1993. Early heart development: dynamics of endocardial cell sorting suggests a common origin with cardiomyo­ cites. Dev. Dyn. 195, 62–69. Locke, M., 1998. Reticular systems and intercellular pymph spaces. In: Harrison, F.W., Woollacott, R.M. (Eds.), Microscopic Anatomy of Invertebrates. vol. 11A, Insecta. Wiley-Liss, New York, NY, pp. 51–71. Lough, J., Sugi, Y., 2000. Endoderm and heart development. Dev. Dyn. 217, 327–342. Magee, J.C., Stone, A.E., Oldham, K.T., Guice, K.S., 1994. Isolation, cul­ ture, and characterization of rat lung microvascular endothelial cells. Am. J. Physiol. 267, L433–L441. Maisonpierre, P.C., Suri, C., Jones, P.F., Bartunkova, S., Wiegand, S.J., Radziejewski, C., Compton, D., McClain, J., Aldrich, T.H., Papadopoulos, N., Daly, T.J., Davis, S., Sato, T.N., Yancopoulos, G.D.,

484

1997. Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis. Science 277, 55–60. Manasek, F.J., 1968. Embryonic development of the heart. I. A light and electron microscopic study of myocardial development in the early chick embryo. J. Morphol. 125, 329–365. Marletaz, F., Holland, L.Z., Laudet, V., Schubert, M., 2006. Retinoic acid signaling and the evolution of chordates. Int. J. Biol. Sci. 2, 38–47. Martin, G.G., Hose, J.E., 1992. Vascular elements and blood (hemolymph). In: Harrison, F.W., Humes, A.G. (Eds.), Microscopic Anatomy of Invertebrates. Vol. 10, Decapod Crustacea. Wiley-Liss, New York, NY, pp. 117–146. Martynova, M.G., Chaga, O.Y., 1997. Ultrastructural study of the heart and its relationship with the connective tissue sheath of the stomach in Rhynchonella psittacea (Tentaculata, Brachiopoda). J. Morph. 234, 69–77. Martynova, M.G., Chaga, O.Y., 2002. Heart and the peritoneal cover of the gut sinus in the polychaete Arenicola marina: an ultrastructural and autoradiographic study. J. Morph. 254, 312–319. Mato, M., Aikawa, E., Kishi, K., 1964. Some observations on interstices between mesoderm and endoderm in the area vasculosa of chick blastoderm. Exp. Cell Res. 35, 426–428. McDonald, D.A., 1960. Blood Flow in Arteries. Williams and Wilkins, Baltimore, MD. Medvinsky, A.L., Samoylina, N.L., Muller, A.M., Dzierzak, E.A., 1993. An early pre-liver intraembryonic source of CFU-S in the developing mouse. Nature 364, 64–67. Medvinsky, A., Dzierzak, E., 1996. Definitive hematopoiesis is autono­ mously initiated by the AGM region. Cell 86, 897–906. Meinhardt, H., 1982. Models of Biological Pattern Formation. Academic Press, London, UK. Mikawa, T., Fischman, D.A., 1992. Retroviral analysis of cardiac mor­ phogenesis: discontinuous formation of coronary vessels. Proc. Natl. Acad. Sci. USA 89, 9504–9508. Mikawa, T., Gourdie, R.G., 1996. Pericardial mesoderm generates a pop­ ulation of smooth muscle cells migrating into the heart along with ingrowth of the epicardial organ. Dev. Biol. 174, 221–232. Millauer, B., Wizigmann-Voos, S., Schnurch, H., Martinez, R., Moller, N.P., Risau, W., Ullrich, A., 1993. High affinity VEGF binding and devel­ opmental expression suggests Flk-1 as a major regulator of vasculo­ genesis and angiogenesis. Cell 72, 835–846. Moldovan, N.I., Goldschmidt-Clermont, P.J., Parker-Thornburg, J., Shapiro, S.D., Kolattukudy, P.E., 2000. Contribution of monocytes/ macrophages to compensatory neovascularization: the drilling of metalloelastase-positive tunnels in ischemic myocardium. Circ. Res. 87, 378–384. Moller, P.C., Philpott, C.W., 1973. The circulatory system of amphioxus (Branchiostoma floridae). I. Morphology of the major vessels of the pharyngeal area. J. Morph. 139, 389–406. Moore, A.W., McInnes, L., Kreidberg, J., Hastie, N.D., Schedl, A., 1999. YAC complementation shows a requirement for Wt1 in the develop­ ment of epicardium, adrenal gland and throughout nephrogenesis. Development 126, 1845–1857. Morello, J.P., Plamondon, J., Meyrick, B., Hoover, R., O’ConnorMccourt, M.D., 1995. Transforming growth factor-beta receptor expression on endothelial cells: heterogeneity of type III receptor expression. J. Cell Physiol. 165, 201–211. Motoike, T., Markham, D.W., Rossant, J., Sato, T.N., 2003. Evidence for novel fate of Flk1 progenitor: contribution to muscle lineage. Genesis 35, 153–159.

PART  |  8  Making Vessels

Moyon, D., Pardanaud, L., Yuan, L., Breant, C., Eichmann, A., 2001. Plasticity of endothelial cells during arterial–venous differentiation in the avian embryo. Development 128, 3359–3370. Munier, A.I., Doucet, D., Perrodou, E., Zachary, D., Meister, M., Hoffmann, J.A., Janeway, C.A., Lagueux, M., 2002. PVF2, a PDGF/VEGF-like growth factor, induces hemocyte proliferation in Drosophila larvae. EMBO Rep. 3, 1195–1200. Muñoz-Chápuli, R., Pérez-Pomares, J.M., Macias, D., García-Garrido, L., Carmona, R., González, M., 1999. Differentiation of hemangioblasts from embryonic mesothelial cells? A model on the origin of the ver­ tebrate cardiovascular system. Differentiation 64, 133–141. Muñoz-Chápuli, R., González-Iriarte, M., Carmona, R., Atencia, G., Macias, D., Pérez-Pomares, J.M., 2002. Cellular precursors of the coronary arteries. Tex. Heart Inst. J. 29, 243–249. Muñoz-Chápuli, R., Carmona, R., Guadix, J.A., Macías, D., Pérez-Pomares, J.M., 2005. The origin of the endothelial cells: an evo–devo approach for the invertebrate/vertebrate transition of the circulatory system. Evol. Dev. 7, 351–358. Murasawa, S., Asahara, T., 2005. Endothelial progenitor cells for vascu­ logenesis. Physiology (Bethesda) 20, 36–42. Murray, G., Garcia-Arrarás, J.E., 2004. Myogenesis during holothurian intestinal regeneration. Cell Tissue. Res. 318, 515–524. Nielsen, C., 1995. Animal Evolution: Interrelationships of the Living Phyla. Oxford University Press, Oxford, UK. Noden, D.M., 1990. Origins and assembly of avian embryonic blood vessels. Ann. NY Acad. Sci. 588, 236–249. Noden, D.M., 1991. Development of craniofacial blood vessels. In: Feinberg, R.N., Sherer, G.K., Auerbach, R. (Eds.), The Development of the Vascular System. Karger, Basel, Switzerland, pp. 1–24. North, T., Gu, T.L., Stacy, T., Wang, Q., Howard, L., Binder, M., Marin-Padilla, M., Speck, N.A., 1999. Cbfa2 is required for the forma­ tion of intra-aortic hemopoietic clusters. Development 126, 2563–2575. Ogden, J.A., 1988. The origin of the coronary arteries (abstract). Circulation 38 (Suppl. 4), 150. Oike, Y., Ito, Y., Hamada, K., Zhang, X.Q., Miyata, K., Arai, F., Inada, T., Araki, K., Nakagata, N., Takeya, M., Kisanuki, Y.Y., Yanagisawa, M., Gale, N.W., Suda, T., 2002. Regulation of vascu­ logenesis and angiogenesis by EphB/ephrin-B2 signaling between endothelial cells and surrounding mesenchymal cells. Blood 100, 1326–1333. Olah, I., Medgyes, J., Glick, B., 1988. Origin of aortic cell clusters in the chicken embryo. Anat. Rec. 222, 60–68. Orkin, S.H., 1992. GATA-binding transcription factors in hemopoietic cells. Blood 80, 575–581. Ostadal, B., 1979. Developmental relationships between the structure, blood supply and metabolic pattern of the vertebrate heart. Cor. Vasa. 21, 380–386. Ostadal, B., Schiebler, T.H., Rychter, Z., 1975. Relations between devel­ opment of the capillary wall and myoarchitecture of the rat heart. Adv. Exp. Med. Biol. 53, 375–388. Pan, C.M., Watabe, N., 1989. Periostracum formation and shell regenera­ tion in the lingulid Glottidia Pyramidata (Brachiopoda, Inarticulata). Trans. Amer. Microsc. Soc. 108, 283–298. Pardanaud, L., Luton, D., Prigent, M., Bourcheix, L.M., Catala, M., Dieterlen-Lievre, F., 1996. Two distinct endothelial lineages in ontogeny, one of them related to hemopoiesis. Development 122, 1363–1371. Pass, G., 1998. Accesory pulsatile organs. In: Harrison, F.W., Locke, M. (Eds.), Microscopic Anatomy of Invertebrates. vol. 11B, Insecta. Wiley-Liss, New York, NY, pp. 621–640.

Chapter  |  8.1  Origin of the Vertebrate Endothelial Cell Lineage: Ontogeny and Phylogeny

Pennisi, D.J., Mikawa, T., 2005. Normal patterning of the coronary capil­ lary plexus is dependent on the correct transmural gradient of FGF expression in the myocardium. Dev. Biol. 279, 378–390. Pérez-Pomares, J.M., Macías, D., Garcia-Garrido, L., Muñoz-Chápuli, R., 1997. Contribution of the primitive epicardium to the subepicar­ dial mesenchyme in hamster and chick embryos. Dev. Dyn. 210, 96–105. Pérez-Pomares, J.M., Macías, D., Garcia-Garrido, L., Muñoz-Chápuli, R., 1998a. The origin of the subepicardial mesenchyme in the avian embryo: an immunohistochemical and quail-chick chimera study. Dev. Biol. 200, 57–68. Pérez-Pomares, J.M., Macías, D., Garcia-Garrido, L., Muñoz-Chápuli, R., 1998b. Immunolocalization of the vascular endothelial growth factor receptor-2 in the subepicardial mesenchyme of hamster embryos: identification of the coronary vessel precursors. Histochem. J. 30, 627–634. Pérez-Pomares, J.M., Macías, D., García-Garrido, L., Muñoz-Chápuli, R., 1999. Immunohistochemical evidence for a mesothelial contribution to the ventral wall of the avian aorta. Histochem. J. 31, 771–779. Pérez-Pomares, J.M., Muñoz-Chápuli, R., 2002. Epithelial–mesenchymal transitions: a mesodermal cell strategy for evolutive innovation in Metazoans. Anat. Rec. 268, 343–351. Pérez-Pomares, J.M., Carmona, R., González-Iriarte, M., Atencia, G., Wessels, A., Muñoz-Chápuli, R., 2002a. Origin of coronary endothe­ lial cells from epicardial mesothelium in avian embryos. Int. J. Dev. Biol. 46, 1005–1013. Pérez-Pomares, J.M., Phelps, A., Sedmerova, M., Carmona, R., González-Iriarte, M., Muñoz-Chápuli, R., Wessels, A., 2002b. Experimental studies on the spatiotemporal expression of WT1 and RALDH2 in the embryonic avian heart: a model for the regulation of myocardial and valvuloseptal development by epicardially derived cells (EPDCs). Dev. Biol. 247, 307–326. Pérez-Pomares, J.M., Carmona, R., Gonzalez-Iriarte, M., Macías, D., Guadix, J.A., Muñoz-Chápuli, R., 2004. Contribution of mesotheliumderived cells to liver sinusoids in avian embryos. Dev. Dyn. 229, 465–474. Pérez-Pomares, J.M., Foty, R.A., 2006. Tissue fusion and cell sorting in embryonic development and disease: biomedical implications. Bioessays 28, 809–821. Pober, J.S., Gimbrone, M.A., Lapierre, L.A., Mendrick, D.L., Fiers, W., Rothlein, R., Springer, T.A., 1986. Overlapping patterns of activation of human endothelial cells by interleukin 1, tumor necrosis factor, and immune interferon. J. Immunol. 137, 1893–1896. Poole, T.J., Coffin, J.D., 1989. Vasculogenesis and angiogenesis: two dis­ tinct morphogenetic mechanisms establish embryonic vascular pat­ tern. J. Exp. Zool. 251, 224–231. Ranger, A.M., Grusby, M.J., Hodge, M.R., Gravallese, E.M., de la Brousse, F.C., Hoey, T., Mickanin, C., Baldwin, H.S., Glimcher, L.H., 1998. The transcription factor NF-ATc is essential for cardiac valve formation. Nature 392, 186–190. Rhodes, C.P., Ratcliffe, N.A., Rowley, A.F., 1982. Presence of coelomo­ cytes in the primitive chordate amphioxus (Branchiostoma lanceolatum). Science 217, 263–265. Rieger, R.M., Lombardi, J., 1987. Ultrastructure of celomic lining in echinoderm podia. Significance for concepts in the evolution of mus­ cle and peritoneal cells. Zoomorphology 107, 191–208. Risau, W., 1997. Mechanisms of angiogenesis. Nature 386, 671–674. Risau, W., Lemmon, V., 1988. Changes in the vascular extracellular matrix during embryonic vasculogenesis and angiogenesis. Dev. Biol. 125, 441–450.

485

Risau, W., Flamme, I., 1995. Vasculogenesis. Annu. Rev. Cell Dev. Biol. 11, 73–91. Ruppert, E.E. (1997). Cephalochordata (Acrania). In: Harrison, F.W., Ruppert, E.E. (Eds.), Microscopic Anatomy of Invertebrates. vol. 15, Hemichordata, Chaethognatha and the invertebrate Chordates. WileyLiss, New York, NY, pp. 349–504. Ruppert, E.E., Carle, K.J., 1983. Morphology of the metazoan circulatory systems. Zoomorphology 103, 193–208. Rychter, Z., Ostadal, B., 1971. Fate of “sinusoidal” intertrabecular spaces of the cardiac wall after development of the coronary vascular bed in the chick embryo. Folia. Morphol. 19, 31–44. Sato, T.N., Tozawa, Y., Deutsch, U., Wolburg-Buchholz, K., Fujiwara, Y., Gendron-Maguire, M., Gridley, T., Wolburg, H., Risau, W., Qin, Y., 1995. Distinct roles of the receptor tyrosine kinases Tie-1 and Tie-2 in blood vessel formation. Nature 376, 70–74. Schipp, R., 1987. General morphological and functional-characteristics of the cephalopod circulatory-system – an introduction. Experientia 43, 474–477. Sharpe, M.J., Lee, D.L., 1981. Observations on the structure and function of the hemoglobin from the cuticle of Nippostrongylus brasiliensis (Nematoda). Parasitology 83, 411–424. Sherer, G. K., 1991. Vasculogenic mechanisms and epithelio–mesenchymal specificity in endodermal organs. In: Feinberg, R.N., Sherer, G.K., Auerbach, R. (Eds.), The Development of the Vascular System. Kar­ ger, Basel, Switzerland, pp. 37–57. Shigei, T., Tsuru, H., Ishikawa, N., Yoshioka, K., 2001. Absence of endothelium in invertebrate blood vessels: significance of endothe­ lium and sympathetic nerve/medial smooth muscle in the vertebrate vascular system. Jpn. J. Pharmacol. 87, 253–260. Smart, N., Risebro, C.A., Melville, A.A., Moses, K., Schwartz, R.J, Chien, K.R., Riley, P.R., 2007. Thymosin beta4 induces adult epi­ cardial progenitor mobilization and neovascularization. Nature 445, 177–182. Smiley, S., 1994. Holothurioidea. In: Harrison, F.W., Chia, F.S. (Eds.), Microscopic Anatomy of Invertebrates. vol. 14, Echinodermata. Wiley-Liss, New York, NY, pp. 401–472. Strong, P.A., Bogitsh, B.J., 1973. Ultrastructure of lymph system of trematode Megalodiscus-temperatus. Trans. Amer. Microsc. Soc. 92, 570–578. Stuckmann, I., Evans, S., Lassar, A.B., 2003. Erythropoietin and retinoic acid, secreted from the epicardium, are required for cardiac myocyte proliferation. Dev. Biol. 255, 334–349. Sugi, Y., Lough, J., 1994. Anterior endoderm is a specific effector of ter­ minal cardiac myocyte differentiation of cells from the embryonic heart forming region. Dev. Dyn. 200, 155–162. Sugi, Y., Lough, J., 1995. Activin-A and FGF-2 mimic the inductive effects of anterior endoderm on terminal cardiac myogenesis in vitro. Dev. Biol. 168, 567–574. Suri, C., Jones, P.F., Patan, S., Bartunkova, S., Maisonpierre, P.C., Davis, S., Sato, T.N., Yancopoulos, G.D., 1996. Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis. Cell 87, 1171–1180. Tettamanti, G., Grimaldi, A., Valvassori, R., Rinaldi, L., de Eguileor, M., 2003. Vascular endothelial growth factor is involved in neoangio­ genesis in Hirudo medicinalis (Annelida, Hirudinea). Cytokine 22, 168–179. Tevosian, S.G., Deconinck, A.E., Tanaka, M., Schinke, M., Litovsky, S.H., Izumo, S., Fujiwara, Y., Orkin, S.H., 2000. FOG-2, a cofactor for GATA transcription factors, is essential for heart morphogenesis and development of coronary vessels from epicardium. Cell 101, 729–739.

486

Timmerman, L.A., Grego-Bessa, J., Raya, A., Bertran, E., Perez-Pomares, J.M., Diez, J., Aranda, S., Palomo, S., McCormick, F., Izpisua-Belmonte, J.C., de la Pompa, J.L., 2004. Notch promotes epithelial–mesenchymal transition during cardiac development and oncogenic transformation. Genes. Dev. 18, 99–115. Tufro, A., Norwood, V.F., Carey, R.M., Gomez, R.A., 1999. Vascular endothelial growth factor induces nephrogenesis and vasculogenesis. J. Am. Soc. Nephrol. 10, 2125–2134. Turbeville, J.M., 1986. An ultrastructural analysis of coelomogenesis in the hoplonemertine Prosorhochmus americanus and the polychaete Magelona sp. J. Morph. 187, 51–60. Turbeville, J.M., 1991. Nemertinea. In: Harrison, F.W., Bogitsch, B.J. (Eds.), Microscopic Anatomy of Invertebrates. vol. 3, Platyhelminthes and Nemertinea. Wiley-Liss, New York, NY, pp. 285–328. Vernall, D.B., 1962. The human embryonic heart in the seventh week. Am. J. Anat. 111, 17–24. Vinogradov, S.N., Walz, D.A., Pohajdak, B., Moens, L., Kapp, O.H., Suzuki, T., Trotman, C.N., 1993. Adventitious variability. The amino-acid sequences of nonvertebrate globins. Comp. Biochem. Physiol. B. 106, 1–26. Viragh, S., Challice, C.E., 1981. The origin of the epicardium and the embryonic myocardial circulation in the mouse. Anat. Rec. 201, 157–168. Viragh, S., Gittenberger-de Groot, A.C., Poelmann, R.E., Kalman, F., 1993. Early development of quail heart epicardium and associated vascular and glandular structures. Anat. Embryol. 188, 391–393. Vokes, S.A., Krieg, P.A., 2002. Endoderm is required for vascular endothelial tube formation, but not for angioblast specification. Development 129, 775–785. Vokes, S.A., Yatskievych, T.A., Heimark, R.L., McMahon, J., McMahon, A.P., Antin, P.B., Krieg, P.A., 2004. Hedgehog signal­ ing is essential for endothelial tube formation during vasculogenesis. Development 131, 4371–4380. Voyta, J.C., Via, D.P., Butterfield, C.E., Zetter, B.R., 1984. Identification and isolation of endothelial cells based on their increased uptake of acetylated-low density lipoprotein. J. Cell Biol. 99, 2034–2040. Vrancken Peeters, M.P.F.M., Gittenberger-de Groot, A.C., Mentink, M.M.T., Hungerford, J.E., Little, C.D., Poelmann, R.E., 1997a. Differences in development of coronary arteries and veins. Cardiovasc. Res. 36, 101–110. Vrancken Peeters, M.P.F.M., Gittenberger-de Groot, A.C., Mentink, M.M.T., Hungerford, J.E., Little, C.D., Poelmann, R.E., 1997b.

PART  |  8  Making Vessels

The development of the coronary vessels and their differentiation into arteries and veins in the embryonic quail heart. Dev. Dyn. 208, 338–348. Vrancken Peeters, M.J., Gittenberger-de Groot, A.C., Mentink, M.M., Poelmann, R.E., 1999. Smooth muscle cells and fibroblasts of the coronary arteries derive from epithelial–mesenchymal transformation of the epicardium. Anat. Embryol. 199, 367–378. Watt, S.M., Gschmeissner, S.E., Bates, P.A., 1995. PECAM-1: its expres­ sion and function as a cell adhesion molecule on hemopoietic and endothelial cells. Leuk. Lymphoma. 17, 229–244. Wessels, A., Pérez-Pomares, J.M., 2004. The epicardium and epicardially derived cells (EPDCs) as cardiac stem cells. Anat. Rec. 276, 43–57. Wilm, B., Ipenberg, A., Hastie, N.D., Burch, J.B., Bader, D.M., 2005. The serosal mesothelium is a major source of smooth muscle cells of the gut vasculature. Development 132, 5317–5328. Wilt, F.H., 1965. Erythropoiesis in the chick embryo: the role of the endoderm. Science 147, 1588–1590. Wood, H.B., May, G., Healy, L., Enver, T., Morriss-Kay, G.M., 1997. CD34 expression patterns during early mouse development are related to modes of blood vessel formation and reveal additional sites of hematopoiesis. Blood 90, 2300–2311. Wunsch, A., Little, C.D., Markwald, R.R., 1994. Cardiac endothelial heterogeneity defines valvular development as demonstrated by the diverse expression of JB3. Dev. Biol. 165, 585–601. Yamaguchi, T.P., Dumont, D.J., Conlon, R.A., Breitman, M.L., Rossant, J., 1993. flk-1, an flt-related receptor tyrosine kinase is an early marker for endothelial cell precursors. Development 118, 489–498. Yamashita, J., Itoh, H., Hirashima, M., Ogawa, M., Nishikawa, S., Yurugi, T., Naito, M., Nakao, K., Nishikawa, S., 2000. Flk1-positive cells derived from embryonic stem cells serve as vascular progenitors. Nature 408, 92–96. Young, P.E., Baumhueter, S., Lasky, L.A., 1995. The sialomucin CD34 is expressed on hemopoietic cells and blood vessels during murine development. Blood 85, 96–105. Zhang, H., Huang, Z., Yamaguchi, K., Tomonaga, S., 1992. Granulocytes and macrophages in amphioxus. Zool. Sci. (Tokyo) 9, 113–118. Zhang, X.Q., Takakura, N., Oike, Y., Inada, T., Gale, N.W., Yancopoulos, G.D., Suda, T., 2001. Stromal cells expressing ephrinB2 promote the growth and sprouting of ephrin-B2() endothelial cells. Blood 98, 1028–1037.

Chapter 8.2

Vascular Development Ondine Cleaver1 and Paul A. Krieg2 1

Department of Molecular Biology, University of Texas Southwestern Medical Center, Dallas, TX, USA Department of Cell Biology and Anatomy, University of Arizona College of Medicine, Molecular Cardiovascular Research Program, Tucson, AZ, USA 2

I.  Introduction to vascular development The cardiovascular system is the first functional organ formed during embryogenesis, arising long before other organs and tissues. Abnormalities in its assembly or function almost invariably lead to embryonic lethality. The reason that cardiovascular function is critical to the survival of higher organisms lies in the fact that every cell must receive nutrition and eliminate wastes via blood vessels. Studies have demonstrated that the cells in complex tissues are generally located within about 100–200 m of an endothelial-lined blood vessel, which is the diffusion limit for oxygen, otherwise they perish from starvation and/or asphyxiation (Folkman, 1971). Therefore, endothelial cells, together with other cardiovascular cell types such as heart muscle and blood, are without a doubt the most critical cell types in the developing embryo. During formation, the general pattern of the embryonic vascular system is strikingly stereotyped within a species, and highly conserved between different vertebrate species (Fig. 1). The defining cell type of the vascular system is the endothelial cell, which forms the seamless lining of the entire circulatory system, including the heart and all arteries and veins, large and small. Given the architectural conservation of the vasculature between species, it is likely that endothelial cells arise and assemble following genetically hard-wired molecular cues that are conserved across vertebrates. The morphogenesis of the embryonic vasculature begins with the appearance of angioblasts in most mesodermal tissues (Fig. 2A). Angioblasts are defined as endothelial precursor cells which have not yet incorporated into the endothelial lining of blood vessels (Fig. 2A, 2A) (His, 1900), and throu­ ghout this chapter we will use the terms angio-blast and endothelial precursor cell interchangeably. Some time after Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

their specification in the mesoderm, the angioblasts associate into lumenless vascular “cords” (Fig. 2B, B). These subsequently rearrange into tubes, and mature into the blood vessels that form the primitive vascular system of the embryo and soon carry blood. The process of angio-blast assembly into a vessel occurs via a process termed “vasculogenesis”, which refers to the coalescence of individual angioblasts into vascular primordia (Fig. 2C) (Risau and Flamme, 1995). After the initial vasculature, or “primary vascular plexus”, is formed, it is elaborated and extended throughout the embryo via a process termed “angiogenesis”, which includes both the formation of new vessels via sprouting angiogenesis (Fig. 2C) and the morphogenetic changes in vessel size or arrangement which occur via angiogenic remodeling (Fig. 2D) (Risau, 1997). Sprouting angiogenesis refers to the de novo formation of new vessels via local proliferation and extension of endothelial cells from the wall of an existing vessel. Angiogenic remodeling involves a range of different cellular rearrangements, including the splitting, enlargement and/or regression of existing vessels, generally resulting in an overall morphological alteration in the vascular network. The final stage of vascular development is vessel stabilization, or maturation, which involves a dramatic reduction in endothelial cell proliferation, morphological changes to endothelial cell shape and organization, and the recruitment of vascular wall components, particularly smooth muscle cells (Fig. 2D, 2D). Each of these developmental processes will be covered in greater detail in this chapter. Driven in part by the study of blood vessel growth during tumor angiogenesis, a great deal of research in the last few decades has focused on the function of growth factors and their receptors in regulating the proliferation and development of vascular tissue. Despite rapid progress in these tumor studies, our understanding of vascular development in the embryo is surprisingly limited. For example, little is known 487

488

PART  |  8  Making Vessels

Dorsal aorta Lung bud

Vitelline/ omphalomesenteric Common artery cardinal vein Vitelline Posterior vein cardinal vein Chorianic villus

Aortic arch III

Anterior cardinal vein

Ventral aortic root Placenta Umbilical/allantoic vein

Internal carotid artery

Yolk sac

Umbilical/allantoic artery

Figure 1  Major features of the vertebrate cardiovascular system. Endothelial cells create the inner lining of all vessels in the vertebrate embryonic vasculature, including the major vessels (shown here) and the minor vessels, such as small arterioles, venules and capillaries (not shown). Major vessels include the endocardium, the dorsal aortae, the cardinal veins (anterior and posterior), the carotid arteries, the vitelline arteries and veins, the common cardinal veins (or Ducts of Cuvier) and the allantoic/umbilical vein and artery. Generally, the blood flows from the heart, up the ventral aorta, through the aortic arches, down the dorsal aorta and to the rest of the embryo. Blood then returns to the sinus venosus of the heart via the vitelline veins and common cardinal veins.

concerning the precise origins of angioblasts, or the nature of the signals responsible for patterning the embryonic vasculature. Consequently, we cannot explain with any certainty why a particular blood vessel forms precisely at one location rather than another, or why certain embryonic tissues remain avascular. Ultimately, knowledge of the function and regulation of the molecules involved in vascular development will lead not only to an understanding of this critical embryonic process, but it will also facilitate the development of diagnostic and therapeutic applications relevant to a wide range of vascular pathologies, from cancer to diabetic retinopathies, to congenital vascular malformations. This chapter will focus on what is known regarding the cellular and molecular mechanisms that regulate embryonic vascular development, and will also endeavor to point out those questions that remain poorly understood (see also Chapters 8.3 and 8.4).

II.  The origin of endothelial cells II.A.  Endothelial Origins in the Mesoderm Where do endothelial cells come from? Although their precise origin has long remained elusive, it is clear that

endothelial cell differentiation occurs exclusively in the mesoderm (Coffin and Poole, 1988; Noden, 1989). Tissues arising from endoderm and ectoderm, in contrast, are initially avascular. The brain which derives from neurectoderm, for instance, must be vascularized by ingrowth of vessels (sprouting angiogenesis), since no resident angioblasts arise in this tissue (Pardanaud and DieterlenLievre, 1993). Initial in situ differentiation of endothelial cells occurs in mesoderm that is in contact with underlying endoderm, differentiating first in splanchnic rather than somatic mesoderm (Mato et al., 1964; Wilt, 1965; Meier, 1980; Pardanaud et al., 1989a; Palis et al., 1995). This has led to the suggestion that endoderm may play a role in endothelial specification (Wilt, 1965; Pardanaud et al., 1989a,b). However, despite the proximity of these tissues, an inductive role for the endoderm cannot be essential for angio-blast specification, since several organs such as the allantois, the kidney and the coronary mesenchyme, give rise to angioblasts in the absence of endoderm (PerezPomares et al., 1998; Robert et al., 1998; Cumano et al., 2000). Furthermore, angioblasts still differentiate in mesoderm that has been experimentally separated from

489

Chapter  |  8.2  Vascular Development

A

A� Ectoderm Mesenchyme

Endoderm

Free angioblasts develop in the mesoderm

B

B�

Angioblast cells

Angioblasts assemble into cords

C�

C

Primary capillary plexus

Endothelial cells

Cord angioblasts form tubes and differentiate into endothelial cells

D

Non-sprouting angiogenesis (intussusception)

Sprouting angiogenesis

D�

Mesenchyme/ smooth muscle cells

Further remodeling and maturation; recruitment of pericytes and smooth muscle.

Smooth muscle cells Pericytes

Figure 2  Overview of vascular development. (A–D) Schematic of vessel formation at the cellular level (lateral view of single vessel). (A–D) Overview of vascular network formation and elaboration, during vasculogenesis and angiogenesis. (A, A) Initially, angioblasts differentiate as individual cells within the mesoderm. (B, B) Then, vascular “cords” form by the aggregation of angioblasts either at the location where they emerge, or at a distance, following migration. These cords of angioblasts are initially lumenless aggregates. (C, C) The endothelial cells in the cords differentiate, reorganize and form tubes with distinct lumens (i.e., acquire patency). In addition, new vessels also form from sprouting angiogenesis and intussusception. (D, D) The primary vascular plexus is reshaped by angiogenic remodeling, resulting in the formation of a hierarchical network of large and small vessels. During remodeling, the endothelium matures and is stabilized by the recruitment of mesenchymal cells that differentiate into smooth muscle cells and pericytes and become components of the vascular wall.

490

endoderm, but subsequent blood vessel growth is disorganized (Vokes and Krieg, 2002). Therefore, although it is likely that endodermal signals influence vascular development, these signals do not appear to be required for specifying angio-blast fate.

II.B.  Molecules Regulating Differentiation of Endothelial Cells II.B.i.  Fibroblast Growth Factor (FGF) Despite the fundamental importance of the question, rema­rkably little is known about the molecular mechanisms regulating angio-blast specification and differentiation. At present, a limited body of evidence points to a role for FGF signaling, but this is unlikely to represent the complete picture. Members of the FGF family (especially bFGF) play a critical role in the induction of the mesodermal germ layer during the earliest stages of embryogenesis. In Xenopus, FGFs are potent inducers of ventral mesoderm which will contribute a subpopulation of cells to form the blood islands, and will also contribute to the endothelial lineage (Isaacs et al., 1992; Mills et al., 1999). Experiments using an in vitro system of avian epiblast culture have shown that FGF induces expression of endothelial markers (Flamme et al., 1995a). The ability to induce marker expression, however, is limited to the first 24 hours of culture, suggesting that FGFs are important very early in the establishment of the endothelial lineage. General support for an FGF induction model is provided by studies in the frog embryo, where the vasculogenic mesoderm and endothelial cells fail to develop in Xenopus embryos in which FGF signaling has been blocked using a dominant negative receptor (Amaya et al., 1991). However, in both the avian and frog studies it remains unclear whether FGF has a specific role in the specification and differentiation of the angio-blast lineage, or whether FGF primarily functions to promote mesoderm formation. Recent studies have addressed this issue by examining FGF signaling in the avian embryo after mesoderm has been specified (Nakazawa et al., 2006). Insertion of beads soaked with FGF4 into the lateral mesoderm stimulated strong expression of endothelial markers in the vicinity of the beads. Conversely, inhibition of FGF signaling resulted in reduction in expression of endothelial markers. These experiments suggest that FGFs are indeed likely to participate directly in the specification of mesodermal cells to an endothelial cell fate.

II.C.  Extraembryonic and Intraembryonic Angioblasts Until recently our knowledge of the origins and development of the vascular system was based largely on studies

PART  |  8  Making Vessels

employing classic embryological techniques. Early studies, often using the avian embryo, suggested that the embryonic vasculature might originate by invasion of the embryo by vessels from extraembryonic tissues (His, 1868, 1900). This hypothesis was in part based on the fact that the first overt signs of blood vessel development are evident on the yolk sac, outside the embryo proper, and that directed branching of vessels occurs towards the embryo. In classical experiments using quail embryos, the area pellucida (embryo proper) was separated from the area opaca (yolk sac) prior to the formation of vessels and, significantly, blood vessel development was still observed within the embryo (Regan, 1915). This demonstrated not only that endothelial cells arise from both intra- and extraembryonic tissues, but also that the two vascular systems are not developmentally dependent on one another. These results have been confirmed in other organisms, demonstrating that endothelial cells originate both in the early mesoderm which contributes to extraembryonic tissues, such as the yolk sac, and also in later mesoderm which contributes to intraembryonic tissues (Lawson et al., 1991; Palis et al., 1995; Kinder et al., 1999). In amniotes, there are important differences between extra- and intraembryonic angioblasts. In extraembryonic tissues, angioblasts arise in close association with developing blood within structures called “blood islands” (see below). In contrast, early intraembryonic angioblasts are scattered throughout the mesoderm, and are not closely associated with blood-forming tissue (Noden, 1989; Pardanaud et al., 1989a). These observations suggest that, from the earliest stages of embryogenesis, intra- and extraembryonic endothelial cells have different origins and developmental potential.

II.D.  Vascular Studies: Classical Embryology to Molecular Breakthroughs The early stages of vascular development in the embryo were first revealed in studies using either light or electron microscopy to examine developing blood vessels directly. Initial development of vessels was observed at the mesoderm–endoderm interface with the assembly of clusters and vesicles of endothelial precursor cells immediately preceding actual vessel formation (Meier, 1980; Hirakow and Hiruma, 1981, 1983). However, microscopy studies were fundamentally limited by the fact that endothelial cells could not be identified until they associated into groups. The characterization of antibodies specific to endothelial cells in the quail, such as MB1 (Peault et al., 1983; Labastie et al., 1986) and QH-1 (Pardanaud et al., 1987; Coffin and Poole, 1988) provided a major breakthrough in the study of vascular development. These antibodies facilitate the identification of endothelial cells soon after they arise in the mesoderm, long before their association into

491

Chapter  |  8.2  Vascular Development

vascular tissue. Subsequent molecular studies, using in situ hybridization of vascular genes or transgenic animals with reporter genes expressed from vascular promoters, have also allowed the detection of individual angioblasts long before they are morphologically detectable. Using QH-1 to detect quail endothelial cells, transplants of various quail tissues into chick embryos showed that all intraembryonic mesodermal tissues except the prechordal plate contain migrating endothelial precursor cells (Noden, 1989). Furthermore, in vivo lineage analysis reveals that some myogenic and endothelial cells in the somites arise from common progenitors in the lateral dermomyotome (Kardon et al., 2002). Other avian experiments show that angioblasts derived from the splanchnopleuric mesoderm are capable of developing into hematopoietic cells, while those derived from the somatopleuric mesoderm are more restricted to the development of endothelial cells (Pardanaud et al., 1996). These studies suggest that angioblasts arising in different embryonic tissues are likely to have different developmental properties.

III.  Blood islands Among the earliest endothelial cells are those that arise within vascular structures called the “blood islands” (Fig. 3A) (Wilt, 1974; Pardanaud et al., 1987; Peault et al., 1988). They have been defined as “mesodermal cell aggregates”, in which the inner cells are blood or hematopoietic stem cells and the outer cells are angioblasts (Fig. 3B,C) (Ferkowicz and Yoder, 2005). In other words, the blood islands are essentially small, closed, endothelial bags of blood. Blood islands have thus been thought to represent transitional structures that presage blood vessel development. Following their formation, individual blood islands extend towards each other and undergo anastomosis (fusion and connection), forming a continuous primitive plexus of vascular tubes (Haar and Ackerman, 1971). Blood islands have primarily been described in the extraembryonic tissues of amniotes. Blood islands were first observed in avian embryos by Florence Sabin and Alexander Maximow, who described their formation in the area vasculosa of the yolk sac (Sabin, 1917; Sabin, 1920). Analogous structures were later identified in the murine yolk sac (Haar and Ackerman, 1971). Blood island observations contributed to the classical view that blood cells always originate intravascularly. This idea, however, has been challenged by a number of more recent studies in the mouse. For instance, isolation of cells from gastrulating embryos showed the presence of blood cell precursors within the primitive streak, long before morphological evidence of blood island formation. Blood cells were detected both in the proximal region of the streak that gives rise to the yolk sac, and in the distal region that gives rise to the embryo proper (Palis et al., 1999). In addition, lineage

tracing studies of gastrulating mouse embryos demonstrate that primitive blood progenitors exit early from the posterior primitive streak and arise independently of the majority of endothelial cells (Kinder et al., 1999). Additional insights into the origins of both endothelium and blood can be gained from the analysis of blood islands in other species. Blood island morphology varies greatly across different classes of vertebrates, and even between different species within a class (Goss, 1928). While the blood islands of mouse and avian embryos arise in the mesodermal layer of the extraembryonic yolk sac, in fish and Amphibia, which lack extraembryonic tissues, blood islands arise in intraembryonic mesoderm. In teleosts, including zebrafish, embryonic hematopoiesis occurs in the intermediate cell mass (ICM) and in the rostral blood island (RBI) (de Jong and Zon, 2005). The intermediate cell mass is thought to be equivalent to mammalian blood islands, as it gives rise to both blood and endothelial cells, while the rostral blood island predominantly gives rise to macrophages. In frogs, blood islands develop in the ventral-most intraembryonic mesoderm (Ackerman, 1971; Kau and Turpen, 1983; Bertwistle et al., 1996; Haar and Walmsley et al., 2002). In Xenopus, a single large blood island is found in the ventralmost mesoderm, where a patch of hematopoietic precursors expressing globin is surrounded at the edges by a wide ring of angioblasts. Although the amphibian blood island is presumably functionally equivalent to those of the chick and mouse, it differs in morphology since the endothelial cells do not enclose the blood cells within a lumen. Overall, these studies suggest that although blood islands are intriguing localized structures, exhibiting tantalizing proximity of endothelial and blood precursors, this association may not be representative of the mechanisms by which blood and vasculature form in the embryo itself.

IV.  The hemangioblast The intimate temporal and spatial association of hemato­ poietic and endothelial cell development in the blood islands led to the hypothesis that both lineages arise from a common precursor called the “hemangioblast” (His, 1900; Sabin, 1920; Murray, 1932; Wagner, 1980; Risau and Flamme, 1995). This hypothesis has been reinforced by the fact that endothelial and blood precursors share a number of markers, and mutation of a number of genes in mouse and zebrafish affects both cell types. Surprisingly, although the hemangioblast theory is nearly a century old, the existence of a cell with both endothelial and hemato­ poietic potential has not yet been conclusively confirmed in the embryo. However, evidence continues to accumulate in support of the existence of a hemangioblast, however rare and transient this cell type may be. It now seems likely that, although most blood and endothelial cells arise from independent progenitors, at least some blood and endothelial cells are derived from a bipotential precursor cell.

492

PART  |  8  Making Vessels

(A) Ectoderm Hematopoietic cells Endothelial cells

Undifferentiated Hemangiomesoderm blasts

Hemangioblastic aggregate

Blood island

Capillary

Endoderm Embryonic area

(B)

Extraembryonic area

(C)

Figure 3  Blood islands. (A) Angio-blast and hematopoietic cell differentiation in embryonic mesoderm. Both endothelial and hematopoietic cells emerge in undifferentiated mesoderm, possibly from a common precursor called a hemangioblast. Precursors are thought to aggregate and differentiate into blood islands, which consist of both endothelial (outer cells) and blood cells (inner cells). Blood islands enlarge and fuse, resulting in continuous vessels containing blood. (B) Blood vessels and blood in the early mouse embryo. Recent observations have noted that initial yolk sac vessels are distant from the majority of yolk sac blood. It has been proposed that endothelial cells migrate proximally from the distal portion of the embryo and encapsulate the extraembryonic blood. (C) Schematic of blood islands in the mouse yolk sac. Endothelial cells surround hematopoietic cells. (B, C) Adapted from Ferkowicz and Yoder (2005).

Some of the strongest support for the existence of the hemangioblast came from expression and genetic studies of the vascular endothelial growth factor (VEGF) receptor Flk-1 (also VEGFR2 or KDR). Flk-1 is a receptor tyrosine kinase initially expressed in both hematopoietic and endothelial lineages (Matthews et al., 1991; Eichmann et al., 1993; Millauer et al., 1993a; Yamaguchi et al., 1993; Dumont et al., 1995). This expression, however, is only maintained in the endothelial lineage. Gene ablation studies demonstrated that

mice mutant for Flk-1 develop neither blood nor vascular tissue (Shalaby et al., 1995), suggesting that a single Flk-1expressing cell type may be necessary for both the hemato­ poietic and endothelial cell types. However, despite this dramatic phenotype, it remains unclear whether Flk-1 marks the hemangioblast precursor. First, Flk-1 protein is not detected in the inner hematopoietic cells of the blood islands (Drake and Fleming, 2000). Second, Flk-1 is not essential for hematopoiesis, since Flk-1-mutant embryonic stem cells can

493

Chapter  |  8.2  Vascular Development

give rise to hemoglobin-expressing blood islands (Shalaby et al., 1995). Together, these data suggest that Flk-1 is not required cell-autonomously for hematopoiesis, but is perhaps necessary for positioning the endothelial and blood lin­ eages in the proper microenvironment to allow blood island development.

IV.A.  In Vitro Studies: Embryonic Stem Cell Differentiation The existence of a cell with hemangioblast properties has been directly demonstrated by in vitro differentiation of embryonic stem (ES) cells (Kennedy et al., 1997; Choi et al., 1998; Nishikawa et al., 1998; Park et al., 2005). Single ES cells can form colonies known as embryoid bodies (EBs), which can mimic in vivo events, such as yolk sac blood island-like formation and generation of both hematopoietic and endothelial progeny (Doetschman et al., 1985; Keller et al., 1993; Palis et al., 1999). Cells isolated from embryonic stem cell-derived embryoid bodies have been shown to give rise to both endothelial and blood cells in clonal cultures (Kennedy et al., 1997; Choi et al., 1998). The clonal nature of these experiments provided the most compelling evidence for a hemangioblast-like precursor for both lineages. However, it remains unclear if these experiments truly identified a bipotential hemangioblast, or a more multipotent mesodermal precursor, with the capacity to form a wider variety of derivatives.

IV.B.  In Vitro Studies: Embryonic Cells Further in vitro experiments supporting the hemangioblast theory have been carried out using cells isolated from embryonic tissues. When the avian epiblast is isolated and cultured in vitro, hematopoietic and angioblastic cell lineages arise spontaneously (Murray, 1932; Zagris, 1980; Mitrani et al., 1990). These experiments suggest that committed hemangioblast precursors may already be present in the epiblast prior to gastrulation. However, when avian epiblast cells are dissociated, cultured in vitro and allowed to form embryoid bodies, no blood island differentiation is observed (Flamme and Risau, 1992; Eichmann et al., 1997). Addition of FGF to dissociated avian epiblast cells restores the capacity to form blood islands, suggesting that a single growth factor is sufficient for the induction of both cell lineages, or possibly for the induction of a common hemangioblast precursor (Flamme and Risau, 1992; Krah et al., 1994). This is in contrast to mouse embryonic stem cell-derived EBs, in which blood islands develop without the addition of FGF (Risau et al., 1988). The difference in the development of blood islands in the avian and mouse systems has not been explained, but it may involve different endogenous levels of FGF.

Cell mixing and limiting dilution analysis of cells isolated from gastrulating mouse embryos has identified progenitors with hemangioblast capability located in the posterior portion of the primitive streak (Palis et al., 1995, 1999; Huber et al., 2004). These experiments demonstrate that hematopoietic and vascular cell fate commitment occurs long before blood island development in the yolk sac. Most importantly, however, they demonstrate that bipotential precursor cells are exceedingly rare, with one study finding only one-to-five identifiable hemangioblastlike cells per embryo (Huber et al., 2004).

IV.C.  In Vivo Studies Perhaps the most compelling studies have been carried out in whole animals, and these confirm that the hemangio­ blast is an incredibly rare cell type unlikely to give rise to whole blood islands or intraembryonic endothelial cells. Clonal analysis of mouse yolk sac blood islands using multichimeric mice, unequivocally demonstrated that the origin of blood and endothelial cells within the blood islands is polyclonal (Ueno and Weissman, 2006). Specifically, these authors state that: “contribution by hemangioblast progenitors to both endothelial and hematopoietic lineages within an island, if it happens at all, is an infrequent event”. Together with the studies of isolated embryonic cells (described above), these data suggest that the hemangioblast is a rare and transient cell type and is unlikely to be the precursor of most endothelial cells in the embryo. Despite these findings, the fact that endothelial cells and blood cells are often found in close association implies either a lineage relationship, or reciprocal signaling between the two cell types. For example, intraembryonic blood vessel formation is associated with hematopoiesis during epicardial-driven coronary blood vessel development (Mikawa and Gourdie, 1996; Perez-Pomares et al., 1998). In another example, endothelial cells in the ventral portion of the dorsal aorta have been shown to give rise to hematopoietic precursors (Jaffredo et al., 2005; Park et al., 2005; Cumano and Godin, 2007). Specifically, both marker expression and chick-quail chimera studies demonstrated the presence of para- and intra-aortic clusters of rounded cells in the floor of the dorsal aorta, and these have been associated with the production of intraembryonic hematopoietic cells (Tavian et al., 1996; DieterlenLievre, 1997; Minko et al., 2003). However, since these blood-producing cells appear to derive from resident aortic endothelial cells they have been designated as “transdifferentiated hematopoietic precursors”, or as “hemogenic endothelium”, rather than hemangioblasts. Another possibility is that hematopoietic precursors actually arise in the mesenchyme ventral to the aorta, but give the appearance of emerging from aortal endothelial cells as they cross the endothelium (Cumano and Godin, 2007).

494

V.  The endothelial cell The fundamental cell type that makes up the vascular system is the endothelial cell. In an adult human, the vasculature consists of approximately 1  1013 endothelial cells, making up a complex yet coherent organ system (Sumpio et al., 2002). How does one define an endothelial cell? Only two unifying characteristics can truly be applied to all endothelial cells (Aird, 2003). The first is anatomic, in that endothelial cells form the seamless lining of all blood vessels, or the “endothelium”, which constitutes the “plumbing” for the entire cardiovascular system. The second characteristic is functional, in that endothelial cells provide the selectively-permeable interface that separates all tissues from oxygen-carrying blood. However, beyond these two characteristics no definition applies universally to all endothelial cells. Even at the level of marker expression, it is difficult to identify a gene that is expressed uniformly in all endothelial cells and does not show expression in other cell types. Overall, endothelial cells are dynamic, adaptive, heterogeneous and plastic cells that coordinate in a seamless way to form the closed and functional vasculature.

V.A.  Cellular Mechanisms of Blood Vessel Assembly The development of a functional vasculature requires precise spatio–temporal regulation. Initially, angioblasts arise as isolated cells and then proliferate in order to accommodate the growing vascular network. A proportion of angio­ blasts migrate towards one another, over and past other nonendothelial cell types, to precise, genetically-determined locations where they gather together prior to forming actual vessels. On contact, angioblasts recognize and adhere to each other, forming a lumenless strand, or cord (Fig. 4A). Following these basic (and poorly-understood) cellular events, angioblasts establish apicobasal polarity, form tight junctions and create a central lumen that can accommodate blood, a process termed tubulogenesis. Once tubes are established, they elaborate further via sprouting of endothelial cells from pre-existing vessels, which requires dramatic shape changes and extensive cell migration (Fig. 4B). These coordinated cellular processes result in the initial assembly of endothelial cells into a complex and functional network of endothelial-cell-lined blood vessels, which subsequently grows and extends to all tissues.

V.A.i.  Endothelial Migration Migration of endothelial cells is essential for much early blood vessel development (Schmidt et al., 2007a). Many studies have confirmed the migratory ability of angioblasts prior to the formation of the primitive vascular plexus (Noden, 1989, 1990; Poole and Coffin, 1989; Christ et al.,

PART  |  8  Making Vessels

1990; Wilting et al., 1995). Experiments using chick-quail chimeras show that angioblasts originating in transplanted tissue are highly invasive and can migrate quickly over long distances (Noden, 1989, 1990). These cells invade the surrounding mesenchyme and contribute indiscriminately to the formation of veins, arteries and capillaries in the vicinity of the site of implantation. Individual angioblasts have the potential to migrate significant distances from their site of origin, and migratory distances up to 400 m have been observed (Klessinger and Christ, 1996). Studies in Xenopus have also demonstrated the importance of endothelial cell migration in the formation of the primary embryonic vasculature (Cleaver and Krieg, 1998). Using expression of Flk-1 to mark vascular precursor cells, angioblasts are first detected in the late neurula stage embryo, in a number of regions including the dorsal-most lateral plate mesoderm where the posterior cardinal veins will form. At this time, no endothelial cells are present at the site of the future dorsal aorta, the embryonic midline. During subsequent development, a subset of the lateral endothelial cells actively migrates under the somites to the midline, where they aggregate to form a single dorsal aorta (Gibson, 1910; Lofberg and Collazo, 1997; Cleaver and Krieg, 1998). A similar pattern of angio-blast migration towards the midline is observed during formation of axial vessels in zebrafish (Fouquet et al., 1997; Jin et al., 2005). The ability of endothelial cells to migrate, prior to vessel formation, is therefore inherent and critical to proper vascular network morphogenesis.

V.A.ii.  Tubulogenesis Formation of a vascular tube from the linear cord of angio­ blasts has been observed to occur by two distinct mechanisms (Fig. 5A). In the first mechanism, the lumen is formed by the enlargement of an extracellular space located between individual endothelial cells. Alternatively, the lumen may form by the alignment of intracellular space, in the form of large vacuoles. Classical observations in the avian embryo suggest the second mechanism, where angioblasts that have gathered into loose cords initiate the formation of tubes with the appearance of “slit-like spaces” within angioblasts (Houser et al., 1961). These spaces then enlarge when intracellular vacuoles fuse with each other at the center of the cell, resulting in the formation of a long, continuous, lumen (Fig. 5B,C) (Folkman and Haudenschild, 1980; Meyer et al., 1997; Davis et al., 2002; Lubarsky and Krasnow, 2003). Recently, this vacuole fusion mechanism has also been shown to occur during angiogenesis, in growing intersomitic vessels in zebrafish (Kamei et al., 2006). The alternative mechanism, in which cellular rearrangements underlie the transformation of a cord into a tube, has been described in zebrafish aorta formation (Fig. 5D) (Jin et al., 2005). Observations of staged embryos show that cells in the pre-aorta cord first make junctions between each

Chapter  |  8.2  Vascular Development

495

(A). Vasculogenesis

(B) Sprouting Angiogenesis

(C) Angiogenic Remodeling

(D) Vasculogenesis plus Angiogenesis

Figure 4  Vasculogenesis, sprouting angiogenesis and angiogenic remodeling. Schematic representation of the basic mechanisms of vascular development. (A) Vasculogenesis is the aggregation of angioblasts in the mesoderm to form blood vessels. Angioblasts coalesce either at the location where they emerge from the mesoderm, or they migrate through tissues and form blood vessels at a distant site. (B) Sprouting angiogenesis involves the formation of new vessels from pre-existing vessels by extension of new sprouts. Sprouting angiogenesis is responsible for the growth of blood vessels into many developing organs. It involves both the proliferation of the vessel stalk and migration of endothelial cells at the tips of the angiogenic sprouts. (C) Angiogenic remodeling involves the reshaping of a homogeneous vascular plexus into a hierarchical array of large and small vessels. (D) In some cases, for example the lung, both vasculogenesis and angiogenesis occur concurrently.

other, and then actively organize to form a lumen between them. This mechanism presumably involves active pumping of fluid into the lumen, possibly in a manner similar to zygote cavitation (Watson and Barcroft, 2001). It is possible that different vessels undergo different mechanisms of lumen formation during vasculogenesis and angiogenesis.

VI.  Vasculogenesis and angiogenesis “Vasculogenesis” and “angiogenesis”, mentioned above, are the two best-described and characterized cellular mechanisms underlying vascular system development (Fig. 4) (Risau et al., 1988; Pardanaud et al., 1989a; Risau

496

PART  |  8  Making Vessels

(A). Endothelial cell rearrangemant

Junctions cluster at the center of endothelial group

Early junction formation (B) a

flk:GFP F-actin ZO-1

b

flk:GFP F-actin ZO-1

Lumen formation c NT

flk:GFP F-actin ZO-1

NT NT NC NC

NC

17hpf

18hpf

30hpf

(C).Vacuole fusion

Angioblasts form into cord. Cells already contain vacuoles

Cells form junctions. Vacuoles enlarge to fill most of cell.

(D)

Figure 5  Vascular tubulogenesis. Schematic representation of vascular tube formation via alternative cellular mechanisms. (A) Endothelial cell rearrangement: patent vessels can form as a result of reorganization of endothelial cells within a primitive cord. Endothelial cells aggregate into a cord of cells. These cells then adhere to each other, forming cell–cell junctions. Junctions are either dissolved in between some cells, or they migrate along the plasma membrane to become localized within the cord center. Tight junctions then form as a lumen is generated between cells. The resulting arrangement of cells is an endothelial-lined tube. (B) Immunofluorescent images of endothelial cell rearrangements in the developing zebrafish aorta. In fish, endothelial cells undergo tubulogenesis via the model presented in (A). Images reproduced from Jin et al. (2005). (C) Vacuole fusion: an alternative mechanism involves both endothelial cell rearrangement and fusion of intracellular vacuoles. In this model, endothelial cells generate large intracellular vacuoles. Cells then adhere to one another and form intercellular junctions. Finally, vacuoles move to the center of the cord where they fuse, resulting in formation of a central, patent lumen. (D) Electron microscope images of endothelial cells during formation of the dorsal aorta in the chicken embryo. The pictures show a progression from early development (at left) to a fully-formed endothelial tube (at right). Images courtesy of Carlos Moran.

497

Chapter  |  8.2  Vascular Development

and Flamme, 1995; Wilting and Christ, 1996; Risau, 1997; Drake, 2003). Unfortunately, both terms strictly mean “the genesis of blood vessels”, however they have come to describe very different mechanisms by which blood vessels form. As mentioned above, vasculogenesis primarily refers to the de novo formation of the primary vascular plexus in the embryo, while angiogenesis describes the extension and remodeling of the primitive embryonic vasculature that occurs during normal growth of embryonic tissues. Vasculogenesis and angiogenesis therefore refer to distinct cellular events, and it is likely that they are regulated by different molecular mechanisms. Here, we will review each process in detail.

VI.A.  Vasculogenesis Vasculogenesis is generally defined as the formation of vessels via the aggregation of free angioblasts into cords, followed by tube formation (Figs 2A,B; 4A), although it can also involve the fusion of blood islands (Sabin, 1917; Risau et al., 1988; Poole and Coffin, 1989; Risau and Flamme, 1995). Vasculogenesis is therefore responsible for the formation of the primordia of major blood vessels, including the dorsal aorta and the endocardium, as well as the rather homogeneous capillary networks found in tissues such as the amniote yolk sac or the amphibian mesoderm, which will eventually be modeled into a more hierarchical and mature vascular network (Coffin and Poole, 1988; Poole and Coffin, 1989). Overall therefore, vasculogenesis encompasses a closely-coordinated and sequential series of steps, including differentiation, migration, adhesion and maturation, which together result in the coalescence of individual migratory angioblasts into a continuous tubular network. Based on experiments in quail, Poole and Coffin distinguished two types of vasculogenesis (Coffin and Poole, 1991). In vasculogenesis type I, angioblasts associate to form a mature vessel in situ at the location where they differentiate in the mesoderm. In vasculogenesis type I, there is no significant migration of angioblasts. In vasculogenesis type II, angioblasts may migrate significant distances from their original location and then associate into a vessel at a distant location. In the quail embryo, the dorsal aorta arises via vasculogenesis type I, while the endocardium and posterior cardinal veins arise via vasculogenesis type II. Longrange migration of angioblasts during initial formation of the vascular plexus in the avian embryo has been beautifully visualized in time-lapse movies prepared by the Little laboratory (Rupp et al., 2004). The formation of major vessels by vasculogenesis type I or II can differ between organisms. For example, unlike the situation in birds and mammals, the dorsal aorta forms by vasculogenesis type II in frog and fish embryos, with angioblasts migrating from lateral to medial regions in the embryo before undergoing tubulogenesis (Cleaver et al., 1997; Fouquet et al., 1997; Sumoy et al., 1997).

VI.A.i.  Organ Vasculogenesis Until fairly recently it was believed that the vascularization of most developing organs occurred through angiogenic sprouting and invasion from pre-existing vessels. However, improved methods for visualization of angioblasts, using molecular markers and vascular reporters, have revealed that many organs acquire blood vessels via de novo vasculogenesis (Drake, 2003). For example, vascularization of the liver, lung, pancreas, stomach/intestine, spleen and kidney have all been shown to occur, in large part, by vasculogenic aggregation of local angioblasts (Pardanaud et al., 1989a; deMello et al., 1997; Robert et al., 1998; Tufro et al., 1999; Gebb and Shannon, 2000; Matsumoto et al., 2001). As development proceeds, it seems likely that angiogenic branching from existing vessels also contributes to maintenance and extension of the primitive organ vasculature.

VI.B.  Molecules Regulating Vasculogenesis VI.B.i.  Vascular Endothelial Growth Factor (VEGF) and its Receptors Vascular Endothelial Growth Factor The VEGF family of growth factors consists of VEGFA, B, C, D, E and placental growth factor (PlGF). All are mitogens for endothelial cells that also regulate cell migration and survival (Ferrara et al., 2003; Roy et al., 2006). However, gene targeting shows that only VEGF-A plays an essential role in early vascular development, so we will limit our discussion to this particular family member. VEGF-A expression in the embryo is dynamic and occurs in all germ layers, often in tissues immediately adjacent to developing vascular structures (Breier et al., 1992; Dumont et al., 1995; Flamme et al., 1995a; Cleaver et al., 1997). VEGF-mediated signaling is responsible for both primary vessel formation by vasculogenesis, and angiogenic invasion of developing organs. In gene ablation studies, mice lacking a single VEGF allele die at about E10.5. These animals show gross abnormalities in vascular development, including defects in initial differentiation of endothelial cells, sprouting angiogenesis, lumen formation, the formation of large vessels and the spatial organization of the vasculature (Carmeliet et al., 1996; Ferrara et al., 1996). The striking vascular disruption in heterozygous-mutant embryos illustrates that tight regulation of VEGF levels is essential for correct vascular morphogenesis. It is important to acknowledge, however, that angioblasts are present in both VEGF and VEGF receptor knockout embryos, albeit at low numbers. This indicates that VEGF signaling is not strictly required for initial specification of the endothelial lineage (Duan et al., 2003). In addition, although the role of VEGF has primarily been understood as a paracrine signal that drives vessel assembly and growth, a recent study

498

has identified autocrine endothelial VEGF as critical for endothelial cell survival (Lee et al., 2007). Alternative splicing generates a number of different VEGF-A protein isoforms. These proteins differ in their bioactivity, due to differences in their receptor-binding affinity and their differential ability to diffuse in the extracellular environment. The larger forms of the protein (VEGF164, VEGF188 and VEGF205 in mouse) possess a domain that binds to extracellular matrix, and thereby reduces or prevents extensive diffusion of the protein. The smallest form of the protein lacks this domain and is freely diffusible. Although the molecular properties of the different VEGF-A proteins must indeed play a role in fine tuning VEGF signaling in specific tissues, they are not essential for general formation of the embryonic vasculature. Animals genetically-modified to express only the smallest VEGF-A isoform (VEGF120) are born alive and are generally healthy (Carmeliet et al., 1999). Similarly, animals that only express the larger, extracellular matrix (ECM)-binding isoforms (VEGF164 and VEGF188) are also viable and show grossly normal blood vessel development (Maes et al., 2004). VEGF expression can be modulated by physiological conditions. In response to low oxygen levels (hypoxia), VEGF expression is upregulated through the activity of the hypoxia inducible 1- (HIF1) transcription factor (Liu et al., 1995; Campochiaro, 2000). In the retina, hypoxia upregulates VEGF expression in astrocytes, and this regionalized expression serves to attract capillary growth from nearby vessels (Stone et al., 1995). Once the tissue is vascularized and the oxygen need is met, the production of VEGF declines. Studies of mouse embryos showed that hypoxic regions correlated with regions of high VEGF expression (Lee et al., 2001). Similarly, studies using chemical indicators of hypoxia in the quail embryo showed that the developing kidney and neural tissues exhibited low-oxygen status and corresponding upregulation of VEGF expression (Nanka et al., 2006). In both cases, the increased VEGF levels correlated with accelerated growth of blood vessels in the affected tissues. As the embryo increases in size, it is likely that VEGF upregulation in response to low oxygen is a normal developmental mechanism that drives blood vessel growth. Flk-1 (Vascular Endothelial Growth Factor Receptor 2) The receptor tyrosine kinase Flk-1 is a high affinity receptor for VEGF, and is critical for both vasculogenesis and angiogenesis. Although not strictly endothelial-specific (it is also expressed in some neuronal and blood cell precursors), Flk-1 is one of the most reliable markers of angio­ blasts and differentiated endothelial cells. Expression of Flk-1 is particularly high during embryonic vascularization and during tumor angiogenesis (Millauer et al., 1993b; Yamaguchi et al., 1993; Flamme et al., 1995a; Cleaver et al., 1997; Fouquet et al., 1997). The importance of Flk-1 for vascular development is revealed in gene-targeting

PART  |  8  Making Vessels

experiments. Mice that lack Flk-1 function die between E8.5 and E9.5 from dramatic defects in vascular development (Shalaby et al., 1995). Mutant animals show a greatly reduced number of angioblasts, and these fail to form into vascular cords. It is interesting that loss-of-function of the Flk-1 gene in zebrafish reveals a significantly less dramatic phenotype than the mouse knockout. In zebrafish, Flk-1-mutant animals show major disruption of angiogenic growth of blood vessels, but formation of the original vascular network and establishment of the blood lineage occurs normally (Habeck et al., 2002). These results suggest that other signaling pathways can substitute for VEGF/Flk-1 during the vasculogenic stages of zebrafish development. Flt-1 (Vascular Endothelial Growth Factor Receptor 1) The receptor tyrosine kinase Flt-1 shows similarities to Flk-1 in overall structure and expression pattern, however their functions during vascular development are very different. Like Flk-1, Flt-1 is a high-affinity receptor for VEGF and also for the VEGF-related growth factor, placental growth factor (PlGF) (de Vries et al., 1992; Waltenberger et al., 1994). Unlike the Flk-1-null mutants, targeted mutation of the mouse Flt-1gene results in normal development of endothelial cells in both intra- and extraembryonic tissues; however, these cells fail to properly assemble and organize into vessels (Fong et al., 1995). Rather than the reduction in endothelial cell number observed in Flk-1-mutant animals, Flt-1-deficient embryos show an increase in endothelial cell number in most tissues. Although the Flt-1 protein possesses an intracellular tyrosine kinase domain, disruption of this domain does not interfere with normal vascular assembly (Fong et al., 1999), suggesting that the intracellular portion of the receptor does not transduce active signaling, at least in the embryo. These observations have led to a model where Flt-1 normally functions to sequester excess VEGF ligand, and perhaps to limit the size of the population of endothelial precursor cells and subsequent endothelial cell proliferation.

VI.B.ii.  TGF Signaling Pathways Signaling by TGF1 through its high-affinity receptor TGFRII is essential for embryonic vasculogenesis. Mice mutant for either TGF1 or TGFRII die at midgestation, due to severe defects in assembly of the original capillary plexus of the yolk sac (Dickson et al., 1995; Oshima et al., 1996b). Although endothelial cell numbers appeared normal, many of the yolk sac vessels failed to assemble correctly and vessels that did assemble were leaky and weak compared to controls. These defects may be due to defects in establishing correct cell adhesion and failure to recruit smooth muscle cells. In contrast to the yolk sac, vasculogenesis in the embryo itself occurs more or less normally in mutant animals, indicating that a redundant pathway can

499

Chapter  |  8.2  Vascular Development

substitute for TGF signaling. At present the specific signaling pathway responsible for the rescuing activity is not known. Neuropilins Neuropilin-1 and -2 (NRP1 and NRP2 respectively) are transmembrane receptors for soluble semaphorin ligands. In neural tissues, NRP1 acts with semaphorin-3A in growth cone guidance. NRP1 and NRP2 are also expressed in endothelial cells, implying a role in vascular development, and NRP1 has been shown to function with VEGFR-2 as a specific co-receptor for the VEGF-165 isoform of VEGF-A (Kawasaki et al., 1999; Soker et al., 2002). NRP1-knockout mice exhibit numerous defects in vascular patterning and development, including failure of angiogenic investment into neural tissues, disorganization of the branchial arch vessels and deficient growth of the yolk sac vasculature (Kawasaki et al., 1999). NRP2-knockout animals are viable, but double NRP1; NRP2-knockout animals show defects significantly more severe than NRP1-mutants alone (Takashima et al., 2002). Double mutant embryos failed to generate an initial vascular plexus and exhibited almost complete absence of vascular development. Despite their demonstrated importance, the function of these proteins has been difficult to define. For example, mice expressing a mutant form of NRP1 that cannot bind semaphorins develop normally, suggesting that aberrant semaphorin/NRP1 signaling is not responsible for the defective vascular phenotype (Gu et al., 2003). Moreover, mice expressing only VEGF-120, which does not interact with NRP1, also undergo largely normal vascular development (Carmeliet et al., 1999), again suggesting that neuropilin function as a VEGF-165 receptor is not essential for vascular development. Based on the dramatic vascular phenotype of the NRP1;NRP2 double mutant, it will be extremely interesting to learn more about the signaling pathways mediated by the neuropilin proteins.

VI.C.  Extracellular Matrix and Cell   Adhesion Molecules Extracellular matrix components make up the microenvironment in which angioblasts migrate and org­ anize into the cords that will form the primary vascular plexus. Angioblasts themselves play an active role in depos­ iting and organizing the extracellular matrix, and composition of the matrix changes dynamically as vascular development proceeds (Schmidt et al., 2007a). The matrix may function as the scaffold that allows clusters of angio­ blasts to anchor themselves and exert mechanical forces on other clusters, thereby forming the interconnected poly­ gonal arrays that characterize the initial vascular plexus (Czirok et al., 2008). Similarly, association with the extracellular matrix is likely to provide a substrate for polarization of cells preceding vascular lumen formation. In vitro

studies, using two-dimensional and three-dimensional gel assays, have directly examined the ability of extracellular matrix components to stimulate endothelial cell proliferation, migration, differentiation or recruitment of pericytes or smooth muscle cells to the vascular wall (Davis and Senger, 2005). Just as the dynamic changes in the composition of the extracellular matrix substrate are important for endothelial behavior, so are the adhesive receptors that regulate the interactions of endothelial cells with their environment. A number of cell adhesion molecules, particularly cadherins, have been shown to be critical for proper vascular development (Vestweber, 2008). A summary of observations concerning selected extracellular matrix components and cell adhesion molecules, and their influence on vascular development, is presented below. Despite the huge body of evidence suggesting important regulation of vascular growth by extracellular matrix, primarily from studies of cells in culture, these effects have been remarkably difficult to define in the whole animal. This may be due to the multiplicity of the proteins representing certain classes of extracellular matrix components, or to different signaling pathways functioning redundantly to regulate cell behavior.

VI.C.i.  Fibronectin The original assembly of vascular cords from free angio­ blasts takes place in a fibronectin-rich extracellular matrix (Hynes, 1990, 2007). In the chick yolk sac, as neighboring angio-blast clusters fuse into the primary vascular plexus, they approach each other along fibrils containing fibronectin protein (Mayer et al., 1981). As soon as the basic vascular network is established, fibronectin decreases in the vicinity of developing blood vessels, and endothelial cells begin to produce laminin and collagen IV in increasing amounts. This dynamic pattern of expression of extracellular matrix components has been particularly well-studied during development of specific vascular structures, such as the endocardium (Risau and Lemmon, 1988; Drake et al., 1990), and the chick chorioallantoic membrane (Ausprunk et al., 1991). Mice lacking a functional fibronectin gene specify angioblasts normally, but exhibit severe defects in blood vessel development – in the most extreme cases, a complete disruption of assembly of the initial vascular network, including the dorsal aortae (George et al., 1993, 1997). Although the results were variable, depending on the genetic background, these studies indicate that fibronectin is essential for the earliest events in blood vessel development.

VI.C.ii.  Collagens Endothelial cells secrete a number of collagen proteins into the extracellular matrix (St Croix et al., 2000) and these in turn bind endothelial integrins located on the cell surface. A large number of mostly in vitro studies support a function for extracellular matrix collagen proteins

500

in regulation of endothelial cell behavior and vascular development. In fact, different members of the collagen family appear to effect different regulatory activities of endothelial cells. For example, endothelial tube formation in vitro is associated with deposition of collagens type I, III, IV and V (Iruela-Arispe et al., 1991). When capillary endothelial cells are cultured on interstitial collagens, such as collagens type I and III, rapid proliferation occurs in all directions (Madri and Williams, 1983). However, when basement membrane collagens, such as collagen type IV, are used as the culture substrate endothelial cells aggregate and form tube-like structures. In similar experiments, endothelial cells grown in a three-dimensional collagen type I matrix organized into networks of branching and anastomosing tubes (Montesano et al., 1983). Overall, evidence suggests that interaction of extracellular matrix collagens with integrin cell surface proteins results in a number of changes in actin organization in endothelial cells (Whelan and Senger, 2003), and this is likely to be responsible for the alterations in endothelial cell behavior. Mouse embryos lacking function of the collagen type I -chain gene show rupture of embryonic blood vessels (Lohler et al., 1984). This occurs late in development and reflects a structural role for collagen in maintenance of the vascular wall, rather than a function during assembly of the vascular network. Because of the large number of collagen proteins deposited in the extracellular matrix, the precise functions of individual components have been difficult to define, but in vitro and in vivo studies both argue that collagen proteins are required for early assembly of the vascular network.

VI.C.iii.  Integrins Integrins serve as receptors for a number of extracellular matrix components, including collagen, laminin and fibronectin, and are among the most extensively-studied cell adhesion molecules involved in vascular development (Stromblad and Cheresh, 1996; Serini et al., 2006). Integrins generally mediate cell–extracellular matrix interactions and occasionally cell–cell adhesion. They function as heterodimers that consist of an -subunit and a noncovalently associated -subunit, both of which are integral membrane proteins. Many different  and  subunits exist (15 and 8, respectively), and they can associate to form different functional receptors (Baldwin, 1996; Arnaout, 2002). Endothelial cells of large vessels express 21, 31, 51 and v3, while microvascular endothelial cells express 11, 61, 64 and v5 (Luscinskas and Lawler, 1994). Integrin 51 is the receptor for fibronectin, and blocking the function of either subunit results in major defects in early vasculogenesis. Mouse embryos which are lacking integrin 5-function die between E10 and E11, at least partly due to defective blood vessel and blood island formation (Yang et al., 1993). These defects resemble

PART  |  8  Making Vessels

those vascular abnormalities observed in fibronectin-null embryos. The importance of integrin 51 is also demonstrated by experiments in quail embryos, where blocking the binding of 1 to its ligands with an anti-integrin antibody results in vasculogenic defects, including failure of lumen formation in the dorsal aorta (Drake et al., 1992). Overall, it appears that loss of integrin 51 function cau­ ses vasculogenesis to be arrested after angio-blast cord formation, but prior to tubulogenesis. Another integrin implicated in vascular development is v3, which interacts with a wide variety of extracellular matrix components. Integrins v3 and v5 are expressed by endothelial cells during vascular development, and studies using quail embryos strongly suggested a function during angiogenic growth (Brooks et al., 1994a,b). Mouse knockout studies, however, have shown that loss of v3 function has limited effects in the mammalian embryo. Ablation of v results in vascular defects, but these are limited to leakage of brain and gut vessels, while overall development of the vasculature is normal (Bader et al., 1998). Similarly, mice lacking function of 3, 5 or both 3 and 5, are born alive, with no obvious effects on blood vessel development (Hodivala-Dilke et al., 1999; Huang et al., 2000; Reynolds et al., 2002). Recent studies have indicated that the most important function of integrin v is to regulate apoptosis of endothelial cells (Stupack et al., 2001; Kim et al., 2002). It appears that when integrins are not bound to an appropriate ligand (e.g., on another endothelial cell), a cell death pathway is initiated. In this way, integrins may regulate the spatial distribution of the vascular network by killing endothelial cells that stray into inappropriate locations.

VI.C.iv.  Vascular Endothelial-Cadherin Vascular endothelial cadherin (VE-cadherin or cadherin-5) is expressed specifically in endothelial cells, and mediates calcium-dependent homophilic binding at adherens junctions between cells (Suzuki et al., 1991; Lampugnani et al., 1992; Breier et al., 1996). Intracellularly, VE-cadherin associates with catenins and the actin cytoskeleton (Dejana, 1996). Expression analysis in mouse embryos reveals VE-cadherin gene expression in endothelial precursor cells and in the earliest blood vessels throughout the embryo (Breier et al., 1996). Knockout of the VE-cadherin gene in mouse embryos results in widespread disruption of the initial vascular network and early embryonic lethality (Gory-Faure et al., 1999). Strangely, these effects are more pronounced in anterior regions of the embryo, while posterior vessels appear normal. In the yolk sac, the blood islands remain as isolated structures which do not coalesce into an organized vascular plexus. More recent studies have suggested that VE-cadherin is not in fact required for initial assembly of the vascular network, but instead is required for maintenance of immature vessels (Crosby

501

Chapter  |  8.2  Vascular Development

et al., 2005). Therefore, although early vascular development occurs normally, the primitive vessels fall apart in the absence of VE-cadherin protein. It is interesting to note that VE-cadherin expression is regulated by another cell adhesion protein, N-cadherin. Specific ablation of Ncadherin function in endothelial cells results in embryonic death at midgestation due to major disruption of vascular structures (Luo and Radice, 2005). Molecular characterization revealed that loss of N-cadherin function led to a significant decrease in expression of VE-cadherin.

VI.D.  Molecules Regulating Tubulogenesis VII.D.i.  Extracellular Matrix and Rho GTPases The molecular and cellular mechanisms specifically responsible for vascular tube formation remain largely unknown. However, the general cellular pathways that appear to be important for angioblasts to undergo tubulogenesis have been identified. One recurrent theme is the requirement for endothelial cells in the vascular cord to rearrange their cytoskeletal architecture and then to establish apical–basal polarity, with the inside of the vascular tube functioning as the apical surface. Both of these events appear to require interactions between the endothelial cells and the surrounding extracellular matrix (Davis and Senger, 2005). For example, it has been shown that interactions between collagens in the extracellular matrix and integrins on the endothelial cell surface result in activation of the Rho GTPases (RhoA, Rac I and Cdc42). The importance of Rho GTPase function for vascular tube formation has been demonstrated in several in vivo and cell culture assays. As described earlier, one mechanism for lumen formation in endothelial tubes involves the fusion of multiple intracellular vacuoles. When chemical inhibitors of Rho GTPase activity were added to endothelial cells in culture, vacuole formation and subsequent lumen formation was inhibited (Bayless and Davis, 2002). In vivo studies using the mouse skin vascular assembly model support this requirement for Rho GTPase activity during vascular lumen formation, because expression of dominant negative RhoA in endothelial cells severely inhibits the formation of new blood vessels (Hoang et al., 2004).

VI.D.ii.  Endothelial Growth Factor-like   Domain 7 (Egfl7) Egfl7 is a secreted protein expressed strongly in embryonic endothelial cells and downregulated in mature vessels (Fitch et al., 2004; Parker et al., 2004). The Egfl7 protein contains EGF-like domains and an EMI domain that often appears in extracellular matrix-associated proteins that regulate cell adhesion. Morpholino-knockdown studies in the zebrafish embryo reveal an essential function for Egfl7 in the control of tubulogenesis (Parker et al.,

2004). Although endothelial cells differentiate normally and assemble into vascular cords at the correct locations, these cords are unable to undergo the subsequent steps of tube formation. It remains unclear precisely which cellular events of tubulogenesis are defective in the Egfl7 deficient zebrafish embryos; however, it is possible that adhesion of endothelial cells is one of the altered properties. Further characterization of Egfl7 function has been gained from mouse knockout studies (Schmidt et al., 2007b). Embryos lacking Egfl7 function exhibit defective vascular development and approximately 50% die before birth. Defects are far less severe than those observed in zebrafish, and are primarily restricted to angiogenic branching and blood vessel morphology, especially affecting the organization of cells within the vascular sprouts. Therefore, while Egfl7 may be important during zebrafish vascular tubulogenesis, it does not appear to regulate tube formation in the mammalian embryo.

VI.D.iii.  Hedgehogs and Patched Receptor Hedgehog proteins, Sonic hedgehog (Shh), Indian hedgehog (Ihh) and Desert hedgehog (Dhh) play an essential role during early formation of the vasculature (Nagase et al., 2008). When hedgehog growth factors bind to the transmembrane receptor, patched (Ptc), an intracellular protein called smoothened (Smo) is mobilized which activates the hedgehog signaling pathway (Wang et al., 2007). Mouse embryos lacking Shh, Ihh or Smo activity exhibit vascular defects (Pepicelli et al., 1998; Dyer et al., 2001; Byrd et al., 2002; Vokes et al., 2004). In the case of the Shh-null mutants, defects are largely limited to lung tissue, potentially due to rescue activity by Ihh and Dhh proteins in other tissues (Pepicelli et al., 1998). In contrast, the zebrafish Shh mutant, sonic-you, exhibits generalized defects in development of the embryonic vasculature, including failure of the vascular cords to undergo tube formation (Brown et al., 2000). Correct arterial–venous specification is also disrupted in sonic-you mutant zebrafish (Lawson et al., 2002). Knockout of the Ihh gene in mouse results in a high frequency of embryonic death due to incomplete formation of the yolk sac vasculature (Byrd et al., 2002). The blood vessels of the yolk sac are small in diameter, and the total number of vessels is reduced relative to wild-type. Similar to the Shh knockout, it is possible that the presence of other hedgehog proteins reduces the severity of the observed defects. Experiments using the hedgehog pathway inhibitor, cyclopamine, circumvent the issues of hedgehog redundancy. Avian embryos treated with cyclopamine exhibit severe defects in vasculogenesis, including a failure of angioblasts to form cords during assembly of the initial vascular plexus and subsequently to form tubes (Vokes et al., 2004). Similarly, mouse Smo mutants that lack all hedgehog signaling exhibit severe vascular defects (Byrd et al., 2002). In the yolk sac, the

502

primary capillary plexus forms, but fails to undergo normal remodeling. Within the embryo itself, many major blood vessels, including the dorsal aortae, fail to undergo tube formation (Vokes et al., 2004). There is still some question as to whether hedgehog molecules signal directly to endothelial precursor cells, or whether their effects are indirect, perhaps via modulation of the VEGF, Notch or angiopoietin signaling pathways (Byrd and Grabel, 2004).

PART  |  8  Making Vessels

may serve to seek out and fuse with other such processes (Klosovskii and Zhukova, 1963). Recent work on “tip cells” has focused on blood vessel growth in the retina (Fig. 7A) (Hellstrom et al., 2007). Tip cell filopodia are the primary location of VEGF response, and have been shown to drive vessel growth and migration (Fig. 7B,C) (Gerhardt and Betsholtz, 2005). As the new vessel extends and takes shape, the basement membrane forms along the newly-sprouting structure and a central lumen is generated (Ausprunk and Folkman, 1977).

VI.E.  Angiogenesis After the formation of the first vessels via vasculogenesis, the simple circulatory system is elaborated and extended via angiogenesis. One angiogenic mechanism is called “sprouting angiogenesis”, where pre-existing vessels branch into avascular tissues (Fig. 4B). In addition, primitive blood vessels can be modified by splitting or fusion, or by enlargement or shrinking of vessel diameter, via a process called “angiogenic remodeling” (Fig. 4C) (Folkman and Klagsbrun, 1987; Klagsbrun and D’Amore, 1991; Patan et al., 1996). Other aspects of angiogenic remodeling include pruning, vessel regression, anastomosis and changes in response to hemodynamic forces. The features which distinguish each type of angiogenesis are described below.

VI.E.i.  Sprouting Angiogenesis One of the angiogenesis mechanisms involves true sprouting of capillaries from pre-existing blood vessels of the primary vascular plexus (Fig. 4B). Initially, quiescent cells at a specific location along the vessel wall become different from surrounding endothelial cells and initiate a cascade of targeted cellular activities. First, proteolytic degradation of the extracellular matrix is coupled with proliferation of the sprouting endothelial cells, to allow the formation of a coherent extension from the primary vessel (Fig. 6A). In the chick, endothelial cells of the perineural plexus form capillary sprouts that degrade the perineural basement membrane and then migrate into and invade the neurectoderm (Fig. 6B) (Bar, 1980). Eventually, these new angiogenic vessels are stabilized by the formation of a basement membrane and ensheathing by smooth muscle cells or pericytes (Fig. 6C). The cells located at the tip of the extending angiogenic sprouts (tip cells) have attracted attention recently. New sprouts have been observed to grow into the interstitium by ameboid migratory motion of the distal endothelial tip cells, either invading surrounding tissue or fusing with the endothelium of an adjacent vessel (Wagner, 1980). As the sprouts extend, new cells are added by mitotic proliferation of the pre-existing endothelial cells along the stalk of a blood vessel (Gerhardt et al., 2003). In angiogenic extensions in the brain, the endothelial cells at the tip of the sprout exhibit a number of “filiform” processes that

VI.E.ii.  Organ Angiogenesis Recent studies have shown that blood vessel formation of at least some organ systems involves in situ assembly of vasculature from endothelial precursor cells (see Section VI.A above). However, numerous vessels are clearly formed by sprouting angiogenesis, including intersomitic veins and arteries, and vessels of the neural and retinal vasculature (Coffin and Poole, 1988; Gerhardt and Betsholtz, 2005; Kamei et al., 2006). Sprouting angiogenesis is also likely the predominant mechanism of vessel formation in the later embryo and in the adult. This includes the normal processes of somatic growth, corpus luteum formation, placental formation and tissue regeneration (Sariola et al., 1983; Folkman and Klagsbrun, 1987; Kadokawa et al., 1990; Klagsbrun and D’Amore, 1991; Augustin et al., 1995). In adults, sprouting angiogenesis is also linked to pathological processes such as tumor growth, inflammatory reactions, wound healing and diabetic retinopathies (Sholley et al., 1984; Folkman and Shing, 1992; Ferrara, 1995; Folkman, 1995; Hanahan and Folkman, 1996). It seems likely that angiogenesis and vasculogenesis occur simultaneously during embryonic development (Fig. 4D). For example, during formation of lung vasculature, it appears that circulating endothelial cells integrate into growing angiogenic sprouts, contributing as much to vessel growth as endothelial proliferation (Baldwin, 1996).

VI.E.iii.  Angiogenic Remodeling The primary capillary plexus is initially formed as a rather homogeneous array of vessels; however, it is rapidly remodeled and modified into the familiar hierarchical continuum of larger and smaller blood vessels. These changes occur via a second type of angiogenesis called “angiogenic remodeling” (Fig. 4C) (Beck and D’Amore, 1997; Risau, 1997). Angiogenic remodeling generally remains poorly-understood, despite the fact that a large number of genes have been identified that when mutated result in failure of vascular remodeling. This is particularly striking in the mouse yolk sac, where the initial array of vessels fails to reorganize into large and small vessels in a number of mutants. The resulting hemodynamic insufficiency is almost always lethal. Angiogenic remodeling

503

Chapter  |  8.2  Vascular Development

Angiogenic sprouting

(A). Removal of pericytes and thinning of basement membrane Outer basement membrane Pericytes

Endothelial cells

Thinning of basement membrane

(B). Breakdown of basement membrane and formation of new sprout Degradation of basement membrane and ECM

Tip cell

(C). New stabilized vessel

Tip of growing sprout

Figure 6  Angiogenic sprouting. Schematic of a new vessel forming via the sprouting of endothelial cells from a pre-existing vessel. (A) Initially, the formation of a new vessel requires destabilization of associated mural cells (blue). (B) Then, endothelial cells within the vessel wall undertake a number of activities, including degrading the endothelial basement membrane, via the production of matrix metalloproteases (MMPs), changing the orientation of their cell division, leaking plasma factors into the extravascular space, and migration of tip cells into that space. (C) Once tip cells have migrated and generated a new endothelial sprout, stalk cells proliferate, extending the new vessel. The vessel becomes stabilized by the formation of a new basement membrane and the recruitment of mural cells.

generally involves a wide range of morphogenetic changes that dynamically alter blood vessel size or architecture. Capillaries of the initial embryonic plexus either remodel or regress during development, accommodating the coordinated growth and differentiation of other tissues. Once the vascular system is mature, it is relatively stable and undergoes angiogenic remodeling only in female reproductive

tissues, during wound healing, or during pathological processes such as tumor growth. Here, we discuss specific types of remodeling that occur in developing tissues, each involving distinct cellular mechanisms. In a number of developing tissues, pre-existing vessels can be split and reorganized by the process of intussusception, resulting in expansion of capillary networks

504

PART  |  8  Making Vessels

(A). Endothelial Filopodia

(B)

Tip cell Very low proliferation

(C).

Figure 7  Endothelial filopodia. Sprout­ ing angiogenic vessels have distinct, specialized “tip” cells that display long filopodial extensions. (A) Cells located at the tip of growing vessels sense and navigate the environment via filopodia, while cells of the vessel stalk proliferate and undergo tubulogenesis to form the vascular lumen. Tip cells have been shown to respond to gradients of VEGF-A in the microenvironment, via activated Flk-1 receptors on the tip cell filopodia. (B, C) Endothelial tip cells in the mouse retina. (B) Tip cells at growing edge of retinal vascular plexus. (C) Close-up showing individual tip cells. Arrows indicate long filopodial extensions. Images from Gerhardt et al. (2003).

Stalk cells Proliferation of stalk cells

(Caduff et al., 1986; Burri and Tarek, 1990; Patan et al., 1992; Patan et al., 1996; Burri and Djonov, 2002). During intussusception, proliferation of endothelial cells within a vessel results in the formation of a large lumen that is then split by the insertion of endothelial columns, called tissue “pillars” or “‘posts” (Fig. 8A). These pillars are formed when a disk-like zone of contact is established between opposite walls of a vessel, in a manner reminiscent of a stalactite extending from the roof of a cave towards the floor. In the zone of contact intercellular junctions are formed between endothelial cells, and the contact zone becomes perforated centrally (Fig. 8B). The perforated region becomes stabilized by recruitment of pericytes and the deposition of connective tissue fibers, such as collagen. The pillar can now enlarge along the length of the vessel, splitting a single vessel into two (Fig. 8B,C). In the developing lung, both intussusception and sprouting angiogenesis occur, however intussusception is believed to predominate (Burri and Djonov, 2002). This is in contrast to the developing brain, where sprouting angiogenesis predominates. Another angiogenic remodeling process has been described as “the intercalated growth of blood vessels” (Folkman and Klagsbrun, 1987). In this process, mitosis of endothelial cells within a vessel leads to an increase in vessel diameter or length. This mechanism is important during healing of endothelial wounds (Reidy and Schwartz, 1981), and in the development of the coronary arteries (Bogers et al., 1989).

VI.E.iv.  Endothelial Regression Another variety of vessel remodeling involves endothelial regression (Im and Kazlauskas, 2006). Key steps in vascular regression include changes in endothelial cell shape, lumen narrowing, increased vacuolation, cessation of blood flow, detachment from the basement membrane and nuclear condensation associated with cell death (Fig. 9). Regression of vessels often occurs as a result of either reduction of blood flow, cessation of VEGF-mediated maintenance, or via other genetically-determined processes, such as changes in expression of angiogenic or repulsive cues in surrounding nonendothelial tissues. Blood vessel regression is an intrinsic feature of vascular remodeling and has been studied in a number of systems, including embryonic and postnatal development (Latker et al., 1986; Bartel and Lametschwandtner, 2000), limb vasculature (Feinberg et al., 1986; Hallmann et al., 1987), ocular capillaries (Ausprunk et al., 1978; Meeson et al., 1996), ovarian angiogenesis (Augustin et al., 1995; Modlich et al., 1996; Suzuki et al., 1998) and models of VEGF inhibition in tumors and normal tissues (Inai et al., 2004; Baffert et al., 2006). A landmark study involved the implantation of an FGF-expressing pellet into the avascular cornea of a mouse, which induced growth of vessels, followed by its removal and subsequent regression of those vessels (Ausprunk et al., 1978). In eye and trachea models, empty “sleeves” of vascular basement membrane have

505

Chapter  |  8.2  Vascular Development

(A)

B

A C Intussusception (B)

(C)

Figure 8  Intussusception. Intussusception, or non-sprouting angiogenesis, generates an increase in vessel numbers by splitting of pre-existing vessels. In primitive capillary networks sprouting angiogenesis or intussusception can occur simultaneously. Intussusception involves the formation of transcapillary “pillars” or “posts” that split a capillary blood vessel into two. Initially the pillar creates a small intervascular space (A), but the space subsequently enlarges (B and C) and forms a much larger intervascular region. After they have formed, the resulting intervascular spaces cannot be distinguished from those created by sprouting angiogenesis. (C) The result of intussusception is that two vessels are created from a single pre-existing vessel.

been observed to remain following endothelial regression. These structures were described as “acellular capillaries” or “basement membrane ghosts”, and marked the prev­ ious locations of vessels (el-Hifnawi et al., 1994; Baffert et al., 2004; Baffert et al., 2006). Other examples of vessel regression during vascular remodeling include the elimination of capillaries in prechondrogenic regions to allow the differentiation of cartilage (Hallmann et al., 1987) and the regression of the hyaloid vasculature to allow the development of the vitreous body in the eye, which is necessary for proper vision (Latker and Kuwabara, 1981; Meeson et al., 1996). An interesting aspect of endothelial regression in the ocular vasculature is the active role of macrophages during vessel remodeling (Lang and Bishop, 1993). Macrophages secrete a Wnt ligand that acts as a short-range paracrine signal that directly induces apoptosis of endothelial cells during the first weeks following birth (Meeson et al., 1996: Lobov et al., 2005). In some cases of vascular regression, cell death is not involved. This type of regression has been referred to as “pruning”, by analogy to the process of trimming a tree (Risau, 1997). Pruning was first described in the embryonic retina and involves the removal of redundant channels (Ashton, 1966). Blood flow ceases in these excess

capillaries, the lumens collapse and the endothelial cells retract toward adjacent capillaries, leaving behind thin extensions called “intercapillary bridges”. These endothelial cells do not appear to die by apoptosis (Augustin et al., 1995), and it is possible that they reassemble into other vessels or dedifferentiate and contribute to alternative tissues (Risau, 1997).

VI.E.v.  Hemodynamic Forces A key factor regulating remodeling of the primary vasculature is the force associated with blood flow (Jones et al., 2006). Classic embryological studies of the chick yolk sac showed that vessels with heavy blood flow enlarged, while those with little flow regressed. Subsequently, Murray surmised that vessels adapted to flow in order to control the shear stress that they experience as a result of circulation (Murray, 1932). In support of this idea, surgical removal of the heart from the chick embryo which completely eliminated circulation was also shown to block remodeling of vessels (Chapman, 1918; Manner et al., 1995). Similarly, when Ncx1 is genetically-ablated in mouse, the heart forms but does not beat (Wakimoto et al., 2000). In these mutant embryos, the initial vascular plexus forms normally

506

PART  |  8  Making Vessels

(A) Naked ECs

(B) Vessel stabilization

(C) Vessel destabilization Macrophage

SMC recruitment

Ang-1, PGDF TGFβ, PIGF

Bleeding TGFβ

EC dysfunction

ECM

withdrawal of VEGF, Ang-1, PlGF, PDGF

Mature vessel SMCs

Vessel regression

Figure 9  Vessel stabilization versus vessel regression. One mechanism of angiogenic remodeling is reduction of vessel network complexity via vessel regession. (A) Primitive vessels initially consist of “naked” endothelium. (B) Vessel stabilization and maturation involve a coordination of both angiogenic factors and mural cell coverage. In addition, vessels require consistent hemodynamic flow for proper maintenance. (C) In contrast, a number of factors can cause the destabilization and subsequent regression of vessels. Insufficient angiogenic factors, flow or mural coverage can lead to fragile, leaky vessels that are prone to hemorrhage, apoptosis, or necrosis. Extracellular matrix (ECM), smooth muscle cells (SMCs), endothelial cell (EC). Image adapted from Conway et al. (2001).

(A)

(B)

Figure 10  Role of flow on vascular patterning and arteriovenous cell fate. Experimental manipulation of blood flow in avian embryos by Eichmann and colleagues. (A) Blood flow can be altered in chick embryos by ligation of the major yolk sac artery. This results in drop of flow in major arteries (red) on the ligated side. (B) Following ligation, blood flow is rerouted from the unligated side, and thus reversed, such that venous flow (blue) now occurs ectopically in arterial vessels. In response, arterial endothelial cells in these vessels turn off arterial markers and begin to express venous markers.

507

Chapter  |  8.2  Vascular Development

(A)

(B)

(C)

(D)

Figure 11  Endothelial anastomosis. Schematic of putative cellular rearrangements required for the fusion of separate vascular networks. (A) During vascular anastomosis, endothelial cells (ECs) of two different vessels recognize each other, degrade their basement membranes (BM), and establish new cell–cell junctions. (B) Once junctions have formed and stabilized, intracellular rearrangements occur, so that lumens from the different vascular networks can be connected. This might occur either inter- or intracellularly. (C) The bridging cells form openings to connect the lumens. (D) The resulting fusion of lumens creates a new vessel that allows blood flow between vessel networks.

in the yolk sac, but never remodels into large and small vessels. Similarly, a more recent study demonstrated that vascular remodeling fails to occur properly in mice lacking the function of MLC2a, as a direct result of impaired blood flow caused by defective atrial contraction (Lucitti et al., 2007). Embryological studies have also addressed the role of blood flow during arterial–venous differentiation. Eichmann and colleagues carried out ligation of the vitelline artery in the chick yolk sac to prevent perfusion of the vessel by blood (le Noble et al., 2004). They found that blood flow would circumnavigate the blockage and find an alternative vessel as a principal conduit (Fig. 10A). The newly-commandeered vessel then enlarged and functionally replaced the original blocked artery (Fig. 10B). When they assayed the expression of artery and vein markers, they found that arterial markers turned off downstream of the ligation, while venous markers turned on. Transcriptional changes in response to blood pressure have been observed for a number of endothelial genes, including PDGF, FGF, TGF and tissue factor (Resnick and Gimbrone, 1995). Together, these studies demonstrate that although there is a genetic basis for vessel fate, it is plastic and subject to remodeling by hemodynamic forces.

VI.E.vi.  Anastomosis Existing blood vessels can also alter their overall architecture by anastomosis. Anastomosis is defined as the joining or coalescing of different branching networks of tubes into a single continuous network. The cellular and molecular mechanisms of anastomosis are not well-characterized or understood, and few studies have addressed its role during development. It is clearly a complex process by which tubes must initially contact each other, then recognize and adhere to one another. Within the endothelial tubes existing cell–cell adhesions must be disrupted in order to form

a continuous new lumen (Fig. 11). This process implies that endothelial cells have an inherent mechanism for recognizing each other and facilitating interendothelial fusion. An excellent example of anastomosis during development is the formation of the dorsal aorta in mammals and birds, when two parallel dorsal aortas contact each other at the midline and fuse to form a single vessel. It is also important in a clinical setting, since after organ transplantation the vascular beds of grafted tissues must become integrated with the host vasculature through anastomosis (Greenwald and Berry, 2000).

VI.F.  Molecules Regulating Angiogenic Remodeling VI.F.i.  Vascular Endothelial Growth Factor Extensive analysis has shown that sprouting of new vessels and remodeling of the vasculature are both influenced by levels of VEGF. In the early quail embryo, VEGF is strongly expressed in the yolk sac, where it promotes the sprouting and reorganization of blood vessels (Flamme et al., 1995b). In mouse embryos, VEGF is expressed in primordia of organs such as the kidney, adrenal gland, pancreatic islets and the pituitary gland, which are not immediately juxtaposed to Flk-1-expressing endothelial cells (Dumont et al., 1995). It is likely that VEGF contributes to the establishment of blood vessels in these organ primordia, at least partly by stimulating branching from adjacent vascular structures. In addition, VEGF in these organs is likely to stimulate proliferation of early angioblasts, and may regulate the connection of vessels formed in situ to those growing in via sprouting. One tissue that is primarily vascularized by sprouting angiogenesis is the nervous system. Spatial and temporal expression of VEGF in neural tissues correlates tightly with the ingrowth of nascent

508

blood vessel sprouts into the ventricular neurectoderm layer (Breier et al., 1992; Millauer et al., 1993b; Breier and Risau, 1996).

VI.F.ii.  Notch and its Ligands The mammalian genome contains four Notch receptor genes (Notch1–4) and two of these, Notch1 and Notch4, are expressed in endothelial cells during early blood vessel formation. Knockout of Notch1 results in embryonic lethality at E9.5–10 with major defects in vascular structures (Krebs et al., 2000). Endothelial cells differentiate normally and the primary plexus is formed, but subsequent remodeling and branching of vascular structures, including the intersomitic vessels, is severely compromised. Notch4mutant animals show no obvious defects in vascular development, but embryos that are mutant for both Notch1 and Notch4 show more severe vascular defects than those observed in Notch1-single-mutants (Krebs et al., 2000). It can be demonstrated that Notch function in endothelial cells is indeed responsible for the defects in vascular development, since tissue-specific ablation of Notch1 in endothelial cells yields the same vascular phenotype (Lim­bourg et al., 2005). Five genes encoding Notch ligands are present in the mammalian genome; Jagged1 and 2 and the Delta-like sequences, Dll1, 3 and 4. Of these, Jagged1 and Dll4 are both highly-expressed in endothelial cells during early vascular development and are therefore in a position to mediate Notch signaling. Knockout of Jagged1 results in embryonic death around E11.5, significantly later than the Notch1-mutant mouse. These embryos again show major defects in remodeling of the original vascular network, although not as pronounced as in the Notch1-mutant embryos. During early embryogenesis, Dll4 is preferentially expressed in developing arterial vessels and is almost absent from developing veins (Villa et al., 2001). Ablation of Dll4 is embryonic-lethal even in the heterozygous condition (Duarte et al., 2004; Gale et al., 2004; Krebs et al., 2004). Embryos die at E10.5 with severe vascular defects in both the embryo and the yolk sac, consistent with an inability to remodel the original vascular plexus. As might be expected from the expression pattern, defects were pronounced in developing arteries, while venous structures (e.g., the posterior cardinal veins) appeared largely normal. Additional studies have provided clues to the mechanism by which Notch/Delta signaling regulates vessel growth. It has been noted that expression of both Notch1 and Dll4 is activated by VEGF in endothelial cells (Liu et al., 2003), particularly in regions where blood vessels are branching into previously avascular tissue, such as in the postnatal retina (Lobov et al., 2007). In these circumstances, it appears that VEGF induces expression of Dll4 in small vessels and veins where it is not normally expressed, and the resultant activation of Notch signaling

PART  |  8  Making Vessels

at these locations serves to regulate angiogenesis negatively, and thereby to limit the number of vascular branch sites. Consistent with this model, inhibition of Notch signaling in the retina leads to an excess of vascular sprouts and a subsequent overproliferation of blood vessels.

VII.  Vascular patterning One of the most fascinating features of the embryonic vasculature is the conservation of the stereotyped pattern of vessels from one embryo to the next (Nelsen, 1953). During early development blood vessels are found in the same relative position to each other and to surrounding embryonic tissues, implying that cues regulating the formation and patterning of the vasculature are genetically hard-wired and largely invariant. In addition, the overall pattern of organ-specific vascular beds, for instance in the skin, retina or skeletal muscle, is highly conserved. The coordination of endothelial migration and formation of cords at specific locations during blood vessel morphogenesis is called “vascular patterning”. Patterning results in a primary vascular network that is highly-reproducible within a species, both in space and time (Hogan and Bautch, 2004). To date, little is known about the cues that regulate where angioblasts emerge within the mesoderm and how these cues drive cell migration and assembly into blood vessels at specific locations. It is generally believed that vessels are patterned in response to extracellular signals produced by other embryonic tissues. Some cues repel endothelial cells while others attract, resulting in a balance of signals that sculpt the stereotypical architecture of the initial vascular plexus. This idea that the source of patterning signals resides in embryonic tissues, rather than in endothelial cells themselves, is supported by observations of transplantated endothelial cell behavior (Noden, 1989, 1990; Poole and Coffin, 1989). When tissues are transplanted into new host tissues, endothelial cells from the transplant sprout into host tissues and integrate into host vessels. Regardless of the source of the tissue comprising the transplant, and despite the invasive character of migratory angioblasts, they do not migrate randomly but are subject to patterning cues provided by the host embryo. For example, studies have shown that transplanted endothelial cells never cross the midline of the embryo and that the notochord is the source of signals that create this barrier (Wilting et al., 1995; Klessinger and Christ, 1996; Wilting and Christ, 1996) (see Section VII.A.i below).

VII.A.  Patterning during Vasculogenesis: Primary Plexus Formation One of the least-understood aspects of blood vessel development is the control of vessel location during vasculogenesis. Do most angioblasts arise at predetermined locations

509

Chapter  |  8.2  Vascular Development

and fuse without undergoing significant migration (vasculogenesis type I)? Or do they arise scattered throughout the mesoderm and then migrate in response to localized cues to form vessels at new locations? As the embryo grows and tissues expand in size, do angioblasts continue to arise de novo in the mesoderm, or are they primarily derived by proliferation of cells in existing vessels? Many of these questions remain unresolved in mammalian embryos; however, work in frog, fish and bird model systems has advanced our understanding of the morphogenesis and patterning of the vasculature. Below, we review examples of studies that address the patterning of specific vessels during vasculogenesis.

VII.A.i.  Patterning of the Aorta Structures at the embryonic midline play an active role in patterning the dorsal aortae (Fig. 12). In Xenopus, the diffusible form of VEGF is expressed at the embryonic midline and appears to stimulate chemotactic migration of angioblasts from the lateral plate to the midline, where a single dorsal aorta assembles (Cleaver and Krieg, 1998). When VEGF-expressing cells were transplanted to an ectopic location, aortic precursor cells migrated towards the new source of VEGF. Work in zebrafish also emphasizes the importance of midline structures for vascular patterning. Zebrafish mutant embryos that lack a notochord also lack an aorta, and when wild-type notochord is transplanted into mutants, aorta formation is rescued (Fouquet et al., 1997). Interestingly, the dorsal aorta forms in a strikingly different manner in higher vertebrates. Rather than the aggregation of a single midline tube observed in fish and frogs, the aorta in birds and mammals arises via the fusion of two parallel tubes that arose on each side of the notochord. Classical studies noted that the midline created a barrier impassable to endothelial cells (Klessinger and Christ, 1996). Elegant work in chick confirmed this observation, and showed that the notochord inhibits EC migration to the midline and prevents the paired aortae from fusing (Reese et al., 2004). These studies also demonstrated that notochord repulsion of ECs is due, at least in part, to the BMP antagonists Noggin and Chordin. In this case therefore, nonendothelial tissues express factors that negatively-regulate the migration and assembly of angio­ blasts as they form blood vessels.

VII.A.ii.  Patterning of the Perineural Vascular Plexus (PNVP) Other studies have characterized the role of the neural tube in directing the formation of the plexus that surrounds it – the perineural vascular plexus, or PNVP (Hogan et al., 2004). The PNVP is a capillary bed that initially surrounds the relatively avascular brain and spinal cord. During later development, angiogenic sprouts from the PNVP invade

and vascularize neural tissues. Using embryonic transplant experiments plus different knockout mouse strains, Bautch and colleagues demonstrated that the neural tube acts to pattern the PNVP by recruiting endothelial cells from surrounding mesoderm (Hogan et al., 2004). Moreover, it was shown that VEGF-A expressed by the neural tube mediates this recruitment. This is an example of a tissue expressing a positive cue that influences migration of angioblasts and vessel formation.

VII.B.  Patterning during Angiogenesis: Sculpting Growing Blood Vessels For many years, much of the research on angiogenic signaling has focused on factors regulating the angiogenic growth of vessels into tumors. For example, a huge number of studies have shown the importance of tumor-expressed VEGF for regulation of tumor angiogenesis (Ferrara, 1995; Carmeliet, 2005). Relatively less attention has been directed towards the patterning factors that are active during embryonic angiogenesis. Recent work, however, has investigated the role of many different factors during angiogenic growth and remodeling of the embryonic vasculature and we will focus on these factors in the following section, with particular attention to the growth of the intersomitic vessels.

VII.B.i.  Patterning of the Intersegmental Vessels (ISV): Sprouting Angiogenesis The intersegmental, or intersomitic, vessels (ISVs) grow dorsally from the aorta and posterior cardinal veins, threading their way between adjacent somites. This sprouting from the dorsal aorta is a highly-stereotyped process that begins at an early somite stage and proceeds in a rostrocaudal wave, as new somites develop. Studies of zebrafish mutants have significantly advanced our understanding of the mechanisms by which the ISVs form (Fig. 13). First, sprout formation depends on the presence of an aortic lumen, since mutants such as sonic-you that never undergo tubulogenesis also fail to form ISVs (Brown et al., 2000). Second, VEGF is an important regulatory factor, since morpholinos targeting VEGF abolish ISV formation (Nasevicius et al., 2000). Third, fusion of intracellular vacuoles underlies lumen formation within the ISVs and possibly acts as a driving force during sprout extension (Kamei et al., 2006). Recently, it has been discovered that a number of molecules regulating neural guidance are also involved in vascular patterning (Carmeliet and Tessier-Lavigne, 2005; Eichmann et al., 2005; Suchting et al., 2006). In general terms, these molecules help to guide the growth of axons and blood vessels along precisely-defined paths, resulting in proper connections with distant targets. Most of these factors regulate growth of the ISVs between the

510

PART  |  8  Making Vessels

(A) Frog/Zebrafish Vertebral artery

Neural tube Notochord Intersomitic artery Dorsal aorta

Perineural vascular plexus Somite

Cardinal vein Hypochord Lateral plate

Signals

Endoderm

shh VEGF Repulsive guidance signals

semaphorin and ephrin ligand

(B) Mouse/Chicken

noggin/chordin

Lateral plate Endoderm

Figure 12  Vascular patterning: Axial blood vessels. Overview of aortal development in lower (single primitive aorta) versus higher vertebrates (double primitive aorta). (A) In frog and fish, a single axial artery, the dorsal aorta, forms at the embryonic midline. Sources of cues, such as VEGF, identified as being important in patterning axial vessels, include the hypochord, notochord and somites. Somites are also a source of ephrin ligands and semaphorins, which act as repulsive cues that direct the guidance and patterning of intersomitic vessels. The neural tube is a source of slit and netrin ligands, which also repulse and guide intersomitic vessels. (B) In chick and mouse, two parallel aortae form during vasculogenesis. These vessels are later remodeled, resulting in the formation of a single midline dorsal aorta, as a result of fusion of the bilateral aortae by anastomosis. Initially, formation of the paired aortae is driven by localized patterning cues. The notochord has been shown to be a strong source of repulsive cues. It secretes noggin and chordin, which act to repel endothelial cells from the midline. In contrast, the neural tube and the endoderm are sources of VEGF, which attract endothelial cells and promote proliferation.

511

Chapter  |  8.2  Vascular Development

(A)

(B)

(C)

(D)

(E)

Figure 13  Vascular patterning: Intersomitic vessels. (A) Intersomitic/intersegmental vessels, such as arteries (ISA) and veins (ISV), initiate growth from the aorta and cardinal veins by angiogenic sprouting and extend between the somites. During their extension and growth, they are guided by somite-derived semaphorin and ephrins. (B) Once sprouts have extended past the somites, they fuse dorsal to the neural tube. (C) Normal intersomitic vessels in zebrafish embryos, as visualized in Fli-EGFP transgenic fish. (D) Intersomitic vessels in zebrafish lacking the semaphorin receptor plexinD1. In these mutants, vessels sprout and extend in random directions, ignoring the somite boundaries. (E) Later, parachordal (P) and vertebral (V) vessels extend from the ISVs and ISAs, and (F) fuse to general vessels running along the anterio–posterior axis. Netrin expression in the neural tube is believed to inhibit premature development of these vessels. Images reproduced from Torres-Vasquez et al. (2004).

somites (Helbling et al., 2000; Ambler et al., 2001; Childs et al., 2002; Gu et al., 2005; Suchting et al., 2005; Shaw et al., 2006; Jin et al., 2007). Families of molecules associated with both neural and vascular patterning include ephrins, semaphorins, slits and netrins. Below, we introduce each family of guidance molecules and briefly describe the roles they play during patterning of the embryonic vasculature.

VII.C.  Molecules Regulating Vascular Patterning VII.C.i.  Vascular Endothelial Growth Factor Perhaps the best-understood regulator of vascular patterning is VEGF. VEGF is expressed in embryonic tissues at most sites where blood vessels form, and is distributed in a gradient to direct sprouting of vessels (Lundkvist et al., 2007). VEGF is also known to direct chemotactic migration

of angioblasts and endothelial cells, as well as to stimulate their proliferation, resulting in proper vascular patterning of embryonic vessels (Cleaver and Krieg, 1998; Hogan et al., 2004). In developing skin, peripheral nerves have been shown to express VEGF and to drive patterning of the skin vasculature (Mukouyama et al., 2002, 2005). Specifically, nerve-derived VEGF directs remodeling and arteriogenesis of the primitive capillary plexus in skin via endothelial NRP1. Those vessels closest to nerves and to high levels of VEGF enlarge and undergo arterial differentiation. This indicates that endothelial cell fate, with respect to arterio–venous differentiation, is at least partially plastic and dependent on environmental signals. It also demonstrates the role of VEGF in both vascular patterning and directing endothelial cell fate.

VII.C.ii.  Semaphorins and Plexin Receptors The plexin proteins (about 10 family members in vertebrates) are transmembrane receptors for semaphorin

512

ligands (Bussolino et al., 2006). Extensive study of semaphorin–plexin signaling during neural development has demonstrated that activation of the signaling pathway generally leads to repulsive interactions between cells. One of the plexin proteins, PlexinD1, is expressed at high levels in endothelial cells during mouse vascular development, and knockout of the PlexinD1 gene results in heart defects and numerous disruptions to vascular patterning (Gitler et al., 2004). Similarly, the zebrafish out-of-bounds mutant, that exhibits defects in vascular patterning, results from mutations in the PlexinD1 gene (Torres-Vazquez et al., 2004). In both mouse and zebrafish PlexinD1-mutant embryos, specification and differentiation of endothelial cells occurs normally, but defects occur in the assembly of certain vessels (e.g., pharyngeal arch vessels) and the errant branching of intersomitic vessels into somites occurs. This last observation can be explained by the fact that Semaphorin 3E is highly-expressed in the somites and normally acts as a repulsive signal for branching vessels (Gu et al., 2005). Consistent with this model, knockout of Semaphorin 3E eliminated the repulsive activity and allowed vessels to extend ectopically into the somites.

VII.C.iii.  Ephrins and Eph Receptors The Ephs are transmembrane tyrosine kinase proteins that act as receptors for cell surface-tethered ephrin ligands. Ephrins have been implicated both as repellents and attractants for axons. In neural tissues, Eph/ephrin signaling acts to establish compartment boundaries and to restrict the migration pathways of neurons. Conditional knockout studies in mouse have shown that Eph signaling is also essential for vascular development in both endothelial cells and the smooth muscle mural cells that protect and stabilize blood vessels (Kuijper et al., 2007). EphrinB2 is an endothelial-expressed high affinity ligand for the EphB receptors. When ephrinB2 expression is ablated specifically in endothelial cells using a Tie2-driven Cre recombinase, mutant animals exhibit severe defects in vascular development primarily associated with angiogenic remodeling (Gerety and Anderson, 2002). More specifically, the yolk sac vasculature remains as a homogeneous vascular plexus, and fails to exhibit the enlargement and regression of vessels associated with normal development. Further defects are observed in development of the anterior cardinal vein, the vessels of the head and the intersomitic vessels. In some cases, intersomitic vessels make errant branches into the somite bundles. This is equivalent to the effects observed in previous studies using the frog embryo, which showed that mis-expression of ephrin ligand or a dominant negative EphB receptor resulted in misguided growth of vessels into the somites (Helbling et al., 2000).

VII.C.iv.  Slits and Robo Receptors During development of the central nervous system, cell surface receptors of the roundabout (robo) family interact

PART  |  8  Making Vessels

with ligands of the slit family to activate cell signaling pathways leading to cell–cell repulsion (Dickson and Gil­estro, 2006). One member of the robo family, robo4, is expressed specifically in vascular endothelial cells, and in vitro studies show that robo4 can bind to slit ligands, resulting in inhibition of endothelial cell migration (Park et al., 2003; Fujiwara et al., 2006). In the zebrafish embryo, both overexpression and knockdown methods demonstrate that robo4 is required for the normal regulation of sprouting of the intersegmental blood vessels (Bedell et al., 2005). Other studies suggest that robo4/slit signaling may activate Cdc42 and Rac1 Rho GTPases, which may play a role in regulating directionality of vessel sprouting (Kaur et al., 2006). In this latter example, robo/slit interactions are believed to mediate attractive signaling. Knockout studies have yet to confirm an essential function for robo/slit signaling during mouse vascular patterning, but this may be due to redundant function of multiple robo and slit family members.

VII.C.v.  Netrin and Unc/DCC Receptors Netrins are an evolutionarily-conserved family of molecules with homology to laminins (Baker et al., 2006). There are three secreted and two cell-tethered (GPI linked) netrins which act at long-range and short-range, respectively. Netrins are generally described as neuronal guidance cues which can act bifunctionally. Genetic studies and in vitro cultures have shown that netrins can either attract axons, when binding DCC or Neogenin receptors, or in contrast, repel axons via UNC5 receptors (Dickson, 2002; Dickson and Keleman, 2002). Together, netrins and Slits have been shown to cooperate in their guidance of commissural axons through the embryonic midline (Stein and Tessier-Lavigne, 2001; Dickson, 2002; Bhat et al., 2007). In addition to their role in neuronal pathfinding, netrins have also been shown to direct other developmental processes, including vascular patterning (Eichmann et al., 2005). Recent work in fish and mouse has shown the potential role for netrin signaling in both attracting and promoting vessel growth (Park et al., 2004; Navankasattusas et al., 2008), and in repulsion and retraction of vessels (Lu et al., 2004; Larrivee et al., 2007).

VIII.  Vessel maturation and vascular wall formation Early embryonic vessels are characterized by thick, plump endothelial cells which have tenuous adherence and incomplete basement membrane formation (Wagner, 1980). However, these characteristics rapidly change as blood flow increases and endothelial cells mature; anastomoses disappear, capillaries may split by intussusception and adherence between endothelial cells increases dramatically (Sabin, 1920; Clark, 1939; Wagner, 1980).

513

Chapter  |  8.2  Vascular Development

As vessels mature, a basement membrane forms and gradually thickens (Wolff and Bar, 1972). Large vessels also become shaped by the mechanical forces generated by circulation (Resnick and Gimbrone, 1995). Overall, final maturation of the vasculature requires interaction of endothelial cells with each other, with the extracellular matrix environment, and with adjacent mesenchymal support cells such as pericytes and smooth muscle cells. Major arteries and veins therefore acquire very different morphological and molecular characteristics, leading to significant differences in their final architecture (Fig. 14). Developmental remodeling of the vasculature results in a heterogeneous hierarchical array of blood vessels,

including large veins and arteries and smaller vessels such as venules, arterioles and capillaries. In addition, the endothelia that line these different blood vessels develop an array of different properties (Kumar et al., 1987; Aird, 2003). For example, the endothelium of large vessels has an important role in controlling vasoconstriction and vasodilation, and in the regulation of blood pressure. On the other hand, the endothelium of the small vessels mediates the exchange of gases and nutrients with the tissues (Risau and Flamme, 1995). Within small vessels, the capillary endothelium can be subdivided into three different phenotypes – continuous, discontinuous and fenestrated (Risau and Flamme, 1995) – and these morphological

(A)

(B)

Endothelium (tunica intima)

Endothelium (tunica intima)

Internal elastic tissue

Internal elastic tissue

Smooth muscle (tunica media)

Flow

Smooth muscle (tunica media)

Flow

External elastic tissue

External elastic tissue Vein

Fibrous connective tissue (tunica adventitia)

Artery

Fibrous connective tissue (tunica adventitia)

Lumen

Capillary bed Venule Endothelial cell

Arteriole

(C) Basal lamina

Pericyte

Smooth muscle cell Smooth muscle cell Basal lamina Pericyte Capillaries Figure 14  Schematic drawing of the mural components associated with major vessels of the circulatory system. (A) Vascular wall components of the veins, including the endothelial tunica intima, the smooth muscle tunica media and the extracellular matrix tunica adventitia. Note the wider diameter of veins and the presence of valves. (B) Vascular wall components of the arteries, including the same vascular wall layers as the veins. Arteries, however, have a much thicker tunica media and often a narrower lumen. (C) The capillary network consists largely of a more naked endothelial layer. Pericytes associate with the capillaries and cover only a small fraction of their surface area.

514

PART  |  8  Making Vessels

PDGF–B PDGF receptor

Figure 15  Pericyte-endothelial interface. Pericytes associate closely with endothelial cells of small vessels, including growing angiogenic sprouts. This association has been shown to depend on PDGF signaling by stalk endothelial cells to receptors located on mural cells. Tip cells of growing angiogenic sprouts are not covered with mural cells.

differences reflect different permeability requirements of these vessels in different tissues. While the molecular mechanism remains unknown, local signaling is believed to be responsible for the emergence of these different vascular phenotypes (Risau, 1991).

VIII.A.  The Endothelial Basement Membrane As the vascular endothelium matures, endothelial cells synthesize extracellular matrix proteins that form a basement membrane. This membrane forms a supportive sleeve around the basal surface of endothelial cells which functions to maintain cell polarity and regulate endothelial cell behaviors, such as proliferation, adhesion and differentiation (Grant et al., 1990). Basement membranes are composed primarily of fibronectin, laminin, entactin/nidogen, collagen and a heparin sulfate proteoglycan. During embryological blood vessel development, extracellular matrix deposition helps to establish the basic patterning of the primary vascular plexus and is an early indication of blood vessel maturation. In the adult, basement membranes provide stability to most blood vessels, which are modified only occasionally, such as in the event of injury, cyclical changes in the reproductive system of females, or in response to pathological conditions. Following the morphological changes associated with vascular remodeling, mesenchymal support cells are recruited to provide mechanical and physiological support

to the endothelium. Pericytes are recruited to the capillaries, and the vascular wall forms around larger vessels by the addition of smooth muscle cells and adventitial fibro­ blasts (Le Lievre and Le Douarin, 1975; Schwartz and Liaw, 1993). Pericytes and endothelial cells are the only cell types included in the mature capillaries and postcapillary venules (Orlidge and D’Amore, 1987). Pericytes are cells which exist in close association with the endothelium of the capillaries, but which cover only a fraction of their surface (Figs. 14 and 15). Their specific function is unclear, but they may modulate the behavior of endothelial cells, probably by regulating permeability, proliferation and integrity (de Oliveira, 1966; Rhodin, 1968; AntonelliOrlidge et al., 1989b). The absence of pericytes has been correlated with growth and proliferation of endothelial cells during neovascularization (de Oliveira, 1966) while in contrast, a higher density of pericytes is observed on quiescent capillaries (Tilton et al., 1985). In vitro experiments have shown that pericytes can inhibit capillary endothelial cell growth (Orlidge and D’Amore, 1987) and that this inhibition is mediated by TGF (Antonelli-Orlidge et al., 1989a).

VIII.B.  The Vascular Wall Large arteries and veins are surrounded by a thick protective sheath called the vascular wall (Fig. 16). Larger vessels recruit a vascular supportive cell called the smooth muscle cell (SMC), which makes up a significant portion

515

Chapter  |  8.2  Vascular Development

(Fig. 14A). Capillaries, which experience little blood pressure, are only sparsely covered by pericytes (Fig. 14C).

VIII.C.  Molecules Regulating Mural Cell Recruitment and Vessel Maturation VIII.C.i.  Angiopoietins and Tie Receptors

Figure 16  Vascular wall of arteries and veins. Image of the thick vascular wall associated with arteries and the thinner vascular wall associated with veins. Arteries and veins are often closely-associated with each other in tissues, as seen here. Image courtesy of Michelle Tallquist.

of the vascular wall and is essential for the physiological properties of these vessels. In contrast, smaller vessels recruit a different type of supportive cell called pericytes. Together, smooth muscle cells and pericytes are collectively referred to as “mural cells”. The type of cells giving rise to pericytes and smooth muscle cells, and the signaling mechanisms recruiting them to blood vessels, are currently largely unknown. Smooth muscle cells of the vascular wall initially express markers such as smooth muscle -actin (Gabbiani et al., 1981; Owens and Thompson, 1986) and later express additional differentiation genes such as SM22 and calponin (Duband et al., 1993). The primary function of the vascular wall is to maintain the integrity of the endothelium and the blood pressure present in large vessels (Fig. 14). Three main layers have been identified in the major blood vessels (Rhodin, 1980). The tunica intima is the innermost layer and is composed of the endothelium, which lines the lumen of the blood vessel, the basement membrane and the internal elastic tissue. Surrounding the tunica intima is the tunica media, which is quite thick and composed mostly of smooth muscle cells with some elastic tissue. The tunica media is already rather well-formed in the fetus and does not significantly change into adulthood. Lastly, the tunica adventitia surrounds the inner layers with connective tissue, elastic tissue and fibroblasts. The composition of the vascular wall is specific for different types of blood vessels. Arteries, which experience high blood pressure, are surrounded by a thick smooth muscle cell layer (Fig. 14B). Veins, which in contrast conduct blood under much lower pressure, have much less smooth muscle and elastic tissue in their vascular walls and can stretch significantly

Angiopoietins are specific ligands for the endothelial receptor tyrosine kinase Tie2, which is expressed almost exclusively in vascular endothelial cells. Binding of Angio­ poietin1 (Ang1) ligand results in Tie2 receptor dimerization, autophosphorylation and activation of downstream signaling pathways. Signaling by angiopoietin ligands and Tie receptors is not required for specification of endothelial cells or for assembly of the original vascular plexus, but is important for subsequent remodeling, integrity and maturation of blood vessels. For example, mice lacking Ang1 die between E9.5 and E12.5 due to vascular hemorrhage, at least partly because the embryonic blood vessels fail to recruit smooth muscle mural cells (Suri et al., 1996). In gene-targeting experiments, Tie2-null animals exhibit vascular defects similar to the Ang1 mutants, strongly suggesting that Ang1 and Tie2 constitute a nonredundant signaling pathway essential for vascular maturation (Dumont et al., 1994; Sato et al., 1995). The most likely model is that Ang1 signals to endothelial cells via Tie2 and they respond by recruiting cells from the surrounding mesenchyme to become vascular smooth muscle (Fig. 17). This smooth muscle recruitment enhances the stability and integrity of the new vessels. The role of Angiopoietin2 (Ang2), which is expressed most prominently in tissues actively undergoing vascular remodeling, is less clear. It is intriguing that although Ang2 binds to the Tie2 receptor, this binding fails to activate downstream signaling pathways (Maisonpierre et al., 1997). This observation has led to the suggestion that Ang2 may function as an antagonist of Ang1 signaling, by competing for binding to Tie2. Mice lacking Ang2 function die about two weeks after birth, primarily due to defects in lymphatic function however, these animals also exhibit defects in vascular patterning, especially in the retinal vasculature (Gale et al., 2002). These results suggest that the normal function of Ang1 signaling through Tie2 is to promote maturation and stability of newly-formed vessels. In contrast, Ang2 appears to function as a destabilizing factor. By interfering with the stabilizing effects of Ang1, Ang2 permits blood vessels to respond to proangiogenic factors such as VEGF, and to undergo growth and remodeling (Holash et al., 1999). Finally, the Tie2-related transmembrane protein, Tie1, is also essential for normal vascular development. Genetic ablation of Tie1 in the mouse causes breakdown of blood vessels with associated hemorrhaging, and embryos die from about E13 onwards (Puri et al., 1995; Sato et al., 1995). The mechanism of Tie1 action is uncertain, because neither Ang1 nor Ang2

516

PART  |  8  Making Vessels

The receptor PDGFR- is expressed in vascular smooth muscle cells surrounding arteries and arterioles, and in pericytes closely associated with capillary vessels. Knockout studies in mouse have shown an essential role for PDGF-B function in recruitment of pericytes to the walls of capillary vessels. Both PDGF-B- and PDGFR--mutant animals die from widespread hemorrhaging from capillary vessels, usually late during embryonic development (Leveen et al., 1994; Soriano, 1994). The formation of the smooth muscle layers of larger vessels is apparently unaffected in PDGF-B or PDGFR- animals, although reduced numbers of mural cells are observed in several organs (Lindahl et al., 1997; Hellstrom et al., 1999). The effect on the number of capillary pericytes is not equivalent for all organs, and the tissues with the most significant reductions are skin, retina, kidneys, brain and heart (Armulik et al., 2005). Since some mural cells still develop in these mutants, albeit in reduced numbers, it has been proposed that the normal function of PDGF-B signaling from endothelial cells is to activate proliferation and migration of mural cells as they are recruited to newly-formed capillaries. Additional mouse studies have shown that endothelial specific ablation of PDGF-B expression results in a much more modest vascular phenotype than the global knockout (Enge et al., 2002). At present it is not clear whether this is due to alternative sources of PDGF-B expression, or more likely to incomplete ablation of PDGF-B function in endothelial cells because of limited Cre-recombinase activity. Figure 17  Model for the recruitment of the cellular vascular wall components. (A) Angiopoietin-1 is secreted by mesenchymal/fibroblast/ mural cells and binds to the Tie-2 receptor located on endothelial cells. This receptor activation triggers the release of factors from the endothelium that cause a chemotactic attraction of mesenchymal cells. These factors may include PDGF, known to be involved in the recruitment of smooth muscle cells and pericytes. (B) Mesenchymal cells are stimulated to migrate to the vessel. (C) When these mesenchymal cells contact the endothelium, release of endothelial TGF- is activated and drives vessel maturation. (D) In the mature vessel, endothelial cells are surrounded by a basement membrane and a layer of smooth muscle cells. Adapted from Folkman and D’Amore (1996).

show detectable binding to Tie1 (Maisonpierre et al., 1997). It is possible, however, that Tie1 dimerizes with Tie2 and modulates the signaling pathways downstream of Tie2 activation (Marron et al., 2000).

VIII.C.ii.  Platelet Derived Growth   Factor (PDGF) Platelet derived growth factor-B (PDGF-B) is a ligand for the receptor tyrosine kinases PDGFR- and PDGFR- (Heldin and Westermark, 1999; Betsholtz, 2004). During embryonic development, PDGF-B is expressed by most vascular endothelial cells, including particularly high levels in the tip cells of capillary sprouts (Gerhardt et al., 2003).

VIII.C.iii.  Ephrins and Eph Receptors In addition to the endothelial-specific role in regulation of remodeling angiogenesis, signaling via Eph receptors and ephrin ligands is essential for correct recruitment of mural cells to blood vessels (Foo et al., 2006; Kuijper et al., 2007). Antibody studies demonstrate that ephrinB2 is expressed in the pericytes and smooth muscle cells surrounding both veins and arteries. The functional relevance of this expression is demonstrated by mural-cell-selective knockout of ephrinB2 using Cre-recombinase under control of the PDGFR- promoter, which is strongly expressed in mural cells. Animals lacking mural cell ephrinB2 exhibit extensive hemorrhaging from the smallest vessels of the skin, intestine and lungs. Although the number of pericyte cells in mutant animals is normal, the cells display aberrant morphology and fail to make tight contact with the capillary vessels. Mechanistic studies show that pericytes lacking ephrinB2 exhibit aberrant migrational behavior and also altered interactions with extracellular matrix (Foo et al., 2006).

VIII.C.iv.  Sphingosine 1-Phosphate and the Sphingosine 1-Phosphate (Edg) Receptors Sphingosine 1-phosphate, S1P, is a by-product of sphingo­ lipid metabolism and is involved in numerous cell signaling

517

Chapter  |  8.2  Vascular Development

processes (Spiegel and Milstien, 2002). Mammals possess five S1P receptor proteins, S1P1–5, which show a broad distribution of tissue expression, but receptor S1P1 (previously called Edg-1) is expressed at quite high levels in the endothelia of embryonic arteries and capillaries. Knockout of S1P1 reveals an essential vascular function for S1P signaling (Liu et al., 2000). Embryos die at around E13.5 from severe hemorrhaging, due to failure to form a properlyorganized smooth muscle layer surrounding some, but not all, of the major vessels. Capillaries also show a reduced number of pericytes. Therefore, although smooth muscle cells are recruited to the vessels, they fail to proliferate and organize correctly, resulting in failure of blood vessel integrity. Since S1P1 receptor is expressed in endothelial cells, the effects on vascular smooth muscle cells must be indirect, possibly involving the PDGF-B pathway.

VIII.C.v.  Transforming Growth Factor-Beta (TGF-) TGF-1 is expressed in both endothelial cells and smooth muscle cells during vascular development. Knockout of either TGF-1 or the receptor, ALK1, results in problems with differentiation of endothelial cells and severe disruption of vasculogenesis in the yolk sac, resulting in embryonic death (Dickson et al., 1995; Oshima et al., 1996a; Goumans et al., 1999; Oh et al., 2000). However, early vascular development in the embryo itself is largely normal. Other components of the TGF- signaling pathway function during later vascular development for angiogenic elaboration of the original vascular plexus and recruitment of mural cells (Lebrin et al., 2005). Knockouts of the broadlyexpressed TGF- family receptor ALK5, the co-receptor endoglin and the signal transduction component, Smad5, all lead to embryonic lethality due to vascular defects, including failure to remodel the yolk sac vasculature and hemorrhage because of faulty vessel maturation (Chang et al., 1999; Li et al., 1999; Yang et al., 1999; Larsson et al., 2001). At least in the case of Smad5 the effects appear to be indirect, since tissue-specific ablation of Smad5 function in endothelial or smooth muscle cells results in normal vascular development (Umans et al., 2007).

VIII.D.  Transcriptional Regulation of Endothelial Gene Expression Transcription factors are among the earliest markers of the endothelial lineage. Numerous studies have confirmed that transcription factors of the Ets family play an essential role in regulation of vascular genes (Dejana et al., 2007; Pham et al., 2007). Proteins of the forkhead (Fox), COUP-TF and GATA families are also important (Kappel et al., 2000; You et al., 2005; Seo et al., 2006). However, due at least partly to the fact that these factors are members of multigene

families, mouse knockout data has been somewhat difficult to interpret and it is clear that transcriptional regulation of vascular development requires additional study. Results from mouse multi-gene knockouts, coupled with morpholino knockdowns of multiple genes in zebrafish and Xenopus are likely to be particularly informative. Discussion of the role of different transcription factors during endothelial differentiation and development is beyond the scope of this chapter, but has been covered thoroughly in reviews (Oettgen, 2001; Dejana et al., 2007).

IX.  Endothelial heterogeneity   and plasticity IX.A.  Endothelial Cell Heterogeneity One of the difficulties in formulating a precise definition for endothelial cells is the significant molecular, cellular and functional differences between endothelial cells in different tissues (Aird, 2003). Molecularly, it is evident that few, if any, genes are ideal signatures for endothelium. Even accepted vascular markers, such as Flk-1 (VEGFR2), Tie2, PECAM (CD31) and VE-cadherin are not expressed completely uniformly throughout the endothelium, and they are also found in other cell types. For instance, Flk-1 transiently marks blood (Yamaguchi et al., 1993), PECAM is rather abundantly expressed in macrophages (Lee, 1991) and both markers are also found in lymphatic vessels (Enholm et al., 2001). In fact, most endothelial markers can also be found in a range of other tissues, including hematopoietic, lymphatic, myogenic and neural tissues. Similarly, one finds that ultrastructural cellular features associated with endothelium, such as Weibel–Palade bodies (organelles that act as storage compartments for factors to be secreted into the serum) or caveolae are either not present in all endothelial cells, or are not unique to the endothelial lineage. In addition, endothelial cells cannot simply be characterized by their overall morphology, since they take on a wide range of geometric shapes, from elongated in microvasculature of the placenta to polygonal in umbilical veins. Functional heterogeneity is also a striking characteristic of the endothelium, which displays a conspicuous division of labor, with regionally-specific roles carried out by either vessels of certain vascular beds, or certain sizes. For instance, the endothelium of capillaries carries out gas exchange between blood and resident tissue cells, while the endothelium of large vessels is load-bearing, carries large volumes of blood and must resist the forces of sheer stress and high pressure. Arterioles regulate endothelial-mediated vasorelaxation of the vascular wall, while venules are responsible for most leukocyte trafficking during immune responses (Minami and Aird, 2005). Increasingly, it is clear that “there are almost as many varieties of capillaries as there are organs and tissues” (Suter and Majno, 1965).

518

Unraveling the molecular basis of endothelial cell biology and heterogeneity will facilitate the development of targeted therapies for a wide array of diseases (Conway and Carmeliet, 2004).

IX.B.  Endothelial Cell Plasticity In addition to being highly-heterogeneous in nature, endothelial cells are also highly-plastic in their ability to adapt to new environments. This is demonstrated most strikingly when tissue is transplanted from one region to another (Noden, 1989, 1990). Regardless of the source or origin of the transplanted tissue, endothelial cells sprout from the transplant into the surrounding host tissue, and they become normally integrated into vessels in adjacent tissues in the host. This behavior supports the notion that regulatory signals responsible for patterning the vasculature reside within host embryonic tissues, and not in the transplanted endothelial cells themselves. Another example of endothelial plasticity is the ability of endothelial cells to readily switch arteriovenous fates (Fig. 18). This was demonstrated in elegant chimera experiments in which either artery or cardinal vein endothelial cells from quail embryos were transplanted into the coelom of chick embryos (Moyon et al., 2001). Results clearly show that, regardless of artery or vein origin, the endothelial cells are able to contribute to either arteries or veins, and switch their expression of molecular markers, demonstrating that

PART  |  8  Making Vessels

their fate is plastic and dependent on environmental context. Another interesting example of endothelial plasticity is the potential for endothelial cells to give rise to blood. Endothelial cells with this potential have been called “hemogenic endothelium” and have been identified in the floor of the dorsal aorta in most organisms examined (Jaffredo et al., 2005; Dieterlen-Lievre, 2007). This ability of endothelial cells to transdifferentiate into blood clearly demonstrates that not only can endothelial cells change their endothelial program, but they can also, under some circumstances, undergo more radical cell fate changes.

X.  Conclusion Cellular and molecular characterization of embryonic endothelial cells has significantly improved our understanding of blood vessel development and behavior. Together, classical observations and more recent molecular and genetic approaches have helped to elucidate many aspects of vascular specification, differentiation and remodeling. However, many questions remain to be answered. What signals initiate and regulate angio-blast cell fate? What are the local patterning cues that direct the formation of vessels at specific locations? Why are some embryonic tissues avascular? Are different populations of angioblasts destined to become different parts of the vascular system, or is angio-blast fate generally plastic, dependent solely on environmental cues? If so, what are these

Figure 18  Endothelial plasticity. Schematic of experimental chick-quail chimera endothelial cell transplantation experiments. (A) When quailderived aortic endothelial cells are transplanted to a chick host, the cells integrate into both chick arteries and veins. If the quail arterial cells integrate into the cardinal veins, they turn off artery markers and turn on venous markers, in response to the new environment. (B) Similarly, if quail-derived cardinal vein endothelial cells are transplanted, these cells can integrate into both arteries and veins, adapting to their new locations. Adapted from Jones et al. (2006).

Chapter  |  8.2  Vascular Development

environmental cues? How do endothelial cells reorganize their subcellular components to migrate, recognize and adhere to one another and form coherent vascular structures? What subtle orchestration of extracellular matrix, cell–cell adhesion and growth factor signals is necessary for blood vessel proliferation and maturation in vivo? What controls the remodeling of the embryonic vasculature to form the complex and hierarchical array of arteries, veins and capillaries found in the mature organism? Clearly, our knowledge of the mechanisms underlying vascular development is continually evolving. Therefore, further analysis of the cellular and molecular events that drive endothelial morphogenesis and vessel formation is required to further our understanding of these critical embryological processes. Perhaps most importantly, the lessons learned from the study of embryonic vascular development in model organisms are likely to be directly applicable to understanding the vascular growth and physiology associated with a wide range of human pathologies.

Acknowledgments We would like to thank Didier Stainier, Brant Weinstein, Christer Betsholtz and Michelle Tallquist for providing images; and Michelle Tallquist and Rolf Brekken for helpful comments on the manuscript. We would also like to apologize to all the authors of primary work not cited due to space constraints. This work was supported by NIH grants HL074184 to PAK and DK079862-01 to OC, and AHA grant 0755054Y to OC.

References Aird, W.C., 2003. Endothelial cell heterogeneity. Crit. Care. Med. 31, S221–S230. Amaya, E., Musci, T.J., Kirschner, M.W., 1991. Expression of a dominant negative mutant of the FGF receptor disrupts mesoderm formation in Xenopus embryos. Cell 66, 257–270. Ambler, C.A., Nowicki, J.L., Burke, A.C., Bautch, V.L., 2001. Assembly of trunk and limb blood vessels involves extensive migration and vasculogenesis of somite-derived angioblasts. Dev. Biol. 234, 352–364. Antonelli-Orlidge, A., Saunders, K.B., Smith, S.R., D’Amore, P.A., 1989a. An activated form of transforming growth factor beta is produced by cocultures of endothelial cells and pericytes. Proc. Natl. Acad. Sci. USA 86, 4544–4548. Antonelli-Orlidge, A., Smith, S.R., D’Amore, P.A., 1989b. Influence of pericytes on capillary endothelial cell growth. Am. Rev. Respir. Dis. 140, 1129–1131. Armulik, A., Abramsson, A., Betsholtz, C., 2005. Endothelial/pericyte interactions. Circ. Res. 97, 512–523. Arnaout, M.A., 2002. Integrin structure: new twists and turns in dynamic cell adhesion. Immunol. Rev. 186, 125–140. Ashton, N., 1966. Oxygen and the growth and development of retinal vessels. In vivo and in vitro studies. The XX Francis I. Proctor Lecture. Am. J. Ophthalmol. 62, 412–435. Augustin, H.G., Braun, K., Telemenakis, I., Modlich, U., Kuhn, W., 1995. Ovarian angiogenesis. Phenotypic characterization of endothelial cells

519

in a physiological model of blood vessel growth and regression. Am. J. Pathol. 147, 339–351. Ausprunk, D.H., Folkman, J., 1977. Migration and proliferation of endothelial cells in preformed and newly formed blood vessels during tumor angiogenesis. Microvasc. Res. 14, 53–65. Ausprunk, D.H., Falterman, K., Folkman, J., 1978. The sequence of events in the regression of corneal capillaries. Laboratory Iinvestigation. J. Tech. Meth. Pathol. 38, 284–294. Ausprunk, D., Dethlefsen, S.M., Higgins, E.R., 1991. Distribution of fibronectin, laminin and type IV collagen during development of blood vessels in the chick chorioallantoic membrane. In: Feinberg, R.N., Sherer, G.K., Auerbach, R. (Eds.), The Development of the Vascular System. Karger, Basel, Switzerland, pp. 93–108. Bader, B.L., Rayburn, H., Crowley, D., Hynes, R.O., 1998. Extensive vasculogenesis, angiogenesis, and organogenesis precede lethality in mice lacking all alpha v integrins. Cell 95, 507–519. Baffert, F., Thurston, G., Rochon-Duck, M., Le, T., Brekken, R., McDonald, D.M., 2004. Age-related changes in vascular endothelial growth factor dependency and angiopoietin-1-induced plasticity of adult blood vessels. Circ. Res. 94, 984–992. Baffert, F., Le, T., Sennino, B., Thurston, G., Kuo, C.J., Hu-Lowe, D., McDonald, D.M., 2006. Cellular changes in normal blood capillaries undergoing regression after inhibition of VEGF signaling. Am. J. Physiol. 290, H547–H559. Baker, K.A., Moore, S.W., Jarjour, A.A., Kennedy, T.E., 2006. When a diffusible axon guidance cue stops diffusing: roles for netrins in adhesion and morphogenesis. Curr. Opin. Neurobiol. 16, 529–534. Baldwin, H.S., 1996. Early embryonic vascular development. Cardiovasc. Res. 31 Spec No, E34–E45. Bar, T., 1980. The vascular system of the cerebral cortex. Adv. Anat. Embryol. Cell. Biol. 59, I–VI, 1–62. Bartel, H., Lametschwandtner, A., 2000. Regression of blood vessels in the ventral velum of Xenopus laevis Daudin during metamorphosis: light microscopic and transmission electron microscopic study. J. Anat. 197 (Pt 2), 157–166. Bayless, K.J., Davis, G.E., 2002. The Cdc42 and Rac1 GTPases are required for capillary lumen formation in three-dimensional extracellular matrices. J. Cell Sci. 115, 1123–1136. Beck L., Jr, D’Amore, P.A., 1997. Vascular development: cellular and molecular regulation. Faseb. J. 11, 365–373. Bedell, V.M., Yeo, S.Y., Park, K.W., Chung, J., Seth, P., Shivalingappa, V., Zhao, J., Obara, T., Sukhatme, V.P., Drummond, I.A., Li, D.Y., Ramchandran, R., 2005. Roundabout4 is essential for angiogenesis in vivo. Proc. Natl. Acad. Sci. USA 102, 6373–6378. Bertwistle, D., Walmsley, M.E., Read, E.M., Pizzey, J.A., Patient, R.K., 1996. GATA factors and the origins of adult and embryonic blood in Xenopus: responses to retinoic acid. Mech. Dev. 57, 199–214. Betsholtz, C., 2004. Insight into the physiological functions of PDGF through genetic studies in mice. Cytokine and Growth Factor Rev. 15, 215–228. Bhat, K.M., Gaziova, I., Krishnan, S., 2007. Regulation of axon guidance by slit and netrin signaling in the Drosophila ventral nerve cord. Genetics 176, 2235–2246. Bogers, A.J., Gittenberger-de Groot, A.C., Poelmann, R.E., Peault, B.M., Huysmans, H.A., 1989. Development of the origin of the coronary arteries, a matter of ingrowth or outgrowth? Anat. Embryol. 180, 437–441. Breier, G., Albrecht, U., Sterrer, S., Risau, W., 1992. Expression of vascular endothelial growth factor during embryonic angiogenesis and endothelial cell differentiation. Development 114, 521–532.

520

Breier, G., Risau, W., 1996. The role of vascular endothelial growth factor in blood vessel formation. Trends. Cell Biol. 6, 454–456. Breier, G., Breviario, F., Caveda, L., Berthier, R., Schnurch, H., Gotsch, U., Vestweber, D., Risau, W., Dejana, E., 1996. Molecular cloning and expression of murine vascular endothelial-cadherin in early stage development of cardiovascular system. Blood 87, 630–641. Brooks, P.C., Clark, R.A., Cheresh, D.A., 1994a. Requirement of vascular integrin alpha v beta 3 for angiogenesis. Science 264, 569–571. Brooks, P.C., Montgomery, A.M., Rosenfeld, M., Reisfeld, R.A., Hu, T., Klier, G., Cheresh, D.A., 1994b. Integrin alpha v beta 3 antagonists promote tumor regression by inducing apoptosis of angiogenic blood vessels. Cell 79, 1157–1164. Brown, L.A., Rodaway, A.R., Schilling, T.F., Jowett, T., Ingham, P.W., Patient, R.K., Sharrocks, A.D., 2000. Insights into early vasculogenesis revealed by expression of the ETS-domain transcription factor Fli-1 in wild-type and mutant zebrafish embryos. Mech. Dev. 90, 237–252. Burri, P.H., Tarek, M.R., 1990. A novel mechanism of capillary growth in the rat pulmonary microcirculation. Anat. Rec. 228, 35–45. Burri, P.H., Djonov, V., 2002. Intussusceptive angiogenesis – the alternative to capillary sprouting. Mol. Aspects Med. 23, S1–S27. Bussolino, F., Valdembri, D., Caccavari, F., Serini, G., 2006. Semaphoring vascular morphogenesis. Endothelium 13, 81–91. Byrd, N., Becker, S., Maye, P., Narasimhaiah, R., St-Jacques, B., Zhang, X., McMahon, J., McMahon, A., Grabel, L., 2002. Hedgehog is required for murine yolk sac angiogenesis. Development 129, 361–372. Byrd, N., Grabel, L., 2004. Hedgehog signaling in murine vasculogenesis and angiogenesis. Trends Cardiovasc. Med. 14, 308–313. Caduff, J.H., Fischer, L.C., Burri, P.H., 1986. Scanning electron microscope study of the developing microvasculature in the postnatal rat lung. Anat. Rec. 216, 154–164. Campochiaro, P.A., 2000. Retinal and choroidal neovascularization. J. Cell. Physiol. 184, 301–310. Carmeliet, P., 2005. VEGF as a key mediator of angiogenesis in cancer. Oncology 69 (Suppl 3), 4–10. Carmeliet, P., Ferreira, V., Breier, G., Pollefeyt, S., Kieckens, L., Gertsenstein, M., Fahrig, M., Vandenhoeck, A., Harpal, K., Eberhardt, C., Declercq, C., Pawling, J., Moons, L., Collen, D., Risau, W., Nagy, A., 1996. Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 380, 435–439. Carmeliet, P., Ng, Y.S., Nuyens, D., Theilmeier, G., Brusselmans, K., Cornelissen, I., Ehler, E., Kakkar, V.V., Stalmans, I., Mattot, V., Perriard, J.C., Dewerchin, M., Flameng, W., Nagy, A., Lupu, F., Moons, L., Collen, D., D’Amore, P.A., Shima, D.T., 1999. Impaired myocardial angiogenesis and ischemic cardiomyopathy in mice lacking the vascular endothelial growth factor isoforms VEGF164 and VEGF188. Nat. Med. 5, 495–502. Carmeliet, P., Tessier-Lavigne, M., 2005. Common mechanisms of nerve and blood vessel wiring. Nature 436, 193–200. Chang, H., Huylebroeck, D., Verschueren, K., Guo, Q., Matzuk, M.M., Zwijsen, A., 1999. Smad5 knockout mice die at mid-gestation due to multiple embryonic and extraembryonic defects. Development 126, 1631–1642. Chapman, W.B., 1918. The effect of the heart-beat upon the development of the vascular system in the chick. Am. J. Anat. 23, 175–203. Childs, S., Chen, J.N., Garrity, D.M., Fishman, M.C., 2002. Patterning of angiogenesis in the zebrafish embryo. Development 129, 973–982. Choi, K., Kennedy, M., Kazarov, A., Papadimitriou, J.C., Keller, G., 1998. A common precursor for hematopoietic and endothelial cells. Development 125, 725–732.

PART | 8

Making Vessels

Christ, B., Poelmann, R.E., Mentink, M.M., Gittenberger-de Groot, A.C., 1990. Vascular endothelial cells migrate centripetally within embryonic arteries. Anat. Embryol. 181, 333–339. Clark, E.R., Clark, E.L., 1939. Microscopic observations on the growth of blood capillaries in the living mammal. Am. J. Anat. 64, 251–301. Cleaver, O., Tonissen, K.F., Saha, M.S., Krieg, P.A., 1997. Neovascularization of the Xenopus embryo. Dev. Dyn. 210, 66–77. Cleaver, O., Krieg, P.A., 1998. VEGF mediates angio-blast migration during development of the dorsal aorta in Xenopus. Development 125, 3905–3914. Coffin, J.D., Poole, T.J., 1988. Embryonic vascular development: immunohistochemical identification of the origin and subsequent morphogenesis of the major vessel primordia in quail embryos. Development 102, 735–748. Coffin, J.D., Poole, T.J., 1991. Endothelial cell origin and migration in embryonic heart and cranial blood vessel development. Anat. Rec. 231, 383–395. Conway, E.M., Carmeliet, P., 2004. The diversity of endothelial cells: a challenge for therapeutic angiogenesis. Genome Biol. 5, 207. Crosby, C.V., Fleming, P.A., Argraves, W.S., Corada, M., Zanetta, L., Dejana, E., Drake, C.J., 2005. VE-cadherin is not required for the formation of nascent blood vessels but acts to prevent their disassembly. Blood 105, 2771–2776. Cumano, A., Dieterlen-Lievre, F., Godin, I., 2000. The splanchnopleura/ AGM region is the prime site for the generation of multipotent hemopoietic precursors, in the mouse embryo. Vaccine 18, 1621–1623. Cumano, A., Godin, I., 2007. Ontogeny of the hematopoietic system. Ann. Rev. Immunol. 25, 745–785. Czirok, A., Zamir, E.A., Szabo, A., Little, C.D., 2008. Multicellular sprouting during vasculogenesis. Curr. Top. Dev. Biol. 81, 269–289. Davis, G.E., Bayless, K.J., Mavila, A., 2002. Molecular basis of endothelial cell morphogenesis in three-dimensional extracellular matrices. Anat. Rec. 268, 252–275. Davis, G.E., Senger, D.R., 2005. Endothelial extracellular matrix: biosynthesis, remodeling, and functions during vascular morphogenesis and neovessel stabilization. Circ. Res. 97, 1093–1107. de Jong, J.L., Zon, L.I., 2005. Use of the zebrafish system to study primitive and definitive hematopoiesis. Ann. Rev. Genet. 39, 481–501. de Oliveira, F., 1966. Pericytes in diabetic retinopathy. Br. J. Ophthalmol. 50, 134–143. de Vries, C., Escobedo, J.A., Ueno, H., Houck, K., Ferrara, N., Williams, L.T., 1992. The fms-like tyrosine kinase, a receptor for vascular endothelial growth factor. Science 255, 989–991. Dejana, E., 1996. Endothelial adherens junctions: implications in the control of vascular permeability and angiogenesis. J. Clin. Invest. 98, 1949–1953. Dejana, E., Taddei, A., Randi, A.M., 2007. Foxs and Ets in the transcriptional regulation of endothelial cell differentiation and angiogenesis. Biochimica. et. Biophysica. Acta. 1775, 298–312. deMello, D.E., Sawyer, D., Galvin, N., Reid, L.M., 1997. Early fetal development of lung vasculature. Am. J. Resp. Cell. Mol. Biol. 16, 568–581. Dickson, B.J., 2002. Molecular mechanisms of axon guidance. Science 298, 1959–1964. Dickson, B.J., Keleman, K., 2002. Netrins. Curr. Biol. 12, R154–R155. Dickson, B.J., Gilestro, G.F., 2006. Regulation of commissural axon pathfinding by slit and its Robo receptors. Annu. Rev. Cell Dev. Biol. 22, 651–675. Dickson, M.C., Martin, J.S., Cousins, F.M., Kulkarni, A.B., Karlsson, S., Akhurst, R.J., 1995. Defective haematopoiesis and vasculogenesis in transforming growth factor-beta 1 knock out mice. Development 121, 1845–1854.

Chapter | 8.2 Vascular Development

Dieterlen-Lievre, F., 1997. Intraembryonic hematopoietic stem cells. Hematol./Oncol. Clin. North Am. 11, 1149–1171. Dieterlen-Lievre, F., 2007. Emergence of haematopoietic stem cells during development. Comptes. Rendus. Biologies 330, 504–509. Doetschman, T.C., Eistetter, H., Katz, M., Schmidt, W., Kemler, R., 1985. The in vitro development of blastocyst-derived embryonic stem cell lines: formation of visceral yolk sac, blood islands and myocardium. J. Embryol. Exper. Morphol. 87, 27–45. Drake, C.J., 2003. Embryonic and adult vasculogenesis. Birth Defects Res. C Embryo Today 69, 73–82. Drake, C.J., Davis, L.A., Walters, L., Little, C.D., 1990. Avian vasculogenesis and the distribution of collagens I, IV, laminin, and fibronectin in the heart primordia. J. Exp. Zool. 255, 309–322. Drake, C.J., Davis, L.A., Little, C.D., 1992. Antibodies to beta 1-integrins cause alterations of aortic vasculogenesis, in vivo. Dev. Dyn. 193, 83–91. Drake, C.J., Fleming, P.A., 2000. Vasculogenesis in the day 6.5 to 9.5 mouse embryo. Blood 95, 1671–1679. Duan, L.J., Nagy, A., Fong, G.H., 2003. Gastrulation and angiogenesis, not endothelial specification, is sensitive to partial deficiency in vascular endothelial growth factor-a in mice. Biol. Repro. 69, 1852–1858. Duarte, A., Hirashima, M., Benedito, R., Trindade, A., Diniz, P., Bekman, E., Costa, L., Henrique, D., Rossant, J., 2004. Dosage-sensitive requirement for mouse Dll4 in artery development. Genes Dev. 18, 2474–2478. Duband, J.L., Gimona, M., Scatena, M., Sartore, S., Small, J.V., 1993. Calponin and SM 22 as differentiation markers of smooth muscle: spatiotemporal distribution during avian embryonic development. Differentiation 55, 1–11. Dumont, D.J., Gradwohl, G., Fong, G.H., Puri, M.C., Gertsenstein, M., Auerbach, A., Breitman, M.L., 1994. Dominant-negative and targeted null mutations in the endothelial receptor tyrosine kinase, tek, reveal a critical role in vasculogenesis of the embryo. Genes Dev. 8, 1897–1909. Dumont, D.J., Fong, G.H., Puri, M.C., Gradwohl, G., Alitalo, K., Breitman, M.L., 1995. Vascularization of the mouse embryo: a study of flk-1, tek, tie, and vascular endothelial growth factor expression during development. Dev. Dyn. 203, 80–92. Dyer, M.A., Farrington, S.M., Mohn, D., Munday, J.R., Baron, M.H., 2001. Indian hedgehog activates hematopoiesis and vasculogenesis and can respecify prospective neurectodermal cell fate in the mouse embryo. Development 128, 1717–1730. Eichmann, A., Marcelle, C., Breant, C., Le Douarin, N.M., 1993. Two molecules related to the VEGF receptor are expressed in early endothelial cells during avian embryonic development. Mech. Dev. 42, 33–48. Eichmann, A., Corbel, C., Nataf, V., Vaigot, P., Breant, C., Le Douarin, N.M., 1997. Ligand-dependent development of the endothelial and hemopoietic lineages from embryonic mesodermal cells expressing vascular endothelial growth factor receptor 2. Proc. Natl. Acad. Sci. USA 94, 5141–5146. Eichmann, A., Makinen, T., Alitalo, K., 2005. Neural guidance molecules regulate vascular remodeling and vessel navigation. Genes Dev. 19, 1013–1021. el-Hifnawi, E., el-Hifnawi, A., Frankenberg, C., Keeler, C., 1994. Ultrastructure and regression of the tunica vasculosa lentis in newborn Wistar rats. Ann. Anat. 176, 143–149. Enge, M., Bjarnegard, M., Gerhardt, H., Gustafsson, E., Kalen, M., Asker, N., Hammes, H.P., Shani, M., Fassler, R., Betsholtz, C., 2002. Endothelium-specific platelet-derived growth factor-B ablation mimics diabetic retinopathy. Embo. J. 21, 4307–4316.

521

Enholm, B., Karpanen, T., Jeltsch, M., Kubo, H., Stenback, F., Prevo, R., Jackson, D.G., Yla-Herttuala, S., Alitalo, K., 2001. Adenoviral expression of vascular endothelial growth factor-C induces lymphangiogenesis in the skin. Circ. Res. 88, 623–629. Feinberg, R.N., Latker, C.H., Beebe, D.C., 1986. Localized vascular regression during limb morphogenesis in the chicken embryo. I. Spatial and temporal changes in the vascular pattern. Anat. Rec. 214, 405–409. Ferkowicz, M.J., Yoder, M.C., 2005. Blood island formation: longstanding observations and modern interpretations. Exp. Hematol. 33, 1041–1047. Ferrara, N., 1995. The role of vascular endothelial growth factor in pathological angiogenesis. Breast Cancer Res. Treat. 36, 127–137. Ferrara, N., Carver-Moore, K., Chen, H., Dowd, M., Lu, L., O’Shea, K.S., Powell-Braxton, L., Hillan, K.J., Moore, M.W., 1996. Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 380, 439–442. Ferrara, N., Gerber, H.P., LeCouter, J., 2003. The biology of VEGF and its receptors. Nat. Med. 9, 669–676. Fitch, M.J., Campagnolo, L., Kuhnert, F., Stuhlmann, H., 2004. Egfl7, a novel epidermal growth factor-domain gene expressed in endothelial cells. Dev. Dyn. 230, 316–324. Flamme, I., Risau, W., 1992. Induction of vasculogenesis and hematopoiesis in vitro. Development 116, 435–439. Flamme, I., Breier, G., Risau, W., 1995a. Vascular endothelial growth factor (VEGF) and VEGF receptor 2 (flk-1) are expressed during vasculogenesis and vascular differentiation in the quail embryo. Dev. Biol. 169, 699–712. Flamme, I., von Reutern, M., Drexler, H.C., Syed-Ali, S., Risau, W., 1995b. Overexpression of vascular endothelial growth factor in the avian embryo induces hypervascularization and increased vascular permeability without alterations of embryonic pattern formation. Dev. Biol. 171, 399–414. Folkman, J., 1971. Tumor angiogenesis: therapeutic implications. New Eng. J. Med. 285, 1182–1186. Folkman, J., 1995. Angiogenesis inhibitors generated by tumors. Mol. Med. 1(2), 120–122. Folkman, J., Haudenschild, C., 1980. Angiogenesis in vitro. Nature 288, 551–556. Folkman, J., Klagsbrun, M., 1987. Angiogenic factors. Science 235, 442–447. Folkman, J., Shing, Y., 1992. Angiogenesis. J. Biol. Chem. 267, 10931–10934. Fong, G.H., Rossant, J., Gertsenstein, M., Breitman, M.L., 1995. Role of the Flt-1 receptor tyrosine kinase in regulating the assembly of vascular endothelium. Nature 376, 66–70. Fong, G.H., Zhang, L., Bryce, D.M., Peng, J., 1999. Increased hemangioblast commitment, not vascular disorganization, is the primary defect in flt-1 knock-out mice. Development 126, 3015–3025. Foo, S.S., Turner, C.J., Adams, S., Compagni, A., Aubyn, D., Kogata, N., Lindblom, P., Shani, M., Zicha, D., Adams, R.H., 2006. Ephrin-B2 controls cell motility and adhesion during blood-vessel-wall assembly. Cell 124, 161–173. Fouquet, B., Weinstein, B.M., Serluca, F.C., Fishman, M.C., 1997. Vessel patterning in the embryo of the zebrafish: guidance by notochord. Dev. Biol. 183, 37–48. Fujiwara, M., Ghazizadeh, M., Kawanami, O., 2006. Potential role of the Slit/Robo signal pathway in angiogenesis. Vasc. Med. (London) 11, 115–121. Gabbiani, G., Schmid, E., Winter, S., Chaponnier, C., de Ckhastonay, C., Vandekerckhove, J., Weber, K., Franke, W.W., 1981. Vascular smooth

522

muscle cells differ from other smooth muscle cells: predominance of vimentin filaments and a specific alpha-type actin. Proc. Natl. Acad. Sci. USA 78, 298–302. Gale, N.W., Thurston, G., Davis, S., Wiegand, S.J., Holash, J., Rudge, J.S., Yancopoulos, G.D., 2002. Complementary and coordinated roles of the VEGFs and angiopoietins during normal and pathologic vascular formation. Cold Spring Harb. Symp. Quant. Biol. 67, 267–273. Gale, N.W., Dominguez, M.G., Noguera, I., Pan, L., Hughes, V., Valenzuela, D.M., Murphy, A.J., Adams, N.C., Lin, H.C., Holash, J., Thurston, G., Yancopoulos, G.D., 2004. Haploinsufficiency of deltalike 4 ligand results in embryonic lethality due to major defects in arterial and vascular development. Proc. Natl. Acad. Sci. USA 101, 15949–15954. Gebb, S.A., Shannon, J.M., 2000. Tissue interactions mediate early events in pulmonary vasculogenesis. Dev. Dyn. 217, 159–169. George, E.L., Georges-Labouesse, E.N., Patel-King, R.S., Rayburn, H., Hynes, R.O., 1993. Defects in mesoderm, neural tube and vascular development in mouse embryos lacking fibronectin. Development 119, 1079–1091. George, E.L., Baldwin, H.S., Hynes, R.O., 1997. Fibronectins are essential for heart and blood vessel morphogenesis but are dispensable for initial specification of precursor cells. Blood 90, 3073–3081. Gerety, S., Anderson, D.J., 2002. Cardiovascular ephrinB2 function is essential for embryonic angiogenesis. Development 129, 1397–1410. Gerhardt, H., Golding, M., Fruttiger, M., Ruhrberg, C., Lundkvist, A., Abramsson, A., Jeltsch, M., Mitchell, C., Alitalo, K., Shima, D., Betsholtz, C., 2003. VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia. J. Cell Biol. 161, 1163–1177. Gerhardt, H., Betsholtz, C., 2005. How do endothelial cells orientate? Exs. 94, 3–15. Gibson, W.T., 1910. The development of the hypochord in Raia batis; with a note upon the occurence of the epibranchial groove in amniote embryos. Anat. Anz. 35, 407–428. Gitler, A.D., Lu, M.M., Epstein, J.A., 2004. PlexinD1 and semaphorin signaling are required in endothelial cells for cardiovascular development. Dev. Cell 7, 107–116. Gory-Faure, S., Prandini, M.H., Pointu, H., Roullot, V., Pignot-Paintrand, I., Vernet, M., Huber, P., 1999. Role of vascular endothelial-cadherin in vascular morphogenesis. Development 126, 2093–2102. Goss, C.M., 1928. Experimental removal of the blood island of Amblystoma punctatum embryos. J. Exp. Zool. 52, 45–61. Goumans, M.J., Zwijsen, A., van Rooijen, M.A., Huylebroeck, D., Roelen, B.A., Mummery,, C.L, 1999. Transforming growth factorbeta signalling in extraembryonic mesoderm is required for yolk sac vasculogenesis in mice. Development 126, 3473–3483. Grant, D.S., Kleinman, H.K., Martin, G.R., 1990. The role of basement membranes in vascular development. Ann. N. Y. Acad. Sci. 588, 61–72. Greenwald, S.E., Berry, C.L., 2000. Improving vascular grafts: the importance of mechanical and haemodynamic properties. J. Pathol. 190, 292–299. Gu, C., Rodriguez, E.R., Reimert, D.V., Shu, T., Fritzsch, B., Richards, L.J., Kolodkin, A.L., Ginty, D.D., 2003. Neuropilin-1 conveys semaphorin and VEGF signaling during neural and cardiovascular development. Dev. Cell 5, 45–57. Gu, C., Yoshida, Y., Livet, J., Reimert, D.V., Mann, F., Merte, J., Henderson, C.E., Jessell, T.M., Kolodkin, A.L., Ginty, D.D., 2005. Semaphorin 3E and plexin-D1 control vascular pattern independently of neuropilins. Science 307, 265–268.

PART | 8

Making Vessels

Haar, J.L., Ackerman, G.A., 1971. A phase and electron microscopic study of vasculogenesis and erythropoiesis in the yolk sac of the mouse. Anat. Rec. 170, 199–223. Habeck, H., Odenthal, J., Walderich, B., Maischein, H., Schulte-Merker, S., 2002. Analysis of a zebrafish VEGF receptor mutant reveals specific disruption of angiogenesis. Curr. Biol. 12, 1405–1412. Hallmann, R., Feinberg, R.N., Latker, C.H., Sasse, J., Risau, W., 1987. Regression of blood vessels precedes cartilage differentiation during chick limb development. Differentiation 34, 98–105. Hanahan, D., Folkman, J., 1996. Patterns and emerging mechanisms of the angiogenic switch during tumorigenesis. Cell 86, 353–364. Helbling, P.M., Saulnier, D.M., Brandli, A.W., 2000. The receptor tyrosine kinase EphB4 and ephrin-B ligands restrict angiogenic growth of embryonic veins in Xenopus laevis. Development 127, 269–278. Heldin, C.H., Westermark, B., 1999. Mechanism of action and in vivo role of platelet-derived growth factor. Physiol. Rev. 79, 1283–1316. Hellstrom, M., Kalen, M., Lindahl, P., Abramsson, A., Betsholtz, C., 1999. Role of PDGF-B and PDGFR-beta in recruitment of vascular smooth muscle cells and pericytes during embryonic blood vessel formation in the mouse. Development 126, 3047–3055. Hellstrom, M., Phng, L.K., Hofmann, J.J., Wallgard, E., Coultas, L., Lindblom, P., Alva, J., Nilsson, A.K., Karlsson, L., Gaiano, N., Yoon, K., Rossant, J., Iruela-Arispe, M.L., Kalén, M., Gerhardt, H., Betsholtz, C., 2007. Dll4 signalling through Notch1 regulates formation of tip cells during angiogenesis. Nature 445, 776–780. Hirakow, R., Hiruma, T., 1981. Scanning electron microscopic study on the development of primitive blood vessels in chick embryos at the early somite-stage. Anat. Embryol. 163, 299–306. Hirakow, R., Hiruma, T., 1983. TEM-studies on development and canalization of the dorsal aorta in the chick embryo. Anat. Embryol. 166, 307–315. His, W., 1868. Untersuchungen uber die erste Anlage des Wirbeltierleibes. Vogel, Leipzig, Germany. His, W., 1900. Lecithoblast und Angiobalst der Wirbeltiere. Abh. MathPhys. Kl Saechs. Ges. 26, 171–328. Hoang, M.V., Whelan, M.C., Senger, D.R., 2004. Rho activity critically and selectively regulates endothelial cell organization during angiogenesis. Proc. Natl. Acad. Sci. USA 101, 1874–1879. Hodivala-Dilke, K.M., McHugh, K.P., Tsakiris, D.A., Rayburn, H., Crowley, D., Ullman-Cullere, M., Ross, F.P., Coller, B.S., Teitelbaum, S., Hynes, R.O., 1999. Beta3-integrin-deficient mice are a model for Glanzmann thrombasthenia showing placental defects and reduced survival. J. Clin. Invest. 103, 229–238. Hogan, K.A., Bautch, V.L., 2004. Blood vessel patterning at the embryonic midline. Curr. Top. Dev. Biol. 62, 55–85. Hogan, K.A., Ambler, C.A., Chapman, D.L., Bautch, V.L., 2004. The neural tube patterns vessels developmentally using the VEGF signaling pathway. Development 131, 1503–1513. Holash, J., Maisonpierre, P.C., Compton, D., Boland, P., Alexander, C.R., Zagzag, D., Yancopoulos, G.D., Wiegand, S.J., 1999. Vessel cooption, regression, and growth in tumors mediated by angiopoietins and VEGF. Science 284, 1994–1998. Houser, J.W., Ackerman, G.A., Knouff, R.A., 1961. Vasculogenesis and erythropoiesis in the living yolk sac of the chick embryo. A phase microscopic study. Anat. Rec. 140, 29–43. Huang X., Griffiths, M., Wu, J., Farese, R.V., Jr, Sheppard, D., 2000. Normal development, wound healing, and adenovirus susceptibility in beta5-deficient mice. Mol. Cell Biol. 20, 755–759.

Chapter | 8.2 Vascular Development

Huber, T.L., Kouskoff, V., Fehling, H.J., Palis, J., Keller, G., 2004. Haemangioblast commitment is initiated in the primitive streak of the mouse embryo. Nature 432, 625–630. Hynes, R.O., 1990. Fibronectins. Springer-Verlag, New York, NY. Hynes, R.O., 2007. Cell-matrix adhesion in vascular development. J. Thromb. Haemost. 5 (Suppl 1), 32–40. Im, E., Kazlauskas, A., 2006. New insights regarding vessel regression. Cell Cycle (Georgetown, TX) 5, 2057–2059. Inai, T., Mancuso, M., Hashizume, H., Baffert, F., Haskell, A., Baluk, P., Hu-Lowe, D.D., Shalinsky, D.R., Thurston, G., Yancopoulos, G.D., McDonald, D.M., 2004. Inhibition of vascular endothelial growth factor (VEGF) signaling in cancer causes loss of endothelial fenestrations, regression of tumor vessels, and appearance of basement membrane ghosts. Am. J. Pathol. 165, 35–52. Iruela-Arispe, M.L., Hasselaar, P., Sage, H., 1991. Differential expression of extracellular proteins is correlated with angiogenesis in vitro. Laboratory investigation; J. Tech. Meth. Pathol. 64, 174–186. Isaacs, H.V., Tannahill, D., Slack, J.M., 1992. Expression of a novel FGF in the Xenopus embryo. A new candidate inducing factor for mesoderm formation and anteroposterior specification. Development 114, 711–720. Jaffredo, T., Nottingham, W., Liddiard, K., Bollerot, K., Pouget, C., de Bruijn, M., 2005. From hemangioblast to hematopoietic stem cell: an endothelial connection? Exp. Hematol. 33, 1029–1040. Jin, S.W., Beis, D., Mitchell, T., Chen, J.N., Stainier, D.Y., 2005. Cellular and molecular analyses of vascular tube and lumen formation in zebrafish. Development 132, 5199–5209. Jin, S.W., Herzog, W., Santoro, M.M., Mitchell, T.S., Frantsve, J., Jungblut, B., Beis, D., Scott, I.C., D’Amico, L.A., Ober, E.A., Verkade, H., Field, H.A., Chi, N.C., Wehman, A.M., Baier, H., Stainier, D.Y., 2007. A transgene-assisted genetic screen identifies essential regulators of vascular development in vertebrate embryos. Dev. Biol. 307, 29–42. Jones, E.A., le Noble, F., Eichmann, A., 2006. What determines blood vessel structure? Genetic prespecification vs. hemodynamics. Physiology (Bethesda, MD) 21, 388–395. Kadokawa, Y., Suemori, H., Nakatsuji, N., 1990. Cell lineage analyses of epithelia and blood vessels in chimeric mouse embryos by use of an embryonic stem cell line expressing the beta-galactosidase gene. Cell. Differ. Dev. 29, 187–194. Kamei, M., Saunders, W.B., Bayless, K.J., Dye, L., Davis, G.E., Weinstein, B.M., 2006. Endothelial tubes assemble from intracellular vacuoles in vivo. Nature 442, 453–456. Kappel, A., Schlaeger, T.M., Flamme, I., Orkin, S.H., Risau, W., Breier, G., 2000. Role of SCL/Tal-1, GATA, and ets transcription factor binding sites for the regulation of flk-1 expression during murine vascular development. Blood 96, 3078–3085. Kardon, G., Campbell, J.K., Tabin, C.J., 2002. Local extrinsic signals determine muscle and endothelial cell fate and patterning in the vertebrate limb. Dev. Cell 3, 533–545. Kau, C.L., Turpen, J.B., 1983. Dual contribution of embryonic ventral blood island and dorsal lateral plate mesoderm during ontogeny of hemopoietic cells in Xenopus laevis. J. Immunol. 131, 2262–2266. Kaur, S., Castellone, M.D., Bedell, V.M., Konar, M., Gutkind, J.S., Ramchandran, R., 2006. Robo4 signaling in endothelial cells implies attraction guidance mechanisms. J. Biol. Chem. 281, 11347–11356. Kawasaki, T., Kitsukawa, T., Bekku, Y., Matsuda, Y., Sanbo, M., Yagi, T., Fujisawa, H., 1999. A requirement for neuropilin-1 in embryonic vessel formation. Development 126, 4895–4902.

523

Keller, G., Kennedy, M., Papayannopoulou, T., Wiles, M.V., 1993. Hematopoietic commitment during embryonic stem cell differentiation in culture. Mol. Cell Biol. 13, 473–486. Kennedy, M., Firpo, M., Choi, K., Wall, C., Robertson, S., Kabrun, N., Keller, G., 1997. A common precursor for primitive erythropoiesis and definitive haematopoiesis. Nature 386, 488–493. Kim, S., Bakre, M., Yin, H., Varner, J.A., 2002. Inhibition of endothelial cell survival and angiogenesis by protein kinase. A. J. Clin. Invest. 110, 933–941. Kinder, S.J., Tsang, T.E., Quinlan, G.A., Hadjantonakis, A.K., Nagy, A., Tam, P.P., 1999. The orderly allocation of mesodermal cells to the extraembryonic structures and the anteroposterior axis during gastrulation of the mouse embryo. Development 126, 4691–4701. Klagsbrun, M., D’Amore, P.A., 1991. Regulators of angiogenesis. Annu. Rev. Physiol. 53, 217–239. Klessinger, S., Christ, B., 1996. Axial structures control laterality in the distribution pattern of endothelial cells. Anat. Embryol. 193, 319–330. Klosovskii, B.N., Zhukova, T.P., 1963. [Effect of colchicine on remote phases of growing capillaries in the brain.] Arkhiv. patologii. 35 (3), 38–44. Krah, K., Mironov, V., Risau, W., Flamme, I., 1994. Induction of vasculogenesis in quail blastodisc-derived embryoid bodies. Dev. Biol. 164, 123–132. Krebs, L.T., Xue, Y., Norton, C.R., Shutter, J.R., Maguire, M., Sundberg, J.P., Gallahan, D., Closson, V., Kitajewski, J., Callahan, R., Smith, G.H., Stark, K.L., Gridley, T., 2000. Notch signaling is essential for vascular morphogenesis in mice. Genes. Dev. 14, 1343–1352. Krebs, L.T., Shutter, J.R., Tanigaki, K., Honjo, T., Stark, K.L., Gridley, T., 2004. Haploinsufficient lethality and formation of arteriovenous malformations in Notch pathway mutants. Genes. Dev. 18, 2469–2473. Kuijper, S., Turner, C.J., Adams, R.H., 2007. Regulation of angiogenesis by Eph-ephrin interactions. Trends Cardiovasc. Med. 17, 145–151. Kumar, S., West, D.C., Ager, A., 1987. Heterogeneity in endothelial cells from large vessels and microvessels. Differentiation 36, 57–70. Labastie, M.C., Poole, T.J., Peault, B.M., Le Douarin, N.M., 1986. MB1, a quail leukocyte-endothelium antigen: partial characterization of the cell surface and secreted forms in cultured endothelial cells. Proc. Natl. Acad. Sci. USA 83, 9016–9020. Lampugnani, M.G., Resnati, M., Raiteri, M., Pigott, R., Pisacane, A., Houen, G., Ruco, L.P., Dejana, E., 1992. A novel endothelial-specific membrane protein is a marker of cell–cell contacts. J. Cell Biol. 118, 1511–1522. Lang, R.A., Bishop, J.M., 1993. Macrophages are required for cell death and tissue remodeling in the developing mouse eye. Cell 74, 453–462. Larrivee, B., Freitas, C., Trombe, M., Lv, X., Delafarge, B., Yuan, L., Bouvree, K., Breant, C., Del Toro, R., Brechot, N., Germain, S., Bono, F., Dol, F., Claes, F., Fischer, C., Autiero, M., Thomas, J.L., Carmeliet, P., Tessier-Lavigne, M., Eichmann, A., 2007. Activation of the UNC5B receptor by Netrin-1 inhibits sprouting angiogenesis. Genes. Dev. 21, 2433–2447. Larsson, J., Goumans, M.J., Sjostrand, L.J., van Rooijen, M.A., Ward, D., Leveen, P., Xu, X., ten Dijke, P., Mummery, C.L., Karlsson, S., 2001. Abnormal angiogenesis but intact hematopoietic potential in TGFbeta type I receptor-deficient mice. Embo. J. 20, 1663–1673. Latker, C.H., Kuwabara, T., 1981. Regression of the tunica vasculosa lentis in the postnatal rat. Invest Ophthalmol. Vis. Sci. 21, 689–699. Latker, C.H., Feinberg, R.N., Beebe, D.C., 1986. Localized vascular regression during limb morphogenesis in the chicken embryo: II.

524

Morphological changes in the vasculature. Anat. Rec. 214, 410–417, 392–413. Lawson, K.A., Meneses, J.J., Pedersen, R.A., 1991. Clonal analysis of epiblast fate during germ layer formation in the mouse embryo. Development 113, 891–911. Lawson, N.D., Vogel, A.M., Weinstein, B.M., 2002. Sonic hedgehog and vascular endothelial growth factor act upstream of the Notch pathway during arterial endothelial differentiation. Dev. Cell 3, 127–136. Le Lievre, C.S., Le Douarin, N.M., 1975. Mesenchymal derivatives of the neural crest: analysis of chimaeric quail and chick embryos. J. Embryol. Exper. Morph. 34, 125–154. le Noble, F., Moyon, D., Pardanaud, L., Yuan, L., Djonov, V., Matthijsen, R., Breant, C., Fleury, V., Eichmann, A., 2004. Flow regulates arterial–venous differentiation in the chick embryo yolk sac. Development 131, 361–375. Lebrin, F., Deckers, M., Bertolino, P., Ten Dijke, P., 2005. TGF-beta receptor function in the endothelium. Cardiovasc. Res. 65, 599–608. Lee, S.H., 1991. Phenotypic analysis of human bone marrow macrophages. Blood Cells 17, 45–54; discussion 54–58. Lee, Y.M., Jeong, C.H., Koo, S.Y., Son, M.J., Song, H.S., Bae, S.K., Raleigh, J.A., Chung, H.Y., Yoo, M.A., Kim, K.W., 2001. Determination of hypoxic region by hypoxia marker in developing mouse embryos in vivo: a possible signal for vessel development. Dev. Dyn. 220, 175–186. Leveen, P., Pekny, M., Gebre-Medhin, S., Swolin, B., Larsson, E., Betsholtz, C., 1994. Mice deficient for PDGF B show renal, cardiovascular, and hematological abnormalities. Genes. Dev. 8, 1875–1887. Li, D.Y., Sorensen, L.K., Brooke, B.S., Urness, L.D., Davis, E.C., Taylor, D.G., Boak, B.B., Wendel, D.P., 1999. Defective angiogenesis in mice lacking endoglin. Science 284, 1534–1537. Limbourg, F.P., Takeshita, K., Radtke, F., Bronson, R.T., Chin, M.T., Liao, J.K., 2005. Essential role of endothelial Notch1 in angiogenesis. Circulation 111, 1826–1832. Lindahl, P., Johansson, B.R., Leveen, P., Betsholtz, C., 1997. Pericyte loss and microaneurysm formation in PDGF-B-deficient mice. Science 277, 242–245. Liu, Y., Cox, S.R., Morita, T., Kourembanas, S., 1995. Hypoxia regulates vascular endothelial growth factor gene expression in endothelial cells. Identification of a 5 enhancer. Circ. Res. 77, 638–643. Liu, Y., Wada, R., Yamashita, T., Mi, Y., Deng, C.X., Hobson, J.P., Rosenfeldt, H.M., Nava, V.E., Chae, S.S., Lee, M.J., Liu, C.H., Hla, T., Spiegel, S., Proia, R.L., 2000. Edg-1, the G protein-coupled receptor for sphingosine-1-phosphate, is essential for vascular maturation. J. Clin. Invest. 106, 951–961. Liu, Z.J., Shirakawa, T., Li, Y., Soma, A., Oka, M., Dotto, G.P., Fairman, R.M., Velazquez, O.C., Herlyn, M., 2003. Regulation of Notch1 and Dll4 by vascular endothelial growth factor in arterial endothelial cells: implications for modulating arteriogenesis and angiogenesis. Mol. Cell Biol. 23, 14–25. Lobov, I.B., Rao, S., Carroll, T.J., Vallance, J.E., Ito, M., Ondr, J.K., Kurup, S., Glass, D.A., Patel, M.S., Shu, W., Morrisey, E.E., McMahon, A.P., Karsenty, G., Lang, R.A., 2005. WNT7b mediates macrophage-induced programmed cell death in patterning of the vasculature. Nature 437, 417–421. Lobov, I.B., Renard, R.A., Papadopoulos, N., Gale, N.W., Thurston, G., Yancopoulos, G.D., Wiegand, S.J., 2007. Delta-like ligand 4 (Dll4) is induced by VEGF as a negative regulator of angiogenic sprouting. Proc. Natl. Acad. Sci. USA 104, 3219–3224.

PART | 8

Making Vessels

Lofberg, J., Collazo, A., 1997. Hypochord, an enigmatic embryonic structure: study of the axolotl embryo. J. Morphol. 232, 57–66. Lohler, J., Timpl, R., Jaenisch, R., 1984. Embryonic lethal mutation in mouse collagen I gene causes rupture of blood vessels and is associated with erythropoietic and mesenchymal cell death. Cell 38, 597–607. Lu, X., Le Noble, F., Yuan, L., Jiang, Q., De Lafarge, B., Sugiyama, D., Breant, C., Claes, F., De Smet, F., Thomas, J.L., Autiero, M., Carmeliet, P., Tessier-Lavigne, M., Eichmann, A., 2004. The netrin receptor UNC5B mediates guidance events controlling morphogenesis of the vascular system. Nature 432, 179–186. Lubarsky, B., Krasnow, M.A., 2003. Tube morphogenesis: making and shaping biological tubes. Cell 112, 19–28. Lucitti, J.L., Jones, E.A., Huang, C., Chen, J., Fraser, S.E., Dickinson, M.E., 2007. Vascular remodeling of the mouse yolk sac requires hemodynamic force. Development 134, 3317–3326. Lundkvist, A., Lee, S., Iruela-Arispe, L., Betsholtz, C., Gerhardt, H., 2007. Growth factor gradients in vascular patterning. Novartis Found. Symp. 283, 194–201; discussion 201–196, 238–241. Luo, Y., Radice, G.L., 2005. N-cadherin acts upstream of VE-cadherin in controlling vascular morphogenesis. J. Cell Biol. 169, 29–34. Luscinskas, F.W., Lawler, J., 1994. Integrins as dynamic regulators of vascular function. Faseb. J. 8, 929–938. Madri, J.A., Williams, S.K., 1983. Capillary endothelial cell cultures: phenotypic modulation by matrix components. J. Cell. Biol. 97, 153–165. Maes, C., Stockmans, I., Moermans, K., Van Looveren, R., Smets, N., Carmeliet, P., Bouillon, R., Carmeliet, G., 2004. Soluble VEGF isoforms are essential for establishing epiphyseal vascularization and regulating chondrocyte development and survival. J. Clin. Invest. 113, 188–199. Maisonpierre, P.C., Suri, C., Jones, P.F., Bartunkova, S., Wiegand, S.J., Radziejewski, C., Compton, D., McClain, J., Aldrich, T.H., Papadopoulos, N., Daly, T.J., Davis, S., Sato, T.N., Yancopoulos, G.D., 1997. Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis. Science 277, 55–60. Manner, J., Seidl, W., Steding, G., 1995. Embryological observations on the morphogenesis of double-outlet right ventricle with subaortic ventricular septal defect and normal arrangement of the great arteries. Thorac Cardiovasc. Surg. 43, 307–312. Marron, M.B., Hughes, D.P., Edge, M.D., Forder, C.L., Brindle, N.P., 2000. Evidence for heterotypic interaction between the receptor tyrosine kinases TIE-1 and TIE-2. J. Biol. Chem. 275, 39741–39746. Mato, M., Aikawa, E., Kishi, K., 1964. Some observations on interstice between mesoderm and endoderm in the area vasculosa of chick blastoderm. Exp. Cell Res. 35, 426–428. Matsumoto, K., Yoshitomi, H., Rossant, J., Zaret, K.S., 2001. Liver organogenesis promoted by endothelial cells prior to vascular function. Science 294, 559–563. Matthews, W., Jordan, C.T., Wiegand, G.W., Pardoll, D., Lemischka, I.R., 1991. A receptor tyrosine kinase specific to hematopoietic stem and progenitor cell-enriched populations. Cell 65, 1143–1152. Mayer B.W., Jr, Hay, E.D., Hynes, R.O., 1981. Immunocytochemical localization of fibronectin in embryonic chick trunk and area vasculosa. Dev. Biol. 82, 267–286. Meeson, A., Palmer, M., Calfon, M., Lang, R., 1996. A relationship between apoptosis and flow during programmed capillary regression is revealed by vital analysis. Development 122, 3929–3938. Meier, S., 1980. Development of the chick embryo mesoblast: pronephros, lateral plate, and early vasculature. J. Embryol. Exper. Morphol. 55, 291–306.

Chapter | 8.2 Vascular Development

Meyer, G.T., Matthias, L.J., Noack, L., Vadas, M.A., Gamble, J.R., 1997. Lumen formation during angiogenesis in vitro involves phagocytic activity, formation and secretion of vacuoles, cell death, and capillary tube remodelling by different populations of endothelial cells. Anat. Rec. 249, 327–340. Mikawa, T., Gourdie, R.G., 1996. Pericardial mesoderm generates a population of coronary smooth muscle cells migrating into the heart along with ingrowth of the epicardial organ. Dev. Biol. 174, 221–232. Millauer, B., Wizigmann-Voos, S., Schnurch, H., Martinez, R., Moller, N.P., Risau, W., Ullrich, A., 1993a. High affinity VEGF binding and developmental expression suggest Flk-1 as a major regulator of vasculogenesis and angiogenesis. Cell 72, 835–846. Mills, K.R., Kruep, D., Saha, M.S., 1999. Elucidating the origins of the vascular system: a fate map of the vascular endothelial and red blood cell lineages in Xenopus laevis. Dev. Biol. 209, 352–368. Minami, T., Aird, W.C., 2005. Endothelial cell gene regulation. Trends Cardiovasc. Med. 15, 174–184. Minko, K., Bollerot, K., Drevon, C., Hallais, M.F., Jaffredo, T., 2003. From mesoderm to blood islands: patterns of key molecules during yolk sac erythropoiesis. Gene Expr. Patterns 3, 261–272. Mitrani, E., Gruenbaum, Y., Shohat, H., Ziv, T., 1990. Fibroblast growth factor during mesoderm induction in the early chick embryo. Development 109, 387–393. Modlich, U., Kaup, F.J., Augustin, H.G., 1996. Cyclic angiogenesis and blood vessel regression in the ovary: blood vessel regression during luteolysis involves endothelial cell detachment and vessel occlusion. Laboratory investigation; J. Tech. Methods Pathol. 74, 771–780. Montesano, R., Orci, L., Vassalli, P., 1983. In vitro rapid organization of endothelial cells into capillary-like networks is promoted by collagen matrices. J. Cell Biol. 97, 1648–1652. Moyon, D., Pardanaud, L., Yuan, L., Breant, C., Eichmann, A., 2001. Plasticity of endothelial cells during arterial–venous differentiation in the avian embryo. Development 128, 3359–3370. Mukouyama, Y.S., Shin, D., Britsch, S., Taniguchi, M., Anderson, D.J., 2002. Sensory nerves determine the pattern of arterial differentiation and blood vessel branching in the skin. Cell 109, 693–705. Mukouyama, Y.S., Gerber, H.P., Ferrara, N., Gu, C., Anderson, D.J., 2005. Peripheral nerve-derived VEGF promotes arterial differentiation via neuropilin 1-mediated positive feedback. Development 132, 941–952. Murray, P.D.F., 1932. The development in vitro of the blood of the early chick embryo. Proc. R. Soc. Lond. 111, 497–521. Nagase, T., Nagase, M., Machida, M., Fujita, T., 2008. Hedgehog signalling in vascular development. Angiogenesis 11(1), 71–77. Nakazawa, F., Nagai, H., Shin, M., Sheng, G., 2006. Negative regulation of primitive hematopoiesis by the FGF signaling pathway. Blood 108, 3335–3343. Nanka, O., Valasek, P., Dvorakova, M., Grim, M., 2006. Experimental hypoxia and embryonic angiogenesis. Dev. Dyn. 235, 723–733. Nasevicius, A., Larson, J., Ekker, S.C., 2000. Distinct requirements for zebrafish angiogenesis revealed by a VEGF-A morphant. Yeast (Chichester, England) 17, 294–301. Navankasattusas, S., Whitehead, K.J., Suli, A., Sorensen, L.K., Lim, A.H., Zhao, J., Park, K.W., Wythe, J.D., Thomas, K.R., Chien, C.B., Li, D.Y., 2008. The netrin receptor UNC5B promotes angiogenesis in specific vascular beds. Development 135, 659–667. Nelsen, O.E., 1953. Comparative Embryology of the Vertebrates. McGraw Hill Book Company, Inc., Blaskiston, NY.

525

Nishikawa, S.I., Nishikawa, S., Hirashima, M., Matsuyoshi, N., Kodama, H., 1998. Progressive lineage analysis by cell sorting and culture identifies FLK1VE-cadherincells at a diverging point of endothelial and hemopoietic lineages. Development 125, 1747–1757. Noden, D.M., 1989. Embryonic origins and assembly of blood vessels. Am. Rev. Resp. Dis. 140, 1097–1103. Noden, D.M., 1990. Origins and assembly of avian embryonic blood vessels. Ann. NY Acad. Sci. 588, 236–249. Oettgen, P., 2001. Transcriptional regulation of vascular development. Circ. Res. 89, 380–388. Oh, S.P., Seki, T., Goss, K.A., Imamura, T., Yi, Y., Donahoe, P.K., Li, L., Miyazono, K., ten Dijke, P., Kim, S., Li, E., 2000. Activin receptor-like kinase 1 modulates transforming growth factor-beta 1 signaling in the regulation of angiogenesis. Proc. Natl. Acad. Sci. USA 97, 2626–2631. Orlidge, A., D’Amore, P.A., 1987. Inhibition of capillary endothelial cell growth by pericytes and smooth muscle cells. J. Cell Biol. 105, 1455–1462. Oshima, M., Oshima, H., Taketo, M.M., 1996a. TGF-beta receptor type II deficiency results in defects of yolk sac hematopoiesis and vasculogenesis. Dev. Biol. 179, 297–302. Owens, G.K., Thompson, M.M., 1986. Developmental changes in isoactin expression in rat aortic smooth muscle cells in vivo. Relationship between growth and cytodifferentiation. J. Biol. Chem. 261, 13373–13380. Palis, J., McGrath, K.E., Kingsley, P.D., 1995. Initiation of hematopoiesis and vasculogenesis in murine yolk sac explants. Blood 86, 156–163. Palis, J., Robertson, S., Kennedy, M., Wall, C., Keller, G., 1999. Development of erythroid and myeloid progenitors in the yolk sac and embryo proper of the mouse. Development 126, 5073–5084. Pardanaud, L., Altmann, C., Kitos, P., Dieterlen-Lievre, F., Buck, C.A., 1987. Vasculogenesis in the early quail blastodisc as studied with a monoclonal antibody recognizing endothelial cells. Development 100, 339–349. Pardanaud, L., Yassine, F., Dieterlen-Lievre, F., 1989a. Relationship between vasculogenesis, angiogenesis and haemopoiesis during avian ontogeny. Development 105, 473–485. Pardanaud, L., Yassine, F., Dieterlen-Lievre, F., 1989b. Relationship between vasculogenesis, angiogenesis and haemopoiesis during avian ontogeny. Development 105, 473–485. Pardanaud, L., Dieterlen-Lievre, F., 1993. Emergence of endothelial and hemopoietic cells in the avian embryo. Anat. Embryol. 187, 107–114. Pardanaud, L., Luton, D., Prigent, M., Bourcheix, L.M., Catala, M., Dieterlen-Lievre, F., 1996. Two distinct endothelial lineages in ontogeny, one of them related to hemopoiesis. Development 122, 1363–1371. Park, C., Ma, Y.D., Choi, K., 2005. Evidence for the hemangioblast. Exp. Hematol. 33, 965–970. Park, K.W., Morrison, C.M., Sorensen, L.K., Jones, C.A., Rao, Y., Chien, C.B., Wu, J.Y., Urness, L.D., Li, D.Y., 2003. Robo4 is a vascularspecific receptor that inhibits endothelial migration. Dev. Biol. 261, 251–267. Park, K.W., Crouse, D., Lee, M., Karnik, S.K., Sorensen, L.K., Murphy, K.J., Kuo, C.J., Li, D.Y., 2004. The axonal attractant Netrin-1 is an angiogenic factor. Proc. Natl. Acad. Sci. USA 101, 16210–16215. Parker, L.H., Schmidt, M., Jin, S.W., Gray, A.M., Beis, D., Pham, T., Frantz, G., Palmieri, S., Hillan, K., Stainier, D.Y., De Sauvage, F.J., Ye, W., 2004. The endothelial-cell-derived secreted factor Egfl7 regulates vascular tube formation. Nature 428, 754–758.

526

Patan, S., Alvarez, M.J., Schittny, J.C., Burri, P.H., 1992. Intussusceptive microvascular growth: a common alternative to capillary sprouting. Archives Histol. Cytol. 55 (Suppl), 65–75. Patan, S., Munn, L.L., Jain, R.K., 1996. Intussusceptive microvascular growth in a human colon adenocarcinoma xenograft: a novel mechanism of tumor angiogenesis. Microvasc. Res. 51, 260–272. Peault, B.M., Thiery, J.P., Le Douarin, N.M., 1983. Surface marker for hemopoietic and endothelial cell lineages in quail that is defined by a monoclonal antibody. Proc. Natl. Acad. Sci. USA 80, 2976–2980. Peault, B., Coltey, M., Le Douarin, N.M., 1988. Ontogenic emergence of a quail leukocyte/endothelium cell surface antigen. Cell Differentiation 23, 165–174. Pepicelli, C.V., Lewis, P.M., McMahon, A.P., 1998. Sonic hedgehog regulates branching morphogenesis in the mammalian lung. Curr. Biol. 8, 1083–1086. Perez-Pomares, J.M., Macias, D., Garcia-Garrido, L., Munoz-Chapuli, R., 1998. The origin of the subepicardial mesenchyme in the avian embryo: an immunohistochemical and quail-chick chimera study. Dev. Biol. 200, 57–68. Pham, V.N., Lawson, N.D., Mugford, J.W., Dye, L., Castranova, D., Lo, B., Weinstein, B.M., 2007. Combinatorial function of ETS transcription factors in the developing vasculature. Dev. Biol. 303, 772–783. Poole, T.J., Coffin, J.D., 1989. Vasculogenesis and angiogenesis: two distinct morphogenetic mechanisms establish embryonic vascular pattern. J. Exp. Zool. 251, 224–231. Puri, M.C., Rossant, J., Alitalo, K., Bernstein, A., Partanen, J., 1995. The receptor tyrosine kinase TIE is required for integrity and survival of vascular endothelial cells. Embo. J. 14, 5884–5891. Reagan, F.P., 1915. Vasculization phenomena in fragments of embryonic bodies completely devoid from yolk-sac entoderm. Anat. Rec. 9, 329–341. Reese, D.E., Hall, C.E., Mikawa, T., 2004. Negative regulation of midline vascular development by the notochord. Dev. Cell 6, 699–708. Reidy, M.A., Schwartz, S.M., 1981. Endothelial regeneration. III. Time course of intimal changes after small defined injury to rat aortic endothelium. Laboratory investigation; J. Tech. Meth. Pathol. 44, 301–308. Resnick N., Gimbrone, M.A., Jr., 1995. Hemodynamic forces are complex regulators of endothelial gene expression. Faseb. J. 9, 874–882. Reynolds, L.E., Wyder, L., Lively, J.C., Taverna, D., Robinson, S.D., Huang, X., Sheppard, D., Hynes, R.O., Hodivala-Dilke, K.M., 2002. Enhanced pathological angiogenesis in mice lacking beta3 integrin or beta3 and beta5 integrins. Nat. Med. 8, 27–34. Rhodin, J.A., 1968. Ultrastructure of mammalian venous capillaries, venules, and small collecting veins. J. Ultrastruct. Res. 25, 452–500. Rhodin, J. A. (1980). Architecture of the vessel wall. In “Handbook of Physiology” (S. R. Geiger, Ed.). Am Physiol Soc., Bethesda, MD. Risau, W., 1991. Embryonic angiogenesis factors. Pharmacol. Ther. 51, 371–376. Risau, W., 1997. Mechanisms of angiogenesis. Nature 386, 671–674. Risau, W., Lemmon, V., 1988. Changes in the vascular extracellular matrix during embryonic vasculogenesis and angiogenesis. Dev. Biol. 125, 441–450. Risau, W., Sariola, H., Zerwes, H.G., Sasse, J., Ekblom, P., Kemler, R., Doetschman, T., 1988. Vasculogenesis and angiogenesis in embryonic-stem-cell-derived embryoid bodies. Development 102, 471–478.

PART | 8

Making Vessels

Risau, W., Flamme, I., 1995. Vasculogenesis. Annu. Rev. Cell Dev. Biol. 11, 73–91. Robert, B., St John, P.L., Abrahamson, D.R., 1998. Direct visualization of renal vascular morphogenesis in Flk1 heterozygous mutant mice. Am. J. Physiol. 275, F164–F172. Roy, H., Bhardwaj, S., Yla-Herttuala, S., 2006. Biology of vascular endothelial growth factors. FEBS Letts. 580, 2879–2887. Rupp, P.A., Czirok, A., Little, C.D., 2004. Alphavbeta3 integrindependent endothelial cell dynamics in vivo. Development 131, 2887–2897. Sabin, F.R., 1917. Origin and development of the primitive vessels of the chick and the pig. Carnegie Contrib Embryol 6, 61–124. Sabin, F.R., 1920. Studies on the origin of the blood vessels and of red blood corpuscles as seen in the living blastoderm of chick during the second day of incubation. Carnegie Contrib. Embryol. 9, 215–262. Sariola, H., Ekblom, P., Lehtonen, E., Saxen, L., 1983. Differentiation and vascularization of the metanephric kidney grafted on the chorioallantoic membrane. Dev. Biol. 96, 427–435. Sato, T.N., Tozawa, Y., Deutsch, U., Wolburg-Buchholz, K., Fujiwara, Y., Gendron-Maguire, M., Gridley, T., Wolburg, H., Risau, W., Qin, Y., 1995. Distinct roles of the receptor tyrosine kinases Tie-1 and Tie-2 in blood vessel formation. Nature 376, 70–74. Schmidt, A., Brixius, K., Bloch, W., 2007a. Endothelial precursor cell migration during vasculogenesis. Circ. Res. 101, 125–136. Schmidt, M., Paes, K., De Maziere, A., Smyczek, T., Yang, S., Gray, A., French, D., Kasman, I., Klumperman, J., Rice, D.S., Ye, W., 2007b. EGFL7 regulates the collective migration of endothelial cells by restricting their spatial distribution. Development 134, 2913–2923. Schwartz, S.M., Liaw, L., 1993. Growth control and morphogenesis in the development and pathology of arteries. J. Cardiovasc. Pharmacol. 21 (Suppl 1), S31–S49. Seo, S., Fujita, H., Nakano, A., Kang, M., Duarte, A., Kume, T., 2006. The forkhead transcription factors, Foxc1 and Foxc2, are required for arterial specification and lymphatic sprouting during vascular development. Dev. Biol. 294, 458–470. Serini, G., Valdembri, D., Bussolino, F., 2006. Integrins and angiogenesis: a sticky business. Exp. Cell Res. 312, 651–658. Shalaby, F., Rossant, J., Yamaguchi, T.P., Gertsenstein, M., Wu, X.F., Breitman, M.L., Schuh, A.C., 1995. Failure of blood-island formation and vasculogenesis in Flk-1-deficient mice. Nature 376, 62–66. Shaw, K.M., Castranova, D.A., Pham, V.N., Kamei, M., Kidd, K.R., Lo, B.D., Torres-Vasquez, J., Ruby, A., Weinstein, B.M., 2006. fused-somites-like mutants exhibit defects in trunk vessel patterning. Dev. Dyn. 235, 1753–1760. Sholley, M.M., Ferguson, G.P., Seibel, H.R., Montour, J.L., Wilson, J.D., 1984. Mechanisms of neovascularization. Vascular sprouting can occur without proliferation of endothelial cells. Laboratory investigation; J. Tech. Meth. Pathol. 51, 624–634. Soker, S., Miao, H.Q., Nomi, M., Takashima, S., Klagsbrun, M., 2002. VEGF165 mediates formation of complexes containing VEGFR-2 and neuropilin-1 that enhance VEGF165-receptor binding. J. Cell. Biochem. 85, 357–368. Soriano, P., 1994. Abnormal kidney development and hematological disorders in PDGF beta-receptor mutant mice. Genes Dev. 8, 1888–1896. Spiegel, S., Milstien, S., 2002. Sphingosine 1-phosphate, a key cell signaling molecule. J. Biol. Chem. 277, 25851–25854.

Chapter | 8.2 Vascular Development

St Croix, B., Rago, C., Velculescu, V., Traverso, G., Romans, K.E., Montgomery, E., Lal, A., Riggins, G.J., Lengauer, C., Vogelstein, B., Kinzler, K.W., 2000. Genes expressed in human tumor endothelium. Science 289, 1197–1202. Stein, E., Tessier-Lavigne, M., 2001. Hierarchical organization of guidance receptors: silencing of netrin attraction by slit through a Robo/ DCC receptor complex. Science 291, 1928–1938. Stone, J., Itin, A., Alon, T., Pe’er, J., Gnessin, H., Chan-Ling, T., Keshet, E., 1995. Development of retinal vasculature is mediated by hypoxia-induced vascular endothelial growth factor (VEGF) expression by neuroglia. J. Neurosci. 15, 4738–4747. Stromblad, S., Cheresh, D.A., 1996. Cell adhesion and angiogenesis. Trends Cell Biol. 6, 462–468. Stupack, D.G., Puente, X.S., Boutsaboualoy, S., Storgard, C.M., Cheresh, D.A., 2001. Apoptosis of adherent cells by recruitment of caspase-8 to unligated integrins. J. Cell Biol. 155, 459–470. Suchting, S., Heal, P., Tahtis, K., Stewart, L.M., Bicknell, R., 2005. Soluble Robo4 receptor inhibits in vivo angiogenesis and endothelial cell migration. Faseb. J. 19, 121–123. Suchting, S., Bicknell, R., Eichmann, A., 2006. Neuronal clues to vascular guidance. Exp. Cell Res. 312, 668–675. Sumoy, L., Keasey, J.B., Dittman, T.D., Kimelman, D., 1997. A role for notochord in axial vascular development revealed by analysis of phenotype and the expression of VEGR-2 in zebrafish flh and ntl mutant embryos. Mech. Dev. 63, 15–27. Sumpio, B.E., Riley, J.T., Dardik, A., 2002. Cells in focus: endothelial cell. Int. J. Biochem. Cell Biol. 34, 1508–1512. Suri, C., Jones, P.F., Patan, S., Bartunkova, S., Maisonpierre, P.C., Davis, S., Sato, T.N., Yancopoulos, G.D., 1996. Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis. Cell 87, 1171–1180. Suter, E.R., Majno, G., 1965. Passage of lipid across vascular endothelium in newborn rats. An electron microscopic study. J. Cell Biol. 27, 163–177. Suzuki, S., Sano, K., Tanihara, H., 1991. Diversity of the cadherin family: evidence for eight new cadherins in nervous tissue. Cell Regul. 2, 261–270. Suzuki, T., Sasano, H., Takaya, R., Fukaya, T., Yajima, A., Nagura, H., 1998. Cyclic changes of vasculature and vascular phenotypes in normal human ovaries. Human Reproduction (Oxford, England) 13, 953–959. Takashima, S., Kitakaze, M., Asakura, M., Asanuma, H., Sanada, S., Tashiro, F., Niwa, H., Miyazaki Ji, J., Hirota, S., Kitamura, Y., Kitsukawa, T., Fujisawa, H., Klagsbrun, M., Hori, M., 2002. Targeting of both mouse neuropilin-1 and neuropilin-2 genes severely impairs developmental yolk sac and embryonic angiogenesis. Proc. Natl. Acad. Sci. USA 99, 3657–3662. Tavian, M., Coulombel, L., Luton, D., Clemente, H.S., Dieterlen-Lievre, F., Peault, B., 1996. Aorta-associated CD34 hematopoietic cells in the early human embryo. Blood 87, 67–72. Tilton, R.G., Miller, E.J., Kilo, C., Williamson, J.R., 1985. Pericyte form and distribution in rat retinal and uveal capillaries. Invest Ophthalmol. Vis. Sci. 26, 68–73. Torres-Vazquez, J., Gitler, A.D., Fraser, S.D., Berk, J.D., Van, N.P., Fishman, M.C., Childs, S., Epstein, J.A., Weinstein, B.M., 2004. Semaphorin-plexin signaling guides patterning of the developing vasculature. Dev. Cell 7, 117–123. Tufro, A., Norwood, V.F., Carey, R.M., Gomez, R.A., 1999. Vascular endothelial growth factor induces nephrogenesis and vasculogenesis. J. Am. Soc. Nephrol. 10, 2125–2134.

527

Ueno, H., Weissman, I.L., 2006. Clonal analysis of mouse development reveals a polyclonal origin for yolk sac blood islands. Dev. Cell 11, 519–533. Umans, L., Cox, L., Tjwa, M., Bito, V., Vermeire, L., Laperre, K., Sipido, K., Moons, L., Huylebroeck, D., Zwijsen, A., 2007. Inactivation of Smad5 in endothelial cells and smooth muscle cells demonstrates that Smad5 is required for cardiac homeostasis. Am. J. Pathol. 170, 1460–1472. Vestweber, D., 2008. VE-cadherin: the major endothelial adhesion molecule controlling cellular junctions and blood vessel formation. Arterioscler. Thromb. Vasc. Biol. 28, 223–232. Villa, N., Walker, L., Lindsell, C.E., Gasson, J., Iruela-Arispe, M.L., Weinmaster, G., 2001. Vascular expression of Notch pathway receptors and ligands is restricted to arterial vessels. Mech. Dev. 108, 161–164. Vokes, S.A., Krieg, P.A., 2002. Endoderm is required for vascular endothelial tube formation, but not for angio-blast specification. Development 129, 775–785. Vokes, S.A., Yatskievych, T.A., Heimark, R.L., McMahon, J., McMahon, A.P., Antin, P.B., Krieg, P.A., 2004. Hedgehog signaling is essential for endothelial tube formation during vasculogenesis. Development 131, 4371–4380. Wagner, R.C., 1980. Endothelial cell embryology and growth. Adv. Microcirc. 9, 45–75. Wakimoto, K., Kobayashi, K., Kuro, O.M., Yao, A., Iwamoto, T., Yanaka, N., Kita, S., Nishida, A., Azuma, S., Toyoda, Y., Omori, K., Imahie, H., Oka, T., Kudoh, S., Kohmoto, O., Yazaki, Y., Shigekawa, M., Imai, Y., Nabeshima, Y., Komuro, I., 2000. Targeted disruption of Na/Ca2 exchanger gene leads to cardiomyocyte apoptosis and defects in heartbeat. J. Biol. Chem. 275, 36991–36998. Walmsley, M., Ciau-Uitz, A., Patient, R., 2002. Adult and embryonic blood and endothelium derive from distinct precursor populations which are differentially programmed by BMP in Xenopus. Development 129, 5683–5695. Waltenberger, J., Claesson-Welsh, L., Siegbahn, A., Shibuya, M., Heldin, C.H., 1994. Different signal transduction properties of KDR and Flt1, two receptors for vascular endothelial growth factor. J. Biol. Chem. 269, 26988–26995. Wang, Y., McMahon, A.P., Allen, B.L., 2007. Shifting paradigms in Hedgehog signaling. Curr. Opin. Cell Biol. 19, 159–165. Watson, A.J., Barcroft, L.C., 2001. Regulation of blastocyst formation. Front Biosci. 6, D708–D730. Whelan, M.C., Senger, D.R., 2003. Collagen I initiates endothelial cell morphogenesis by inducing actin polymerization through suppression of cyclic AMP and protein kinase A. J. Biol. Chem. 278, 327–334. Wilt, F.H., 1965. Regulation of the initiation of chick embryo hemoglobin synthesis. J. Mol. Biol. 12, 331–341. Wilt, F.H., 1974. The beginnings of erythropoiesis in the yolk sac of the chick embryo. Ann. NY Acad. Sci. 241, 99–112. Wilting, J., Brand-Saberi, B., Kurz, H., Christ, B., 1995. Development of the embryonic vascular system. Cell. Mol. Biol. Res. 41, 219–232. Wilting, J., Christ, B., 1996. Embryonic angiogenesis: a review. Die Naturwissenschaften 83, 153–164. Wolff, J.R., Bar, T., 1972. “Seamless” endothelia in brain capillaries during development of the rat’s cerebral cortex. Brain Res. 41, 17–24. Yamaguchi, T.P., Dumont, D.J., Conlon, R.A., Breitman, M.L., Rossant, J., 1993. Flk-1, an flt-related receptor tyrosine kinase is an early marker for endothelial cell precursors. Development 118, 489–498.

528

Yang, J.T., Rayburn, H., Hynes, R.O., 1993. Embryonic mesodermal defects in alpha 5 integrin-deficient mice. Development 119, 1093–1105. Yang, X., Castilla, L.H., Xu, X., Li, C., Gotay, J., Weinstein, M., Liu, P. P., Deng, C.X., 1999. Angiogenesis defects and mesenchymal apoptosis in mice lacking SMAD5. Development 126, 1571–1580.

PART | 8

Making Vessels

You, L.R., Lin, F.J., Lee, C.T., DeMayo, F.J., Tsai, M.J., Tsai, S.Y., 2005. Suppression of Notch signalling by the COUP-TFII transcription factor regulates vein identity. Nature 435, 98–104. Zagris, N., 1980. Erythroid cell differentiation in unincubated chick blastoderm in culture. J. Embryol. Exper. Morphol. 58, 209–216.

Chapter 8.3

Arteriovenous Patterning in the Vascular System Thomas Gridley The Jackson Laboratory, Bar Harbor, ME, USA

I.  Introduction One of the primary distinctions in the mammalian vasculature is its division into arterial and venous systems (see also Chapter 8.2). As children, we become familiar with diagrams of the human body that depict arteries in red and veins in blue. These simple diagrams underscore the functional, anatomical and structural differences that exist between arteries and veins. While both vessel types are lined by a thin inner layer of endothelial cells, arteries have a thicker vessel wall with more elastic fibers and vascular smooth muscle cells to support the higher blood pressure in arteries. Veins, on the other hand, contain valves to prevent retrograde flow of blood. Acquisition of these morphological and structural differences has long been attributed to the physiological factors that differ between these two vessel types. Recent work has established that, contrary to the historic view, genetic prepatterning prior to the onset of circulation is a primary determinant in regulating the differentiation of arteries and veins. However, this genetic prepattern is plastic, and is modified and optimized by environmental cues such as hemodynamic flow.

II.  Arteriovenous patterning: genetic prepatterning II.A.  Ephrins and EPH Receptors It had long been believed that the primary factor regulating differentiation of arteries and veins was blood flow. Endothelial cells lining arteries experience higher blood pressures, higher rates of hemodynamic flow, and higher oxygen tensions than endothelial cells lining veins. Heart Development and Regeneration Copyright © 2010, 2010 Elsevier Inc. All rights of reproduction in any form reserved.

However, work over the last eight years has established that genetic prepatterning plays a primary role in regulating arteriovenous differentiation and patterning (Fig. 1). The first indication of genetic patterning during arteriovenous differentiation in early embryos came from analysis of ephrinB2 mutant mice (Wang et al., 1998). Ephrin family proteins are membrane-bound ligands for the Eph receptors, the largest family of receptor tyrosine kinases in vertebrates. Ephrin/Eph signaling regulates multiple tissue patterning and morphogenic events, including axon guidance, somite formation and cell migration (Poliakov et al., 2004; Pasquale, 2005). Analysis of mice heterozygous for a targeted mutation of the ephrinB2 gene, in which expression of the lacZ gene encoding the b-galactosidase protein was under transcriptional control of ephrinB2 coding sequences, revealed that the ephrinB2-lacZ allele was expressed only in arterial endothelial cells (Wang et al., 1998). In situ hybridization with an ephrinB2 riboprobe demonstrated that the endogenous ephrinB2 gene was expressed in the same pattern. Notably, ephrinB2 expression was restricted to arterial endothelial cells from very early developmental stages, prior to the initiation of blood flow in the embryo. Conversely, it was shown that the EphB4 receptor was preferentially, although not exclusively, expressed on veins (Wang et al., 1998; Adams et al., 1999; Gerety et al., 1999). This expression pattern suggested that EphB4 could be the functional receptor for the ephrinB2 ligand in the developing vasculature. Mouse embryos homozygous for the ephrinB2-lacZ allele exhibited an embryonic-lethal phenotype in which, while initial formation of the primary vascular plexus occurred relatively normally in both the embryo and the extraembryonic yolk sac, angiogenic remodeling of that

529

530

PART  |  8  Making Vessels

(A) Artery

Arterioles

(B)

Capillaries

Venules

Vegf

Foxc1/2

?

EphrinB2 Hey1/2

Vegf

VegfR2, Neuropilin1

DII4, Notch1/4

Vein

PLC� PKC

EphB4

Foxc1/2

?

VegfR2, Neuropilin1

DII4, Notch1/4 EphrinB2 Hey1/2

PLC�

COUP-TFII

PI3K

PKC EphB4

Raf

Raf

MEK

MEK

ERK

ERK

Akt

Figure 1  Genetic prepatterning of the embryonic vasculature. The top figure is a schematic depiction of the arterial–venous network. Arteries communicate with veins only through an intervening capillary bed. At the bottom, a model for genetic regulation of artery-vein differentiation is shown. The left side depicts artery differentiation, while the right side depicts vein differentiation. Two main signaling pathways operate downstream of Vegf to induce artery differentiation: the Notch pathway (green box); and the PLCg/MAPK pathway (gray box). The transcription factors Foxc1 and Foxc2 induce Dll4 gene expression, but it is not known whether expression of the Foxc1 and Foxc2 genes is regulated by Vegf. During vein differentiation, two different mechanisms inhibit artery differentiation. The orphan nuclear receptor COUP-TFII suppresses neuropilin1 expression, thereby suppressing reception of the Vegf signal and activation of Notch signaling. In addition, activation of PI3K/Akt signaling antagonizes promotion of arterial cell differentiation by blocking ERK activation.

plexus to form a more mature vascular system did not occur (Wang et al., 1998). EphB4 homozygous mutant mouse embryos exhibited a nearly identical phenotype, substantiating the hypothesis that the EphB4 protein serves as the physiologically-relevant receptor of the ephrinB2 protein during early vascular development in mice (Gerety et al., 1999). Subsequent to the identification of the ephrinB2 and ephB4 genes as markers of the arterial and venous endothelial cell lineages, respectively, a number of other genes in several different vertebrate model systems have been identified that are preferentially expressed in either the arterial or venous lineage. Arterial markers include those encoded by the ephrinB2 (Efnb2), neuropilin1 (Nrp1), connexin 37 (Gja4) and 40 (Gja5) and Acvrl1 genes, as well as numerous Notch pathway components,

including the Dll4, Notch1 and Notch4 genes. There are fewer venous markers than the arterially-restricted markers, and they include those encoded by the Ephb4, neuropilin2 (Nrp2), COUP-TFII (Nr2f2), Flt4 and angiotensin receptor-like 1 (Agtrl1) genes. Not all of these markers are unique arterial or venous markers in all species. However some, such as ephrinB2 gene- and protein-expression as an arterial marker, are remarkably robust and appear to be arterial markers in all species examined. While evidence for genetic prepatterning of arterio­ venous differentiation was first discovered during mouse embryogenesis, work in the zebrafish and chick model systems has added substantially to our understanding of this process during vertebrate development. For example, analysis of the expression of the neuropilin1 and neuropilin2 genes in blood islands of the chick embryo yolk sac

Chapter  |  8.3  Arteriovenous Patterning in the Vascular System

revealed that, even prior to the formation of a vascular plexus, some blood islands displayed restricted expression patterns of these genes (Herzog et al., 2005). This suggests that genetic prepatterning could occur even prior to vessel formation, although no fate mapping studies have yet been performed to demonstrate that, for example, angio­ blasts from neuropilin1-expressing blood islands give rise to arteries.

II.B.  Vascular Endothelial Growth Factor   and the Notch Pathway in Zebrafish Two signaling pathways that play critical roles in vascular development and arteriovenous patterning are the vascular endothelial growth factor (Vegf) and the Notch pathways. Vegf (also termed Vegfa) is a secreted glycoprotein that is a potent inducer of angiogenesis which also regulates multiple other aspects of blood vessel homeostasis (Byrne et al., 2005; Coultas et al., 2005; Cebe-Suarez et al., 2006; Shibuya, 2006; Shibuya and Claesson-Welsh, 2006), while the Notch pathway is an evolutionarily-conserved intercellular signaling mechanism (Bray, 2006; Ehebauer et al., 2006). Genes of the Notch family encode large transmembrane receptors that interact with membrane-bound ligands encoded in vertebrates by Delta-like (Dll) and Jagged (Jag) family genes. The signal induced by ligand binding is transmitted intracellularly by a process involving proteolytic cleavage of the receptor and nuclear translocation of the intracellular domain of the Notch family protein (Notch-IC). Once in the nucleus, Notch-IC forms a complex with the RBP-J protein, a sequence-specific DNAbinding protein that is the primary transcriptional mediator of Notch signaling. The Notch-IC/RBP-J complex then activates transcription of downstream target genes, particularly certain members of the Hes/Hey family of basichelix-loop-helix (bHLH) transcriptional repressors. The roles of the Notch and Vegf pathways in regulating formation of the large trunk axial blood vessels, the dorsal aorta and the posterior cardinal vein, was studied first in zebrafish (Lawson et al., 2001, 2002). Embryos in which Notch signaling was reduced exhibited a poorly-formed dorsal aorta and posterior cardinal vein accompanied by formation of arteriovenous malformations, the fusion of arteries and veins without an intervening capillary bed. These embryos exhibited a loss of expression of arterial markers such as ephrinB2 from arterial vessels, with an accompanying expansion of venous markers into normally arterial domains. Embryos in which Notch signaling had been ectopically activated exhibited the reverse phenotype; suppression of vein-specific markers with ectopic expression of arterial markers in venous vessels (Lawson et al., 2001). A broadly similar phenotype was observed in embryos mutant for some Notch target genes. A target gene family commonly activated by Notch signaling

531

is the Hes/Hey family of bHLH transcriptional repressors. Zebrafish homozygous for a Hey2 (referred to as the gridlock gene in zebrafish) hypomorphic mutation exhibit a malformed aorta that prevents blood circulation to the trunk and tail (Weinstein et al., 1995; Zhong et al., 2000). The Hey2 gene is expressed in arteries; in embryos with reduced Hey2 activity expression of the arterial marker ephrinB2 was reduced and the venous marker EphB4 was increased (Zhong et al., 2001). Analysis of formation of the major trunk vessels of the zebrafish embryo revealed a signaling cascade respons­ ible for determining arterial and venous cell fates in these vessels (Lawson et al., 2002). Reduction of Vegf activity resulted in loss of expression of arterial markers from the dorsal aorta and ectopic arterial expression of vein markers. Injection of Vegf mRNA induced ectopic expression of the ephrinB2 gene in the posterior cardinal vein. Vegf expression was regulated by expression of the secreted morphogen Sonic hedgehog (Shh) along the axial midline. Similar to that which was observed in Vegf-deficient embryos, Shh mutant zebrafish also exhibited loss of arterial differentiation, while injection of Shh mRNA caused ectopic expression of arterial markers. Shh acted upstream of Vegf, since injection of Vegf mRNA into the Shh mutant embryos rescued arterial differentiation. This approach also demonstrated that the Notch pathway acted downstream of the Vegf pathway. While injection of Vegf mRNA into Notch signaling-deficient zebrafish embryos could not rescue arterial marker gene expression, expression of an activated Notch1 transgene in Vegf-deficient embryos could rescue expression of arterial markers (Lawson et al., 2002). A recently-described component of the Shh-Vegf-Notch signaling cascade is the calcitonin receptor-like (Calcrl) receptor (Nicoli et al., 2008). The Calcrl gene encodes a seven transmembrane G-protein-coupled receptor that is the primary receptor for the peptide vasodilator adreno­ medullin during embryonic development. Zebrafish Calcrl morphants exhibited no apparent blood circulation, and displayed a disorganized trunk vascular plexus with a severely reduced dorsal aorta. Arterial endothelial cell markers were not expressed in the Calcrl morphants, while expression of venous markers was not affected. Vegf RNA expression was also downregulated in the Calcrl morphants, and injection of Vegf RNA into Calcrl-morphant embryos was sufficient to restore expression of the arterial marker ephrinB2 in the dorsal aorta of the injected morphant embryos. Expression of the Calcrl gene was lost in zebrafish embryos homozygous for a Shh-null mutation. These results demonstrate that the Calcrl gene encodes a novel component of the Shh-Vegf-Notch signaling cascade responsible for controlling arteriovenous specification of the major axial vessels of the zebrafish embryo (Nicoli et al., 2008). Formation of the different blood vessel types in zebrafish embryos is also regulated through distinct genetic interactions among genes encoding the multiple

532

Vegf receptors (Covassin et al., 2006). A screen for zebrafish embryos that lack segmental arteries resulted in the isolation of a missense mutant in the gene encoding the Vegfr2 receptor. This mutation (termed kdray17) eliminated the kinase activity of the Vegfr2 protein. Embryos homozygous for the kdray17 mutation had a poorly-formed dorsal aorta, circulatory defects such as arteriovenous malformations, and reduced expression of arterial markers such as ephrinB2. These defects were milder than other loss-of-function mutants in the Vegf signaling pathway in zebrafish. For example, exposure of embryos to a drug that blocks the catalytic activity of all Vegf receptors completely eliminates segmental artery formation, indicating that Vegf receptors other than Vegfr2 also play essential roles in artery formation. Analysis of embryos with reduced Vegfr3 (termed flt4 in zebrafish) or Vegfc expression revealed genetic interactions with the kdray17 mutation. These Vegf receptors were also required for formation of veins, but revealed distinct genetic interactions that differed from those required for artery development. These results suggest that expression of genes encoding the different Vegf receptors is responsible, at least in part, for generating blood vessel diversity during zebrafish development.

II.C.  Vascular Endothelial Growth Factor   and the Notch Pathway in Mice In addition to the work described above in zebrafish embryos (Lawson et al., 2002), studies of mammalian cells in culture have also placed the Notch pathway genetically downstream of the Vegf pathway. Vegf administration could induce expression of mRNA for the Notch1 receptor and the Dll4 ligand in human arterial endothelial cells, but not in venous endothelial cells (Liu et al., 2003). Vegf is essential for vascular development in mice. Embryos heterozygous for a Vegf-targeted mutation exhibited lethal haploinsufficiency (Carmeliet et al., 1996; Ferrara et al., 1996). Blood vessels formed in these embryos, but were severely constricted or atretic. It is not known whether artery–vein differentiation is affected in Vegf/ embryos. However, other gain-of-function transgenic experiments have demonstrated a role for Vegf in regulating arterial endothelial cell differentiation in mice. Alternative splicing of the Vegf gene in mice results in production of several different protein isoforms (Vegf 120, Vegf 164 and Vegf188). Genetically-engineered mice expressing only the Vegf 164 isoform exhibited normal retinal vascular development. However, mice expressing only Vegf 120 exhibited severe defects in vascular outgrowth, while mice expressing only Vegf 188 exhibited impaired retinal arterial development, but normal venous and capillary development (Stalmans et al., 2002). Overexpression of the Vegf 164 isoform in cardiac muscle increased the number of ephrinB2-positive

PART  |  8  Making Vessels

capillaries in the heart, while reducing the number of EphB4-positive venules (Visconti et al., 2002). Vegf could induce ephrinB2 expression in mouse primary embryonic endothelial cells, and Vegf derived from sensory neurons, motor neurons and Schwann cells was required for arterial differentiation of small diameter nerve-associated vessels in mice (Mukouyama et al., 2002, 2005). In mice, targeted mutagenesis and transgenic studies have identified numerous Notch pathway components with a demonstrable role in embryonic vascular development. These include the receptors Notch1 (Huppert et al., 2000; Krebs et al., 2000; Limbourg et al., 2005) and Notch4 (Krebs et al., 2000; Uyttendaele et al., 2001; Carlson et al., 2005), the ligands Jag1 (Xue et al., 1999) and Dll4 (Duarte et al., 2004; Gale et al., 2004; Krebs et al., 2004), the transcriptional regulator RBP-J (Krebs et al., 2004), the E3 ubiquitin ligase Mib1 (Barsi et al., 2005; Koo et al., 2005), components of the g-secretase complex nicastrin (Li et al., 2003) and presenilin 1 and 2 (Herreman et al., 1999), and the downstream effector bHLH proteins Hey1 and Hey2 (Fischer et al., 2004; Kokubo et al., 2005). The Dll4 gene encodes the key Notch ligand required for early vascular development in mice. Surprisingly, similar to Vegf/ heterozygous embryos (Carmeliet et al., 1996; Ferrara et al., 1996), Dll4/ heterozygous embryos exhibited embryonic-lethal haploinsufficiency on certain inbred genetic backgrounds (Duarte et al., 2004; Gale et al., 2004; Krebs et al., 2004). These Dll4/ embryos died due to vascular defects. Some Dll4/ mice were viable on an outbred background, enabling the examination of Dll4/ embryos. The phenotype of the Dll4/ homozygotes was similar, although more severe, than that of the Dll4/ heterozygous embryos (Duarte et al., 2004; Gale et al., 2004; Krebs et al., 2004). Similar to that which was observed in Notch signaling-deficient zebrafish embryos, both Dll4-deficient embryos and other types of Notch signaling-deficient mouse embryos lost expression of arterial markers such as ephrinB2 (Duarte et al., 2004; Fischer et al., 2004; Gale et al., 2004; Krebs et al., 2004; Kokubo et al., 2005). Recent work has demonstrated that Dll4mediated Notch signaling induces ephrinB2 expression in cultured endothelial cells (Iso et al., 2006), and that the ephrinB2 gene is a direct Notch target gene (Grego-Bessa et al., 2007). As mentioned previously, Notch signaling-deficient zebrafish embryos exhibit arteriovenous malformations (Lawson et al., 2001, 2002). Arteries normally connect to veins only through an intervening capillary bed. An aberrant direct communication between an artery and vein is termed an arteriovenous malformation. It has been suggested that an inability to establish or maintain distinct arterial and venous vascular beds contributes to the formation of arteriovenous malformations. In mouse embryos, injection of ink into the heart is an effective way to visualize the presence of arteriovenous malformations. Notch

Chapter  |  8.3  Arteriovenous Patterning in the Vascular System

533

Figure 2  Arterio venous malformations in Notch pathway mouse mutants. (A) In a wild-type embryo, India ink injected into the heart in order to visualize blood flow exits through the aortic arch arteries and enters the descending dorsal aorta, traversing to the posterior end of the embryo. (B) In a Notch1/ embryo, on exiting the heart the injected ink flows directly into the venous system via arterio venous malformations.

signaling-deficient embryos (e.g., Dll4/ or Notch1/ embryos) exhibited gross arteriovenous malformations (Fig. 2) (Krebs et al., 2004) which were also detectable by histological analysis (Duarte et al., 2004; Gale et al., 2004; Krebs et al., 2004). It should be noted that the types of severe arteriovenous malformations exhibited by these mouse mutants contribute to early embryonic lethality, and are therefore never observed in clinical settings in humans. Such severe arteriovenous malformations in humans would result in spontaneous abortion of the fetuses at early gestational stages. Interestingly, inducible expression of an activated Notch4 transgene in adult mice resulted in vessel arterialization, such as induction of venous expression of ephrinB2, and caused arteriovenous malformations in several organs, including liver, uterus and skin (Carlson et al., 2005). Surprisingly, these malformations were reversible if activated Notch4 transgene expression was repressed. These studies demonstrate that the ability of Notch signaling to arterialize blood vessels is not confined to the embryonic period. Another major function of Notch signaling during vascular development is to suppress formation of vascular tip cells (Sainson et al., 2005; Noguera-Troise et al., 2006; Ridgway et al., 2006; Gridley, 2007; Hellstrom et al., 2007; Leslie et al., 2007; Lobov et al., 2007; Scehnet et al., 2007; Siekmann and Lawson, 2007; Suchting et al., 2007). Tip cells are specialized endothelial cells at the leading edge of vascular sprouts. Tip cells extend filopodia to sense local environmental conditions, and guide growth of vascular sprouts along gradients of Vegf protein. Dll4-mediated Notch signaling regulates tip cell numbers, filopodia extension in tip cells and the branching of angiogenic sprouts in several model systems, including human umbilical vein endothelial cells, the mouse retina and hindbrain, the zebrafish embryo and xenograft tumor models. A finding common to all of these studies is that inhibition of Notch signaling leads to increased sprouting and branching of blood vessels. The Notch pathway regulates sprouting and branching behaviors in these vessels

by influencing the differentiation, migration and proliferation of vascular tip cells; reduced Notch signaling leads to increases in tip cell numbers, filopodia extension and vessel branching. Suppression of tip cell formation and angiogenic sprouting by Notch signaling is downstream of the Vegf signal, since pharmacological or genetic manipulations that block Vegf function reduce both Dll4 expression and blood vessel sprouting. Regulation of tip cell differentiation by Notch signaling is likely independent of its role in arteriovenous specification.

II.D.  FOXC1/FOXC2 The winged helix/forkhead/Fox family of DNA-binding proteins is a large and diverse set of evolutionarilyconserved transcription factors (Kaestner et al., 2000). Mouse embryos with compound mutations of the Foxc1 and Foxc2 genes, two closely-related Fox family transcription factors, exhibited defects in vascular remodeling in the yolk sac and embryo (Kume et al., 2001). Compound Foxc1;Foxc2 mutant embryos exhibited reduced or absent expression of arterial markers, while expression of venous markers was unaffected (Seo et al., 2006). These embryos also exhibited arteriovenous malformations. The mechanism for this failure of arterial specification was likely through disrupted regulation of Dll4 transcription. The Foxc1 and Foxc2 proteins directly activate Dll4 transcription through a Foxc-binding element in the upstream region of the Dll4 gene. These results demonstrate that the Foxc proteins are key transcriptional regulators that act upstream of the Notch pathway during arteriovenous differentiation (Seo et al., 2006). It is not known if expression of the Foxc1 and Foxc2 genes is induced by Vegf signaling.

II.E.  SOXF Subgroup Genes The Sox7, Sox17 and Sox18 genes form the SoxF subgroup of Sox (Sry-related HMG box) genes (Bowles et al., 2000).

534

These genes are highly-expressed in vascular endothelial cells in both zebrafish and mice. The roles of the Sox7 and Sox18 genes during embryonic vascular development in zebrafish have been analyzed recently (Cermenati et al., 2008; Herpers et al., 2008; Pendeville et al., 2008). Knockdown of the Sox7 or Sox18 gene alone by injection of antisense morpholinos had no obvious effects on vascular development in the injected embryos. However, simultaneous knockdown of the Sox7 and Sox18 genes resulted in arteriovenous malformations with resulting loss of circulation to the posterior of the injected embryos. Endothelial cells forming the major axial vessels of the double-morphant embryos failed to acquire a correct arter­ iovenous identity, exhibiting downregulation of arterial markers accompanied by upregulation of venous markers. Analysis of the role of the SoxF subgroup genes Sox17 and Sox18 in mice revealed partially redundant roles in cardiovascular development (Sakamoto et al., 2007). Sox17-null embryos exhibited aberrant heart looping, enlarged cardinal veins and defects in formation of the anterior dorsal aorta. While Sox18-null embryos did not exhibit cardiovascular abnormalities (Pennisi et al., 2000), Sox17;Sox18 double null mutant embryos exhibited more severe defects in the formation of the anterior dorsal aorta and head microvasculature than Sox17 single mutant embryos. Some double mutant embryos exhibited aberrant differentiation of heart tube endocardial cells and defective fusion of the endocardial tube (Sakamoto et al., 2007). However, expression of arterial and venous markers was not analyzed in these mutant embryos; therefore, the state of arteriovenous specification in Sox17;Sox18 double mutant mouse embryos has not yet been determined.

II.F.  COUP-TFII Much more is understood about the acquisition of arterial identity than of venous identity. The finding that VegfNotch signaling specified arterial identity led to the suggestion that acquisition of venous identity was simply the default state of angioblast differentiation. However, recent work has shown that the COUP-TFII (also known as Nr2f2) protein, a member of the orphan nuclear receptor superfamily, regulates vein identity in an instructive manner in mice (You et al., 2005; Lin et al., 2007). The COUP-TFII protein is expressed in venous, but not arterial, endothelial cells. Analysis of loss-of-function mutants revealed that the COUP-TFII gene is required for formation of the venous endothelium. Endothelial cell-specific deletion of the COUP-TFII gene revealed that veins in the mutant embryos acquired arterial characteristics, including expression of arterial markers such as Jag1, Notch1, ephrinB2 and neuropilin1. Transgenic overexpression of the COUP-TFII protein in endothelial cells resulted in formation of arteriovenous malformations, the loss of expression of arterial

PART  |  8  Making Vessels

markers and the upregulation of the venous marker EphB4 (You et al., 2005; Lin et al., 2007). These results support a model in which COUP-TFII functions upstream of both neuropilin1 and Notch signaling instructively to regulate acquisition of the venous cell fate.

II.G.  Chemical Genetic Screens in Zebrafish: Pathways Downstream of Vascular Endothelial Growth Factor Several features of the zebrafish model system have popularized its use in chemical genetic screens for smallmolecule drugs (Chan et al., 2002; Zon and Peterson, 2005; Bayliss et al., 2006). These features include the optical transparency of zebrafish embryos, the relatively low cost (compared to mice or other vertebrates) of maintenance and their adaptability to high-throughput screening assays. These features have been put to advantage in recent screens to identify small molecules that can rescue vascular defects in mutant zebrafish embryos. As mentioned previously, zebrafish homozygous for a Hey2 (gridlock) hypomorphic mutation exhibit a malformed aorta that prevents blood circulation to the trunk and tail (Weinstein et al., 1995; Zhong et al., 2000). These defects resemble human congenital cardiovascular anomalies, such as coarctation of the aorta (Towbin and McQuinn, 1995). A chemical screen for compounds that could restore trunk and tail circulation to Hey2 homozygous mutant zebrafish embryos resulted in the identification of two small-molecule suppressors of the mutant phenotype (Peterson et al., 2004). These compounds increased Vegf mRNA expression, both in zebrafish embryos and in human umbilical vein endothelial cells. Injection of an expression plasmid of the mouse Vegf cDNA was also able to rescue the Hey2 mutant phenotype. In a more recent screen utilizing the same strategy, another class of compounds was identified that could rescue the Hey2 mutation (Hong et al., 2006). The compound identified was an inhibitor of phosphatidylinositol-3 kinase (PI3K). Other previously-characterized PI3K inhibitors, such as wortmannin, could similarly suppress the Hey2 mutant phenotype. PI3K activation results in the phosphorylation and activation of the serine/threonine protein kinase Akt. The PI3K/Akt pathway is one of the major signaling pathways activated by Vegf receptor signaling, particularly through Vegfr2 (Byrne et al., 2005; Cebe-Suarez et al., 2006; Rahimi, 2006; Shibuya, 2006). Another major effector pathway of the Vegf receptor is the phospholipase Cg/mitogen-activated protein kinase/extracellular signal-regulated kinase (PLCg/MAPK/ERK) pathway. Activated ERK protein was preferentially localized to arterial endothelial precursors. Stimulation of ERK activation shifted the artery/vein decision in favor of artery formation during zebrafish embryogenesis, while inhibition of ERK

Chapter  |  8.3  Arteriovenous Patterning in the Vascular System

had the opposite effect. This work revealed that two downstream effector pathways of Vegf signaling have opposite effects on arteriovenous specification of angioblasts. ERK signaling promoted adoption of the arterial cell fate, while PI3 kinase signaling exerted an opposite effect by blocking ERK activation (Hong et al., 2006).

II.H.  Phospholipase C Gamma A genetic zebrafish screen for mutants lacking segmental blood vessels provided further support for the model that PLCg functions downstream of Vegf, and is required for arterial vessel formation. This screen resulted in the isolation of a loss-of-function mutant of the PCg1 (Plcg1) gene (Lawson et al., 2003). Mutant embryos exhibited defects in the formation of arteries, and loss of artery-specific gene expression in the dorsal aorta. However, development of veins in these mutants was unaffected. While biochemical studies in endothelial cell lines had demonstrated that the Plcg1 protein could function as an effector of Vegf signaling (Takahashi and Shibuya, 1997; Takahashi et al., 2001), in vivo evidence that the Plcg1 gene functioned downstream of Vegf was lacking. Injection of exogenous Vegf mRNA into wild-type zebrafish embryos led to the induction of expression of both the ephrinB2 gene and the Vegfr2 gene. Induction of ephrinB2 expression required Notch signaling activity, while Vegfr2 induction was Notch-independent. Injection of exogenous Vegf mRNA into the Plcg1 mutant embryos failed to induce expression of either the Vegfr2 or ephrinB2 gene. These data provided in vivo evidence that Plcg1 function was required for both Notch-dependent and Notch-independent signaling downstream of the Vegf signal.

II.I.  TGFb Signaling, Endoglin/Acvrl1 and Hereditary Hemorrhagic Telangiectasia TGFb signaling also plays an important role in vascular development in both mice and humans. Mutations in the TGFb pathway components activin receptor-like kinase 1 (Acvrl1; also termed Alk1) and endoglin (Eng) cause two forms of the autosomal dominant vascular disease hereditary hemorrhagic telangiectasia (HHT) (Tille and Pepper, 2004; Wang, 2005; Fernandez et al., 2006). The Acvrl1 gene encodes a type I TGFb receptor serine-threonine kinase, while the Eng gene encodes a type III TGFb coreceptor. Heterozygous loss-of-function mutations of the Eng gene cause HHT1 (McAllister et al., 1994), while heterozygous loss-of-function mutations of the Acvrl1 gene cause HHT2 (Johnson et al., 1996). These diseases are characterized by the formation of telangiectases (clusters of abnormally-dilated blood vessels), typically in the skin and mucocutaneous tissues, and arteriovenous malformations in major internal organs such as the lungs, liver and brain.

535

Mice which are homozygous for Acvrl1 or Eng mutations exhibit defects in embryonic vascular remodeling, and form arteriovenous malformations (Bourdeau et al., 1999; Li et al., 1999; Arthur et al., 2000; Oh et al., 2000; Urness et al., 2000; Sorensen et al., 2003). Formation of arteriovenous malformations was more extensive and severe in Acvrl1/ embryos than in Eng/ embryos. While expression of the arterial marker ephrinB2 was downregulated in Acvrl1/ embryos, its expression was not downregulated in Eng/ embryos (Sorensen et al., 2003). Unfortunately, this was the only arterial- or venousspecific marker whose expression was assessed in these embryos. It is not clear, therefore, whether there is a general loss of arterial specification in Acvrl1/ embryos. However, analysis of the Eng/ embryos suggests that loss of arterial identity may not be essential for formation of arteriovenous malformations. Heterozygous Eng/ and Acvrl1/ mice have been assessed as potential animal models for HHT1 and HHT2, respectively. Eng/ mice with incomplete penetrance developed features of HHT1, such as telangiectases, dilated vessels, and vascular lesions in lungs, brain, liver and gastrointestinal tract (Bourdeau et al., 1999; Arthur et al., 2000; Bourdeau et al., 2001; Torsney et al., 2003). Interestingly, development of vascular lesions in these mice was strongly dependent on the genetic background, with Eng/ mice on the 129/Ola strain having the highest incidence of defects. Similarly, Acvrl1/ mice developed vascular lesions in the skin, oral cavity and internal organs such as the lungs, liver, gastrointestinal tract, spleen and brain, which resembled those observed in HHT2 patients (Srinivasan et al., 2003). Further analysis of these models should yield important insights into the onset and progression of HHT.

II.J.  In Vitro Differentiation of Endothelial Cells from Stem Cells An exciting recent development has been in vitro differentiation of endothelial cells from stem cells. Both embryonic stem (ES) cells of human (Levenberg et al., 2002; GerechtNir et al., 2003; Wang et al., 2007) and mouse (Yamashita et al., 2000; Nakagami et al., 2006; Yurugi-Kobayashi et al., 2006; Lanner et al., 2007) origin, and human multipotent adult progenitor cells (MAPCs) (Aranguren et al., 2006) have been used to generate endothelial cells in culture. These cells will have potential therapeutic uses in regenerative medicine, for example in producing engineered blood vessels for the treatment of vascular disease, or for the treatment of regional areas of ischemia. However, production and analysis of these cells are also providing novel insights into endothelial cell biology. For example, mouse ES cells cultured on collagen type IV coated dishes were used to generate Vegfr2-positive endothelial

536

cell precursors (Lanner et al., 2007). Subculture of these precursors yielded mature endothelial cells in a Vegf dosedependent manner. Arterial markers were expressed in cultures differentiated in the presence of high Vegf concentrations, while in cultures differentiated in the presence of low-to-intermediate Vegf concentrations expression of the venous marker COUP-TFII was upregulated. Another study with mouse ES cells revealed a role for the cyclic AMP pathway in synergizing with Vegf in the in vitro differentiation of arterial endothelial cells (Yurugi-Kobayashi et al., 2006).

III.  Arteriovenous patterning: role of hemodynamic flow and other environmental cues III.A.  Endothelial Cell Plasticity Revealed   by Cell Transplantation While the experiments described above clearly demonstrate that artery–vein specification occurs early in embryogenesis, prior to the initiation of blood flow, other work has shown that this specification is plastic and can be influenced by environmental cues. The plasticity of endothelial cell differentiation has been assessed by constructing quail-chick chimeras (Moyon et al., 2001). Fragments of arteries and veins from quail embryos were grafted into the coelom of host chick embryos, and the resulting chimeras were analyzed. Through embryonic day (E) 7, both arterial and venous grafts yielded quail endothelial cells capable of colonizing both arteries and veins of the chick hosts. Subsequent to E7, the grafted vessel fragments progressively lost their ability to colonize both types of vessels: arterial grafts preferentially colonized host arteries; and venous grafts preferentially colonized veins. After E11, endothelial cell plasticity was almost entirely lost. However, a role for the vessel wall in restricting plasticity was also observed. If endothelial cells were isolated from quail aorta by collagenase digestion at E11, the cells regained the ability to colonize both host arteries and veins (Moyon et al., 2001). Another study reported similar results (Othman-Hassan et al., 2001). Grafts of quail endothelial cells from either the internal carotid artery or the vena cava were performed into E4 chick embryos. Arterial endothelial cells integrated into arteries retained ephrinB2 expression, while arterial endothelial cells integrated into vessels other than arteries did not express ephrinB2. In contrast, quail venous endothelial cells became ephrinB2-positive if they integrated into arteries, but remained ephrinB2-negative if integrated into other types of vessels (Othman-Hassan et al., 2001). Both of these studies revealed that endothelial cells retain the ability to adapt to environmental cues, despite initial expression of differentiation markers specific for the arterial endothelium.

PART  |  8  Making Vessels

III.B.  Endothelial Cell Plasticity Revealed   by Disruption of Hemodynamic Flow A series of elegant experiments has demonstrated the plasticity of endothelial cell differentiation in response to changes in hemodynamic flow in the chick embryo yolk sac (le Noble et al., 2004). Careful time-lapse videomicroscopy of formation of the yolk sac vasculature during normal remodeling of the primary capillary plexus into a network of arteries and veins revealed that small caliber vessels become disconnected from the growing arterial tree. Expression of arterial markers, such as ephrinB2 and neuropilin1, are rapidly downregulated in these disconnected vessels, which are subsequently incorporated into the forming venous system of the yolk sac. Thus, the plasticity of arteriovenous identity is a normal property of blood vessel formation in the yolk sac. To examine in more detail the role of hemodynamic flow during arteriovenous differentiation of the yolk sac, physical methods were used to alter the perfusion pattern (le Noble et al., 2004) (Fig. 3). Total inhibition of blood flow to the yolk sac by destruction of the heart resulted in the formation of a primary vascular plexus, but a complete absence of vascular remodeling to form arteries and veins. In these embryos, regions of the yolk sacs expressed the ephrinB2

(A)

Normal Blood Flow Arterial markers expressed

(B)

Reversed Blood Flow Arterial markers downregulated; venous markers expressed

Figure 3  Refinement of genetic prepatterning by blood flow. In the vasculature of the chick embryo yolk sac, acquisition of arterial cell fate is plastic and can be regulated by blood flow (le Noble et al., 2004). (A) Vessels experiencing normal arterial blood flow express arterial markers. (B) Reversal of blood flow by vessel ligation and rerouting leads to rapid downregulation of arterial markers and a more gradual upregulation of venous markers.

Chapter  |  8.3  Arteriovenous Patterning in the Vascular System

gene, while other regions were ephrinB2-negative. Thus, the initiation of ephrinB2 expression does not require perfusion of the yolk sac. If blood flow was obstructed on one side of the yolk sac after formation of the vitelline artery, blood flow was rerouted. Within 24 hours of ligation, the ligated side of the yolk sac had become entirely venularized. Preexisting arterioles on the ligated side were incorporated into the newly-developing venous vasculature if they were oriented in the direction of the newly-induced venous flow. Similarly, pre-existing veins could be incorporated into expanding arteries. Arterial markers such as ephrinB2 and neuropilin1 were downregulated on the ligated side of the embryos, and venous markers such as neuropilin2 and Tie2 were upregulated. Interestingly, downregulation of arterial markers occurred more rapidly than upregulation of venous markers. For more detailed reviews of these experiments, and of the role of hemodynamic flow in refining genetically-specified arteriovenous patterning, see Eichmann et al. (2005), le Noble et al. (2005) and Jones et al. (2006). In summary, these data clearly demonstrate that, at least in the yolk sac of the chick embryo, expression of arterial markers and incorporation into a functional arterial network does not irreversibly commit endothelial cells to an arterial fate. Instead, the prespecified vascular network can be refined by hemodynamic flow, and arterial and venous cellular identities can be altered. A similar picture emerges from analysis of the development of the trunk vasculature of zebrafish embryos. During formation of the trunk vasculature (Isogai et al., 2001), a complete artery-derived primary vascular network initially forms, followed by formation of a set of secondary, veinderived sprouts (Isogai et al., 2003). Formation of this network is independent of circulatory flow. However, flow dynamics do appear to play a critical role in refining the pattern of connections between vessels that permit this vascular network to function properly. The arteriovenous identity of the trunk intersegmental vessels is not fixed until after emergence of the secondary sprouts and their connection to the primary network. Primary segments that develop robust connections to secondary segments become intersegmental veins, while primary segments that do not develop such connections become intersegmental arteries. Thus, formation of the zebrafish trunk vasculature results from a combination of genetic prepatterning of the basic vascular architecture and refinement of this prepattern by hemodynamic flow.

III.C.  Hypoxia and Oxygen Tension Hypoxia is another key regulator of arteriovenous differentiation. Depending on the type of cell analyzed, expression of genes encoding the Dll4 ligand and the Notch targets Hey1, Hey2 and Hes1 is upregulated by hypoxia (Mailhos et al., 2001; Jogi et al., 2002; Gustafsson et al., 2005; Patel et al., 2005; Diez et al., 2007). Hypoxia-inducible factor 1

537

(HIF-1) is a heterodimeric transcriptional regulator composed of HIF-1a and HIF-1b (also referred to as ARNT, for aryl hydrocarbon receptor nuclear translocator) subunits (Hickey and Simon, 2006; Semenza, 2006). HIF-1 is a master regulator of oxygen homeostasis and activates transcription of hypoxia-responsive genes by binding to a specific nucleotide sequence, termed the hypoxia response element. Transcription of the Hey1, Hey2 and Dll4 genes was controlled by both the HIF system and the Notch pathway, which acted in concert to regulate transcription of these genes (Diez et al., 2007). In a mouse embryonic endothelial precursor cell line, expression of the COUP-TFII gene was repressed by hypoxia and Notch activation. Together, these results indicate that hypoxia can drive differentiation of endothelial progenitors towards an arterial fate by activating Dll4 and Hey2 gene expression, while concomitantly suppressing expression of the COUP-TFII gene (Diez et al., 2007). The mechanism for hypoxia-induced upregulation of Notch signaling may be through direct interaction between Notch protein intracellular domains and the HIF-1a sub­unit (Gustafsson et al., 2005). In these experiments, hypoxia inhibited differentiation of myogenic cell lines, muscle satellite cells and cortical neural stem cells, and transcription of Notch target genes such as Hes1 was induced. The intracellular domain of the cleaved, activated form of the Notch1 protein bound directly to the HIF-1a subunit. This interaction enhanced the stability of the Notch1 intracellular domain, and recruited the HIF-1a subunit to Notchresponsive promoters. This work provided evidence of an unexpected direct link between Notch signaling and the hypoxic response, and provided insights into how hypoxia maintains cells in an undifferentiated state. In addition to being exposed to higher rates of hemodynamic flow and shear stress, endothelial cells lining arteries also experience higher levels of oxygen than endothelial cells lining veins. To test whether oxygen tension could be an instructive signal for arteriovenous differentiation, neonatal mice were raised in a moderately hypoxic environment, and the effects on differentiation of the developing retinal vasculature were assessed (Claxton and Fruttiger, 2005). The mouse retina possesses a number of advantages for the analysis of developmental angiogenesis (Dorrell and Friedlander, 2006; Uemura et al., 2006). Development of the vascular system of the mouse retina occurs postnatally in a highlyreproducible spatial and temporal pattern. The retinal vascular system emerges first in the region of the optic nerve head and grows radially towards the periphery. The primitive vascular plexus that forms initially is remodeled into large and small arterial and venous vessels. During these stages, the retinal vasculature is accessible both for observation and for experimental administration of exogenous agents. When mice were raised in a moderately hypoxic environment containing 10% oxygen, expression of the arteryspecific Dll4 and ephrinB2 genes was lost. However, expression of alpha-smooth muscle actin by mural cells

538

recruited to the developing retinal arteries was retained. In addition, retinal veins continued to express the angiotensin receptor-like 1 (Agtrl1, also termed Msr or Apj) gene, a vein-specific marker (Saint-Geniez et al., 2003). These results suggest that while moderate hypoxia can prevent expression of some arterial markers during development of the retinal vasculature, it does not convert arteries into veins.

IV.  Conclusions and perspectives The picture emerging from recent work on the regulation of arteriovenous differentiation in vertebrates is that a genetically-determined prepattern for the embryonic vasculature is established prior to the initiation of blood flow, and possibly prior to the formation of vessels. However, endothelial cells at this stage are not yet committed to an arterial or venous cell fate. The prepattern is plastic, and the final determination of endothelial cell fate can be influenced strongly by environmental factors such circulatory flow and hypoxia. Although we have learned much about both the genetic and environmental control of artery/vein differentiation, numerous questions still remain. It is clear, at least in zebrafish, that the Vegf/Notch pathways and the PLCg/ MAPK/ERK pathway play major roles in regulating arterial endothelial cell differentiation. However, it is not known whether these pathways function independently, or whether they communicate in some fashion. Also unknown is the extent to which certain aspects of the regulation of arteriovenous patterning in zebrafish, such as Shh regulation of Vegf expression and PLCg/MAPK/ERK pathway regulation of arterial differentiation, are conserved in mammals. An exciting area in which there will certainly be major advances in the future is differentiation of endothelial cells from various types of stem cells, and their therapeutic utilization. We can expect improvements in the efficiency of generation of endothelial cells from stem cells, and in regulating differentiation of arterial versus venous endothelial cells. Development of these cells should provide a renewable source of endothelial cells for regenerative medicine.

Acknowledgments I thank Luke Krebs for the images for Fig. 2. Work in my laboratory on Notch signaling during vascular development in mice was supported by a grant from the National Institutes of Health (NS036437).

References Adams, R.H., Wilkinson, G.A., Weiss, C., Diella, F., Gale, N.W., Deutsch, U., Risau, W., Klein, R., 1999. Roles of ephrinB ligands and EphB receptors in cardiovascular development, demarcation of arterial/venous domains, vascular morphogenesis, and sprouting angiogenesis. Genes Dev. 13, 295–306.

PART  |  8  Making Vessels

Aranguren, X.L., Luttun, A., Clavel, C., Moreno, C., Abizanda, G., Barajas, M.A., Pelacho, B., Uriz, M., Arana, M., Echavarri, A., Soriano, M., Andreu, E.J., Merino, J., Garcia-Verdugo, J.M., Verfaillie, C.M., Prosper, F., 2006. In vitro and in vivo arterial differentiation of human multipotent adult progenitor cells. Blood 109 (6), 2634–2642. Arthur, H.M., Ure, J., Smith, A.J., Renforth, G., Wilson, D.I., Torsney, E., Charlton, R., Parums, D.V., Jowett, T., Marchuk, D.A., Burn, J., Diamond, A.G., 2000. Endoglin, an ancillary TGFbeta receptor, is required for extraembryonic angiogenesis and plays a key role in heart development. Dev. Biol. 217, 42–53. Barsi, J.C., Rajendra, R., Wu, J.I., Artzt, K., 2005. Mind bomb1 is a ubiquitin ligase essential for mouse embryonic development and Notch signaling. Mech. Dev. 122, 1106–1117. Bayliss, P.E., Bellavance, K.L., Whitehead, G.G., Abrams, J.M., Aegerter, S., Robbins, H.S., Cowan, D.B., Keating, M.T., O’Reilly, T., Wood, J.M., Roberts, T.M., Chan, J., 2006. Chemical modulation of receptor signaling inhibits regenerative angiogenesis in adult zebrafish. Nat. Chem. Biol. 2, 265–273. Bourdeau, A., Dumont, D.J., Letarte, M., 1999. A murine model of hereditary hemorrhagic telangiectasia. J. Clin. Invest. 104, 1343–1351. Bourdeau, A., Faughnan, M.E., McDonald, M.L., Paterson, A.D., Wanless, I.R., Letarte, M., 2001. Potential role of modifier genes influencing transforming growth factor-beta1 levels in the development of vascular defects in endoglin heterozygous mice with hereditary hemorrhagic telangiectasia. Am. J. Pathol. 158, 2011–2020. Bowles, J., Schepers, G., Koopman, P., 2000. Phylogeny of the SOX family of developmental transcription factors based on sequence and structural indicators. Dev. Biol. 227, 239–255. Bray, S.J., 2006. Notch signalling, a simple pathway becomes complex. Nat. Rev. Mol. Cell. Biol. 7, 678–689. Byrne, A.M., Bouchier-Hayes, D.J., Harmey, J.H., 2005. Angiogenic and cell survival functions of vascular endothelial growth factor (VEGF). J. Cell. Mol. Med. 9, 777–794. Carlson, T.R., Yan, Y., Wu, X., Lam, M.T., Tang, G.L., Beverly, L.J., Messina, L.M., Capobianco, A.J., Werb, Z., Wang, R., 2005. Endothelial expression of constitutively active Notch4 elicits revers­ ible arteriovenous malformations in adult mice. Proc. Natl. Acad. Sci. U.S.A. 102, 9884–9889. Carmeliet, P., Ferreira, V., Breier, G., Pollefeyt, S., Kieckens, L., Gertsenstein, M., Fahrig, M., Vandenhoeck, A., Harpal, K., Eberhardt, C., Declercq, C., Pawling, J., Moons, L., Collen, D., Risau, W., Nagy, A., 1996. Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 380, 435–439. Cebe-Suarez, S., Zehnder-Fjallman, A., Ballmer-Hofer, K., 2006. The role of VEGF receptors in angiogenesis; complex partnerships. Cell. Mol. Life Sci. 63, 601–615. Cermenati, S., Moleri, S., Cimbro, S., Corti, P., Del Giacco, L., Amodeo, R., Dejana, E., Koopman, P., Cotelli, F., Beltrame, M., 2008. Sox18 and Sox7 play redundant roles in vascular development. Blood 111, 2657–2666. Chan, J., Bayliss, P.E., Wood, J.M., Roberts, T.M., 2002. Dissection of angiogenic signaling in zebrafish using a chemical genetic approach. Cancer Cell 1, 257–267. Claxton, S., Fruttiger, M., 2005. Oxygen modifies artery differentiation and network morphogenesis in the retinal vasculature. Dev. Dyn. 233, 822–828. Coultas, L., Chawengsaksophak, K., Rossant, J., 2005. Endothelial cells and VEGF in vascular development. Nature 438, 937–945.

Chapter  |  8.3  Arteriovenous Patterning in the Vascular System

Covassin, L.D., Villefranc, J.A., Kacergis, M.C., Weinstein, B.M., Lawson, N.D., 2006. Distinct genetic interactions between multiple Vegf receptors are required for development of different blood vessel types in zebrafish. Proc. Natl. Acad. Sci. U.S.A. 103, 6554–6559. Diez, H., Fischer, A., Winkler, A., Hu, C.J., Hatzopoulos, A.K., Breier, G., Gessler, M., 2007. Hypoxia-mediated activation of Dll4-Notch-Hey2 signaling in endothelial progenitor cells and adoption of arterial cell fate. Exp. Cell Res. 313, 1–9. Dorrell, M.I., Friedlander, M., 2006. Mechanisms of endothelial cell guidance and vascular patterning in the developing mouse retina. Prog. Retin. Eye Res. 25, 277–295. Duarte, A., Hirashima, M., Benedito, R., Trindade, A., Diniz, P., Bekman, E., Costa, L., Henrique, D., Rossant, J., 2004. Dosage-sensitive require­ment for mouse Dll4 in artery development. Genes Dev. 18, 2474–2478. Ehebauer, M., Hayward, P., Martinez-Arias, A., 2006. Notch, a universal arbiter of cell fate decisions. Science 314, 1414–1415. Eichmann, A., Yuan, L., Moyon, D., Lenoble, F., Pardanaud, L., Breant, C., 2005. Vascular development, from precursor cells to branched arterial and venous networks. Int. J. Dev. Biol. 49, 259–267. Fernandez, L.A., Sanz-Rodriguez, F., Blanco, F.J., Bernabeu, C., Botella, L.M., 2006. Hereditary hemorrhagic telangiectasia, a vascular dysplasia affecting the TGF-beta signaling pathway. Clin. Med. Res. 4, 66–78. Ferrara, N., Carver-Moore, K., Chen, H., Dowd, M., Lu, L., O’Shea, K.S., Powell-Braxton, L., Hillan, K.J., Moore, M.W., 1996. Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 380, 439–442. Fischer, A., Schumacher, N., Maier, M., Sendtner, M., Gessler, M., 2004. The Notch target genes Hey1 and Hey2 are required for embryonic vascular development. Genes Dev. 18, 901–911. Gale, N.W., Dominguez, M.G., Noguera, I., Pan, L., Hughes, V., Valenzuela, D.M., Murphy, A.J., Adams, N.C., Lin, H.C., Holash, J., Thurston, G., Yancopoulos, G.D., 2004. Haploinsufficiency of deltalike 4 ligand results in embryonic lethality due to major defects in arterial and vascular development. Proc. Natl. Acad. Sci. U.S.A. 101, 15949–15954. Gerecht-Nir, S., Ziskind, A., Cohen, S., Itskovitz-Eldor, J., 2003. Human embryonic stem cells as an in vitro model for human vascular development and the induction of vascular differentiation. Lab. Invest. 83, 1811–1820. Gerety, S.S., Wang, H.U., Chen, Z.F., Anderson, D.J., 1999. Symmetrical mutant phenotypes of the receptor EphB4 and its specific transmembrane ligand ephrin-B2 in cardiovascular development. Mol. Cell. 4, 403–414. Grego-Bessa, J., Luna-Zurita, L., Del Monte, G., Bolos, V., Melgar, P., Arandilla, A., Garratt, A.N., Zang, H., Mukouyama, Y.S., Chen, H., Shou, W., Ballestar, E., Esteller, M., Rojas, A., Perez-Pomares, J.M., de la Pompa, J.L., 2007. Notch signaling is essential for ventricular chamber development. Dev. Cell. 12, 415–429. Gridley, T., 2007. Vascular biology, vessel guidance. Nature 445, 722–723. Gustafsson, M.V., Zheng, X., Pereira, T., Gradin, K., Jin, S., Lundkvist, J., Ruas, J.L., Poellinger, L., Lendahl, U., Bondesson, M., 2005. Hypoxia requires notch signaling to maintain the undifferentiated cell state. Dev. Cell. 9, 617–628. Hellstrom, M., Phng, L.K., Hofmann, J.J., Wallgard, E., Coultas, L., Lindblom, P., Alva, J., Nilsson, A.K., Karlsson, L., Gaiano, N., Yoon, K., Rossant, J., Iruela-Arispe, M.L., Kalen, M., Gerhardt, H., Betsholtz, C., 2007. Dll4 signalling through Notch1 regulates formation of tip cells during angiogenesis. Nature 445, 776–780.

539

Herpers, R., van de Kamp, E., Duckers, H.J., Schulte-Merker, S., 2008. Redundant roles for sox7 and sox18 in arteriovenous specification in zebrafish. Circ. Res. 102, 12–15. Herreman, A., Hartmann, D., Annaert, W., Saftig, P., Craessaerts, K., Serneels, L., Umans, L., Schrijvers, V., Checler, F., Vanderstichele, H., Baekelandt, V., Dressel, R., Cupers, P., Huylebroeck, D., Zwijsen, A., Van Leuven, F., De Strooper, B., 1999. Presenilin 2 deficiency causes a mild pulmonary phenotype and no changes in amyloid precursor protein processing but enhances the embryonic lethal phenotype of presenilin 1 deficiency. Proc. Natl. Acad. Sci. U.S.A. 96, 11872–11877. Herzog, Y., Guttmann-Raviv, N., Neufeld, G., 2005. Segregation of arterial and venous markers in subpopulations of blood islands before vessel formation. Dev. Dyn. 232, 1047–1055. Hickey, M.M., Simon, M.C., 2006. Regulation of angiogenesis by hypoxia and hypoxia-inducible factors. Curr. Top. Dev. Biol. 76, 217–257. Hong, C.C., Peterson, Q.P., Hong, J.Y., Peterson, R.T., 2006. Artery/vein specification is governed by opposing phosphatidylinositol-3 kinase and MAP kinase/ERK signaling. Curr. Biol. 16, 1366–1372. Huppert, S.S., Le, A., Schroeter, E.H., Mumm, J.S., Saxena, M.T., Milner, L.A., Kopan, R., 2000. Embryonic lethality in mice homozygous for a processing-deficient allele of Notch1. Nature 405, 966–970. Iso, T., Maeno, T., Oike, Y., Yamazaki, M., Doi, H., Arai, M., Kurabayashi, M., 2006. Dll4-selective Notch signaling induces ephrinB2 gene expression in endothelial cells. Biochem. Biophys. Res. Commun. 341, 708–714. Isogai, S., Horiguchi, M., Weinstein, B.M., 2001. The vascular anatomy of the developing zebrafish, an atlas of embryonic and early larval development. Dev. Biol. 230, 278–301. Isogai, S., Lawson, N.D., Torrealday, S., Horiguchi, M., Weinstein, B.M., 2003. Angiogenic network formation in the developing vertebrate trunk. Development 130, 5281–5290. Jogi, A., Ora, I., Nilsson, H., Lindeheim, A., Makino, Y., Poellinger, L., Axelson, H., Pahlman, S., 2002. Hypoxia alters gene expression in human neuroblastoma cells toward an immature and neural crest-like phenotype. Proc. Natl. Acad. Sci. U.S.A. 99, 7021–7026. Johnson, D.W., Berg, J.N., Baldwin, M.A., Gallione, C.J., Marondel, I., Yoon, S.J., Stenzel, T.T., Speer, M., Pericak-Vance, M.A., Diamond, A., Guttmacher, A.E., Jackson, C.E., Attisano, L., Kucherlapati, R., Porteous, M.E., Marchuk, D.A., 1996. Mutations in the activin receptor-like kinase 1 gene in hereditary haemorrhagic telangiectasia type 2. Nat. Genet. 13, 189–195. Jones, E.A., le Noble, F., Eichmann, A., 2006. What determines blood vessel structure? Genetic prespecification vs. hemodynamics. Physiology 21, 395–399. Kaestner, K.H., Knochel, W., Martinez, D.E., 2000. Unified nomenclature for the winged helix/forkhead transcription factors. Genes Dev. 14, 142–146. Kokubo, H., Miyagawa-Tomita, S., Nakazawa, M., Saga, Y., Johnson, R.L., 2005. Mouse hesr1 and hesr2 genes are redundantly required to mediate Notch signaling in the developing cardiovascular system. Dev. Biol. 278, 301–309. Koo, B.K., Lim, H.S., Song, R., Yoon, M.J., Yoon, K.J., Moon, J.S., Kim, Y.W., Kwon, M.C., Yoo, K.W., Kong, M.P., Lee, J., Chitnis, A.B., Kim, C.H., Kong, Y.Y., 2005. Mind bomb 1 is essential for generating functional Notch ligands to activate Notch. Development 132, 3459–3470. Krebs, L.T., Shutter, J.R., Tanigaki, K., Honjo, T., Stark, K.L., Gridley, T., 2004. Haploinsufficient lethality and formation of arteriovenous malformations in Notch pathway mutants. Genes Dev. 18, 2469–2473.

540

Krebs, L.T., Xue, Y., Norton, C.R., Shutter, J.R., Maguire, M., Sundberg, J.P., Gallahan, D., Closson, V., Kitajewski, J., Callahan, R., Smith, G.H., Stark, K.L., Gridley, T., 2000. Notch signaling is essential for vascular morphogenesis in mice. Genes Dev. 14, 1343–1352. Kume, T., Jiang, H., Topczewska, J.M., Hogan, B.L., 2001. The murine winged helix transcription factors, Foxc1 and Foxc2, are both required for cardiovascular development and somitogenesis. Genes Dev. 15, 2470–2482. Lanner, F., Sohl, M., Farnebo, F., 2007. Functional arterial and venous fate is determined by graded VEGF signaling and notch status during embryonic stem cell differentiation. Arterioscler. Thromb. Vasc. Biol. 27, 487–493. Lawson, N.D., Scheer, N., Pham, V.N., Kim, C.H., Chitnis, A.B., Campos-Ortega, J.A., Weinstein, B.M., 2001. Notch signaling is required for arterial–venous differentiation during embryonic vascular development. Development 128, 3675–3683. Lawson, N.D., Vogel, A.M., Weinstein, B.M., 2002. Sonic hedgehog and vascular endothelial growth factor act upstream of the Notch pathway during arterial endothelial differentiation. Dev. Cell 3, 127–136. Lawson, N.D., Mugford, J.W., Diamond, B.A., Weinstein, B.M., 2003. Phospholipase C gamma-1 is required downstream of vascular endothelial growth factor during arterial development. Genes Dev. 17, 1346–1351. le Noble, F., Moyon, D., Pardanaud, L., Yuan, L., Djonov, V., Matthijsen, R., Breant, C., Fleury, V., Eichmann, A., 2004. Flow regulates arterial– venous differentiation in the chick embryo yolk sac. Development 131, 361–375. le Noble, F., Fleury, V., Pries, A., Corvol, P., Eichmann, A., Reneman, R.S., 2005. Control of arterial branching morphogenesis in embryogenesis, go with the flow. Cardiovasc. Res. 65, 619–628. Leslie, J.D., Ariza-McNaughton, L., Bermange, A.L., McAdow, R., Johnson, S.L., Lewis, J., 2007. Endothelial signalling by the Notch ligand Delta-like 4 restricts angiogenesis. Development 134, 839–844. Levenberg, S., Golub, J.S., Amit, M., Itskovitz-Eldor, J., Langer, R., 2002. Endothelial cells derived from human embryonic stem cells. Proc. Natl. Acad. Sci. U.S.A. 99, 4391–4396. Li, D.Y., Sorensen, L.K., Brooke, B.S., Urness, L.D., Davis, E.C., Taylor, D.G., Boak, B.B., Wendel, D.P., 1999. Defective angiogenesis in mice lacking endoglin. Science 284, 1534–1537. Li, T., Ma, G., Cai, H., Price, D.L., Wong, P.C., 2003. Nicastrin is required for assembly of presenilin/gamma-secretase complexes to mediate Notch signaling and for processing and trafficking of betaamyloid precursor protein in mammals. J. Neurosci. 23, 3272–3277. Limbourg, F.P., Takeshita, K., Radtke, F., Bronson, R.T., Chin, M.T., Liao, J.K., 2005. Essential role of endothelial Notch1 in angiogenesis. Circulation 111, 1826–1832. Lin, F.J., Tsai, M.J., Tsai, S.Y., 2007. Artery and vein formation, a tug of war between different forces. EMBO Rep. 8, 920–924. Liu, Z.J., Shirakawa, T., Li, Y., Soma, A., Oka, M., Dotto, G.P., Fairman, R.M., Velazquez, O.C., Herlyn, M., 2003. Regulation of Notch1 and Dll4 by vascular endothelial growth factor in arterial endothelial cells, implications for modulating arteriogenesis and angiogenesis. Mol. Cell. Biol. 23, 14–25. Lobov, I.B., Renard, R.A., Papadopoulos, N., Gale, N.W., Thurston, G., Yancopoulos, G.D., Wiegand, S.J., 2007. Delta-like ligand 4 (Dll4) is induced by VEGF as a negative regulator of angiogenic sprouting. Proc. Natl. Acad. Sci. U.S.A. 104 (9), 3219–3224. Mailhos, C., Modlich, U., Lewis, J., Harris, A., Bicknell, R., IshHorowicz, D., 2001. Delta4, an endothelial specific notch ligand

PART  |  8  Making Vessels

expressed at sites of physiological and tumor angiogenesis. Diff­ erentiation 69, 135–144. McAllister, K.A., Grogg, K.M., Johnson, D.W., Gallione, C.J., Baldwin, M.A., Jackson, C.E., Helmbold, E.A., Markel, D.S., McKinnon, W.C., Murrell, J., et al., 1994. Endoglin, a TGF-beta binding protein of endothelial cells, is the gene for hereditary haemorrhagic telangiectasia type 1. Nat. Genet. 8, 345–351. Moyon, D., Pardanaud, L., Yuan, L., Breant, C., Eichmann, A., 2001. Plasticity of endothelial cells during arterial–venous differentiation in the avian embryo. Development 128, 3359–3370. Mukouyama, Y.S., Shin, D., Britsch, S., Taniguchi, M., Anderson, D.J., 2002. Sensory nerves determine the pattern of arterial differentiation and blood vessel branching in the skin. Cell 109, 693–705. Mukouyama, Y.S., Gerber, H.P., Ferrara, N., Gu, C., Anderson, D.J., 2005. Peripheral nerve-derived VEGF promotes arterial differentiation via neuropilin 1-mediated positive feedback. Development 132, 941–952. Nakagami, H., Nakagawa, N., Takeya, Y., Kashiwagi, K., Ishida, C., Hayashi, S., Aoki, M., Matsumoto, K., Nakamura, T., Ogihara, T., Morishita, R., 2006. Model of vasculogenesis from embryonic stem cells for vascular research and regenerative medicine. Hypertension 48, 112–119. Nicoli, S., Tobia, C., Gualandi, L., De Sena, G., Presta, M., 2008. Calcitonin receptor-like receptor guides arterial differentiation in Zebrafish. Blood 111 (10), 4965–4972. Noguera-Troise, I., Daly, C., Papadopoulos, N.J., Coetzee, S., Boland, P., Gale, N.W., Lin, H.C., Yancopoulos, G.D., Thurston, G., 2006. Blockade of Dll4 inhibits tumour growth by promoting non-productive angiogenesis. Nature 444, 1032–1037. Oh, S.P., Seki, T., Goss, K.A., Imamura, T., Yi, Y., Donahoe, P.K., Li, L., Miyazono, K., ten Dijke, P., Kim, S., Li, E., 2000. Activin receptorlike kinase 1 modulates transforming growth factor-beta 1 signaling in the regulation of angiogenesis. Proc. Natl. Acad. Sci. U.S.A. 97, 2626–2631. Othman-Hassan, K., Patel, K., Papoutsi, M., Rodriguez-Niedenfuhr, M., Christ, B., Wilting, J., 2001. Arterial identity of endothelial cells is controlled by local cues. Dev. Biol. 237, 398–409. Pasquale, E.B., 2005. Eph receptor signalling casts a wide net on cell behaviour. Nat. Rev. Mol. Cell. Biol. 6, 462–475. Patel, N.S., Li, J.L., Generali, D., Poulsom, R., Cranston, D.W., Harris, A.L., 2005. Up-regulation of delta-like 4 ligand in human tumor vasculature and the role of basal expression in endothelial cell function. Cancer Res. 65, 8690–8697. Pendeville, H., Winandy, M., Manfroid, I., Nivelles, O., Motte, P., Pasque, V., Peers, B., Struman, I., Martial J. A., Voz, M.L, 2008. Zebrafish Sox7 and Sox18 function together to control arterial–venous identity. Dev. Biol. 317 (2), 405–416. Pennisi, D., Bowles, J., Nagy, A., Muscat, G., Koopman, P., 2000. Mice null for sox18 are viable and display a mild coat defect. Mol. Cell. Biol. 20, 9331–9336. Peterson, R.T., Shaw, S.Y., Peterson, T.A., Milan, D.J., Zhong, T.P., Schreiber, S.L., MacRae, C.A., Fishman, M.C., 2004. Chemical suppression of a genetic mutation in a zebrafish model of aortic coarctation. Nat. Biotechnol. 22, 595–599. Poliakov, A., Cotrina, M., Wilkinson, D.G., 2004. Diverse roles of eph receptors and ephrins in the regulation of cell migration and tissue assembly. Dev. Cell 7, 465–480. Rahimi, N., 2006. Vascular endothelial growth factor receptors, molecular mechanisms of activation and therapeutic potentials. Exp. Eye Res. 83, 1005–1016.

Chapter  |  8.3  Arteriovenous Patterning in the Vascular System

Ridgway, J., Zhang, G., Wu, Y., Stawicki, S., Liang, W.C., Chanthery, Y., Kowalski, J., Watts, R.J., Callahan, C., Kasman, I., Singh, M., Chien, M., Tan, C., Hongo, J.A., de Sauvage, F., Plowman, G., Yan, M., 2006. Inhibition of Dll4 signalling inhibits tumour growth by deregulating angiogenesis. Nature 444, 1083–1087. Sainson, R.C., Aoto, J., Nakatsu, M.N., Holderfield, M., Conn, E., Koller, E., Hughes, C.C., 2005. Cell-autonomous notch signaling regulates endothelial cell branching and proliferation during vascular tubulogenesis. Faseb J. 19, 1027–1029. Saint-Geniez, M., Argence, C.B., Knibiehler, B., Audigier, Y., 2003. The msr/apj gene encoding the apelin receptor is an early and specific marker of the venous phenotype in the retinal vasculature. Gene Expr. Patterns 3, 467–472. Sakamoto, Y., Hara, K., Kanai-Azuma, M., Matsui, T., Miura, Y., Tsunekawa, N., Kurohmaru, M., Saijoh, Y., Koopman, P., Kanai, Y., 2007. Redundant roles of Sox17 and Sox18 in early cardiovascular development of mouse embryos. Biochem. Biophys. Res. Commun. 360, 539–544. Scehnet, J.S., Jiang, W., Kumar, S.R., Krasnoperov, V., Trindade, A., Benedito, R., Djokovic, D., Borges, C., Ley, E.J., Duarte, A., Gill, P.S., 2007. Inhibition of Dll4 mediated signaling induces proliferation of immature vessels and results in poor tissue perfusion. Blood 109 (11), 4753–4760. Semenza, G.L., 2006. Regulation of physiological responses to continuous and intermittent hypoxia by hypoxia-inducible factor 1. Exp. Physiol. 91, 803–806. Seo, S., Fujita, H., Nakano, A., Kang, M., Duarte, A., Kume, T., 2006. The forkhead transcription factors, Foxc1 and Foxc2, are required for arterial specification and lymphatic sprouting during vascular development. Dev. Biol. 294, 458–470. Shibuya, M., 2006. Differential roles of vascular endothelial growth factor receptor-1 and receptor-2 in angiogenesis. J. Biochem. Mol. Biol. 39, 469–478. Shibuya, M., Claesson-Welsh, L., 2006. Signal transduction by VEGF receptors in regulation of angiogenesis and lymphangiogenesis. Exp. Cell Res. 312, 549–560. Siekmann, A.F., Lawson, N.D., 2007. Notch signalling limits angiogenic cell behaviour in developing zebrafish arteries. Nature 445, 781–784. Sorensen, L.K., Brooke, B.S., Li, D.Y., Urness, L.D., 2003. Loss of distinct arterial and venous boundaries in mice lacking endoglin, a vascular-specific TGFbeta coreceptor. Dev. Biol. 261, 235–250. Srinivasan, S., Hanes, M.A., Dickens, T., Porteous, M.E., Oh, S.P., Hale, L.P., Marchuk, D.A., 2003. A mouse model for hereditary hemorrhagic telangiectasia (HHT) type 2. Hum. Mol. Genet. 12, 473–482. Stalmans, I., Ng, Y.S., Rohan, R., Fruttiger, M., Bouche, A., Yuce, A., Fujisawa, H., Hermans, B., Shani, M., Jansen, S., Hicklin, D., Anderson, D.J., Gardiner, T., Hammes, H.P., Moons, L., Dewerchin, M., Collen, D., Carmeliet, P., D’Amore, P.A., 2002. Arteriolar and venular patterning in retinas of mice selectively expressing VEGF isoforms. J. Clin. Invest. 109, 327–336. Suchting, S., Freitas, C., le Noble, F., Benedito, R., Breant, C., Duarte, A., Eichmann, A., 2007. The Notch ligand Delta-like 4 negatively regulates endothelial tip cell formation and vessel branching. Proc. Natl. Acad. Sci. U.S.A. 104 (9), 3225–3230. Takahashi, T., Shibuya, M., 1997. The 230 kDa mature form of KDR/Flk-1 (VEGF receptor-2) activates the PLC-gamma pathway and partially induces mitotic signals in NIH3T3 fibroblasts. Oncogene 14, 2079–2089. Takahashi, T., Yamaguchi, S., Chida, K., Shibuya, M., 2001. A single autophosphorylation site on KDR/Flk-1 is essential for

541

VEGF-A-dependent activation of PLC-gamma and DNA synthesis in vascular endothelial cells. Embo. J. 20, 2768–2778. Tille, J.C., Pepper, M.S., 2004. Hereditary vascular anomalies, new insights into their pathogenesis. Arterioscler. Thromb. Vasc. Biol. 24, 1578–1590. Torsney, E., Charlton, R., Diamond, A.G., Burn, J., Soames, J.V., Arthur, H.M., 2003. Mouse model for hereditary hemorrhagic telangiectasia has a generalized vascular abnormality. Circulation 107, 1653–1657. Towbin, J.A., McQuinn, T.C., 1995. Gridlock, a model for coarctation of the aorta? Nat. Med. 1, 1141–1142. Uemura, A., Kusuhara, S., Katsuta, H., Nishikawa, S., 2006. Angiogenesis in the mouse retina, a model system for experimental manipulation. Exp. Cell Res. 312, 676–683. Urness, L.D., Sorensen, L.K., Li, D.Y., 2000. Arteriovenous malformations in mice lacking activin receptor-like kinase-1. Nat. Genet. 26, 328–331. Uyttendaele, H., Ho, J., Rossant, J., Kitajewski, J., 2001. Vascular patterning defects associated with expression of activated Notch4 in embryonic endothelium. Proc. Natl. Acad. Sci. U.S.A. 98, 5643–5648. Visconti, R.P., Richardson, C.D., Sato, T.N., 2002. Orchestration of angiogenesis and arteriovenous contribution by angiopoietins and vascular endothelial growth factor (VEGF). Proc. Natl. Acad. Sci. U.S.A. 99, 8219–8224. Wang, H.U., Chen, Z.F., Anderson, D.J., 1998. Molecular distinction and angiogenic interaction between embryonic arteries and veins revealed by ephrin-B2 and its receptor Eph-B4. Cell 93, 741–753. Wang, Q.K., 2005. Update on the molecular genetics of vascular anomalies. Lymphat. Res. Biol. 3, 226–233. Wang, Z.Z., Au, P., Chen, T., Shao, Y., Daheron, L.M., Bai, H., Arzigian, M., Fukumura, D., Jain, R.K., Scadden, D.T., 2007. Endothelial cells derived from human embryonic stem cells form durable blood vessels in vivo. Nat. Biotechnol. 25, 317–318. Weinstein, B.M., Stemple, D.L., Driever, W., Fishman, M.C., 1995. Gridlock, a localized heritable vascular patterning defect in the zebrafish. Nat. Med. 1, 1143–1147. Xue, Y., Gao, X., Lindsell, C.E., Norton, C.R., Chang, B., Hicks, C., Gendron-Maguire, M., Rand, E.B., Weinmaster, G., Gridley, T., 1999. Embryonic lethality and vascular defects in mice lacking the Notch ligand Jagged1. Hum. Mol. Genet. 8, 723–730. Yamashita, J., Itoh, H., Hirashima, M., Ogawa, M., Nishikawa, S., Yurugi, T., Naito, M., Nakao, K., Nishikawa, S., 2000. Flk1-positive cells derived from embryonic stem cells serve as vascular progenitors. Nature 408, 92–96. You, L.R., Lin, F.J., Lee, C.T., DeMayo, F.J., Tsai, M.J., Tsai, S.Y., 2005. Suppression of Notch signalling by the COUP-TFII transcription factor regulates vein identity. Nature 435, 98–104. Yurugi-Kobayashi, T., Itoh, H., Schroeder, T., Nakano, A., Narazaki, G., Kita, F., Yanagi, K., Hiraoka-Kanie, M., Inoue, E., Ara, T., Nagasawa, T., Just, U., Nakao, K., Nishikawa, S., Yamashita, J.K., 2006. Adrenomedullin/cyclic AMP pathway induces Notch activation and differentiation of arterial endothelial cells from vascular progenitors. Arterioscler. Thromb. Vasc. Biol. 26, 1977–1984. Zhong, T.P., Rosenberg, M., Mohideen, M.A., Weinstein, B., Fishman, M.C., 2000. Gridlock, an HLH gene required for assembly of the aorta in zebrafish. Science 287, 1820–1824. Zhong, T.P., Childs, S., Leu, J.P., Fishman, M.C., 2001. Gridlock signalling pathway fashions the first embryonic artery. Nature 414, 216–220. Zon, L.I., Peterson, R.T., 2005. In vivo drug discovery in the zebrafish. Nat. Rev. Drug Discov. 4, 35–44.

Chapter 8.4

Lymphatic Vascular Development Natasha L. Harvey Division of Haematology, Centre for Cancer Biology, SA Pathology, Adelaide, South Australia

I.  Introduction Lymphatic vessels are a vital but often overlooked component of the cardiovascular system. In contrast to blood vessels, lymphatic vessels do not deliver oxygen and nutrients to tissues, but instead collect and return interstitial fluid and protein (lymph) to the bloodstream. In addition, lymphatic vessels provide an important trafficking route for cells of the immune system during immune surveillance and infection, and facilitate the absorption of lipids from the digestive tract. Lymphatic vascular function is critical for both embryonic development and adult homeostasis, reflected by the fact that abnormalities in the growth and development of lymphatic vessels (lymphangiogenesis) are associated with an ever-expanding catalog of human pathologies. Defects in embryonic lymphangiogenesis that result in dysfunctional lymphatic vessels are associated with congenital lymphoedema syndromes, as well as Down, Noonan’s and Turner syndromes. It is likely that the most severe disturbances in embryonic lymphatic vascular development are incompatible with life. Aberrant postnatal lymphangiogenesis has recently been associated with inflammatory pathologies including graft rejection, asthma, psoriasis and arthritis, while the stimulation of lymphangiogenesis by tumors has been demonstrated to promote tumor metastasis in mouse models and has been correlated with poor patient prognosis in several types of human cancers. A major focus of lymphatic vascular research is to delineate the mechanisms by which the lymphatic vasculature is constructed, in order to identify opportunities to intervene in this process and thereby develop better treatments of lymphatic vascular diseases. This chapter will focus on what we currently know about the events that initiate and control construction of the lymphatic vasculature during embryonic development, and how these events are recapitulated or go wrong in disease processes. Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

II.  Lymphatic vessels: a historical perspective Definitive references to the description of lymphatic vessels date back to the ancient Greeks around 300 bc (Lord, 1968). Interestingly, it appears that the lipid transport properties of lymphatic vessels aided their recognition as a distinct vascular entity; mesenteric lymphatic vessels in suckling young were illuminated by the milk they carried. This feature of lymphatic vessels no doubt highlighted their existence at a much earlier time in the course of history than otherwise would have occurred. Despite these early descriptions, further studies characterizing lymphatic vascular anatomy and function were not documented until the 1500s (Lord, 1968). While the major lymphatic collecting and transport duct of the lymphatic vasculature, the thoracic duct, was discovered and described to join the subclavian vein by Eustachius in 1563, it was not until later work by a number of anatomists in the 1650s that the lymphatic vasculature was demonstrated to be comprised of a connected system of vessels; the lacteals, cisterna chyli and the thoracic duct (Fig. 1). Peripheral lymphatic vessels carrying transparent fluid or serum were subsequently described in many organs, with injection techniques being instrumental in illustrating the connectivity of an entire lymphatic vascular circulatory system (Lord, 1968). The first indication of the function of this specialized vascular network was described in 1622, by Italian anatomist Gaspar Aselli (Asellius, 1627). Aselli recognized the role of lymphatic vessels in the absorption of lipids when he was studying a dog that had consumed a fatty meal and thereby displayed illuminated, lipid-filled intestinal lymphatic vessels. By the late eighteenth century, the anatomy and absorptive nature of the lymphatic vasculature, together with the fact that it comprised a distinct network to the blood vasculature, was firmlyestablished, largely as a result of the work of William Hunter and his students William Cruikshank and William Hewson.

543

544

PART  |  8  Making Vessels

Figure 1  Drawings of the human intestinal lymphatic vessels, the lacteals (L), by Joannis Vesling, 1647 (A) and of the visceral lymphatic vasculature from the lower limbs to the thoracic duct, by Olof Rudbeck, 1659 (B). Reproduced from Lord (1968), with permission from Wellcome Library, London.

Comparative studies of lymphatic ­ vascular anatomy in fish, birds and mammals were documented as early as 1774 by William Hewson (Lord, 1968; Yoffey and Courtice, 1970), who documented the presence and location of intestinal lymphatic vessels (the lacteals), as well as the thoracic duct in all of these animal models. In addition, Hewson is credited with noting that lymph nodes, the organizing centers of immune response, were absent in fish and the turtle, limited in number in birds, and prominent in number in mammals, reflecting the increase in complexity of the immune system across phyla. This groundwork set the scene for later investigations into the mechanism by which the lymphatic vasculature is generated during embryogenesis.

III.  A comparison of lymphatic vessel and blood vessel architecture The lymphatic and blood vascular networks are exquisitely designed to fulfill distinct functional roles, and consequently exhibit distinct structural features that enable them to be distinguished from one another (see Chapters 8.2 and 8.3). The lymphatic vascular network is comprised of superficial plexi of lymphatic capillaries that absorb and channel lymph to progressively larger collecting vessels (Fig. 2). The thoracic duct is the penultimate vessel through which lymph passes before joining the bloodstream, this occurs at the junction of the thoracic duct with the great veins of the neck. Until very recently,

Figure 2  Drawing of the superficial lymphatic vessels of the human torso draining to local lymph nodes, by Sappey (1874). Reproduced from Suami (Suami et al., 2008), with permission of Springer Science and Business Media.

l­ymphatic vessels and blood vessels were ­distinguished from one another purely on the basis of morphology. Beside the fact that lymphatic vessels do not carry red blood cells, several additional differences in vessel morphology exist.

Chapter  |  8.4  Lymphatic Vascular Development

III.A.  Lymphatic versus Blood Capillaries In terms of vessel caliber, lymphatic capillaries tend to app­ ear more dilated than blood capillaries, and than venules which in turn are of greater caliber than arterioles (Figs 3; 4). The structure of lymphatic capillaries is perfectly suited to their absorptive nature; lymphatic capillaries are comprised of a continuous layer of attenuated endothelial cells that loosely interlock by virtue of long, overlapping endothelial junctions, or “inlet valves” (Casley-Smith, 1962) (Fig. 5). Lymphatic capillaries are anchored to the connective tissue by specialized filaments that extend from the abluminal surface of endothelial cells into the connective tissue, a feature that allows the capillaries to respond to changes in interstitial pressure by opening overlapping endothelial junctions to

Figure 3  Comparative structure of arterioles, venules and lymphatic capillaries. Arterial (red) and venous (blue) endothelial cells are surrounded by supporting mural cells including pericytes and smooth muscle cells (white), as well as extracellular matrix (purple). In contrast, lymphatic capillaries are not associated with pericytes or smooth muscle cells, but are invested with anchoring filaments that facilitate the opening of overlapping endothelial cell junctions and thereby influx of lymph to the capillaries. Lymphatic capillaries express very low levels of extracellular matrix in comparison to blood vascular counterparts. While blood vessels traffic both red blood cells and leukocytes, lymphatic vessels are devoid of red blood cells.

545

allow the influx of interstitial fluid and protein (Pullinger and Florey, 1935; Leak and Burke, 1968; Trzewik et al., 2001). Interestingly, leukocytes have been observed to transmigrate across the lymphatic endothelium at the sites of overlapping intercellular junctions (Leak, 1980; Baluk et al., 2007), although whether they are also able to transmigrate directly across endothelial cells has not been explored to date. In the case of leukocyte transmigration across blood vascular endothelium, both junctional and transcellular migration have been observed (Carman and Springer, 2008). Two interesting features of lymphatic endothelial cells described in elegant electron microscopic studies performed by Leak and colleagues, are the presence of fine

Figure 5  Electron micrographs depicting overlapping endothelial junctions between adjacent lymphatic endothelial cells of guinea pig dermal lymphatic capillaries. Black arrows indicate points of close apposition between endothelial cell membranes, while asterisks indicate variability in the spacing between overlapping cells. White arrows indicate membrane invaginations. Reproduced from Leak (1971), with permission.

Figure 4  Lymphatic vessels and blood vessels in embryonic skin. Lymphatic vessels express the transcription factor Prox1 in their nuclei (A: blue; B: red) and are devoid of smooth muscle that surrounds arterioles and venules (at a high level) and blood capillaries (at a low level) (A). Blood vessels do not express Prox1 (A, B). Lymphatic vessels express a lower level of CD31 than do blood vessels (A) and also express neuropilin 2 (B, C). Lymphatic capillaries are of a larger caliber than blood capillaries (A–C). Both lymphatic and blood vascular networks display a similar arborized pattern as the peripheral nervous system (C: red).

546

cytoplasmic filaments (Leak and Burke, 1968) and of pinocytotic vesicles (Leak, 1971). Pinocytotic vesicles perhaps represent an additional mechanism by which lymphatic endothelial cells transport interstitial fluid and protein across the endothelial barrier, in addition to influx through the overlapping endothelial cell junctions. The relative contribution of each of these transport mechanisms to total tissue fluid homeostasis is currently unexplored, although studies by Casley-Smith (1962) have demonstrated that lipid absorption across the lymphatic endothelium of the intestinal villi is mediated both through intercellular junctions, and via the uptake of chylomicrons and lipoproteins into endothelial caveolae and large vacuoles. It is likely that the major proportion of lymph influx is mediated via the opening of overlapping, intercellular junctions. As might be expected, transport across the blood and lymphatic vascular endothelium generally occurs in opposing directions; in the case of the blood vascular endothelium, movement usually proceeds from the lumen towards the abluminal tissue, while transport across the lymphatic vascular endothelium usually occurs from the interstitium towards the lumen. Whether different transport machinery is employed in each circumstance is currently unexplored. In contrast to blood capillaries, lymphatic capillaries have very little surrounding basement membrane, extracellular matrix, or pericyte support, a structural attribute that aids their absorptive function. Lymphatic endothelial cells are also devoid of fenestrae (Casley-Smith, 1987), specialized endothelial cell pores that are involved in filtration and are a feature of specialized blood vascular endothelial cells.

III.B.  Lymphatic Collectors Collecting lymphatic vessels are usually situated more viscerally than the superficial lymphatic capillaries, and are invested with more extracellular support in the form of pericytes/smooth muscle cells together with basement membrane (Yoffey and Courtice, 1970) (Fig. 6). The innervated smooth muscle investment of these larger collecting vessels facilitates rhythmic vessel contraction, acting to propel lymph onwards in its return to the bloodstream. Lymphatic collectors, like veins, also have valves that act to prevent lymph backflow (Fig. 7). All of these features assist in the directional propulsion of lymph from the periphery to the viscera. While lymphatic collectors are invested with smooth muscle, the level of investment does not approach that seen in veins and arteries which carry fluid under much greater pressure. As described below in further detail, the nature of intercellular junctions within the collecting lymphatic vessels differs from that found in the lymphatic capillaries; more extensive intercellular junctions, together with a greater level of extracellular support account for the transport, rather than absorptive, role of the collecting lymphatic vessels.

PART  |  8  Making Vessels

(A)

(B)

Capillary plexus

Flow

(C)

Collecting vessel

(D)

Figure 6  Lymphatic capillaries versus collecting lymphatic vessels. In contrast to lymphatic capillaries, collecting lymphatic vessels are invested with pericellular support in the form of pericytes/smooth muscle cells, as well as valves, to prevent lymph backflow (A). Recent work from Baluk and colleagues has demonstrated that the endothelial cells comprising lymphatic capillaries are different in shape (oakleaf) compared to those in collecting vessels (more columnar) (B) and display a different pattern of endothelial cell junctions; “buttons” (C) versus “zippers” (D). These features contribute to the absorptive versus transport nature of lymphatic capillaries and collecting lymphatic vessels. Red lines in (C) and (D) indicate distribution of VE-Cadherin in lymphatic endothelial junctions between capillary endothelial cells (buttons: C) and collecting endothelial cells (zippers: D). Based on figures from Baluk et al. (2007).

III.C.  Junctions between Lymphatic Endothelial Cells The junctions between endothelial cells act not only as sites of cell–cell adhesion to maintain vascular ­integrity, but also play important roles in events such as vessel growth, signaling and vascular homeostasis (Dejana, 2004). Several types of intercellular junctions are found between endothelial cells. These include tight junctions and adherens junctions, which are common to blood vascular endothelial cells, as well as complexus adhaerentes, a unique type of endothelial cell junction that comprises proteins found in both tight and adherens junctions which is more prominent in lymphatic endothelial cells (Dejana, 2004). In addition, like blood vascular endothelial cells, lymphatic endothelial cells express platelet endothelial cell adhesion molecule (PECAM), a cell surface ­ receptor that mediates homophilic cell adhesion,

547

Chapter  |  8.4  Lymphatic Vascular Development

Figure 7  Electron micrographs depicting valves in collecting lymphatic vessels. Arrows indicate endothelial cell folds projecting into the lymphatic vessel lumen (L). Reproduced from Leak (1980), with permission.

although the level of PECAM expressed on lymphatic endothelial cells is much lower than that in veins and arteries, respectively (Fig. 4). An interesting feature of isolated blood and lymphatic endothelial cells, despite the fact that they share the expression of common cell surface adhesion proteins including PECAM, is that they do not undergo heterotypic cell adhesion between one another. When placed in culture, isolated human lymphatic endothelial cells have been shown to coalesce together to form lymphatic islands, or lymphatic tubes, dependent on the culture conditions, while blood endothelial cells form blood vascular islands or tubes (Kriehuber et al., 2001). While gene expression profiles of lymphatic and blood vascular endothelial cells have revealed differences in the expression of genes involved in cell adhesion between these populations of endothelial cells, the identity of the specific adhesion molecules responsible for maintaining separation of the blood and lymphatic vascular networks remains ambiguous (Petrova et al., 2002; Hirakawa et al., 2003). Within lymphatic vessels themselves, recent work has elegantly demonstrated that the intercellular junctions between initial lymphatic capillaries and collecting lymphatic vessels are distinct (Baluk et al., 2007), with initial lymphatic capillary endothelial cells displaying loose, discontinuous “button-like” junctions, while collecting vessel endothelial cells display more closed “zipper-like” junctions (Fig. 6). Interestingly, both types of junctions contain vascular endothelial cadherin (VE-cadherin) (Baluk et al., 2007), a molecule critical for endothelial cell survival, remodeling and maturation (Carmeliet et al., 1999). Baluk and colleagues also determined that, like collecting lymphatic vessels, the junctions of blood capillaries exhibited continuous “zipper-like” junctions (Baluk et al., 2007). These data suggest that the intercellular junctions between lymphatic endothelial cells are specialized depending on the identity and function of the lymphatic vessel; lymphatic capillaries have intercellular junctions that more readily allow the uptake of interstitial fluid, proteins and cells, while collecting vessels have continuous intercellular junctions that suit their lymph transport function. What are

the signals that determine whether lymphatic vessels adopt a capillary versus collecting vessel identity? Currently they remain enigmatic, but no doubt are being hotly pursued.

IV.  The embryonic origin of lymphatic endothelial cells Studies on the embryonic origin of lymphatic vessels, prominent early in the twentieth century, employed mainly mammalian and avian model systems to study lymphatic vascular development. These early studies utilized morphological differences between lymphatic and blood vessels, together with selective injection techniques, to identify and examine the embryonic lymphatic vasculature. Two major schools of thought as to how the lymphatic vasculature originated in the embryo were actively debated at this time and remain incompletely resolved today. The recent generation of molecular and genetic tools with which to recognize and specifically manipulate the lymphatic vasculature, together with the employment of animal models including the zebrafish and the Xenopus tadpole, has tremendously aided our ability to study the earliest events during lymphatic vascular genesis in the embryo. The recent “rediscovery” of lymphatic vessels in fish and tadpoles will no doubt result in a substantial increase in the use of zebrafish and Xenopus species as valuable models for the study of lymphangiogenesis, particularly as models in which to conduct high throughput genetic and/or chemical screens.

IV.A.  A Venous Origin of Lymphatic Vessels Florence Sabin, an anatomist who utilized the technique of ink injection into pig embryos to document the pattern of growing lymphatic vessels throughout embryonic development, proposed that lymphatic vessels originate by budding from the embryonic veins (Sabin, 1902, 1904) (Fig. 8). Sabin, together with others, proposed that the lymph sacs, the first lymphatics observed within the embryo, originated by endothelial cell

548

PART  |  8  Making Vessels

Figure 8  The growth of lymphatic vessels in pig embryos as visualized by the ink injection method of Sabin, 1904. Embryos 3 cm (A), 4.3 cm (B) and 5.5 cm (C) in length. Note the appearance of a progressively-sprouting lymphatic network, initially from the jugular lymphatic sacs (A) and later from pre-existing vessels (C). Reproduced from Sabin (1904), with permission of John Wiley & Sons, Inc.

(A)

Artery

IV.B.  A Mesenchymal Origin of   Lymphatic Vessels

(B)

Vein

Lymph sacs

“The Venous Origin”

Artery

Vein

Lymph sacs

“The Local Origin”

Figure 9  Model depicting “The Venous Origin” (A) versus “The Local Origin” (B) of lymphatic vascular development.

budding and sprouting from the embryonic veins towards the periphery, and that lymphatic vessels further developed from the lymph sacs by subsequent sprouting and migration to form an extensive lymphatic vascular network (Figs 8; 9). These experiments agreed with the hypothesis proposed almost 50 years earlier by Langer, that lymphatic vascular growth in the tadpole tail proceeded by sprouts extending from pre-existing lymphatic vessels, in a manner similar to that documented for blood capillary growth (Yoffey and Courtice, 1970). Recent support of a venous origin of embryonic lymphatic vessels, aided tremendously by the generation of molecular tools to identify lymphatic vessels, has been garnered in mouse (Wigle and Oliver, 1999; Srinivasan et al., 2007), while the genesis of lymphatic vessels in zebrafish (Yaniv et al., 2006) and Xenopus (Ny et al., 2005), appears at least partially attributable to venous progenitors.

In contrast to the venous origin hypothesis supported by Sabin, Huntington and colleagues proposed that lymphatic channels arise de novo from mesenchymal spaces via a mesenchymal-to-endothelial cell differentiation program, and that lymphatic vessels formed by this mechanism secondarily establish connections to the venous system to facilitate lymph return to the bloodstream (Huntington and McClure, 1910; Kampmeier, 1912) (Fig. 9). While this has been a less-favored explanation for lymphatic vascular genesis, evidence for this model, or for a model in which lymphatic vessels arise from both venous and mesenchymal components, has been documented in recent years in avian model systems (Schneider et al., 1999; Wilting et al., 2000) and Xenopus (Ny et al., 2005). Whether or not the hemopoietic compartment harbors a pool of lymphatic endothelial progenitor cells, either during embryogenesis or during postnatal neo-lymphangiogenesis, has been actively investigated in recent years. Macrophages have recently been proposed to act as a pool of lymphatic endothelial progenitor cells and differentiate into lymphatic endothelia in response to inflammatory stimuli (Kerjaschki et al., 2004; Maruyama et al., 2005). The differentiation of macrophages into lymphatic endothelial cells in response to inflammatory stimuli is a somewhat controversial proposal, as previous studies of tumor-stimulated lymphangiogenesis failed to detect any contribution of potential hemopoietic-derived lymphatic endothelial progenitor cells to the generation of new lymphatic vessels (He et al., 2004). In this particular study, it was concluded that the tumor-stimulated growth of new lymphatic vessels occurred solely by sprouting from pre-existing

549

Chapter  |  8.4  Lymphatic Vascular Development

Figure 10  Lymphatic endothelial cells of the jugular lymph sacs and dermal lymphatic vessels express VEGFR-3 (A, B) and LYVE-1 (C, D). In contrast to blood vessels (white arrows: B, D), lymphatic vessels (blue arrows: B, D) express very low levels of laminin and collagen IV.

lymphatics. Definitive lineage tracing studies to determine whether cells of the monocyte/macrophage lineage contribute to the generation of lymphatic endothelial cells during both embryonic and postnatal lymphangiogenesis remain to be done. The origin and identity of embryonic and postnatal lymphatic endothelial progenitor cells will no doubt remain an area of active investigation in the field of lymphatic vascular development.

V.  Molecular markers of lymphatic endothelial cells As mentioned above, the identification of molecular markers specific for lymphatic endothelial cells, as well as the generation of mouse models of perturbed lymphatic vascular development, have been major breakthroughs in the field of lymphangiogenesis research. Below some of the markers currently used to distinguish between lymphatic endothelial cells and blood endothelial cells are summar­ ized; while this catalog is by no means complete, there is little doubt it will be expanded now that the identification and manipulation of lymphatic endothelial cells is readily achievable. As is often the case in biology, the use of combinations of these markers is currently the best way to distinguish the identity of lymphatic vessels unambiguously.

V.A.  Cell Surface Markers of Lymphatic Endothelium V.A.i.  Vegfr-3 The first recognized molecule to be expressed preferentially in lymphatic vessels was the fms-like tyrosine kinase 4 (Flt-4), also known as vascular endothelial growth factor receptor 3 (Vegfr-3) (Kaipainen et al., 1995) (Fig. 10). Interestingly, despite the progressive restriction of Vegfr-3 expression to the lymphatic vasculature throughout embryonic development, targeted ablation of Vegfr-3 expression in the mouse resulted in early blood vascular defects, pericardial oedema and embryonic death at approximately E9.5, a time-point

prior to the initiation of lymphatic vascular development in the mouse (Dumont et al., 1998). This demonstrated that Vegfr-3 is critical for early blood vascular development, but precluded the analysis of Vegfr-3 function during lymphatic vascular development. VEGFR-3 has subsequently been demonstrated to be expressed on fenestrated blood capillaries in the adult (Kamba et al., 2006), on angiogenic capillaries in settings of inflammation (Witmer et al., 2004) and tumor development (Clarijs et al., 2002; Tammela et al., 2008) and on the remodeling postnatal retinal blood vasculature (Tammela et al., 2008), scenarios that likely recapitulate early embryonic events. Recent work demonstrated that VEGFR-3 is expressed at high levels in angiogenic vascular sprouts and that signaling mediated via VEGFR-3 is critical for blood vascular sprouting angiogenesis (Tammela et al., 2008). Several lines of evidence reveal a critical role for Vegfr-3 in lymphatic vascular development; heterozygous inactivating mutations in Vegfr-3 have been identified in human patients suffering from congenital lymphoedema (Irrthum et al., 2000; Karkkainen et al., 2000) and in the Chy mouse mutant that displays postnatal chylous ascites and lymphoedema (Table 1) (Karkkainen et al., 2001), signaling via VEGFR3 has been shown to be sufficient for lymphangiogenesis in mice (Veikkola et al., 2001) and inactivation of a VEGFR-3 ligand, Vegf-c, in mice results in severely arrested embryonic lymphatic vascular development (Karkkainen et al., 2004).

V.A.ii.  Podoplanin Podoplanin (Breiteneder-Geleff et al., 1997), first known variously as T1 (Dobbs et al., 1988), gp38 (Farr et al., 1992) and E11 (Wetterwald et al., 1996), was recognized to be a cell surface marker selective for lymphatic, but not blood, endothelial cells (Wetterwald et al., 1996). Podoplanin is a transmembrane glycoprotein expressed on a range of tissues in addition to lymphatic endothelia which is critical for embryonic lymphatic development (Schacht et al., 2003). Inactivation of podoplanin in mice resulted in early postnatal death due to respiratory failure, with podoplanin/ pups displaying mispatterned, dilated and dysfunctional lymphatic vessels and resultant ­ subcutaneous lymphoedema (Table 1)

550

PART  |  8  Making Vessels

Table 1  Mouse Knockout Models Exhibiting Developmental Lymphatic Vascular Defects Gene knockout

Phenotype

Prox1/

Failure to specify lymphatic endothelial cell (LEC) fate, complete absence of lymphatic vessels, embryonic lethality (Wigle and Oliver, 1999). Prox1/ exhibit mispatterned lymphatic vessels, postnatal chylous ascites, obesity (Harvey et al., 2005).

Sox18/

Failure to specify LEC fate, absence of lymphatic vessels, embryonic lethality (C57Bl/6) (Francois et al., 2008). Sox18Ra mice exhibit postnatal chylous ascites, edema, cardiovascular defects (Pennisi et al., 2000).

FoxC2/

Abnormal lymphatic vascular patterning, absence of collecting lymphatic vessel valves, aberrant recruitment of smooth muscle to lymphatic capillaries (Petrova et al., 2004).

Net/

Embryonic lymphangiectasia, postnatal chylothorax, postnatal lethality (Ayadi et al., 2001b).

Vezf1

Embryonic-lethal due to cardiovascular defects prior to lymphatic endothelial cell fate commitment. Vezf1/ embryos display hyperplastic jugular lymph sacs (Kuhnert et al., 2005).

Vegfr3/

Embryonic-lethal due to cardiovascular defects prior to lymphatic endothelial cell fate commitment (Dumont et al., 1998). Vegfr3Chy/ exhibit postnatal chylous ascites, hypoplastic superficial lymphatic vessels, lymphoedema (Karkkainen et al., 2001).

Angiopoietin2/

Postnatal chylous ascites, hypoplastic, mispatterned lymphatic capillaries, impaired lymphatic transport, defective smooth muscle cell recruitment to collecting lymphatic vessels (Gale et al., 2002).

Neuropilin2/

Transient embryonic and postnatal hypoplasia of superficial lymphatic vessels (Yuan et al., 2002).

Podoplanin/

Postnatal lethality, lymphangiectasia, lymphoedema, impaired lymphatic transport (Schacht et al., 2003).

EphrinB2V/V

Postnatal lethality, chylothorax, impaired lymphatic transport, hyperplastic collecting lymphatic vessels, absence of collecting lymphatic vessel valves, failure to remodel capillary plexus (Makinen et al., 2005).

Integrin 9/

Postnatal lethality, edema, chylothorax (Huang et al., 2000).

Emilin1/

Mild lymphoedema, impaired lymphatic transport, hyperplastic lymphatic vessels (Danussi et al., 2008).

Vegfc/

Embryonic lethality, failure of venous lymphatic endothelial progenitors to migrate away from embryonic veins. Vegfc/ mice display postnatal chylous ascites, hypoplastic lymphatic capillaries, impaired lymph transport, lymphoedema (Karkkainen et al., 2004).

Adrenomedullin/, Calcrl/, Ramp2/

Embryonic lethality, edema (Fritz-Six et al., 2008; Ichikawa-Shindo et al., 2008).

Aspp1/

Embryonic edema, mispatterned collecting lymphatic vessels (Hirashima et al., 2008).

Pi3kcaRBD/RBD

Postnatal lethality, chylous ascites, hypoplastic, mispatterned lymphatic vessels (Gupta et al., 2007).

/

Postnatal lethality, chylous ascites (Fruman et al., 2000).

Pik3r1

/

Spred1

; Spred2

/

/

Embryonic lethality, edema, hyperplastic lymphatic vessels, erythrocytes within lymphatic vessels (Taniguchi et al., 2007).

Syk/, Slp76/

Failure to separate lymphatic and blood vascular networks (Abtahian et al., 2003).

Fiaf /

Failure to separate postnatal intestinal lymphatic and blood vascular networks (Backhed et al., 2007).

(Schacht et al., 2003). Intriguingly, the identity of podoplanin ligands important for lymphatic vascular development, together with the mechanism of podoplanin signaling, remain uncharacterized, although in vitro assays suggest a role for podoplanin in lymphatic endothelial cell ­adhesion, ­migration

and tube formation (Schacht et al., 2003). Podoplanin is not expressed in blood vessels, or in lymphatic endothelial progenitor cells of the embryonic cardinal veins, but it is expressed in lymphatic endothelial cells once they have budded off the veins (Schacht et al., 2003).

551

Chapter  |  8.4  Lymphatic Vascular Development

V.A.iii.  LYVE-1 The lymphatic vascular endothelial HA receptor 1 (LYVE-1) was identified on the basis of shared structural and functional homology with the hyaluronan (HA) receptor CD44 (Banerji et al., 1999; Prevo et al., 2001) and is currently one of the markers most commonly used to identify lymphatic endothelial cells. LYVE-1 expression was initially found to be almost exclusively on lymphatic vessels, and lymph node and spleen sinusoidal endothelium in adult human and mouse tissues (Banerji et al., 1999; Prevo et al., 2001), although recent work has demonstrated the presence of LYVE-1 on endothelial sinusoids of the liver (Mouta Carreira et al., 2001), lung capillaries (Prevo et al., 2001) and a population of macrophages (Dadras et al., 2003; Maruyama et al., 2005; Schledzewski et al., 2006). The first example of LYVE-1 expression by blood vascular endothelial cells was noted in studies of early lymphatic vascular development; LYVE-1 expression was observed in the embryonic cardinal veins, at the sites where the first budding of venous endothelial cells to form the jugular lymph sacs occurs (Wigle et al., 2002) (Fig. 11). In addition, LYVE-1 expression has recently been described on isolated arterial and venous endothelial cells within the early mouse embryo (at E9.5 and E10.5), as well as extensively throughout the mouse yolk sac blood vasculature during embryogenesis (Gordon et al., 2008). Interestingly, in both the yolk sac blood vascular setting and the setting of the postnatal lymphatic vasculature, LYVE-1 expression has been shown to be downregulated on arteries (Gordon et al., 2008) and collecting lymphatic vessels (Makinen et al., 2005), respectively, on recruitment of vascular smooth muscle. LYVE-1 expression has also been shown to be reversibly downmodulated on exposure to the proinflammatory cytokine TNF- (Johnson et al., 2007). These recent data suggest that LYVE-1 is best used in conjunction with additional markers of the lymphatic vasculature to precisely identify lymphatic vessels in both embryonic and neolymphangiogenic settings. The precise role of LYVE-1 in lymphatic vascular development and function remains enigmatic; targeted inactivation of Lyve-1 in the mouse revealed that LYVE-1 is dispensable for hyaluronan transport and for embryonic and postnatal lymphangiogenesis (Valenzuela et al., 2003; Huang et al., 2006; Gale et al., 2007).

V.A.iv.  Neuropilin-2 Neuropilin-2 was first identified as a receptor for the class III family of semaphorins, and plays an important role in the process of axon guidance during neural development (Chen et al., 1997; Chen et al., 2000; Giger et al., 2000). A role for neuropilin-2 and the related cell surface glycoprotein, neuropilin-1, in embryonic angiogenesis of blood vessels (Takashima et al., 2002) was revealed following the major discovery that neuropilin-1 could act as a receptor for VEGF (Soker et al., 1998). Subsequent analyses ­ demonstrated that neuropilin-2

could bind the VEGF family members VEGF145, VEGF165, placental growth factor (PlGF) (Gluzman-Poltorak et al., 2000) and the major lymphangiogenic VEGF family member, VEGF-C (Karkkainen et al., 2001). Analyses performed by Yuan and colleagues demonstrated that neuropilin-2 is expressed on both embryonic and postnatal veins and at a higher level on lymphatic vessels, and that neuropilin-2 function is required for the development of lymphatic capillaries during embryogenesis (Fig. 4; Table 1) (Yuan et al., 2002). Interestingly, the genesis of blood vessels and collecting lymphatic vessels proceeds normally in neuropilin-2/ mice, and the generation of lymphatic vessels recover postnatally, highlighting a selective and stage-specific requirement for neuropilin-2 signaling in lymphatic vascular development (Yuan et al., 2002). Recent work from Caunt and colleagues demonstrated that blocking neuropilin-2 function by using an antibody that disrupts neuropilin-2 binding to VEGF-C prevented tumor-stimulated lymphangiogenesis and subsequent tumor metastasis without affecting established lymphatic vessels (Caunt et al., 2008). These observations suggest that neuropilin-2 plays a role in lymphatic vascular sprouting and highlights the potential use of neuropilin-2 blocking agents as antimetastatic therapeutics.

V.B.  Lymphatic Endothelial Cell Surface Markers Involved in Immune Cell Trafficking V.B.i.  D6 The cell surface chemokine receptor D6 was recently shown to be expressed in a subset of lymphatic vessels in the skin, gut and secondary lymphoid tissue, as well as on a proportion of vascular tumors (Nibbs et al., 2001). Targeted inactivation of the D6 receptor in mice demonstrated that D6 is important for the resolution of cutaneous inflammation by virtue of its proinflammatory chemokine binding activity (Jamieson et al., 2005; Martinez de la Torre et al., 2005). D6 receptor binding to inflammatory chemokines including macrophage inflammatory protein1 (MIP-1) does not initiate receptor signaling events but instead, results in rapid chemokine internalization and degradation (Weber et al., 2004). The major role of D6 on lymphatic endothelial cells appears to be in the resolution of inflammation; while no investigations into the structure and function of the lymphatic vasculature have been performed in D6 knockout mice to date, D6/ mice are viable and healthy (Jamieson et al., 2005).

V.B.ii.  Ccl21 Ccl21, also known as secondary lymphoid tissue chemo­ kine (Slc), is expressed in lymphatic endothelial cells, as well as in the specialized endothelium of high endothelial venules (HEV) in lymph nodes (Gunn et al., 1998).

552

The production of Ccl21 by lymphatic endothelial cells is important for lymphocyte and dendritic cell migration into the afferent lymphatic vessels and subsequently to the lymph nodes, where these cells act to initiate and coordinate subsequent immune responses (Gunn et al., 1999). Like podoplanin, the expression of Ccl21 in embryonic lymphatic endothelial cells is initiated once Prox1-positive cells have budded from the cardinal veins (Wigle et al., 2002). Whether or not Ccl21 plays a role in the orchestration of embryonic lymphatic vascular development has not been investigated.

V.C.  Macrophage Mannose Receptor The macrophage mannose receptor (MMR) is expressed on the cell surface of macrophages, where it functions to recognize and engulf pathogens (Stahl and Ezekowitz, 1998). MMR has more recently been shown to be expressed on the cell surface of lymphatic endothelial cells, where it is able to interact with L-selectin on lymphocytes and med­ iate lymphocyte binding to, and trafficking via, afferent lymphatic vessels (Irjala et al., 2001; Marttila-Ichihara et al., 2008). Like D6 and Ccl21, MMR does not appear to be required for lymphatic vascular development; mice deficient in MMR display lymphatic vascular morphology indistinguishable from wild-type counterparts (MarttilaIchihara et al., 2008).

V.D.  Nuclear Markers of Lymphatic Endothelium V.D.i.  Prox1 Prox1 is a homeobox transcription factor that was first identified on the basis of shared homology with the Drosophila melanogaster homeobox gene, prospero (Oliver et al., 1993). Prox1 is expressed in a wide variety of tissues, but

PART  |  8  Making Vessels

importantly, with respect to endothelial cells, is a marker specific for lymphatic endothelial cells (Fig. 4). Prox1 was the first gene identified to be absolutely required for generation of the lymphatic vasculature during embryogenesis; Prox1/ mouse embryos are completely devoid of lymphatic vessels (Table 1) (Wigle and Oliver, 1999) and Prox1/ mice display a range of lymphatic vascular defects (Harvey et al., 2005). Prox1 was also the first specific molecular marker available to examine the earliest recognizable lymphatic endothelial progenitor cells in the embryo. Analyses of Prox1 expression in the early mouse embryo revealed Prox1 to be present in a polarized population of endothelial cells in the embryonic cardinal veins; these cells were observed to bud off and migrate away from the veins to form the jugular lymph sacs (Wigle and Oliver, 1999) (Fig. 11). This was the first molecular evidence to support Florence Sabin’s hypothesis of a venous origin for the lymphatic vasculature. Subsequent genetic lineage tracing analyses using endothelial and hemopoeitic Cre/loxP mouse lines have demonstrated that the venous pool of Prox1-positive progenitor cells appears to be the major source of lymphatic endothelial cells in the mouse embryo (Srinivasan et al., 2007).

V.D.ii.  FoxC2 The closely-related forkhead/winged helix transcription factors FoxC1 and FoxC2 play roles in the genesis of multiple organ systems, including the cardiovascular system, during embryonic development (Winnier et al., 1997, 1999; Kume et al., 1998, 2000). A role for FoxC2 in lymphatic vascular development was revealed when inactivating mutations in FoxC2 were found to be responsible for the hereditary lymphoedema/distichiasis syndrome (Fang et al., 2000; Bell et al., 2001; Finegold et al., 2001). Subsequent analyses of FoxC2/ mice demonstrated that FoxC2 is required for the embryonic patterning of the lymphatic vasculature, for the development of

Figure 11  Expression of Prox1 in lymphatic endothelial progenitor cells. Prox1 is expressed in a polarized population of cells within, and budding from, the embryonic cardinal veins at E11.5 (A: arrows) and continues to be expressed in lymphatic endothelial cells comprising the jugular lymph sacs (JLS) at E14.5 (C). By E14.5, expression of Prox1 in venous endothelial cells of the jugular veins (JV) decreases (C). LYVE-1 is expressed by lymphatic endothelial progenitor cells in the cardinal veins (B), as well as lymphatic endothelial cells in the jugular lymph sacs (D). Lymphatic endothelial cells, once they have budded off the cardinal veins, express neuropilin 2 (C). Prox1-positive lymphatic endothelial cells express a lower level of CD31 than do arterial and venous endothelial cells (A, C) and are not invested with smooth muscle cells as the arteries and, to a lesser extent the veins, are (A, B, D).

553

Chapter  |  8.4  Lymphatic Vascular Development

valves in collecting lymphatic vessels, and for maintaining a pericyte-free lymphatic capillary plexus (Table 1) (Petrova et al., 2004). These data provided a molecular explanation for the association of FoxC2 mutations with the failure to properly transport lymph in lymphoedemadistichiasis patients. Interestingly, FoxC2 expression in embryonic endothelial cells is not restricted to lymphatic endothelium; FoxC2 is also expressed in the early embryonic blood vasculature and closely associated mesenchyme, although FoxC2 does not appear to be expressed in blood vessels in the skin (Kume et al., 2001; Petrova et al., 2004). FoxC2 is expressed continuously in lymphatic endothelial cells from the initiation of lymphatic vascular development in the mouse (approximately E10.0), through to the adult (Dagenais et al., 2004), with particularly high expression in the lymphatic valves compared to neighboring endothelial cells (Petrova et al., 2004). The recent discovery that FoxC2 mutations are also present in patients with venous primary valve failure (Mellor et al., 2007) suggests that FoxC2 is utilized in both venous and lymphatic vascular networks for the construction and function of valves. The expression and function of FoxC2 in development of the blood vasculature, as well as the lymphatic vasculature, means that it is important to use FoxC2 in combination with additional markers of the lymphatic and blood vasculature, to identify vessel type precisely. The transcriptional targets of FoxC2 that are important for valve formation remain uncharacterized. To date, the only characterized marker that is reported to be specific for the endothelium of venous and lymphatic valves is class III tubulin (TuJ1), although the role of TuJ1 in valve formation and function is not currently known (Kang and Lee, 2006). The critical role of valves for the transport function of lymphatic vessels will no doubt result in inroads being made to the identity of signaling systems important for valve formation.

V.D.iii.  Sox18 Multiple lines of evidence implicated the SRY-related transcription factor Sox18 in lymphatic vascular development. The phenotype of ragged (Ra) mutant mice that display the lymphatic vascular defects of peritoneal chylous ascites and lymphoedema, as well as hair follicle and blood vascular defects, was found to be caused by mutations in Sox18 (Table 1) (Pennisi et al., 2000). The similarity of the Ra mouse phenotype to the human condition hypotrichosis-lymphoedema-telangiectasia (HLT), in which patients display sparse hair and lymphoedema, prompted Irrthum and colleagues to investigate whether mutations in SOX18 might underlie this syndrome; both dominant and recessive inactivating mutations in SOX18 were found in HLT patients (Irrthum et al., 2003). Sox18 has recently been shown to initiate lymphatic vascular development in the mouse; studies by Francois and colleagues demonstrated

that Sox18 activity is required upstream of Prox1 for the specification of lymphatic endothelial cell fate. Sox18 initiates lymphatic vascular development by directly binding to the Prox1 promoter and activating Prox1 transcription (Francois et al., 2008). Sox18 activity is not restricted to the lymphatic vasculature, but is also involved, along with fellow Sox F subgroup members Sox7 and Sox17, in early blood vascular development (Sakamoto et al., 2007; Cermenati et al., 2008; Herpers et al., 2008). It therefore seems likely that spatially-restricted Sox18 co-factors might be important for the localized induction of Prox1 expression in a discrete population of Sox18-expressing venous endothelial cells.

V.D.iv.  NFB The generation of mice expressing the reporter gene lacZ under the control of nuclear factor B (NF-B) activity revealed unexpected, constitutive transcriptional NF-B activity in lymphatic endothelial cells, but not in blood endothelial cells, of a variety of adult mouse tissues (Saban et al., 2004). Interestingly, like the expression of FoxC2 and Prox1 transcription factors, expression of the NF-B-driven reporter gene in adult lymphatic vessels was visible in a beaded pattern, suggesting that NF-B activity is likely to be high in lymphatic valves. Until NF-B transcriptional activity is examined specifically in the lymphatic vasculature during embryogenesis, a role for this transcriptional regulator during lymphatic vascular development will remain uncharacterized. The downstream targets of NFB in lymphatic endothelial cells have not yet been investigated.

V.D.v.  Net The ternary complex factor family member Net is an Etsdomain transcription factor that is expressed throughout the embryonic vasculature and in developing cartilage (Ayadi et al., 2001a). Even though Net is expressed in embryonic blood vascular endothelium, gene-targeted mice expressing a hypomorphic Net mutation revealed a specific requirement for Net function during lymphatic vascular development. Homozygous Net-mutant mice died of respiratory failure due to chylothorax within the first few weeks after birth (Table 1) (Ayadi et al., 2001b), and displayed highly dilated thoracic lymphatic vessels. Intriguingly, the lymphatic vascular phenotype observed in homozygous Net-mutant mice was reported to be restricted to the lymphatic vessels of the thoracic region; the morphology of lymphatic vessels in the skin and pericardium of postnatal mutant mice appeared normal. The specific transcriptional targets of Net in lymphatic endothelial cells remain uncharacterized, although increased expression of the immediate early gene, egr-1, was observed in a restricted compartment of the thoracic blood vasculature in Net-mutant mice (Ayadi et al., 2001b).

554

The role of Net in the transmission of MAP kinase pathway signals suggests that Net could be activated downstream of a wide variety of signaling cascades initiated at the endothelial cell surface.

VI.  Genesis and remodeling of the lymphatic vasculature While some of the proteins expressed by lymphatic endothelial cells play a functional role in embryonic lymphangiogenesis (as discussed above), the catalog of genes important for lymphatic vascular development is not restricted only to those expressed cell-autonomously in lymphatic endothelium, nor does the process of lymphatic vascular development arrest at the completion of embryogenesis. Here, the role of additional signals that influence lymphatic vascular growth and remodeling is discussed.

VI.A.  Lymphangiogenesis Promoting Factors VI.A.i.  Angiopoietin-2 Angiopoietin-2 was established to play a role in blood vessel angiogenesis prior to its role in lymphatic vascular development being recognized. Angiopoietin-2 was first isolated on the basis of shared homology with the secreted angiogenic factor angiopoietin-1, but in contrast to angiopoietin-1 it was shown to be an antagonist of signaling through the Tie2 tyrosine kinase receptor expressed on vascular endothelial and early hemopoietic cells (Maisonpierre et al., 1997). Targeted inactivation of angiopoietin-2 in mice demonstrated that angiopoietin-2 was not required for embryonic blood vascular development, but was required for blood vascular remodeling, and unexpectedly, for lymphatic vascular patterning and integrity (Table 1) (Gale et al., 2002). Evidence of an important role for angiopoietin-2 in lymphatic vascular integrity was observed soon after the birth of angiopoietin-2/ pups; the majority of pups displayed milky chylous ascites in the peritoneal cavity after suckling, a result of the failure of intestinal lymphatic vessels to transport lipids away from the digestive tract. Analyses of lymphatic vascular pattern in other tissues of angiopoietin-2/ mice also revealed strikingly mispatterned lymphatic vessels and a failure of smooth muscle association with collecting lymphatic vessels. Surprisingly, the lymphatic phenotype of angiopoietin-2/ mice was rescued by introducing angio­ poietin-1 expression from the angiopoietin-2 locus (Gale et al., 2002). These data demonstrated that a likely agonistic activity of angiopoietin-2 is important for lymphatic vascular development, while an antagonistic activity is important for its blood vascular remodeling role. In an extension of these initial studies, recent work demonstrated that the lymphatic abnormalities in

PART  |  8  Making Vessels

angiopoietin-2/ mice largely involve the collecting lymphatic vessels; vessels that should have exhibited a collector phenotype instead resembled lymphatic capillaries, indicating a failure of the lymphatic vasculature to remodel (Dellinger et al., 2008). The mechanism by which angiopoietin-2 acts on lymphatic endothelial cells remains incompletely deciphered; while the angiopoeitin-2 receptor, Tie2, is expressed on blood vascular endothelial cells, whether or not Tie2 is present on lymphatic endothelial cells remains controversial. Tie2-driven reporter gene expression has not been convincingly demonstrated on embryonic or adult lymphatic endothelial cells (Saban et al., 2004; Srinivasan et al., 2007; Dellinger et al., 2008), while an antibody to the Tie2 receptor has been shown to bind lymphatic endothelial cells in adult mouse tissues (Morisada et al., 2005; Tammela et al., 2005), and mRNA for Tie2 has been detected in embryonic and adult tissue LEC (Kriehuber et al., 2001; Petrova et al., 2002; Morisada et al., 2005). Interestingly, while reporter gene expression driven from the angiopoietin-2 locus is detected in the smooth muscle cells surrounding arteries and veins in Ang2/LacZ mice, as well as in endothelial cells at sites of vessel remodeling (Gale et al., 2002; Dellinger et al., 2008), reporter gene expression is also observed in what appear to be lymphatic vessels in Ang2LacZ/LacZ mice (Gale et al., 2002; Dellinger et al., 2008). It is therefore plausible that the expression of angiopoietin-2, or indeed Tie2, is normally very low in lymphatic endothelial cells, but might be induced in response to environmental remodeling stimuli.

VI.A.ii.  Angiopoietin-1 Recent data have revealed that angiopoietin-1 promotes lymphangiogenesis when ectopically-expressed in mouse tissues, although the mechanism of lymphangiogenic growth induced by angiopoietin-1 remains incompletely resolved (Morisada et al., 2005; Tammela et al., 2005). It has been suggested that angiopoietin-1 stimulation of lymphatic endothelial cells results in increased expression of Vegfr-3, activation of which is associated with prolymphangiogenic signaling (Tammela et al., 2005). A prolymphangiogenic role for angiopoietin-1 is perhaps unexpected, given that angiopoietin-1 was shown to stabilize the lymphatic vascular defects induced as a result of angiopoietin-2 disruption (Gale et al., 2002), although the rescue data of Gale and colleagues does suggest that angiopoietin-1 acts as a receptor agonist on lymphatic endothelial cells. It is, however, possible that angiopoietin1 stimulates lymphangiogenesis indirectly, rather than by direct binding to lymphatic endothelial cells. An alternative potential explanation for the prolymphangiogenic role of angiopoietin-1 is that angiopoietins could mediate their effect on lymphatic endothelial cells not through binding to Tie receptors, but potentially via binding to integrin family members including integrin 1 (Morisada et al., 2005).

Chapter  |  8.4  Lymphatic Vascular Development

VI.B.  EphrinB2 and Postnatal Lymphatic Remodeling Like the angiopoietins, an important role for the ephrinB2 transmembrane ligand, together with its Eph receptor counterpart EphB4, had been established in blood vascular development prior to a role for ephrinB2 being recognized in lymphatic vascular development. In an elegant set of experiments published by Makinen and colleagues (Makinen et al., 2005), regions in the cytoplasmic tail of ephrinB2 encoding a PDZ interaction site, or phosphotyrosine signaling residues, were mutated to remove their respective functions, and introduced into the ephrinB2 locus by gene targeting to create knockin mice (Makinen et al., 2005). While neither line of ephrinB2 homozygous mutant mice displayed developmental blood vascular defects, mice homozygous for removal of the PDZ interaction site (ephrinB2V/V) died within the first few weeks of birth exhibiting major defects in lymphatic vascular remodeling, obvious due to the accumulation of chylous ascites in the thoracic cavity (Table 1). Further analysis of these mice revealed that the lymphatic vasculature failed to undergo remodeling to generate the characteristic hierarchical organization of superficial capillary plexus and collecting lymphatic network; hyperplastic lymphatic vessels devoid of valves were present in place of collecting lymphatic vessels. Interestingly, ephrinB25F/5F mice exhibited only a mild lymphatic phenotype, demonstrating that the interaction of ephrinB2 with signaling components that bind the PDZ domain is essential for the development of collecting lymphatic vessels. EphrinB2 expression was demonstrated to be dynamic during the process of lymphatic vascular remodeling in the dermis; in wild-type mice, ephrinB2 was expressed in the primary capillary plexus of P0–P4 pups as the plexus sprouted to upper dermal layers to generate a more superficial plexus, but was then downregulated in the most superficial plexus by P5. Expression remained high on the initial capillary plexus. The signals important for the specification of the superficial versus collecting lymphatic vessels thus far remain uncharacterized, although the PDZ binding proteins RGS3 and Dvl2 were identified as potential downstream mediators of ephrinB2 signaling in the specification of valve formation (Makinen et al., 2005).

VI.C.  The Vascular Endothelial Growth Factor (VEGF) Family: VEGF-A, -B, -C   and -D and PlGF Members of the vascular endothelial growth factor (VEGF) family play critical roles in both embryonic and postnatal vascular development by virtue of the fact that they bind to VEGF receptors expressed by endothelial cells. While the importance of VEGF-A for blood vascular ­ development,

555

angiogenesis (Carmeliet et al., 1996; Ferrara et al., 1996) and endothelial cell survival (Lee et al., 2007) has been extensively characterized and is reviewed in the previous chapter, members of the VEGF family have more recently been shown to play a role in lymphangiogenesis. The ability of VEGF family members to stimulate both angiogenesis and lymphangiogenesis is due to the expression of VEGF receptors by both blood and lymphatic endothelial cells. While VEGFR-1 expression is almost exclusively restricted to blood vascular endothelial cells, VEGFR-2 and -3 are expressed on both blood and lymphatic endothelial cells. VEGF-A can therefore signal via VEGFR-2 on lymphatic endothelial cells, and has recently been shown to act as a lymphangiogenic factor (Nagy et al., 2002; Hirakawa et al., 2005). As discussed above, while VEGFR-3 expression is mostly restricted to lymphatic vessels, immature blood vessels during both embryonic and postnatal angiogenesis express VEGFR-3, therefore VEGF-C has the potential to promote angiogenesis of blood vessels. A dual opportunity for VEGF-C to promote blood vessel angiogenesis is provided by virtue of the ability of proteolyticallyprocessed forms of VEGF-C to bind VEGFR-2 (Joukov et al., 1996). VEGF-D, identified by virtue of homology to VEGF-A, and the closest relative of the lymphangiogenic factor VEGF-C, also binds both VEGFR-2 and -3 (Achen et al., 1998) and is a potent lymphangiogenic factor (Stacker et al., 2001; Rissanen et al., 2003). Surprisingly, VEGFD was also shown to have the strongest proangiogenic effect in an assay in which adenoviral vectors expressing VEGF family members were delivered to skeletal muscle (Rissanen et al., 2003). VEGF-D appears to be dispens­ able for embryonic lymphatic vascular development in the mouse (Baldwin et al., 2005), but is able to compensate for the deficiency of VEGF-C and rescue lymphangiogenesis in Vegf-c/ or Vegfc/ embryos (Table 1) (Karkkainen et al., 2004; Haiko et al., 2008). A role for VEGF-D in the migration of embryonic lymphatic endothelial cells has recently been demonstrated in the Xenopus laevis tadpole (Ny et al., 2008). Interestingly, VEGF-E reported to bind only to VEGFR2, has recently been shown to promote lymphatic vessel hyperplasia but not sprouting, suggesting that signals transduced by different receptors mediate vessel growth versus sprouting decisions (Wirzenius et al., 2007). While PlGF binds only to VEGFR-1, a receptor not expressed by lymphatic endothelial cells, recent abrogation of PlGF activity in tumors using an anti-PlGF antibody resulted in the inhibition of tumor-initiated lymphangiogenesis (Fischer et al., 2007). The mechanism of lymphangiogenesis inhibition by 10 anti-PlGF was proposed to be via the prevention of prolymphangiogenic macrophage recruitment to the tumor environment (Fischer et al., 2007), although the potential binding of PlGF directly to neuropilin-2 (Gluzman-Poltorak et al., 2000) on lymphatic endothelial cells was not examined.

556

VI.D.  Alternative Splicing and Proteolytic Processing of Vascular Endothelial Growth Factor (VEGF) Family Members An additional level of complexity in VEGF signaling pathways occurs as a consequence of the fact that the production and activity of VEGF family members is controlled both by alternative splicing and proteolytic processing. While little attention has focused on the physiological control of these processes to date, this area of research will no doubt be a hot topic in the near future. Alternatively, spliced isoforms of VEGF-A that result in VEGF-A being retained at the cell surface or in the extracellular matrix, or being freely diffusible, have been best-characterized and are discussed in the previous chapter. Signaling initiated by alternatively-spliced VEGF-A isoforms results in distinct biological outcomes for endothelial cells (Ferrara et al., 2003), which implies that the regulation of alternative splicing is likely to be critical during vascular development, as well as in settings of postnatal angiogenesis. Proteolytic processing of VEGF-A by proteases including plasmin (Houck et al., 1992) and matrix metalloproteases (MMPs) (Lee et al., 2005) also regulates VEGF-A bioavailability, due to proteolytic release of extracellular matrixbound VEGF-A. The release of extracellular matrix-bound VEGF-A is associated with vessel dilation due to endothelial cell proliferation, while the maintenance of extracellular matrix binding is associated with vessel branching. Like VEGF-A, the lymphangiogenic factors VEGF-C and VEGF-D are proteolytically processed by proteases, including plasmin and members of the proprotein convertase (PC) family (Joukov et al., 1997; Baldwin et al., 2001; McColl et al., 2003, 2007; Siegfried et al., 2003). In the case of these VEGF family members, proteolysis of the N- and C-terminal propeptides is required for maximum VEGFR binding and bioactivity (McColl et al., 2003; Rissanen et al., 2003). Like VEGF-A, several isoforms of VEGF-D protein are produced by alternative splicing (Baldwin et al., 2001). It will be intriguing to determine whether a restricted temporal or spatial expression pattern of VEGF-activating proteases is important for orchestrating the precise control of VEGF family member signaling during vascular development and postnatal angiogenesis/lymphangiogenesis.

VI.E.  PDGF-BB, FGF-2, HGF, IGF-1   and -2, Adm Many growth factors that have been implicated in stimulating blood vascular angiogenesis also show prolymphangiogenic activity, most likely due to the shared expression of receptors on both endothelial cell types. The identification of factors that promote lymphangiogenesis is important not only from a developmental perspective, but also from a disease perspective; by defining the signals that stimulate

PART  |  8  Making Vessels

the growth of new lymphatic vessels in settings of inflammation, it is hoped that agents can be developed to block these signals and thereby provide novel anti-inflammatory and antimetastatic therapeutics. While the following growth factors have been shown to possess lymphangiogenic activity, the role of each during embryonic lymphatic vascular development has not been fully-explored. The platelet derived growth factor (PDGF) family member Pdgfb has been reported to play two roles in lymphangiogenesis. Consistent with the critical role of Pdgfb and Pdgfrb for smooth muscle cell/pericyte recruitment during blood vascular development (Lindahl et al., 1997; Hellstrom et al., 1999), increased Pdgfb expression in the superficial capillary plexus of FoxC2/ mice was associated with the aberrant recruitment of smooth muscle cells to lymphatic capillaries (Petrova et al., 2004). This suggests that Pdgfb is normally a gene that is repressed, at least in the superficial lymphatic capillary plexus, during lymphatic vascular development. Whether Pdgfb is normally expressed during lymphatic maturation by collecting lymphatic vessels, in order to recruit smooth muscle cells, has not yet been characterized. PDGF-BB has also been shown to stimulate lymphangiogenesis in vitro and in vivo in mouse models, reportedly by binding to PDGF receptors on lymphatic endothelial cells (Cao et al., 2004). Implantation of PDGF-BB to the mouse cornea stimulated both angiogenesis and lymphangiogenesis, while ectopic expression of PDGF-BB in fibrosarcoma cells stimulated tumor-associated lymphangiogenesis and metastasis of tumor cells via lymphatic vessels (Cao et al., 2004). PDGF family members PDGF-AA and PDGF-AB were also shown to stimulate angiogenesis and lymphangiogenesis in the mouse cornea model (Cao et al., 2004), as did FGF-2 (Kubo et al., 2002; Cao et al., 2004; Chang et al., 2004). How can one growth factor play pleiotropic roles in lymphangiogenesis? One possibility is that the levels of PDGF-BB ectopically-expressed in mouse tumor models are far higher than those normally seen by lymphatic endothelial cells during development; saturation of receptor binding could result in different biological outcomes. Indirect effects as a result of PDGF-BB stimulation of additional cell types in the tumor environment could also result in the liberation of secondary lymphangiogenic signals. Dissection of the normal, physiological role of PDGF family members and their receptors during embryonic lymphatic vascular development awaits characterization of the lymphatic vasculature in mice harboring mutations in Pdgfb and Pdgfrb. The development of reagents that will allow the conditional inactivation of candidate PDGFs and their receptors specifically in lymphatic endothelial cells will aid these studies. A comparison of gene expression profiles of human LEC and blood endothelial cells (BEC) illustrated that the receptor for hepatocyte growth factor (HGF)/scatter factor was expressed at a higher level in LEC than BEC, prompting the role of HGF signaling in lymphangiogenesis to be investigated (Kajiya et al., 2005). HGF is a growth ­factor highly-associated

557

Chapter  |  8.4  Lymphatic Vascular Development

with cell migration during embryonic development, and has been shown to stimulate angiogenesis in a number of assays (Birchmeier et al., 2003). In studies performed by Kajiya and colleagues, HGF was demonstrated to stimulate lymphangiogenesis in vitro and HGF receptor was shown to be expressed on lymphatic endothelial cells in vivo, both during embryonic lymphatic vascular sprouting as well as in settings of tissue repair and inflammation (Kajiya et al., 2005). Taken together, these data suggest that HGF signaling through the HGF receptor could be important for both embryonic and inflammation-stimulated lymphangiogenesis, although these processes are yet to be investigated in mice harboring mutations in either HGF or HGF-R. Mutations in both HGF and the HGF receptor cMET have recently been found in patients with primary lymphoedema, lymphoedema/lymphangioectasia and breast cancer-associated secondary lymphoedema (Finegold et al., 2008). Like HGF, FGF-2 and the PDGF family members discussed above, the insulin-like growth factors Igf-1 and Igf-2 have been shown to promote blood vascular angiogenesis (Nicosia et al., 1994; Volpert et al., 1996). Igf-1 and Igf2 were also recently shown to promote lymphangiogenesis in a mouse corneal neovascularization assay, and to stimulate the proliferation and migration of isolated LEC (Bjorndahl et al., 2005). The prolymphangiogenic effects of IGF-1 and IGF-2 were proposed to occur directly via IGF receptors -1 and -2 expressed on lymphatic endothelial cells, although the involvement of accessory signals in the corneal milieu can not be excluded. Whether or not IGF signaling plays a physiological role during embryonic lymphatic vascular development is yet to be reported. The peptide adrenomedullin (Adm) was recently shown to play roles in both lymphatic and blood vascular development in the mouse (Table 1) (Fritz-Six et al., 2008; IchikawaShindo et al., 2008). Targeted inactivation of Adm, or of Adm receptors Calcrl or Ramp2 resulted in midgestational lethality, with Adm/, Calcrl/ and Ramp2/ embryos displaying massive edema. The magnitude of edema observed in these mutant embryos was reportedly much more severe than that observed in Vegf-c/ embryos in which the lymphatic vasculature is selectively affected, or those in which the blood vasculature or cardiac function is selectively affected. This is likely due to the probability that edema in Adm/, Calcrl/ and Ramp2/ embryos occurs as a result of a combined increase in the leakage of fluid to the interstitium due to blood vascular integrity defects, combined with a failure in efficient interstitial fluid transport back to the blood vasculature, due to dysfunctional lymphatic vessels.

VI.F.  Integrins Integrins are heterodimeric cell surface glycoproteins composed of  and  subunits that bind both to the ­extracellular matrix (ECM) and to members of the immunoglobulin

superfamily. Several integrins have well-established roles in blood vessel angiogenesis, being important for endothelial cell survival, proliferation and migration (Avraamides et al., 2008). More recently, select members of the integrin family have been demonstrated to be important for lymphangiogenesis. Integrin 91 was recently shown to be important for embryonic lymphatic vascular development; mice with targeted inactivation of 9 integrin developed respiratory failure as a result of chylothorax and died within the first few postnatal weeks (Table 1) (Huang et al., 2000). Intriguingly, 9 integrin expression in the major collecting lymphatic vessel, the thoracic duct, was shown to be transient during embryogenesis, suggesting that integrin 9 function plays a temporally-restricted role in lymphatic vascular development. Integrin 9 could perhaps be critical for the migration and initial coalescence of the thoracic duct, but may not be required for later vascular maturation (Huang et al., 2000). Recent work suggested that integrin 9 is a downstream target gene of the master lymphatic regulator Prox1 (Mishima et al., 2007), although Prox1 is continually expressed in embryonic and adult lymphatic endothelial cells, suggesting that additional factors are likely to be involved in the regulation of 9 integrin expression in LEC. While no lymphangiogenic roles for the documented 91 ligands tenascin C, osteopontin and vascular cell adhesion molecule-1 (VCAM-1) have been reported to date, 91 has recently been demonstrated to bind the lymphangiogenic factors VEGF-C, VEGF-D and VEGF-A to effect cell migration (Vlahakis et al., 2005, 2007). Integrins 1 and 2 have also been shown to mediate VEGF-A induced lymphangiogenesis during wound repair (Hong et al., 2004), and small molecule inhibitors of integrin 5 have been demonstrated to ablate inflammation-stimulated lymphangiogenesis (Dietrich et al., 2007). As integrins do not signal directly through their intracellular termini, but via their association with kinases and adaptor proteins in focal adhesion complexes, it could be postulated that integrins that respond to lymphangiogenic signals might associate with and signal through the VEGF receptors VEGFR-2 and VEGFR-3. Alternatively, 91, 1, 2 and 5 integrins might signal via as yet undetermined lymphatic endothelial cell signaling complexes.

VI.G.  Hemopoietic Cells and Lymphangiogenic Signals Given that one of the major roles of lymphatic vessels is to traffic leukocytes during immune surveillance and immune responses, it is not surprising that leukocytes communicate with lymphatic endothelial cells. Several groups studying inflammation-stimulated lymphangiogenesis in adult mice have recently demonstrated that macrophages play a role in stimulating the growth of new lymphatic vessels by liberating lymphangiogenic stimuli including VEGF-C and

558

VEGF-D (Schoppmann et al., 2002; Cursiefen et al., 2004; Baluk et al., 2005). B-lymphocytes have recently been shown to stimulate lymphangiogenesis in inflamed lymph nodes post-immunization (Angeli et al., 2006). The pro­ lymphangiogenic signal released by B-lymphocytes in this study was proposed to be VEGF-A, although the contribution of VEGF-C and other lymphangiogenic factors to the expansion of the lymph node lymphatic vasculature could not be ruled out (Angeli et al., 2006). An interesting downstream effect of increased lymph node lymphangiogenesis was the increased trafficking of dendritic antigen-presenting cells to the lymph node, implying that the stimulation of lymph node lymphangiogenesis is likely to be an important event in mounting efficient immune responses. A second, independent study has also implicated B-lymphocytes as a source of prolymphangiogenic VEGF. Expansion of the number of B-lymphocytes in the lymph nodes of E-c-myc mice was shown to result in increased lymph node lymphangiogenesis and increased lymph flow (Ruddell et al., 2003). Whether or not B-lymphocytes produce additional lymphangiogenic factors remains to be established.

VI.H.  Signaling Downstream of VEGFR3 VI.H.i.  PI-3 Kinase Signaling in Lymphatic Vascular Development Class IA phosphoinositide 3-kinases (PI-3 kinase) are enzymes that translocate to cell membranes on the receipt of tyrosine kinase initiated signals, and then act to liberate second messengers that propagate these initial signals further. PI-3 kinases are composed of a regulatory and a catalytic subunit, both of which are required for activity. Disruption of the ability of the catalytic p110 sub­ unit of PI-3 kinase to interact with Ras has recently been shown to result in aberrant embryonic lymphatic vascular development (Table 1) (Gupta et al., 2007). Pups homozygous for the p110-Ras-binding mutation display chylous ascites accumulation in the peritoneal cavity and die soon after birth. Lymphatic vascular development in these mutants was shown to be defective, with hypoplastic, mispatterned lymphatic vessels obvious in the skin, mesentery and thorax, although the thoracic duct formed normally in homozygous mutants. The similar phenotype of these mutant mice to Vegfc/ mice was suggested by the authors to perhaps reflect a role for p110-Ras interaction in the propagation of VEGF-C initiated signaling via VEGFR-2 and/or VEGFR-3 receptors. However, the fact that mice harboring targeted mutations in genes including Prox1 and Ang2 exhibit similar phenotypes of postnatal chylous ascites and lethality suggests that Ras-PI-3 kinase signaling could be important downstream of additional, or alternative lymphangiogenic stimuli. Interestingly, Pik3r1/ mice with a targeted mutation in the regulatory subunits of PI-3 kinase also display postnatal chylous

PART  |  8  Making Vessels

ascites and die shortly after birth (Table 1) (Fruman et al., 2000). Whether or not the lymphatic vascular phenotype in these mice occurs as a result of a reduction in p110 interaction with Ras, or as a consequence of alterations in other signaling events, has not been characterized.

VI.H.ii.  Spreds Spred/Sprouty family members have been characterized as negative regulators of growth factor- and cytokinestimulated kinase activation. A role for two of the three Spred family members in embryonic lymphatic vascular development was recently discovered when mice harboring targeted inactivation of both Spred-1 and Spred-2 were generated (Table 1) (Taniguchi et al., 2007). Spred-1/; Spred-2/ mutants die between E12.5 and E15.5, and display lymphatic vascular aberrations including subcutaneous oedema, hyperplastic dermal lymphatic vessels and the presence of erythrocytes within lymphatic vessels (Taniguchi et al., 2007). The mechanism by which Spreds control lymphatic vascular development was proposed to be via dampening signaling through the VEGF-C/VEGFR3 axis, as lymphatic endothelial cells isolated from Spred1/; Spred-2/ mice exhibited greater levels of ERK phosphorylation in response to VEGF-C stimulation compared to their wild-type counterparts. Loss of Spred1 and Spred2 function therefore readily accounts for the hyperplastic lymphatic vessel phenotype observed in Spred-1/; Spred-2/ mice, but does not account for the phenotypes of edema and blood-filled lymphatic vessels. Whether these effects are due to hyperactive signaling via VEGFR-3, or alternative signaling pathways, remains to be explored.

VII.  Separation of the lymphatic   and blood vascular networks VII.A.  Syk, SLP-76 and PLC2 The importance of executing separation of the blood and lymphatic vascular networks was realized relatively recently on characterization of the phenotype of mice harboring mutations in the intracellular signaling molecules Syk, SLP-76 or PLC2 (Table 1) (Abtahian et al., 2003). Slp-76/ mice exhibit abnormal arterial–venous connections, as well as abnormal connections between lymphatic and blood vessels, revealing that lymphatic vessels and blood vessels fail to maintain separate circulatory networks in this mutant background. In addition to the presence of lymph within the mesenteric blood vessels of postnatal Slp-76/ pups, erythrocytes were present within the lymphatic vessels of Slp-76/ and Syk/ embryos. Particularly intriguing was the observation that the SLP-76 signaling adaptor molecule was not expressed at detectable levels in embryonic or adult endothelial cells, but appeared

559

Chapter  |  8.4  Lymphatic Vascular Development

to be selectively expressed in hemopoietic cells that circulate throughout the vasculature (Abtahian et al., 2003). In addition, the phenotype of lymphatic and blood vascular mixing could be conferred to irradiated wild-type mice following the transplant of slp76/ bone marrow (Abtahian et al., 2003). How can genes expressed in hemopoietic, but not endothelial, cells confer a vascular phenotype? One possibility is that circulating endothelial progenitor cells expressing Slp-76 and Syk are required for genesis of the lymphatic and blood vasculature, at least to effect the separation of these circulatory systems. However, this possibility is actively debated, with conflicting data supporting (Sebzda et al., 2006), or negating (Srinivasan et al., 2007), the contribution of hemopoietic cells to the genesis of lymphatic vascular endothelium. The definitive answer to this question is still not apparent, but is an intriguing topic in the field of lymphatic vascular development.

VII.A.i.  Fiaf Intriguingly, a defect in partitioning of the blood and lymphatic vascular networks is also observed in mice deficient in fasting-induced adipose factor (Fiaf) (Table 1) (Backhed et al., 2007). The lymphatic/blood vascular connections present in this mouse mutant are not global, as with Slp76/ and Syk/ mutants, but are restricted to the postnatal intestinal vasculature. The onset of the blood-filled lymphatic phenotype in Fiaf/ pups at postnatal day 2 but not prior to this stage of development, suggests perhaps that signals are continuously required for the maintenance of separate blood and lymphatic vascular networks.

VIII.  How does lymphatic vascular development go wrong in disease? How do the genes and signals discussed above together orchestrate construction of the lymphatic vasculature, and how do these processes go wrong in human disease? Our knowledge to date indicates that development of the lymphatic vasculature can be broken down into discrete stages, at least in the mouse: these stages are lymphatic endothelial cell fate specification; formation of the embryonic lymphatic sacs and visceral vessels; generation of the superficial lymphatic vascular plexus; and remodeling and maturation of the lymphatic vasculature. To date, the earliest indication that lymphatic endothelial cell fate has been specified is the onset of Prox1 expression in a subpopulation of cells in the embryonic cardinal veins, this occurs at approximately E9.5–E10.0 in the mouse embryo (Fig. 12) (Wigle and Oliver, 1999) and is dependent on Sox18 (Francois et al., 2008). Following the initiation of Prox1 expression in the cardinal veins, Prox1-positive endothelial cells begin to bud and migrate centrifugally from the veins to form the embryonic lymph sacs; this process occurs in a temporal fashion

VEGF-C

CV

CV

CV

E9.5

E10.5-13.5

E14.5

Figure 12  Model of embryonic lymphatic vascular development. The first sign of lymphatic endothelial cell fate commitment is expression of the homebox transcription factor, Prox1, in a polarized population of cells in the embryonic cardinal veins at approximately E9.5 in the mouse. The induction of Prox1 expression is dependent on Sox18 activity. Following the induction of Prox1 expression, Prox1-positive cells bud and migrate from the cardinal veins in a VEGF-C-dependent manner, to form the jugular lymph sacs (E10.5–E14.5). From E14.5 onwards, Prox1 expression in the jugular veins is reduced, and Prox1-positive lymphatic endothelial cells continue to sprout and migrate from the jugular lymph sacs to form the superficial lymphatic plexus.

along the rostral–caudal axis between E11.5–E14.5 in the mouse (van der Putte, 1975; Srinivasan et al., 2007) and is dependent on Vegf-c (Karkkainen et al., 2004). Following the establishment of lymphatic vascular sacs, endothelial cell sprouting, migration and proliferation is proposed to generate the superficial lymphatic vascular plexus that progressively remodels during late stages of mouse embryogenesis/early postnatal stages, to form both the superficial capillary network and the lymphatic collecting vessels that are invested with smooth muscle and valves. These processes are coordinately regulated by the action of genes including FoxC2 (Petrova et al., 2004), Ang2 (Gale et al., 2002) and ephrinB2 (Makinen et al., 2005). Defects in lymphatic development that occur at various developmental stages outlined above are reflected in a variety of human disorders. These include Down syndrome (trisomy 21) and Turner syndrome, in which embryonic jugular lymph sacs and superficial lymphatic vessels are hyperplastic or hypoplastic respectively, resulting in lymphatic vascular dysfunction and fetal nuchal edema (Bekker et al., 2008). The gene or genes responsible for aberrant lymphatic vascular development in these syndromes have not yet been identified, but could be multiple. Congenital lymphoedema syndromes in which lymphatic vascular function is greatly reduced due to the formation of hypoplastic lymphatic vessels have been attributed to inactivating mutations in Vegfr3 (Irrthum et al., 2000; Karkkainen et al., 2000) and Sox18 (Irrthum et al., 2003), while mutations in FoxC2, important for lymphatic vascular remodeling and maturation (Petrova et al., 2004) have been associated with lymphoedemadistichiasis (Fang et al., 2000; Bell et al., 2001; Finegold et al., 2001). To date, human lymphatic vascular disorders have

560

not been associated with mutations in Prox1, although the critical dependence on Prox1 dosage for normal lymphatic vascular development might indicate that mutations in Prox1 are not likely to be compatible with development. In contrast to the situations outlined above that are associated with hypoplastic lymphatic vessels and reduced lymphatic vascular function, are human diseases associated with the aberrant stimulation of postnatal neolymphangiogenesis. These include inflammatory diseases such as psoriasis (Kunstfeld et al., 2004) and inflammatory arthritis (Zhang et al., 2007), as well as inflammatory settings such as the tumor microenvironment. In each of these cases, the stimulation of neolymphangiogenesis has been associated with disease progression, either by further exacerbating the influx of immune cells to inflammatory sites and thereby increasing tissue damage, or by providing a route for the passage of metastasis of tumor cells to lymph nodes and subsequent distant sites in the body. Increasing our knowledge of the signals that regulate lymphatic vessel construction will be crucial in order to develop novel therapeutics able to repair hypoplastic or damaged lymphatic vessels, or conversely, prevent the growth of neolymphangiogenic lymphatic vessels.

IX.  Future perspectives in lymphatic vascular development Despite the rapid recent progress that has been made in delineating the mechanisms by which the lymphatic vasculature is constructed, many questions remain to be answered before we fully-understand lymphatic vascular development. Is the venous pool of lymphatic endothelial progenitor cells the sole source of lymphatic endothelial cells in the mammalian embryo? Do the veins continue to be a source of lymphatic endothelial progenitor cells postnatally? How do lymphatic vessels navigate to their correct location during development? How do lymphatic vessels mediate their separation from blood vessels? What signals are important for directing whether lymphatic vessels will attain a capillary plexus versus collecting lymphatic vessel identity? What are the signals that direct immune cell traffic through the lymphatic vasculature? What signals initiate and direct maturation of the lymphatic vasculature? The answers to these questions will fill important gaps in our knowledge of embryonic lymphatic vascular development, and should provide the basis on which our understanding of the origin of human lymphatic vascular pathologies is improved, as is our potential to better treat human lymphatic vascular diseases.

Acknowledgments Thanks to the members of my laboratory for constructive comments on the manuscript. This work was supported by a RAH Florey Fellowship (NH) and NHMRC grant 453493 (NH).

PART  |  8  Making Vessels

References Abtahian, F., Guerriero, A., Sebzda, E., Lu, M.M., Zhou, R., Mocsai, A., Myers, E.E., Huang, B., Jackson, D.G., Ferrari, V.A., Tybulewicz, V., Lowell, C.A., Lepore, J.J., Koretzky, G.A., Kahn, M.L., 2003. Regulation of blood and lymphatic vascular separation by signaling proteins SLP-76 and Syk. Science 299, 247–251. Achen, M.G., Jeltsch, M., Kukk, E., Makinen, T., Vitali, A., Wilks, A.F., Alitalo, K., Stacker, S.A., 1998. Vascular endothelial growth factor D (VEGF-D) is a ligand for the tyrosine kinases VEGF receptor 2 (Flk1) and VEGF receptor 3 (Flt4). Proc. Natl. Acad. Sci. USA 95, 548–553. Angeli, V., Ginhoux, F., Llodra, J., Quemeneur, L., Frenette, P.S., Skobe, M., Jessberger, R., Merad, M., Randolph, G.J., 2006. B celldriven lymphangiogenesis in inflamed lymph nodes enhances dendritic cell mobilization. Immunity 24, 203–215. Asellius, G., 1627. De Lactibus Sive Lacteis Venis. Mediolani, Milan, Italy. Avraamides, C.J., Garmy-Susini, B., Varner, J.A., 2008. Integrins in angiogenesis and lymphangiogenesis. Nat. Rev. Cancer 8, 604–617. Ayadi, A., Suelves, M., Dolle, P., Wasylyk, B., 2001a. Net, an Ets ternary complex transcription factor, is expressed in sites of vasculogenesis, angiogenesis, and chondrogenesis during mouse development. Mech. Dev. 102, 205–208. Ayadi, A., Zheng, H., Sobieszczuk, P., Buchwalter, G., Moerman, P., Alitalo, K., Wasylyk, B., 2001b. Net-targeted mutant mice develop a vascular phenotype and up-regulate egr-1. Embo. J. 20, 5139–5152. Backhed, F., Crawford, P.A., O’Donnell, D., Gordon, J.I., 2007. Postnatal lymphatic partitioning from the blood vasculature in the small intestine requires fasting-induced adipose factor. Proc. Natl. Acad. Sci. USA 104, 606–611. Baldwin, M.E., Halford, M.M., Roufail, S., Williams, R.A., Hibbs, M.L., Grail, D., Kubo, H., Stacker, S.A., Achen, M.G., 2005. Vascular endothelial growth factor D is dispensable for development of the lymphatic system. Mol. Cell. Biol. 25, 2441–2449. Baldwin, M.E., Roufail, S., Halford, M.M., Alitalo, K., Stacker, S.A., Achen, M.G., 2001. Multiple forms of mouse vascular endothelial growth factor-D are generated by RNA splicing and proteolysis. J. Biol. Chem. 276, 44307–44314. Baluk, P., Tammela, T., Ator, E., Lyubynska, N., Achen, M.G., Hicklin, D.J., Jeltsch, M., Petrova, T.V., Pytowski, B., Stacker, S.A., YläHerttuala, S., Jackson, D.G., Alitalo, K., McDonald, D.M., 2005. Pathogenesis of persistent lymphatic vessel hyperplasia in chronic airway inflammation. J. Clin. Invest. 115, 247–257. Baluk, P., Fuxe, J., Hashizume, H., Romano, T., Lashnits, E., Butz, S., Vestweber, D., Corada, M., Molendini, C., Dejana, E., McDonald, D.M., 2007. Functionally specialized junctions between endothelial cells of lymphatic vessels. J. Exp. Med. 204, 2349–2362. Banerji, S., Ni, J., Wang, S.X., Clasper, S., Su, J., Tammi, R., Jones, M., Jackson, D.G., 1999. LYVE-1, a new homologue of the CD44 glyco­ protein, is a lymph-specific receptor for hyaluronan. J. Cell. Biol. 144, 789–801. Bekker, M.N., van den Akker, N.M., de Mooij, Y.M., Bartelings, M.M., van Vugt, J.M., Gittenberger-de Groot, A.C., 2008. Jugular lymphatic maldevelopment in Turner syndrome and trisomy 21: different anomalies leading to nuchal edema. Reprod. Sci. 15, 295–304. Bell, R., Brice, G., Child, A.H., Murday, V.A., Mansour, S., Sandy, C.J., Collin, J.R., Brady, A.F., Callen, D.F., Burnand, K., Mortimer, P., Jeffery, S., 2001. Analysis of lymphoedema-distichiasis families for FOXC2 mutations reveals small insertions and deletions throughout the gene. Hum. Genet. 108, 546–551.

Chapter  |  8.4  Lymphatic Vascular Development

Birchmeier, C., Birchmeier, W., Gherardi, E., Vande Woude, G.F., 2003. Met, metastasis, motility and more. Nat. Rev. Mol. Cell. Biol. 4, 915–925. Bjorndahl, M., Cao, R., Nissen, L.J., Clasper, S., Johnson, L.A., Xue, Y., Zhou, Z., Jackson, D., Hansen, A.J., Cao, Y., 2005. Insulin-like growth factors 1 and 2 induce lymphangiogenesis in vivo. Proc. Natl. Acad. Sci. USA 102, 15593–15598. Breiteneder-Geleff, S., Matsui, K., Soleiman, A., Meraner, P., Poczewski, H., Kalt, R., Schaffner, G., Kerjaschki, D., 1997. Podoplanin, novel 43kd membrane protein of glomerular epithelial cells, is down-regulated in puromycin nephrosis. Am. J. Pathol. 151, 1141–1152. Cao, R., Bjorndahl, M.A., Religa, P., Clasper, S., Garvin, S., Galter, D., Meister, B., Ikomi, F., Tritsaris, K., Dissing, S., Ohhashi, T., Jackson, D.G., Cao, Y., 2004. PDGF-BB induces intratumoral lymphangiogenesis and promotes lymphatic metastasis. Cancer Cell 6, 333–345. Carman, C.V., Springer, T.A., 2008. Trans-cellular migration: cell–cell contacts get intimate. Curr. Opin. Cell Biol. 20 (5), 533–540. Carmeliet, P., Ferreira, V., Breier, G., Pollefeyt, S., Kieckens, L., Gertsenstein, M., Fahrig, M., Vandenhoeck, A., Harpal, K., Eberhardt, C., Declercq, C., Pawling, J., Moons, L., Collen, D., Risau, W., Nagy, A., 1996. Abnormal blood vessel development and lethality in embryos lacking a single VEGF allele. Nature 380, 435–439. Carmeliet, P., Lampugnani, M.G., Moons, L., Breviario, F., Compernolle, V., Bono, F., Balconi, G., Spagnuolo, R., Oosthuyse, B., Dewerchin, M., Zanetti, A., Angellilo, A., Mattot, V., Nuyens, D., Lutgens, E., Clotman, F., de Ruiter, M.C., Gittenberger-de Groot, A.C., Poelmann, R.E., Lupu, F., Herbert, J.M., Collen, D., Dejana, E., 1999. Targeted deficiency or cytosolic truncation of the VE-cadherin gene in mice impairs VEGFmediated endothelial survival and angiogenesis. Cell 98, 147–157. Casley-Smith, J.R., 1962. The identification of chylomicra and lipoproteins in tissue sections and their passage into jejunal lacteals. J. Cell Biol. 15, 259–277. Casley-Smith, J.R., 1987. The phylogeny of the fine structure of blood vessels and lymphatics: similarities and differences. Lymphology 20, 182–188. Caunt, M., Mak, J., Liang, W.C., Stawicki, S., Pan, Q., Tong, R.K., Kowalski, J., Ho, C., Reslan, H.B., Ross, J., Berry, L., Kasman, I., Zlot, C., Cheng, Z., Le Couter, J., Filvaroff, E.H., Plowman, G., Peale, F., French, D., Carano, R., Koch, A.W., Wu, Y., Watts, R.J., Tessier-Lavigne, M., Bagri, A., 2008. Blocking neuropilin-2 function inhibits tumor cell metastasis. Cancer Cell 13, 331–342. Cermenati, S., Moleri, S., Cimbro, S., Corti, P., Del Giacco, L., Amodeo, R., Dejana, E., Koopman, P., Cotelli, F., Beltrame, M., 2008. Sox18 and Sox7 play redundant roles in vascular development. Blood 111, 2657–2666. Chang, L.K., Garcia-Cardena, G., Farnebo, F., Fannon, M., Chen, E.J., Butterfield, C., Moses, M.A., Mulligan, R.C., Folkman, J., Kaipainen, A., 2004. Dose-dependent response of FGF-2 for lymphangiogenesis. Proc. Natl. Acad. Sci. USA 101, 11658–11663. Chen, H., Chedotal, A., He, Z., Goodman, C.S., Tessier-Lavigne, M., 1997. Neuropilin-2, a novel member of the neuropilin family, is a high affinity receptor for the semaphorins Sema E and Sema IV but not Sema III. Neuron 19, 547–559. Chen, H., Bagri, A., Zupicich, J.A., Zou, Y., Stoeckli, E., Pleasure, S.J., Lowenstein, D.H., Skarnes, W.C., Chedotal, A., Tessier-Lavigne, M., 2000. Neuropilin-2 regulates the development of selective cranial and sensory nerves and hippocampal mossy fiber projections. Neuron 25, 43–56. Clarijs, R., Schalkwijk, L., Hofmann, U.B., Ruiter, D.J., de Waal, R.M., 2002. Induction of vascular endothelial growth factor receptor-3 expression on tumor microvasculature as a new progression marker in human cutaneous melanoma. Cancer Res. 62, 7059–7065.

561

Cursiefen, C., Chen, L., Borges, L.P., Jackson, D., Cao, J., Radziejewski, C., D’Amore, P.A., Dana, M.R., Wiegand, S.J., Streilein, J.W., 2004. VEGF-A stimulates lymphangiogenesis and hemangiogenesis in inflammatory neovascularization via macrophage recruitment. J. Clin. Invest. 113, 1040–1050. Dadras, S.S., Paul, T., Bertoncini, J., Brown, L.F., Muzikansky, A., Jackson, D.G., Ellwanger, U., Garbe, C., Mihm, M.C., Detmar, M., 2003. Tumor lymphangiogenesis: a novel prognostic indicator for cutaneous melanoma metastasis and survival. Am. J. Pathol. 162, 1951–1960. Dagenais, S.L., Hartsough, R.L., Erickson, R.P., Witte, M.H., Butler, M.G., Glover, T.W., 2004. Foxc2 is expressed in developing lymphatic vessels and other tissues associated with lymphedema-distichiasis syndrome. Gene. Expr. Patterns 4, 611–619. Danussi, C., Spessotto, P., Petrucco, A., Wassermann, B., Sabatelli, P., Montesi, M., Doliana, R., Bressan, G.M., Colombatti, A., 2008. Emilin1 deficiency causes structural and functional defects of lymphatic vasculature. Mol. Cell. Biol. 28, 4026–4039. Dejana, E., 2004. Endothelial cell–cell junctions: happy together. Nat. Rev. Mol. Cell. Biol. 5, 261–270. Dellinger, M., Hunter, R., Bernas, M., Gale, N., Yancopoulos, G., Erickson, R., Witte, M., 2008. Defective remodeling and maturation of the lymphatic vasculature in Angiopoietin-2 deficient mice. Dev. Biol. 319, 309–320. Dietrich, T., Onderka, J., Bock, F., Kruse, F.E., Vossmeyer, D., Stragies, R., Zahn, G., Cursiefen, C., 2007. Inhibition of inflammatory lymphangiogenesis by integrin alpha5 blockade. Am. J. Pathol. 171, 361–372. Dobbs, L.G., Williams, M.C., Gonzalez, R., 1988. Monoclonal antibodies specific to apical surfaces of rat alveolar type I cells bind to surfaces of cultured, but not freshly isolated, type II cells. Biochim. Biophys. Acta. 970, 146–156. Dumont, D.J., Jussila, L., Taipale, J., Lymboussaki, A., Mustonen, T., Pajusola, K., Breitman, M., Alitalo, K., 1998. Cardiovascular failure in mouse embryos deficient in VEGF receptor-3. Science 282, 946–949. Fang, J., Dagenais, S.L., Erickson, R.P., Arlt, M.F., Glynn, M.W., Gorski, J.L., Seaver, L.H., Glover, T.W., 2000. Mutations in FOXC2 (MFH1), a forkhead family transcription factor, are responsible for the hereditary lymphedema-distichiasis syndrome. Am. J. Hum. Genet. 67, 1382–1388. Farr, A.G., Berry, M.L., Kim, A., Nelson, A.J., Welch, M.P., Aruffo, A., 1992. Characterization and cloning of a novel glycoprotein expressed by stromal cells in T-dependent areas of peripheral lymphoid tissues. J. Exp. Med. 176, 1477–1482. Ferrara, N., Carver-Moore, K., Chen, H., Dowd, M., Lu, L., O’Shea, K.S., Powell-Braxton, L., Hillan, K.J., Moore, M.W., 1996. Heterozygous embryonic lethality induced by targeted inactivation of the VEGF gene. Nature 380, 439–442. Ferrara, N., Gerber, H.P., LeCouter, J., 2003. The biology of VEGF and its receptors. Nat. Med. 9, 669–676. Finegold, D.N., Kimak, M.A., Lawrence, E.C., Levinson, K.L., Cherniske, E.M., Pober, B.R., Dunlap, J.W., Ferrell, R.E., 2001. Truncating mutations in FOXC2 cause multiple lymphedema syndromes. Hum. Mol. Genet. 10, 1185–1189. Finegold, D.N., Schacht, V., Kimak, M.A., Lawrence, E.C., Foeldi, E., Karlsson, J.M., Baty, C.J., Ferrell, R.E., 2008. HGF and MET mutations in primary and secondary lymphedema. Lymphat. Res. Biol. 6, 65–68. Fischer, C., Jonckx, B., Mazzone, M., Zacchigna, S., Loges, S., Pattarini, L., Chorianopoulos, E., Liesenborghs, L., Koch, M., De Mol, M., Autiero, M., Wyns, S., Plaisance, S., Moons, L., van Rooijen, N., Giacca, M., Stassen, J.M., Dewerchin, M., Collen, D., Carmeliet, P.,

562

2007. Anti-PlGF inhibits growth of VEGF(R)-inhibitor-resistant tumors without affecting healthy vessels. Cell 131, 463–475. Francois, M., Caprini, A., Hosking, B., Orsenigo, F., Wilhelm, D., Browne, C., Paavonen, K., Karnezis, T., Shayan, R., Downes, M., Davidson, T., Tutt, D., Cheah, K.S., Stacker, S.A., Muscat, G.E., Achen, M.G., Dejana, E., Koopman, P., 2008. Sox18 induces development of the lymphatic vasculature in mice. Nature 456, 643–647. Fritz-Six, K.L., Dunworth, W.P., Li, M., Caron, K.M., 2008. Adrenomedullin signaling is necessary for murine lymphatic vascular development. J. Clin. Invest. 118, 40–50. Fruman, D.A., Mauvais-Jarvis, F., Pollard, D.A., Yballe, C.M., Brazil, D., Bronson, R.T., Kahn, C.R., Cantley, L.C., 2000. Hypoglycaemia, liver necrosis and perinatal death in mice lacking all isoforms of phosphoinositide 3-kinase p85 alpha. Nat. Genet. 26, 379–382. Gale, N.W., Prevo, R., Espinosa, J., Ferguson, D.J., Dominguez, M.G., Yancopoulos, G.D., Thurston, G., Jackson, D.G., 2007. Normal lymphatic development and function in mice deficient for the lymphatic hyaluronan receptor LYVE-1. Mol. Cell. Biol. 27, 595–604. Gale, N.W., Thurston, G., Hackett, S.F., Renard, R., Wang, Q., McClain, J., Martin, C., Witte, C., Witte, M.H., Jackson, D., Suri, C., Campochiaro, P.A., Wiegand, S.J., Yancopoulos, G.D., 2002. Angiopoietin-2 is required for postnatal angiogenesis and lymphatic patterning, and only the latter role is rescued by Angiopoietin-1. Dev. Cell 3, 411–423. Giger, R.J., Cloutier, J.F., Sahay, A., Prinjha, R.K., Levengood, D.V., Moore, S.E., Pickering, S., Simmons, D., Rastan, S., Walsh, F.S., Kolodkin, A.L., Ginty, D.D., Geppert, M., 2000. Neuropilin-2 is required in vivo for selective axon guidance responses to secreted semaphorins. Neuron 25, 29–41. Gluzman-Poltorak, Z., Cohen, T., Herzog, Y., Neufeld, G., 2000. Neuropilin-2 is a receptor for the vascular endothelial growth factor (VEGF) forms VEGF-145 and VEGF-165 [corrected]. J. Biol. Chem. 275, 18040–18045. Gordon, E.J., Gale, N.W., Harvey, N.L., 2008. Expression of the hyaluronan receptor LYVE-1 is not restricted to the lymphatic vasculature; LYVE-1 is also expressed on embryonic blood vessels. Dev. Dyn. 237, 1901–1909. Gunn, M.D., Kyuwa, S., Tam, C., Kakiuchi, T., Matsuzawa, A., Williams, L.T., Nakano, H., 1999. Mice lacking expression of secondary lymphoid organ chemokine have defects in lymphocyte homing and dendritic cell localization. J. Exp. Med. 189, 451–460. Gupta, S., Ramjaun, A.R., Haiko, P., Wang, Y., Warne, P.H., Nicke, B., Nye, E., Stamp, G., Alitalo, K., Downward, J., 2007. Binding of ras to phosphoinositide 3-kinase p110alpha is required for ras-driven tumorigenesis in mice. Cell 129, 957–968. Haiko, P., Makinen, T., Keskitalo, S., Taipale, J., Karkkainen, M.J., Baldwin, M.E., Stacker, S.A., Achen, M.G., Alitalo, K., 2008. Deletion of vascular endothelial growth factor C (VEGF-C) and VEGF-D is not equivalent to VEGF receptor 3 deletion in mouse embryos. Mol. Cell. Biol. 28, 4843–4850. Harvey, N.L., Srinivasan, R.S., Dillard, M.E., Johnson, N.C., Witte, M.H., Boyd, K., Sleeman, M.W., Oliver, G., 2005. Lymphatic vascular defects promoted by Prox1 haploinsufficiency cause adult-onset obesity. Nat. Genet. 37, 1072–1081. He, Y., Rajantie, I., Ilmonen, M., Makinen, T., Karkkainen, M.J., Haiko, P., Salven, P., Alitalo, K., 2004. Preexisting lymphatic endothelium but not endothelial progenitor cells are essential for tumor lymphangiogenesis and lymphatic metastasis. Cancer Res. 64, 3737–3740. Hellstrom, M., Kalen, M., Lindahl, P., Abramsson, A., Betsholtz, C., 1999. Role of PDGF-B and PDGFR-beta in recruitment of vascular smooth muscle cells and pericytes during embryonic blood vessel formation in the mouse. Development 126, 3047–3055.

PART  |  8  Making Vessels

Herpers, R., van de Kamp, E., Duckers, H.J., Schulte-Merker, S., 2008. Redundant roles for sox7 and sox18 in arteriovenous specification in zebrafish. Circ. Res. 102, 12–15. Hirakawa, S., Hong, Y.K., Harvey, N., Schacht, V., Matsuda, K., Libermann, T., Detmar, M., 2003. Identification of vascular lineagespecific genes by transcriptional profiling of isolated blood vascular and lymphatic endothelial cells. Am. J. Pathol. 162, 575–586. Hirakawa, S., Kodama, S., Kunstfeld, R., Kajiya, K., Brown, L.F., Detmar, M., 2005. VEGF-A induces tumor and sentinel lymph node lymphangiogenesis and promotes lymphatic metastasis. J. Exp. Med. 201, 1089–1099. Hirashima, M., Sano, K., Morisada, T., Murakami, K., Rossant, J., Suda, T., 2008. Lymphatic vessel assembly is impaired in Aspp1-deficient mouse embryos. Dev. Biol. 316, 149–159. Hong, Y.K., Lange-Asschenfeldt, B., Velasco, P., Hirakawa, S., Kunstfeld, R., Brown, L.F., Bohlen, P., Senger, D.R., Detmar, M., 2004. VEGF-A promotes tissue repair-associated lymphatic vessel formation via VEGFR-2 and the alpha1beta1 and alpha2beta1 integrins. Faseb. J. 18, 1111–1113. Houck, K.A., Leung, D.W., Rowland, A.M., Winer, J., Ferrara, N., 1992. Dual regulation of vascular endothelial growth factor bioavailability by genetic and proteolytic mechanisms. J. Biol. Chem. 267, 26031–26037. Huang, S.S., Liu, I.H., Smith, T., Shah, M.R., Johnson, F.E., Huang, J.S., 2006. CRSBP-1/LYVE-l-null mice exhibit identifiable morphological and functional alterations of lymphatic capillary vessels. FEBS Lett. 580, 6259–6268. Huang Jr, X.Z., Wu, J.F., Ferrando, R., Lee, J.H., Wang, Y.L., Farese, R. V., Sheppard, D., 2000. Fatal bilateral chylothorax in mice lacking the integrin alpha9beta1. Mol. Cell. Biol. 20, 5208–5215. Huntington, G.S., McClure, C.F.W., 1910. The anatomy and development of the jugular lymph sacs in the domestic cat. Am. J. Anat. 10, 177–311. Ichikawa-Shindo, Y., Sakurai, T., Kamiyoshi, A., Kawate, H., Iinuma, N., Yoshizawa, T., Koyama, T., Fukuchi, J., Iimuro, S., Moriyama, N., Kawakami, H., Murata, T., Kangawa, K., Nagai, R., Shindo, T., 2008. The GPCR modulator protein RAMP2 is essential for angiogenesis and vascular integrity. J. Clin. Invest. 118, 29–39. Irjala, H., Johansson, E.L., Grenman, R., Alanen, K., Salmi, M., Jalkanen, S., 2001. Mannose receptor is a novel ligand for L-selectin and mediates lymphocyte binding to lymphatic endothelium. J. Exp. Med. 194, 1033–1042. Irrthum, A., Karkkainen, M.J., Devriendt, K., Alitalo, K., Vikkula, M., 2000. Congenital hereditary lymphedema caused by a mutation that inactivates VEGFR3 tyrosine kinase. Am. J. Hum. Genet. 67, 295–301. Irrthum, A., Devriendt, K., Chitayat, D., Matthijs, G., Glade, C., Steijlen, P.M., Fryns, J.P., Van Steensel, M.A., Vikkula, M., 2003. Mutations in the transcription factor gene SOX18 underlie recessive and dominant forms of hypotrichosis-lymphedema-telangiectasia. Am. J. Hum. Genet. 72, 1470–1478. Jamieson, T., Cook, D.N., Nibbs, R.J., Rot, A., Nixon, C., McLean, P., Alcami, A., Lira, S.A., Wiekowski, M., Graham, G.J., 2005. The chemokine receptor D6 limits the inflammatory response in vivo. Nat. Immunol. 6, 403–411. Johnson, L.A., Prevo, R., Clasper, S., Jackson, D.G., 2007. Inflammationinduced uptake and degradation of the lymphatic endothelial hyaluronan receptor LYVE-1. J. Biol. Chem. 282, 33671–33680. Joukov, V., Pajusola, K., Kaipainen, A., Chilov, D., Lahtinen, I., Kukk, E., Saksela, O., Kalkkinen, N., Alitalo, K., 1996. A novel vascular endothelial growth factor, VEGF-C, is a ligand for the Flt4 (VEGFR3) and KDR (VEGFR-2) receptor tyrosine kinases. Embo. J. 15, 1751.

Chapter  |  8.4  Lymphatic Vascular Development

Joukov, V., Sorsa, T., Kumar, V., Jeltsch, M., Claesson-Welsh, L., Cao, Y., Saksela, O., Kalkkinen, N., Alitalo, K., 1997. Proteolytic processing regulates receptor specificity and activity of VEGF-C. Embo. J. 16, 3898–3911. Kaipainen, A., Korhonen, J., Mustonen, T., van Hinsbergh, V.W., Fang, G.H., Dumont, D., Breitman, M., Alitalo, K., 1995. Expression of the fms-like tyrosine kinase 4 gene becomes restricted to lymphatic endothelium during development. Proc. Natl. Acad. Sci. USA 92, 3566–3570. Kajiya, K., Hirakawa, S., Ma, B., Drinnenberg, I., Detmar, M., 2005. Hepatocyte growth factor promotes lymphatic vessel formation and function. Embo. J. 24, 2885–2895. Kamba, T., Tam, B.Y., Hashizume, H., Haskell, A., Sennino, B., Mancuso, M.R., Norberg, S.M., O’Brien, S.M., Davis, R.B., Gowen, L.C., Anderson, K.D., Thurston, G., Joho, S., Springer, M. L., Kuo, C.J., McDonald, D.M., 2006. VEGF-dependent plasticity of fenestrated capillaries in the normal adult microvasculature. Am. J. Physiol. Heart Circ. Physiol. 290, H560–H576. Kampmeier, O.F., 1912. The development of the thoracic duct in the pig. Am. J. Anat. 13, 401–476. Kang, J., Lee, I., 2006. TuJ1 (class III beta-tubulin) as phenotypic marker of lymphatic and venous valves. Cardiovasc. Pathol. 15, 218–221. Karkkainen, M.J., Ferrell, R.E., Lawrence, E.C., Kimak, M.A., Levinson, K.L., McTigue, M.A., Alitalo, K., Finegold, D.N., 2000. Missense mutations interfere with VEGFR-3 signalling in primary lymphoedema. Nat. Genet. 25, 153–159. Karkkainen, M.J., Saaristo, A., Jussila, L., Karila, K.A., Lawrence, E.C., Pajusola, K., Bueler, H., Eichmann, A., Kauppinen, R., Kettunen, M.I., Ylä-Herttuala, S., Finegold, D.N., Ferrell, R.E., Alitalo, K., 2001. A model for gene therapy of human hereditary lymphedema. Proc. Natl. Acad. Sci. USA 98, 12677–12682. Karkkainen, M.J., Haiko, P., Sainio, K., Partanen, J., Taipale, J., Petrova, T.V., Jeltsch, M., Jackson, D.G., Talikka, M., Rauvala, H., Betsholtz, C., Alitalo, K., 2004. Vascular endothelial growth factor C is required for sprouting of the first lymphatic vessels from embryonic veins. Nat. Immunol. 5, 74–80. Kerjaschki, D., Regele, H.M., Moosberger, I., Nagy-Bojarski, K., Watschinger, B., Soleiman, A., Birner, P., Krieger, S., Hovorka, A., Silberhumer, G., Laakkonen, P., Petrova, T., Langer, B., Raab, I., 2004. Lymphatic neoangiogenesis in human kidney transplants is associated with immunologically active lymphocytic infiltrates. J. Am. Soc. Nephrol. 15, 603–612. Kriehuber, E., Breiteneder-Geleff, S., Groeger, M., Soleiman, A., Schoppmann, S.F., Stingl, G., Kerjaschki, D., Maurer, D., 2001. Isolation and characterization of dermal lymphatic and blood endothelial cells reveal stable and functionally specialized cell lin­ eages. J. Exp. Med. 194, 797–808. Kubo, H., Cao, R., Brakenhielm, E., Makinen, T., Cao, Y., Alitalo, K., 2002. Blockade of vascular endothelial growth factor receptor-3 signaling inhibits fibroblast growth factor-2-induced lymphangiogenesis in mouse cornea. Proc. Natl. Acad. Sci. USA 99, 8868–8873. Kuhnert, F., Campagnolo, L., Xiong, J.W., Lemons, D., Fitch, M.J., Zou, Z., Kiosses, W.B., Gardner, H., Stuhlmann, H., 2005. Dosagedependent requirement for mouse Vezf1 in vascular system development. Dev. Biol. 283, 140–156. Kume, T., Deng, K.Y., Winfrey, V., Gould, D.B., Walter, M.A., Hogan, B.L., 1998. The forkhead/winged helix gene Mf1 is disrupted in the pleiotropic mouse mutation congenital hydrocephalus. Cell 93, 985–996. Kume, T., Deng, K., Hogan, B.L., 2000. Murine forkhead/winged helix genes Foxc1 (Mf1) and Foxc2 (Mfh1) are required for the early

563

organogenesis of the kidney and urinary tract. Development 127, 1387–1395. Kume, T., Jiang, H., Topczewska, J.M., Hogan, B.L., 2001. The murine winged helix transcription factors, Foxc1 and Foxc2, are both required for cardiovascular development and somitogenesis. Gene. Dev. 15, 2470–2482. Kunstfeld, R., Hirakawa, S., Hong, Y.K., Schacht, V., LangeAsschenfeldt, B., Velasco, P., Lin, C., Fiebiger, E., Wei, X., Wu, Y., Hicklin, D., Bohlen, P., Detmar, M., 2004. Induction of cutaneous delayed-type hypersensitivity reactions in VEGF-A transgenic mice results in chronic skin inflammation associated with persistent lymphatic hyperplasia. Blood 104, 1048–1057. Leak, L.V., 1971. Studies on the permeability of lymphatic capillaries. J. Cell. Biol. 50, 300–323. Leak, L.V., 1980. Lymphatic removal of fluids and particles in the mammalian lung. Environ. Health Perspect. 35, 55–76. Leak, L.V., Burke, J.F., 1968. Ultrastructural studies on the lymphatic anchoring filaments. J. Cell Biol. 36, 129–149. Lee, S., Jilani, S.M., Nikolova, G.V., Carpizo, D., Iruela-Arispe, M.L., 2005. Processing of VEGF-A by matrix metalloproteinases regulates bioavailability and vascular patterning in tumors. J. Cell Biol. 169, 681–691. Lee, S., Chen, T.T., Barber, C.L., Jordan, M.C., Murdock, J., Desai, S., Ferrara, N., Nagy, A., Roos, K.P., Iruela-Arispe, M.L., 2007. Autocrine VEGF signaling is required for vascular homeostasis. Cell 130, 691–703. Lindahl, P., Johansson, B.R., Leveen, P., Betsholtz, C., 1997. Pericyte loss and microaneurysm formation in PDGF-B-deficient mice. Science 277, 242–245. Lord, R.S., 1968. The white veins: conceptual difficulties in the history of the lymphatics. Med. Hist. 12, 174–184. Maisonpierre, P.C., Suri, C., Jones, P.F., Bartunkova, S., Wiegand, S.J., Radziejewski, C., Compton, D., McClain, J., Aldrich, T.H., Papadopoulos, N., Daly, T.J., Davis, S., Sato, T.N., Yancopoulos, G. D., 1997. Angiopoietin-2, a natural antagonist for Tie2 that disrupts in vivo angiogenesis. Science 277, 55–60. Makinen, T., Adams, R.H., Bailey, J., Lu, Q., Ziemiecki, A., Alitalo, K., Klein, R., Wilkinson, G.A., 2005. PDZ interaction site in ephrinB2 is required for the remodeling of lymphatic vasculature. Gene. Dev. 19, 397–410. Martinez de la Torre, Y., Locati, M., Buracchi, C., Dupor, J., Cook, D.N., Bonecchi, R., Nebuloni, M., Rukavina, D., Vago, L., Vecchi, A., Lira, S.A., Mantovani, A., 2005. Increased inflammation in mice deficient for the chemokine decoy receptor D6. Eur. J. Immunol. 35, 1342–1346. Marttila-Ichihara, F., Turja, R., Miiluniemi, M., Karikoski, M., Maksimow, M., Niemela, J., Martinez-Pomares, L., Salmi, M., Jalkanen, S., 2008. Macrophage mannose receptor on lymphatics controls cell trafficking. Blood 112, 64–72. Maruyama, K., Ii, M., Cursiefen, C., Jackson, D.G., Keino, H., Tomita, M., Van Rooijen, N., Takenaka, H., D’Amore, P.A., Stein-Streilein, J., Losordo, D.W., Streilein, J.W., 2005. Inflammation-induced lymphangiogenesis in the cornea arises from CD11b-positive macrophages. J. Clin. Invest. 115, 2363–2372. McColl, B.K., Baldwin, M.E., Roufail, S., Freeman, C., Moritz, R. L., Simpson, R.J., Alitalo, K., Stacker, S.A., Achen, M.G., 2003. Plasmin activates the lymphangiogenic growth factors VEGF-C and VEGF-D. J. Exp. Med. 198, 863–868. McColl, B.K., Paavonen, K., Karnezis, T., Harris, N.C., Davydova, N., Rothacker, J., Nice, E.C., Harder, K.W., Roufail, S., Hibbs, M.L.,

564

Rogers, P.A., Alitalo, K., Stacker, S.A., Achen, M.G., 2007. Proprotein convertases promote processing of VEGF-D, a critical step for binding the angiogenic receptor VEGFR-2. Faseb. J. 21, 1088–1098. Mellor, R.H., Brice, G., Stanton, A.W., French, J., Smith, A., Jeffery, S., Levick, J.R., Burnand, K.G., Mortimer, P.S., 2007. Mutations in FOXC2 are strongly associated with primary valve failure in veins of the lower limb. Circulation 115, 1912–1920. Mishima, K., Watabe, T., Saito, A., Yoshimatsu, Y., Imaizumi, N., Masui, S., Hirashima, M., Morisada, T., Oike, Y., Araie, M., Niwa, H., Kubo, H., Suda, T., Miyazono, K., 2007. Prox1 induces lymphatic endothelial differentiation via integrin alpha9 and other signaling cascades. Mol. Biol. Cell 18, 1421–1429. Morisada, T., Oike, Y., Yamada, Y., Urano, T., Akao, M., Kubota, Y., Maekawa, H., Kimura, Y., Ohmura, M., Miyamoto, T., Nozawa, S., Koh, G.Y., Alitalo, K., Suda, T., 2005. Angiopoietin-1 promotes LYVE-1-positive lymphatic vessel formation. Blood 105, 4649–4656. Mouta Carreira, C., Nasser, S.M., di Tomaso, E., Padera, T.P., Boucher, Y., Tomarev, S.I., Jain, R.K., 2001. LYVE-1 is not restricted to the lymph vessels: expression in normal liver blood sinusoids and downregulation in human liver cancer and cirrhosis. Cancer Res. 61, 8079–8084. Nagy, J.A., Vasile, E., Feng, D., Sundberg, C., Brown, L.F., Detmar, M.J., Lawitts, J.A., Benjamin, L., Tan, X., Manseau, E.J., Dvorak, A.M., Dvorak, H.F., 2002. Vascular permeability factor/vascular endothelial growth factor induces lymphangiogenesis as well as angiogenesis. J. Exp. Med. 196, 1497–1506. Nibbs, R.J., Kriehuber, E., Ponath, P.D., Parent, D., Qin, S., Campbell, J. D., Henderson, A., Kerjaschki, D., Maurer, D., Graham, G.J., Rot, A., 2001. The beta-chemokine receptor D6 is expressed by lymphatic endothelium and a subset of vascular tumors. Am. J. Pathol. 158, 867–877. Nicosia, R.F., Nicosia, S.V., Smith, M., 1994. Vascular endothelial growth factor, platelet-derived growth factor, and insulin-like growth factor-1 promote rat aortic angiogenesis in vitro. Am. J. Pathol. 145, 1023–1029. Ny, A., Koch, M., Schneider, M., Neven, E., Tong, R.T., Maity, S., Fischer, C., Plaisance, S., Lambrechts, D., Heligon, C., Terclavers, S., Ciesiolka, M., Kälin, R., Man, W.Y., Senn, I., Wyns, S., Lupu, F., Brändli, A., Vleminckx, K., Collen, D., Dewerchin, M., Conway, E.M., Moons, L., Jain, R.K., Carmeliet, P., 2005. A genetic Xenopus laevis tadpole model to study lymphangiogenesis. Nat. Med. 11, 998–1004. Ny, A., Koch, M., Vandevelde, W., Schneider, M., Fischer, C., DiezJuan, A., Neven, E., Geudens, I., Maity, S., Moons, L., Plaisance, S., Lambrechts, D., Carmeliet, P., Dewerchin, M., 2008. Role of VEGFD and VEGFR-3 in developmental lymphangiogenesis, a chemicogenetic study in Xenopus tadpoles. Blood 112, 1740–1749. Oliver, G., Sosa-Pineda, B., Geisendorf, S., Spana, E.P., Doe, C.Q., Gruss, P., 1993. Prox 1, a prospero-related homeobox gene expressed during mouse development. Mech. Dev. 44, 3–16. Pennisi, D., Gardner, J., Chambers, D., Hosking, B., Peters, J., Muscat, G., Abbott, C., Koopman, P., 2000. Mutations in Sox18 underlie cardiovascular and hair follicle defects in ragged mice. Nat. Genet. 24, 434–437. Petrova, T.V., Makinen, T., Makela, T.P., Saarela, J., Virtanen, I., Ferrell, R. E., Finegold, D.N., Kerjaschki, D., Ylä-Herttuala, S., Alitalo, K., 2002. Lymphatic endothelial reprogramming of vascular endothelial cells by the Prox-1 homeobox transcription factor. Embo. J. 21, 4593–4599. Petrova, T.V., Karpanen, T., Norrmen, C., Mellor, R., Tamakoshi, T., Finegold, D., Ferrell, R., Kerjaschki, D., Mortimer, P., Ylä-Herttuala,

PART  |  8  Making Vessels

S., Miura, N., Alitalo, K., 2004. Defective valves and abnormal mural cell recruitment underlie lymphatic vascular failure in lymphedema distichiasis. Nat. Med. 10, 974–981. Prevo, R., Banerji, S., Ferguson, D.J., Clasper, S., Jackson, D.G., 2001. Mouse LYVE-1 is an endocytic receptor for hyaluronan in lymphatic endothelium. J. Biol. Chem. 276, 19420–19430. Pullinger, B., Florey, H., 1935. Some observations on the structure and function of lymphatics: their behaviour in local oedema. Br. J. Exp. Path. 16, 49–61. Rissanen, T.T., Markkanen, J.E., Gruchala, M., Heikura, T., Puranen, A., Kettunen, M.I., Kholova, I., Kauppinen, R.A., Achen, M.G., Stacker, S.A., Alitalo, K., Ylä-Herttuala, S., 2003. VEGF-D is the strongest angiogenic and lymphangiogenic effector among VEGFs delivered into skeletal muscle via adenoviruses. Circ. Res. 92, 1098–1106. Ruddell, A., Mezquita, P., Brandvold, K.A., Farr, A., Iritani, B.M., 2003. B lymphocyte-specific c-Myc expression stimulates early and functional expansion of the vasculature and lymphatics during lymphoma­ genesis. Am. J. Pathol. 163, 2233–2245. Saban, M.R., Memet, S., Jackson, D.G., Ash, J., Roig, A.A., Israel, A., Saban, R., 2004. Visualization of lymphatic vessels through NFkappaB activity. Blood 104, 3228–3230. Sabin, F.R., 1902. On the origin of the lymphatic system from the veins, and the development of the lymph hearts and thoracic duct in the pig. Am. J. Anat. 1, 367–389. Sabin, F.R., 1904. On the development of the superficial lymphatics in the skin of the pig. Am. J. Anat. 3, 183–195. Sakamoto, Y., Hara, K., Kanai-Azuma, M., Matsui, T., Miura, Y., Tsunekawa, N., Kurohmaru, M., Saijoh, Y., Koopman, P., Kanai, Y., 2007. Redundant roles of Sox17 and Sox18 in early cardiovascular development of mouse embryos. Biochem. Biophys. Res. Commun. 360, 539–544. Sappey, M., 1874. Anatomie, Physiologie, Patholgie des vaisseaux Lymphatiques consideres chez L’homme et les Vertebres. A. Delahaye, Paris. Schacht, V., Ramirez, M.I., Hong, Y.K., Hirakawa, S., Feng, D., Harvey, N., Williams, M., Dvorak, A.M., Dvorak, H.F., Oliver, G., Detmar, M., 2003. T1alpha/podoplanin deficiency disrupts normal lymphatic vasculature formation and causes lymphedema. Embo. J. 22, 3546–3556. Schledzewski, K., Falkowski, M., Moldenhauer, G., Metharom, P., Kzhyshkowska, J., Ganss, R., Demory, A., Falkowska-Hansen, B., Kurzen, H., Ugurel, S., Geginat, G., Arnold, B., Goerdt, S., 2006. Lymphatic endothelium-specific hyaluronan receptor LYVE-1 is expressed by stabilin-1  , F4/80  , CD11b macrophages in malignant tumours and wound healing tissue in vivo and in bone marrow cultures in vitro: implications for the assessment of lymphangiogenesis. J. Pathol. 209, 67–77. Schneider, M., Othman-Hassan, K., Christ, B., Wilting, J., 1999. Lymphangioblasts in the avian wing bud. Dev. Dyn. 216, 311–319. Schoppmann, S.F., Birner, P., Stockl, J., Kalt, R., Ullrich, R., Caucig, C., Kriehuber, E., Nagy, K., Alitalo, K., Kerjaschki, D., 2002. Tumorassociated macrophages express lymphatic endothelial growth factors and are related to peritumoral lymphangiogenesis. Am. J. Pathol. 161, 947–956. Sebzda, E., Hibbard, C., Sweeney, S., Abtahian, F., Bezman, N., Clemens, G., Maltzman, J.S., Cheng, L., Liu, F., Turner, M., Tybulewicz, V., Koretzky, G.A., Kahn, M.L., 2006. Syk and Slp-76 mutant mice reveal a cell-autonomous hematopoietic cell contribution to vascular development. Dev. Cell 11, 349–361. Siegfried, G., Basak, A., Cromlish, J.A., Benjannet, S., Marcinkiewicz, J., Chretien, M., Seidah, N.G., Khatib, A.M., 2003. The secretory

Chapter  |  8.4  Lymphatic Vascular Development

p­ roprotein convertases furin, PC5, and PC7 activate VEGF-C to induce tumorigenesis. J. Clin. Invest. 111, 1723–1732. Soker, S., Takashima, S., Miao, H.Q., Neufeld, G., Klagsbrun, M., 1998. Neuropilin-1 is expressed by endothelial and tumor cells as an isoform-specific receptor for vascular endothelial growth factor. Cell 92, 735–745. Srinivasan, R.S., Dillard, M.E., Lagutin, O.V., Lin, F.J., Tsai, S., Tsai, M.J., Samokhvalov, I.M., Oliver, G., 2007. Lineage tracing demonstrates the venous origin of the mammalian lymphatic vasculature. Gene. Dev. 21, 2422–2432. Stacker, S.A., Caesar, C., Baldwin, M.E., Thornton, G.E., Williams, R. A., Prevo, R., Jackson, D.G., Nishikawa, S., Kubo, H., Achen, M.G., 2001. VEGF-D promotes the metastatic spread of tumor cells via the lymphatics. Nat. Med. 7, 186–191. Stahl, P.D., Ezekowitz, R.A., 1998. The mannose receptor is a pattern recognition receptor involved in host defense. Curr. Opin. Immunol. 10, 50–55. Suami, H., Pan, W.R., Mann, G.B., Taylor, G.I., 2008. The lymphatic anatomy of the breast and its implications for sentinel lymph node biopsy: a human cadaver study. Ann. Surg. Oncol. 15, 863–871. Takashima, S., Kitakaze, M., Asakura, M., Asanuma, H., Sanada, S., Tashiro, F., Niwa, H., Miyazaki, Ji,J., Hirota, S., Kitamura, Y., Kitsukawa, T., Fujisawa, H., Klagsbrun, M., Hori, M., 2002. Targeting of both mouse neuropilin-1 and neuropilin-2 genes severely impairs developmental yolk sac and embryonic angiogenesis. Proc. Natl. Acad. Sci. USA 99, 3657–3662. Tammela, T., Saaristo, A., Lohela, M., Morisada, T., Tornberg, J., Norrmen, C., Oike, Y., Pajusola, K., Thurston, G., Suda, T., YläHerttuala, S., Alitalo, K., 2005. Angiopoietin-1 promotes lymphatic sprouting and hyperplasia. Blood 105, 4642–4648. Tammela, T., Zarkada, G., Wallgard, E., Murtomaki, A., Suchting, S., Wirzenius, M., Waltari, M., Hellstrom, M., Schomber, T., Peltonen, R., Freitas, C., Duarte, A., Isoniemi, H., Laakkonen, P., Christofori, G., Ylä-Herttuala, S., Shibuya, M., Pytowski, B., Eichmann, A., Betsholtz, C., Alitalo, K., 2008. Blocking VEGFR-3 suppresses angiogenic sprouting and vascular network formation. Nature 454, 656–660. Taniguchi, K., Kohno, R., Ayada, T., Kato, R., Ichiyama, K., Morisada, T., Oike, Y., Yonemitsu, Y., Maehara, Y., Yoshimura, A., 2007. Spreds are essential for embryonic lymphangiogenesis by regulating vascular endothelial growth factor receptor 3 signaling. Mol. Cell. Biol. 27, 4541–4550. Trzewik, J., Mallipattu, S.K., Artmann, G.M., Delano, F.A., SchmidSchonbein, G.W., 2001. Evidence for a second valve system in lymphatics: endothelial microvalves. Faseb. J. 15, 1711–1717. 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., Chernomorsky, R., Boucher, M., Elsasser, A.L., Esau, L., Zheng, J., Griffiths, J.A., Wang, X., Su, H., Xue, Y., Dominguez, M.G., Noguera, I., Torres, R., Macdonald, L.E., Stewart, A.F., DeChiara, T.M., Yancopoulos, G.D., 2003. High-throughput engineering of the mouse genome coupled with high-resolution expression analysis. Nat. Biotechnol. 21, 652–659. van der Putte, S.C., 1975. The early development of the lymphatic system in mouse embryos. Acta. Morphol. Neerl. Scand. 13, 245–286. Veikkola, T., Jussila, L., Makinen, T., Karpanen, T., Jeltsch, M., Petrova, T.V., Kubo, H., Thurston, G., McDonald, D.M., Achen, M.G., Stacker, S.A., Alitalo, K., 2001. Signalling via vascular endothelial

565

growth factor receptor-3 is sufficient for lymphangiogenesis in transgenic mice. Embo. J. 20, 1223–1231. Vlahakis, N.E., Young, B.A., Atakilit, A., Sheppard, D., 2005. The lymphangiogenic vascular endothelial growth factors VEGF-C and -D are ligands for the integrin alpha9beta1. J. Biol. Chem. 280, 4544–4552. Vlahakis, N.E., Young, B.A., Atakilit, A., Hawkridge, A.E., Issaka, R. B., Boudreau, N., Sheppard, D., 2007. Integrin alpha9beta1 directly binds to vascular endothelial growth factor (VEGF)-A and contributes to VEGF-A-induced angiogenesis. J. Biol. Chem. 282, 15187–15196. Volpert, O., Jackson, D., Bouck, N., Linzer, D.I., 1996. The insulinlike growth factor II/mannose 6-phosphate receptor is required for proliferin-induced angiogenesis. Endocrinology 137, 3871–3876. Weber, M., Blair, E., Simpson, C.V., O’Hara, M., Blackburn, P.E., Rot, A., Graham, G.J., Nibbs, R.J., 2004. The chemokine receptor D6 constitutively traffics to and from the cell surface to internalize and degrade chemokines. Mol. Biol. Cell 15, 2492–2508. Wetterwald, A., Hoffstetter, W., Cecchini, M.G., Lanske, B., Wagner, C., Fleisch, H., Atkinson, M., 1996. Characterization and cloning of the E11 antigen, a marker expressed by rat osteoblasts and osteocytes. Bone 18, 125–132. Wigle, J.T., Oliver, G., 1999. Prox1 function is required for the development of the murine lymphatic system. Cell 98, 769–778. Wigle, J.T., Harvey, N., Detmar, M., Lagutina, I., Grosveld, G., Gunn, M. D., Jackson, D.G., Oliver, G., 2002. An essential role for Prox1 in the induction of the lymphatic endothelial cell phenotype. Embo. J. 21, 1505–1513. Wilting, J., Papoutsi, M., Schneider, M., Christ, B., 2000. The lymphatic endothelium of the avian wing is of somitic origin. Dev. Dyn. 217, 271–278. Winnier, G.E., Hargett, L., Hogan, B.L., 1997. The winged helix transcription factor MFH1 is required for proliferation and patterning of paraxial mesoderm in the mouse embryo. Gene. Dev. 11, 926–940. Winnier, G.E., Kume, T., Deng, K., Rogers, R., Bundy, J., Raines, C., Walter, M.A., Hogan, B.L., Conway, S.J., 1999. Roles for the winged helix transcription factors MF1 and MFH1 in cardiovascular development revealed by nonallelic noncomplementation of null alleles. Dev. Biol. 213, 418–431. Wirzenius, M., Tammela, T., Uutela, M., He, Y., Odorisio, T., Zambruno, G., Nagy, J.A., Dvorak, H.F., Ylä-Herttuala, S., Shibuya, M., Alitalo, K., 2007. Distinct vascular endothelial growth factor signals for lymphatic vessel enlargement and sprouting. J. Exp. Med. 204, 1431–1440. Witmer, A.N., van Blijswijk, B.C., van Noorden, C.J., Vrensen, G.F., Schlingemann, R.O., 2004. In vivo angiogenic phenotype of endothelial cells and pericytes induced by vascular endothelial growth factorA. J. Histochem. Cytochem. 52, 39–52. Yaniv, K., Isogai, S., Castranova, D., Dye, L., Hitomi, J., Weinstein, B. M., 2006. Live imaging of lymphatic development in the zebrafish. Nat. Med. 12, 711–716. Yoffey, J.M., Courtice, F.C., 1970. Lymphatics, Lymph and the Lymphomyeloid Complex. Academic Press, London–New York. Yuan, L., Moyon, D., Pardanaud, L., Breant, C., Karkkainen, M.J., Alitalo, K., Eichmann, A., 2002. Abnormal lymphatic vessel development in neuropilin 2 mutant mice. Development 129, 4797–4806. Zhang, Q., Lu, Y., Proulx, S.T., Guo, R., Yao, Z., Schwarz, E.M., Boyce, B.F., Xing, L., 2007. Increased lymphangiogenesis in joints of mice with inflammatory arthritis. Arthritis. Res. Ther. 9, R118.

Chapter 9.1

NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis David A. Elliott1, Edwin P. Kirk2,3,4, Daniel Schaft2 and Richard P. Harvey2,5 1

Monash Immunology and Stem Cell Laboratories, Monash University, Clayton, VIC, Australia Victor Chang Cardiac Research Institute, Sydney, NSW, Australia 3 Department of Medical Genetics, Sydney Children’s Hospital 4 School of Women’s and Children’s Health, Faculty of Medicine, University of New South Wales 5 Faculties of Science and Medicine, University of New South Wales, Kensington, Australia 2

I.  Introduction Evolution of the mammalian heart from the simple contractile tube used in ancestral species for nutrient dispersal into a complex, multi-chambered organ as seen in mammals, relied on a conserved network of transcription factors. The discovery of these factors has been an area of intense interest over the last two decades. The identification of genes encoding cardiac transcription factors from a number of invertebrate and vertebrate species has revealed an underlying genetic architecture of cardiac development that dates back over 500 million years to prior to the emergence of the bilateria (Cripps and Olson, 2002; Davidson and Erwin, 2006). An ancient gene regulatory network (GRN) acts to specify the regulatory state within

(A)

a cohort of cells, and further participates in morphogenesis of the final organ. Importantly, in terms of the evolution of cardiac structures, the core transcriptional network, or “kernel”, is inviolate, as removal of one component of the kernel leads to catastrophic effects on heart development (Davidson and Erwin, 2006). In retrospect, cardiac developmental biologists have benefited greatly from the observation that, despite the gross morphological variation in the hearts of different phyla (Fig. 1), the fundamental genetic circuitry is relatively inflexible and, therefore, highly conserved. A prominent example of this was the discovery that members of the NK-2 class of homeobox genes are essential for cardiac development in both Drosophila and vertebrates (Azpiazu and Frasch, 1993; Bodmer, 1993; Lyons et al., 1995; Cleaver et al., 1996).

(B)

h

T

RA

LA

Pc RV LV 1mm Figure 1  Invertebrate and vertebrate heart morphology. (A) Scanning electron micrograph of a Drosophila dorsal vessel (heart) showing cardiac cells (h) forming the heart tube and pericardial cells (pc) whose function is unknown (see Volume I, Chapter 1.2) (T: Trachea). (B) Scanning electron micrograph of a dissected chick heart highlighting the relative complexity of remodeling of the embryonic heart tube. All four cardiac chambers are indicated (LV: left ventricle; RV: right ventricle; LA: left atria; RA: right atria). Reproduced with permission of Wiley-Liss, Inc., a subsidiary of John Wiley and Sons, Inc. from (A) Curtis et al. (1999); (B) Butcher et al. (2007). Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

569

570

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

The dorsal vessel, or heart analog, of Drosophila is a simple contractile tube (Fig. 1A) and, while specialized in its own right, can also be considered ancestral to the multichambered vertebrate heart (reviewed in Bodmer, 1995). Tinman is an important regulator of dorsal vessel development (Azpiazu and Frasch, 1993; Bodmer, 1993), and is a founding member of the NK-2 class of homeobox genes (Kim and Nirenberg, 1989). Flies lacking the tin gene fail to develop visceral muscle, some body wall muscles and cardiac cells (Azpiazu and Frasch, 1993; Bodmer, 1993). However, mesoderm formation is unaffected, as dorsal spreading of the mesodermal layer at gastrulation still occurs. Although tin is essential for dorsal mesodermal fate, overexpression of tin under the control of a heat shock promoter does not force ventral mesoderm into a cardiac or visceral fate (Azpiazu and Frasch, 1993; Bodmer, 1993). No master regulator of heart development has been found and tin is, therefore, proposed to maintain the competence of cells to receive external cues, which guide their differentiation and development (Bodmer and Frasch, 1999; Frasch, 1999; Zaffran and Frasch, 2002). The isolation of mammalian homologs of the tinman gene was an important milestone in the genetic and molecular dissection of vertebrate cardiogenesis. Nkx2-5, also referred to as cardiac-specific homeobox gene (Csx), was first identified in screens for mouse homologs of tin (Komuro and Izumo, 1993; Lints et al., 1993). In the mouse, Nkx2-5 is expressed continually in the forming myocardium from 7 d pc (days post coitum), as well as in the pharyngeal floor, including the thyroid diverticulum, tongue myocytes, spleen and stomach (Lints et al., 1993). Nkx2-5 transcripts are initially found in the pharyngeal endoderm, overlying the lateral cardiac progenitor pools but are gradually lost from this tissue as development proceeds. In mammals, Nkx2-5 is the only NK-2 gene expressed throughout cardiogenesis; however, Nkx2-6 is transiently expressed from 8 dpc in the posterior myocardium, sinus venosa and dorsal pericardium and later at 9.5 dpc in the outflow myocardium (Biben et al., 1998). Linkage analysis of a single large family with six children displaying persistent truncus arteriosus (or common arterial trunk) identified a missense mutation (F151L) in the NKX2-6 homeodomain that co-segregated with disease (Heathcote et al., 2005), implying a functional role for NKX2-6 in septation of the cardiac outflow tract into the pulmonary artery and the aorta. In Xenopus, three NK-2 family members, XNkx2-3, XNkx2-5 and XNkx2-10, are expressed in the developing cardio–pharyngeal complex (Tonissen et al., 1994; Evans et al., 1995; Newman and Krieg, 1998). XNkx2-3 and XNkx2-5 expression overlaps during late gastrulation with transcripts for both genes detected in the paired cardiac progenitors and anterior endoderm. As development proceeds, both genes are maintained in the cardiac field and subsequently in fully-­differentiated myocardial tissue (Evans et al., 1995; Cleaver et al., 1996). XNkx2-3 and XNkx2-5 are also expressed together in the pharyngeal region, while XNkx2-3 alone is expressed in the

anterior endoderm of the cement gland (Cleaver et al., 1996). In two other model organisms, chick and zebrafish, multiple NK-2 genes are expressed during cardiogenesis. In the chick, cNkx-2.5 transcripts are detected in the cardiac progenitors and maintained during heart development (Schultheiss et al., 1995; Buchberger et al., 1996). Slightly late, cNkx-2.8 is activated at the linear heart tube stage (Reecy et al., 1997). After looping morphogenesis, cNkx-2.3 expression is also observed in the heart (Buchberger et al., 1996). Zebrafish nkx2.5 is expressed in the myocardium throughout cardiac development from the paired progenitor pool stage (Lee and Breitbart, 1996). A related gene, nkx2.7, is also expressed from this early stage (Lee and Breitbart, 1996). Thus, NK-2 genes are expressed from the earliest stages of cardiac development in all vertebrate species examined. Recent studies have shown that mammalian cardiac progenitors arise from two distinct sources, the first and second heart fields (reviewed in Buckingham et al., 2005) (see Chapter 2.2, Vol. I). The first heart field forms the classical cardiac crescent, differentiating early to form the primary heart tube which is composed largely of left ventricular precursors. Cells from the second heart field contribute to the outflow tract, right ventricle and atria, and potentially to a portion of the left ventricle at its outflow aspect. Nkx2-5 is expressed throughout the first and second heart fields (Stanley et al., 2002), and is essential for the integration of patterning information that leads to specification of all five of the anatomical cardiac regions. While Nkx2-5 is expressed in myocardial cells throughout the heart, it is noticeably absent during early formation of myocardium of the sinus horns (Christoffels et al., 2006), an embryological structure that gives rise to the persistent left caval vein, right superior and inferior caval veins, and sinus venarum. It is also absent from the sinoatrial node. Myocardium lining these inflow regions is derived from mesenchymal progenitor cells that do not express Nkx2-5 during their formation, but are positive for the T-box factor Tbx18 (Christoffels et al., 2006). Nkx2-6 is unlikely to function in the production of these myocardial tissues, as they form normally in Nkx2-6-deficient animals (Tanaka et al., 2000). Thus, while Nkx2-5 and its ancestral genes are highly-conserved components of the cardiac transcriptional program during evolution, a variant of the cardiac kernel may have developed, and indeed other solutions to cardiomyogenesis may have arisen during evolution. Nkx2-5 is, however, likely to be expressed transiently in these caudal precursors at an earlier time, and may have a role in establishing the lineage (Zhou et al., 2008). Nonetheless, the adaptability of the role of Nkx2-5/tin within the cardiogenic program highlights how control pathways can evolve within cardiac constraints. Until recently the adult myocardium was considered to be post-mitotic, with no regenerative capacity. However, this paradigm has been challenged by the isolation of a small population of cardiac stem-like cells from the mature mammalian heart (Beltrami et al., 2003; Oh et al., 2003; Urbanek et al., 2003; Linke et al., 2005; Bearzi et al., 2007) (see

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

Sections 4-6, Vol. II). Cardiac stem cells (CSCs) can be visualized in a supportive niche consisting of myocytes and fibroblasts (Urbanek et al., 2006) and can be identified and enriched using cell surface markers. At least a portion of the CSC population is Linvec-Kitve (Beltrami et al., 2003; Bearzi et al., 2007) and some of these CSCs are self-renewing, clonogenic and differentiate into cardiomyocytes, smooth muscle cells and endothelial cells in vitro, and after cell transplantation in cardiac infarct beds (Beltrami et al., 2003). In the clinical setting, CSCs hold much promise for the development of novel, cell-based therapies both for myocardial repair using endogenous CSCs and for myocardial replacement (reviewed in Ballard and Edelberg, 2007). Nkx2-5 is expressed in a small percentage of cardiac interstitial cells, although whether these cells represent a distinct type of CSCs or merely reflect their developmental journey is unclear. Nevertheless, Nkx2-5 is an important transcription factor during heart development, and Nkx2-5 expression in CSCs suggests hitherto unsuspected functions for Nkx2-5 during adult life. Clearly, Nkx2-5 occupies a central position within the GRN governing cardiac morphogenesis. The ­discovery that mutation in NKX2-5 (and other members of the cardiac GRN) is causative for congenital heart disease heightened interest in this transcriptional network from the medical perspective. This chapter summarizes the multiple functions of the best-characterized NK-2 class homeo­domain proteins, Nkx2–5 and tin, during heart development, and the molecular mechanisms through which they execute their functions.

II.  Molecular nature of NKX2-5 The defining feature of the NK-2 family is the presence of a “homeodomain.” The homeodomain consists, in most cases, of a 60 amino acid DNA-binding motif (Scott et al., 1989; Duboule, 1994). Homeodomains are encoded in the DNA by the “homeobox,” so named because the founding members of this protein family were capable of inducing “homeotic” transformations of body segments, a classic example being the formation of legs in place of antennae induced by misexpression of the D. melanogaster Antennapedia gene (Kaufman et al., 1980; Schneuwly et al., 1987). The homeodomain is an ancient motif found in all eukaryotes. The functional utility of the homeodomain is illustrated by the diverse developmental pathways they control and the conservation, despite primary sequence variation, of the helixturn-helix structure in divergent homeodomains (Gehring et al., 1994; Fraenkel and Pabo, 1998; Gruschus et al., 1999; Passner et al., 1999; Piper et al., 1999). Structural studies on two members of the NK-2 family, TTF-1 and ventral nervous system defective (vnd or NK-2), demonstrate that NK-2 homeodomains also form the conserved helix-turn-helix motif which mediates DNA binding (Viglino et al., 1993; Tsao et al., 1994) and that the

571

third helix of the homeodomain lies in the major groove of the DNA (Gruschus et al., 1997, 1999; Tell et al., 1999). However, the highest affinity DNA-binding site of NK-2 homeodomains contains a 5-CAAG-3 core motif (Guazzi et al., 1990; Damante et al., 1994; Chen and Schwartz, 1995; Sparrow et al., 2000), while the majority of homeodomains bind sequences with a 5-TAAT-3 motif (Duboule, 1994). The NK-2 homeodomain DNA binding site is found in many cardiac promoters and is referred to as an NK element or NKE. Structural analysis of the NK-2 and TTF-1 homeodomains has revealed that, on DNA binding, the third helix extends in length from 10 amino acids to 20 amino acids, moving from a “disordered to ordered state” (Tsao et al., 1994; Gruschus et al., 1999; Tell et al., 1999). This extension has a critical functional implication, as the helix thus formed now includes the tyrosine at position 54 of the homeodomain (Y54), a hallmark of this homeodomain class (Fig. 2A). Once incorporated into the helix III extension, Y54 mediates contact with the major groove of the DNA (Gruschus et al., 1999). Furthermore, replacement of the Y54 of TTF-1 with a methionine, which is the amino acid found in Antp-like homeodomains, converts the DNA binding specificity from the NK-2 form (5-CAAG-3) to an Antp form (5-TAAT-3) (Tell et al., 1999), and the same mutation in the homeodomain of vnd abolishes transcriptional repressor function in a Drosophila transgenic assay (Koizumi et al., 2003). Another important factor for determining DNA binding site specificity may be the angle of helix 3. Gruchus and colleagues (1999) describe the angle of NK-2 helix 3 with respect to helix 2 to be 5° less than that of other homeodomains studied to date. Again, this conformational difference enhances the opportunity of Y54 to make contact with the C-residue that is complementary to the G-residue of the 5-CAAG-3 motif. Thus, the unique nature of the NK-2 homeodomain binding site is due to a combination of the helical extension and an alteration of relative helical angle, which facilitates DNA contacts by the conserved Y54 residue. An early structural divergence of this type, with profound implications for specificity of DNA binding and function, may represent an evolutionarily-important constraint limiting the use of NK2 class homeodomain proteins to cardiogenesis and a few other lineage processes in different metazoa. A second evolutionary constraint on the Nkx2-5 homeodomain is that it mediates interactions with many other transcription factors; for example, GATA4, Tbx5 and 20, CAMTA2 and SRF (discussed in Section III). Therefore, the homeodomain is important as an interface for co-factor contacts, as well as DNA binding. However, such interactions have been demonstrated only in vitro, and the structure of the Nkx2-5 homeodomain in a complex with both DNA and any of the array of co-factors with which it interacts remains to be solved. Further structural analysis is required to reveal which of the conserved residues within the HD are critical for physical interactions with other transcription factors.

GIRAW

(A) NKX2-5

TN

HD

NK2 SD

Nkx2-5 Box

YRD

NKX2-5 mNkx2-5 amphiNk2-tin CnNk2

RRKPRVLFSQAQVYELERRFKQQRYLSAPERDQLASVLKLTSTQVKIWFQNRRYKCKRQR ------------------------------------------------------------------------F---------------------Q------------------------K---I----S--M--GKK--D-K----S--D-I--K-N--P---------K-----K-T tin K------------L---C--RLKK--TGA---II-QK-P-SA-------------S--GD Helix 2

Helix 1

TN

HD

NK2

YRD

N G

Lophotrochozoa

TN

HD

NK2

YRD

G

HD

Ecdysozoa

Cnidaria

(C)

TN

HD

TN

HD

TN

HD

TN

HD

Activation

Repression

Jmj

Nkx2-5

TN

Gro/Tup1 HDAC1

TAD

YRD

% Identity

Example species

Deuterostomes

Protostomes

H3 ext.

Helix 3

NK2 protein structure

Phylogenetic relationship

Amphioxus; Mouse; Human Octopus; Cuttlefish; Squid C. elegans Drosophila (Dipteran insect)

YRD

Mosquito (Dipteran insect)

NK2

YRD

Honeybee (Hymenopteran insect)

NK2

YRD

Hydra; Sea anemone

PKC/PKD signaling

Repression

Class II HDACs CAMTA2 NK2 SD

HD

HDAC1 Sin3A RbAp46/8

Smad1/4

(B)

T

100 96.6 70 66.6

GIRAW

YRD

Nkx2-5 Box

HIPK

BMP signalling

Repression

Repression

Figure 2  Cardiac-class NK2 homeodomain proteins show a high degree of phylogenetic conservation and interact with a range of co-factors to influence transcription. (A) Alignment of homeodomain sequences of cardiac NK2 proteins from human (NKX2-5), mouse (mNkx2-5), amphioxus (amphiNk2-tin), cnidaria (CnNk2) and Drosophila (tin). Note the common presence of tyrosine at amino acid position 54 (red), unique to this homeodomain class and essential for binding site specificity. Location of homeodomain helices 1, 2 and 3 are shown below amino acid sequences. The helix 3 extension (H3 ext.) forms on DNA binding. (B) Phylogenetic view of the domain structure of NK2-class homeodomain proteins. Diagrammatic representation of the domain structure of cardiac NK2 proteins showing the conserved homeodomain (HD), tinman/Nkx2-5 domain (TN), NK2-­specific domain (NK2), YRD (tyrosine rich domain), Nkx2-5 box (N) and GIRAW motif (G), across different phyla. (C) Summary of reported Nkx2-5 interactions with transcriptional co-factors and histone-modification enzymes, and the resultant transcriptional response. Dashed lines represent implied effects (see text). The N-terminal transactivation domain (TAD) is also shown.

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

II.A.  Conserved Domains within   NK-2 Genes The NK-2 family can be grouped into three classes based on the presence of two conserved domains, the tin/Nkx25-domain (TN-domain) and the NK-2 Specific Domain (NK2-SD) (Harvey, 1996). Class I NK-2 proteins such as Nkx2-5 contain both the TN-domain and the NK2-SD, class II is defined by the presence of the TN-domain only and includes tinman, while class III members have neither of these motifs (Fig. 2B). The degree of conservation shown in these domains throughout evolution implies an important role in protein function. At present, the precise molecular details of the function of these domains remain elusive. Most data for the TN-domain support a role for it in mediating protein interactions with the Groucho family of co-repressors, while the NK2-SD may also function as an intramolecular repressor.

II.B.  The TN-Domain In the majority of NK-2 proteins the TN-domain is highlyconserved, suggesting an important role. A key insight into the molecular mechanism of TN-domain function is the observation that it shares homology with a repressor domain found in the Drosophila homeodomain protein engrailed, called EH-1 (Smith and Jaynes, 1996). The EH-1 motif is found in an array of transcription factor families including homeodomain, forkhead, zinc-­finger, T-box and double sex domain proteins (Smith and Jaynes, 1996; Copley, 2005). The widespread presence of this domain in many different transcription factors implies that it mediates interactions with common, or possibly universal, co-factors. Biochemical analysis demonstrates that the EH-1 motif mediates interactions with the members of the Groucho family of co-repressors, which recruit class I histone deacetylases (HDACs) (Tolkunova et al., 1998; Gasperowicz and Otto, 2005) (Fig. 2C). In both yeast two-hybrid and GST-pull down experiments the EH-1 motif interacted with Groucho (Gro), and mutagenesis of the invariant Phe residue in the EH-1 motif severely reduced the GRO-en interaction (Tolkunova et al., 1998). Subsequently, the TN-domains of Nkx2-2 and Nkx6.1 were found to mediate interactions with the Groucho-related protein Grg4 during neural tube specification and patterning (Muhr et al., 2001). Taken together, these findings suggest that NK-2 proteins act as repressors via the TN-domain-dependent recruitment of Groucho corepressors to target promoters. However, Groucho interactions may not be restricted to the TN-domain, as tin also binds to Groucho via the homeodomain in GST-pull-down experiments (Choi et al., 1999). While most data support the notion that the TN-domain acts as a repressor via interaction with the Groucho co-repressor family, limited data also suggest that in some contexts the TN-domain may activate transcription (Masson et al., 1998). Therefore, the TN-domain

573

may function in a modular and context-dependent manner to mediate interactions with co-repressor and co-activator proteins.

II.C.  NK-2 Specific Domain (NK2-SD) The NK-2 specific domain (NK2-SD) is unique to the NK-2 class of homeobox proteins, and is found carboxylterminal to the homeodomain in many NK-2 family members (Harvey, 1996; Evans, 1999). The key feature of the NK2-SD is the hydrophobic core sequence, V/IAVPVLV (Harvey, 1996; Watada et al., 2000), which is flanked by basic amino acids and separated from the homeodomain by a short proline-rich linker polypeptide. The conserved hydrophobic core is postulated to form a “pocket” to which accessory proteins bind (Price et al., 1992). The NK2-SD of Nkx2-2 acts as an intramolecular transcriptional repressor by inhibiting the C-terminal transactivation domain (Watada et al., 2000); furthermore, altering the hydrophobicity of the NK2-SD by mutagenesis of two residues within the hydrophobic core results in a significant increase in transactivation (Watada et al., 2000). Thus, the NK2-SD masks a strong transactivation domain within the C-terminus of Nkx2-2, and the repressor effect is reliant on the hydrophobic nature of the motif. Recent studies of vnd reveal that a possible mechanism for the intramolecular repressor activity of the NK2-SD is the augmentation of protein interactions between the TN-domain and the Groucho co-repressor (Uhler et al., 2007). However, this has yet to be tested in many NK-2 proteins. Interestingly, in experiments using Nkx2-5 to rescue the tinman phenotype in D. melanogaster, removal of the NK2-SD results in the partial rescue of some cells of the dorsal vessel, which was not seen with the wildtype protein (Ranganayakulu et al., 1998). The enhanced rescue may result from this protein having slightly improved transcriptional activity, because the intramolecular repressor effect of the NK2-SD has been removed. NK2-SD transcriptional repression can also be mediated by the Smad1-dependent recruitment of the histone deacetylase HDAC1 (Kim and Lassar, 2003). Nkx3.2, a regulator of chondrocyte differentiation, binds to phospho-Smad1 via the NK2-SD, and in this manner forms a complex with pSmad1 and 4 (Fig. 2C). While the Nkx3.2 homeodomain physically interacts with HDAC1 in vitro, both the homeodomain and the NK2-SD are required for this interaction and subsequent transcriptional repression by this complex (Kim and Lassar, 2003). Thus, the repressor activity of the NK2-SD, in some contexts, is likely to be regulated by BMP signaling-induced nuclear translocation of Smad1/4 which binds the NK2-SD, augmenting the homeodomain-mediated binding of HDAC1. Several other regions of homology are found within subsets of the NK-2 gene family (see Fig. 2A). One of the most striking is the GIRAW motif found at the C-­terminus

574

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

of all Nkx2-3 and Nkx2-5 homologs (Evans, 1999). Furthermore, a tryptophan residue is also found at the C-terminus of members of the Nkx2-1 and Nkx2-2 subclass of NK-2 proteins. Nothing is known about how this motif functions, however, tryptophan residues have been demonstrated to be important mediators of protein–protein interactions in two classes of transcription factors, the hairy-related bHLH proteins (Paroush et al., 1994; Dawson et al., 1995; Fisher et al., 1996) and the Hox proteins (Passner et al., 1999; Piper et al., 1999). The conserved nature of the C-terminal tryptophan and the surrounding residues suggests that the motif may play an important role in protein interactions. Another conserved motif is termed the “Nkx2-5 Box” (Evans, 1999), which is found in all Nkx2-5 cognates and in a divergent form in Nkx2-3 proteins supporting other sequence and expression data that suggest these genes shared a common ancestor. Perhaps the Nkx2-3 sequence is vestigial and Nkx2-3 proteins no longer require the Nkx2-5 box for function. Little functional data exists for this motif; however, point mutations in the Nkx2-5 box reduce transcriptional activity of the Nkx2-5 C-terminal activation domain by 50% in an heterologous assay (Elliott et al., 2006). Finally, a tyrosine-rich domain (YRD) is found at the C-terminal of the NK2-SD in many NK-2 proteins, including all members of the cardiac clade and those related to bagpipe, involved in the development of visceral muscles (Elliott et al., 2006). A YRD is found in NK2 proteins from cephalopod molluscs and Cnidarians suggesting that this is an ancient motif that has been conserved across 550 million years of evolution. In heterologous transcriptional assays the YRD can function as part of a transcriptional activation domain, and activity was dependent on the presence of the conserved tyrosines. Mutation of all nine conserved tyrosines to alanine (Nkx2-5Y-A:IRESlacZ) by gene-targeting in both wild-type embryonic stem cells and Nkx2-5GFP cells, in which sequences encoding GFP replaced the Nkx2-5 coding sequence in the second allele, demonstrate that the YRD is essential for Nkx2-5 function during cardiogenesis (Elliott et al., 2006). In chimeric embryos generated with the Nkx25GFP/Y-A:IRESlacZ embryonic stem cells, a high level of chimerism blocked cardiogenesis at the cardiac looping stage, with embryos displaying a close phenocopy of Nkx2-5-null embryos (see Section V). High-level Nkx2-5Y-A:IRESlacZ/ chimeras have multiple cardiac malformations including cardiomegaly, a rounded apex, dilated right atria, dilated ventricles with thin and hypertrabeculated walls, finger-like projections extruding from the external surface of the heart and ventricular septal defects. These embryos die in utero, indicating a potent dominant-negative effect of the Y-A mutation in the heterozygous context. Furthermore, hearts of high-level Nkx2-5Y-A:IRESlacZ/ chimeras display absent or hypoplastic tricuspid valves in which the leaflets are partially fused to the interventricular septum or ventricular free wall, a classical feature of Ebstein’s anomaly. Ebstein’s anomaly

is occasionally associated with mutations in NKX2-5 in humans (Benson et al., 1999). The evolutionary constraints preserving motifs such as the TN-Domain, NK2-SD and the YRD suggest that these domains are critical components of Nkx2-5 function. The TN-domain and NK2-SD are important mediators of interactions with the Groucho family of co-repressors. However, the precise molecular mechanisms and contexts through which these and the other conserved domains act are yet to be revealed. The simplest hypothesis is that they govern interactions with specific co-factors, chromatin remodeling proteins or the basal transcriptional machinery. The interactions between the homeodomain and other transcription factors and co-factors are described below (Section III).

II.D.  Post-translational Modifications   of Nkx2-5 Nkx2-5 is post-translationally modified by phosphorylation, acetylation and sumoylation (Kasahara and Izumo, 1999; Li et al., 2007; Wang et al., 2008). Phosphorylated Nkx2-5 is observed in protein extracts from neonatal rat hearts, and a conserved serine at position 27 (S27) of the homeodomain is phosphorylated by casein kinase II (CK2) in vitro (Kasahara and Izumo, 1999). CK2 phosphorylation of S27 increases Nkx2-5 DNA-binding affinity (Kasahara and Izumo, 1999). Furthermore, mutation of this serine to an alanine abolishes Nkx2-5 transcriptional activity on the Nppa promoter and a multimerized NK2 binding site/ promoter combination in transient transfection assays. A second group of kinases, the conserved nuclear serine/threonine homeodomain-interacting protein kinases (HIPKs), are suspected to phosphorylate NK-2 proteins, and can do so in vitro (Kim et al., 1998). A double knockout of HIPK1 and 2 in mice suggests critical roles for these kinases in cell migration and apoptosis (Isono et al., 2006). With respect to NK-2 proteins, HIPK2 augments the DNA binding and transcriptional repressive activity of Drosophila bap (NK3), however, repression is only partially-­dependent on the kinase activity of HIPK2 (Kim et al., 1998). In vertebrate development, blood formation and cardiogenesis occur in adjacent domains of the mesoderm. Hemopoiesis is actively repressed in heart progenitors and the mechanism appears to involve transcriptional repression of the Gata1 gene by Nkx2-5 via an association between Nkx2-5 and HIPK1 (Dan Garry and RPH, unpublished data). Recent studies demonstrate that the p300 co-activator acetylates Nkx2-5 (Li et al., 2007). The Nkx2-5/p300 interaction appears to be strongly conserved, as tin and p300 form a complex in vitro and are able to synergistically activate transcription from the E2 autoregulatory element of the tin gene, which contains two NKEs (Choi et al., 1999), although it is unknown whether tin is a target of the p300 acetyl-­transferase activity. Mutation of two residues, Gln21 and Gln22, reduces

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

Nkx2-5-p300 binding by 50%, suggesting that Nkx2-5 interacts with p300 through the N-terminal activation domain (AD) (Fig. 2C) (Li et al., 2007), as had been previously shown for tin; however, there is no sequence homology between the N-terminal ADs of Nkx2-5 and tin. As for phosphorylation, the exact action of acetylation of NK-2 class homeodomain proteins remains to be determined. In vitro, Nkx2-5 is sumoylated on lysine 51 and this appears essential for DNA binding, transcriptional activity and transcriptional synergy with serum response factor (SRF) (Wang et al., 2008). Sumoylation alters the sedimentation characteristics of Nkx2-5 transcriptional complexes in size exclusion chromatography, suggesting that the modification may represent a key switch for Nkx2-5 transcriptional activity.

III.  Biochemical interactions between nkx2-5 and other members of the cardiac gene regulatory network Biochemical analysis of the cardiac transcription factors has shown that NK-2 homeodomain proteins physically interact with many other members of the cardiac gene regulatory network. Much of our understanding of collaborative transcriptional regulation by members of the cardiac GRN has built on promoter analysis of chamber-restricted genes to identify and characterize critical cis-acting elements (Small and Krieg, 2004). The presence of known transcription factor binding sites within these sequences can reveal putative trans-acting factors, and the promoters can be used to screen for novel transcriptional regulators. A classic example of this approach is the regulation of the cardiac-expressed gene, Nppa, which encodes the atrial natriuretic factor (ANF) (Grepin et al., 1994; Durocher et al., 1996, 1997). In the adult, Nppa expression is also a marker of cardiac wall stress, and its well-characterized promoter has been used to define the pathways underpinning the transcriptional response to hypertrophy (discussed below and in Liu and Olson, 2006). In a developmental enhancer, NKE and GATA sites are closely located within a conserved proximal 250 bp element, suggesting that Nkx2-5 and the GATA factors are part of the genetic regulatory mechanism controlling Nppa expression in the forming atria and atrioventricular canal (Durocher et al., 1996, 1997; Lee et al., 1998; Sepulveda et al., 1998). However, expression in fetal and diseased ventricular myocardium is controlled by more distal elements (Horsthuis et al., 2008).

575

including the vertebrate heart, liver and gut (Molkentin, 2000) (see Chapter 1.2, Vol. II). GATA factors contain two highly-conserved zinc fingers of the Cys-X2-Cys-X17-CysX2-Cys type which bind the DNA sequence, (A/T)GATA(A/ G), from which GATA factors derived their name (Ko and Engel, 1993; Merika and Orkin, 1993). Mammalian Nkx2-5 and GATA-4 interact to synergistically activate transcription from the Nppa promoter in cell culture experiments (Durocher et al., 1997; Lee et al., 1998; Sepulveda et al., 1998). Structural studies demonstrate that the C-terminal zinc-finger binds to the homeodomain of Nkx2-5 (Durocher et al., 1997; Lee et al., 1998; Sepulveda et al., 1998). Interaction with GATA-4 is proposed to cause a conformational change in Nkx2-5 which results in the aminoterminal activation domain becoming exposed (Durocher et al., 1997). Nkx2-5 is also capable of binding GATA-5 and -6, but the transcriptional output varies, implying a role for Nkx2-5 in GATA factor functional specificity. The Nkx25-GATA factor transcriptional complex is important for the regulation of several cardiac genes, including cardiac ankyrin repeat protein (Kuo et al., 1999), Mef-2c (Dodou et al., 2004; von Both et al., 2004) and Gja5 (connexin40) (Linhares et al., 2004). Similarly, NK2 proteins and GATA factors have a role in transcriptional regulation in smooth muscle cells (Nishida et al., 2002). Furthermore, the NK-2 homeodomainGATA factor synergistic interaction appears to be conserved from flies to mammals (Gajewski et al., 1999, 2001). In Drosophila the concomitant overexpression of tin and the Drosophila GATA-4 homolog, pannier, synergistically activates a D-mef2 enhancer-lacZ transgene in ectopic locations (Gajewski et al., 1999, 2001). Characterization of the D-mef 2 cardiac enhancer and genetic analysis has shown that D-mef2 is downstream of both pnr and tin in the cardiac genetic program. It remains to be determined if GATA factors and Nkx2-5 collaborate to regulate expression of Mef-2c in mammals, however, both are required to establish the appropriate expression pattern. Mef-2c expression is downregulated in Nkx2-5-null embryos (Tanaka et al., 1999a). Further, NKE and GATA sites are found in two transcriptional enhancers which control Mef-2c expression in the secondary heart field (Dodou et al., 2004; von Both et al., 2004). However, at least in the secondary heart field, Mef-2c expression is also dependent on two other transcription factors, the LIM-containing homeodomain protein Isl1 (Dodou et al., 2004) and the forkhead factor Foxh1 acting on a distinct enhancer (von Both et al., 2004). Nevertheless, it is clear that GATA proteins and NK-2 homeodomain proteins have key roles in Mef-2c regulation in both insect and mammalian cardiogenesis.

III.A.  GATA Factors

III.B.  Foxh1

Three GATA transcription factors, GATA-4, -5 and -6, are expressed in tissue of mesodermal and endodermal origin,

Foxh1 (or FAST2) contains a forkhead DNA-binding domain and is strongly expressed in the heart, including the second

576

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

heart field, during looping morphogenesis (Weisberg et al., 1998; von Both et al., 2004). Foxh1-insufficiency impairs development of both the right ventricle and outflow tract, which are formed from the second heart field, and specification of the outer curvature as working chamber myocardium (von Both et al., 2004). The Foxh1-mutant phenotype in mice is similar to embryos deficient for the MADS-box protein Mef2c, which is absent in the second heart field region in Foxh1-null embryos. Furthermore, a 132 base pair intronic Mef2c enhancer element, which is sufficient to recapitulate endogenous gene expression in the second heart field, contains a Foxh1 DNA-binding site, an NKE and a SMADbinding element. Foxh1 transcriptional activity is regulated by the TGF--Smad signaling pathway through biophysical interactions with R-Smad2 and R-Smad3 in the nucleus, implying a role for TGF- signaling during second heart field development and cardiac chamber formation (Labbe et al., 1998). Foxh1 and Nkx2-5 synergistically activate the intronic Mef 2c enhancer element, simultaneously occupy their cognate binding sites, and physically interact. Thus, in the second heart field and its derivatives, Nkx2-5 is part of a genetic pathway linking TGF- signaling and myogenic differentiation through the interaction with Foxh1 and subsequent transcriptional control of Mef2c (von Both et al., 2004).

III.C.  T-box Proteins Mutations in TBX5, a T-box transcription factor, are causative for Holt-Oram syndrome, which is characterized by cardiac malformations including atrial septal defect (ASD), ventricular septal defect (VSD) and tetralogy of Fallot (TOF), as well as defects in the anterior preaxial limb (Li et al., 1997; Basson et al., 1997, 1999). A yeast-two hybrid screen for NKX2-5 interacting proteins identified TBX5 as a putative NKX2-5 co-factor (Hiroi et al., 2001); the interaction domains map to the NKX2-5 homeodomain and to the T-box of TBX5 (Hiroi et al., 2001). TBX5 binding to NKX25 involves two amino acids within the T-box, G80 and R237, both of which are mutated in patients with Holt-Oram syndrome (Garg et al., 2003). Furthermore, NKX2-5 and TBX5 display cooperative DNA binding at both cognate sites, and synergistically activate the promoters of the chamber­specific genes Nppa and Gja5 (which encodes Connexin40) (Bruneau et al., 2001; Hiroi et al., 2001). A second T-box protein, Tbx20, also interacts with Nkx2-5 (Stennard et al., 2003). As for Tbx5, the Nkx2-5-Tbx20 physical interaction is coordinated by the homeodomain and T-box domain, respectively. Tbx20 is capable of synergistically activating transcription from the Nppa promoter in collaboration with Nkx2-5 and GATA4 and the Gja5 promoter when overexpressed with Nkx2-5 and GATA5. Taken together, these data support a role for the Nkx2-5 and Tbx pathways in the positive regulation of the cardiogenic transcriptional program. By contrast, Tbx2 appears to function as a transcriptional repressor with a critical role in limiting the ­expression

pattern of Nppa to the working myocardium (Habets et al., 2002). Nkx2-5 and Tbx2 cooperatively repress transcription from the Nppa promoter in tissue culture experiments, and repression requires binding of both Nkx2-5 and Tbx2 to the promoter. Further, the Tbx2–Nkx2-5 repressive complex competes with the positive Nkx2-5–Tbx5 complex at the Nppa promoter. Importantly, mutation of the T-box factorbinding element (TBE) or the NKE in the Nppa promoter transgene removes repression in vivo, resulting in ectopic expression of Nppa in the atrioventricular canal (AVC). These data suggest that a key mechanism restricting Nppa expression to the chamber myocardium is the presence of Tbx2 in nonchamber myocardium including the atrioventricular canal, which competes with Tbx5 resulting in a dominant Nkx2-5–Tbx2 repressor complex in the atrioventricular canal inhibiting development of working myocardium in this region (Habets et al., 2002).

III.D.  Serum Response Factor The broadly-expressed serum response factor, a MADS-box protein, is an Nkx2-5 co-factor. In addition, Nkx2-5 can interact with serum response factor binding sites (Serum Response Elements or SRE), albeit weakly, in the cardiac a-actin promoter (Chen et al., 1996; Chen and Schwartz, 1996). Serum response factor (SRF) and Nkx2-5 are capable of synergistically activating a cardiac -actin promoter reporter gene, and co-activation is dependent on SRF binding of SREs, and subsequent recruitment of Nkx2-5 (Chen et al., 1996; Chen and Schwartz, 1996). SRF recruitment of cardiac transcription factors is not limited to Nkx2-5, but also includes GATA4 (Sepulveda et al., 2002). Furthermore, SRF, Nkx2-5 and GATA4 form ternary complexes in electromobility shift DNA-binding assays (Sepulveda et al., 2002). Nkx2-5 also indirectly regulates cardiac SRF activity through control of the expression of HOP, a small homeodomain-only protein which inhibits the transcriptional activity of SRF by recruitment of class I HDACs (Chen et al., 2002; Shin et al., 2002; Kook et al., 2003). Taken together, these data demonstrate that Nkx2-5 and SRF can act as co-factors, and that Nkx2-5 regulates SRF activity via genetic interactions with HOP.

III.E.  Nkx2-5 and Chromatin Remodeling Factors CAMTA2 is a novel Nkx2-5 co-activator isolated in a screen for activators of the Nppa promoter. Calmodulin binding transcription activators (CAMTAs) are highly-conserved cofactors found in species ranging from plants to humans (Bouche et al., 2002), and there are two CAMTA proteins, CAMTA1 and 2, in mammals (Liu and Olson, 2006). Both CAMTA1 and 2 are expressed in cardiac tissue with CAMTA1 more predominant during embryogenesis and CAMTA2 expressed at a higher level in the mature heart. CAMTA2 and Nkx2-5

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

synergistically activate the Nppa promoter in an NKE sitedependent manner (Song et al., 2006). Furthermore, this interaction relies on a physical interaction between the proteins that is mediated by the conserved CG-1 domain of CAMTA2, and chromatin immunoprecipitation (ChIP) experiments demonstrate that CAMTA2 is present at the Nppa NKE element in native chromatin (Song et al., 2006). Overexpression of CAMTA2 in transgenic animals induces a cardiac overgrowth phenotype which is dependent on Nkx2-5 binding, since mice expressing a mutant CAMTA2 lacking the CG-1 domain maintain normal heart size. CAMTA2null mice are defective for the hypertrophic response to both hormonal and pressure overload stimuli. CAMTA2 functions by mediating binding to the class II histone deacetylase, HDAC5, which is a negative regulator of cardiac growth (Song et al., 2006) (Fig. 2C). During hypertrophic growth, activated protein kinase C (PkC) and protein kinase D signaling promotes nuclear export of HDAC5 (Vega et al., 2004), alleviating the repression of CAMTA2-­dependent hypertrophic response genes. Thus, CAMTAs link the signaling cascades induced by various hypertrophic stimuli to the resultant Nkx2-5-dependent transcriptional response. It remains to be determined whether CAMTA1 plays a similar role during embryonic cardiac development to that described for CAMTA2 in hypertrophy (Liu and Olson, 2006). Nkx2-5 transcriptional activity is inhibited by the Jumonji transcriptional repressor (Fig. 2C). Jumonji (Jmj) proteins bind DNA through an A/T-rich interaction domain (ARID), share a conserved repressor domain (JmjN) and can act as histone demethylases through a conserved JmjC domain (Takeuchi et al., 2006). Homozygote-null mice for Jumonji (Jmj), also called JARID2, display a range of cardiac abnormalities including VSD, thinning of ventricular walls and double-outlet right ventricle (Lee et al., 2000). In Jmj mutants the Nppa gene is downregulated, which led Kim and colleagues to explore whether Jmj regulates Nppa expression through an Nkx2-5 or GATA-4-­dependent mechanism (Lee et al., 2000). The Nppa promoter contains two Jmj binding sites. However, at least in vitro, these sites are not required for Jmj repression of transcription from this promoter. Jmj physically interacts with both Nkx2-5 (Fig. 2C) and GATA4, through the homeobox and zincfinger domains respectively, via its ARID domain, and inhibits the activity of both these transcription factors on the Nppa promoter. Mutants of Jmj that abolish protein– protein binding with Nkx2-5 and GATA4 also prevent transcriptional repression. Therefore, the transcriptional activity of Nkx2-5 may be repressed by a context-­dependent interaction with Jmj during cardiogenesis.

IV.  NKX2-5 and heart disease Congenital heart disease is the most common of all birth defects, occurring in 6 out of every 1000 live births, and is a leading cause of infant mortality (Hoffman and Kaplan,

577

2002). Heart malformations arise from perturbation of the morphogenetic programs controlling heart development, which are regulated at the transcriptional level by the ­evolutionary-conserved cardiac GRN, during embryogenesis. Genetic mapping of familial congenital heart disease has demonstrated a link between mutations in members of the GRN and the disease state (Basson et al., 1997; Schott et al., 1998; Garg et al., 2003). The first example of the use of linkage mapping to find genes causing nonsyndromic congenital heart disease was the identification of mutations in NKX2-5 from families with a history of atrioventricular (AV) conduction block and secundum atrial septal defects (ASD) (Schott et al., 1998). Classical genetic mapping of three families with congenital atrioventricular block and associated ASD demonstrated a linkage of the disease phenotype with a marker located at chromosome 5q35. NKX25 localizes to the same region, and subsequent sequencing revealed sequence changes that cause amino acid changes Met178Thr, Gln170ter and Gln198ter (Fig. 3). The mutations at amino acid 178 and 170 lie within the homeodomain, while the termination at 198 results in a protein with a complete homeodomain, but lacking the entire C-terminus. Further studies have identified some 38 mutations in total within NKX2-5, which are associated with a range of cardiac conditions (Fig. 3 and Table 1). The importance of NKX2-5 for atrial septation is illustrated by the observation that ASDs are associated with 26 of the 38 (68.4%) mutations found to date. Genotype–phenotype correlations also reveal the critical function of NKX2-5 in the formation of the conduction system, with 25 of the 38 (65.7%) identified mutations resulting in atrioventricular block. NKX2-5 mutations are also found in patients with ventricular septal defects (VSDs), Ebstein’s anomaly, double-outlet right ventricle (DORV), hypoplastic left heart syndrome, heterotaxia and tetralogy of Fallot. The spectrum of cardiac abnormalities represented by these patients implies that NKX2-5 has roles in the development of the atrial and ventricular septum, the conduction system and the tricuspid valve. Missense mutations in NKX2-5 are the most common type of mutation associated with congenital heart disease, with 21 cases described. While the 12 missense mutations found within the homeodomain (Fig. 3) are fully-penetrant, there is limited evidence for the pathogenicity of the nine mutations located outside the homeodomain (i.e., K15I, E21Q, Q22P, R25C, A63V, R126C, A127E, A219V and P275T). Evidence of nonpenetrance is available for several of these nine mutations, while others have only been described for the affected individual with no information regarding the status of other family members showing no congenital heart disease in the family history. Pathogenicity may be demonstrated in further studies on other family members, or by analysis of additional families. Alternatively, these mutations may represent benign polymorphisms. However, given the multifactorial nature of congenital heart disease, it remains possible that these mutations influence the disease state under certain genetic or ­environmental ­conditions. For example, the R25C

578

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

R142C

R25C Q22P

∆2bp 75-trunc

E21Q

∆7bp 72-trunc

K15I

A63V

A127E A119S ∆312G A88X Int1

20 aa

Q149X Q160P R161P Ins498 Q170X L171P T178M W185L Ins701 Q187H N188K R189G R190C/H Y191C Q198X ∆605-6 R216C A219V

TN

NK

HD h1

h2

Y256X Y259X C264X P275T ∆291N A323T YRD

h3

Figure 3  Localization of human NKX2-5 mutations shown or suggested to be associated with congenital heart disease. Mutations are plotted on a schematic representation of the NKX2-5 protein showing conserved and/or functional domains (homeodomain: HD; tinman/Nkx2-5 domain: TN; NK2 specific domain: NK; tyrosine rich domain: YRD). Two other conserved domains, the Nkx2-5 box and the GIRAW motif, are indicated by the yellow and green boxes, respectively. The relative positions of the three helices within the homeodomain are shown (h1, h2 and h3) beneath the homeodomain. The dashed line of h3 denotes the extension of helix 3 after DNA binding. Note the clustering of most disease-associated mutations within the HD, and that 28 of the 38 mutations lie within the conserved domains. Schematic of NKX2-5 protein is drawn to scale, scale bar shown on upper right. For further details and references see Table 1.

mutation has been reported in individuals with tetralogy of Fallot; however, the mutation was not found in 50 healthy individuals selected without bias for racial background, but was present in two of 43 healthy African–American individuals (Goldmuntz et al., 2001; McElhinney et al., 2003). Of the seven probands identified with the mutation, five were African–American, one Hispanic and one Caucasian (McElhinney et al., 2003). Of the three families reported with the R25C change, healthy heterozygous parents were observed in two families and the father of the other proband had VSD. At a molecular level, NKX2-5 R25C has slightly reduced DNA-binding affinity compared to the wild-type protein, and dimer formation is also compromised (Kasahara et al., 2000). Therefore, the possibility that the R25C mutation is pathogenic is supported by the in vitro DNA-binding data, as well as the finding of congenital heart disease segregating with the mutation in one family. However, pathogenicity is clearly determined by other genetic interactions, as R25C is observed in healthy individuals. Identification of disease-causing mutations has guided structure–function analysis of NKX2-5 (Kasahara et al., 2000; Zhu et al., 2000; Kasahara and Benson, 2004; Dentice et al., 2006). The overriding feature of all disease-causing homeodomain mutations examined is a reduced ability to bind DNA (Kasahara et al., 2000; Kasahara and Benson, 2004), suggesting that the primary reason for disease progression is an overall reduction in DNA-bound Nkx2-5. Furthermore, several missense mutations in the homeodomain (R142C, L171P, Q187H, R190H) abolish binding to GATA4 (Kasahara and Benson, 2004). Interestingly, mutation of homeodomain

residue Y54 to cysteine (Y191C) only slightly reduces binding to GATA4, and does not affect binding to Tbx5, suggesting the predominant role of Y54 is in DNA binding. Missense mutations within the homeodomain also impair Nkx2-5 dimerization when binding to DNA in vitro (Kasahara and Benson, 2004). The transcriptionally-active C-terminal truncation mutant Q198X causes increased apoptosis when overexpressed in cardiomyocytes, suggesting that the phenotypes (atrioventricular block, ASD, tetralogy of Fallot and VSD) found associated with this mutation may have arisen from increased cell death during cardiogenesis (Zhu et al., 2000). Pathological hypertrophy in the postnatal heart is characterized by reactivation of the fetal genetic program (Oka et al., 2007). Therefore, one may expect that NKX2-5 plays a role in the response to hypertrophy. In a feline right ventricular pressure overload model of hypertrophy, expression levels of both Nkx2-5 and the Nkx2-5 target genes Nppa and a-CA are dramatically increased in the RV when compared to the LV (Thompson et al., 1998). Nkx2-5 expression is also increased in response to adrenoreceptor agonist-induced hypertrophy (Saadane et al., 1999). While Nkx2-5 levels are upregulated in response to hypertrophic stimuli, transgenic overexpression of Nkx2-5 is insufficient to induce hypertrophy despite the activation of the hypertrophic markers, Nppa and BNP (Takimoto et al., 2000). These data suggest that Nkx2-5 may act as a modulator of the hypertrophic growth response controlling transcriptional activity by context-dependent physical interactions with co-factors such as CAMTA and GATA-4 (see Section III). Further functions of Nkx2-5 in the postnatal heart, in

Mutation*

Location

#

ASD

VSD

AV block

Other CHD

CHD nonpenetrance

Comment

Reference

K15I

TN domain

2

1

–

–

–

1

Nonpenetrance – pathological significance?

McElhinney et al., 2003

E21Q

TN domain

3

–

–

–

PVS, RIV

2

Nonpenetrance – pathological significance?

Elliott et al., 2003; Goldmuntz et al., 2001

Q22P

5 CDS

1

–

–

–

TOF

?

No family history of CHD, parents not geneotyped

McElhinney et al., 2003

R25C

5 CDS

15

–

1

–

TOF (6), TA, IAA, HLHS

4

2 families and 7 isolated cases with cardiac anomalies, 2 families with normal cardiac status, but probands with athyreosis of the gland and thyroid ectopy.

Benson et al., 1999; Dentice et al., 2006; Goldmuntz et al., 2001; McElhinney et al., 2003

A63V

5 CDS

1

–

–

–

L-TGA

?

No family history of CHD, parents not geneotyped

McElhinney et al., 2003

215–221

5 CDS

5

5

1

5

situs inversus (poly-splenia)

1

Nonpenetrant individual had AF aged 46 but no AV block or CHD

Watanabe et al., 2002

223–224

5 CDS

4

4

–

5

–

1

Watanabe et al., 2002

A88X

5 CDS

4

3

–

4

Arrythmia

1

Hirayama-Yamada et al., 2005

312G

5 CDS

3

3

1

2

COA

–

Konig et al., 2006

Int1 DSG  1 T

5 CDS

1

–

–

1

–

–

Benson et al., 1999

A119S

5’ CDS

2

–

–

–

–

2

No CHD; proband had severe hypothyroidism

Dentice et al., 2006

A127E

5’ CDS

3

1

–

–

BAV

1

Nonpenetrance, one affected has BAV only; pathological significance?

McElhinney et al., 2003

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

Table 1  Summary of Reported NKX2-5 Mutations

(Continued)

579

580

Table 1  (Continued) Location

#

ASD

VSD

AV block

Other CHD

CHD nonpenetrance

Comment

Reference

R142C

HD

13

10

3

11

TOF, PS, PDA

3

4 subjects had CHD but no AV block, but 3 were children at time of study

Gutierrez-Roelens et al., 2002

Q149X

HD

6

4

3

5

–

1

Benson et al., 1999

Q160P

HD

4

4

–

4

–

–

Rifai et al., 2007

R161P

HD

2

–

–

–

–

2

No CHD; Proband had thyroid ectopy and severe hypothyroidism; pathological significance?

Dentice et al., 2006

498insC

HD

1

1

–

1

–

–

De novo mutation

Sarkozy et al., 2005

Q170X

HD

4

4

–

3

–

–

L171P

HD

9

7

1

9

–

2

7 deceased individuals had CHD, type not confirmed

Kasahara and Benson, 2004

T178M

HD

14

21

2

24

TOF (2), SVAHS, PA

2

3 families of which 1 had 2 unaffected mutation carriers

Elliott et al., 2003; HirayamaYamada et al., 2005; Schott et al., 1998

W185L

HD

3

4

3

2

MVP, LV noncompaction

–

Sarkozy et al., 2005

Q187H

HD

6

6

–

7

Anomalous systemic venous return (2)

1

Gutierrez-Roelens et al., 2002

N188K

HD

5

5

–

5

Ebstein’s anomaly (3)

R189G

HD

5

4

–

5

TV abn

R190C

HD

1

2

–

1

R190H

HD

3

3

2

3

Schott et al., 1998

1 with decreased LV function

Benson et al., 1999

1

3 with decreased LV function

Benson et al., 1999

–

1

Proband’s father died aged 63 of subaracnoid hemorrhage, cousin with ASD not genotyped, uncle not studied

Hirayama-Yamada et al., 2005

–

–

2 deceased individuals had CHD, type not confirmed

Kasahara and Benson, 2004

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Mutation*

HD

1

1

1

1

–

–

De novo mutation

Benson et al., 1999

Q198X

HD

4

8

–

5

MVF

–

2 families

Hosoda et al., 1999; Schott et al., 1998

605–606Del

3 CDS

3

7

2

7

–

–

R216C

NK2SD

1

–

–

–

TOF

1

No family history of CHD, parents not geneotyped

McElhinney et al., 2003

A219V

NK2SD

2

–

–

–

TOF (2)

–

Nonpenetrance – pathological significance?

Goldmuntz et al., 2001

Ins TCCCT 701

3 CDS

2

1

–

1

–

1

McElhinney et al., 2003

Y256X

YRD

5

2

–

5

MVP

2

Gutierrez-Roelens et al., 2002

Y259X

YRD

7

6

3

7

–

1

Benson et al., 1999

C264X

YRD

1

4

–

6

–

2

Ikeda et al., 2002

P275T

3 CDS

1

–

–

–

COA

–

No family history of CHD, parents not geneotyped

McElhinney et al., 2003

291N

Nkx2.5 Box

2

–

–

–

DORV

1

Nonpenetrance – pathological significance?

McElhinney et al., 2003

A323T

GIRAW

1

–

–

–

TOF

?

No family history of CHD, parents not geneotyped

McElhinney et al., 2003

Sarkozy et al., 2005

# Heterozygous: number proven for the heterozygous mutation. Number of individuals with ASD or other congenital heart disease (CHD) may be greater than this, if multiple affected individuals within a pedigree could not be genotyped. ASD: atrial septal defect. VSD: ventricular septal defect. AV block: confirmed AV conduction block. CHD nonpenetrance: confirmed mutation or obligate heterozygote with known normal cardiac status. Numbers in shaded boxes refer to individuals with AV block but structurally normal heart. Other CHD: May overlap with ASD/VSD categories if more than one affected in pedigree (5CDS: 5 coding sequence; 3 CDS: 3 coding sequence; AA: amino acid; BAV: bicuspid aortic valve; COA: coarctation of the aorta; DORV: double-outlet right ventricle; HD: homeodomain; IAA: interrupted aortic arch; L-TGA: transposition of the great arteries; LV: left ventricle; MVF: mitral valve fenestration; MVP: mitral valve prolapse; NK2SD: NK2 specific domain; PA: pulmonary atresia; PS: pulmonary stenosis; PVS: pulmonary valve stenosis; RIV: retroaortic innominate vein; SVAHS: supravalvular stenosis; TA: truncus arteriosus; TOF: tetralogy of Fallot; TV: tricuspid valve abnormality; YRD: tyrosine rich domain). * Mutations are listed according to their distribution along NKX2.5. Multiple independent reports of the same mutation have been combined.

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

Y191C

581

582

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

particular any role in the regulation of CSCs in response to cardiac injury, remain to be determined. Studies in mouse models have been informative about the important role of NKX2-5 in the formation of the conduction system and the atrial septum, in particular. For example, using Nkx2-5-knockout mice as a model for the human condition, Biben et al. (2000) demonstrate that haploinsufficiency results in a higher prevalence of septal defects and a mild conduction defect. Full-blown ASD is only observed in 1% of mutant mice, and the conduction abnormality is limited to a prolongation of the P-R interval of the electrocardiogram. However, heterozygote-mutant mice have a higher incidence of Patent Foramen Ovale (PFO) and atrial septal aneurysm due to septal dysmorphogenesis. Interestingly, genetic background is a major influence on both the number and size of the PFO observed (Biben et al., 2000). Given that PFO and ASD can be considered to exist within a spectrum of septal abnormalities, murine and human Nkx2-5 genes probably play similar roles in septation during cardiogenesis (Biben et al., 2000). Further genome wide association studies identified seven significant and six suggestive modifier loci that influence septal morphology in the mouse that may also have modifier roles in human septal morphogenesis (Kirk et al., 2006). Transgenic mice overexpressing the mutant protein Nkx2-5 I183P under the control of the -MHC promoter develop conduction defects leading to heart block (Kasahara et al., 2001). Further work using mouse models shows that Nkx2-5 is required for the correct formation and maintenance of the cardiac conduction system (see Section V). During embryogenesis NKX2-5 is expressed in the pharyngeal floor, in particular the thyroid primordium (Lints et al., 1993). Congenital hypothyroidism (CH) is a common endocrine disease primarily caused by impaired thyroid gland morphogenesis resulting in the absence, hypoplasia or ectopic localization of the thyroid, classified as thyroid dysgeneses (De Felice and Di Lauro, 2004). The NKX2-5 expression pattern and its established role in guiding cardiac development make it an ideal candidate gene for a role in thyroid dysgeneses (Dentice et al., 2006). A screen of 241 TD patients identified three NKX2-5 mutations in patients with thyroid dysgeneses (Dentice et al., 2006); two of the mutations were novel, A119S and R161P. The third mutation R25C, which had previously been found in patients with congenital heart disease (McElhinney et al., 2003), is not causative as it was also identified in nonaffected controls, but nonetheless it may increase disease susceptibility in a polygenic setting (see above). Biochemical analysis suggests that the A119S and R161P mutations may function as ­dominantnegative proteins. While these findings illustrate that NKX2-5 mutations are associated with thyroid dysgeneses, at least one member of each family identified was clinically normal demonstrating that the penetrance of these mutations in thyroid dysgeneses is relatively low. Nevertheless, in a permissive genetic background NKX2-5 is likely to have a clinically relevant role in thyroid development.

V.  Phenotypes of NKX2-5 mutants   in mice Modern mouse genetics have been a powerful tool for dissecting the role of Nkx2-5 in cardiogenesis. The Nkx2-5 locus is readily amenable to gene targeting by homologous recombination, and both conventional and conditional gene knockouts have been successfully used to elucidate different functions of Nkx2-5. To gain further insights into Nkx2-5 function, the locus was disrupted by the introduction of a neomycin resistance cassette into the homeobox (Lyons et al., 1995). Homozygous-null Nkx2-5 embryos failed to survive until term, dying at 10 dpc due to hemodynamic insufficiency. Nkx2-5 was not essential for initial cardiomyocyte specification or formation of the linear heart tube. However, cardiogenesis in Nkx2-5-mutant embryos is blocked at the looping morphogenesis stage, and mutant hearts showed a primitive primary chamber and a grossly truncated outflow tract (Lyons et al., 1995; Tanaka et al., 1999a). Subsequently, the bHLH transcription factor HAND1 (formerly eHAND) was found to be downregulated in Nkx2-5null embryos (Biben and Harvey, 1997), suggesting that Nkx2-5 is required for correct spatial and temporal expression of HAND1. Indeed, the morphological defects seen in Nkx2-5-null embryos are similar to that found in HAND1 mutants and mutant chimeras (Firulli et al., 1998; Riley et al., 1998). HAND1 plays a critical role in regulating the balance between proliferation and differentiation of myocardial precursors (Risebro et al., 2006). The working myocardium of the heart derives from regions at the outer curvature of the heart tube (Christoffels et al., 2000). Outer curvature markers, including Chisel, Nppa and MLC2v, expressed in the ventricular region, are markedly reduced or absent in homozygous Nkx2-5-null embryos (Lyons et al., 1995; Biben and Harvey, 1997; Tanaka et al., 1999a; Palmer et al., 2001). Thus, differentiation of the outer curvature into the working myocardium is impaired in Nkx2-5-null embryos. Dysregulation of key markers suggests that the linear heart tube in these embryos remains in a primitive state and cannot support further chamber maturation. The Nkx2-5-null phenotype has been confirmed by two subsequent knockin experiments, one in which the entire coding sequence of Nkx2-5 was replaced by sequences encoding lacZ (Tanaka et al., 1999a), and a second which generated a fusion protein between the first 35 amino acids on Nkx2-5 and the green fluorescent protein (GFP) (Biben et al., 2000). Results from the initial knockout studies raise a number of interesting issues. First, while the mouse Nkx2-5 gene was isolated via homology to Drosophila tin, clearly Nkx2-5mutant mice do not have a tin-like phenotype (discussed in Section VII). Formally, genetic/functional redundancy could account for the less severe mammalian phenotype; however, this is unlikely as no other mammalian NK-2 gene has been identified with an expression pattern overlapping that of Nkx2-5 early in cardiogenesis. Second,

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

Nkx2-5 is critical for the correct interpretation of early patterning information received by cardiac progenitors during formation of the heart tube (Harvey et al., 1998). Within the apparently simple linear heart tube there exist anterior– ­posterior (AP) and dorsal–ventral (DV) positioning cues which influence the formation of the heart’s chambers. In Nkx2-5-mutants, the expansion and elaboration of both ventricles is impaired, thus demonstrating that DV patterning is abnormal (Lyons et al., 1995; Harvey et al., 1998; Tanaka et al., 1999a). The ability of cardiomyocytes to respond to patterning and regionalization signals is critical not only for chamber formation, but also during development of the conduction system.

V.A.  Nkx2-5 and the Patterning of the Vertebrate Heart The broad expression pattern of Nkx2-5 in the cardiac progenitors of both the first and second heart fields implies an early role for Nkx2-5 in cardiogenesis. Analysis of a series of transgenes driving regionalized expression of lacZ demonstrates that the second heart field in Nkx2-5-null embryos is severely compromised, resulting in truncation of the outflow tract and right ventricle (Fig. 4A) (Prall et al., 2007). The transcriptional pathways perturbed in Nkx2-5-mutants control cardiac specification and maintenance of the progenitor state (Fig. 4B). Nkx2-5-dependent genes include the transcription factors Tbx5 and Pbx3, the cytokine receptor platelet-derived growth factor receptor a (Pdgfra), a component of the IGF signaling cascade, insulin-like growth factor binding protein (Igfbp5), and the cardiac inducing cytokines Bmp2 and Fgf10. Furthermore, expression of Isl1, a transiently expressed marker of cardiac progenitors which is rapidly lost after cardiomyocyte differentiation (Cai et al., 2003), is maintained in the cardiac crescent and heart tube in Nkx2-5-null embryos. The loss of Nkx2-5 results in an initial two-fold increase in the number of progenitor cells. This was followed by a collapse of SHF numbers due to defects in progenitor cell proliferation. Nkx2-5GFP/GFP homozygotes have much higher levels of GFP per cell than heterozygotes suggesting that Nkx2-5 itself is part of a negative feedback regulatory loop in cardiac progenitors. The outflow tract in Nkx2-5-mutant embryos is severely truncated and narrowed, which may result from either impaired proliferation or recruitment of second heart field cells. Both Bmp2 and phosphorylated Smad1/5/8 are elevated in the precardiac mesoderm of Nkx2-5-null embryos. Nkx2-5 mutants also display dramatically reduced cell proliferation in the dorsal second heart field. Conditional deletion of Smad1 in the anterior mesoderm results in increased cellular proliferation in the second heart field, and subsequent elongation and widening of the outflow tract. Therefore, mutations in Nkx2-5 and Bmp/Smad1 signaling have opposite effects on outflow tract morphogenesis.

583

Furthermore, the second heart field phenotype observed in Nkx2-5GFP/GFP embryos is rescued, at least partially, by the deletion of Smad1, i.e., proliferation is restored to near wild-type levels in Smad1/;Nkx2-5GFP/GFP embryos. In addition, in a hypomorphic model of Nkx2-5 deficiency, Smad1 deletion was able to rescue, to some degree, a suite of cardiac defects which resemble those found in patients with NKX2-5 mutations, including ASD, VSD, double­outlet right ventricle and absent or reduced mitral and tricuspid valves (Fig. 4C–J). Taken together, these data support a model in which Nkx2-5, Bmp2 and phospho-Smad1 act in a negative feedback loop in early cardiogenesis, in particular in the progenitors within the second heart field (Fig. 4B). In this model, secreted BMP activate SMAD signaling which has two effects: (1) the induction of Nkx2-5 expression as part of the larger cardiogenic program; and (2) the suppression of proliferation in progenitors of the second heart field. Subsequently, Nkx2-5 represses Bmp2 expression and Smad1 activation, thereby regulating the delicate temporal balance between cardiac specification and differentiation, and progenitor proliferation. The above studies highlight the important role of Nkx2-5 as a transcriptional repressor in the early phases of cardiogenesis. Its repressor action is likely to be direct, given the number of conserved domains, including the homeodomain implicated in co-repressor binding; unraveling the specific function of these interactions will expand our understanding of early cardiac circuitry. Recent, WHSC1, a histone H3 lysine 36 trimethylase implicated in Wolf-Hirshhorn syndrome, has been linked to the repressive functions of Nkx2-5 on progenitor genes including Pdgfr, TnC and Isl1 (Nimura et al., 2009). The WHSC1 complex contains repressor proteins such as HDAC1. This important finding opens the way for understanding chromatin regulation during heart development (see also Chapters 10.1 and 10.2, Vol. II). The single ventricular chamber seen in homozygous Nkx2-5-null embryos lacks expression of HAND1 (or eHAND) (Biben and Harvey, 1997). HAND2 is a related bHLH protein which is expressed throughout the developing heart, gradually becoming enriched in the right ventricle, and HAND2/ embryos display RV hypoplasia (Srivastava et al., 1997). Compound Nkx2-5/HAND2-mutant embryos, which lack expression of both HAND1 and HAND2, have only a remnant cardiac chamber and molecular characterization of mutants reveals that this is atrial in character (Yamagishi et al., 2001). Expression of two ventricular markers, HET and MLC2v, suggest that in the double mutant embryos a small number of ventricular cells are specified, however, there are too few cells for normal ventricular development to proceed (Yamagishi et al., 2001). Upregulation of endogenous HAND1 expression results in elevated proliferation of cardioblasts, leading to an extended heart tube and extraneous looping (Risebro et al., 2006). The increased cell proliferation, mediated by a block in cell-cycle exit, effectively delays cardiomyocyte differentiation in the outflow

584

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

(A)

mef2c-LacZ

Nkx2-5 GFP/+

Nkx2-5 GFP/GFP

LV

(B)

Input: repression

Input: activation

Smad1/5

Gata 1

Bmp2 TnC Isl1

Nkx2-5

Nppa

SHF progenitor proliferation

Smpx

Etsrp71 Tie2

Other Progenitor genes

Nppa Cn40 Myl2 Id2

Nkx2-5 gfp/IRES-cre/Smad fl/+ Nkx2-5 gfp/IRES-cre/Smad +/+

Heart progenitors and cardiomyoblasts

Endocardial progenitors

Chamber myocardium conduction tissue

(C)

(D)

(E)

(F)

(G)

(H)

(I)

(J)

Figure 4  Murine models of Nkx2-5 deficiency. Heart looping is impaired in Nkx2-5-null mutants due, in part, to the disruption of an Nkx25-Smad1/5-Bmp2 regulatory circuit. (A) Expression of Mef2c-lacZ transgene shows contribution of first heart field cells (unstained for LacZ) and anterior second heart field cells (stained positive for LacZ) to the hearts of E9.0 Nkx2-5 heterozygote (Nkx2-5GFP/) and null (Nkx2-5GFP/GFP) embryos. The mutant heart contains myogenic and endocardial layers, and beats, but morphogenesis is arrested at the linear heart tube stage and differentiation of chamber myocardium and endocardial cushions is blocked. The LacZ pattern demonstrates that deployment of cells from the second heart field is severely compromised (LV: left ventricle; OFT: outflow tract; RV: right ventricle; V: ventricle-like chamber in null embryo). (B) Dual role of Nkx2-5 in positive and negative regulatory circuits in the developing heart. An Nkx2-5-Smad1/5-Bmp2 negative feedback loop controls second heart field progenitor cell proliferation and deployment, while chamber and conduction system specification and differentiation is controlled by transcriptional activation (in collaboration with co-factors). Nkx2-5 also inhibits blood formation via direct repression of GATA1 in the cardiac fields (Dan Garry, personal communication). (C–J) Outflow tract, septal and valve defects in embryos hypomorphic for Nkx2-5 (Nkx2-5gfp/IREScre) are partially rescued by mutation of a single Smad1 allele (Smad1fl/) (see Prall et al., 2007) (Ao: aortic root; ASD: atrial septal defect; DORV: double-outlet right ventricle; LV: left ventricle; MV: mitral valve; OA: overriding aorta; PA: pulmonary artery; RV: right ventricle; TV: tricuspid valve; VSD: ventricular septal defect). Immature atrioventricular cushions in hypomorphs are indicated by an asterisk. Adapted with permission of Cell Press from Prall et al., 2007.

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

tract and left ventricle, disrupting heart patterning (Risebro et al., 2006). Furthermore, cardiomyocyte differentiation in an in vitro embryoid body system was diminished in HAND1 overexpressing cells and five-fold higher in HAND1-null cells. Therefore, the HAND genes are critical regulators of cardiomyocyte proliferation and survival, and HAND1 expression in chamber myocardium is dependent on Nkx2-5. Whether HAND1 is a direct target of Nkx2-5 remains to be seen. Nevertheless, Nkx2-5 regulates proliferation of both progenitor cells (through repression of Bmp2) and cardiomyocytes (via HAND1), and has a profound influence on heart patterning via this mechanism. Independently of its role in the regulation of cardiac progenitor cell proliferation through repression of Bmp2, Nkx2-5 also plays an additional role, apparently as a repressor, in inhibiting the expression of a host of other progenitor genes in differentiating myocardium. It seems that this is a mechanism that ensures the separation of progenitor and differentiated states, a fundamental property of most, if not all, differentiating systems. As discussed further below, the expression signature of Nkx2-5-null cardiomyocytes suggests that they resemble the most caudal and least differentiated myocardium, that of the sinoatrial node and sinus horns.

V.B.  Nkx2-5 in the Establishment and Maintenance of Boundaries Nkx2-5 plays an important role in the formation and maintenance of boundaries between discrete compartments within the developing heart. Examples of this function include the specification and formation of the sinoatrial node and the myocardium of the systemic and pulmonary returns of the developing heart (Christoffels et al., 2006; Mommersteeg et al., 2007a,b). Recent work has shown that the sinus horns, a transient structure that gives rise to the myocardium of the right superior and inferior caval veins and the sinoatrial node, forms after heart looping and is derived from the differentiation of Nkx2-5-negative mesenchyme into myocardium (Christoffels et al., 2006). The juxtaposition of Nkx2-5-negative and Nkx2-5-positive myocardium has implications for the correct patterning of the venous pole of the mammalian heart. The sinoatrial node (SA node) is a specialized myocardial structure that initiates the electrical impulses to stimulate contraction, and is found in the atrial wall at the junction of superior caval vein and the right atrium (Mikawa and Hurtado, 2007). Nkx2-5 plays an important, albeit indirect, role in maintaining the boundary between the SA node and the adjacent atrial myocardium, and the distinct differentiated states of these two sublineages. Genetic experiments demonstrate that Nkx2-5 is required to repress both the pacemaker channel gene Hcn4 and Tbx3 in the atria (Mommersteeg et al., 2007b). The repression of Tbx3 in the atrial myocardium is an important patterning mechanism,

585

since transgenic overexpression of Tbx3 in the developing heart blocks expression of the atrial myocardial markers Nppa and Cx40, i.e., the myocardial molecular phenotype was altered to resemble that of the systemic venous return and SA node. Therefore, Nkx2-5-dependent repression of Tbx3 throughout the atrial myocardium prevents atrial cells adopting a conductive cell phenotype. Furthermore, the position of the SA node is facilitated by the formation and maintenance of a boundary between the Nkx2-5-positive atrial myocardium and that of the sinus horns from which the pacemaker cells of the SA node derive. Later in development Nkx2-5 expression is activated in the SA node, however, these cells remain Hcn4 positive and Cx40 negative, unlike myocardial cells within the sinus horns which appear to adopt an atrial gene program post-Nkx2-5 expression (Mommersteeg et al., 2007b). These data suggest that the molecular program of SA nodal cells is recalcitrant to certain Nkx2-5 pathways, perhaps mediated by the formation of a Tbx3/Nkx2-5 repressor complex in these cells. Defects in the formation of pulmonary venous myocardium underlie congenital sinus venous defects, and atrial fibrillation often originates from ectopic electrical stimulation occurring within the pulmonary veins (Haissaguerre et al., 1998). Cell-labeling experiments show that the pulmonary myocardium develops in two waves, initiated by the de novo differentiation of mesenchyme surrounding the pulmonary vein (Mommersteeg et al., 2007a). The differentiating mesenchyme then undergoes a period of proliferation to form the myocardial sleeve surrounding the pulmonary vein. Importantly, these studies demonstrate that the pulmonary myocardium arises as a distinct and independent tissue, which is not derived from migrating atrial myocardium. The molecular program controlling pulmonary myogenesis includes the homeodomain proteins Pitx2c and Nkx2-5. In Pitx2-null embryos both the sinus venosus and the pulmonary return drain into a common medially-located sinus venosus (Franco and Campione, 2003). Furthermore, in these embryos the pulmonary myocardial precursors are not specified, and subsequently the myocardial sleeve fails to form despite the normal morphology of the lumen of the vein and the surrounding endothelium (Mommersteeg et al., 2007a). Like Pitx2c, Nkx2-5 is expressed in the developing pulmonary myocardium; however, Nkx2-5-null embryos die before the onset of pulmonary myogenesis. To circumvent this, Mommersteeg and colleagues modulated Nkx2-5 expression level using an allelic series consisting of a null allele (Nkx2-5GFP) and a hypomorphic allele of Nkx2-5 (Nkx2-5IRESCre) to generate embryos which survive until term, but die at birth. The pulmonary myocardium forms normally in Nkx2-5IRESCre/GFP embryos. However, the sinus horn marker Hcn4 is expressed throughout the pulmonary myocardium while Cx40, which is normally expressed in the pulmonary myocardium, is completely absent. Therefore, reduced Nkx2-5 levels result in the

586

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

myocardial sleeve of the pulmonary vein adopting a similar phenotype to that of the sinus horn myocardium. These experiments suggest that with reduced Nkx2-5, both chamber and pulmonary myocardium tend towards the default of a conduction tissue fate, and boundaries between chamber and nonchamber myocardium are blurred.

V.C.  Nkx2-5 and the Developing Ventricular Conduction System Coordination of myocardial contraction during the heartbeat is controlled by the conduction system, a specialized network of elements derived from the myocardium (see Chapter 2.3, Vol. I). Electrophysiological activity in the heart is established and propagated via the cardiac conduction system (CCS). The sinoatrial and atrioventricular nodes are highly-conserved anatomical components derived from the inflow and atrial components, respectively (Mikawa and Hurtado, 2007). Electrical signals in the ventricle spread from the atrioventricular node to the ventricular apex via the His-Purkinje network, which consists of an atrioventricular bundle (bundle of His) and the left and right bundle branches (Mikawa and Hurtado, 2007). Nkx2-5 is critical for the specialization of conduction cells from myocardium. The discovery of NKX2-5 mutations underlying cardiac conduction disease (see Section III) and the relatively high levels of Nkx2-5 in presumptive precursors of the ventricular conduction system (Thomas et al., 2001) suggest that differentiation of these specialized cells relies on Nkx2-5. Initial reports hinted that development of the CCS in mice is, in part, Nkx2-5-dependent (Biben et al., 2000). Subsequently, the availability of molecular markers of the atrioventricular conduction system has facilitated the dissection of Nkx2-5 function during formation of the CCS. Cell number in the CCS is related to Nkx2-5 gene dosage (Jay et al., 2004); specifically, Nkx2-5-null mice appear to lack atrioventricular node progenitors, while haploinsufficiency for Nkx2-5 results in the specification of only half the normal number of cells in the atrioventricular node and HisPurkinje network (Jay et al., 2004). The minK-lacZ allele is an early marker of the conduction system (Kupershmidt et al., 1999) which is not expressed in the atrioventricular node region of Nkx2-5-null embryos, suggesting that these cells have not been properly specified in mutant hearts. However, it remains possible that lacZ expression is lacking because Nkx2-5 regulates expression of minK in those cells; in heterozygous Nkx2-5-mice carrying the minK-lacZ allele, -galactosidase staining revealed hypoplasia of the CCS when compared with wild-type mice. Furthermore, in Nkx2-5 heterozygous mice a subpopulation of Cx40 negative/Cx45 positive cells were missing from the atrioventricular node (Jay et al., 2004). The cellular and molecular phenotype described manifests in neonates and adult mice as prolonged

PR interval and prolonged QRS interval, reflecting impaired atrioventricular node and His/Purkinje function. Studies using Cx40-GFP mice to mark the Purkinje fibers also reveal a requirement for correct Nkx2-5 levels for formation and function of the Purkinje network (Meysen et al., 2007). These experiments reveal that Nkx2-5 heterozygotes have hypoplasia of Purkinje fibers and disorganization of the Purkinje network at the ventricular apex, resulting in abnormal ventricular electrical activation. Reduced Nkx2-5 levels cause a delay in cell-cycle withdrawal, and thereby may result in cells fated to form the peripheral conduction system continuing to cycle and adopting a ventricular myocyte fate. The molecular circuitry regulating CCS development is beginning to be revealed, and the Nkx2-5-regulated gene Hop is an important component of the CCS transcriptional network (Ismat et al., 2005). Deletion of Hop results in conduction delay and a decrease in Cx40 expression throughout the CCS, with the greatest reduction at its distal reaches (Ismat et al., 2005). The role of Nkx2-5 during lineage specification and diversification of ventricular myocytes has been addressed using the CRE/lox system to delete Nkx2-5 in the developing ventricle (Pashmforoush et al., 2004). Mice lacking ventricular Nkx2-5 expression develop conduction system disease, with 12-week-old animals displaying first degree atrioventricular block, and 50% of the mutant animals progressing to complete heart block by 12 months. Control mice are completely unaffected. The atrioventricular node and His-bundle of Nkx2-5 ventricular-null newborns are smaller than wild-type, however, the gross morphology of the heart is normal with no appearance of septal abnormalities. Adult mutants have a smaller and atrophic CCS, and fibrosis is evident throughout the conduction system. Three CCS markers, HCN-1, minK and Cx40 are downregulated in the atrioventricular node and distal His-Purkinje system in the Nkx2-5-mutant mice. Ventricular ablation of Nkx2-5 also results in myocardial overgrowth and hypertrabeculation which may be due to an increase in BMP signaling, since Bmp-10 expression is upregulated and maintained into adulthood in the trabeculae of mutant mice. In addition, mutant mice display cardiomyopathy as ventricular ejection fraction was greatly reduced. The phenotype of ventricular Nkx2-5-null mice resembles the human disease condition in which degeneration of the CCS and hypertrabeculation are components (Pashmforoush et al., 2004). The genetic program underlying differentiation of the ventricular conduction cells is, in part, dependent on the cooperative regulation of Id2 by Tbx5 and Nkx2-5 (Moskowitz et al., 2007). The Id proteins (Id1-4) are members of the helixloop-helix family of transcriptional co-repressors, with roles in development, tumorigenesis and senescence (Ruzinova and Benezra, 2003). Id2-deficient mice display both structural and functional conduction system abnormalities, including abnormal His bundles, left bundle branch block and a prolonged QRS interval (Moskowitz et al., 2007). A 1.2 kb promoter

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

Nkx2-5 expression

587

Overlap

Tbx5 expression

Cell-cycle withdrawal

Nkx2-5

Id2 Conduction system genetic program

Tbx5 Cja5, etc.

Figure 5  Nkx2-5 acts in concert with Tbx5 to regulate formation of the cardiac conduction system. Developed expression domains of Nkx2-5 (green) and Tbx5 (yellow) within the midgestation heart are shown (top panels). Nkx2-5 is at highest levels in the ventricular conduction system. Overlapping Nkx2-5 and Tbx5 expression is shown in blue, and is highest in the ventricular conduction system. A model for molecular control of the ventricular conduction system specification is shown below. High and overlapping Nkx2-5 and Tbx5 expression act cooperatively to promote cell-cycle withdrawal in a regionally-restricted manner via activation of Id2. Further, these two cardiac transcription factors activate a suite of conduction system genes, including Gja5. Adapted with permission of Cell Press from Moskowitz et al. (2007).

element of Id2, able to recapitulate endogenous gene expression in transgenic mice, contains NKE, TBE and GATA sites, and is cooperatively regulated by Nkx2-5 and Tbx5 (Lim et al., 2007; Moskowitz et al., 2007). Haploinsufficency for both Tbx5 and Nkx2-5 impairs CCS development and prevents expression of the minK-lacZ marker in the atrioventricular bundle and bundle branch regions. Id proteins have an established role in the cell-cycle through interactions with Retinoblastoma (Rb) and ETS transcription family of ­transcription factors (Iavarone and Lasorella, 2006). The finding that reduced Nkx2-5 levels delays cell-cycle withdrawal (Meysen et al., 2007) can be interpreted in the light of Id2 regulation by Nkx2-5 (Moskowitz et al., 2007). Higher levels of both Tbx5 and Nkx2-5 within conduction system precursor cells results in an upregulation of Id2 which in turn promotes the cell-cycle withdrawal, specification and differentiation of the conduction cells from cardiomyocytes (Fig. 5). Thus, Nkx2-5, in part, regulates formation of the ventricular conduction system through an Id2 dependent differentiation program.

V.D.  Nkx2-5 Regulates Formation of the Endocardium Endocardial cells may form from a common precursor with myocardial cells (Kattman et al., 2006; Moretti et al., 2006). Cre lineage analysis shows that Nkx2-5 is expressed in precursors of endocardial cells when or before they first become evident in-between the myocardiogenic plate and subjacent endoderm (Stanley et al., 2002). Although Nkx2-5 expression does not persist in endocardium, a prominent feature of the Nkx2-5 knockout phenotype is lack of endocardial cushions, despite upregulation of cushion-inducing Bmp (Lyons et al., 1995; Prall et al., 2007). Nkx2-5 has

been shown to directly and positively regulate endocardial expression of the gene for Ets-related protein 71 (Etsrp71), itself essential for endocardial cell formation (Ferdous et al., 2009). Etsrp71 may directly regulate the Tie2 gene, encoding an angiopoietin receptor. These studies document an early and transient role for Nkx2-5 in specification of a nonmyocyte lineage of the heart, and reinforce the notion that Nkx2-5 has multiple roles in the lineage outcomes of multipotent cardiac progenitor cells.

VI.  Studies of NKX2-5 in other vertebrate model systems Overexpression by RNA injection of either XNkx2-3, or XNkx2-5, results in enlargement of the Xenopus heart (Fu and Izumo, 1995; Cleaver et al., 1996). Hearts are approximately 1.8 times larger, and contained 1.5 times as many cells. However, overexpression did not lead to ectopic expression of cardiac markers, suggesting that the enlarged heart phenotype is due to recruitment of more cells from the cardiac field. In addition, hypertrophy of cardiomyocytes is also observed. Injection of dominant-negative NK-2 genes has been used to assess the overall contribution of NK-2 genes to Xenopus cardiogenesis. Since these factors can form dimers, and possibly share co-factors, a dominant negative protein can, in principle, be designed to disrupt all three cardiac NK-2 Xenopus genes (Kasahara et al., 2000). Mutation of the conserved leucine at position 40 between helix II and helix III to proline abolishes DNA binding, but maintains protein interactions of two homeobox proteins, Mix.1 and Xvent-2 (Mead et al., 1996; Onichtchouk et al., 1998). The same mutation was introduced into XNkx2-3 and XNkx2-5, and the effects monitored by overexpression in Xenopus embryos

588

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

(Grow and Krieg, 1998). Embryos injected with XNkx23L40P or XNkx2-5L40P display a range of defects from normal (45%) to complete lack of myocardial tissue (18%). Expression of several myofilament genes is disrupted, indicating a general impairment of myogenesis. Furthermore, the myocardium fails to delaminate from the surrounding tissues to enclose the endocardial tube, suggesting that these dominant inhibitory mutants block cell migration or epithelial structure (Grow and Krieg, 1998). Injection of mRNA encoding dominant-negative forms of XNkx2-3 and XNkx2-5 fused to the repressor domain of engrailed results in a range of mutant phenotypes including unfused hearts, asymmetrically located small hearts and a decrease in the cardiac mesodermal area as judged by expression of the markers MLC2 and cardiac troponin I (cTnI) (Fu et al., 1998). Both the number of embryos displaying cardiac defects and severity of the phenotype was increased if XNkx2-3 and XNkx2-5 engrailed fusions were coinjected. In the most severe cases a complete absence of cardiac mesoderm is observed. Taken together these data support an important role for XNkx2-3 and XNkx2-5 in Xenopus cardiogenesis. Recently, the Xenopus injection system has been used to analyze the effect of mutations in NKX2-5 which are associated with congenital heart disease. Two truncation mutants XNkx2-5His156ter and XNkx2-5Gln185ter, which correspond to the human mutations Gln170ter and Gln198ter (Schott et al., 1998; Benson et al., 1999), were injected into Xenopus embryos to assess the impact of early transient expression of these mutant proteins on heart development. Frogs expressing the mutant XNkx2-5 proteins display dilated atrial chambers, impaired septation, valve dysplasia and conduction defects (Bartlett et al., 2007), phenotypes observed in patients with NKX2-5 mutations. In zebrafish, the earliest cardiac field is marked by expression of the NK-2 gene nkx2.7, which precedes expression of nkx2.5 and nkx2.3 (Lee and Breitbart, 1996). Injection of nkx2.5 into zebrafish embryos results in cardiac hyperplasia (Chen and Fishman, 1996). The enlarged hearts may be due to increased cellular proliferation, or the recruitment of a larger number of cells from the cardiogenic field. Higher doses of nkx2.5 cause ectopic expression of cardiac markers, and in 50% of embryos extra beating tissue is found close to the heart. However, these embryos are grossly malformed so the appearance of ectopic contractile tissue may be due to impaired morphogenesis. These overexpression studies demonstrate that nkx2.5 is an important regulator of the cardiac cell fate decision in fish, and is critical for correct patterning within the emerging heart field. Morpholino knockdown studies directed at nkx2.5 and its close relative nkx2-7 indicate roles in heart tube morphogenesis and chamber cell proliferation (Targoff et al., 2008). Analysis of the futka (ftk) mutant in zebrafish revealed that ftk lies upstream of nkx2.5 in the genetic regulatory cascade controlling fish cardiogenesis. The ftk phenotype results from a loss-of-function mutation in a connexin

gene, cx36.7 or early cardiac connexin (ecx) (Sultana et al., 2008). Ftk mutants have a reduced heart contractile rate as a result of abnormal myofibrillogenesis and morphogenesis. In ftk embryos nkx2.5 is severely downregulated. Furthermore, restoration of nkx2.5 expression levels in ftk mutants rescues the mutant phenotype. Interestingly, cardiac conduction is normal in ftk mutants, suggesting that other signaling or cytoplasmic activities of cx36.7 regulate nkx2.5 expression. Based on sequence homology and chromosomal localization, the likely mammalian cx36.7 ortholog is Cx31.9. It remains to be determined if signaling from Cx31.9 is required for the onset of Nkx2-5 expression in mammals. Nevertheless, this study suggests a novel, possibly conserved, feedback mechanism linking nkx2.5 expression to contractile function in the developing embryo.

VII.  Tinman and the drosophila dorsal vessel The dorsal vessel of Drosophila is a midline-located linear tube that is often referred to as the Drosophila “heart” (Rizki, 1978; Rugendorff et al., 1994; Bodmer, 1995). Although simple in comparison to vertebrate cardiac structures, the Drosophila heart does demonstrate a considerable degree of specialization (Fig. 1). First, it is patterned in an anterior– ­posterior manner, with the anterior end of the tube narrower than the posterior. The anterior end of the heart is referred to as the aorta. Second, later in larval development a valve forms between the aorta and the posterior end of the heart tube. Coincident with valve formation is the development of small valvular openings in the posterior heart wall known as ostia (Rizki, 1978), which may be analogous to the inflow regions of the vertebrate heart. The ostia and aortic valve work in concert during a heartbeat, with the ostia closing as the valve opens, to allow a unidirectional flow of hemolymph from the posterior tube through to the aorta. Conversely, when the ostia open the aortic valve closes allowing the posterior chamber to fill with hemolymph (Bodmer and Frasch, 1999). The heart of Drosophila is progressively defined from the mesoderm until the highly-specialized cells reach the dorsal midline and form a functioning organ in a manner highly analogous to that described in vertebrates (Rizki, 1978; Rugendorff et al., 1994; Bodmer, 1995; Bodmer and Frasch, 1999; Tao and Schulz; 2007). Many of the molecular pathways underpinning dorsal vessel formation and patterning, including the requirement for NK-2 homeobox genes, are homologous to those occurring in vertebrate cardiogenesis. Initially the NK-2 gene tinman (tin) is activated throughout the mesoderm by the bHLH transcription factor twist (Yin et al., 1997). Subsequently, combinatorial Dpp (a Bmp homolog) (Xu et al., 1998; Yin and Frasch, 1998) and wingless (wg) (Wu et al., 1995; Park et al., 1996) signaling from the overlying ectoderm specifies cardiac progenitors, and

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

gradually restricts tin expression in a bilateral, segmentallyrepeated pattern within the dorsal mesoderm. These progenitors generate four myocardial cells (cardioblasts), which go on to form the muscular inner layer of cells surrounding the heart lumen and two pericardial cells, which become the outer layer of the heart (Zaffran and Frasch, 2002; Tao and Schulz, 2007). Therefore, the Drosophila heart, like that of its vertebrate counterpart, is specified from the precardiac mesoderm as paired progenitor pools by the action of BMP and Wnt signaling, which then fuse at the midline and go on to form the mature vessel. Tin has several roles during Drosophila cardiogenesis. First, tin mutants lack all dorsal mesodermal derivatives, including the dorsal vessel, visceral muscles and a subset of body wall muscles (Azpiazu and Frasch, 1993; Bodmer, 1993). Therefore, expression of tin ensures that cells of the dorsal mesoderm are competent to respond to cardiac induction signals. Expression in the dorsal mesoderm is induced by Dpp signaling via the Mad protein, and tin positively autoregulates itself through the combined action of tin and Mad (Yin et al., 1997; Xu et al., 1998). Second, tin is required to activate the GATA family member pannier (pnr), which is crucial in the specification of the myocardial and pericardial progenitors within the cardiogenic mesoderm (Gajewski et al., 1999; Alvarez et al., 2003; Klinedinst and Bodmer, 2003). A parallel genetic pathway involving the T-box proteins encoded by the Dorsocross locus (Doc1, 2 and 3) also activates pnr during specification of the cardiac progenitors (Reim and Frasch, 2005). Doc expression is induced at the intersection of Wg and Dpp signaling in the mesoderm, and is required for formation of myocardial and pericardial cell lineages (Reim and Frasch, 2005). Subsequently, pnr transcriptionally activates the Drosophila Tbx20 homolog midline (mid) in all cardioblasts (Miskolczi-McCallum et al., 2005; Qian et al., 2005; Reim et al., 2005), and mid acts in concert with its paralog H15 to promote late-stage tin expression in the cardioblasts (Reim et al., 2005). Conversely, in two myocardial cells tin acts as part of a feedback loop in which it directly activates expression of the seven-up (sup) gene that encodes a COUP-TF transcription factor, which in turn represses tin (Lo and Frasch, 2001; Ryan et al., 2007). Loss of tin in these cells, which are the precursors to the inflow structures called ostia, is permissive for continued Doc gene expression and ensuing cell specification (Reim and Frasch, 2005). Therefore, tin is part of numerous transcriptional feedback loops involving GATA factors and T-box proteins, and acts both to specify and pattern the cardiac lineages. Further, as with vertebrate Nkx2-5, tin acts as a transcriptional repressor or activator. Finally, tin may be required for the further differentiation of the dorsal vessel progenitors through the direct regulation of the transcription factors Mef2 and Hand and myofilament genes such as 3-tubulin (Gajewski et al., 1997; Kremser et al., 1999; Han and Olson, 2005).

589

The most striking phenotype of tin mutants is the complete lack of cardiac tissue, which is not seen in mutants of the verterbrate tin homolog Nkx2-5. To address the later roles of tin a “cardiac tissue-specific knockout” of tin was generated (Zaffran et al., 2006). Remarkably, in these flies myocardial cells are specified, the dorsal vessel forms and viable adults are observed. However, several aspects of heart development and activity are perturbed by the absence of myocardial tin. First, the morphology of the embryonic dorsal vessel is disorganized with the cells displaying abnormal shapes and the cardioblasts rows are incomplete. Second, cardioblast diversification is impaired, for example Doc is ectopically expressed in all cardioblasts whereas in wild-type hearts Doc is restricted to the two tin negative cells per hemisegment. Interestingly, the expression pattern of svp, which is required for Doc expression in the wild-type setting, is unaltered in tin-null myocardium, suggesting that Doc is repressed by tin. Finally, tin has an important role in the formation of the adult and larval heart. Dorsal vessels form from tin-deficient myocardium, but display morphological defects such as abnormal contractile apparatus, thin aortas and impaired remodeling at the adult stage. Furthermore, the physiological function of tin-deficient hearts is impaired. From an evolutionary perspective, it is interesting to note that Nkx2-5 functionally compensates for tin in the developing dorsal vessel, at least at the level of repressing both Doc and wg in embryonic cardioblasts.

VIII.  Regulatory components of the NKX2-5 locus Onset of Nkx2-5 expression is concomitant with the appearance of the earliest cardiac precursors within the paired cardiac progenitor pools. During development, Nkx2-5 expression is maintained throughout the formation of the linear heart tube, looping morphogenesis and during maturation of the four-chambered heart (Lints et al., 1993). That Nkx2-5 expression persists during these complex morphogenetic processes is noteworthy, and suggests that cis-­regulatory regions are precisely controlled by trans-acting factors during expression initiation and the subsequent regionalization and patterning of the heart tube and in the adult heart. Transgenic approaches using serial deletions of the genomic sequences flanking Nkx2-5 to control expression of either lacZ or GFP have illuminated the genetic pathways coordinating Nkx2-5 expression during mouse embryogenesis (Searcy et al., 1998; Lien et al., 1999; Reecy et al., 1999; Tanaka et al., 1999b; Lien et al., 2002; Brown et al., 2004; Chi et al., 2005). These studies reveal that the genomic architecture of the Nkx2-5 locus includes a number of positive and negative regulatory modules that have discrete roles during cardiogenesis (Schwartz and Olson, 1999). Several Nkx2-5 enhancers are found within the 14kb 5 of the start

590

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

codon (reviewed in Schwartz and Olson, 1999). However, a 3 flanking element known as activating region 5 is proposed to drive expression in the right ventricle (Tanaka et al., 1999a), and expression in the tongue is regulated by a conserved enhancer termed UH4 found 20 kb upstream of exon 1, which is shared with the neighboring, ­myogenicallyexpressed, gene called texas (Chi et al., 2005). Other noncardiac Nkx2-5 expression sites (i.e., thyroid primordium, spleen, pharynx and stomach) are controlled by a proximal enhancer region termed activating region 2 (AR2) (reviewed in Schwartz and Olson, 1999). In the heart, transgenic analysis using fragments of the 14 kb of 5 flanking sequence revealed that no single element was capable of faithfully recapitulating the endogenous pattern, demonstrating the collaborative action of these elements (Searcy et al., 1998; Reecy et al., 1999; Tanaka et al., 1999b). Five activating modules and three inhibitory regions have been mapped within the Nkx2-5 5 regulatory sequences (Schwartz and Olson, 1999; Lien et al., 2002; Brown et al., 2004; Chi et al., 2005). At 3 kb from the transcriptional start site lies activating region 2 (AR2) (Searcy et al., 1998) and a further 2 kb upstream is activating region 3 (also called the GATA-SMAD (G-S) element), both of which are required for early expression, i.e., in the cardiac crescent (Reecy et al., 1999; Brown et al., 2004). Further upstream (9 kb) lies activating region 1 (AR1) which promotes expression in the cardiac crescent, the linear and looping heart tube before becoming restricted to the outflow tract and right ventricle (Lien et al., 1999); upstream homology region 5 which drives transgene expression in both atria, the atrioventricular canal and the interventricular septum (IVS) (Chi et al., 2005); and upstream homology region 6 which activates expression in the ventricular chambers, the atrioventricular canal and interventricular septum (Chi et al., 2005). Inhibitory region 1 (IR1) is located within AR1, while IR 2 and 3 are found adjacent to AR2 and 3, respectively. Factors that bind at the three IRs are unknown so the mechanisms by which they act have yet to be determined. For the AR and UH regions histone acetylation appears to be important in maintaining the active chromatin state, as the temporal and spatial histone acetylation pattern matches that of the mapped Nkx2-5 enhancers (Chi et al., 2005). For example, the tongue enhancer UH4 is only associated with acetylated histones in chromatin isolated from tongue tissue. Refined mapping of the cardiac Nkx2-5 enhancers will facilitate the search for the transacting factors and signaling pathways that regulate these elements. A key step in heart development is the response of cells within the lateral plate mesoderm cells to signaling from the overlying endoderm leading to the adoption of a cardiac cell fate. Studies in the frog and chick indicate that secreted BMPs mediate this inductive signal (Schultheiss et al., 1995, 1997). Furthermore, Nkx2-5 expression is induced in response to BMP signaling, raising the possibility that Nkx2-5 is a direct transcriptional target of active SMADs (Schultheiss et al., 1997; Schlange et al., 2000).

Indeed, the AR2 enhancer element of Nkx2-5 is a direct transcriptional target of activated SMADs (Liberatore et al., 2002; Lien et al., 2002). Mutation of an ­ evolutionarily­conserved SMAD site within AR2 abolishes transgene expression in the cardiac crescent (Liberatore et al., 2002; Lien et al., 2002). SMAD activity at the AR2 enhancer is further refined by interactions with the high mobility group (HMG) protein HMGA2, a nonhistone chromatin associated factor (Monzen et al., 2008). HMGA2 directly binds to both the AR2 and the Smad1/5/8 complex, and the HMG2A binding site within AR2 is required for ­expression of AR2 reporter genes in Xenopus embryos. In addition, HMGA2 is essential for cardiogenesis in Xenopus embryos (Monzen et al., 2008). Thus, chromatin remodeling and the subsequent alteration of gene expression at the Nkx2-5 locus may be mediated by the recruitment of HMGA2 by Smad1 at AR2. The importance of SMAD activation for Nkx2-5 induction is further highlighted by the requirement of SMAD binding for AR3/G-S enhancer-dependent expression (Brown et al., 2004). Therefore, early induction of Nkx2-5 by the BMP-signaling cascade is supported by the presence of possibly redundant SMAD responsive elements within the enhancer sequences. Nkx2-5 enhancer fragments also contain conserved GATA binding sites, implying a critical role for GATA factors in the regulation of Nkx2-5 (Searcy et al., 1998; Brown et al., 2004). Interestingly, many of these are found in the AR3 or G-S enhancer region, suggesting that GATA and Smad factors cooperatively regulate Nkx2-5. Deletion of AR3 in the context of a transgenic construct containing the 5 flanking 10 kb enhancer, which includes both AR1 and AR2, severely diminishes transgene expression in the early embryo (Brown et al., 2004). Furthermore, GATA4 and the Smad1/4 heterodimer can synergistically activate expression from the G-S enhancer and GATA4 bound to Smad2, 3 and 4, in vitro (Brown et al., 2004). Taken together these data suggest that the onset of Nkx2-5 expression relies on initial activation of the G-S enhancer by a transcriptional complex comprising GATA4 and transcriptionally-active SMADs. The identified Nkx2-5 enhancers are being used as reagents to examine lineage specification during cardiogenesis. For example, the directed expression of the large SV-40 T-antigen in cardiac precursors to generate novel cardiac cell lines using the Nkx2-5 proximal 3 kb enhancer (Brunskill et al., 2001). A second enhancer element driving enhanced yellow fluorescent protein was used to establish transcription profiles of cardiomyocytes from the cardiac crescent, linear tube and looped heart tube stages (Masino et al., 2004). Nkx2-5 enhancer-directed expression of eGFP has also been used to examine the developmental potential of Nkx2-5 positive cells (Wu et al., 2006; Christoforou et al., 2008). Cardiac progenitors isolated from 9.5 dpc embryos by FACS on the basis of eGFP expression gave rise not only to cardiomyocytes and conduction cells, but also to smooth muscle cells (Wu et al., 2006), indicating the presence of bipotential progenitor cells. Analysis of cell surface stem cell markers

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

revealed a subset of Nkx2-5ve/c-Kitve cells capable of generating both cardiac and smooth muscle cell types (Wu et al., 2006). In an emerging model of progressive cardiac lineage restriction, these bipotential cells arise from multipotent progenitors positive for the cell surface marker Flk1 and the transcription factor Isl-1 (Kattman et al., 2006; Moretti et al., 2006; Wu et al., 2006). In the context of cardiac stem cell therapy, it is promising to note that these multipotent progenitors can be cultured and, furthermore, can be isolated in vitro from differentiating embryonic stem cell-derived cultures (Kattman et al., 2006; Moretti et al., 2006).

IX.  Conclusions NK-2 homeobox genes are essential for cardiogenesis in a wide range of species, reflecting the early deployment of these genes during the evolution of cardiac structures. As the complexity of the heart has increased during evolution NK-2 genes have been utilized to control different processes. In vertebrates, Nkx2-5 represents a critical node in the complex transcriptional network governing the early specification and proliferation of the cardiac lineage in both the first and second heart fields of vertebrates, and during the subsequent molding of these progenitor cells into the mature organ. As outlined in this chapter, Nkx2-5 plays a role in almost all facets of heart development, including the regulation of cardiac progenitor cell number, the formation and patterning of the cardiac conduction system, septation, valve development, patterning the venous returns and the myogenic program. Importantly, Nkx2-5 does not act alone, but in concert with other highly­conserved transcription factors and signaling cascades to regulate cardiogenesis. Genome-wide expression profiling has begun to reveal Nkx2-5-dependent genes, and establishing which of these genes are direct targets remains the next challenge. Similar data for members of the conserved cardiac GRN would allow the construction of a transcriptional network that regulates early cardiogenesis, in a manner analogous to that for early patterning of the Drosophila mesoderm (Sandmann et al., 2007; Zeitlinger et al., 2007). Investigation of network topology may reveal the molecular circuitry underlying the cardiac developmental process in greater depth, and perhaps give clues to some of the regulatory transitions that led to the evolution of the four-chambered mammalian heart. Despite our increased understanding of NK-2 gene function, at both the molecular and developmental levels, much remains to be determined. Since mutations in NKX2-5 underlie congenital heart disease, NKX2-5 transcriptional targets and the molecular pathways acting on NKX2-5 expression and function may represent fields of increasing importance in the clinical setting. This research is given an added incentive by the possible clinical ramifications of further discoveries, not only in the identification and understanding of congenital heart disease, but also in the drive to develop strategies in cardiac regeneration and tissue engineering by harnessing endogenous cardiac stem cells.

591

References Alvarez, A.D., Shi, W., Wilson, B.A., Skeath, J.B., 2003. Pannier and pointedP2 act sequentially to regulate Drosophila heart development. Development 130, 3015–3026. Azpiazu, N., Frasch, M., 1993. Tinman and bagpipe: two homeo box genes that determine cell fates in the dorsal mesoderm of Drosophila. Genes Dev. 7, 1325–1340. Ballard, V.L., Edelberg, J.M., 2007. Stem cells and the regeneration of the aging cardiovascular system. Circ. Res. 100, 1116–1127. Bartlett, H.L., Sutherland, L., Kolker, S.J., Welp, C., Tajchman, U., Desmarais, V., Weeks, D.L., 2007. Transient early embryonic ­expression of Nkx2-5 mutations linked to congenital heart defects in human cause heart defects in Xenopus laevis. Dev. Dyn. 236, 2475–2484. Basson, C.T., Bachinsky, D.R., Lin, R.C., Levi, T., Elkins, J.A., Soults, J., Grayzel, D., Kroumpouzou, E., Traill, T.A., Leblanc-Straceski, J., Renault, B., Kucherlapati, R., Seidman, J.G., Seidman, C.E., 1997. Mutations in human TBX5 cause limb and cardiac malformation in Holt-Oram syndrome. Nat. Genet. 15, 30–35. Basson, C.T., Huang, T., Lin, R.C., Bachinsky, D.R., Weremowicz, S., Vaglio, A., Bruzzone, R., Quadrelli, R., Lerone, M., Romeo, G., Silengo, M., Pereira, A., Krieger, J., Mesquita, S.F., Kamisago, M., Morton, C.C., Pierpont, M.E., Müller, C.W., Seidman, J.G., Seidman, C.E., 1999. Different TBX5 interactions in heart and limb defined by Holt-Oram syndrome mutations. Proc. Natl. Acad. Sci. USA 96, 2919–2924. Bearzi, C., Rota, M., Hosoda, T., Tillmanns, J., Nascimbene, A., De Angelis, A., Yasuzawa-Amano, S., Trofimova, I., Siggins, R.W., Lecapitaine, N., Cascapera, S., Beltrami, A.P., D’Alessandro, D.A., Zias, E., Quaini, F., Urbanek, K., Michler, R.E., Bolli, R., Kajstura, J., Leri, A., Anversa, P., 2007. Human cardiac stem cells. Proc. Natl. Acad. Sci. USA 104, 14068–14073. Beltrami, A.P., Barlucchi, L., Torella, D., Baker, M., Limana, F., Chimenti, S., Kasahara, H., Rota, M., Musso, E., Urbanek, K., Leri, A., Kajstura, J., Nadal-Ginard, B., Anversa, P., 2003. Adult cardiac stem cells are multipotent and support myocardial regeneration. Cell 114, 763–776. Benson, D.W., Silberbach, G.M., Kavanaugh-McHugh, A., Cottrill, C., Zhang, Y., Riggs, S., Smalls, O., Johnson, M.C., Watson, M.S., Seidman, J.G., Seidman, C.E., Plowden, J., Kugler, J.D., 1999. Mutations in Nkx2-5, a cardiac transcription factor, affect diverse cardiac developmental pathways. J. Clin. Invest. 104, 1567–1573. Biben, C., Harvey, R.P., 1997. Homeodomain factor Nkx2-5 controls left–right asymmetric expression of bHLH eHand during murine heart development. Genes Dev. 11, 1357–1369. Biben, C., Hatzistavrou, T., Harvey, R.P., 1998. Expression of NK-2 class homeobox gene Nkx2-6 in foregut endoderm and heart. Mech. Dev. 73, 125–127. Biben, C., Weber, R., Kesteven, S., Stanley, E., McDonald, L., Elliott, D.A., Barnett, L., Köentgen, F., Robb, L., Feneley, M., Harvey, R.P., 2000. Cardiac septal and valvular dysmorphogenesis in mice heterozygous for mutations in the homeobox gene Nkx2-5. Circ. Res. 87, 888–895. Bodmer, R., 1993. The gene tinman is required for specification of the heart and visceral muscles in Drosophila. Development 118, 719–729. Bodmer, R., 1995. Heart development in Drosophila and its relationship to vertebrates. Trends Cardiovasc. Med. 5, 21–28. Bodmer, R., Frasch, M., 1999. Genetic determination of Drosophila heart development. In: Harvey, R., Rosenthal, N. (Eds.), Heart Development. Academic Press, San Diego, CA, pp. 65–90. Bouche, N., Scharlat, A., Snedden, W., Bouchez, D., Fromm, H., 2002. A novel family of calmodulin-binding transcription activators in multicellular organisms. J. Biol. Chem. 277, 21851–21861.

592

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Brown, C.O., III, Chi, X., Garcia-Gras, E., Shirai, M., Feng, X.H., Schwartz, R.J., 2004. The cardiac determination factor, Nkx2-5, is activated by mutual cofactors GATA-4 and Smad1/4 via a novel upstream enhancer. J. Biol. Chem. 279, 10659–10669. Bruneau, B.G., Nemer, G., Schmitt, J.P., Charron, F., Robitaille, L., Caron, S., Conner, D.A., Gessler, M., Nemer, M., Seidman, C.E., Seidman, J.G., 2001. A murine model of Holt-Oram syndrome defines roles of the T-box transcription factor Tbx5 in cardiogenesis and disease. Cell 106, 709–721. Brunskill, E.W., Witte, D.P., Yutzey, K.E., Potter, S.S., 2001. Novel cell lines promote the discovery of genes involved in early heart development. Dev. Biol. 235, 507–520. Buchberger, A., Pabst, O., Brand, T., Seidl, K., Arnold, H.-H., 1996. Chick NKx-2.3 represents a novel family member of vertebrate homologues to Drosophila homeobox gene tinman: differential expression of cNKx-2.3 and cNKx-2.5 during heart and gut development. Mech. Dev. 56, 151–163. Buckingham, M., Meilhac, S., Zaffran, S., 2005. Building the mammalian heart from two sources of myocardial cells. Nat. Rev. Genet. 6, 826–835. Cai, C.L., Liang, X., Shi, Y., Chu, P.H., Pfaff, S.L., Chen, J., Evans, S., 2003. Isl1 identifies a cardiac progenitor population that proliferates prior to differentiation and contributes a majority of cells to the heart. Dev. Cell. 5, 877–889. Chen, C.Y., Schwartz, R.J., 1995. Identification of novel DNA binding targets and regulatory domains of a murine tinman homeodomain factor, Nkx-2.5. J. Biol. Chem. 270, 15628–15633. Chen, C.Y., Schwartz, R.J., 1996. Recruitment of the tinman homologue Nkx-2.5 by serum response factor activates cardiac a-actin gene transcription. Mol. Cell. Biol. 16, 6372–6384. Chen, C.Y., Croissant, J., Majesky, M., Topouzis, S., McQuinn, T., Frankovsky, M.J., Schwartz, R.J., 1996. Activation of the cardiac alpha-actin promoter depends upon serum response factor, Tinman homologue, Nkx-2.5, and intact serum response elements. Dev. Genet. 19, 119–130. Chen, F., Kook, H., Milewski, R., Gitler, A.D., Lu, M.M., Li, J., Nazarian, R., Schnepp, R., Jen, K., Biben, C., Runke, G., Mackay, J.P., Novotny, J., Schwartz, R.J., Harvey, R.P., Mullins, M.C., Epstein, J.A., 2002. Hop is an unusual homeobox gene that modulates cardiac development. Cell 110, 713–723. Chen, J.N., Fishman, M.C., 1996. Zebrafish tinman homolog demarcates the heart field and initiates myocardial differentiation. Development 122, 3809–3816. Chi, X., Chatterjee, P.K., Wilson, W., Zhang, S.X., Demayo, F.J., Schwartz, R.J., 2005. Complex cardiac Nkx2-5 gene expression activated by noggin-sensitive enhancers followed by chamber-specific modules. Proc. Natl. Acad. Sci. USA 102, 13490–13495. Choi, C.Y., Lee, Y.M., Kim, Y.H., Park, T., Jeon, B.H., Schulz, R.A., Kim, Y., 1999. The homeodomain transcription factor NK-4 acts as either a transcriptional activator or repressor and interacts with the p300 coactivator and the groucho corepressor. J. Biol. Chem. 274, 31543–31552. Christoffels, V.M., Habets, P.E., Franco, D., Campione, M., de Jong, F., Lamers, W.H., Bao, Z.Z., Palmer, S., Biben, C., Harvey, R.P., Moorman, A.F., 2000. Chamber formation and morphogenesis in the developing mammalian heart. Dev. Biol. 223, 266–278. Christoffels, V.M., Mommersteeg, M.T., Trowe, M.O., Prall, O.W., de Gier-de Vries, C., Soufan, A.T., Bussen, M., Schuster-Gossler, K., Harvey, R.P., Moorman, A.F., Kispert, A., 2006. Formation of the venous pole of the heart from an Nkx2-5-negative precursor population requires Tbx18. Circ. Res. 98, 1555–1563.

Christoforou, N., Miller, R.A., Hill, C.M., Jie, C.C., Mccallion, A.S., Gearhart, J.D., 2008. Mouse ES cell-derived cardiac precursor cells are multipotent and facilitate identification of novel cardiac genes. J. Clin. Invest. 118, 894–903. Cleaver, O.B., Patterson, K.D., Krieg, P.A., 1996. Overexpression of the tinman-related genes XNkx-2.5 and XNkx-2.3 in Xenopus embryos results in myocardial hyperplasia. Development 122, 3549–3556. Copley, R.R., 2005. The EH1 motif in metazoan transcription factors. BMC Genomics 6, 169. Cripps, R.M., Olson, E.N., 2002. Control of cardiac development by an evolutionarily conserved transcriptional network. Dev. Biol. 246, 14–28. Damante, G., Fabbro, D., Pellizzari, L., Civitareale, D., Guazzi, S., Polycarpou-Schwartz, M., Cauci, S., Quadrifoglio, F., Formisano, S., Di Lauro, R., 1994. Sequence-specific DNA recognition by the thyroid transcription factor-1 homeodomain. NAR 22, 3075–3083. Davidson, E.H., Erwin, D.H., 2006. Gene regulatory networks and the evolution of animal body plans. Science 311, 796–800. Dawson, S.R., Turner, D.L., Weintraub, H., Parkhurst, S.M., 1995. Specificity for the hairy/enhancer of split basic helix-loop-helix (bHLH) proteins maps outside the bHLH domain and suggests two separable modes of transcriptional repression. Mol. Cell Biol. 15, 6923–6931. De Felice, M., Di Lauro, R., 2004. Thyroid development and its disorders: genetics and molecular mechanisms. Endocr. Rev. 25, 722–746. Dentice, M., Cordeddu, V., Rosica, A., Ferrara, A.M., Santarpia, L., Salvatore, D., Chiovato, L., Perri, A., Moschini, L., Fazzini, C., Olivieri, A., Costa, P., Stoppioni, V., Baserga, M., De Felice, M., Sorcini, M., Fenzi, G., Di Lauro, R., Tartaglia, M., Macchia, P.E., 2006. Missense mutation in the transcription factor NKX2-5: a novel molecular event in the pathogenesis of thyroid dysgenesis. J. Clin. Endocrinol. Metab. 91, 1428–1433. Dodou, E., Verzi, M.P., Anderson, J.P., Xu, S.M., Black, B.L., 2004. Mef2c is a direct transcriptional target of ISL1 and GATA factors in the anterior heart field during mouse embryonic development. Development 131, 3931–3942. Duboule, D., 1994. Guidebook of Homeobox Genes. IRL Press, Oxford, UK. Durocher, D., Charron, F., Warren, R., Schwartz, R.J., Nemer, M., 1997. The cardiac transcription factors Nkx2-5 and GATA-4 are mutual cofactors. EMBO. J. 16, 5687–5696. Durocher, D., Chen, C.-Y., Ardati, A., Schwartz, R.J., Nemer, M., 1996. The atrial natriuretic factor promoter is a downstream target for Nkx2.5 in the myocardium. Mol. Cell Biol. 16, 4648–4655. Elliott, D.A., Kirk, E.P., Yeoh, T., Chandar, S., McKenzie, F., Taylor, P., Grossfeld, P., Fatkin, D., Jones, O., Hayes, P., Feneley, M., Harvey, R.P., 2003. Cardiac homeobox gene NKX2-5 mutations and congenital heart disease: associations with atrial septal defect and hypoplastic left heart syndrome. J. Am. Coll. Cardiol. 41, 2072–2076. Elliott, D.A., Solloway, M.J., Wise, N., Biben, C., Costa, M.W., Furtado, M.B., Lange, M., Dunwoodie, S., Harvey, R.P., 2006. A tyrosine-rich domain within homeodomain transcription factor Nkx2-5 is an essential element in the early cardiac transcriptional regulatory machinery. Development 133, 1311–1322. Evans, S., 1999. Vertebrate tinman homologues and cardiac differentiation. Sem. Cell Dev. Biol. 10, 73–83. Evans, S., Yan, W., Murillo, M.P., Ponce, J., Papalopulu, N., 1995. Tinman, a Drosophila homeobox gene required for heart and visceral mesoderm specification, may be represented by a family of genes in vertebrates: XNkx-2.3, a second homologue of tinman. Development 121, 3889–3899. Ferdous, A., Caprioli, A., Iacovino, M., Martin, C.M., Morris, J., Richardson, J.A., Latif, S., Hammer, R.E., Harvey, R.P., Olson, E.N.,

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

Kyba, M., Garry, D.J., 2009. Nkx2-5 transactivates the Ets-related protein 71 gene and specifies an endothelial/endocardial fate in the developing embryo. Proc. Natl. Acad. Sci. (USA) 106 (3), 814–819. Firulli, A.B., Mcfadden, D.G., Lin, Q., Srivastava, D., Olson, E.N., 1998. Heart and extra-embryonic mesodermal defects in mouse embryos lacking the bHLH transcription factor Hand1. Nat. Genet. 18, 266–270. Fisher, A.L., Ohsako, S., Caudy, M., 1996. The WRPW motif of the hairy-related basic helix-loop-helix repressor proteins acts as a 4amino-acid transcription repression and protein-protein interaction domain. Mol. Cell Biol. 16, 2670–2677. Fraenkel, E., Pabo, C.O., 1998. Comparison of X-ray and NMR structures for the Antennapedia homeodomain-DNA complex. Nat. Struct. Biol. 5, 692–697. Franco, D., Campione, M., 2003. The role of Pitx2 during cardiac development. Linking left–right signaling and congenital heart diseases. Trends Cardiovasc. Med. 13, 157–163. Frasch, M., 1999. Intersecting signalling and transcriptional pathways in Drosophila heart specification. Semin. Cell Dev. Biol. 10, 61–71. Fu, Y., Izumo, S., 1995. Cardiac myogenesis: overexpression of XCsx2 or XMef2a in whole Xenopus embryos induces the precocious expression of XMHCa gene. Roux’s Arch. Dev. Biol. 205, 198–202. Fu, Y., Yan, W., Mohun, T.J., Evans, S.M., 1998. Vertebrate tinman homologues XNkx2-3 and XNkx2-5 are required for heart formation in a functionally redundant manner. Development 125, 4439–4449. Gajewski, K., Kim, Y., Lee, Y.M., Olson, E.N., Schulz, R.A., 1997. Dmef2 is a target for Tinman activation during Drosophila heart development. EMBO. J. 16, 515–522. Gajewski, K., Fossett, N., Molkentin, J.D., Schulz, R.A., 1999. The zinc finger proteins Pannier and GATA4 function as cardiogenic factors in Drosophila. Development 126, 5679–5688. Gajewski, K., Zhang, Q., Choi, C.Y., Fossett, N., Dang, A., Kim, Y.H., Kim, Y., Schulz, R.A., 2001. Pannier is a transcriptional target and partner of Tinman during Drosophila cardiogenesis. Dev. Biol. 233, 425–436. Garg, V., Kathiriya, I.S., Barnes, R., Schluterman, M.K., King, I.N., Butler, C.A., Rothrock, C.R., Eapen, R.S., Hirayama-Yamada, K., Joo, K., Matsuoka, R., Cohen, J.C., Srivastava, D., 2003. GATA4 mutations cause human congenital heart defects and reveal an interaction with TBX5. Nature 424, 443–447. Gasperowicz, M., Otto, F., 2005. Mammalian Groucho homologs: redundancy or specificity? J. Cell Biochem. 95, 670–687. Gehring, W.J., Qian, Y.Q., Billeter, M., Furukubo-Tokunaga, K., Schier, A.F., Resendez-Perez, D., Affolter, M., Otting, G., Wüthrich, K., 1994. Homeodomain-DNA recognition. Cell 78, 211–223. Goldmuntz, E., Geiger, E., Benson, D.W., 2001. NKX2.5 mutations in patients with tetralogy of Fallot. Circulation 104, 2565–2568. Grepin, C., Dagnino, L., Robitaille, L., Haberstroh, L., Antakly, T., Nemer, M., 1994. A hormone-encoding gene identifies a pathway for cardiac but not skeletal muscle gene transcription. Mol. Cell Biol. 14, 3115–3129. Grow, M.W., Krieg, P.A., 1998. Tinman function is essential for vertebrate heart development: elimination of cardiac differentiation by dominant inhibitory mutants of the tinman-related genes, XNkx2-3 and XNkx2-5. Dev. Biol. 204, 187–196. Gruschus, J., Tsao, D., Wang, L.-H., Nirenberg, M., Ferretti, J., 1997. Interactions of the vnd/NK-2 homeodomain with DNA by nuclear magnetic resonance spectroscopy: basis of binding specificity. Biochemistry 36, 5372–5380. Gruschus, J., Tsao, D., Wang, L.-H., Nirenberg, M., Ferretti, J., 1999. The three dimensional structure of the vnd/NK-2 homeodomain-DNA complex by NMR spectroscopy. J. Mol. Biol. 289, 529–545.

593

Guazzi, S., Price, M., De Felice, M., Damante, G., Mattei, M.-G., Di Lauro, R., 1990. Thyroid nuclear factor 1 (TTF-1) contains a homeodomain and displays a novel DNA binding specificity. EMBO. J. 9, 3631–3639. Gutierrez-Roelens, I., Sluysmans, T., Gewillig, M., Devriendt, K., Vikkula, M., 2002. Progressive AV-block and anomalous venous return among cardiac anomalies associated with two novel missense mutations in the CSX/NKX2-5 gene. Hum. Mutat. 20, 75–76. Habets, P.E., Moorman, A.F., Clout, D.E., van Roon, M.A., Lingbeek, M., van Lohuizen, M., Campione, M., Christoffels, V.M., 2002. Cooperative action of Tbx2 and Nkx2.5 inhibits ANF expression in the atrioventricular canal: implications for cardiac chamber formation. Genes Dev. 16, 1234–1246. Haissaguerre, M., Jais, P., Shah, D.C., Takahashi, A., Hocini, M., Quiniou, G., Garrigue, S., Le Mouroux, A., Le Métayer, P., Clémenty, J., 1998. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N. Engl. J. Med. 339, 659–666. Han, Z., Olson, E.N., 2005. Hand is a direct target of Tinman and GATA factors during Drosophila cardiogenesis and hematopoiesis. Development 132, 3525–3536. Harvey, R.P., 1996. NK-2 homeobox genes and heart development. Dev. Biol. 178, 203–216. Harvey, R.P., Biben, C., Elliott, D.A., 1998. Transcriptional control and pattern formation in the vertebrate heart: studies on NK-2 class homeodomain factors. In: Harvey, R.P., Rosenthal, N. (Eds.), Heart Development. Academic Press, San Diego, CA, pp. 111–129. Heathcote, K., Braybrook, C., Abushaban, L., Guy, M., Khetyar, M.E., Patton, M.A., Carter, N.D., Scambler, P.J., Syrris, P., 2005. Common arterial trunk associated with a homeodomain mutation of NKX2.6. Hum. Mol. Genet. 14, 585–593. Hirayama-Yamada, K., Kamisago, M., Akimoto, K., Aotsuka, H., Nakamura, Y., Tomita, H., Furutani, M., Imamura, S., Takao, A., Nakazawa, M., Matsuoka, R., 2005. Phenotypes with GATA4 or NKX2.5 mutations in familial atrial septal defect. Am. J. Med. Genet. A. 135, 47–52. Hiroi, Y., Kudoh, S., Monzen, K., Ikeda, Y., Yazaki, Y., Nagai, R., Komuro, I., 2001. Tbx5 associates with Nkx2-5 and synergistically promotes cardiomyocyte differentiation. Nat. Genet. 28, 276–280. Hoffman, J.I., Kaplan, S., 2002. The incidence of congenital heart disease. J. Am. Coll. Cardiol. 39, 1890–1900. Horsthuis, T., Houweling, A.C., Habets, P.E., de Lange, F.J., el Azzouzi, H., Clout, D.E., Moorman, A.F., Christoffels, V.M., 2008. Distinct regulation of developmental and heart disease-induced atrial natriuretic factor expression by two separate distal sequences. Circ. Res. 102, 849–859. Iavarone, A., Lasorella, A., 2006. ID proteins as targets in cancer and tools in neurobiology. Trends Mol. Med. 12, 588–594. Ikeda, Y., Hiroi, Y., Hosoda, T., Utsunomiya, T., Matsuo, S., Ito, T., Inoue, J., Sumiyoshi, T., Takano, H., Nagai, R., Komuro, I., 2002. Novel point mutation in the cardiac transcription factor CSX/NKX2.5 associated with congenital heart disease. Circ. J. 66, 561–563. Ismat, F.A., Zhang, M., Kook, H., Huang, B., Zhou, R., Ferrari, V.A., Epstein, J.A., Patel, V.V., 2005. Homeobox protein Hop functions in the adult cardiac conduction system. Circ. Res. 96, 898–903. Isono, K., Nemoto, K., Li, Y., Takada, Y., Suzuki, R., Katsuki, M., Nakagawara, A., Koseki, H., 2006. Overlapping roles for homeodomain-interacting protein kinases hipk1 and hipk2 in the mediation of cell growth in response to morphogenetic and genotoxic signals. Mol. Cell Biol. 26, 2758–2771. Jay, P.Y., Harris, B.S., Maguire, C.T., Buerger, A., Wakimoto, H., Tanaka, M., Kupershmidt, S., Roden, D.M., Schultheiss, T.M.,

594

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

O’Brien, T.X., Gourdie, R.G., Berul, C.I., Izumo, S., 2004. Nkx2-5 mutation causes anatomic hypoplasia of the cardiac conduction system. J. Clin. Invest. 113, 1130–1137. Kasahara, H., Izumo, S., 1999. Identification of the in vivo casein kinase II phosphorylation site within the homeodomain of the cardiac tissue specifying homeobox gene product Csx/Nkx2-5. Mol. Cell Biol. 19, 526–536. Kasahara, H., Lee, B., Schott, J.J., Benson, D.W., Seidman, J.G., Seidman, C.E., Izumo, S., 2000. Loss of function and inhibitory effects of human CSX/NKX2.5 homeoprotein mutations associated with congenital heart disease. J. Clin. Invest. 106, 299–308. Kasahara, H., Wakimoto, H., Liu, M., Maguire, C.T., Converso, K.L., Shioi, T., Huang, W.Y., Manning, W.J., Paul, D., Lawitts, J., Berul, C.I., Izumo, S., 2001. Progressive atrioventricular conduction defects and heart failure in mice expressing a mutant Csx/Nkx2.5 homeoprotein. J. Clin. Invest. 108, 189–201. Kasahara, H., Benson, D.W., 2004. Biochemical analyses of eight NKX2.5 homeodomain missense mutations causing atrioventricular block and cardiac anomalies. Cardiovasc. Res. 64, 40–51. Kattman, S.J., Huber, T.L., Keller, G.M., 2006. Multipotent flk-1 cardiovascular progenitor cells give rise to the cardiomyocyte, endothelial, and vascular smooth muscle lineages. Dev. Cell 11, 723–732. Kaufman, T.C., Lewis, R., Wakimoto, B., 1980. Cytogenetic analysis of chromosome 3 in Drosophila melanogaster: the homeotic gene complex in polytene chromosome interval 84A-B. Genetics 94, 115–133. Kim, D.W., Lassar, A.B., 2003. Smad-dependent recruitment of a ­histone deacetylase/Sin3A complex modulates the bone morphogenetic protein-dependent transcriptional repressor activity of Nkx3.2. Mol. Cell Biol. 23, 8704–8717. Kim, Y., Nirenberg, M., 1989. Drosophila NK-homeobox genes. Proc. Natl. Acad. Sci. USA 86, 7716–7720. Kim, Y.H., Choi, C.Y., Lee, S.-J., Conti, M.A., Kim, Y., 1998. Homeodomain-interacting protein kinases, a novel family of corepressors for homeodomain transcription factors. J. Biol. Chem. 273, 25875–25879. Kirk, E.P., Hyun, C., Thomson, P.C., Lai, D., Castro, M.L., Biben, C., Buckley, M.F., Martin, I.C., Moran, C., Harvey, R.P., 2006. Quantitative trait loci modifying cardiac atrial septal morphology and risk of patent foramen ovale in the mouse. Circ. Res. 98, 651–658. Klinedinst, S.L., Bodmer, R., 2003. Gata factor Pannier is required to establish competence for heart progenitor formation. Development 130, 3027–3038. Ko, L.J., Engel, J.D., 1993. DNA-binding specificities of the GATA transcription factor family. Mol. Cell Biol. 13, 4011–4022. Koizumi, K., Lintas, C., Nirenberg, M., Maeng, J.S., Ju, J.H., Mack, J.W., Gruschus, J.M., Odenwald, W.F., Ferretti, J.A., 2003. Mutations that affect the ability of the vnd/NK-2 homeoprotein to regulate gene expression: transgenic alterations and tertiary structure. Proc. Natl. Acad. Sci. USA 100, 3119–3124. Komuro, I., Izumo, S., 1993. Csx: A murine homeobox-containing gene specifically expressed in the developing heart. Proc. Natl. Acad. Sci. USA 90, 8145–8149. Konig, K., Will, J.C., Berger, F., Muller, D., Benson, D.W., 2006. Familial congenital heart disease, progressive atrioventricular block and the cardiac homeobox transcription factor gene NKX2.5: identification of a novel mutation. Clin. Res. Cardiol. 95, 499–503. Kook, H., Lepore, J.J., Gitler, A.D., Lu, M.M., Wing-Man Yung, W., Mackay, J., Zhou, R., Ferrari, V., Gruber, P., Epstein, J.A., 2003. Cardiac hypertrophy and histone deacetylase-dependent transcriptional repression mediated by the atypical homeodomain protein Hop. J. Clin. Invest. 112, 863–871.

Kremser, T., Gajewski, K., Schulz, R.A., Renkawitz-Pohl, R., 1999. Tinman regulates the transcription of the beta3 tubulin gene (betaTub60D) in the dorsal vessel of Drosophila. Dev. Biol. 216, 327–339. Kuo, H., Chen, J., Ruiz-Lozano, P., Zou, Y., Nemer, M., Chien, K.R., 1999. Control of segmental expression of the cardiac-restricted ankyrin repeat protein gene by distinct regulatory pathways in murine cardiogenesis. Development 126, 4223–4234. Kupershmidt, S., Yang, T., Anderson, M.E., Wessels, A., Niswender, K.D., Magnuson, M.A., Roden, D.M., 1999. Replacement by homologous recombination of the minK gene with lacZ reveals restriction of minK expression to the mouse cardiac conduction system. Circ. Res. 84, 146–152. Labbe, E., Silvestri, C., Hoodless, P.A., Wrana, J.L., Attisano, L., 1998. Smad2 and Smad3 positively and negatively regulate TGF betadependent transcription through the forkhead DNA-binding protein FAST2. Mol. Cell. 2, 109–120. Lee, K.H., Breitbart, R.E., 1996. A new tinman-related gene, nxk2.7, anticipates the expression of nkx2.5 and nkx2.3 in zebrafish heart and pharyngeal endoderm. Dev. Biol. 180, 722–731. Lee, Y., Shioi, T., Kasahara, H., Jobe, S.M., Wiese, R.J., Markham, B.E., Izumo, S., 1998. The cardiac tissue-restricted homeobox protein Csx/ Nkx2.5 physically associates with the zinc finger protein GATA4 and cooperatively activates Atrial Natriuretic Factor gene expression. Mol. Cell Biol. 18, 3120–3129. Lee, Y., Song, A.J., Baker, R., Micales, B., Conway, S.J., Lyons, G.E., 2000. Jumonji, a nuclear protein that is necessary for normal heart development. Circ. Res. 86, 932–938. Li, Q.Y., Newbury-Ecob, R.A., Terrett, J.A., Wilson, D.I., Curtis, A.R., Yi, C.H., Gebuhr, T., Bullen, P.J., Robson, S.C., Strachan, T., Bonnet, D., Lyonnet, S., Young, I.D., Raeburn, J.A., Buckler, A.J., Law, D.J., Brook, J.D., 1997. Holt-Oram syndrome is caused by mutations in TBX5, a member of the Brachyury (T) gene family. Nat. Genet. 15, 21–29. Li, T., Li, Y.M., Jia, Z.Q., Chen, P., Ma, K.T., Zhou, C.Y., 2007. Carboxyl terminus of Nkx2.5 impairs its interaction with p300. J. Mol. Biol. 370, 976–992. Liberatore, C.M., Searcy-Schrick, R.D., Vincent, E.B., Yutzey, K.E., 2002. Nkx-2.5 gene induction in mice is mediated by a Smad consensus regulatory region. Dev Biol. 244, 243–256. Lien, C.L., Wu, C., Mercer, B., Webb, R., Richardson, J.A., Olson, E.N., 1999. Control of early cardiac-specific transcription of Nkx2-5 by a GATA-dependent enhancer. Development 126, 75–84. Lien, C.L., McAnally, J., Richardson, J.A., Olson, E.N., 2002. Cardiacspecific activity of an Nkx2-5 enhancer requires an evolutionarily conserved Smad binding site. Dev. Biol. 244, 257–266. Lim, J.Y., Kim, W.H., Kim, J., Park, S.I., 2007. Induction of Id2 expression by cardiac transcription factors GATA4 and Nkx2.5. J. Cell Biochem. 103, 182–194. Linhares, V.L., Almeida, N.A., Menezes, D.C., Elliott, D.A., Lai, D., Beyer, E.C., Campos de Carvalho, A.C., Costa, M.W., 2004. Transcriptional regulation of the murine Connexin40 promoter by cardiac factors Nkx2-5, GATA4 and Tbx5. Cardiovasc. Res. 64, 402–411. Linke, A., Muller, P., Nurzynska, D., Casarsa, C., Torella, D., Nascimbene, A., Castaldo, C., Cascapera, S., Böhm, M., Quaini, F., Urbanek, K., Leri, A., Hintze, T.H., Kajstura, J., Anversa, P., 2005. Stem cells in the dog heart are self-renewing, clonogenic, and multipotent and regenerate infarcted myocardium, improving cardiac function. Proc. Natl. Acad. Sci. USA 102, 8966–8971. Lints, T.J., Parsons, L.M., Hartley, L., Lyons, I., Harvey, R.P., 1993. Nkx2.5: a novel murine homeobox gene expressed in early heart progenitor cells and their myogenic descendants. Development 119, 419–431.

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

Liu, N., Olson, E.N., 2006. Coactivator control of cardiovascular growth and remodeling. Curr. Opin. Cell Biol. 18, 715–722. Lo, P.C., Frasch, M., 2001. A role for the COUP-TF-related gene sevenup in the diversification of cardioblast identities in the dorsal vessel of Drosophila. Mech. Dev. 104, 49–60. Lyons, I., Parsons, L.M., Hartley, L., Li, R., Andrews, J.E., Robb, L., Harvey, R.P., 1995. Myogenic and morphogenetic defects in the heart tubes of murine embryos lacking the homeobox gene Nkx2-5. Genes Dev. 9, 1654–1666. Masino, A.M., Gallardo, T.D., Wilcox, C.A., Olson, E.N., Williams, R.S., Garry, D.J., 2004. Transcriptional regulation of cardiac progenitor cell populations. Circ. Res. 95, 389–397. Masson, N., Greene, W.K., Rabbitts, T.H., 1998. Optimal activation of an endogenous gene by HOX11 requires the NH2-terminal 50 amino acids. Mol. Cell Biol. 18, 3502–3508. McElhinney, D.B., Geiger, E., Blinder, J., Benson, D.W., Goldmuntz, E., 2003. NKX2.5 mutations in patients with congenital heart disease. J. Am. Coll. Cardiol. 42, 1650–1655. Mead, P.E., Brivanlou, I.H., Kelley, C.M., Zon, L.I., 1996. BMP4-responsive regulation of dorsal-ventral patterning by the homeobox protein Mix.1. Nature 382, 357–360. Merika, M., Orkin, S.H., 1993. DNA-binding specificity of GATA family transcription factors. Mol. Cell Biol. 13, 3999–4010. Meysen, S., Marger, L., Hewett, K.W., Jarry-Guichard, T., Agarkova, I., Chauvin, J.P., Perriard, J.C., Izumo, S., Gourdie, R.G., Mangoni, M.E., Nargeot, J., Gros, D., Miquerol, L., 2007. Nkx2.5 cell-autonomous gene function is required for the postnatal formation of the peripheral ventricular conduction system. Dev. Biol. 303, 740–753. Mikawa, T., Hurtado, R., 2007. Development of the cardiac conduction system. Semin. Cell Dev. Biol. 18, 90–100. Miskolczi-Mccallum, C.M., Scavetta, R.J., Svendsen, P.C., Soanes, K.H., Brook, W.J., 2005. The Drosophila melanogaster T-box genes midline and H15 are conserved regulators of heart development. Dev. Biol. 278, 459–472. Molkentin, J.D., 2000. The zinc finger-containing transcription factors GATA-4,-5 and -6. J. Biol. Chem. 275, 38949–38952. Mommersteeg, M.T., Brown, N.A., Prall, O.W., de Gier-de Vries, C., Harvey, R.P., Moorman, A.F., Christoffels, V.M., 2007a. Pitx2c and Nkx2-5 are required for the formation and identity of the pulmonary myocardium. Circ. Res. 101, 902–909. Mommersteeg, M.T., Hoogaars, W.M., Prall, O.W., de Gier-de Vries, C., Wiese, C., Clout, D.E., Papaioannou, V.E., Brown, N.A., Harvey, R.P., Moorman, A.F., Christoffels, V.M., 2007b. Molecular pathway for the localized formation of the sinoatrial node. Circ. Res. 100, 354–362. Monzen, K., Ito, Y., Naito, A.T., Kasai, H., Hiroi, Y., Hayashi, D., Shiojima, I., Yamazaki, T., Miyazono, K., Asashima, M., Nagai, R., Komuro, I., 2008. A crucial role of a high mobility group protein HMGA2 in cardiogenesis. Nat. Cell Biol. 10, 567–574. Moretti, A., Caron, L., Nakano, A., Lam, J.T., Bernshausen, A., Chen, Y., Qyang, Y., Bu, L., Sasaki, M., Martin-Puig, S., Sun, Y., Evans, S.M., Laugwitz, K.L., Chien, K.R., 2006. Multipotent embryonic isl1 progenitor cells lead to cardiac, smooth muscle, and endothelial cell diversification. Cell 127, 1151–1165. Moskowitz, I.P., Kim, J.B., Moore, M.L., Wolf, C.M., Peterson, M.A., Shendure, J., Nobrega, M.A., Yokota, Y., Berul, C., Izumo, S., Seidman, J.G., Seidman, C.E., 2007. A molecular pathway including Id2, Tbx5, and Nkx2-5 required for cardiac conduction system development. Cell 129, 1365–1376. Muhr, J., Andersson, E., Persson, M., Jessell, T.M., Ericson, J., 2001. Groucho-mediated transcriptional repression establishes progenitor cell pattern and neuronal fate in the ventral neural tube. Cell 104, 861–873.

595

Newman, C.S., Krieg, P.A., 1998. Tinman-related genes expressed during heart development in Xenopus. Dev. Genet. 22, 230–238. Nimura, K., Ura, K., Shiratori, H., Ikawa, M., Okabe, M., Schwartz, R.J., Kaneda, Y., 2009. A histone H3 lysine 36 trimethylase links Nkx2-5 to Wolf-Hirschhorn syndrome. Nature 460, 287–292. Nishida, W., Nakamura, M., Mori, S., Takahashi, M., Ohkawa, Y., Tadokoro, S., Yoshida, K., Hiwada, K., Hayashi, K., Sobue, K., 2002. A triad of serum response factor and the GATA and NK ­ families ­governs the transcription of smooth and cardiac muscle genes. J. Biol. Chem. 277, 7308–7317. Oh, H., Bradfute, S.B., Gallardo, T.D., Nakamura, T., Gaussin, V., Mishina, Y., Pocius, J., Michael, L.H., Behringer, R.R., Garry, D.J., Entman, M.L., Schneider, M.D., 2003. Cardiac progenitor cells from adult myocardium: homing, differentiation, and fusion after infarction. Proc. Natl. Acad. Sci. USA 100, 12313–12318. Oka, T., Xu, J., Molkentin, J.D., 2007. Re-employment of developmental transcription factors in adult heart disease. Semin. Cell Dev. Biol. 18, 117–131. Onichtchouk, D., Glinka, A., Niehrs, C., 1998. Requirement for Xvent1 and Xvent-2 gene function in dorsoventral patterning of Xenopus mesoderm. Development 125, 1447–1456. Palmer, S., Groves, N., Schindeler, A., Yeoh, T., Biben, C., Wang, C.C., Sparrow, D.B., Barnett, L., Jenkins, N.A., Copeland, N.G., Koentgen, F., Mohun, T., Harvey, R.P., 2001. The small muscle-specific protein Csl modifies cell shape and promotes myocyte fusion in an insulin-like growth factor 1-dependent manner. J. Cell Biol. 153, 985–998. Park, M., Wu, X., Golden, K., Axelrod, J.D., Bodmer, R., 1996. The wingless pathway is directly involved in Drosophila heart development. Dev. Biol. 177, 104–116. Paroush, Z., Finley, R.L., Kidd, T., Wainwright, S.M., Ingham, P.W., Brent, R., Ish-Horowicz, D., 1994. Groucho is required for Drosophila neurogenesis, segmentation, and sex determination and interacts directly with hairy-related bHLH proteins. Cell 79, 805–815. Pashmforoush, M., Lu, J.T., Chen, H., Amand, T.S., Kondo, R., Pradervand, S., Evans, S.M., Clark, B., Feramisco, J.R., Giles, W., Ho, S.Y., Benson, D.W., Silberbach, M., Shou, W., Chien, K.R., 2004. Nkx2-5 pathways and congenital heart disease; loss of ventricular myocyte lineage specification leads to progressive cardiomyopathy and complete heart block. Cell 117, 373–386. Passner, J.M., Ryoo, H.D., Shen, L., Mann, R.S., Aggarwal, A.K., 1999. Structure of a DNA-bound Ultrabithorax-Extradenticle homeodomain complex. Nature 397, 714–719. Piper, D.E., Batchelor, A.H., Chang, C.-P., Cleary, M.L., Wolberger, C., 1999. Structure of a HoxB1-Pbx1 heterodimer bound to DNA: Role of the hexapeptide and a fourth homeodomain helix in complex formation. Cell 96, 587–597. Prall, O.W., Menon, M.K., Solloway, M.J., Watanabe, Y., Zaffran, S., Bajolle, F., Biben, C., McBride, J.J., Robertson, B.R., Chaulet, H., Stennard, F.A., Wise, N., Schaft, D., Wolstein, O., Furtado, M.B., Shiratori, H., Chien, K.R., Hamada, H., Black, B.L., Saga, Y., Robertson, E.J., Buckingham, M.E., Harvey, R.P., 2007. An nkx2-5/ bmp2/smad1 negative feedback loop controls heart progenitor specification and proliferation. Cell 128, 947–959. Price, M., Lazzaro, D., Pohl, T., Mattei, M.G., Rüther, U., Olivo, J.C., Duboule, D., Di Lauro, R., 1992. Regional expression of the homeobox gene Nkx-2.2 in the developing mammalian forebrain. Neuron 8, 241–255. Qian, L., Liu, J., Bodmer, R., 2005. Neuromancer Tbx20-related genes (H15/midline) promote cell fate specification and morphogenesis of the Drosophila heart. Dev. Biol. 279, 509–524.

596

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Ranganayakulu, G., Elliott, D.A., Harvey, R.P., Olson, E.N., 1998. Divergent roles for NK-2 class homeobox genes in cardiogenesis in flies and mice. Development 125, 3037–3048. Reecy, J.M., Li, X., Yamada, M., DeMayo, F.J., Newman, C.S., Harvey, R.P., Schwartz, R.J., 1999. Identification of upstream regulatory domains in the heart-expressed homeobox gene Nkx2-5. Development 125, 839–849. Reecy, J.M., Yamada, M., Cummings, K., Sosic, D., Chen, C.Y., Eichele, G., Olson, E.N., Schwartz, R.J., 1997. Chicken Nkx-2.8: A novel homeobox gene expressed in early heart progenitor cells and pharyngeal pouch-2 and -3 endoderm. Dev. Biol. 188, 295–311. Reim, I., Frasch, M., 2005. The Dorsocross T-box genes are key components of the regulatory network controlling early cardiogenesis in Drosophila. Development 132, 4911–4925. Reim, I., Mohler, J.P., Frasch, M., 2005. Tbx20-related genes, mid and H15, are required for tinman expression, proper patterning, and normal differentiation of cardioblasts in Drosophila. Mech. Dev. 122, 1056–1069. Riley, P., Anson-Cartwright, L., Cross, J.C., 1998. The Hand1 bHLH transcription factor is essential for placentation and cardiac morphogenesis. Nat. Genet. 18, 271–275. Risebro, C.A., Smart, N., Dupays, L., Breckenridge, R., Mohun, T.J., Riley, P.R., 2006. Hand1 regulates cardiomyocyte proliferation versus differentiation in the developing heart. Development 133, 4595–4606. Rizki, T.M., 1978. The circulation system and associated cells and tissues. In: Ashburner, M., Wright, T.R.F. (Eds.), The Genetics and Biology of Drosophila, Vol. 26. Academic Press, New York, NY, pp. 397–452. Rugendorff, A., Younossi-Hartenstein, A., Hartenstein, V., 1994. Embryonic origin and differentiation of the Drosophila heart. Roux’s Arch. Dev. Biol. 203, 266–280. Ruzinova, M.B., Benezra, R., 2003. Id proteins in development, cell cycle and cancer. Trends Cell. Biol. 13, 410–418. Ryan, K.M., Hendren, J.D., Helander, L.A., Cripps, R.M., 2007. The NK homeodomain transcription factor Tinman is a direct activator of seven-up in the Drosophila dorsal vessel. Dev. Biol. 302, 694–702. Saadane, N., Alpert, L., Chalifour, L.E., 1999. Expression of immediate early genes, GATA-4, and Nkx-2.5 in adrenergic-induced cardiac hypertrophy and during regression in adult mice. Br. J. Pharmacol. 127, 1165–1176. Sandmann, T., Girardot, C., Brehme, M., Tongprasit, W., Stolc, V., Furlong, E.E., 2007. A core transcriptional network for early mesoderm development in Drosophila melanogaster. Genes Dev. 21, 436–449. Sarkozy, A., Conti, E., Neri, C., D’Agostino, R., Digilio, M.C., Esposito, G., Toscano, A., Marino, B., Pizzuti, A., Dallapiccola, B., 2005. Spectrum of atrial septal defects associated with mutations of NKX2.5 and GATA4 transcription factors. J. Med. Genet. 42, e16. Schlange, T., Andree, B., Arnold, H.H., Brand, T., 2000. BMP2 is required for early heart development during a distinct time period. Mech. Dev. 91, 259–270. Schneuwly, S., Klemenz, R., Gehring, W.J., 1987. Redesigning the body plan of Drosophila by ectopic expression of the homoeotic gene Antennapedia. Nature 325, 816–818. Schott, J.-J., Benson, D.W., Basson, C.T., Pease, W., Silberbach, G.M., Moak, J.P., Maron, B.J., Seidman, C.E., Seidman, J.G., 1998. Congenital heart disease caused by mutations in the transcription factor NKX2-5. Science 281, 108–111. Schultheiss, T.M., Xydas, S., Lassar, A.B., 1995. Induction of avian cardiac myogenesis by anterior endoderm. Development 121, 4203–4214.

Schultheiss, T.M., Burch, J.B., Lassar, A.B., 1997. A role for bone morphogenetic proteins in the induction of cardiac myogenesis. Genes Dev. 11, 451–462. Schwartz, R.J., Olson, E.N., 1999. Building the heart piece by piece: modularity of cis-elements regulating Nkx2-5 transcription. Development 126, 4187–4192. Scott, M.P., Tamkun, J.W., Hartzell, G.W., 1989. The structure and function of the homeodomain. Biochim. Biophys. Acta 989, 25–48. Searcy, R.D., Vincent, E.B., Liberatore, C.M., Yutzey, K.E., 1998. A GATA-dependent nkx-2.5 regulatory element activates early cardiac gene expression in transgenic mice. Development 125, 4461–4470. Sepulveda, J.L., Belaguli, N., Nigam, V., Chen, C.-Y., Nemer, M., Schwartz, R.J., 1998. GATA-4 and Nkx2-5 coactivate Nkx-2 DNA binding targets: role for regulating early cardiac gene expression. Mol. Cell Biol. 18, 3405–3415. Sepulveda, J.L., Vlahopoulos, S., Iyer, D., Belaguli, N., Schwartz, R.J., 2002. Combinatorial expression of GATA4, Nkx2-5, and serum response factor directs early cardiac gene activity. J. Biol. Chem. 277, 25775–25782. Shin, C.H., Liu, Z.P., Passier, R., Zhang, C.L., Wang, D.Z., Harris, T.M., Yamagishi, H., Richardson, J.A., Childs, G., Olson, E.N., 2002. Modulation of cardiac growth and development by HOP, an unusual homeodomain protein. Cell 110, 725–735. Small, E.M., Krieg, P.A., 2004. Molecular regulation of cardiac chamberspecific gene expression. Trends Cardiovasc. Med. 14, 13–18. Smith, S.T., Jaynes, J.B., 1996. A conserved region of engrailed, shared among all en-, gsc-, Nk1-, Nk2- and msh-class homeoproteins, mediates active transcriptional repression in vivo. Development 122, 3141–3150. Song, K., Backs, J., McAnally, J., Qi, X., Gerard, R.D., Richardson, J.A., Hill, J.A., Bassel-Duby, R., Olson, E.N., 2006. The transcriptional coactivator CAMTA2 stimulates cardiac growth by opposing class II histone deacetylases. Cell 125, 453–466. Sparrow, D.B., Cai, C., Kotecha, S., Latinkic, B., Cooper, B., Towers, N., Evans, S.M., Mohun, T.J., 2000. Regulation of the tinman homologues in Xenopus embryos. Dev. Biol. 227, 65–79. Srivastava, D., Thomas, T., Lin, Q., Brown, D., Olson, E.N., 1997. Regulation of cardiac mesodermal and neural crest development by the bHLH transcription factor, dHAND. Nat. Genet. 16, 154–160. Stanley, E.G., Biben, C., Elefanty, A., Barnett, L., Koentgen, F., Robb, L., Harvey, R.P., 2002. Efficient Cre-mediated deletion in cardiac progenitor cells conferred by a 3UTR-ires-Cre allele of the homeobox gene Nkx2-5. Int. J. Dev. Biol. 46, 431–439. Stennard, F.A., Costa, M.W., Elliott, D.A., Rankin, S., Haast, S.J., Lai, D., McDonald, L.P., Niederreither, K., Dolle, P., Bruneau, B.G., Zorn, A.M., Harvey, R.P., 2003. Cardiac T-box factor Tbx20 directly interacts with Nkx2-5, GATA4, and GATA5 in regulation of gene expression in the developing heart. Dev. Biol. 262, 206–224. Sultana, N., Nag, K., Hoshijima, K., Laird, D.W., Kawakami, A., Hirose, S., 2008. Zebrafish early cardiac connexin, Cx36.7/Ecx, regulates myofibril orientation and heart morphogenesis by establishing Nkx2.5 ­expression. Proc. Natl. Acad. Sci. USA 105, 4763–4768. Takeuchi, T., Watanabe, Y., Takano-Shimizu, T., Kondo, S., 2006. Roles of jumonji and jumonji family genes in chromatin regulation and development. Dev. Dyn. 235, 2449–2459. Takimoto, E., Mizuno, T., Terasaki, F., Shimoyama, M., Honda, H., Shiojima, I., Hiroi, Y., Oka, T., Hayashi, D., Hirai, H., Kudoh, S., Toko, H., Kawamura, K., Nagai, R., Yazaki, Y., Komuro, I., 2000. Up-regulation of natriuretic peptides in the ventricle of Csx/Nkx2-5 transgenic mice. Biochem. Biophys. Res. Commun. 270, 1074–1079.

Chapter  |  9.1  NK-2 Class Homeodomain Proteins: Conserved Regulators of Cardiogenesis

Tanaka, M., Chen, Z., Bartunkova, S., Yamasaki, N., Izumo, S., 1999a. The cardiac homeobox gene Csx/Nkx2.5 lies genetically upstream of multiple genes essential for heart development. Development 126, 1269–1280. Tanaka, M., Wechsler, S.B., Lee, I.W., Yamasaki, N., Lawitts, J.A., Izumo, S., 1999b. Complex modular cis-acting elements regulate expression of the cardiac specifying homeobox gene Csx/Nkx2.5. Development 126, 1439–1450. Tanaka, M., Yamasaki, N., Izumo, S., 2000. Phenotypic characterization of the murine Nkx2.6 homeobox gene by gene targeting. Mol. Cell Biol. 20, 2874–2879. Tao, Y., Schulz, R.A., 2007. Heart development in Drosophila. Semin. Cell Dev. Biol. 18, 3–15. Targoff, K.L., Schell, T., Yelon, D., 2008. Nkx genes regulate heart tube extension and exert differential effects on ventricular and atrial cell number. Dev. Biol. 322, 314–321. Tell, G., Acquaviva, R., Formisano, S., Fogolari, F., Pucillo, C., Damante, G., 1999. Comparative stability analysis of the thyroid transcription factor 1 and Antennapedia homeodomains: evidence for residue 54 in controlling the structural stability of the recognition helix. Int. J. Biochem. Cell Biol. 31, 1339–1353. Thomas, P.S., Kasahara, H., Edmonson, A.M., Izumo, S., Yacoub, M.H., Barton, P.J., Gourdie, R.G., 2001. Elevated expression of Nkx-2.5 in developing myocardial conduction cells. Anat. Rec. 263, 307–313. Thompson, J.T., Rackley, M.S., O’Brien, T.X., 1998. Upregulation of the cardiac homeobox gene Nkx2-5 (CSX) in feline right ventricular pressure overload. Am. J. Physiol. 274, H1569–H1573. Tolkunova, E.N., Fujioka, M., Kobayashi, M., Deepali, D., Jaynes, J.B., 1998. Two distinct types of repression domain in Engrailed: One interacts with the Groucho corepressor and is preferentially active on integrated target genes. Mol. Cell Biol. 18, 2804–2814. Tonissen, K.F., Drysdale, T.A., Lints, T.J., Harvey, R.P., Krieg, P.A., 1994. XNkx2.5, a Xenopus gene related to Nkx-2.5 and tinman: Evidence for a conserved role in cardiac development. Dev. Biol. 162, 325–328. Tsao, D., Gruschus, J., Wang, L.-H., Nirenberg, M., Ferretti, J., 1994. Elongation of helix III of the NK-2 homeodomain upon binding to DNA: a secondary structure study by NMR. Biochemistry 33, 15053–15060. Uhler, J., Zhang, H., Syu, L.J., Mellerick, D.M., 2007. The Nk-2 box of the Drosophila homeodomain protein, Vnd, contributes to its repression activity in a Groucho-dependent manner. Mech. Dev. 124, 1–10. Urbanek, K., Quaini, F., Tasca, G., Torella, D., Castaldo, C., NadalGinard, B., Leri, A., Kajstura, J., Quaini, E., Anversa, P., 2003. Intense myocyte formation from cardiac stem cells in human cardiac hypertrophy. Proc. Natl. Acad. Sci. USA 100, 10440–10445. Urbanek, K., Cesselli, D., Rota, M., Nascimbene, A., De Angelis, A., Hosoda, T., Bearzi, C., Boni, A., Bolli, R., Kajstura, J., Anversa, P., Leri, A., 2006. Stem cell niches in the adult mouse heart. Proc. Natl. Acad. Sci. USA 103, 9226–9231. Vega, R.B., Harrison, B.C., Meadows, E., Roberts, C.R., Papst, P.J., Olson, E.N., McKinsey, T.A., 2004. Protein kinases C and D mediate agonist-dependent cardiac hypertrophy through nuclear export of histone deacetylase 5. Mol. Cell Biol. 24, 8374–8385. Viglino, P., Fogolari, F., Formisano, S., Bortolotti, N., Damante, G., Di Lauro, R., Esposito, G., 1993. Structural study of rat thyroid transcription factor 1 homeodomain (TTF-1 HD) by nuclear magnetic resonance. FEBS. Lett. 336, 397–402.

597

Von Both, I., Silvestri, C., Erdemir, T., Lickert, H., Walls, J.R., Henkelman, R.M., Rossant, J., Harvey, R.P., Attisano, L., Wrana, J.L., 2004. Foxh1 is essential for development of the anterior heart field. Dev. Cell 7, 331–345. Wang, J., Zhang, H., Iyer, D., Feng, X.H., Schwartz, R.J., 2008. Regulation of cardiac specific nkx2.5 gene activity by small ubiquitin-like modifier. J. Biol. Chem. 283, 23235–23243. Watada, H., Mirmira, R.G., Kalamaras, J., German, M.S., 2000. Intramolecular control of transcriptional activity by the NK2-specific domain in NK-2 homeodomain proteins. Proc. Natl. Acad. Sci. USA 97, 9443–9448. Watanabe, Y., Benson, D.W., Yano, S., Akagi, T., Yoshino, M., Murray, J. C., 2002. Two novel frameshift mutations in NKX2.5 result in novel features including visceral inversus and sinus venosus type ASD. J. Med. Genet. 39, 807–811. Weisberg, E., Winnier, G.E., Chen, X., Farnsworth, C.L., Hogan, B.L., Whitman, M., 1998. A mouse homologue of FAST-1 transduces TGF beta superfamily signals and is expressed during early embryogenesis. Mech. Dev. 79, 17–27. Wu, S.M., Fujiwara, Y., Cibulsky, S.M., Clapham, D.E., Lien, C.L., Schultheiss, T.M., Orkin, S.H., 2006. Developmental origin of a bipotential myocardial and smooth muscle cell precursor in the mammalian heart. Cell 127, 1137–1150. Wu, X., Golden, K., Bodmer, R., 1995. Heart development in Drosophila requires the segment polarity gene wingless. Dev. Biol. 169, 619–628. Xu, X., Yin, Z., Hudson, J.B., Ferguson, E.L., Frasch, M., 1998. Smad proteins act in combination with synergistic and antagonistic regulators to target Dpp responses to the Drosophila mesoderm. Genes Dev. 12, 2354–2370. Yamagishi, H., Yamagishi, C., Nakagawa, O., Harvey, R.P., Olson, E.N., Srivastava, D., 2001. The combinatorial activities of Nkx2.5 and dHAND are essential for cardiac ventricle formation. Dev. Biol. 239, 190–203. Yin, Z., Xu, X.L., Frasch, M., 1997. Regulation of twist target gene tinman by modular cis-regulatory elements during early mesoderm development. Development 124, 4971–4982. Yin, Z., Frasch, M., 1998. Regulation and function of tinman during dorsal mesoderm induction and heart specification in Drosophila. Dev. Genet. 22, 187–200. Zaffran, S., Frasch, M., 2002. Early signals in cardiac development. Circ. Res. 91, 457–469. Zaffran, S., Reim, I., Qian, L., Lo, P.C., Bodmer, R., Frasch, M., 2006. Cardioblast-intrinsic Tinman activity controls proper diversification and differentiation of myocardial cells in Drosophila. Development 133, 4073–4083. Zeitlinger, J., Zinzen, R.P., Stark, A., Kellis, M., Zhang, H., Young, R.A., Levine, M., 2007. Whole-genome ChIP-chip analysis of Dorsal, Twist, and Snail suggests integration of diverse patterning processes in the Drosophila embryo. Genes Dev. 21, 385–390. Zhou, B., Von Gise, A., Ma, Q., Rivera-Feliciano, J., Pu, W.T., 2008. Nkx2-5- and Isl1-expressing cardiac progenitors contribute to proepicardium. Biochem. Biophys. Res. Commun. 375, 450–453. Zhu, W., Shiojima, I., Hiroi, Y., Zou, Y., Akazawa, H., Mizukami, M., Toko, H., Yazaki, Y., Nagai, R., Komuro, I., 2000. Functional analyses of three Csx/Nkx2-5 mutations that cause human congenital heart disease. J. Biol. Chem. 275, 35291–35296.

Chapter 9.2

GATA4 in Heart Development and Disease Georges Nemer1 and Mona Nemer2 Research Unit in Cardiac Growth and Differentiation, Institut de recherches cliniques de Montréal (IRCM), Montréal, Quebec, Canada 1 Department of Biochemistry, American University of Beirut, Lebanon  2 Department of Biochemistry, University of Ottawa, Ottawa, Ontario, Canada

I.  The GATA family of zinc-finger transcription factors I.A.  Overview GATA proteins belong to the large evolutionary-conserved family of zinc-finger transcription factors characterized by the presence of zinc ions as a major component of their structure (Lowry and Atchley, 2000; Patient and McGhee, 2002). Members of the GATA subfamily are found in fungi, metazoans and plants, where they play essential roles in cellular growth and differentiation. As many as 29 GATA proteins exist in Arabidobsis thaliana, and whereas invertebrates like Caenorhabditis elegans have 11 factors, higher vertebrates have only 6 (Scazzocchio, 2000; Patient et al., 2002; Reyes et al., 2004; Manfield et al., 2007). The phylogenetic analysis of GATA proteins from different organisms has, at least in the case of vertebrates, identified gene duplication through evolution as a plausible mechanism for their existence in relation to one common ancestor (Lowry et al., 2000). GATA proteins share the ability to bind, through their zinc-finger DNA-binding domain, sequence-specific DNA motifs containing a stretch of six nucleotides HGATAR (H  A,C,T and R  A,G). The zinc-finger domains of GATA proteins belong to the C4 subfamily of zinc-finger proteins with the following consensus sequence CX2CX17–20 CX2C. Higher vertebrate GATA factors possess two zincfinger domains encoded by two different exons, while plants and fungi GATA proteins primarily have only one zinc-finger motif. The N-terminal zinc-finger is thought to have been duplicated during evolution, although the fact that some fungi and lower vertebrates have GATA proteins Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

with two zinc-fingers argues against such a hypothesis (Lowry et al., 2000). The C-terminal zinc-finger is the one that shares the highest homology between vertebrates and other species. It is through this region and the stretch of basic amino acids adjacent to it that GATA proteins bind specifically to the HGATAR sequence. The N-terminal zinc-finger of vertebrate GATA factors was shown to possess the ability to bind the HGATAR sequence, albeit with lower affinity, and this is believed to help stabilize the interaction of the carboxy zinc-finger to the DNA. Other than DNA binding, both zinc-finger domains are involved in extensive protein–protein interactions. Finally, the DNA-binding domain is flanked by amino and carboxy terminal domains that are involved in transcriptional activation or repression (Fig. 1).

I.B.  General Properties In general, GATA proteins are expressed in a cell- and temporal-specific manner. Their essential roles in early development are well-established and include lineage differentiation, as well as cell survival and proliferation (Table 1). For example, in plants, GATA proteins are involved in flower development, whereas in C. elegans, fish and mammals, the different GATA proteins are crucial for lineage specification and organ development (Patient et al., 2002; Manfield et al., 2007). This is well-characterized in C. elegans, where the Elt and END proteins are involved in the commitment of cells into the endodermal lineage whereas the MED proteins are involved in specification of the mesendodermal cells (Broitman-Maduro et al., 2005; Maduro et al., 2005). This role in lineage commitment is 599

600

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

(A)

Protein interaction DNA binding 1

177 217

334

440

Zn ++

GATA4

TAD

GATA5

49.7%

87.6%

36.9%

GATA6

49.5%

91.2%

44.2%

Zn

TAD

(B)

Figure 1  GATA proteins. (A) Structure of the GATA4 protein showing two activating domains in the amino and carboxy terminal domains of the protein. The DNA-binding domain is composed of two zinc-finger regions (Zn) that mediate DNA interaction and protein–protein interaction. The basic region () is involved in nuclear localization and stabilizes DNA binding. The high homology between GATA4, 5 and 6 is in the DNA-binding domain (85%) whereas the least homologous region is the carboxy terminal region. (B) The second zinc-finger of the murine GATA4 protein bound to DNA. Homology-based modeling was used to dock the second zinc-finger of GATA4 to DNA, based on the NMR structure of GATA1.

Table 1  Evolutionary Conserved Role of GATA Factors in Development Species

Factor

Function

Plant

CGA1,GNC, BME3, HAN

Sugar metabolism Flower development Seed germination

Worm

elt-1,2,3,4,5,6,7 END1,2,med-1,2

Gut development Mesendoderm development

Fungus

asd4,nreB,areA,areB, Gln3p,DAL80, GZF3,NIL

Nitrogen metabolism Siderophorebiosynthesis Circadian rhythm

Fly

Pannier, Serpent, Grain GATAd, GATAe

Cardiovascular and blood development

Fish

GATA1,2,3,4,5,6

Cardiovascular, blood and gut development

Mouse

GATA1,2,3,4,5,6

Cardiovascular, blood, gonads, gut and lungs development

crucial for the evolution of metazoans, where compartmentalization via cell layer formation is a major event. GATA proteins also play essential roles throughout the lifecycle of the various organisms as they act as sensor/effectors of external threats and stimuli, and thus contribute to the inner defense mechanism of the cell. This particular role is conserved during evolution from fungi to higher vertebrates. For example, the areA protein plays an essential role in nitrogen metabolism in Aspergillus nidulans (Wilson and Arst Jr, 1998) and GATA proteins NIL1 and GLN3 are crucial regulators of carbon and nitrogen metabolism in Saccharomyces cerevisiae (Kuruvilla et al., 2001). As will be discussed in this chapter, mammalian GATA proteins play an important role in cardiac stress responses, and are critical regulators of cell survival.

II.  Expression of GATA proteins in cardiovascular cells II.A.  The Two Subfamilies of Vertebrate   GATA Proteins In vertebrates, the six GATA proteins are divided into two subfamilies based on their sequence homology and tissue distribution. GATA1, 2 and 3 are largely expressed in the hematopoietic system, where they play nonredundant roles in lineage specification and differentiation. Inactivation of any of the three genes in mice results in embryonic lethality at midgestation, and results have identified essential roles for GATA1 in terminal differentiation of erythroid and megakaryotic cells, and for GATA2 in expansion of hematopoietic progenitor cells (Tsai et al., 1994; Fujiwara et al., 1996; Harigae et al., 1998). Within the hematopoietic system, GATA3 is restricted to the lymphocyte lin­eage, and is required for terminal differentiation of T-cells (Pandolfi et al., 1995). In addition to their expression in the hematopoietic system, GATA1, 2 and 3 are differentially-expressed in other organs where they have important functions. In particular, GATA3 is critical for kidney formation and for development of the dopaminergic neurons (Lakshmanan et al., 1999). The important role of GATA factors in blood formation is conserved in fly where serpent acts as the functional homolog of GATA1, 2 and 3 (Waltzer et al., 2002). Finally, the relevance of these GATA factors to normal development is underscored by the finding that mutations in GATA1 and 3 are associated with human diseases (Table 2). The second subfamily includes GATA4, 5 and 6, whose expression is predominant in the heart and in the endodermal cells of the digestive system. This chapter will focus on the regulation and function of members of this subclass in cardiovascular cells. However, it should be noted that GATA4 is also expressed in the gonads where it has

601

Chapter  |  9.2  GATA4 in Heart Development and Disease

Table 2  Mutations in Gata Genes Cause Human Diseases Gene

Human disease

References

Gata1

Thrombocytopenia, Leukemia, Thalassemia

De Gobbi et al., 2006; Hollanda et al., 2006; Yu et al., 2002; Nichols et al., 2000

Gata3

HDR (hypoparathyroidism-deafness-renal) syndrome

Hernandez et al., 2007; Van Esh et al., 2000

Gata4

ASD, VSD, PS, TOF, DORV

Schluterman et al., 2007; Rajagopal et al., 2007; Nemer et al., 2006; Tang et al., 2006; Garg et al., 2003

(A) G4

Figure 2  GATA proteins in the developing heart (E11.5). (A) GATA4 and GATA5 are highly-expressed in the endocardial cushions (EC). Note the high expression of GATA5 in the endocardium and some underlying myocardial cells. GATA4 expression is detected both in the endocardium and myocardium. Photos are taken at 10 magnification. (B) Predominant expression of GATA6 in the left ventricle (LV) at E11.5. Note the difference in the number of GATA6stained nuclei (brown color) between left and right ventricles, while GATA4 expression is the same between the two subcompartments. Photos are taken at 20 magnification.

G5 EC

LV RV

B) G4

RV

LV

G6

an important function in Sertoli cell development (Viger et al., 1998; Tevosian et al., 2002) and GATA6 is found in the lung, where it regulates epithelial cell differentiation (Keijzer et al., 2001; Yang et al., 2002). Detailed analysis of the expression of cardiac GATA proteins in the heart reveals a distinct, but at times overlapping, pattern of expression with GATA4 being virtually omnipresent in all cells.

II.B.  GATA Proteins in the Myocardium Prior to the gastrulation stage GATA4, 5 and 6 are expressed predominantly in the extra-embryonic endoderm, and then restricted to the mesodermal part of the embryo that will form the heart and the underlying endoderm. In the developing heart and as early as the formation of the cardiac crescent, GATA4 and 6 are expressed in the myocardial layer in

an almost perfect overlapping way. Interestingly, the protein expression of GATA4 and GATA6 is strongest at the posterior end (sinus venosus and atrium), while the myocardial expression of GATA5 is rather weak and only becomes evident after looping in few myocardial cells adjacent to the endocardium and endocardial cushions (Nemer and Nemer, 2003). Myocardial GATA5 expression is restricted to few atrial cells and disappears completely at embryonic day (E) 14.5. In contrast, GATA4 and GATA6 expression persists in myocardial cells of the atria and ventricles. Interestingly, not all myocardial cells contain GATA6 but all GATA6-positive cells co-express GATA4; this suggests the existence of different myocyte subpopulations (Fig. 3). As will be discussed in the following sections, the exact functions of GATA5 and GATA6 in the myocardium remain to be established, but GATA4 has emerged as a significant regulator of numerous myocardial processes and cell fate decisions.

602

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

GATA4

GATA6

Hoechst

Merge

Figure 3  Expression of GATA4 and GATA6 during heart morphogenesis. Coexpression of GATA4 and GATA6 in the E10.5 bulbus cordis showing the nuclear staining for both. In blue, control staining for the nuclei with Hoechst, in green GATA4, and in red GATA6. Note how all red nuclei (GATA6) are also positive for GATA4 (green). Magnification is 20.

(A) mGata4 E1a

E1b E2

E3

E4 E5 E6

E7

(B) mGata5 cGata5

E1a E1b E2

E3

E4 E5 E6

E7

(C) mGata6 hGata6

E1a E1b E2

E3

E4 E5 E6

E7

Figure 4  Transcriptional regulation of Gata genes by alternative splicing. (A and C) Both the human (h) Gata4 and the murine (m) and human (h) Gata6 genes have two initiating exons that will lead to two mRNA with different 5 untranslated regions. The chicken (c) and mouse (m) Gata5 genes give two different mRNAs; the alternative splicing occurs by skipping exon 2 to generate a protein that lacks the amino terminal region (B).

II.C.  GATA Proteins in the Endocardium

II.D.  GATA Proteins in the Outflow Tract

Two GATA factors, GATA4 and GATA5, are present in the endocardium and the endocardium is the major site of GATA5 expression in the heart (Fig. 4). GATA5 is first expressed in the endocardial cells of the cardiac crescent; its expression is maintained until midgestation but disappears abruptly thereafter. In the endocardial cushions, GATA5 expression transiently overlaps with GATA4, but only GATA4 expression persists in the mesenchymal cells of the cushions that will form the valves (Nemer et al., 2003). This distribution suggests a very specific role for GATA5 in endocardial development and a broader function for GATA4 at early, as well as late, stages of endocardial differentiation and septal and valve formation.

The outflow tract (OT) and the right ventricle (RV) are composed of different cell types including myocytes and cardiac neural crest cells that are involved in septation (see Chapters 2.2 and 3.1). Recent studies suggest that myocytes of the OT and RV are derived from a distinct embryologic field termed the second heart field which originates from the splanchnic mesoderm underlying the pharynx (Waldo et al., 2001). They are thus different from the myocytes and endocardial cells that form the early heart tube. GATA4 is highly-expressed in the outflow tract, whereas GATA6 is not present in all cells, suggesting a differential role for the two proteins in these compartments. In zebrafish, GATA4 was detected in the cells of the

603

Chapter  |  9.2  GATA4 in Heart Development and Disease

177

217

TAD

Zn

In situ hybridization and immunohistochemical studies have generally revealed concordance between the presence of GATA transcripts and proteins in the developing heart, suggesting that spatial specification is largely determined at the transcriptional level (Morrisey et al., 1996, 1997; Nemer et al., 2003). However, the mechanisms controlling transcription of Gata4, 5 or 6 in cardiac cells have not been elucidated. The available evidence suggests the existence for each gene of an alternate exon 1, hence distinct promoters (Fig. 6). Whether these are differentially used in the various cell types expressing each Gata gene, or during development, remains to be determined. Stimuli that were found to increase GATA4 mRNA accumulation during development include retinoic acid (Arceci et al., 1993; Kostetskii et al., 1999) and BMP4 (Monzen et al., 1999). Another member of the BMP family, BMP2, was also shown to upregulate GATA5 in the zebrafish heart (Reiter et al., 2001).

S4 S419 20 440

TAD

Zn ++

c

b

c

b d

d

TLC + TLEpH 1.9

1

440 ++

200

++

440

Figure 5  Post-translational modification of GATA4. (A) Sites for phosphorylation and sumoylation are indicated on the murine GATA4 protein. (B) Phosphopeptide mapping analysis of 32P-labeled wild-type (left) and N-terminal deleted (right) GATA4; note the absence of spot “a” when the N-terminal domain is removed.

LacZ

ANF

GATA4

III.A.  Regulation of Gene Expression

334

a

Actin F

The highly-specific pattern of expression of GATA factors in the heart is regulated at multiple levels, including transcriptional, post-transcriptional and nuclear/cytoplasmic shuttling.

PKC

(B)

II.E.  GATA Proteins in Other Cardiovascular Cells

III.  Regulation of cardiac GATA factors

K3 6

S2 6

1

a

The epicardium contributes to the formation of numerous cardiac structures, notably the coronary arteries in which endothelial and smooth muscle are the major derivatives (see Chapter 5.1). Both GATA4 and GATA6 are expressed in the epicardium, but only GATA4 expression persists in the coronary arteries into adulthood. Beside the coronary arteries, other vascular cells that contain GATA4 and 6 are present in the smooth muscle of the pulmonary artery and GATA6, but GATA4 is expressed in the smooth muscle cells of the aorta (Nemer et al., 2003). Finally, GATA4 is found in the endothelial cells of the coronary arteries, but other vascular endothelial cells do not have detectable levels of GATA4, 5 or 6.

Sumo

5

PKA

1

MAPK

5

(A)

S1 0

second heart field anterior to the cardiogenic compartment (Serbedzija et al., 1998). This, together with the finding that GATA motifs in the promoter region of the genes for several cardiac TFs (MEF2C, Nkx2-5 and Hand2) are sufficient to drive their expression in the second heart field, suggests that GATA4 may be a transregulator of second heart field cell identity (Jiang et al., 1999; McFadden et al., 2000; Dodou et al., 2004). Finally, GATA6 is present in neural crest cells, where it plays a critical role in vascular patterning.

Figure 6  GATA4 as effector of upstream regulators.

In postnatal myocytes, Gata4 transcription correlates positively with cell growth. A number of myocyte growth factors including endothelin 1 and BMP4 were shown to upregulate GATA4 transcripts (Monzen et al., 1999). GATA4, but not GATA6, mRNA was also increased in a genetic model of pressure-associated cardiac hypertrophy (our unpublished results). Conversely, Gata4 transcription is downregulated by adryamycin, a known cardiotoxic agent (Aries et al., 2004). As further discussed below, GATA4 is thus both a target and an effector of numerous cardioregulators. Production of GATA4, 5 and 6 proteins is also controlled at the RNA splicing level, as well as at the level of

604

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

translation initiation. The first evidence for the existence of distinct GATA protein isoforms due to alternative splicing came from studies of chicken Gata5 where exon skipping results in the production of a shorter isoform lacking the entire N-terminal domain and the first zinc-finger (MacNeill et al., 1997). Interestingly, this isoform retains the ability to bind DNA and to activate target promoters, albeit at a lower level than the full length protein (Nemer et al., 1999). Recently, this alternative splicing was also observed in the murine Gata5 gene. The regulatory mechanism that leads to the generation of a GATA isoform containing one zinc-finger appears to be evolutionaryconserved, as it also has been reported for the Drosophila GATA factor serpent (Waltzer et al., 2002) and the Bombyx mori GATA (Drevet et al., 1995). Different GATA6 protein isoforms are also produced through alternative translation initiation. In this case, the longer protein has an extended N-terminal domain and greater transcriptional activation potential (Brewer et al., 1999; Aries et al., 2004). Immunohistochemical localization using specific antibodies revealed distinct but also co-localization of the two isoforms in cardiovascular cells (Brewer et al., 2002; Aries et al., 2004). The exact function and target genes of the different GATA5 and 6 isoforms are not yet defined, but based on studies of other GATA proteins it is tempting to speculate that they may control distinct processes.

III.B.  Regulation of Protein Activity GATA protein activity can be extensively modulated posttranslationally as a consequence of various modifications, including phosphorylation, acetylation and sumoylation. In this respect, the most extensively-analyzed cardiac GATA factor is GATA4. The GATA4 protein is subject to posttranslational modifications that affect its DNA-binding activity, transactivation potential and subcellular distribution. First, GATA4 contains numerous potential phosphor­ ylation sites within its N- and C-terminal transactivation domains, as well as in its DNA-binding domain. Twodimensional gel analysis of endogenous GATA4 in postnatal cardiomyocytes revealed the presence of at least seven phosphopeptides, indicating that GATA4 is indeed a phosphoprotein (Charron et al., 2001). It is now well-established that the N-terminal transactivation domain is targeted by mitogen-activated protein kinases (MAPK) in response to growth-promoting physical and hormonal stimuli, leading to phosphorylation of S105 and enhanced transcriptional activation potential (Charron et al., 2001; Pikkarainen et al., 2003; Tenhunen et al., 2004). In myocytes, activation of protein kinase C also leads to phosphorylation of the C-terminal transactivation domains at S419,420. This is associated with enhanced DNA-binding activity (Wang et al., 2005). Protein kinase A phosphorylation

of S261 just N-terminal of the second zinc-finger (Fig. 1) also leads to increased transcriptional activation, possibly due to enhanced recruitment of the CBP coactivator (Tremblay and Viger, 2003). Given the essential roles of MAPK, PKC and PKA in myocyte differentiation and in the regulation of cardiac growth and contractility, these findings suggest that GATA4 activity is dynamically regulated in response to physiological and pathophysiological cardiac stimuli. Finally, glycogen synthase kinase 3 (GSK3) was shown to directly interact with GATA4 and phosphorylate its N-terminal domain, causing nuclear export and cytoplasmic redistribution via a Crm1dependent mechanism (Morisco et al., 2001; Philips et al., 2007). At this stage, the physiological relevance of this finding has not been further explored. In addition to phosphorylation, GATA4 was shown to be sumoylated at Lys36 within the C-terminal transactivation domains, resulting in increased transcriptional activity (Wang et al., 2004). Similarly, GATA4 transcriptional properties are enhanced by acetylation, a modification that was found to occur during embryonic stem cell differentiation into cardiomyocytes (Kawamura et al., 2005). In addition to GATA4, GATA6 was also found to be targeted by PKA. The precise PKA phosphorylation site on GATA6 is not yet determined, but PKA activation leads to proteolytic cleavage and degradation of GATA6 via a proteasome-dependent mechanism (Maeda et al., 2005). Thus, activation of the PKA pathway in cardiomyocytes may potentially have opposite effects on GATA4 and GATA6. In summary, GATA protein expression and activity are modulated through diverse mechanisms and at multiple checkpoints. This complex and exquisite regulation is

RA

Growth Adrenergic factors agonists

GATA4

Hyperglycemia Adriamycin

GATA4

Apoptosis Decreased Impaired Contractility Stress Response

Differentiation Proliferation Migration Hypertrophy Figure 7  Effect of GATA4 overexpression in cultured cardiomyocytes. (A) GATA4-infected cells show reorganization of the cytoskeleton, as shown by the expression of actin- (in green) as compared to LacZinfected cells. (B) GATA4- and LacZ-infected cardiomyocytes show similar expression of ANF (in red).

605

Chapter  |  9.2  GATA4 in Heart Development and Disease

Table 3  Cardiac Phenotypes Associated to Loss of GATA4, 5, or 6 Function Species

Gene

Method

Targeted cells

Viability

Cardiac phenotype

Mouse

Gata4 Gata4 Gata4 Gata4 Gata4 Gata4 Gata4 Gata4

Total KO Knock in MLC1 Cre MHC Cre MHC Cre Nkx2.5 Cre Tie2 Cre Flox/floxallele Hypomorph

All All Emb RV Emb/Neo myocytes Emb myocytes Emb myocytes Emb/Neo endothelial All

E8.5–9 E12.5 E8.5–9 Viable Viable E12.5–14 E12.5 E13.5–16

Cardia bifida DORV, TA Cardia bifida Hypoplasia Hypoplasia RV hypoplasia AV cushion defect DORV, CAVC

Zebrafish

Gata4 Gata5

Morpholinos Mutation

All All

E8.5–9 E8.5–9

Looping defects Cardia bifida, endocardial defect

Mouse

Gata5 Gata6 Gata6 Gata6 Gata6 Gata6

Total KO Total KO SM22 Cre Wnt1 Cre SP-C/DN KO tetraploid

All All SM neural crest Neural crest Lung epithelium All

Viable E6.5–7 E18.5–P2 E18.5–P1 P10 E9.5–10.5

None Gastrulation defect VSD, CT defect VSD, CT defect None None

Zebrafish

Gata6

Morpholinos

All

E8.5–9

Cardia bifida

Xenopus

Gata6

Morpholinos

All

E8.5–9

Heartless

consistent with the critical dosage-sensitive function of GATA4, and likely other GATAs, at various cardiac developmental stages.

IV.  Role of GATA factors in embryonic heart development The expression patterns of GATA4, 5 and 6 are consistent with a role for these factors in regulation of heart development, and over the past decade several studies have provided ample evidence for this. The importance of GATA factors in heart development was first evidenced in loss of GATA4 function in the P19 embryonic stem cell line in which cardiac differentiation is induced by DMSO. These studies revealed that knockdown of Gata4 using an antisense approach blocks cardiogenesis at an early stage (Grépin et al., 1995), while overexpression of Gata4 enhances cardiogenesis (Grépin et al., 1997). Consistent with this, injection of GATA4/5/6 mRNA into fertilized Xenopus eggs lead to precocious expression of MHC and cardiac actin (Jiang and Evans, 1996). As shown in Table 3, loss-of-function studies in mice and other species have largely confirmed the critical role of GATA4/5/6 in various stages of heart morphogenesis.

IV.A.  GATA4 It is now well-established that GATA4 is essential for normal heart development, as best evidenced by the fact that GATA4 mutations cause congenital heart disease in

humans (detailed in Section IX below). Targeted deletion of mouse Gata4 during development systematically results in abnormal hearts, in which numerous structures are affected (Table 3). The bulk of the studies indicate that GATA4 is independently required for various stages of myocardial, as well as endocardial, development. Gata4null embryos display Cardia bifida and hypoplastic ventricles (Kuo et al., 1997; Molkentin et al., 1997). Tetraploid complementation rescues the cardia bifida phenotype, indicating that it is likely due to defects in extraembryonic cells; however, in Gata4/ embryos, development arrests around E9.5–10, the hearts do not undergo looping and chamber formation but are characterized by myocardial thinning and absence of proepicardium and endocardial cushion formation (Watt et al., 2004). A similar hypoplastic heart phenotype is produced with a knockin mutation in Gata4, which abrogates interaction with the FOG collaborator (see Section VIII); in this case, embryos survive until E12.5 and display, in addition to myocardial hypoplasia, septation and valve defects such as common atrioventricular canal (CAVC) and double-outlet right ventricle (DORV) (Crispino et al., 2001). Myocardial thinning is a consistent feature whenever Gata4 is deleted from embryonic cardiomyocytes, leading to the suggestion that GATA4 may be required for myocyte proliferation (Pu et al., 2004; Zeisberg et al., 2005). Gata4 was also specific­ ally deleted from endothelial/endocardial cells. This leads to hypoplastic valve leaflets unable to ensure proper blood circulation, and revealed an important role for GATA4 in cushion mesenchyme formation (Rivera-Feliciano et al., 2006). Defective epithelial–mesenchymal transformation

606

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

is consistent with a role in differentiation. Additionally, decreased cellular proliferation of cushion mesenchyme suggests a conserved role for GATA4 in progenitor cell expansion during development. In summary, the essential role of GATA4 for proper embryonic heart development is now well-established. However, the exact mechanism(s) that underlie defective heart morphogenesis in the absence of Gata4 remain to be clarified. They will likely involve stage-dependent alterations in cell survival, proliferation migration and differentiation.

IV.B.  GATA5 The highly-specific expression of GATA5 in endocardial and pericardial cells of all species examined (birds, fish, mammals) suggests a specialized role for GATA5 in endocardial development. Two lines of evidence are consistent with this. First, the finding that the faust locus in zebrafish is encoded by Gata5 (Reiter et al., 1999); faust mutants lack endocardial cells and have cardia bifida, hypoplastic ventricles and reduced expression of several cardiac markers including Nkx2-5. Second, blocking expression of Gata5 in an in vitro model of endocardial differentiation prevents expression of terminal differentiation markers and formation of morphologically and genetically distinguishable endocardial cells (Nemer and Nemer, 2002). However, inactivation of Gata5 in mice did not produce any detectable cardiac phenotype, but revealed a role for GATA5 in urogenital development (Molkentin et al., 2000). GATA4, which is expressed throughout endocardial development, may compensate for lack of GATA5 in these cells. Alternatively, the targeting strategy which results in deletion of exon 2 encoding part of the N-terminal transactivation domains may have produced a deleted rather than a null allele, since production of the short Gata5 allele (discussed in Section III) would be unaffected. Production of a bona fide null-Gata5 allele in transgenic mice may help clarify the role of GATA5 in mammalian heart development.

IV.C.  GATA6 Several lines of evidence suggest that GATA6 participates in cardiovascular morphogenesis, but its exact role in heart development is not fully elucidated. Gata6/ mice have gastrulation defects due to dysregulated extraembryonic endoderm differentiation and arrest development at E6.5–7 prior to heart tube formation (Morrisey et al., 1998; Koutsourakis et al., 1999). Embryos generated from Gata6/ embryonic stem cells by tetraploid embryo complementation (in which extraembryonic endoderm defects are rescued), develop until E10.5, and show no detectable alteration in heart development or in expression

of cardiac genes, leading to the conclusion that, unlike GATA4, GATA6 is dispensable for heart formation (Zhao et al., 2005). However, mice lacking one copy of each Gata4 and Gata6 (double heterozygotes) are embryoniclethal, suggesting that the two GATA factors can partially compensate for each other, and that a threshold of GATA factors is required for normal heart development (Xin et al., 2006). Alternatively, this may reflect combinatorial action of the two GATA factors to regulate specific downstream targets (Charron et al., 1999). Consistent with this notion, mice lacking both factors do not develop recognizable heart tissue (Zhou et al., 2009). These results implicate GATA6, as well as GATA4, in early cardiac cell specification and/or survival. Morpholino-mediated knockdown of Gata6 in Xenopus and zebrafish embryos result in heartless embryos in which gut tissue occupies the heart region; maintenance of cardiomyocyte differentiation rather than specification was suggested as the mechanisms of GATA6 action in the heart, which involves regulation of BMP4 and Nkx2-5 (Peterkin et al., 2003). Conditional inactivation of Gata6 in specific cardiovascular cells, namely vascular smooth muscle cells (VSMC) and neural crest, was achieved using SM22Cre and Wnt Cre transgenic mice, respectively (Lepore et al., 2006). Of note, the SM22 promoter is active in neural-crest-derived smooth muscle cells, mesoderm-derived smooth muscle cells and in cardiac myocytes. Conditional deletion of Gata6 with SM22Cre resulted in perinatal lethality, largely due to aortic arch or cardiac outflow tract abnormality as well as membranous ventricular septal defect. These defects were recapitulated in mice in which Gata6 was deleted specifically in neural crest cells using Wnt1Cremediated recombination. The results support a cell-autonomous function of GATA6 in neural-crest-derived smooth muscle cells where it regulates aortic arch patterning and cardiac outflow tract septation; in part through activation of semaporin 3C (a guidance molecule that controls migration of neural crest cells to the cardiac outflow tract). Thus, deletion of Gata6 from smooth muscle cells recapitulates several forms of congenital heart malformations, including persistent truncus arteriosus and double-outlet right ventricle. They are consistent with in vitro studies showing that GATA6 promotes differentiation of smooth muscle cells, in part due to upregulation of the cyclindependent p21 inhibitor (Perlman et al., 1998; Mano et al., 1999). The fact that these defects were not detected in mice derived from Gata6/ embryonic stem cells which die prior to midgestation is in agreement with an essential role for GATA6 in later developmental events, such as cardiac neural crest cell migration and terminal smooth muscle cell differentiation. Together, the available evidence supports an essential role for GATA6 in heart morphogenesis, and identifies GATA6 as a potential congenital heart disease-causing gene.

607

Chapter  |  9.2  GATA4 in Heart Development and Disease

V.  Role of GATA factors in postnatal heart development Postnatal cardiac development is characterized by trophic growth rather that proliferative cardiomyocyte growth. During this phase of physiologic hypertrophy, increased size of existing myocytes is accompanied by changes in gene expression that involve downregulation of several embryonic genes, the hallmark of which is Nppa (the gene coding for ANP), and upregulation of “adult” genes including those for specific contractile protein isoforms and membrane channels. In response to volume or pressure overload, postnatal cardiomyocytes undergo further compensatory adaptive growth which is also accompanied by changes in gene expression of some embryonic genes, such as Nppa. Thus, adaptive or pathological hypertrophy can be distinguished, among others, by their specific genetic programs. A regulatory role for GATA factors in the postnatal heart was first suggested by the identification of GATAbinding motifs as critical regulatory elements of cardiac natriuretic peptide genes (Grépin et al., 1994); this finding led to the isolation from postnatal cardiomyocytes of rat GATA4 cDNA, and the demonstration that GATA4 is abundantly expressed throughout postnatal heart development and is a potent transcriptional activator of cardiac genes. The list of GATA4 target genes keeps expanding (Table 4). It presently includes numerous embryonic and adult genes that encode membrane receptors and channels, contractile proteins, hormones and growth factors, as well as adhesion molecules and transcription factors.

(A)

−700 bp

−600

A/T-rich

−500

−400

−300

TRE TBE

SRE CARE GATA

−200

NPPA fold activation

140 120 100 80 60 40

0

x5

GATA4

x2

Nkx2.5

x10

Tbx5

V.A.  GATA4 and Cardiomyocyte Hypertrophy A role for GATA4 in transcriptional regulation during adaptive/stress responses was first suggested from analysis of regulatory elements required for in vitro promoter responsiveness to pressure overload. These studies revealed that GATA elements within the -myosin heavy chain (MHC) and angiotensin type 1 receptor (Agtr1a) genes are required for upregulation of the respective promoters following aortic constriction (Hasegawa et al., 1997; Herzig et al., 1997). GATA elements were subsequently shown to be required for the transcriptional response to mechanical stretch, 1-adrenergic stimulation, endothelin-1 and angiotensin II (Marttila et al., 2001; Morin et al., 2001; Pikkarainen et al., 2003; Wang et al., 2005). Moreover, many of these stimuli were shown to induce GATA-binding activity in the heart or in cultured cardiomyocytes (Herzig et al., 1997; Hautala et al., 2001; Wang et al., 2005). Consistent with this, knockdown of Gata4 using an antisense strategy (Charron et al., 2001) or

−100

TBE NKE GATA SRE

(B)

20

GATA4 and 6 are expressed at high levels in postnatal cardiomyocytes, and knockdown of either factor in these cells leads to profound changes in endogenous gene transcription (Charron et al., 1999). Together, these findings strongly argue for an essential role of GATA4/6 in postnatal heart development. As described below, studies so far provide definitive support for a critical function of GATA4 in physiological postnatal cardiac growth and in adaptive stress response.

GATA4 Nkx2.5 Tbx5

NKE TBE

PERE

Figure 8  GATA4 transcriptional role. (A) The 700 bp rat promoter for the Nppa gene showing two GATA-binding sites. (B) The transcriptional synergy by GATA4/ Nkx2-5/Tbx5 on the ANF promoter.

608

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Table 4  Known GATA4 Targets in the Heart Gene ontology

References

Contractile proteins MHC MHC Skactin cTnI cTnC

Molkentin et al., 1994; Charron et al., 1999 Charron et al., 1999 Morin et al., 2000; Belaguli et al., 2000 Di Lisi et al., 1998; Charron et al., 1999 Ip et al., 1994

Hormones and growth factors ANP BNP ET-1 BMP-4

Grépin et al., 1994 Grépin et al., 1994 Nemer et al., 1999 Nemer et al., 2003

Membrane proteins PDGFR PDGFR VCAM 1a adrenergic receptor N-cadherin NCX1 A1 adenosine receptor A3 adenosine receptor M2 muscarinicacetylcholine Angiotensin receptor type 1a Kv4.2k

Wang et al., 1996 Charron et al., 1999 Umetani et al., 2001 Michelotti et al., 2003 Zhang et al., 2003 Cheng et al., 1999 Rivkees et al., 1999 Yaar et al., 2002 Rosoff et al., 1998 Herzig et al., 1997 Jiya et al., 2003

Signaling molecules and transcription factors MEF2 Nkx2.5 Hand2 CARP

Grépin et al., 1997 Grépin et al., 1997 McFadden et al., 2001 Kuo et al., 1999

Mitochondria/Metabolism Carnitinepalmitoyltransferase1 Grp78 BclX Bcl2

Moore et al., 2001 Mao et al., 2006 Aries et al., 2004 Kobayashi et al., 2006

overexpression of a dominant-negative GATA4-engrailed repressor fusion protein, blocked cytoskeletal reorganization, increased cell size and protein synthesis in response to the -adrenergic-agonist phenylephrine (PE) and to endothelin-1 (Charron et al., 2001; Liang et al., 2001). These studies establish the essential role of GATA4 for response to stimuli required for normal functioning of the adult heart and whose signaling is further enhanced during pressure overload. Gain-of-function studies confirmed that GATA4 is necessary and sufficient for cardiac hypertrophy. In cultured postnatal cardiomyocytes, overexpression of GATA4, as well as

GATA6, is sufficient to induce cytoskeletal reorganization and myocyte hypertrophy (Charron et al., 2001; Liang et al., 2001). However, GATA4 overexpression is not accompanied by enhanced Nppa transcription, raising the possibility that GATA4 activation may be associated with physiologic hypertrophy (Fig. 6). In vivo, attempts to test this hypothesis through production of transgenic mice overexpressing GATA4 in the heart under the control of the MHC promoter – which directs expression in the embryonic atria and postnatal ventricles – were largely unsuccessful, resulting in embryonic lethality (Wang and Nemer, unpublished results). Liang et al. were able to produce a single transgenic line in which GATA4 was overexpressed by 2- to 2.5-fold only, using a similar approach but in a different mouse strain; lines expressing higher transgene copy numbers could not be generated or maintained, revealing that embryonic heart development is exquisitely sensitive to GATA4 levels. Nppa and Nppb levels were increased in the adult transgenic line, with 2–5-fold higher GATA4 levels but, until six months of age, no changes in heart function were evident. However, by eight months, transgenic hearts showed dilatation and decreased ejection fraction, for reasons that are not fullyunderstood (Liang et al., 2001). Thus, the importance of GATA4 in postnatal heart growth is now well-established, but more studies are required to determine whether GATA4 mediates physiological or pathological hypertrophy.

V.B.  GATA6 and Vascular Remodeling Unlike cardiomyocytes, vascular smooth muscle cells do not terminally-differentiate, and can reversibly modulate their phenotype in response to growth factors. Differentiated vascular smooth muscle cells are quiescent and express contractile proteins, whereas proliferating vascular smooth muscle cells have a “synthetic” phenotype more akin to fibroblasts with lower levels of contractile proteins. This plasticity is required for wound healing, but is also involved in the pathogenesis of atherosclerosis and arterosis. GATA6 is expressed in quiescent contractile vascular smooth muscle cells and its expression is downregulated in vascular smooth muscle cell cultures in response to mitogen stimulation (Suzuki et al., 1996). GATA6 gain-of-function in proliferating embryonic fibroblasts induces cell-cycle arrest in a p21 (Cip1)-dependent manner (Perlman et al., 1998). Moreover, GATA6 levels are downregulated in rat carotid arteries following balloon injury, and adenovirus-mediated transfer of GATA6 to the vessel wall inhibits lesion formation by nearly 50% (Mano et al., 1999). This, together with the finding that GATA6 activates several contractile protein promoters (Wada et al., 2000), and numerous serum response factor-dependent smooth muscle cell genes (Chang et al., 2003) support a role for GATA6 in promoting and maintaining the differentiated contractile phenotype of vascular smooth muscle

609

Chapter  |  9.2  GATA4 in Heart Development and Disease

cells. They also suggest that dysregulated GATA6 activity may predispose or lead to vascular disease. In conclusion, GATA4 and 6 are essential for embryonic heart formation, and play critical roles in postnatal heart development and maintenance of cardiovascular function. Thus, GATA4 and 6 are part of the regulatory pathways that are shared between the embryonic and the adult heart.

VI.  Role of GATA4 in cardiomyocyte survival Regulation of postnatal cardiomyocyte survival is essential for proper cardiac function. Lost myocytes are replaced by fibroblasts, causing loss of contractility and progressive cardiac dysfunction. In addition to its role in myocyte growth and differentiation, GATA functions as a critical regulator of cardiomyocyte survival. The link between GATA4 and cell survival first came to light in studies aimed at understanding the GATAdependent mechanisms underlying P19 cell differentiation into beating cardiomyocytes. Knockdown of GATA4 expression in these cells causes massive apoptosis of cardioblasts, and blocks expression of downstream targets including contractile protein genes and other differentiation markers (Grépin et al., 1997). Cultured postnatal myocytes lacking GATA4 and hearts of Gata4/ mice have increased levels of apoptotic cells, suggesting that GATA4 is required for postnatal cardiomyocyte survival (Aries et al., 2004). This was further confirmed in transgenic mice with cardiac-specific deletion of Gata4 using MHC-Cre and MHC-CRE (Oka et al., 2006). Importantly, GATA4 appears to be the target of survival factors such as hepatocyte growth factor (Kim et al., 2003; Kitta et al., 2003), as well as proapoptotic substances such as chemotherapeutic agents of the anthracycline class, like daunorubicin and doxorubicin (Kim et al., 2003; Aries et al., 2004). Anthracycline cardiotoxicity is a serious clinical problem as it involves irreversible cardiomyopathy and myocyte loss. Both in vivo and in vitro, treatment with doxorubicin leads to rapid depletion of GATA4, due in large part to decreased transcription. This precedes detectable myocyte apoptosis (Aries et al., 2004) and adenovirus-mediated restoration of GATA4 prevents doxorubicin-induced myocyte apoptosis and cardiac dysfunction, indicating that GATA4 is a key regulator of myocyte survival (Aries et al., 2004). The mechanisms underlying the cardioprotective effects of GATA4 involve upregulation of the anti-apoptotic genes Bcl2 and BclXL (Aries et al., 2004; Kobayashi et al., 2006), as well as genes involved in biogenesis (such as creatine kinase and carnitine palmitoyl transferase) and several stress response genes (Hsp70 glutanetoxin) and Grp78.

Table 5  Known GATA4 Collaborators in the Heart GATA partners

Cell colocalization

Development stage

Fos

Myocardium

Postnatal

NFATc1

Endocardium

Embryo

NFATc4

Myocardium

Postnatal

Smad

Myocardium

Embryo

STAT

Myocardium

Postnatal

CBP

Myocardium

Postnatal

FOG1

Endocardium

Embryo

FOG2

Endocardium, myocardium

Embryo

Hand1

A  LV myocardium

Embryo

Hand2

RV myocardium, outflow tract, endocardium

Embryo

Hey2

Myocardium

Embryo

KLF13

Myocardium

Embryo  postnatal

KLF15

Myocardium

Postnatal

MEF2

Myocardium

Embryo

Nkx2.5

Myocardium

Embryo

p300

Myocardium

Postnatal

SRF

Myocardium

Embryo  postnatal

Tbx2

Myocardium

Embryo

Tbx5

A  LV myocardium

Embryo  postnatal

Tbx20

Endocardium, myocardium

Embryo  postnatal

Zfp260

Myocardium

Embryo  postnatal

VII.  Combinatorial interactions of GATA factors with other transcriptional regulators As discussed above, GATA4 and 6 are involved in several cellular processes, and are required for different developmental stages. This is achieved through combinatorial interactions with other cell-specific or inducible transcription factors. Table 5 provides a list of known GATA4 collaborators. That combinatorial interactions regulate cardiac transcription is now an accepted paradigm.

VII.A.  Cell-Specific GATA Collaborators A cornerstone in understanding transcriptional regulation of heart development was the demonstration that

610

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

GATA4 acts in collaboration with other transcription factors, and that combinatorial interaction between cardiacspecific transcription factors to establish a transcriptome or an enhancesome controls cell- and stage-specific gene expression and cardiac cell fates. The first example of such interactions was reported for GATA4 and Nkx2-5 regulation of the Nppa promoter (Durocher et al., 1997). The GATA4/Nkx2-5 interaction turned out to be evolutionarily-conserved and required for signaling of cardiogenic factors at early stages of cardiac cell differentiation (discussed below). It is based on the finding that GATA4 and Nkx2-5 are collaborators, that we predicted that the Drosophila GATA factor, pannier (see Chapter 1.2), is the likely GATA4 homolog that regulates heart development, in collaboration with the Drosophila NK protein, tinman (Durocher et al., 1997). This prediction indeed panned out. In Drosophila, pannier and tinman cooperatively regulate other transcription factors required for heart development, such as dMEF2 and Hand (Gajewski et al., 1999, 2001; Han and Olson, 2005). In the case of Nppa, binding of both factors to DNA is required for optimal functional cooperativity, although GATA-binding sites are sufficient to recruit Nkx2-5 to target promoters (see Chapter 9.1). GATA4/Nkx2-5 cooperativity involves direct protein–protein interaction that apparently unmasks a C-terminal autorepressive mechanism on Nkx2-5 (Durocher et al., 1997). The second zincfinger of GATA4 and a C-terminal extension is required for the physical interaction with Nkx2-5. Subsequent studies revealed that the second zinc-finger is involved in most,

GFR

GPCR PIP2

JAK

but not all, GATA4 interactions with its partners. GATA4/ Nkx2-5 interaction was further documented for other cardiac promoters, including the cardiac -actin gene and Ankrd1 encoding CARP (Kuo et al., 1997; Sepulveda et al., 1998; Kuo et al., 1999) and requires, in addition to the DNA-binding domain, both N- and C-terminal transactivation domains. This interaction is highly-selective and GATA6 cannot functionally substitute for GATA4 in interacting with Nkx2-5, suggesting that differential interaction may underly specificity of GATA factors (Durocher and Nemer, 1998). Moreover, in vivo promoter analysis in Xenopus embryos revealed that Nkx2-5 and GATA4 binding sites on Nppa are required for its atrial expression, suggesting a role conserved for cooperative GATA4/Nkx2-5 interaction in chamber formation (Small and Krieg, 2003). Another important class of tissue-specific and signal responsive Gata4 collaborators are the MADS-box factors MEF2 and serum response factor. MEF2 proteins are recruited to target promoters lacking high affinity MEF2 sites via GATA binding sites (Morin et al., 2000). In addition to a role in development, GATA4 interaction with MEF2 and serum response factor may also mediate calcium and hormone signaling as shown in Fig. 9 and discussed below. Other partners that are implicated with GATA4 in common pathways in heart development are members of the T-box family, mainly Tbx5 and Tbx20 (Garg et al., 2003; Stennard et al., 2003), as well as bHLH family members like Hand1/2 and Hey proteins (Dai and Markham, 2001; Fischer et al., 2005; Morin et al., 2005). Although most of

IP3

TAK

PLCβ

Gq

Ras

DG MEKK

Raf

MEKK?

SEK

MEK

RKK

JNK/SAPK

ERK1/2

P38/RK

Ca2+

STAT

CaM

SMAD

Calcineurin CaMKII

T STA pGATA

GATA SMAD

GATA

NFAT

GATA

PKC

pGATA

Mef2

pGATA

Fos jun

pGATA

SRF SRF

pGATA

Zfp 260

Figure 9  GATA4 partners: different signaling pathways regulate their cooperative physical and functional interactions.

Mef2

611

Chapter  |  9.2  GATA4 in Heart Development and Disease

the studies carried out focus on Nppa as a model for combinatorial interaction between GATA proteins and their partners, many other genes were shown to be synergistic­ ally activated by these proteins. It is of particular importance to note that the simultaneous interaction of GATA proteins in a given cardiac compartment with different partners provides a molecular explanation for understanding how similar cardiac defects are caused by mutations in different genes. Such cooperativity can be illustrated by the interaction between Tbx5/Nkx2-5/GATA4 (Fig. 5) that might unravel the molecular pathway for atrial septation, since mutations in any of the three genes leads to septal defects in experimental models and in humans. The above discussed interactions occur through the second zinc-finger of GATA proteins, and generally lead to synergistic activation of transcription; this is not the case with FOG (Friend Of GATA) proteins where in vitro analysis mapped the interaction domain to the first zincfinger (Tevosian et al., 1999). The GATA4–FOG2 interaction is important for the development of the valves and the formation of the coronary artery system, as evidenced by a knockin mutation that disrupts this interaction (Table 3) (Crispino et al., 2001). Whether FOG2 acts in vivo as a GATA4/6 co-repressor or a co-activator has not yet been settled, and the in vivo downstream target genes of GATA4/FOG2 have not been described up till now. FOG proteins contain an N-terminal repression domain (Lin et al., 2004), but in vitro the GATA4–FOG2 interaction leads to promoter context-dependent activation or repression (Lu et al., 1999; Svensson et al., 1999; Tevosian et al., 2000). The GATA–FOG interaction is conserved in Drosophila, where the FOG homolog U-shaped interacts with pannier and functions as a negative regulator of heart development (Fossett et al., 2001). Finally, GATA4, 5 and 6 interact with ubiquitouslyexpressed proteins like the kruppel zinc-finger protein YY1 (Bhalla et al., 2001; Mao et al., 2006) and the coactivator p300 (Dai et al., 2001). In these cases, physical interaction occurs through the C-terminal zinc-finger, except for the newly-described member of the kruppel-like family of transcription factor KLF13, where both zinc-fingers are required for optimal interaction (Lavallée et al., 2006).

VII.B.  Inducible GATA Co-Factors In addition to the GATA-interacting partners described above, GATA proteins interact with signal-inducible transcription factors acting as nuclear effectors of signaling pathways (Fig. 9). Examples of such interactions include the interaction of GATA4/5/6 with members of the NFAT family of transciption factors, the nuclear effectors of the calcium–calcinurin signaling pathway, as well as interaction of GATA4/5/6 with STAT proteins. In response to activation of calcium signaling, NFAT family proteins are

dephosphorylated by calcineurin and translocate to the nucleus where they can associate with GATA proteins. In cardiomyocytes, a GATA4–NFATC4 interaction was shown to mediate the hypertrophic response to several stimuli (Molkentin et al., 1998). During endocardial cell differentiation GATA5–NFATC1 was shown to be required for expression of terminal differentiation genes, like endothelin-1 (Nemer et al., 2002). NFAT interaction extends to GATA6, and may be a co-regulator of GATA6 in vascular smooth muscle cells (Wada et al., 2002). In the case of STAT proteins, they are activated following growth factor stimulation acting through JAK- or SRC-mediated phosphorylation, resulting in nuclear translocation. GATA4– STAT was shown to transduce angiotensin II-activation of the Nppa promoter (Wang et al., 2005). Similarly, SMAD proteins, the downstream effectors of the TGF family of growth–differentiation factors, associate with GATA4 to transduce, among others, BMP4 cardiogenic signals and transcription of Nkx2-5 (Brown et al., 2004). In these and other cases, such as with cFos (McBride et al., 2003), interaction of signal-inducible factors with a cell-specific GATA protein provides a mechanism for cell-specific responses to extracellar stimuli.

VIII.  GATA factors as integrators and regulators of cell signaling in the heart It is evident from the above sections that GATA4, presently the most extensively-studied cardiac GATA factor, is both upstream and downstream of several important signaling cascades in the heart. For example, during early heart development, GATA4 is an upstream regulator of BMP4 (Nemer et al., 2003), and an effector of BMP signaling (discussed above). In response to BMP4, GATA4–SMAD cooperatively activate Nkx2-5, and Nkx2-5 in turn acts as a collaborator of GATA4 in activating further downstream targets. This reinforcing loop may explain how GATA4 is sufficient to induce cardiogenesis in Xenopus ectoderm (Latinkic et al., 2003) and how, in Drosophila, pannier acts as a competence factor for heart progenitor formation (Klinedinst and Bodmer, 2003). On the other hand, GATA4 regulates the angiotensin receptor, while acting as an essential nuclear effector of angiotensin signaling through its multiple kinase inducible partners (Fig. 9). GATA4 also activates transcription of endothelin-1, and acts downstream of endothelin signaling. It is noteworthy that both endothelin-1 and angiotensin II are the hormonal mediators of mechanic­al stretch in cardiomyocytes. It is therefore important to note that GATA4 is essential for the adaptive response of the heart to mechanical stretch (Pikkarainen et al., 2003), and that mice lacking one Gata4 allele have impaired response to pressure overload (Bisping et al., 2006; Oka et al., 2006). These data raise the intriguing possibility that

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

3S P1 6

1

A3 4 V3 6P 8 L 4 0M 03 M

R266X V267M N273S T277I R283H N285K C292R A294V G296S H302R

F208 L F211L G214S E216D M223T R229S G234S N239D/S Y244C N248S R252P I255T R260Q L261P

(A)

F

subtle variations in GATA4 activity may represent risk factors for adult cardiac disease. Finally, the finding that upregulation of GATA4 prevents chemotherapy-induced cardiomyopathy suggests that pharmacological manipulation of GATA4 levels or activity might have therapeutic potential for cardioprotection.

S5 2

612

442

IX.  GATA4 and congenital heart disease Many studies previously showed that patients with deletion of the distal arm of human chromosome 8p often have congenital heart disease and other physical anomalies. The human GATA4 gene maps to 8p22–23 and its deletion in heterozygote individuals was shown to be associated with septal and valvular defects (Pehlivan et al., 1999). Lately, different groups have found that point mutations in GATA4 are linked to different forms of CHDs (Fig. 10). Familial cases of ASDs were shown to harbor the G296S mutation that decreases GATA4 DNA-binding activity and abrogates interaction with Tbx5, or a frame shift mutation (E359del) that severely impairs transcriptional activity (Garg et al., 2003). Further studies revealed other missense mutations, including the E216D present in sporadic cases of tetralogy of Fallot that result in a dramatic decrease in GATA4 transcriptional activity (Nemer et al., 2006). In both cases haploinsufficiency is suggested as the underlying mechanism of pathogenesis. This is consistent with analysis of genetically-engineered mice showing that GATA4 is a dosage-sensitive regulator of heart development, and that decreased GATA4 levels cause varying cardiac abnormalities that mimic human congenital heart disease (Pu et al., 2004). So far, only six missense mutations in the GATA4 gene were described in different populations with variable phenotypes including septal and valvular defects (Fig. 10). The mechanisms by which these mutations cause disease are not yet defined. The relatively lower frequency of mutations reported for GATA4 as compared to Nkx2-5 might reflect the broader expression of GATA4, including in placenta, and its essential role in early developmental events. Mutations in a gene encoding a protein broadly expressed in early embryonic development like GATA4 might thus lead more often to aborted fetuses rather than newborns with specific congenital diseases. Interestingly, a retrospective study of tissue samples recently reported numerous missense and nonsense mutations in the exons encoding the first and second zinc-finger domains of GATA4 that are predicted to severely impair its DNAbinding and protein interaction properties, and might well represent disease-causing mutations (Reamon-Buettner et al., 2005). Finally, as discussed above, GATA4 activity is also regulated by its interaction with other collaborators such as Nkx2-5, and mutations in Nkx2-5 that decrease its

Zn Zn ++

TAD

P226fs

TAD

S358RfsX45 E358RfsX44

(B) PTA ASD

DORV

TOF

PS

AVSD

VSD

Figure 10  Gata4 in congenital heart disease. (A) Different mutations that were found so far in patients with congenital heart disease. In blue, mutations that were found while screening genomic DNA from blood samples, whereas mutations in red are from cardiac tissue samples. (B) Different phenotypes were associated to mutations in the Gata4 gene: TOF (tetralogy of Fallot); VSD (ventricular septal defect); ASD (atrial septal defect); DORV (double-outlet right ventricle); PTA (persistent truncus arteriosius); PS (pulmonary stenosis); AVSD (atrioventricular ­septal defect).

transcriptional activity and/or interaction with GATA4 may represent “functional” GATA4 mutants. Thus, decreased GATA4 activity in individuals who harbor a subset of NKX2-5 mutations, and/or possibly mutations in other transcription factors that are known GATA4 collaborators may underlie the pathogenesis of congenital heart disease. Clearly, further genetic and biochemical studies are required, including knockin analysis in transgenic mice, to fully elucidate phenotype–genotype relationships.

X.  Conclusion and perspectives The role of GATA4 in the heart was first recognized in 1994. Since then, spectacular progress has been achieved in delineating the various functions of GATA4 in heart development and disease. Nevertheless, significant gaps remain. For example, the targets of GATA4 in cell proliferation, survival and differentiation remain incompletely understood, and the mechanisms regulating GATA4 interaction with its many collaborators and the relationship of

Chapter  |  9.2  GATA4 in Heart Development and Disease

these interactions to specific cellular functions need to be elucidated. The potential protective role of GATA4 in the adult heart needs to be analyzed, and the possibility that dysregulated GATA4 levels or activity may predispose to human disease should be further examined. The role of GATA6 in cardiomyocytes remains undefined, and more studies are needed to understand the specific role of GATA6 in cardiomyocytes and more generally, the mechanisms underlying specificity of GATA factors. Indeed, despite studies showing similar in vitro activities of GATA4 and GATA6, GATA6, which is co-expressed with GATA4 in cardiomyocytes, is unable to compensate for GATA4 absence in vivo, as evidenced by defective heart development in animals and humans with hypomorphic GATA4 alleles. Finally, the role of GATA5 in the mammalian heart still needs to be elucidated. Over, the next years, the combined power and complementarity of genetic and biochemical approaches will undoubtedly yield exciting new insights into the mechanisms of GATA regulation of heart formation and function.

References Arceci, R.J., King, A.A., Simon, M.C., Orkin, S.H., Wilson, D.B., 1993. Mouse GATA-4: a retinoic acid-inducible GATA-binding transcription factor expressed in endodermally derived tissues and heart. Mol. Cell Biol. 13, 2235–2246. Aries, A., Paradis, P., Lefebvre, C., Schwartz, R.J., Nemer, M., 2004. Essential role of GATA-4 in cell survival and drug-induced cardiotoxicity. Proc. Natl. Acad. Sci. USA 101, 6975–6980. Bhalla, S.S., Robitaille, L., Nemer, M., 2001. Cooperative activation by GATA-4 and YY1 of the cardiac B-type natriuretic peptide promoter. J. Biol. Chem. 276, 11439–11445. Bisping, E., Ikeda, S., Kong, S.W., Tarnavski, O., Bodyak, N., McMullen, J.R., Rajagopal, S., Son, J.K., Ma, Q., Springer, Z., Kang, P.M., Izumo, S., Pu, W.T., 2006. Gata4 is required for maintenance of postnatal cardiac function and protection from pressure overload-induced heart failure. Proc. Natl. Acad. Sci. USA 103, 14471–14476. Brewer, A., Gove, C., Davies, A., McNulty, C., Barrow, D., Koutsourakis, M., Farzaneh, F., Pizzey, J., Bomford, A., Patient, R., 1999. The human and mouse GATA-6 genes utilize two promoters and two initiation codons. J. Biol. Chem. 274, 38004–38016. Brewer, A., Nemer, G., Gove, C., Rawlins, F., Nemer, M., Patient, R., Pizzey, J., 2002. Widespread expression of an extended peptide sequence of GATA-6 during murine embryogenesis and non-equivalence of RNA and protein expression domains. Mech. Dev. 119 (Suppl. 1), S121–S129. Broitman-Maduro, G., Maduro, M.F., Rothman, J.H., 2005. The noncanonical binding site of the MED-1 GATA factor defines differentially regulated target genes in the C. elegans mesendoderm. Dev. Cell 8, 427–433. Brown, C.O., Chi, X., Garcia-Gras, E., Shirai, M., Feng, X.H., Schwartz, R.J., 2004. The cardiac determination factor, Nkx2-5, is activated by mutual cofactors GATA-4 and Smad1/4 via a novel upstream enhancer. J. Biol. Chem. 279, 10659–10669.

613

Chang, D.F., Belaguli, N.S., Iyer, D., Roberts, W.B., Wu, S.P., Dong, X.R., Marx, J.G., Moore, M.S., Beckerle, M.C., Majesky, M.W., Schwartz, R.J., 2003. Cysteine-rich LIM-only proteins CRP1 and CRP2 are potent smooth muscle differentiation cofactors. Dev. Cell 4, 107–118. Charron, F., Paradis, P., Bronchain, O., Nemer, G., Nemer, M., 1999. Cooperative interaction between GATA-4 and GATA-6 regulates myocardial gene expression. Mol. Cell Biol. 19, 4355–4365. Charron, F., Tsimiklis, G., Arcand, M., Robitaille, L., Liang, Q., Molkentin, J.D., Meloche, S., Nemer, M., 2001. Tissue-specific GATA factors are transcriptional effectors of the small GTPase RhoA. Genes Dev. 15, 2702–2719. Crispino, J.D., Lodish, M.B., Thurberg, B.L., Litovsky, S.H., Collins, T., Molkentin, J.D., Orkin, S.H., 2001. Proper coronary vascular development and heart morphogenesis depend on interaction of GATA-4 with FOG cofactors. Genes Dev. 15, 839–844. Dai, Y.S., Markham, B.E., 2001. p300 Functions as a coactivator of transcription factor gata-4. J. Biol. Chem. 276, 37178–37185. Dodou, E., Verzi, M.P., Anderson, J.P., Xu, S.M., Black, B.L., 2004. Mef2c is a direct transcriptional target of ISL1 and GATA factors in the anterior heart field during mouse embryonic development. Development 131, 3931–3942. Drevet, J.R., Swevers, L., Iatrou, K., 1995. Developmental regulation of a silkworm gene encoding multiple GATA-type transcription factors by alternative splicing. J. Mol. Biol. 246, 43–53. Durocher, D., Charron, F., Warren, R., Schwartz, R.J., Nemer, M., 1997. The cardiac transcription factors Nkx2-5 and GATA-4 are mutual cofactors. EMBO J. 16, 5687–5696. Durocher, D., Nemer, M., 1998. Combinatorial interactions regulating cardiac transcription. Dev. Genet. 22, 250–262. Fischer, A., Klattig, J., Kneitz, B., Diez, H., Maier, M., Holtmann, B., Englert, C., Gessler, M., 2005. Hey basic helix-loop-helix transcription factors are repressors of GATA4 and GATA6 and restrict expression of the GATA target gene ANF in fetal hearts. Mol. Cell Biol. 25, 8960–8970. Fossett, N., Tevosian, S.G., Gajewski, K., Zhang, Q., Orkin, S.H., Schulz, R.A., 2001. The Friend of GATA proteins U-shaped. FOG-1, and FOG-2 function as negative regulators of blood, heart, and eye development in Drosophila. Proc. Natl. Acad. Sci. USA 98, 7342–7347. Fujiwara, Y., Browne, C.P., Cunniff, K., Goff, S.C., Orkin, S.H., 1996. Arrested development of embryonic red cell precursors in mouse embryos lacking transcription factor GATA-1. Proc. Natl. Acad. Sci. USA 93, 12355–12358. Gajewski, K., Fossett, N., Molkentin, J.D., Schulz, R.A., 1999. The zinc finger proteins Pannier and GATA4 function as cardiogenic factors in Drosophila. Development 126, 5679–5688. Gajewski, K., Zhang, Q., Choi, C.Y., Fossett, N., Dang, A., Kim, Y.H., Kim, Y., Schulz, R.A., 2001. Pannier is a transcriptional target and partner of Tinman during Drosophila cardiogenesis. Dev. Biol. 233, 425–436. Garg, V., Kathiriya, I.S., Barnes, R., Schluterman, M.K., King, I.N., Butler, C.A., Rothrock, C.R., Eapen, R.S., Hirayama-Yamada, K., Joo, K., Matsuoka, R., Cohen, J.C., Srivastava, D., 2003. GATA4 mutations cause human congenital heart defects and reveal an interaction with TBX5. Nature 424, 443–447. Grépin, C., Dagnino, L., Robitaille, L., Haberstroh, L., Antakly, T., Nemer, M., 1994. A hormone-encoding gene identifies a pathway for cardiac but not skeletal muscle gene transcription. Mol. Cell Biol. 14, 3115–3129.

614

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Grépin, C., Robitaille, L., Antakly, T., Nemer, M., 1995. Inhibition of transcription factor GATA-4 expression blocks in vitro cardiac muscle differentiation. Mol. Cell Biol. 15, 4095–4102. Grépin, C., Nemer, G., Nemer, M., 1997. Enhanced cardiogenesis in embryonic stem cells overexpressing the GATA-4 transcription factor. Development 124, 2387–2395. Han, Z., Olson, E.N., 2005. Hand is a direct target of Tinman and GATA factors during Drosophila cardiogenesis and hematopoiesis. Development 132, 3525–3536. Harigae, H., Takahashi, S., Suwabe, N., Ohtsu, H., Gu, L., Yang, Z., Tsai, F.Y., Kitamura, Y., Engel, J.D., Yamamoto, M., 1998. Differential roles of GATA-1 and GATA-2 in growth and differentiation of mast cells. Genes Cells 3, 39–50. Hasegawa, K., Lee, S.J., Jobe, S.M., Markham, B.E., Kitsis, R.N., 1997. Cis-acting sequences that mediate induction of beta-myosin heavy chain gene expression during left ventricular hypertrophy due to aortic constriction. Circulation 96, 3943–3953. Hautala, N., Tokola, H., Luodonpaa, M., Puhakka, J., Romppanen, H., Vuolteenaho, O., Ruskoaho, H., 2001. Pressure overload increases GATA4 binding activity via endothelin-1. Circulation 103, 730–735. Herzig, T.C., Jobe, S.M., Aoki, H., Molkentin, J.D., Cowley Jr., A.W., Izumo, S., Markham, B.E., 1997. Angiotensin II type1a receptor gene expression in the heart: AP-1 and GATA-4 participate in the response to pressure overload. Proc. Natl. Acad. Sci. USA 94, 7543–7548. Jiang, Y., Evans, T., 1996. The Xenopus GATA-4/5/6 genes are associated with cardiac specification and can regulate cardiac-specific transcription during embryogenesis. Dev. Biol. 174, 258–270. Jiang, Y., Drysdale, T.A., Evans, T., 1999. A role for GATA-4/5/6 in the regulation of Nkx2.5 expression with implications for patterning of the precardiac field. Dev. Biol. 216, 57–71. Kawamura, T., Ono, K., Morimoto, T., Wada, H., Hirai, M., Hidaka, K., Morisaki, T., Heike, T., Nakahata, T., Kita, T., Hasegawa, K., 2005. Acetylation of GATA-4 is involved in the differentiation of embryonic stem cells into cardiac myocytes. J. Biol. Chem. 280, 19682–19688. Keijzer, R., van Tuyl, M., Meijers, C., Post, M., Tibboel, D., Grosveld, F., Koutsourakis, M., 2001. The transcription factor GATA6 is essential for branching morphogenesis and epithelial cell differentiation during fetal pulmonary development. Development 128, 503–511. Kim, Y., Ma, A.G., Kitta, K., Fitch, S.N., Ikeda, T., Ihara, Y., Simon, A.R., Evans, T., Suzuki, Y.J., 2003. Anthracycline-induced suppression of GATA-4 transcription factor: implication in the regulation of cardiac myocyte apoptosis. Mol. Pharmacol. 63, 368–377. Kitta, K., Day, R.M., Kim, Y., Torregroza, I., Evans, T., Suzuki, Y.J., 2003. Hepatocyte growth factor induces GATA-4 phosphorylation and cell survival in cardiac muscle cells. J. Biol. Chem. 278, 4705–4712. Klinedinst, S.L., Bodmer, R., 2003. Gata factor Pannier is required to establish competence for heart progenitor formation. Development 130, 3027–3038. Kobayashi, S., Lackey, T., Huang, Y., Bisping, E., Pu, W.T., Boxer, L.M., Liang, Q., 2006. Transcription factor gata4 regulates cardiac BCL2 gene expression in vitro and in vivo. FASEB J. 20, 800–802. Kostetskii, I., Jiang, Y., Kostetskaia, E., Yuan, S., Evans, T., Zile, M., 1999. Retinoid signaling required for normal heart development regulates GATA4 in a pathway distinct from cardiomyocyte differentiation. Dev. Biol. 206, 206–218. Koutsourakis, M., Langeveld, A., Patient, R., Beddington, R., Grosveld, F., 1999. The transcription factor GATA6 is essential for early extraembryonic development. Development 126, 723–732.

Kuo, C.T., Morrisey, E.E., Anandappa, R., Sigrist, K., Lu, M.M., Parmacek, M.S., Soudais, C., Leiden, J.M., 1997. GATA4 transcription factor is required for ventral morphogenesis and heart tube formation. Genes Dev. 11, 1048–1060. Kuo, H.C., Chen, J., Ruiz-Lozano, P., Zou, H., Nemer, M., Chien, K.R., 1999. Control of segmental expression of the cardiac-restricted ankyrin repeat protein gene by distinct regulatory pathways in murine cardiogenesis. Development 126, 4223–4234. Kuruvilla, F.G., Shamji, A.F., Schreiber, S.L., 2001. Carbon- and nitrogen-quality signaling to translation are mediated by distinct GATA-type transcription factors. Proc. Natl. Acad. Sci. USA 98, 7283–7288. Lakshmanan, G., Lieuw, K.H., Lim, K.C., Gu, Y., Grosveld, F., Engel, J.D., Karis, A., 1999. Localization of distant urogenital system-, central nervous system-, and endocardium-specific transcriptional regulatory elements in the GATA-3 locus. Mol. Cell Biol. 19, 1558–1568. Latinkic, B.V., Kotecha, S., Mohun, T.J., 2003. Induction of cardiomyocytes by GATA4 in Xenopus ectodermal explants. Development 130, 3865–3876. Lavallée, G., Andelfinger, G., Nadeau, M., Lefebvre, C., Nemer, G., Horb, M., Nemer, M., 2006. The Kruppel-like transcription factor KLF13 is a GATA-4 collaborator for heart development. EMBO J. 25, 5201–5213. Lepore, J.J., Mericko, P.A., Cheng, L., Lu, M.M., Morrisey, E.E., Parmacek, M.S., 2006. GATA-6 regulates semaphorin 3C and is required in cardiac neural crest for cardiovascular morphogenesis. J. Clin. Invest. 116, 929–939. Liang, Q., De Windt, L.J., Witt, S.A., Kimball, T.R., Markham, B.E., Molkentin, J.D., 2001. The transcription factors GATA4 and GATA6 regulate cardiomyocyte hypertrophy in vitro and in vivo. J. Biol. Chem. 276, 30245–30253. Lin, A.C., Roche, A.E., Wilk, J., Svensson, E.C., 2004. The N termini of Friend of GATA (FOG) proteins define a novel transcriptional repression motif and a superfamily of transcriptional repressors. J. Biol. Chem. 279, 55017–55023. Lowry, J.A., Atchley, W.R., 2000. Molecular evolution of the GATA f­amily of transcription factors: conservation within the DNA-binding domain. J. Mol. Evol. 50, 103–115. Lu, J.R., McKinsey, T.A., Xu, H.T., Wang, D.Z., Richardson, J.A., Olson, E.N., 1999. FOG-2, a heart- and brain-enriched cofactor for GATA transcription factors. Mol. Cell Biol. 19, 4495–4502. MacNeill, C., Ayres, B., Laverriere, A.C., Burch, J.B., 1997. Transcripts for functionally distinct isoforms of chicken GATA-5 are differentially expressed from alternative first exons. J. Biol. Chem. 272 (13), 8396–8401. Maduro, M.F., Hill, R.J., Heid, P.J., Newman-Smith, E.D., Zhu, J., Priess, J.R., Rothman, J.H., 2005. Genetic redundancy in endoderm specification within the genus Caenorhabditis. Dev. Biol. 284, 509–522. Maeda, M., Ohashi, K., Ohashi-Kobayashi, A., 2005. Further extension of mammalian GATA-6. Dev. Growth Differ. 47, 591–600. Manfield, I.W., Devlin, P.F., Jen, C.H., Westhead, D.R., Gilmartin, P.M., 2007. Conservation, convergence, and divergence of light-responsive, circadian-regulated, and tissue-specific expression patterns during evolution of the Arabidopsis GATA gene family. Plant Physiol. 143, 941–958. Mano, T., Luo, Z., Malendowicz, S.L., Evans, T., Walsh, K., 1999. Reversal of GATA-6 downregulation promotes smooth muscle differentiation and inhibits intimal hyperplasia in balloon-injured rat carotid artery. Circ. Res. 84, 647–654.

Chapter  |  9.2  GATA4 in Heart Development and Disease

Mao, C., Tai, W.C., Bai, Y., Poizat, C., Lee, A.S., 2006. In vivo regulation of Grp78/BiP transcription in the embryonic heart: role of the endoplasmic reticulum stress response element and GATA-4. J. Biol. Chem. 281, 8877–8887. Marttila, M., Hautala, N., Paradis, P., Toth, M., Vuolteenaho, O., Nemer, M., Ruskoaho, H., 2001. GATA4 mediates transcriptional activation of the B-type natriuretic peptide gene expression in response to hemodynamic stress. Endocrinology 142, 4693–4700. McBride, K., Charron, F., Lefebvre, C., Nemer, M., 2003. Interaction with GATA transcription factors provides a mechanism for cell­specific effects of c-Fos. Oncogene 22, 8403–8412. McFadden, D.G., Charité, J., Richardson, J.A., Srivastava, D., Firulli, A.B., Olson, E.N., 2000. A GATA-dependent right ventricular enhancer controls dHAND transcription in the developing heart. Development 127, 5331–5341. Molkentin, J.D., Lin, Q., Duncan, S.A., Olson, E.N., 1997. Requirement of the transcription factor GATA4 for heart tube formation and ventral morphogenesis. Genes Dev. 11, 1061–1072. Molkentin, J.D., Lu, J.R., Antos, C.L., Markham, B., Richardson, J., Robbins, J., Grant, S.R., Olson, E.N., 1998. A calcineurin-dependent transcriptional pathway for cardiac hypertrophy. Cell 93, 215–228. Molkentin, J.D., Tymitz, K.M., Richardson, J.A., Olson, E.N., 2000. Abnormalities of the genitourinary tract in female mice lacking GATA5. Mol. Cell Biol. 20, 5256–5260. Monzen, K., Shiojima, I., Hiroi, Y., Kudoh, S., Oka, T., Takimoto, E., Hayashi, D., Hosoda, T., Habara-Ohkubo, A., Nakaoka, T., Fujita, T., Yazaki, Y., Komuro, I., 1999. Bone morphogenetic proteins induce cardiomyocyte differentiation through the mitogen-activated protein kinase TAK1 and cardiac transcription factors Csx/Nkx-2.5 and GATA-4. Mol. Cell Biol. 19, 7096–7105. Morin, S., Charron, F., Robitaille, L., Nemer, M., 2000. GATA-­dependent recruitment of MEF2 proteins to target promoters. EMBO J. 19, 2046–2055. Morin, S., Paradis, P., Aries, A., Nemer, M., 2001. Serum response ­factor-GATA ternary complex required for nuclear signaling by a G-protein-coupled receptor. Mol. Cell Biol. 21, 1036–1044. Morin, S., Pozzulo, G., Robitaille, L., Cross, J.C., Nemer, M., 2005. MEF2-dependent recruitment of the HAND1 transcription factor results in synergistic activation of target promoters. J. Biol. Chem. 280, 32272–32278. Morisco, C., Seta, K., Hardt, S.E., Lee, Y., Vatner, S.F., Sadoshima, J., 2001. Glycogen synthase kinase 3beta regulates GATA4 in cardiac myocytes. J. Biol. Chem. 276, 28586–28597. Morrisey, E.E., Ip, H.S., Lu, M.M., Parmacek, M.S., 1996. GATA-6 – a zinc finger transcription factor that is expressed in multiple cell lin­ eages derived from lateral mesoderm. Dev. Biol. 177, 309–322. Morrisey, E.E., Ip, H.S., Tang, Z.H., Lu, M.M., Parmacek, M.S., 1997. GATA-5 – a transcriptional activator expressed in a novel temporally and spatially-restricted pattern during embryonic development. Dev. Biol. 183, 21–36. Morrisey, E.E., Tang, Z., Sigrist, K., Lu, M.M., Jiang, F., Ip, H.S., Parmacek, M.S., 1998. GATA6 regulates HNF4 and is required for differentiation of visceral endoderm in the mouse embryo. Genes Dev. 12, 3579–3590. Nemer, G., Qureshi, S.A., Malo, D., Nemer, M., 1999. Functional analysis and chromosomal mapping of GATA5, a gene encoding a zinc finger DNA-binding protein. Mamm. Genome. 10, 993–999. Nemer, G., Nemer, M., 2002. Cooperative interaction between GATA-5 and NF-ATc regulates endothelial–endocardial differentiation of ­cardiogenic cells. Development 129, 4045–4055.

615

Nemer, G., Nemer, M., 2003. Transcriptional activation of BMP-4 and regulation of mammalian organogenesis by GATA-4 and -6. Dev. Biol. 254, 131–148. Nemer, G., Fadlalah, F., Usta, J., Nemer, M., Dbaibo, G., Obeid, M., Bitar, F., 2006. A novel mutation in the GATA4 gene in patients with tetralogy of Fallot. Hum. Mutat. 27, 293–294. Oka, T., Maillet, M., Watt, A.J., Schwartz, R.J., Aronow, B.J., Duncan, S.A., Molkentin, J.D., 2006. Cardiac-specific deletion of Gata4 reveals its requirement for hypertrophy, compensation, and myocyte viability. Circ. Res. 98, 837–845. Pandolfi, P.P., Roth, M.E., Karis, A., Leonard, M.W., Dzierzak, E., Grosveld, F.G., Engel, J.D., Lindenbaum, M.H., 1995. Targeted disruption of the GATA3 gene causes severe abnormalities in the nervous system and in fetal liver haematopoiesis. Nat. Genet. 11, 40–44. Patient, R.K., McGhee, J.D., 2002. The GATA family (vertebrates and invertebrates). Curr. Opin. Genet. Dev. 12, 416–422. Pehlivan, T., Pober, B.R., Brueckner, M., Garrett, S., Slaugh, R., Van Rheeden, R., Wilson, D.B., Watson, M.S., Hing, A.V., 1999. GATA4 haploinsufficiency in patients with interstitial deletion of chromosome region 8p23.1 and congenital heart disease. Am. J. Med. Genet. 83, 201–206. Perlman, H., Suzuki, E., Simonson, M., Smith, R.C., Walsh, K., 1998. GATA-6 induces P21(C1P1) expression and G(1) cell cycle arrest. J. Biol. Chem. 273, 13713–13718. Peterkin, T., Gibson, A., Patient, R., 2003. GATA-6 maintains BMP-4 and Nkx2 expression during cardiomyocyte precursor maturation. EMBO J. 22, 4260–4273. Philips, A.S., Kwok, J.C., Chong, B.H., 2007. Analysis of the signals and mechanisms mediating nuclear trafficking of GATA-4. Loss of DNA binding is associated with localization in intranuclear speckles. J. Biol. Chem. 282, 24915–24927. Pikkarainen, S., Tokola, H., Majalahti-Palviainen, T., Kerkela, R., Hautala, N., Bhalla, S.S., Charron, F., Nemer, M., Vuolteenaho, O., Ruskoaho, H., 2003. GATA-4 is a nuclear mediator of mechanical stretch-activated hypertrophic program. J. Biol. Chem. 278, 23807–23816. Pu, W.T., Ishiwata, T., Juraszek, A.L., Ma, Q., Izumo, S., 2004. GATA4 is a dosage-sensitive regulator of cardiac morphogenesis. Dev. Biol. 275, 235–244. Reamon-Buettner, S.M., Borlak, J., 2005. GATA4 zinc finger mutations as a molecular rationale for septation defects of the human heart. J. Med. Genet. 42 (5), e32. Reiter, J.F., Alexander, J., Rodaway, A., Yelon, D., Patient, R., Holder, N., Stainier, D.Y., 1999. Gata5 is required for the development of the heart and endoderm in zebrafish. Genes Dev. 13, 2983–2995. Reiter, J.F., Kikuchi, Y., Stainier, D.Y., 2001. Multiple roles for Gata5 in zebrafish endoderm formation. Development 128, 125–135. Reyes, J.C., Muro-Pastor, M.I., Florencio, F.J., 2004. The GATA family of transcription factors in Arabidopsis and rice. Plant Physiol. 134, 1718–1732. Rivera-Feliciano, J., Lee, K.H., Kong, S.W., Rajagopal, S., Ma, Q., Springer, Z., Izumo, S., Tabin, C.J., Pu, W.T., 2006. Development of heart valves requires Gata4 expression in endothelial-derived cells. Development 133, 3607–3618. Scazzocchio, C., 2000. The fungal GATA factors. Curr. Opin. Microbiol. 3 (2), 126–131. Schluterman, M.K., Krysiak, A.E., Kathiriya, I.S., Abate, N., Chandalia, M., Srivastava, D., Garg, V., 2007. Screening and biochemical analysis of GATA4 sequence variations identified in patients with congenital heart disease. Am. J. Med. Genet. A 143, 817–823.

616

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Sepulveda, J.L., Belaguli, N., Nigam, V., Chen, C.Y., Nemer, M., Schwartz, R.J., 1998. GATA-4 and Nkx-2.5 coactivate Nkx-2 DNA binding targets: role for regulating early cardiac gene expression. Mol. Cell Biol. 18, 3405–3415. Serbedzija, G.N., Chen, J.N., Fishman, M.C., 1998. Regulation in the heart field of zebrafish. Development 125, 1095–1101. Small, E.M., Krieg, P.A., 2003. Transgenic analysis of the atrialnatriuretic factor (ANF) promoter: Nkx2-5 and GATA-4 binding sites are required for atrial specific expression of ANF. Dev. Biol. 261, 116–131. Stennard, F.A., Costa, M.W., Elliott, D.A., Rankin, S., Haast, S.J., Lai, D., McDonald, L.P., Niederreither, K., Dolle, P., Bruneau, B.G., Zorn, A.M., Harvey, R.P., 2003. Cardiac T-box factor Tbx20 directly interacts with Nkx2-5, GATA4, and GATA5 in regulation of gene expression in the developing heart. Dev. Biol. 262, 206–224. Suzuki, T., Kim, H.S., Kurabayashi, M., Hamada, H., Fujii, H., Aikawa, M., Watanabe, M., Watanabe, N., Sakomura, Y., Yazaki, Y., Nagai, R., 1996. Preferential differentiation of P19 mouse embryonal carcinoma cells into smooth muscle cells. Use of retinoic acid and antisense against the central nervous system-specific POU transcription factor Brn-2. Circ. Res. 78, 395–404. Svensson, E.C., Tufts, R.L., Polk, C.E., Leiden, J.M., 1999. Molecular cloning of FOG-2: a modulator of transcription factor GATA-4 in cardiomyocytes. Proc. Natl. Acad. Sci. USA 96, 956–961. Tenhunen, O., Sarman, B., Kerkela, R., Szokodi, I., Papp, L., Toth, M., Ruskoaho, H., 2004. Mitogen-activated protein kinases p38 and ERK 1/2 mediate the wall stress-induced activation of GATA-4 binding in adult heart. J. Biol. Chem. 279, 24852–24860. Tevosian, S.G., Deconinck, A.E., Cantor, A.B., Rieff, H.I., Fujiwara, Y., Corfas, G., Orkin, S.H., 1999. FOG-2: A novel GATA-family cofactor related to multitype zinc-finger proteins Friend of GATA-1 and U-shaped. Proc. Natl. Acad. Sci. USA 96, 950–955. Tevosian, S.G., Deconinck, A.E., Tanaka, M., Schinke, M., Litovsky, S.H., Izumo, S., Fujiwara, Y., Orkin, S.H., 2000. FOG-2, a cofactor for GATA transcription factors, is essential for heart morphogenesis and development of coronary vessels from epicardium. Cell 101, 729–739. Tevosian, S.G., Albrecht, K.H., Crispino, J.D., Fujiwara, Y., Eicher, E.M., Orkin, S.H., 2002. Gonadal differentiation, sex determination and normal Sry expression in mice require direct interaction between transcription partners GATA4 and FOG2. Development 129, 4627–4634. Tremblay, J.J., Viger, R.S., 2003. Transcription factor GATA-4 is activated by phosphorylation of serine 261 via the cAMP/protein kinase a sig­ naling pathway in gonadal cells. J. Biol. Chem. 278, 22128–22135. Tsai, F.Y., Keller, G., Kuo, F.C., Weiss, M., Chen, J., Rosenblatt, M., Alt, F.W., Orkin, S.H., 1994. An early haematopoietic defect in mice lacking the transcription factor GATA-2. Nature 371, 221–226. Viger, R.S., Mertineit, C., Trasler, J.M., Nemer, M., 1998. Transcription factor GATA-4 is expressed in a sexually dimorphic pattern during

mouse gonadal development and is a potent activator of the Müllerian inhibiting substance promoter. Development 125, 2665–2675. Wada, H., Hasegawa, K., Morimoto, T., Kakita, T., Yanazume, T., Abe, M., Sasayama, S., 2002. Calcineurin-GATA-6 pathway is involved in smooth muscle-specific transcription. J. Cell Biol. 156, 983–991. Wada, H., Hasegawa, K., Morimoto, T., Kakita, T., Yanazume, T., Sasayama, S., 2000. A p300 protein as a coactivator of GATA-6 in the transcription of the smooth muscle-myosin heavy chain gene. J. Biol. Chem. 275, 25330–25335. Waldo, K.L., Kumiski, D.H., Wallis, K.T., Stadt, H.A., Hutson, M.R., Platt, D.H., Kirby, M.L., 2001. Conotruncal myocardium arises from a secondary heart field. Development 128, 3179–3188. Waltzer, L., Bataille, L., Peyrefitte, S., Haenlin, M., 2002. Two isoforms of Serpent containing either one or two GATA zinc fingers have different roles in Drosophila haematopoiesis. EMBO J. 21, 5477–5486. Wang, J., Feng, X.H., Schwartz, R.J., 2004. SUMO-1 modification activated GATA4-dependent cardiogenic gene activity. J. Biol. Chem. 279, 49091–49098. Wang, J., Paradis, P., Lefebvre, C., Nemer, M., 2005. Convergence of protein kinase C and JAK-STAT signaling on transcription factor GATA4. Mol. Cell Biol. 25, 9829–9844. Watt, A.J., Battle, M.A., Li, J., Duncan, S.A., 2004. GATA4 is essential for formation of the proepicardium and regulates cardiogenesis. Proc. Natl. Acad. Sci. USA 101, 12573–12578. Wilson, R.A., Arst Jr., H.N., 1998. Mutational analysis of AREA, a transcriptional activator mediating nitrogen metabolite repression in Aspergillus nidulans and a member of the “streetwise” GATA family of transcription factors. Microbiol. Mol. Biol. Rev. 62, 586–596. Xin, M., Davis, C.A., Molkentin, J.D., Lien, C.L., Duncan, S.A., Richardson, J.A., Olson, E.N., 2006. A threshold of GATA4 and GATA6 expression is required for cardiovascular development. Proc. Natl. Acad. Sci. USA 103, 11189–11194. Yang, H., Lu, M.M., Zhang, L., Whitsett, J.A., Morrisey, E.E., 2002. GATA6 regulates differentiation of distal lung epithelium. Development 129, 2233–2246. Zeisberg, E.M., Ma, Q., Juraszek, A.L., Moses, K., Schwartz, R.J., Izumo, S., Pu, W.T., 2005. Morphogenesis of the right ventricle requires myocardial expression of Gata4. J. Clin. Invest. 115, 1522–1531. Zhao, R., Watt, A.J., Li, J., Luebke-Wheeler, J., Morrisey, E.E., Duncan, S.A., 2005. GATA6 is essential for embryonic development of the liver but dispensable for early heart formation. Mol. Cell Biol. 25, 2622–2631 Zhao R., Watt A.J., Battle M.A., Li J., Bondow B.J., Duncan S.A., 2008. Loss of both GATA4 and GATA6 blocks cardiac myocyte differentiation and results in acardia in mice. Dev. Biol., 317(2), 614–619.

Chapter 9.3

Serum Response Factor and Co-Factors, Roles in Cardiac Development Robert J. Schwartz The Institute of Biosciences and Technology, The Texas A&M University System Health Science Center, Houston, TX, USA

I.  Introduction The myogenic gene switch paradigm promulgated more than 25 years ago suggested that nonmuscle contractile protein isoforms were “switched off” during terminal differention, only to be replaced by the “switched on” sarcomeric specific protein isoforms (Schwartz and Rothblum, 1981; Minty et al., 1982; Bains et al., 1984). This regulatory paradigm holds well for early cardiogenesis, in which nonmuscle -actins were replaced by smooth muscle and cardiac -actin during heart formation in vertebrate embryos (Ruzicka and Schwartz, 1988; Sugi and Lough, 1992; Colas et al., 2000). Replacement of the nonmuscle actins by sequential activation of smooth muscle -actin and -striated actins in the embryonic heart also occurs during somitogenesis, and may be a general property of primitive muscle mesoderm, undergoing a transition from the earliest stages of myogenic commitment in a cardiomyoblast towards terminal differentiation of a cardiac myocyte. Contrary to early expectations that a master gene may specify cardiac lineages, it is now generally acknowledged that expression of cardiac-specified genes requires combinatorial interactions between transcription factors enriched during the emergence of cardiac progenitor cells. Cardiac progenitors receiving the appropriate developmental cues switch on several cardiac-restricted transcription factors, such as Nkx2-5, GATA-4, Tbx5 and myocardin that interact with serum response factor (SRF) or potentiate its activation of many cardiac and smooth muscle structural genes (reviewed in Pipes et al., 2006; Niu et al., 2007). Serum response factor target genes are also involved with contractility, cell movement, cell growth signaling and the recently-discovered microRNAs required for normal heart development (reviewed in Posern and Treisman, 2006; Miano et al., 2007; Niu et al., 2007). Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

This chapter highlights how serum response factor regulates target genes during the elaboration of the embryonic cardiovascular system. First, a focus on recent conditional knockout studies supports an obligatory role for serum response factor in cardiogenesis, which has also led to the discovery of hundreds of serum response factor-dependent coding gene targets and many noncoding microRNAs expressed in the heart (see also Chapter 10.3). Second, serum response factor gene expression is under a growth factor autoregulatory loop in fibroblasts, but is also highly-dependent on T-box factors and its 3UTR enhancer for robust expression in the developing heart (see also Chapter 9.4). Third, serum response factor co-factors that appear during cardiogenesis will be scrutinized for their ability to form combinatorial complexes that activate and or repress serum response factordependent cardiovascular restricted gene programs (see also Chapter 9.5). Finally, we will spotlight a novel regulatory mechanism whereby specific phosphorylation of a single amino acid residue regulates serum response factor-dependent myogenic gene activity during muscle differentiation.

II.  Serum response factor Serum response factor, a 67 kd DNA-binding protein first discovered by Richard Treisman and colleagues (Norman et al., 1998) as a factor that bound to the serum response element in the c-fos promoter (Treisman, 1986), is the founding member of an ancient DNA-binding protein family, which shares a highly-conserved DNA-binding/ dimer­ization domain termed the MADs-box (Sommer et al., 1990) (Fig. 1). MADS stands for the MCM1, Agamous, Deficiens and serum response factor family of transcription factors, which share homology in a motif (MADS-box) that mediates homodimerization and DNA-binding to a dyad 617

618

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

symmetrical A  T-rich DNA consensus sequence (Shore and Sharrocks, 1994). There are numerous MADS-box proteins in plants, but serum response factor and the four members of the myocyte enhancer factor-2 (MEF2) family (MEF2A, MEFB, MEFC and MEFD) are the only MADSbox proteins found in metazoans (Black and Olson, 1998; Chapter 9.5). The crystal structures of serum response factor and MEF2 have revealed commonalities in their modes of DNA-binding (Pellegrini et al., 1995; Santelli and Richmond, 2000), reflected in the similar sequences of their binding sites. Serum response factor target genes are characterized by the presence of single or multiple copies of the serum response factor-binding consensus element CC(A/T)2 A(A/T) 3GG, otherwise known as the CArG box (Minty and Kedes, 1986), with a variant CTA(A/T)4TAG for MEF2 target genes. CArG boxes are found primarily in genes involved with contractility, cell movement and cell growth sig­ naling. The full spectrum of functional CArG elements in the genome was recently named the CArGome (Sun et al., 2006), representing known and novel serum response

factor-binding sequences with preferred base composition across the CArG element, as shown in Fig. 1. Richmond and colleagues elucidated the X-ray crystal structure of the SRF MADS-box bound to DNA (Pelligrini et al., 1995), shown in a schematic diagram in Fig. 1B,D. The conserved N-terminal region of the MADS-box, an -helical structure, becomes oriented in an antiparallel manner within homodimers to form a bipartite DNAbinding domain. The MADS-box I helices (aa153–aa179) align with a narrow DNA major groove, making contacts with the phosphate backbone on a CArG half site. In addition, the unstructured N-terminal extension towards the I helix (aa132–152aa) makes critical base contacts in the minor groove. Dimerization of the MADS-box occurs above the I-helix by a structure composed of two -sheets in the monomer that interact with the same unit in its partner. A second II helix in the C-terminal portion of the MADS-box, stacked above these -sheets, completes the stratified structure. Serum response factor, like other MADS-box transcription factors, interacts with a diverse

DNA Dimer TCF

(A) SRF 1 (B)

140

141

150

160

223 170

406 180

190

508 200

210

220

(C) αI

N-extension

(D)

oil

αll c

β–loop

β I β Ioop βIl

αIl

(E) CArG element

αΙ coil

-5 4 –3 –2–1 1 2 3 4 5 bases

Figure 1  Serum response factor contains an evolutionary-conserved MADS-box domain that allows for dimerization of serum response factor monomers and DNA binding to a CArG element. (A) Schematic diagram of human serum response factor with a transcription activation domain (TAD) and the DNA-binding and dimerization domain between 141aa to 223aa. (B) This diagram shows the amino acid sequence of the N-extension, I coil, 1 sheet, -loop, II sheet and II coil of the MADS-box. (C) Protein sequences of different MADS domains from human (H), mouse (M) and Xenopus (X) serum response factor, MEF2A, -B, and -C, yeast ARG80, MCM1 and plant AGL1 were aligned to show a high degree of sequence conservation across broad evolutionary time. Standard single amino acid letter codes were used in alignments. (D) Diagram of the serum response factor-DNA crystal structure was adapted from Pellegrini et al. (1995). (E) Sequence of serum response factor targets showed a high level of similarity for preferred base composition across the CArG elements named CArGome, as reported by Sun et al. (2006).

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

array of transcriptional regulators to generate tissuespecific and signal-responsive patterns of gene expression. A central problem in development is to understand how serum response factor activity is differentially controlled according to muscle cell type and/or signaling pathway. The discovery that serum response factor activity is controlled to a large extent by its interaction with signal-regulated or tissue-specific regulatory co-factors provided the first mechanistic insights into the problem (reviewed in Posern and Treisman, 2006; Niu et al., 2007). In fibroblasts, serum response factor controls transcription of many cellular “immediate-early” genes, whose expression is activated by growth factor or mitogenic stimuli. In nonmyogenic fibroblasts, serum response factor target genes can be classified into two types by their sensitivity to the Rho-actin or Ras-ERK signaling pathways. Members of the ternary complex factor (TCF) family of Ets domain proteins, Elk-1, SAP-1 and SAP-2/Net, are activated by MAPK phosphorylation. Early studies focused on serum response factor-accessory proteins that stimulate c-fos induction, as endpoints of signal transduction cascades often leading to mitogenesis. For example, the Ets-related ternary complex factor Elk-1 or SAP1 interacts with serum response factor, binding to the c-fos CArG box in a cooperative manner. Phosphorylation of Elk-1 by mitogen responsive JNK and ERK groups of mitogen-activated protein (MAP) kinases causes increased DNA binding, ternary complex formation and transcriptional activation. Activated Elk-1 touches the MADS-box and binds to an adjacent Ets site (reviewed in Treisman, 1994, 1995). More recently, a second family of signal-regulated serum response factor co-factors, the myocardin-related transcription factors (MRTFs), have been well-characterized. Olson and colleagues first identified myocardin, the founding member of this group of co-activators, which binds to and functionally synergizes with serum response factor in the heart (Wang et al., 2001, 2002, 2003, 2004; Chapter 9.5). Expression of myocardin is restricted to smooth muscle and cardiac tissue, but that of the other two myocardin-related transcription factor family members, MRTF-A and MRTF-B, is extensive. Activity of the two myocardin-related transcription factors is regulated by a novel signaling pathway controlled by Rhofamily GTPases and monomeric actin (Sotiropioulos et al., 1999; Miralles et al., 2003), while myocardin itself appears to activate serum response factor directly. The non­myogenic roles of these important pathways have been recently reviewed (Pipes et al., 2006; Posern and Treisman, 2006). The importance of serum response factor in normal developmental processes has been documented in species ranging from yeast to mice (Guillemin et al., 1996; Arsenian et al., 1998; Escalante et al., 2004; Miano et al., 2004; Li et al., 2005; Niu et al., 2005). Despite their similarities, MADSbox proteins have evolved to perform important biological functions, such as specification of mating type haploid cells in yeast, homeotic activities in plants, and muscle

619

specification in C. elegans and vertebrates. Another MADS-family member, the yeast protein MCM1, like serum response factor, is also influenced by the recruitment of an array of accessory factors that either activate or repress genes in a cell type-specific and temporal pattern (Herskowitz, 1989). These pioneering studies set the stage for MADS-box-dependent gene activity to be modified through signal transduction and require combinatorial factor interactions and post-translational modifications. In myogenic cells, positive-acting serum response factor co-factors include the members of the GATA family of zinc-finger transcription factors, and the Nkx2-5 family of homeodomain proteins, which can form complexes both with serum response factor and their own adjacent recognition site (see Chapter 9.1). The cysteine-rich LIM-only proteins may act as bridging factors to facilitate formation of GATA-SRF complexes, thus allowing muscle differentiation. Myocardin is another important factor that drives smooth muscle gene activity by association with serum response factor, without binding directly to DNA. Negative-regulating serum response factor co-factors include YY1, a competitor of serum response factor’s DNA-binding activity, and the heart-enriched homeodomain-only co-factor HOP.

III.  Embryonic serum response factor expression is largely restricted to cardiac and skeletal muscle tissues Although serum response factor was first purified from Hela cells, it is actually expressed in a highly-restrictive pattern throughout chick and mouse development (Croissant et al., 1996; Barron et al., 2005). In situ hybridization analysis of mouse embryos, using a serum response factor RNA probe, indicated that serum response factor transcripts are somewhat diffuse early in development, with concentrated expression in the lateral plate mesoderm and primitive streak (Fig. 2A) at 7.5 days post coitum (dpc). These areas are of significance, because precardiac mesoderm cells migrate through the streak to take up residence in the anterior lateral plate, whereas skeletal muscle originates in myotomes of the paraxial mesoderm. Serum response factor is concentrated in the cardiac crescent (Fig. 2A), and later in the heart tube and developing somites (Fig. 2B,C). These tissues continue to express high levels of serum response factor throughout development. Similar to chick embryos, serum response factor transcripts in the mouse appear highly restricted to myogenic mesoderm and neuroectoderm. Serum response factor -galactosidase knockin reporter mice faithfully reproduce the endogenous serum response factor gene expression pattern (Barron et al., 2005). As seen in Fig. 2D,E, embryos ranging from 7.5 dpc to 9.5 dpc showed serum response factor-LacZ staining in the cardiac

620

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Figure 2  Serum response factor transcripts are highly-expressed in the early heart during early murine embryogenesis. (A) Whole mount in situ hybridization analysis revealed serum response factor mRNA in the lateral plate mesoderm and cardiac crescent of a 7.5 dpc embryo. (B) Enriched serum response factor RNA appeared in the cardiac crescent (arrow) and (C) inflow regions of the primitive heart tube. (D) LacZ expression in a serum response factor-LacZ knockin mouse line shown in 7.5 dpc and (E) 9.5 dpc mouse embryos (whole mount views), and (F) 11 dpc mouse embryo (transverse section) adapted from Barron et al. (2005). (G) Serum response factor expression was highly-enriched in embryonic cardiac myocytes during early mouse development at 10.5 dpc. Serum response factor immunofluorescence staining (green) in atrial and ventricular myocytes marked by an antibody against striated MHC, MF20 (red), as adapted from Niu et al. (2005) (cc: cardiac crescent; ht: heart tube; A: atrium; V: ventricle).

crescent, somites and the precondensed mesenchyme in the tail, consistent with the in situ pattern of the endogenous serum response factor gene. At later stages, 11.5 dpc embryos (Fig. 2F) showed serum response factor-LacZ expression in the myocardium of the heart, the somites and strong staining in the tail region (Fig. 2E). Serum response factor-LacZ expression was also observed in the smooth muscle of blood vessels and the ventral portion of the neural tube (Fig. 2F). Thus, serum response factor transcriptional activity, revealed by LacZ expression, reinforced the notion of restricted developmental expression. The appearance of serum response factor protein was validated by immunofluorescence staining in the hearts of mouse embryos (Fig. 2G), where serum response factor was predominantly localized to myocytes, as demonstrated by the positive co-expression of serum response factor with muscle myosin heavy chain (MHC) protein in atrial and ventricular myocytes. The enrichment of serum response factor in myocytes early in heart development supports a significant role for serum response factor in muscle specification and differentiation.

IV.  Serum response factor orchestrates cardiac myogenesis IV.A.  Myogenic Contractile Proteins are Downregulated in Serum Response   Factor-Null Embryonic Stem Cells Recently, serum response factor inactivation studies in the developing mouse heart were performed through a

conditional knockout strategy using Cre-recombinase driven by several late-expressing transgenic promoters such as SM22a or myosin-heavy chain  or  (Miano et al., 2004; Parlakian et al., 2004; Niu et al., 2005). In each study, cardiac-specific ablation of serum response factor resulted in embryonic lethality due to cardiac insufficiency during chamber maturation. Conditional ablation of serum response factor also reduced cell survival concomitant with increased apoptosis and reduced cellularity. Significant reductions in serum response factor (95%), atrial naturetic factor (80%), and cardiac (60%), skeletal (90%) and smooth muscle (75%) -actin transcripts were also observed in the cardiac-conditional serum response factor knockout heart, consistent with the idea that serum response factor directs de novo cardiac and smooth muscle gene activities (Niu et al., 2005). Although the serum response factor gene was efficiently deleted by 9.5 dpc, serum response factor is rather stable and turns over slowly in the heart allowing for the appearance of actin gene expression and cardiac beating (Niu et al., 2005) Eventually, serum response factor protein dis­appeared, which in part contributed to turnover of contractile proteins and disorganized sarcomeres in serum response factor myocytic knockout embryonic hearts. In addition, a thinner ventricular compact layer was caused by decreased levels of the anti-apoptotic factor Bcl2, a serum response factor target (Schratt et al., 2004), and increased doubling in the number of TUNEL-positive cells undergoing apoptosis in the serum response factor-null myocardium. To block the activation of serum response factor expression before the appearance of beating cardiac myocytes, we engineered a mouse that carried both SrfLacZ and

621

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

(A)

300

# of Embryos

Nkx2-5Cre on chromosome 17, which was then bred to SRFLox/Lox mice to generate a conditional serum response factor knockout (SrfCko) in the heart-forming region. This SrfCko embryonic mutant failed to initiate cardiac beating at 8.0–8.5 dpc (Fig. 3A). Immunofluorescence staining with anti-serum response factor antibodies showed serum response factor present in myocytes of haploid serum response factor mutant embryos, but absent in the SrfCko embryo (Fig. 3B,C). Smooth muscle and cardiac -actin gene RNA transcripts emergent at the late cardiac crescent stage (7.75–8.0 dpc) were blocked in SrfCko mutant hearts (Fig. 3D–K). Immunofluorescence staining

200

confirmed the absence of smooth muscle and striated -actin in the hearts of SrfCko embryos. Analysis by transmission electron microscopy indicated that neither aligned filaments nor Z disks were formed in multiple SrfCko cardiac mutants, correlating well with the nonbeating heart (Fig. 3L,M). These “paralyzed” mutant hearts did not display any sarcomere signatures in multiple SrfCko mutant samples (Fig. 3N). Expression of Myl2 and Myom1 components of the thick filament and M-band of sarcomeres were dependent on serum response factor expression (Fig. 3O–R). The appearance of Smyd1, a cardiac and skeletal muscle-specific SRF

280 279

(B)

(B)

Beating

da da

Embyros

83

100

da

h

h

0 ctr SM α-actin (D)

ctr

cko SM α-actin (F)

(E)

da

da

(G)

(I)

(M)

ctr

cko

CA α-actin (H)

(L)

h

ctr

cko

da

da

h ctr

cko

CA α-actin (J)

(N) 100%

(K)

cko 100%

75%

ctr cko

62%

48%

50% 25% h cko

ctr

h

ctr

0%

0%

41% 0%

Parallel transverse Z-line disarrayed Filaments

cko

(O)

Myl2 (P)

(Q)

Myom1 (R)

(S)

Smyd 1 (T)

(U)

ctr

cko

ctr

cko

ctr

cko

ctr

Hand 1 (V)

cko

(W)

ctr

Kcnmb 1 (X)

cko

Figure 3  Serum response factor insufficiency shuts down cardiac myocyte differentiation. (A) Quantitative summary of the number of beating hearts observed for control and SrfCko embryos at 8.25 dpc. (B, C) Immunofluorescent staining of control Srf Lox/ and SrfCko embryo sections antiserum response factor (green) antibodies. (D–K) Whole mount in situ hybridization (D, E, H, I) and immunofluorescence (F, G, J, K) staining showed the absence of Acta2 (smooth muscle -actin) and Actc1 (cardiac -actin) expression in the SrfCko embryo in comparison to control embryos at 8.5 dpc. Transmission EM revealed organized sarcomere structure in control (L), but not SrfCko sections (M). Quantitative summary of contractile structures recognized in multiple control and SrfCko samples (N). In control samples Z-lines were observed at high incidence (N  10/21 myocytes), while aligned filament bundles were observed in all control samples. “Disarrayed” filament structures shown within the dotted lines were observed in less than half the Srfcko mutant cells (N  11/27 cells) which shared a diameter at 10 nm, the same as non-contractile intermediate filaments (M, N). Whole mount in situ hybridization showed downregulation of Myl2 (O, P), Myom1 (Q, R), Smyd1 (S, T), Hand1 (U, V) and Kcnmb1 (W, X) in SrfCko embryos. Adapted from Niu et al. (2008).

622

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

chromatin remodeling factor (Gottlieb et al., 2002), failed to appear in the nascent SrfCko mutant myocytes (Fig. 3S,T). Expression of the Hand1 gene, a first heart field marker, was reduced in the SrfCko mutant (Fig. 3U,V) and may have also contributed to ventricular dilation. Appearance of calcium-activated potassium channel gene activity (KCNMB1) was blunted in the SrfCko mutant hearts (Fig. 3W,X) as well as other excitation–contraction coupling pathway proteins Gja5 and Jph2. Many of these genes were validated for the presence of conserved CArG elements in their promoters. Niu et al. (2008) provided incontrovertible evidence that serum response factor activity controls sarcomerogenesis in higher vertebrates (as shown in the schematic diagram in Fig. 4). Serum response factor orchestrates cardiac sarcomerogenesis through multiple levels of regulation, from the appearance of cardiogenic chromatin remodeling factors and myofibrillar proteins to excitation contraction coupling proteins leading to the first heartbeat. In addition, the Gene Ontology (GO) project allowed us to analyze the global biological impact of serum response factor on the expression of functionally-related groups of genes (Fig. 5). Ten GO terms fulfilled the Benjamini test (conservative correct p-values), and included muscle contraction, actin cytoskeleton, myosin, contractile fiber, sarcomere, myofibril, cytoskeleton, A band, motor activity, muscle development, cytoskeleton organization and biogenesis, thus correlating well with serum response factor primary role recast as the “Sarcomeric Regulatory Factor.” Recently, Fukushige et al. (2006) demonstrated that UNC-120, the C. elegans equivalent of serum response factor, was required to initiate a body wall muscle program. The role of serum response factor as a universal “myogenic driver” was totally abrogated in the Srf-null cells, supporting the concept that serum response factor resides at a high point in the regulatory hierarchy governing sarcomerogenesis from worms to mammals.

SRF

Assembling Structural comonents protein Ion channel

Sarcomere 1st Heartbeat Figure 4  Serum response factor orchestrates cardiac sarcomerogenesis through multiple levels of regulation from the appearance of cardiogenic chromatin remodeling factors, myofibrillar proteins and excitation contraction coupling proteins leading to the first heartbeat. Niu et al. (2008) provided inconvertible evidence that serum response factor activity controls sarcomerogenesis in higher vertebrates, as modeled in the schematic diagram above.

Surprisingly, we also observed strong upregulation of genes associated with the four GO terms involving bio­ mineral formation, ossification, bone remodeling and extracellular matrix. Microarray analysis revealed genes that were strongly induced in the absence of serum response factor which included GATA6, BMP4, endothelin and periostin, key factors involved with endocardial specification (Fig. 5). GATA6 contributes to septal and valvular development via its direct transcription target, BMP4, through the induction of periostin (Ma et al., 2005). Transformation of atrioventricular (AV) canal endocardium into invasive mesenchyme correlates spatially and temporally with the expression of BMPs in the AV myocardium crucial for the induction of periostin (Okagawa et al., 2007). Also, periostin inhibited cardiac myocyte differentiation (Niu et al., 2008).

IV.B.  Serum Response Factor Directs the Expression of Many MicroRNAs Although serum response factor is generally not considered as an inhibitory transfactor, there is a greater likelihood that serum response factor indirectly exerts gene silencing activities through its regulation over miRNAs. Based on sequence conservation and the ability to fold into a hairpin structure, as many as a thousand miRNAs are predicted in the human genome, which are estimated to regulate as many as 30% of mRNA transcripts (Berezikov et al., 2005). Informatics helped us to determine that approximately 169 miRNA genes in mammalian genomes contain at least one CArG element in their promoter regions, while at least 40 miRNA genes contain three or more CArGs (Niu et al., 2007). Obviously, ablation of serum response factor may block the expression of many miRNAs, causing rampant and complex dysregulation in cardiac development. In support of this idea, Srivastava and colleagues (Zhao et al., 2005) showed that serum response factor regulates the expression of miR-1-2, through its three CArG boxes in the miR-1-2 promoter (Chapter 10.3). MiRNAs control cellular regular pathways through post-transcriptional regulation of protein expression (Ambros, 2004). MiRNA may modify cellular events, function as molecular switches, and even silence superfluous mRNAs in specific cell lineages (Cohen et al., 2006). Serum response factor may also exert gene silencing activities through its regulation over microRNAs (miRNAs). In day 9.5 SrfCko embryonic mutant hearts, at least 20 miRNAs, each of which contains at least one conserved CArG element in their promoters, were downregulated in comparison to control heart samples (Fig. 6A). Among these miRNAs, miR-1 (Zhao et al., 2005) was shown to be controlled by serum response factor through its three CArG boxes (shown schematically in Fig. 6B). In addition, mice lacking miR-1-2 have a spectrum of cardiac abnormalities including the induction of GATA6 (Zhao et al., 2007). More than 90% of miR-1 RNA transcripts were blocked in the SrfCko hearts, whereas GATA6 mRNA levels increased

623

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

Upregulated in SRFCKO

Downregulated in SRFCKO SRFCKO

Embryo ctr

Expected

GO Biological process

SRFCKO Srf* Myom1* Myl3* Myl9* Cnn1* Tnnc2* Tagin*

GO Biological process Biomineral formation

Muscle development Muscle contraction

Ossification

Cell adhesion

Bone remodeling

Intracellular signaling cascade

Tissue remodeling

Sema3C* Acta2* Actg2* Acta1* Sntb1 Pdlim7 Jam2 Tgfb1/1*

Skeletal development

Development 10%

20%

GO Cellular component

Regulation of organismal physiological process Tissue development

Troponin complex

Embryonic development

Z disc

Cell adhesion

Sarcomere

Cell differentiation

Myofibril

Development

Hand1* Smyd1* Jph2 Nppa* Kcnmb1*

10% 20% 30%

Contractile fiber

GO Cellular component

Actin cytoskeleton

Extracellular matrix (sensu metazoa)

Cytoskeleton

Extracellular space

Intracellular 10% 50% 60%

Down regulated

Extracellular region 10% 20% 30% 40%

1 2 –2.5 –2 –1 0 Relative expression

2.5

Embryo Ctr

Cko

Pln Edn1 Dusp4 Gata6 Pdgfra Ndrg1 Postn Bmp4

Up regulated

–2.5 –2 –1 0 1 2 2.5 Relative expression

Figure 5  Analysis of srfCko mutant on global gene transcription revealed downregulation of contractile proteins and ion channels and upregulation of bone-related proteins in embryonic hearts. To dissect out serum response factor’s role during cardiac myocyte commitment and differentiation, we generated lineage-specific deletion of Srf with Cre/loxP system (Niu et al., 2008). RNA samples isolated from embryonic hearts were pooled and hybridized against Affymetrix array 430a2 chip. Microarray raw data analysis was done with dCHIP software (www.dchip.org). Genes affected by the serum response factor knockout were treated for functional classification using DAVID 2007 (http://david.abcc.ncifcrf.gov/). The GO term classification uses a program that analyzes for functional associations. The ablation of serum response factor in the heart revealed 10 GO terms that were associated with downregulated genes and fulfill the Benjamini test (conservative correct p-values) including muscle contraction, actin cytoskeleton, myosin, contractile fiber, sarcomere, myofibril, cytoskeleton, A band, motor activity, muscle development, cytoskeleton organization and biogenesis. Conversely, upregulated genes were associated with the four GO terms biomineral formation, ossification, bone remodeling and extracellular matrix.

over 18-fold (Niu et al., 2008). Thus, tissue-specific expression of miRNAs regulated by serum response factor at the transcriptional level may exert strong regulation of noncontractile endocardial/epicardial gene activity.

V.  Serum response factor in human heart disease V.A.  Inhibitory Serum Response Factor is Generated by Caspase 3 Cleavage in Human Heart Failure Serum response factor also has important roles in adult cardiac myocytes and heart failure. Not surprisingly, the activity of serum response factor must be tightly regulated

since persistantly increased levels of serum response factor may have similar deleterious effects to decreased levels. Zhang et al. (2001) reported that excessive transgenic expression of serum response factor led to dilated cardiomyopathy, with early mortality and impaired ventricular function. The conditional ablation of serum response factor in adult hearts showed that persistant serum response factor activity is required for the long-term maintenance of normal sarcomeric organization and contractility in the adult murine heart (Parlakian et al., 2006). The studies by Zhao et al. (2007) and van Rooij et al. (2007) also indicated that miRNA dosage is important during heart development and hypertrophy, and could constitute potential risk factors for adult cardiovascular disease and heart failure. In addition, Chen et al. (2006) showed that expression of miR-133 repressed myoblast differentiation by repressing

624

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Samples

ctr-1 ctr-2 ctr-3 ctr-4 ctr-5 ctr-6 cto-1 cto-2 cto-3 cto-4 cto-5 cto-6 CArG(s )

(A)

3 2 4 1 3 1 2 1 1 6 2 2 2 2 2 6 1 1 4 2

1.5

0.0

miR-377 miR-143 mir-30c miR208b miR-1 miR-92-b miR-433 miR-126-3p miR-342-3p miR-451 miR-495 miR-127 miR-376a miR-151-5p miR-133b miR-133a miR-210 miR-103 miR-106b miR-503

(B) SRF Ch2(+)

3(CArG)

Ch 18(-)

3(CArG)

lrx4/5 Cardiac Rhythm

-1.5

Relative expression

Kcnd2

miR-1-1

miR-133a-2

miR-1-2

miR-133a-1

GATA-6

Hand2

BMP4 Cell cycle

Cardiac Repair

Periostin Endocardial lineages

Figure 6  Regulation of serum response factor-dependent microRNA and miR-1 targets. (A) Microarray analysis of serum response factor-dependent microRNAs that were downregulated in the srfcko mutant hearts. Genome-wide microRNA expression profiling was accomplished with RNA samples taken from control and SRFCKO mutant hearts and hybridized on a mouse microRNA microarray platform (MRA-1002, LC Sciences, Houston, TX) contained 568 unique mature miRNA sequences. (B) miR1-2 regulates cardiac morphogenesis, cardiac conduction and cell-cycle. Schematic diagram shows the two miR-1 bicistronic genes on chromosomes 2 and 18, and summarizes some of the roles for miR-1-2 and miR-133 in regulating Irx4/5, GATA6 and Hand2 and serum response factor expression that were described in recent studies. Hypothetical regulation of the endocardial cell lineage specification via GATA6, BMP4 and periostin was shown in Zhao et al. and verified in Niu et al. (2008).

serum response factor expression. Thus, miR-133 is controlled by serum response factor, yet directs a negative regulatory loop through inhibiting serum response factor translation (Fig. 6). It is not surprising that serum response factor activity must also be tightly-regulated through an miR-133-dependent negative feedback loop, because enhanced levels of serum response factor in mice have led to dilated cardiomyopathy, while depletion of serum response factor showed that its activity was required for the maintenance of normal sarcomeric organization and contractility (Parlakian et al., 2005). Possibly, upsetting the delicate balance of miR-133 silencing serum response factor-regulated cardiac hypertrophy may have even more profound effects in human heart disease. Indeed, Chang et al. (2003) found serum response factor was cleaved by caspase 3 in human heart failure, acting as a dominant-negative modifier in the failing human myocardium. Examination of serum response factor protein sequences revealed 2 caspase consensus cut sites at aspartate residues 245 and 254, which generated dual 32 kDa fragments, as shown in Fig. 7. The clipping sites were mapped by N and C terminal specific anti­bodies and a mutated SRF species A245/A254 showed a complete block to caspase 3, which indicated that both native sites are subject to caspase 3 cutting. There are at least two mechanisms involved in the suppression of serum response factor-dependent gene activation. First, there is

a striking reduction in naturally-occurring serum response factor. In addition, caspase 3 cleaved SRF-N functions as a dominant-negative inhibitor. This serum response factor fragment contains an intact MADS-box, but lacks the C-terminal transactivation domain. Therefore, cleaved serum response factor competes with full-length serum response factor for binding to CArG boxes, and will not be able to transactivate serum response factor gene targets efficiently. This is consistant with the downregulation of a broad range of genes, including those with serum response factor-binding sites, such as cardiac -actin, calponin and SERCA (sarcoendoplasmic reticulum Ca-ATPase), confirming the critical role of serum response factor in maintaining cardiac contractility. Although caspase 3 cleavage of serum response factor plays a central role in promoting human heart failure and possibly apoptosis, it is not known whether proteolytic processing of serum response factor has a role in cardiac development.

VI.  Serum response factor gene autoregulation The serum response factor gene is an important target for the ERK signaling pathways that converge on ternary complex factors (TCF), some of which are a subfamily of ETS domain transcription factors such as Elk-1, as modeled

625

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

(A)

failing HR normal HR

without LVAD

(D)

with LVAD

SRF cleaved SRF

65 KD 55 KD

cleaved SRF

32 KD

(B)

N

Exon1

Activation domain

2

3

Caspase 3 N-SRF dominant negative

4

Relative to control

Active Caspase 3

MADS Boc

N

3

Exon1

2

3

4

5

6

7

C

7

C

Activation domain 4

5

6

n = 13 *

2

n = 10

Competitive DNA Binder

n=7

1

N

Exon1

?

2

0

(C)

Normal HR Without LVAD with LVAD normal HR

Caspase 3 precurosr Active Caspase 3

failing HR

failing HR with LVAD 32 kDa

SRF gene targets including cardiac specific gene programs contractile proteins, channels and microRNAs

20 kDa

Figure 7  Full-length serum response factor cleaved by Caspase 3 in human failing hearts generates a competitive serum response factor dominant negative inhibitor. (A) Representative Western blot. Each lane represents a single human sample from two normal and nine heart failure patients, five of whom were on LVAD. (B) Densitometry analysis of small fragment (32 kDa) from 13 heart failure patients and 10 LVAD patients was made based on Western blots and demonstrated a significant difference (*P  0.03) compared with seven normal heart samples. No significant difference was found between normal heart and LVAD samples (HR: heart). (C) Active caspase 3 (20 kDa) level was increased in human failing hearts and returned to normal level after LVAD. Proteins were analyzed by immunoblotting. (D) Schematic presentation of two cleavage sites, aspartate 245 and 254, and their serum response factor fragments, 32 kDa each, derived from cleavage by activated caspase 3. Cleaved SRF-N functioned as a dominant-negative transfector that contained an intact MADS-box, but did not retain the C-terminal transactivation domain. Therefore, cleaved N-SRF competes with fulllength serum response factor for binding to CArG boxes, and will not be able to transactivate serum response factor gene targets efficiently, including noncoding microRNA.

in Fig. 8. Phosphorylation of the Elk-1 B-box enhanced its association with serum response factor and through an adjacent ETS binding site to the CArG box, activated the promoters of immediate early target genes. There are two consensus CArGs located at positions -42 and -62 in the serum response factor promoter. In addition, there are two divergent CArG-like sequences located at positions -142 and -22. Potential binding sites for other immediate-early gene products, such as AP-1, Egr-1, and Ets-1 identified in the serum response factor promoter were also required for serum-induced promoter activity (Spencer and Misra, 1996; Belaguli et al., 1997; Spencer et al., 1999). These CArG boxes were targets for the growth factor-driven serum response factor-dependent ternary complex factor, Elk-1. Although serum response factor expression is autoregulated during cellular growth, leading to a positive feedback loop, this does not appear to be the primary mechanism for driving high levels of serum response factor-restricted gene activity in the embryonic heart. In addition, Elk-1 was shown to be a competitive inhibitor of myogenic co-factors such as myocardin and blocked the activation of smooth muscle gene products (Fig. 8) (Wang et al., 2004b).

VI.A.  Tbx Factors Regulate Serum Response Factor Gene Activity through its 3UTR Gene Enhancer Belaguli et al. (1997) showed that in addition to CArG boxes, the serum response factor promoter also has binding sites for the cardiac transcription factors GATA4 and Nkx2-5 (see Chapters 9.1 and 9.2). These three factors can transactivate cardiac specified genes (as will be discussed later); thus, it seemed reasonable to assume that cardiac regulation of serum response factor would be controlled by these transcription factors. To test this hypothesis, Barron et al. (2005) constructed serum response factor promoter deletion constructs, linked to b-galactosidase, and tested for LacZ-expression in transgenic mice (Fig. 9). Although serum response factor transgene activity was observed in several tissues (neural tube and somites), no reporter activity was detected in developing cardiac tissue. This observation is important for two reasons; first, it illustrates that the mechanism of serum response factor regulation in cardiac muscle differs from serum response factor regulatory mechanisms in other cell types, implying the existence of a cardiac-specific muscle enhancer for serum response factor.

626

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

SRF Growth factor targets

Cardiogenic SRF Regulation

T-Box factors

PKCα

ERK-1/2

MEK

ETS P SRF

immediate early genges

ERK-1/2

P SRF

SR F UT R

SRF

SRF Co-factors

Facilitates binding

weak CArG(s)

3’

TIP60

EBS

Myocyte growth

CArG(s)

transfactors sarcomeres ion channels

Cis-elements

miRNAs

Myocyte differentiation

Figure 8  Serum response factor gene activity directed by growth factor signaling and by cardiac-specific transfactors. (A) The first diagram corresponds to growth factor signaling in which the serum response factor gene is a downstream target for the ERK-signaling pathways that converge on ternary complex factors (TCF), such as the ETS domain transcription factor, Elk1. SREs are dependent on TCF Elk-1 and drive a positive feedback loop. In addition, growth factor stimulation leads to PKC-a phosphorylation of S162 in the MADS-box in a manner that suppressed the expression of muscle differentiation genes, while permitting and/or increasing the association between Ets factors and serum response factor to upregulate c-fos transcription. The MADS-box serves as a regulatory platform that mediates interactions between serum response factor and a myriad of transcription co-factors, hence differential phosphorylation of MADS-box residues directed specific serum response factor transcriptional complexes toward the expression of proliferation-dependent genes in myoblasts. (B) The second diagram shows cardiogenic-specific regulation of the serum response factor gene. Tbx2, Tbx5 and TIP60 were shown to strongly activate serum response factor activity (Barron et al., 2005). Also shown is the competitive inhibition of serum response factor–cardiogenic co-factor interactions by ETS factors. In the heart, serum response factor and cardiogenic co-factors facilitate serum response factor binding to promoters that drive the expression of other cardiogenic transcription factors, sarcomeric contractile proteins, ion channels and miRNAs.

Second, it demonstrated that even though the transgene contained functional, accessible binding sites for GATA, Nkx2-5 and SRF, these factors were not sufficient to drive cardiac expression of serum response factor. Serum response factor appeared to be a likely downstream gene target of the Tbx factors for the following reasons. First, many evolutionarily-conserved consensus T-box DNA sequences (Tada and Smith, 2001) were concentrated in the 3-UTR of the serum response factor gene. Second, the cardiac-enriched Tbx5 factor is known to physically interact with the cardiac factor Nkx2-5 (Hiroi et al., 2001), and serum response factor interacts with Nkx2-5 (Chen and Schwartz, 1996). Mutation or ablation of Tbx5 from the embryo causes Holt-Oram syndrome, a condition characterized by defects in limb and heart development (Bruneau et al., 1999; Basson et al., 2001; Ghosh et al., 2001). In addition, Tbx2, Tbx5 and serum response factor are co-expressed in overlapping patterns, indicative of potential co-regulation. Third, serum response factor transcripts were reduced by about half in the cardiac region of Tbx5 knockout mice, which display the Holt-Oram phenotype (Bruneau et al., 1999; Barron et al., 2005; Chapter 9.4). Remaining serum response factor transcripts

could be a result of redundant activation by other Tbx factors in the heart, such as Tbx2. Finally, the embryonic expression of SRF-LacZ in the tail presomitic mesenchyme shown in Fig. 2 was reminiscent of the embryonic pattern of expression of Tbx6 in somite precursor cells involved with the specification of paraxial mesoderm. Chapman and Papaioannou (1998) showed that mutated Tbx6 blocked the differentiation of paraxial mesoderm and the formation of posterior somites, but allowed differentiation along a neural pathway. Their study indicated that Tbx6 was needed for cells to choose between a muscle or a neuronal differentiation pathway during gastrulation. Barron et al. (2005) found that SRF5-flanking sequences, even up to 5.5 kb away from the transcription initiation site, failed to target transgene activity to the heart or the tail. They constructed transgenes that linked the promoter to the Tbx3-UTR region, which resulted in robust cardiac and tail expression of LacZ reporter gene activity, as shown in Fig. 9. They also found that the first of two T-box “clusters” in the 3-UTR (contained within the first 231 bp) was sufficient for transgene expression. Constructs containing 541 bp of the serum response factor promoter linked to the first T-box cluster were capable of driving

627

(e)

(C)

-5kB -1.0kB -0.5kB

NC

Somite

NC

+

-

+

-

-

-

LacZ

0/6

-

-

5/7

PCR +

-

-

β-gal +

-

-

NT

Somite

-

Tail

-

-

4/6

(g)

lll

NT

(D)

UTR

ll

-

4/4

LacZ

(f)

Exon l

-

-

(C)

LacZ

-

Vessel

-0.5KB

(B)

+

+

Tail

(B)

-

-

Heart

(A)

+

Heart

-.5Kb

+

-

β-gal+ PCR +

-1.0Kb

-

-

LacZ -3.5KB

+

5/5

LacZ

(A) -5.5KB

Vessel

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

-

(h)

lV

V

Vl Vll 3‘ UTR PolyA l

Somite enhancer Neural crest enhancers

Atrial/ventricular and tail enhancer

Figure 9  Serum response factor gene transcription in the embryonic heart requires its 5 promoter and 3UTR enhancer. (A) Schematic representation of the serum response factor promoter constructs were analyzed in F0 7.5–10.5 dpc embryos with a summary of their transgene expression activities in embryonic tissues. The “” or “” symbols indicate whether lacZ expression was detected in PCR-positive embryos. (a–d) Panels showed that the 5 flanking sequences from 5.5 kb to 0.54 Kb failed to drive reporter LacZ in the heart. (B) Early cardiac expression of serum response factor was directed by a downstream enhancer, located in the 3UTR of the serum response factor gene. Schematic diagram shows the transgene results of the linkage of 0.54 Kb of the 5 flanking sequence linked to the -galactosidase reporter and SRF 3UTR. (C) Serum response factor genetic regulatory regions were defined by their requirement for consistent transgene expression in F0 embryos. The “somite enhancer” lies within the promoter sequences containing the GATA, Nkx and serum response factor-binding sites, whereas the “A/V and Tail enhancer” lies downstream of the first polyA site, within the UTR T-box sequences, as adapted from Barron et al. (2005) (the following embryonic tissues M are marked: nt: neural tube; h: heart; m: myocardium; e: endocardium; a: allintois).

serum response factor expression in the heart. Also, Tbx5 bound several of the T-boxes located in the 3-UTR DNA. To strengthen this point, transgenic mice were generated using the same 3-UTR sequence, except with all the T-box sites knockedout. These 3UTR mutants failed to drive LacZ expression in transgenic mice (Fig. 9), demonstrating that functional T-box sites were required for cardiac expression. Barron et al. (2005) further showed that Tbx2, Tbx5 and the cardiac-enriched histone acetyltransferase Tatinteractive protein 60 (TIP60), were mutual interactive co-factors through their association with the TIP60 zincfinger and the T-box of the Tbx factors. These Tbx factors and TIP60 strongly activated enhancer serum response

factor promoter reporter gene activity in the presence of the SRF 3-UTR (Fig. 5). Notably Tbx2, along with Nkx2-5, represses expression of ANF (Habets et al., 2002). Competition was observed between Tbx2 and Tbx5 for a single half site in the ANF promoter, whereas Paxton et al. (2002) also found Tbx2 activated the promoter of a reporter gene containing five multimerized T-box elements, similar to the many T-boxes contained within the serum response factor 3-UTR sequences, and consistent with its activation by Tbx2. TIP60, a member of the MYST family of histone acetyltransferases (HATs) contains a heterochromatin-associated protein 1-like chromodomain and a zinc-finger-like domain. TIP60 is developmentally-regulated during early chick

628

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

embryogenesis, highlighted by the transient enrichment of TIP60 protein in developing myocardium (Lough, 2002). TIP60 HAT chromatin-remodeling complexes (Ikura et al., 2000) may support the activation of serum response factor in the heart. Tbx5 interacts with Tbx2 and with TIP60 through the zinc-finger-like domain. Mutant TIP60 proteins, one of which contained a mutation in its HAT domain (TIP60mHAT) and the other which lacked the zincfinger domain (TIP60Zn), inhibited the co-activation of serum response factor reporters with Tbx complexes and TIP60WT. In addition, TIP60 isoproteins also interact with serum response factor and together cooperatively activate the ANF promoter (Kim et al., 2006). Recently, zebrafish mutant lieberskummer (lik) inhibited an ATPase-complex involved with cell-autonomous proliferation of cardiomyocytes (Rottbauer, 2002). Lik is related to Reptin, a DNA helicase found in all organisms as a part of the TIP60 complex (Doyon et al., 2004). The function of TIP60 in regulating serum response factor may involve chromatin remodeling, modulating local Tbx gene activity during cardiogenesis (Fig. 8).

VII.  Identification of serum response factor gene targets In a genome-wide search for serum response factor targets, there are an estimated 1,200 genes that contain conserved serum response factor-binding sites, of which about 250 genes have been validated as bonafide serum response factor targets. An important goal is to determine how serum response factor selects among these myriad of potential targets to primarily express genes that direct tissue-restricted expression. In a recent study, Zhang et al. (2006) used Protein A-TEV-tagged chromatin immunoprecipitation technology to collect direct serum response factor-bound gene targets from P19 cells, induced to form an enriched cardiac cell population by Me2SO treatment. To identify direct serum response factor DNA-binding targets, over 16,000 bp of DNA were sequenced, revealing a great enrichment for serum response factor-binding sites, such as CArG and CArGL. These sequences also included an assortment of consensus binding sites such as NKE, (Nkx2-5 and other Nkx2 homeodomain proteins), GATA (GATA1-6, dual C4 zinc-finger protein), mTATA (muscle TATA box), E box (MyoD/b-HLH family, MyoD, Myf5 and myogenin), HNF1/4 (hepatic nuclear factor 1, 4 and 6), STATs (STAT3, -5 and -6), Smad (Smad1, and Smad4), Comp (co­operates with myogenic proteins in multicomponent complex), mTEF (muscle TEF), Ets (Ets family), NF-κB and YY1. The transfactor binding sites summarized in Table 1 appeared at a much greater frequency than by chance in the mouse genome, suggesting that these co-factors preferentially interact with serum response factor and may be important for influencing serum response factor activity.

Table 1  Enrichment of DNA-binding motifs from SRF and co-accessory factors from SRF-PATChIP pulldown Expected

Observed

Observed/ expected

CarG box

8

89

11

CarG-like box

12

66

5

NKE

12

65

6

Gata family

102

118

1

mTATA box

17

46

3

MyoD family

8

34

4

HNF1/4

7

35

5

STAT family

11

26

2

mTEF motif

1

10

10

Ets box

28

34

1

A total of 16,650 bp were sequenced from cloned 234 fragments pulled out by SRF-PATChIP from P19 cells converted to beating myocytes. Within these sequences we predicted the random appearance of the transcription DNA-binding sites shown above by Matinspector. Observed to expectation ratios indicated enrichment for most of the co-accessory transcription-binding sites. An exception was GATA-binding sites that are more numerous compared to other transfactors.

VII.A.  Serum Response Factor-Dependent Transactivation of DNA Targets Correlates Well With the Quality and Quantity of Serum Response Factor-Binding Sites Serum response factor-binding fragment-Hsp68-luciferase constructs co-transfected with the serum response factor expression vectors into CV-1 cells showed that most of the serum response factor-DNA-binding fragments were activated by 3–20-fold in the presence of co-expressed SRF. Activation of the serum response factor-binding fragments correlated well with the quality of CArG boxes, in that higher levels of serum response factor-induced transactivation were associated with consensus CArG sequences, rather than imperfect CArGL sequences (Fig. 10). Reporter gene activity was also influenced by multimeric CArG and CArGL boxes, indicating dependence of transfected gene activity on serum response factor. A high percentage of the fragments contained multiple CArG boxes and were located close to or within the promoters. Clearly, the activity of these serum response factor-dependent genes was influenced by the increased number and quality of consensus CArG boxes, as shown in Fig. 10.

629

SRF

CArG

Hsp68

Reporter gene

10

15

SRF

E2F5 Etv1 Npm1 Hifa3 Ophn1 Abca4 Lepr Scy1 Snx2 Ldr Calb1 ltga9 Mapk10 lpt1 Batt Pmfbp1 Amd1 Tera Azi2 Ctnnbl1 Spnb3 Map4k4 Cenpb Hess Asb5 Hspg2 Map3k14 Brd3 Txnl2 Bpgm Myl3 Opn3 Bicc1 Lrp4 Patah1b1 Rbbp6 Ski SerpinD1 Sema3a Ehox Actc Raf1 Vegfr3

0

5

Fold luciferase reporter activity

20

25

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

CArG CArG 1 1 1 2 2 2 2 3 3

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3 4 1 1 1 1 1 1 1 1 1 2 2 2

1 2 2 2

1 2

Figure 10  Serum response factor-dependent transactivation of DNA targets depends on the quality and quantity of serum response factor-binding sites. The blue bars are a relative measurement of luciferase activity in CV-1 cells co-transfected with serum response factor-binding fragment-Hsp68luciferase reporter genes and pCGN-SRF. Adapted from Zhang et al. (2005). Four measurements were performed in duplicate, and the mean  SE is shown as an error bar. Ns, 100 bp (in the first column), the numbers of binding motifs per 100 bp.

VII.B.  Serum Response Factor Target Genes Raf1, Map4k4 and Bicc1 Play Roles in Mesoderm Formation Murine embryonic stem cells afford a model system to investigate the function of serum response factor involved with the developmental appearance of serum response factor networks. Previously, c-fos and -actin genes, two known serum response factor target genes, were shown to be dramatically downregulated in serum response factornull mouse embryos (Arsenian et al., 1998) and in serum response factor-null embryonic stem cells (Fig. 11). We asked if serum response factor target gene activity was compromised by the absence of serum response factor in embryonic stem cells, and then defined a “regulated” serum response factor target gene as one that was either up- or downregulated by greater than 2-fold in serum response factor-null embyronic stem cells compared to wild-type embryonic stem cells (see Table 2). For example, serum response factor may activate Raf1, Mak4k4 and other genes to promote or permit mesoderm formation, and repress Bicc1 to prevent overgrowth of endoderm tissue. Raf1, a signaling molecule, is an important

component in the Ras/Raf/MAP signaling cascade in which serum response factor is a mediator (Treisman and Ammere, 1992). Cripto activates the Ras/Raf/MAP pathway in embryogenesis, and the lack of Cripto results in defective precardiac mesoderm unable to differentiate into functional cardiomyocytes (Minchiotti et al., 2002). Cripto was strongly upregulated (perhaps representing another serum response factor-dependent miR target), while Raf1 was severely repressed in serum response factor-null embryonic stem cells; indicating that the Raf1 downregulation results from serum response factor deficiency rather than from repression by Cripto. In mammalian cells, there are three highly-conserved Raf genes, Araf, Braf and Craf (Raf1). Two out three Raf genes are indispensable for life, as their disruption invariably results in embryonic perinatal lethality. The double Braf/;Craf/ mutants (disruption of all four copies of Braf and Craf) were present in the maternal decidua in small clumps of cells. Loss of three of four copies of Braf and Craf (Braf/;Craf/ or Braf/;Craf/) resulted in underdeveloped brain structure, heart and limbs (Xue et al., 2001). Most severely affected embryos were arrested at the early stages of gastrulation. Therefore, Raf1 gene

630

Days 0

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

2

Srf +/4 8 14

0

2

Srf -/4 8

14

H SRF Tgfβ Wnt5a Bmp4 Smad2 GATA2 GATA3 GATA4 GATA6 Nkx2-5 Myocardin Hhex Skα-actin Smα-actin Caα-actin Sm22α GAPDH

downregulation in serum response factor-null embryonic stem cells may be one of the primary causes of defective mesoderm formation. Most interestingly, the serum response factor-binding fragment located within intron-6 of the Raf1 gene also resided at a region 6 kb downstream of a zinc-finger protein-encoding gene, Makorin-2, which runs antisense in the orientation of the Raf-1 gene. In humans, both genes were co-expressed in the same tissues and cell lines, suggesting activated makorin may affect the expression of Raf-1 (Gray et al., 2001).

VII.C.  Serum Response Factor Target Genes Play an Inductive Role in Cardiovascular Development Both Wnt/-catenin and Wnt/JNK cascades play important roles in cardiovascular development (reviewed in Solloway and Harvey, 2003; Chapter 9.1). Among the serum response factor-regulated target genes, Mapk10 (JNK3), Txnl2, Tera and Azi2 were involved in the Wnt/JNK cascade. Mapk10 (JNK3) is expressed in the heart, the brain and the testis, whereas JNK1 and JNK2 are widely-expressed. Txnl2 (thioredoxin-like 2) plays a role in protecting the heart tissue from various types of stress in JNK cascade (Witte et al., 2000). Tera is a Wnt-antagonist identified recently and is expressed in neural tubes and somites (Aubert et al., 2002). We observed that Txnl2 and Mapk10 (JNK3) were downregulated, whereas Tera was upregulated in serum response factor-null embryonic stem cell differentiation, indicating that serum response factor may regulate these genes via cross-talk to other components of Wnt/JNK cascade to play a role in heart formation.

Figure 11  Expression of myocardin and myogenic and smooth muscle contractile protein genes is blocked in serum response factor-null embryonic stem cells. Serum response factor haploid (SRF/) and null-serum response factor (SRF/) embryonic stem cell lines were grown on a gelatinized plate in the presence of leukemia inhibitory factor. Embryoid bodies were made by the hanging drop method. Embryonic stem cells were allowed to aggregate in the hanging drop by gravity for four days in the incubator before transferring them onto gelatinized Petri dishes containing medium with 15% fetal bovine serum to induce differentiation. Embryoid bodies were harvested for RNA extraction and assayed by PCR reactions. This figure was adapted from Niu et al. (2005).

A number of genes expressed in serum response factornull embryonic stem cells were unaffected by the loss of serum response factor, such as the growth factor morphogens, Tgf2 and Wnt5a, and the transcription factors Nkx2–5, GATA3, GATA4, GATA5, GATA6 and Smad2 (Fig. 11). However, we observed reduced transcription of Hhex and the absence of myocardin, both factors important for cardiovascular development. Also, the lack of expression of cardiac, skeletal and smooth muscle -actins, and SM22  transcripts was consistent with the observation that serum response factor directs de novo cardiac and smooth muscle gene activity. Other serum response factor target genes were severely downregulated in serum response factor-null embryonic stem cells, such as Myl3 (myosin light chain, slow), Lrp4 (low density lipoprotein-related protein), Hspg2 (perlecan), Pgm2 (phosphoglucomutase), Hif3a (hypoxiainduced factor 3a), VEGFR3 (Flt4), Itgb9 (integrin 9) and Asb5 (ankyrin repeated and Soc box-containing protein) as shown in Table 2. Notably, Hspg2-null mutant embryos displayed cardiac defects during mouse development (Behar et al., 1996; Costell et al., 1999).

VIII.  Combinatorial interactions of serum response factor-accessory proteins Serum response factor plays an obligatory role in regulating post-replicative muscle gene expression. The promoters of virtually all vertebrate muscle and smooth muscle contractile protein genes contain CArG boxes. Site-directed mutagenesis of the CArGs of the avian striated α-actin promoters revealed that these CArGs act in combination

MGD

Name and description

Transcription factor and nuclear protein Hif3 Npm1 Cenpb Ski Brd3 E2F5 Ipf1 Hes6 Etv1 Ehox Rbbp6 Baft Ctnnbl1

Microarray data (Srf/ embryonic stem cells)

Fold changes of reporter genes

Expression (Unigene database*)

Accession number

Heart Heart Heart Heart Heart Heart

NM_016868 NM_008277 NM_007682 NM_011358 NM_023336 NM_007892 NM_008814 NM_019479 NM_007960 NM_021300 NM_011247 NM_016767 AK014880

D4

D8

hypoxia inducible factor alpha nucleophosmin1 centromere protein B Sloan-Kettering viral bromodomain bi-ing protein E2F5 transcription factor Insulin promoting factor 1 hairy and enhancer of split Ets transcriptor factor family ES derived homeobox retinoblastoma binding protein 6 transcription factor -catenin-like 1

            

            

5 3 4.8 9.1 8.8 4.3 6.1 7.4 4.2 15 10 5.7 9.3

bicaudal C1 V-raf-1 leukemia viral oncogene 1 NIK- or NCK-interacting kinase NF-κB-inducing kinase 5-azacytidine induced gene 2 thioredoxin-like 2 Map kinase 10, JNK3 N-terminal kinase-like semaphorin 3A convert pro-ANF to ANF Vegf receptor3, Flt4 ankyrin repeat and SOCs box-containing serine protease inhibitor, clade D integrin alpha 9 teratocarcinoma expressed serine rich Oligophrenin 1 platelet-activating factor acetylhydrolase Polyamine modulated factor 1 ATP-binding cassette 4 calbindin-28D opsin

                    

                    

18 11 7.3 6.8 5 9.5 6.5 5.9 15 10 23 6 10 7.5 6.1 3 8 7.8 7 4.9 8

Heart Heart Heart

Growth factor, receptor, ligand and signaling molecule Bicc1 Raf1 Map4k4 Map3k14 Azi 2 Txnl2 Mapk10 Scyl1 Sema3a Lrp4 Vegfr3 Asb5 Serpind1 Itga9 Tera Ophn1 Pafah1b1 Pmfbp1 Abca4 Calb1 Opn3

Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart Heart

NM_031397 AA990557 NM_008696 NM_016896 NM_013737 NM_023140 NM_009158 NM_023912 NM_009152 NM_016869 NM_008029 NM_029569 NM_008223 NM_133721 NM_019643 NM_053978 NM_013625 NM_019938 NM_007378 NM_009788 NM_010089

631

(Continued)

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

Table 2  SRF target gene activity is highly regulated in SRF-null embryonic stem cells

632

Table 2  (Continued) MGD

Name and description

Structural and metabolic protein

Microarray data (Srf/ embryonic stem cells) D4

D8

Fold changes of reporter genes

Expression (Unigene database*)

Accession number

myosin regulatory light chain





7

Heart

NM_010859

cardiac α actin





15

Heart

NM_009608

Spnb3

Spectrin β 3





5

Pgm2

Phosphoglucomutase





7

AF026489 Heart

NM_028132

Hspg2

perlecan (heparan sulfate proteoglycan2)





4.4

Heart

NM_008239

Amd1

S-adenosylmethionine decarboxylase 1





5.3

Heart

NM_009665

Snx2

Sorting nexin6





6.8

Heart

NM_026386

Ldlr

LDL receptor





5.9

Heart

NM_010700

Lepr

leptin receptor





7.4

Heart

NM_010704

The right column represents fold changes in gene expression during Srf-null embryonic stem cell differentiation. The triangle down indicates downregulated and the triangle up indicates upregulated by more than two-fold. The diamond indicates no change in activity. * Unigene was used to determine whether the target gene was expressed in the heart cDNA library and does not construe tissue-specificity.

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Myl3 Actc1

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

with each other in a highly-cooperative manner (Lee et al., 1991). In the case of the avian skeletal -actin gene promoter, reporter gene activity was stimulated by halfintegral numbers of helix turns when pairs of neighboring CArG boxes were brought into alignment in the promoter (Chow et al., 1991). Mutations that prevent serum response factor binding severely impaired the expression of these muscle-restricted promoters (Chow and Schwartz, 1990). De novo activation of muscle genes in the early specification of the cardiac lineage may require combinatorial interactions between transcription factors found to be enriched during the emergence of cardiac progenitor cells. Although serum response factor is indispensable for cardiac and vascular gene expression, serum response factor alone is not sufficient for regulating cardiogenic differentiation. This is supported by two facts. First, serum response factor alone is a weak activator of muscle gene promoters. Second, during embryonic development serum response factor is widely-expressed in other mesoderm-derived tissues, including skeletal, cardiac and smooth muscle, vascular endothelial and erythrogenic cell types. To resolve these paradoxes, we proposed that serum response factor associates with a compendium of positive and negative trans-factors (some of which are shown in Fig. 12) that probably influence its activity.

VIII.A.  Recruitment of the Tinman Homolog Nkx2-5 by Serum Response Factor-Activated Cardiac -Actin Gene Transcription Nkx2-5, a homeobox vertebrate homolog of Drosophila tinman (Azpiazu and Frasch, 1993; Bodmer, 1993) is one of the earliest markers of vertebrate heart development

αΙ

N

βΙ

βΙΙ

Homeodomain Nkx2-5 Zinc finger HOP GATA4/6 LIM only CRP1/2 Nkx2-5

GATA-4

LIM only CRP1/2

Myocardin

TN

TAD

+–+–+

TAD

LIM1 g

RP EL

αΙΙ

ETS Elk1, SAP1

ZF1

NK-SD

ZF2

Inhibitory domain

nls

LIM2 g

++

(Komuro and Izumo, 1993 (Lints et al., 1993; Chapter 9.1) and a serum response factor co-factor (Chen and Schwartz, 1996; Chen et al., 1996). Nkx2-5 prefers binding to imperfect CarG-like elements that resemble the NKE (CAAGTG) elucidated by Chen and Schwartz (1995). CArGs of the cardiac -actin promoter serve as both highaffinity (CArG2 and CArG3) and intermediate-strength binding targets (CArG1 and CArG4) of Nkx2-5 (Chen and Schwartz, 1996, 1997). Durocher et al. (1996) showed that two NKE sites on the proximal region of the ANF promoter directed high levels of cardiac-specific promoter activity. The ANF promoter contains two near-consensus NKEs that are each able to bind purified Nkx2-5 and are required for Nkx2-5 activation of the ANF promoter in heterologous cells. Interestingly, in primary cardiomyocyte cultures, the NKE contributed to ANF promoter activity in a chamber- and developmental stage-specific manner, confirming that Nkx2-5 and/or other related cardiac proteins play a role in chamber specification (Durocher et al., 1996, 1997). Gajewski et al. (1997) demonstrated that two NKE promoter sites direct D-mef2 expression in response to tinman. The recruitment of Nkx2-5 to a CArG box is dependent on serum response factor-DNA-binding activity. The dominant negative SRFpm1 mutant, which dimerizes with wild-type serum response factor, blocked the recruitment of Nkx2-5. Although Nkx2-5 can bind weakly to some CArGs, we found that the activation of a minimal promoter consisting of a single CArG box was dependent on increasing the cellular levels of serum response factor (Chen and Schwartz, 1996). When Nkx2-5-binding activity was blocked by a point mutation in the third helix of the homeodomain, serum response factor was still capable of recruiting mutated Nkx2-5 to the cardiac -actin promoter

SAP Myocardin

HD

Q B SAP

LZ

TAD

633

Figure 12  Cardiogenic enriched transcription factors Nkx2-5, GATA4, GATA6, CRP2 and myocardin support serum response factor myogenic activity and associate with the serum response factor MADS-box. The homeodomain factor, HOP, and the Ets factor, ELK-1, are competive inhibitors. Schematic diagram shows positive acting serum response factor co-factors and their regulatory domains. Nkx2-5 contains a conserved NK2 TN domain, a mixed charged activation domain, a homeodomain, an NK-specific domain and a highly hydrophobic inhibitory domain; GATA4 contains two transcription activation domains, two zinc-fingers and a nuclear localization sequence; CRP2, a dual LIM only factor, contains two LIM domains and two flexible glycine rich domains; myocardin contains an actin-binding and negative regulatory RPEL domain, a basic and glutamine rich domain that binds serum response factor and class II HDACs, a SAP domain, a leucine zipper and a transcription activation domain.

634

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

(Chen and Schwartz, 1996). Nkx2-5 bound to serum response factor in the absence of DNA as soluble protein complexes isolated from cardiac myocyte nuclei. A short 30-amino acid peptide (amino acids 142–171), which encompassed the N-terminal extension and the basic region of serum response factor in the 1 coil of the MADS-box, was sufficient to mediate protein–protein contacts with the Nkx2-5. The N-terminus/helix-I and helix-II regions of the Nkx2-5 homeodomain interacted with the MADS-box.

VIII.B.  GATA4 and Nkx2-5 Co-activate   Nkx2-5 DNA-Binding Targets Nkx2-5 also cooperates with GATA4 to co-activate cardiac -actin (Sepulveda et al., 1998, 2002) and ANF (Durocher et al., 1997) promoters containing Nkx2-5 DNA binding sites (Chapter 9.2). The Drosophila zinc-finger protein Pannier, a homolog of GATA4, also functions as a cardiogenic Tinman co-factor (Gajewski et al., 1999). Transcriptional activity requires the N-terminal activation domain of Nkx2-5 and binding activity through its homeodomain, but does not require the activation domain of GATA4. Minimal interactive regions were mapped to the homeodomain of Nkx2-5 and the second zinc-finger of GATA4. The removal of Nkx2-5 C-terminal inhibitory domain stimulates robust transcriptional activity compared with the effects of GATA4 on wild-type Nkx2-5, which in part facilitates Nkx2-5 DNA-binding activity. We postulated a simple model whereby GATA4 induces a conformational change in Nkx2-5 that displaces the C-terminal inhibitory domain, thus eliciting transcriptional activation of promoters containing Nkx2-5 DNA-binding targets (Sepulveda et al., 1998).

VIII.C.  Serum Response Factor and GATA4 are Mutual Co-Regulators Co-regulation of muscle gene expression by serum response factor and GATA4 is suggested by the robust activation of muscle-restricted promoters by co-expressing the two factors in fibroblasts (Belaguli et al., 2000; Chapter 9.2). The GATA4 second zinc-finger binds avidly to the I coil of the MADS-box. Deletion of N-terminal activation domains of GATA4 located between amino acids 1–74 and 130–177 does not affect the ability of GATA4 to co-activate serum response factor gene targets, suggesting that the activation domain of serum response factor may compensate for lack of activation domains on GATA4. Interestingly, deletion of the second N-terminal activation domain and the first zinc-finger of GATA4 increase the ability of GATA4 to synergize with serum response factor. The multizinc-finger proteins FOG1 and FOG2 modulate the transcriptional activity of GATA1 and GATA4 by interacting

with their first zinc-fingers (Tevosian et al., 1999). In a similar manner, Drosophila GATA protein, pannier, interacts with a zinc-finger protein called U-shaped (Fosset et al., 2000) that negatively-regulates the transcriptional activity of pannier towards the expression of proneural basic HLH proteins, achete and scute. Co-activation of the cardiac -actin promoter by serum response factor and GATA4 is mediated through the first CArG box (Belaguli et al., 2000). Deletion and specific point mutations of CArG1 reduced the basal activity of the cardiac a-actin promoter and eliminated the synergistic activation. Co-activation of the cardiac -actin promoter was strictly dependent on serum response factor binding to the most proximal CArG box, as deletion or point mutations that abolish DNA binding of serum response factor also abrogated synergistic activation. Co-activation appears to be independent of GATA4 DNA binding, since no GATA-binding sites were detected in the minimal cardiac -actin promoter (100 bp) responsive to the SRFGATA4 combination. Thus, GATA4 was recruited to the cardiac -actin promoter by serum response factor independently of GATA4 binding to DNA. Transcriptional activation of GATA binding site-dependent genes such as cTnC, ANF, BNP and troponin I required the N-terminal activation domains of GATA4. Furthermore, GATA5 and GATA6, which share extensive homology within the N-terminal activation domains, were also capable of activating these genes and substituting for GATA4. The exact overlap of serum response factor mRNA with Nkx2-5 and GATA4 transcripts was found to be coincident with cardiac -actin gene activity and the appearance of the myocardial cells in the linear heart (Sepulveda et al., 2002). Indeed, the co-expression of all three proteins serum response factor, Nkx2-5 and GATA4 resulted in robust synergistic activation of the cardiac -actin promoter about two orders of magnitude above base line (Fig. 13A). The combination of Nkx2-5 homeodomain and GATA4 enhanced the formation of serum response factor-dependent DNA-binding complexes. It is likely that conformational changes in serum response factor structure facilitated by Nkx2-5 and GATA4 made serum response factor a more efficient DNA-binding factor, allowing it to bind to weaker nonconsensus CArG boxes and stimulating cardiac -actin gene activity under limiting amounts (Fig. 13B). This level of activation required intact SRF-CArG binding, but was even higher when Nkx2-5pm was substituted for wild-type Nkx2-5. The lack of requirement for Nkx2-5 binding to DNA was also observed with Nkx2-5/SRF co-activation of the minimal -actin CArG1 promoter (Chen and Schwartz, 1996). This may indicate that Nkx2-5 acts by providing a strong transcriptional activation domain, whereas the role of serum response factor is to attract Nkx2-5 to the -actin promoter and facilitate the recruitment of GATA4. A schematic model of how serum response factor may interact with both Nkx2-5 and GATA4 is shown in Fig. 13C.

635

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

(A)

CArG Box 4

3 2

1

(C)

LUC reporter

SRF

αCA pormoter CArG1M CArG2M CArG2M

β loop

α1

CArG4M -200Del.

sh ee t

-150Del. -100Del.

80

β-

-100Del.+1M

α2

Transfectants C = control

60

1 = SRF Nkx2-5 2 = SRF Nkx2-5pm 3 = SRF pm Nkx2-5

GATA4 GATA4 GATA4

SRF + GATA4

40

20 Zn

(B)

C123 αCA

C123 C123 C123 C123 C123 C123 C123 C123 CArG1M CArG2M CArG3M CArG4M-200Del. -150Del. -100Del. -100 Del.+1M

+ SRF Nkx2-5 GATA4 -

+ + -

+ +

+

+

+

+

-

-

-

-

+ -

+ +

+ -

+ -

+

+

SRF + Nkx2-5 + GATA4

C3 SRF C1

C2 Zn

Nkx2-5 Nkx2-5 Figure 13  Multiple CArG boxes bestowed cooperative dependent and synergistic cardiac -actin promoter activity. (A) A schematic diagram of the cardiac -actin promoter with its CArG box mutations and serial deletion constructs, as previously described in Chen and Schwartz (1996). CArG boxes were indicated by closed boxes. Transfection analysis of various promoter mutant reporter constructs in CV1 fibroblasts. The luciferase activity of the full-length promoter in these cells was set at 1 and designated as control. Shown is co-transfection analysis of reporter constructs with serum response factor, Nkx2-5, and GATA4 expression vectors designated as condition 1, serum response factor plus defective DNA-binding mutant Nkx25pm plus GATA4 designated as condition 2, and defective serum response factor–DNA-binding mutant SRFpm plus Nkx2-5 plus GATA4 as designated condition 3. (B) Nkx2-5 and GATA4 recruited by serum response factor facilitated multimeric serum response factor binding. Adapted from Sepulveda et al., (2002). A suboptimal amount of wild-type serum response factor (complex1) recruited Nkx2-5 and GATA4 and formed cooperative serum response factor–DNA binding multimeric complexes (C2 and C3, arrowheads) that coincided with occupation of CArG boxes with serum response factor, Nkx2-5 and GATA4. Arrows indicate Nkx2-5 monomer and dimer. (C) Structural models for serum response factor, GATA4 and Nkx2-5 interactions with the CArG box. Serum response factor dimer binding to SRE through helix 1 in the major groove and N-terminal extension wrapping around to the minor groove caused DNA bending. The second zinc-finger of GATA4 bound the N-terminal extension and -coil I of the serum response factor MADS-box. The Nkx2-5 homeodomain through its helix I/II bound the serum response factor monomer and may also interact with GATA4 through helix III. Note that helix III may be positioned to bind the exposed major groove at the center of the CArG box. This model shows GATA4 binding to a serum response factor monomer and Nkx2-5 binding to the other serum response factor monomeric subunit.

VIII.D.  Competition between Negatively Acting YY1 versus Positively Acting Serum Response Factor Regulates α-Actin Promoter Activity A zinc-finger protein named Yin Yang 1 (YY1), a member of the GLI-Krüppel family, was shown to bind to a subset

of CArG boxes (Fig. 14A) (Gualberto et al., 1992; Lee et al., 1992). YY1 is a multifunctional transcription factor that can act as a transcriptional repressor, a transcriptional activator, or a transcriptional initiator (Shi et al., 1991). Previous reports have demonstrated a negative role for YY1 in regulating muscle gene expression through interaction with distinct nucleotides within the CArG box

636

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

(Gualberto et al., 1992; Lee et al., 1992, 1994; MacLellan et al., 1994; Galvagni et al., 1998). Transcriptional activation coincided with replacement of YY1 by serum response factor (Gualberto et al., 1992; Lee et al., 1992), whose interaction with the CArG box was required for musclespecific transcription. Thus, skeletal -actin promoter could be repressed or activated by two functionally-opposite SRE-binding proteins, YY1 and serum response factor, depending on the outcome of their competitive interactions with the most proximal CArG box (Fig. 14B,C). Serum response factor binding sites do not all share equivalent roles, e.g., subtle differences in sequences embedded in each CArG may influence binding of a host of transcription factors, such as YY1, serum response factor and Nkx2-5 (Chen and Schwartz, 1997). YY1 was found to repress the cardiac -actin promoter by binding to CArG2. By contrast, Nkx2-5 and serum response factor,

which appeared to compete with YY1 for binding to SRE2, were positive regulators of the promoter, and co-expression of both factors was able to overcome the inhibitory effect of YY1. Reduced YY1 binding to CArG2 by site-directed mutagenesis stimulated cardiac -actin promoter activity (Chen and Schwartz, 1997). Thus, YY1 is a preferentiallyenriched transcription factor in replicating myoblasts and nonmuscle cells that appears to be a transcriptional repressor responsible for inhibiting cardiac α-actin activity. Recently, the Ezh2 protein was shown to endow the Polycomb PRC2 and PRC3 complexes with histone lysine methyltransferase (HKMT) activity that is associated with transcriptional repression (Caretti et al., 2005). In undifferentiated myoblasts, endogenous Ezh2 was associated with the transcriptional regulator YY1. Both Ezh2 and YY1 were detected, with the deacetylase HDAC1, at genomic regions of silent muscle-specific genes. Their presence

(A)

(C) CArG3 YY1

CArG2 YY1

CArG1

20 Relative activity

CArG4

Cardiac α-actin CArG2 Chicken (–119) GGCCATGAATGGCG (–132) Human (–152) GTCCATGAATGGCC (–139) Mouse (–153) GTCCAAGAATGGCC (–140) Xenopus (–132) GTCCATTAATGGCT (–119) (–167) (–203) (–204) (–183)

YY1 consensus

TGCCTTAGATGGCC TTCCATACATGGTC TTCCTTACATGGTC TTCCATACATGGGC

10 5

Cardiac α-actin CArG3 Chicken Human Mouse Xenopus

15

(–132) (–139) (–140) (–119)

YY1 SRF Nkx2-5 Nkx2-5pm

-

+ +

+ -

+ + -

+ -

+ + -

+ + +

+ + + -

+ + -

AANATGGNC/G

(B)

(D) YY1 + + + + + SRF SRF YY1

SRF Dependent

YY1 Dependent H3 tail K4

)3 (Me

K9

H3 tail

K27

K9(Ac -)

HM

T

HDAC1 YY1

Repressed chromatin

GATA4

EZh2

K4(Me)3

SR F Nkx 2-5 SR F

K27

HAT

s

Activated chromatin

Figure 14  YY1, a competitive inhibitor of serum response factor, blocks striated α-actin gene activity. (A) DNA sequence comparison of vertebrate cardiac -actin CArG2 and CArG3 and the YY1 consensus sequence. Boldface letters indicate conserved YY1 binding sequence. The chick CArG2 is conserved in the reverse orientation in relation to the human, mouse and Xenopus CArG2. The chick, human and Xenopus CArG2 contain the halfYY1-binding site (ATGGN) in both orientations. (B) Mutually-exclusive DNA-binding activities of YY1 versus serum response factor on CArG2. Labeled CArG2 oligonucleotide duplexes were incubated with YY1-transfected 10T1/2 nuclear extracts in the presence of increasing amounts of purified bacterial purified serum response factor. (C) YY1 repression was overcome by co-transfected Nkx2-5 and serum response factor, as shown in Chen and Schwartz (1997). (D) Model of YY1 inhibiton of serum response factor-dependent myogenic gene expression. Regulatory regions of certain muscle-specific genes are occupied by a protein complex containing the DNA-binding protein YY1, the methyltransferase, Ezh2, and the deacetylase, HDAC1. Deacetylation of lysine residues by HDAC1 and di-trimethylation of H3-K27 by Ezh2 prevented transcription (repressed state). Adapted from Caretti et al. (2004). At the onset of transcriptional activation, YY1 is displaced from the chromatin with consequent loss of Ezh2 and HDAC1 and replaced by serum response factor. H3-K27 becomes hypomethylated, allowing for loading of the Nkx2-5, GATA transcription factors and an unidentified histone acetyltransferase (HATs) and histone methyltransferase may activate transcription.

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

correlated with histone H3 methylation at lysine K27. YY1 was required for Ezh2 binding because RNA interference of YY1 abrogated chromatin recruitment of Ezh2 and prevented H3-K27 methylation. On gene activation, Ezh2, HDAC1 and YY1 dissociated from muscle gene loci, histone H3-K27 became hypomethylated and serum response factor was recruited to the chromatin. These findings suggest a two-step activation mechanism, whereby removal of H3-K27 methylation, originally conferred by an active Ezh2-containing protein complex, and followed by recruitment of positive transcriptional regulator, is likely to be required to promote muscle gene expression and cell differentiation (Fig. 14D) (Caretti et al., 2005).

IX.  Hop, a homeobox protein enriched in the heart, inhibits serum response factor myogenic activity The Hop gene encodes a 73 amino acid protein that contains a domain (the 60 amino acid homeodomain) homologous to those seen in homeobox (Hox) transcription factors (Chen et al., 2002; Shin et al., 2002). Unlike Hox homeodomains, however, Hop lacks certain conserved amino acid residues required for protein–DNA interactions, and Hop is also unable to bind DNA. Nevertheless, Hop is a nuclear protein that can function to modulate transcription. Hop physically associates with serum response factor and inhibits serum response factor-dependent cardiac-specific gene activity. For instance, forced expression of serum response factor and myocardin resulting in transactivation of SM 22aANF promoter activity was blocked by Hop, and in Hop-deficient hearts, serum response factor-dependent genes were upregulated. Hop is expressed by cardiac myocytes during gestation and in the adult. In the absence of Hop, neonatal mice display enhanced cardiac myocyte proliferation. Hence, Hop is thought to modulate the delicate balance between cardiac replication and differentiation.

X.  Cysteine-rich protein lim factors bridge serum response factor with gata6 and activate smooth muscle genes The members of the cysteine-rich protein (CRP) family have been implicated in regulatory processes linked to cell growth, proliferation and differentiation, particularly of heart and muscle cells (Weiskirchen et al., 1995). Each member of the CRP family contains two LIM domains with associated glycine-rich repeats. The structural independence and spatial separation of these two motifs support

637

the notion that CRP proteins may act as scaffolds that link different protein partners to form higher-order functional complexes (Arber and Caroni, 1996; Konrat et al., 1998). All three members of the CRP family are expressed in early cardiovascular development (Arber et al., 1997; Jain et al., 1996; Henderson et al., 1999). A distinguishing feature of the three CRP proteins is that they shuttle between the cytoplasm and nucleus (Nix and Beckerle, 1997). Both LIM-only proteins (including the CRPs) and GATA transcription factors contain similar zinc-finger motifs that could potentially heterodimerize. For example, LMO2, a LIM-only protein, was demonstrated as an associating molecule with GATA1 in transcriptional complex assembly important for erythropoiesis (Wadman et al., 1997; Mead, et al., 2001). This suggested the possibility that LIM-only CRP1 or CRP2 may interact with GATA family members enriched in the cardiovascular system, such as GATA4 and GATA6. Previously, synergism was shown between SRF and GATA4, GATA5, and GATA6 in mediating the activation of SRF-dependent gene promoters, including the SMC marker genes SMA, SM22, and SM-actin (Belaguli et al., 2000). The SM22 promoter, which contains two CArG boxes, was activated about 600-fold by the combination of these three factors, while the 549 Calponin-I reporter (Miano et al., 2000), with three intronic SREs, and the smooth muscle -actin SMGA5 promoter, with four CArGs (Broening et al., 1997), were activated about 250- and 300fold, respectively (Fig. 15). In contrast, weak co-activation was seen with the single-SRE-containing c-fos promoter. Furthermore, mutation in either CArG box on the SMA promoter (Blank et al., 1992) reduced the synergistic transactivation by these transcription factors (Chang et al., 2003). In the double-SRE mutant reporter, the responsiveness to CRP2 was completely abolished. These results suggested that most of the synergistic coactivation by CRP2 requires efficient binding of serum response factor to multiple CArGs. Consistent with this idea, substitution of the transactivation-defective SRFC mutant (Belaguli et al., 2000) for wild-type serum response factor in combination with GATA6 and CRP2 resulted in a greater than 95% reduction in coactivation of these multi-CArG-containing smooth muscle gene promoters. CRP1 and CRP2, expressed during cardiovascular development, act as bridging molecules that associate with serum response factor and GATA proteins (Fig. 15) (Chang et al., 2003). SRF-CRP-GATA complexes strongly activated smooth muscle gene targets, in part by strongly facilitating the DNA-binding activity of serum response factor. The combination of GATA6 and CRP2 greatly increased serum response factor-DNA-binding affinity by two to three orders of magnitude (Chang et al., 2003). CRP2, serum response factor and GATA6 can physically associate in cells, and were co-immunoprecipitated with each other. Physical association and complementary

638

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

(A)

CRP1

HH7

(D)

CRP2

HH10

HH7

HH12

HH8

Keel of Pharynx

Ectoderm

Ventricle Epimyocardium

CRP2 + SRF + GATA6 SMA

SMA

SM22α

SM22α

h-Caldesmon

h-Caldesmon

Calponin

Calponin

SM-MHC

SM-MHC

HH12

Pharynx

Endoderm Endoderdium

Control

Dorsal aorta (L)

Dorsal aorta (R)

(B) CRP2

(C) -191

1,200

-1443 Luc

-549

SMA-Luc

Luc

SM22-Luc

+1923

(E)

-549CalpLuc-l

4

5 + +

+ +

GA TA 6

6

7 +

+ +

+

8 + + +

9 + + +

MAD

S

1

3

+

LIM

2

DS

-

P2

1 SRF GATA6 CRP2 SRF∆C

ZF2

0

CR

SMGA5-Luc

MA

Luc

-285

400

LIM

2

800

Luc

ZF1

Fold-activation

1,600

SRF

Figure 15  Cysteine-rich LIM-only proteins CRP1 and CRP2 are potent smooth muscle differentiation co-factors. (A) Whole-mount in situ hybrid­ ization showed CRP1 transcription in the lateral plate precardiac mesoderm (HH7) heart tubes (HH10) and looped tube (HH12). Cross-section of the cardiogenic region at HH10 showed CRP1 expression in the foregut endoderm and the epimyocardium (lower panel). CRP2 is expressed in the region of the developing head fold and outflow tract (HH7 and HH8), left and right dorsal aorta, truncus and bulbus arteriosus (HH12). The cross-section of HH12 embryo (lower panel) showed CRP2 expression to some degree of ventricular overlap (the bulboventricular region) and in a discrete area of expression at the junction of the vitelline veins and in the sinus venosus. (B) Endogenous CRP2 was localized to the nucleus of epicardial cells during the period of initial activation of SMC-specific gene transcription (i) and localized along actin filaments on epithelial-to-mesenchymal transformation (ii) and progression to mature SMCs (iii). Combinatorial interactions among CRPs, serum response factor and GATA factors facilitated robust activation of SMC Target Genes. (C) Transactivation assay of CArG box-containing SMC-restricted genes in CV1 cells. (D) Immunofluorescent staining showed expression of the full complement of SMC markers in 10T1/2 cells that receive the triple transfected combination of factors, compared with the control group. (E) The N-terminal LIM domain mediated serum response factor–CRP2-binding, while the C-terminal LIM was required for CRP2GATA4 co-association. This composite figure was adapted from Chang et al. (2003).

co-transfection analyses identified specific protein–protein interaction domains that are responsible for CRP2-SRF and CRP2-GATA binding. The two LIM domains of CRP2 independently co-associated with serum response factor (N-terminal LIM) and with GATA protein (C-terminal LIM), respectively. The C-terminal LIM motif is necessary for CRP2’s interaction with GATA4 (Fig. 15) (Chang et al., 2003). Intact LIM motif is obligatory for its function as a protein–protein interface; “half-LIM” is not equivalent to a functional zinc-finger motif, which was known for mediating protein–protein and protein–DNA interaction. Second, the ability of each LIM domain of CRP2

to function ­ independently, with the N-terminal and Cterminal LIM co-associated with serum response factor and GATA protein affirms its adaptor role, as previous studies have suggested (Konrat et al., 1998). Lastly, the Nterminal LIM domain of CRP2 has an important functional role, required for synergistic transactivation. On the basis of the observation that synergistic interaction among CRP2, serum response factor and GATA factors produced robust promoter activity, the combined effects of these transcription factors were tested in murine pluripotent 10T1/2 mesenchymal cells (Chang et al., 2003). Within two days of co-transfection with CMV-driven

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

expression plasmids encoding serum response factor, GATA6 in combination with CRP2 strongly induced the expression of SMA, SM22, SM-calponin, SM-MHC and SMA were detected by RT-PCR, immunofluorescence and protein blot analyses (Fig.  15). Muscle-specific sarcomeric -actins were not induced, indicating that the differentiation pattern was uniquely SMC-specific. Together, expression of CRP2 with serum response factor and GATA factors activated an SMC lineage-specific gene program in 10T1/2 fibroblasts. Recent studies by Chang et al. (2007) found nuclear CRP2 complexed with DNA-binding transcription activators and chromatin-modifying factors. Moreover, they discovered a novel property of the CRP2 LIM domain to direct chromatin modification that might generate a DNA template more accessible to the general transcription machinery. Two major classes of complexes, ATP-dependent SWI/SNF remodeling complexes and histone-modifying HAT or HDAC complexes, regulate accessibility of the nucleosomal DNA to binding factors. CRP2 interacted with Brg1-containing SWI/SNF chromatin remodeling complexes. The SWI/ SNF complexes, which use ATP hydrolysis to expose nucleosomal DNA sequences to trans-factors, are targeted to activated SMC gene loci in the transgenic cardiac myocytes. In addition, using specific antibodies to acetylated histones, chromatin immunoprecipitation assay (ChIP) studies revealed that the increased acetylation in proximal promoter regions of upregulated SMC genes correlated with gene activation and the loss of histone deacetylase HDAC1. Thus, CRP2 exerts its potent differentiation role through a combinatorial interaction with serum response factor, involving DNA-binding transactivators, chromatinremodeling and histone modifiers.

XI.  Serum response factor   co-activator myocardin is required for vascular smooth muscle development Myocardin, a member of the SAP family of nuclear proteins, shares homology with myocardin-related transcription factors (Mercher et al., 2001; Wang et al., 2002) which are expressed in a variety of embryonic and adult tissues. Each member of the MRTF family contains a conserved SAP domain that binds to A/T-rich genomic sequences or scaffold attachment regions (SARs) which play a role in high-order transcriptional regulation and chromatin remodeling. The N-terminal regions of myocardin and MRTFs contain three RPEL motifs, which mediate association of MRTFs with actin, thereby providing responsiveness to cytoskeletal signaling (Miralles et al., 2003). Association of myocardin with serum response factor is mediated by a short peptide sequence that includes a basic and glutamine-rich region (Wang et al., 2001, 2002).

639

A coiled-coil motif resembling a leucine zipper mediates homo- and heterodimerization of myocardin and MRTFs, and has been proposed to contribute to the cooperativity between CArG boxes in serum response factor-dependent muscle genes (Wang et al., 2001, 2002; Miralles et al., 2003; Du et al., 2004). The C-terminal regions of myocardin and MRTFs are somewhat divergent in amino acid sequence, and function as transcription activation domains (TADs). Deletion of these regions generates dominantnegative mutants. The transcription activation domains can be replaced by heterologous transcription activation domains, such as that from the viral coactivator VP16, and function normally in reporter activation in vitro, indicating that this domain serves a general function in transcriptional activation, but does not contribute to the specificity of these factors for serum response factor co-activation (Wang et al., 2001). Myocardin recruits chromatin remodeling enzymes to serum response factor target genes. Association with the histone acetyltransferase p300 enhances an interaction with class II histone deacetylases and represses expression of serum response factor target genes (Cao et al., 2005). In support of a role for myocardin in altering chromatin structure, myocardin–serum response factor complexes have been shown to associate with a specific variant of histone H3 on SMC gene loci in vivo (McDonald et al., 2006). Olson and colleagues suggested that the serum response factor-binding regions of myocardin and MRTFs resemble the predicted secondary structure of the B-box, although they lack direct amino acid homology with this region of Elk-1 (Wang et al., 2004). Deletion of this region of myocardin abolishes its ability to associate with serum response factor and activate serum response factordependent genes, and replacement of this domain with the Elk-1 B-box restores these functions (Wang et al., 2004). Myocardin and Elk-1 compete for interaction with a common docking site in the MADS-box of serum response factor (Miralles et al., 2003; Wang et al., 2004). Olson noted that their mutual exclusive association with this site creates a binary switch in which growth signals can modulate the phenotype of SMCs. When SMCs are stimulated with PDGF, Elk-1 becomes phosphorylated by the MAP kinase signaling pathway, facilitating its association with serum response factor and favoring the displacement of myocardin (Wang et al., 2004). Although Elk-1 is a co-activator of serum response factor, its activity is substantially weaker than that of myocardin, such that the displacement of myocardin from serum response factor by Elk-1 results in an overall decrease in the expression of smooth muscle genes; perhaps made easier by the lack of Ets binding sites on most smooth-muscle-specified gene promoters. Consistent with this model, lowering endogenous levels of Elk-1 in SMCs results in an increase in expression of a subset of smooth muscle genes, likely via derepression of serum response factor–myocardin complexes on the promoters of these genes (Zhou et al., 2005).

640

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

GATA factors are capable of repressing myocardin’s co-activation of serum response factor by competing for a common docking site on serum response factor, although this effect is only observed on a subset of smooth muscle enhancer elements. On the CArG-dependent enhancer of the smooth muscle myosin heavy chain gene, GATA6 can synergize with myocardin and serum response factor to co-activate high levels of transcription, highlighting the context-dependent activity of myocardin (Yin and Herring, 2005). Similarly, GATA4 binds directly to myocardin and can synergize with myocardin to activate transcription from the cardiac-specific Nkx2-5 enhancer in a CArG-boxdependent manner, yet it prevents activation of the cardiac ANF promoter by myocardin (Oh et al., 2004).

XII.  Serum response factor mutants block sarcomerogenesis in serum response factor-null embryonic stem cells The idea that serum response factor activity is largely controlled by its interaction with co-factors was tested by a gain-of-function approach applied to serum response factor-null embryonic stem cells. We focused on two mutants: SRF-145 (triple alanine substitutions 145aa-147aa); and SRF-194 (dual alanine substitutions at 194aa and 196aa) (Fig. 16A). These two mutants did not interfere with DNA binding, but prevented their co-association and facilitated binding of serum response factor with Nkx25/GATA4 (SRF-145 mutant), or myocardin (SRF-194 mutant). Increased expression of wild-type serum response factor and mutants by the lentivirus system was sufficient to either rescue completely, and/or in part, the cardiac myogenesis defect of serum response factor-null embryonic stem cells (Fig. 16B–D). Microarray analysis and quantitative RT-PCR confirmed the rescue and enhanced expression of Hand1 and Smyd1, transcription and chromatin remodeling factors, Acta1, Acta2, Myl3 and Myom1, myofibril proteins, and Kcnmb1 excitation contraction coupling proteins by serum response factor replenishment in serum response factor-null embryonic stem cells (Fig. 16B,C). Measurement of GATA4, myocardin and MRTF transcripts indicated that there were sufficient amounts of these factors to allow for gene activation in the presence of the serum response factor mutants. Moderate elevation of some structural gene expression was elicited by either SRF-145 or SRF-194, but neither mutant alone was sufficient to drive terminal cardiac myocyte differentiation, as determined by immunofluorescent staining of organized sarcomeres, microarray analysis and beating cardiac myocytes (Fig. 16D,E). Given the fact that wild-type serum response factor directed the appearance of beating myocytes from serum response factor-null ­ embryonic

stem cell cultures, the inability of the serum response factor interaction-defective mutants to rescue myogenesis highlights the significance of these co-factor associations in cardiac myogenesis and not their absence. Because serum response factor mutants were incapable of restoring sarcomerogenesis, our data supports the claim that serum response factor plays an obligatory role in cardiac myogenesis with a requirement for serum response factor cofactor interactions.

XIII.  Post-translational modification of serum response factor and co-factors are important regulatory switches XIII.A.  Myocardin Sumoylation Transactivates Cardiogenic Genes Small ubiquitin-like modifiers (SUMOs) are implicated in numerous physiological and pathological processes through altering the function of its target proteins. SUMO conjugation is part of an enzymatic cascade basic­ ally involving two enzymes, heterodimer E1-activating enzyme (SAE1/2) and E2-conjugating enzyme (Ubc9). Unlike ubiquitination, some sumoylation assays revealed that in the presence of E1 and E2, the E3 ligase was dispensable to accomplish SUMO conjugation. However, SUMO E3 ligases such as PIAS1 contributed to the efficiency and specificity of SUMO conjugation (Liang et al., 2004; Wang et al., 2004), and was attributed to the RING domain, which is similar to the structure of E3 ligases involved in protein ubiquitination. Recently, Wang et al. (2004) found that GATA4 was modified by SUMO-1 both in vivo and in vitro, with E3 ligase PIAS1 strongly enhancing the sumoylation efficiency. SUMO modification resulted in enhanced GATA4 transcriptional activation, and strongly induced the transcriptional activity of endogenous cardiac-specific genes in pluripotent 10T1/2 fibroblast cells. In addition, serum response factor, a chief co-accessory factor of myocardin and GATA4, was shown to be a SUMO target (Matsuzaki et al., 2003). Since myocardin, serum response factor and GATA factors are co-interactive and enriched in the heart, we asked if myocardin might also be a SUMO target. In fact, bioinformatics revealed a potential SUMO modification consensus sequence in myocardin, suggesting that myocardin could be sumoylated with potential consequences for its activity. Myocardin, a powerful regulator of smooth muscle specified genes in concert with serum response factor, was not sufficient to activate the endogenous cardiogenic program in pluripotent 10T1/2 fibroblast cells. Wang et al. (2007) found myocardin’s activity was strongly enhanced by SUMO-1 via modification of a lysine residue primarily

641

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

(A)

(B) αl

N-extension 140

βl loop βll 190

150

Lentivirus Srf –/– ES cells Null SRF145 SRF 194 SRF rescue Srf Myom1 Myl3 Myl9 Cnn1 Tnnc2 Tagin Sema3C Acta2 Actg2 Acta1 Tgfb1/1 Hand1 Smyd1

200

SFTGHVYTFATR SRF TRGRVKIKMEF SRF-145 TRGRVAAAMEF SETGHAYAFATR

(C)

αll

SRF SRF-194

Relative activity

1.2 Srf–/– Srf-145

0.6

Srf-194 Srf wt

at

a4

–2.5 –2 –1 0 1 2 2.5 Relative expression

G

N

kx

2.

5

1 Sm yd

d1 H

an

6 yh M

tc Ac

2 ta Ac

G

ap

dh

1.0

(D)

(E) Srf-145

Srf-194

antiα-actin

Srf wt

80 EB beating %

Srf –/–

60

SRF null SRF wt SRF-145 SRF-194

40 20 0 7

8

9

10 days

Figure 16  Rescue of cardiac myogenic differentiation defect in Srf-null embryonic stem cells by serum response factor wild-type and interaction defective mutants revealed an obligatory role for serum response factor and co-factor interactions to drive sarcomerogenesis and contractility. (A) Serum response factor contains an evolutionary-conserved MADS-box domain that allows for dimerization of serum response factor monomers and DNA-binding to a CArG element. This diagram shows the N-extension, I coil, I sheet, -loop, II sheet and II coil of the serum response factorDNA crystal structure adapted from Pellegrini et al. (1995). Alanine substitution mutations were made in the MADS-box at aa145–147 (SRF-145) and at aa194 and aa196 (SRF-194). (B) Affymetrix expression microarray showed a subset of genes that were consistently affected by the lentiviral rescue of serum response factor-null embryonic stem cells with wild-type serum response factor during embryoid body formation leading to cardiac cell differentiation (C) Quantitative RT-PCR showing the effects of the rescue of Srf-null embryonic stem cells with lentiviruses expressing serum response factor wild-type and serum response factor point mutants on the expression of serum response factor targets, which included Acta2, Actc1, Myh6, Hand1, Smyd1 and cardiac regulatory factors Nkx2-5 and Gata4. (D) Immunofluorescent detection of sarcomeric -actin following lentivirus-mediated infection and expression of serum response factor and point mutants SRF-145 and SRF-194 in Srf / embryonic stem cells. (E) Beating embryoid bodies were found only in Srf-null embryonic stem cells rescued by lentiviral infection with serum response factor wild-type. Data adapted from Niu et al. (2008).

located at position 445, and that the conversion of this residue to arginine (K445R) impaired myocardin transactivation. PIAS1 was involved in governing myocardin activity via its E3 ligase action that stimulated myocardin sumoylation, and by its physical association with myocardin. Myocardin was unable to induce the expression of cardiac-specific gene tested in fibroblasts, however, in the presence of SUMO-1/PIAS1, myocardin was sufficient to strongly induce the expression of cardiac -actin and cardiac -MHC (Fig. 17) (Wang et al., 2007). In contrast, neither K445R together with SUMO-1, nor myocadin with SUMO-1GG triggered cardiac gene activation (lanes 8 and 9). Sarcomeric -MHC induction by myocardin via

SUMO-1/PIAS1 was confirmed by immunostaining performed on 10T1/2 cells using MF20 that specifically detected sarcomeric -MHC (Fig. 17). Endogenous cardiac gene activation data also revealed that SUMO-1 enhanced myocardin activity more efficiently than SUMO-2/3 (Fig. 17). Taken together, these data argue that SUMO members modulate the function of myocardin differentially, and that sumoylation converted myocardin into a potent transactivator of cardiac gene expression. Myocardin initiated the expression of cardiac muscle-specified genes, such as those encoding cardiac -actin and -myosin heavy chain, in a serum response factor-dependent manner in 10T1/2 fibroblasts, but only in the presence of co-expressed

642

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

(A)

(B) Ca-α-MHC

Dapi

Anti- V5

Anti-α-MHC

merge

Ca-α-Actin Cont. SM-calponin

% of a-MHC positive cells among 100 v5 positive cells

3

4

5

6

7

8

9

Pias1+SUMO-1

myocardin

Myocardin+SUMO-1

Myocardin+Pias1

Myocardin+SUMO-1+Pias1

K445R

K445R+SUMO-1

Myocardin+SUMO-1 ∆GG

10 11 K445R+PlAS1

2

myocardin

K445R+SUMO-1+Pias1

1 Cont

Heart

GAPDH

Pias1+ SUMO-1

SUMO-1+ myocardin

80 60

Pias1+

40

myocardin

20

SUMO-1 +

0

myocardin

-

+

-

+

+

+

Pias 1+

SUMO-1

-

-

-

+

-

+

myocardin

PIAS1 + + Figure 17  Myocardin-induced cardiac-specific gene expression in fibroblast 10T1/2 cells in the presence of SUMO-1/PIAS1. (A) RT-PCR was performed on total RNA from 10T1/2 cells transfected with specified expression vectors alone or in various combinations as indicated. (B) Sarcomeric -myosin heavy chain (-MHC) was detected in 10T1/2 cells transfected with SUMO-1 and or PIAS1, as indicated by using MF20 antibody in immuno­fluorescent staining assays. V5 antibody was used to detect myocardin-positive cells. Data adapted from Wang et al. (2006).

SUMO-1/PIAS1. Thus, SUMO modification acted as a molecular switch to promote myocardin’s role in cardiogenic gene expression. Reversible SUMO-conjugation and de-conjugation regulates the activities of a large pool of proteins controlling cellular proliferation and differentiation. For example, mice with knockdowns of sumoylation pathway components, such as Ubc9, died at the early embryonic postimplantation stage (Hayashi et al., 2002; Nacerddine et al., 2005). More recently, Ubc9 was shown to be required for myotube formation in C2C12 cells and pharyngeal muscle development in C. elegans (Riquelme et al., 2006; Roy, 2006); thus, implicating the sumoylation pathway in muscle development. SUMO-modification of GATA4 elicited cardiac specific genes expression and myocardin sumoylation by SUMO-1/PIAS1 showed induced cardiogenic gene expression. The fact that transcription factors such as myocardin, serum response factor and GATA4 are all SUMOtargeted and physically interact with each other, and that all of them are crucial to heart development, collectively point to the possibility that sumoylation pathway may contribute

significantly to heart development via modification of heart-enriched transcription factors, as well as co-factors.

XIII.B.  Role of Histone Deacetylases (HDACs) and Histone Acetyl-transferases (HATs) in Serum Response Factor-Dependent Muscle Gene Activity HDACs and HATs are dueling factors involved in transcription silencing and activation (Roth et al., 2001; Chapter 10.1). Hypoacetylation by HDAC leads to transcriptional silencing and class II HDACs, which include HDAC4, -5, -7 and -9, exhibit tissue-specific expression. HDAC4 and HDAC5 are expressed at high levels in the heart, skeletal muscle and brain, where they have been suggested to be involved in cell differentiation and development (McKinsey et al., 2000; Chapter 10.2), and were shown to associate with serum response factor, resulting in repression of serum response factor gene activation. In addition, SRF/HDAC4 association is a target for upstream signaling events that increase ­intracellular Ca2 levels and

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

activate CaMK signaling during hypertrophic growth of myocytes (Davis et al., 2003). Histone acetylation catalyzed by HATs promotes gene transcription by destabilizing chromatin structure and facilitating access of transcriptional complexes to their target genes. During differentiation of P19 cells to smooth muscle cells, serum response factor was shown to be recruited to target genes, along with hyperacetylation of histones at CArG elements (Manabe and Owens, 2001). Similarly, serum response factor and the HAT-containing co-activator, cAMP-response element-binding protein (CREB)-binding protein, were recruited to the CArG box of the SM22 promoter during gene activation (Qiu and Li, 2002). Serum response factor also physically interacted with the HATcontaining activator, p300 (Montaner et al., 1999), and myocardin was shown to partner with p300 (Cao et al., 2005). The ubiquitous expression of p300 with an elevated abundance in neural tissues indicates that this protein is not cardiovascular-specific. Indeed, CBP and p300 are known to interact with a multitude of other proteins, including nuclear hormone receptors, c-Jun, c-Myb, c-fos, MyoD and E2F, in addition to serum response factor and/or myocardin (Goodman and Smolik, 2000). Disruption of the murine genes encoding p300 or CBP leads to early embryonic lethality (Xu et al., 2000; Yamauchi et al., 2000; Yao et al., 1998) and defective organogenesis in multiple organ systems including the heart, but did not interfere with the beginning of cardiovascular-specified gene expression.

XIII.C.  Serum Response Factor MADS Box Serine-162 Phosphorylation Switches Proliferation and Myogenic Gene Programs In growth factor-stimulated replicating myoblasts, serum response factor recruits proteins containing an Ets domain (Elk-1, Sap1 and Fli) and forms ternary complexes at the c-fos promoter (Treisman, 1994; Ling et al., 1997). For the most part, Ets factor-associated ternary complexes inhibit transcription of serum response factor-dependent myogenic gene targets, a theme mentioned several times throughout this chapter. Instead, on serum withdrawal, serum response factor associates with many of the co-accessory transfactors described within, such as Nkx2-5, GATA4, CRP2 and myocardin, to activate transcription of serum response factor myogenic target genes, but does so only weakly in replicating myoblasts. What are the triggers that switch serum response factor from modulating the proliferation pathway to modulating the differentiation pathway? Because serum response factor isolated from cell extracts is highly phosphorylated in numerous sites (Iyer et al., 2003) it was likely that site-specific phosphorylation was a mechanism for regulating the function of serum response factor. Indeed, Iyer et al. (2006) demonstrated that phosphorylation of serine-162 in the serum response factor MADS-box

643

I coil is a key determinant for switching serum response factor-dependent gene expression between replication and differentiation of myogenic cell types. As modeled in Fig. 18, the MADS-box serves as a regulatory nexus that mediates interactions between serum response factor and a myriad of transcription co-factors. Hence, differential phosphorylation of MADS-box residues directs specific serum response factor transcriptional complexes toward the expression of either proliferation- or differentiationrelated genes. Most serum response factor accessory proteins that have been identified as co-regulators of c-fos induction are at endpoints of growth factor-induced signal transduction cascades. Data from X-ray crystallographic analysis of the serum response factor MADS-box bound to CArG box directed the attention of Iyer et al. (2006) to two potential phosphorylation sites (Threonine 159 and Serine 162) closely grouped on the I coil of the MADSbox and well-conserved across evolution. A monomer of serum response factor at S162 contacts T8 and A9 on the C strand, and phosphorylation of S162, was predicted to impede DNA-binding directly by phosphate–phosphate repulsion and/or steric hindrance. Iyer et al. (2006) also noted that T159 did not make a significant contact on the CarG box; hence they focused on the role of S162 phosphorylation. Phosphorylation of this residue on the  I coil was mimicked by substitution of S162 with aspartate, and absence of phosphorylation was achieved by substitution with alanine at the same site. SRF-S162D completely abolished serum response factor–DNA binding. In contrast, substitution of the phosphorylatable serine 162 to a nonmodifiable SRF-S162A did not prevent serum response factor–DNA binding.

XIII.D.  Mimicking Phosphorylation of S162 in the MADS-box Permits c-fos Promoter Activity Disruption of serum response factor–DNA binding by SRF-S162D could be overcome by the presence of the ternary complex factor Elk-1 in the context of the c-fos promoter. Serum response factor alone bound to the c-fos SRE probe and with Elk-1 formed a ternary complex (Fig. 18). SRF-S162D displayed markedly diminished binding to the c-fos SRE, but in the company of Elk-1 this serum response factor mutant was able to form a stable ternary complex. Mutation of the Ets binding site (EBS) motif adjacent to the c-fos SRE, to which Elk-1 binds, prevented Elk-1 from rescuing SRF-S162D–DNA binding, indicating that interaction with Elk-1 was necessary and sufficient for formation of a ternary complex involving the c-fos SRE and SRF-S162D. As shown in Fig. 18, homozygous serum response factor-null mouse embryonic stem cells, unable to express serum response factor-dependent myogenic genes and immediate early genes c-fos and Egr-1 was rescued by

644

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

(A)

Cell replication serum, growth factors, PKCα

162 αΙ coil NKLRRYTTFSKRKTGIMKKAYELSTL PO4 + Elk-1 Myogenic genes α−actins and SM22α (B)

C-Fos

Egr1

SM22α

(109 bo) (81 bo) (384 bo) DO D1 D3 D7 DO D1 D3 D7 DO D1 D3 D7

SRF+/–

Immediate early genes C-Fos and Egr1

Ca α-Actin

(C)

(435 bo) DO D1 D3 D7

SRF:RED Sarcomeric Actin:GREEN

SRF–/–

SRF–/– +162D

+SRF +162D

+162A

Figure 18  Phosphorylation of serine-162 in the serum response factor MADS-box I coil is a key determinant for switching serum response factor-dependent gene expression between cell replication and differentiation. The MADS-box served as a regulatory nexus that mediates interactions between serum response factor and a myriad of transcription co-factors, hence differential phosphorylation of MADS-box residues directed specific serum response factor transcriptional complexes toward the expression of either proliferation- or differentiation-dependent genes. Most serum response factor accessory proteins that have been identified as co-regulators of c-fos induction are at endpoints of growth factor-induced signal transduction cascades. Growth factor-stimulated PKC pathways converge on the c-fos SRE to promote myoblast proliferation. PKC is involved mainly in the proliferation process in myoblasts. Phosphorylation of S162, as mimicked by replacement of this residue by aspartate, prevented binding of serum response factor to the c-fos SRE in the absence of an Ets factor protein, but SRF–SRE binding was restored to normal in the presence of Elk-1, which formed a ternary complex with serum response factor on the c-fos SRE. This model showed that growth factor stimulation led to phosphorylation of S162 in the MADS-box in a manner that suppressed the expression of muscle differentiation genes, while permitting and or increasing the association between Ets factors and serum response factor to upregulate c-fos transcription. Model was adapted from Iyer et al. (2006). (B) Rescue of serum response factor-null embryonic stem cells by serum response factor phospho-mimetic S162D permitted expression of immediate early genes, but not myogenic gene targets. Wild-type embryonic stem cells expressed the immediate early genes c-fos and Egr1, the terminal differentiation myogenic contractile genes SMA and cardiac -actin, as shown by RT-PCR of these transcripts on the indicated days of cell culture (top row). Serum response factor-null embryonic stem cells failed to express any of these transcripts (second row). Introduction of HA-tagged human serum response factor induced expression of immediate early genes and terminal differentiation myogenic contractile genes (third row). Introduction of HA-tagged SRF-S162D induced expression only of the immediate early genes (bottom row), but not of the serum response factor myogenic gene targets. (C) Immunofluorescence staining with anti-striated -actin antibody serum response factor showed sarcomere staining by rescue with the S162 to alanine conversion mutant in serum response factor-null embryonic stem cells. Data was adapted from Iyer et al. (2006).

wild-type human serum response factor and restored transcriptional activity of these murine serum response factor gene targets. Transfection of serum response factor-null embryonic stem cells with SRF-S162A displayed sarcomerogenesis (Fig. 18C). However, SRF-S162D restored transcriptional activity only of the immediate early genes exclusively, which require ternary complex formation with an Ets factor. None of the serum response factor-dependent myogenic genes were ­transactivated after introduction

of the SRF-162D mutant, nor activated terminal differentiation as shown by the absence of sarcomeres (Fig. 18C). These results indicate that phosphorylation of S162 in the MADS-box I coil may impede activation of serum response factor-dependent myogenic gene programs without interfering with activation of serum response factordependent proliferation genes. What are the upstream signals that could induce these functionally important covalent modifications in the

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

MADS-box  I coil? The PKC family has been implicated in the control of proliferation and differentiation of muscle cells (Zhu et al., 1991; Capiati et al., 1999; Gliki et al., 2002). Growth factor-stimulated PKC pathways converge on the c-fos SRE to promote myoblast proliferation (Soh et al., 1999), and by phosphorylating a conserved site on myogenin, a basic helix-loop-helix protein, also de-repress the myogenic differentiation program. Capiati et al. (1999) have shown that PKC- is involved mainly in the proliferation process in myoblasts. Phosphorylation of S162, as mimicked by replacement of this residue by aspartate, prevents binding of serum response factor to the c-fos SRE in the absence of an Ets factor protein, but SRF–CArG binding is restored in the presence of Elk-1, which forms a ternary complex with serum response factor. Thus, phosphorylation of the highly-conserved serine-162 residue in the I coil of the serum response factor MADS-box represents a new regulatory nexus or switch that could direct serum response factor target gene expression into proliferation or differentiation programs.

References Ambros, V., 2004. The functions of animal microRNAs. Nature 431, 350–355. Arber, S., Caroni, P., 1996. Specificity of single LIM motifs in targeting and LIM/LIM interactions in situ. Genes Dev. 10, 289–300. Arber Jr, S., Hunter, J.J., Ross, J., Hongo, M., Sansig, G., Borg, J., Perriard, J.C., Chien, K.R., Caroni, P., 1997. MLP-deficient mice exhibit a disruption of cardiac cytoarchitectural organization, dilated cardiomyopathy, and heart failure. Cell 88, 393–403. Arsenian, S., Weinhold, B., Oelgeschlager, M., Ruther, U., Nordheim, A., 1998. Serum response factor is essential for mesoderm formation during mouse embryogenesis. EMB J. 17, 6289–6299. Aubert, J., Dunstan, H., Chambers, I., Smith, A., 2002. Functional gene screening in embryonic stem cells implicates Wnt antagonism in neural differentiation. Nat. Biotechnol. 20, 1240–1245. Azpiazu, N., Frasch, M., 1993. Tinman and bagpipe, two homeo box genes that determine cell fates in the dorsal mesoderm of Drosophila. Genes Dev. 7, 1325–1340. Bains, W., Ponte, P., Blau, H., Kedes, L., 1984. Cardiac actin is the major actin gene product in skeletal muscle cell differentiation in vitro. Mol. Cell Biol. 8, 1449–1453. Barron, M.R., Belaguli, N.S., Trinh, M., Iyer, D., Lough, J.W., Parmacek, M.S., Bruneau, B.G., Schwartz, R.J., 2005. Serum response factor, an enriched cardiac mesoderm obligatory factor is a downstream gene target TBX genes. J. Biol. Chem. 280, 11816–11828. Basson, C.T., Huang, T., Lin, R.C., Bachinsky, D.R., Weremowicz, S., Vaglio, A., Bruzzone, R., Quadrelli, R., Lerone, M., Romeo, G., Silengo, M., Pereira, A., Krieger, J., Mesquita, S.F., Kamisago, M., Morton, C.C., Pierpont, M.E., Muller, C.W., Seidman, J.G., Seidman, C.E., 2001. Different TBX5 interactions in heart and limb defined by Holt-Oram syndrome mutations. Proc. Natl. Acad. Sci. USA 96, 2919–2924. Behar, O., Golden, J.A., Mashimo, H., Schoen, F.J., Fishman, M.C., 1996. Semaphorin III is needed for normal patterning and growth of nerves, bones and heart. Nature 383, 525–528.

645

Belaguli, N.S., Schildmeyer, L.A., Schwartz, R.J., 1997. Organization and myogenic restricted expression of the murine serum response factor gene, a role for autoregulation. J. Biol. Chem. 272, 18222–18231. Belaguli, N.S., Sepulveda, J.L., Nigam, V., Charron, F., Nemer, M., Schwartz, R.J., 2000. Cardiac tissue enriched factors, SRF and GATA-4 are mutual coregulators. Mol. Cell Biol. 20, 7550–7558. Berezikov, E., Berezikov, E., Guryev, V., van de Belt, J., Wienholds, E., Plasterk, R.H., Cuppen, E., 2005. Phylogenetic shadowing and computational identification of human microRNA genes. Cell 120, 21–24. Black, B., Olson, E.N., 1998. Transcriptional control of muscle development by myocyte enhancer factor-2 (MEF2) proteins. Annu. Rev. Cell Dev. Biol. 14, 167–196. Blank, R.S., McQuinn, T.C., Yin, K.C., Thompson, M.M., Takeyasu, K., Schwartz, R.J., Owens, G.K., 1992. Elements of the smooth muscle α-actin promoter required in cis for transcriptional activation in smooth muscle. Evidence for cell type-specific regulation. J. Biol. Chem. 267, 984–989. Bodmer, R., 1993. The gene tinman is required for specification of the heart and visceral muscles in Drosophila. Development 118, 719–729. Bruneau, B.G., Logan, M., Davis, N., Levi, T., Tabin, C.J., Seidman, J.G., Seidman, C.E., 1999. Chamber-specific cardiac expression of Tbx5 and heart defects in Holt-Oram syndrome. Dev. Biol. 211, 100–108. Cao, D., Wang, Z., Zhang, C.L., Oh, J., Xing, W., Li, S., Richardson, J. A., Wang, D.Z., Olson, E.N., 2005. Modulation of smooth muscle gene expression by association of histone acetyltransferases and deacetylases with myocardin. Mol. Cell Biol. 25, 364–376. Capiati, D.A., Tellez-Inon, M.T., Boland, R.L., 1999. Participation of protein kinase C alpha in 1,25-dihydroxy-vitamin D3 regulation of chick myo­ blast proliferation and differentiation. Mol. Cell Endocrinol. 153, 39–45. Chang, J., Wei, J., Otani, T., Youker, K.A., Entman, M.L., Schwartz, R. J., 2003. Inhibitory cardiac transcription SRF-N, is generated by caspase 3 cleavage in human heart failure and attenuated by ventricular unloading. Circulation 108, 405–411. Chang, D.F., Belaguli, N.S., Roberts, W.B., Iyer, D., Wu, S.P., Dong, X. R., Marx, J., Moore, M.S., Beckerle, M.C., Majesky, M., Schwartz, R.J., 2003. Vascular smooth muscle LIM-only proteins CRP1 and CRP2 are potent cardiovascular differentiation cofactors. Dev. Cell 4, 107–118. Chang, D.F., Belaguli, N.S., Schwartz, R.J., 2007. LIM-only protein, CRP2, switched on smooth muscle gene activity in adult cardiac myocytes. Proc. Natl. Acad. Sci. USA 104, 157–162. Chapman, D.L., Papaioannou, V.E., 1998. Three neural tubes in mouse embryos with mutations in the T-box gene Tbx6. Nature 391, 695–697. Chen, C.Y., Schwartz, R.J., 1995. Identification of novel DNA binding targets and regulatory domains of a murine tinman homeodomain factor, Nkx2-5. J. Biol. Chem. 270, 15628–15633. Chen, C.Y., Schwartz, R.J., 1996. Recruitment of the tinman homolog Nkx2-5 by serum response factor activates cardiac α-actin gene transcription. Mol. Cell Biol. 16, 6372–6384. Chen, C.Y., Croissant, J., Majesky, M., Topouzis, S., McQuinn, T., Frankovsky, M.J., Schwartz, R.J., 1996. Activation of the cardiac actin promoter depends upon serum response factor, tinman homologue, Nkx2-5, and intact serum response elements. Dev. Genet. 19, 119–130. Chen, C.Y., Schwartz, R.J., 1997. Competition between negative acting YY1 versus positive acting serum response factor and tinman homologue Nkx2-5 regulates cardiac -actin promoter activity. Mol. Endocrinol. 11, 812–822.

646

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Chen, F., Kook, H., Milewski, R., Gitler, A.D., Lu, M.M., Li, J., Nazarian, R., Schnepp, R., Jen, K., Biben, C., Runke, G., Mackay, J., Novotny, J., Schwartz, R.J., Richard, P., Harvey, R.P., Mullins, M.C., Epstein, J.A., 2002. Hop is an unusual homeobox gene that modulates cardiac development. Cell 110, 713–723. Chen, J.F., Mandel, E.M., Thomson, J.M., Wu, Q., Callis, T.E., Hammond, S.M., Conlon, F.L., Wang, D.Z., 2006. The role of microRNA-1 and microRNA-133 in skeletal muscle proliferation and differentiation. Nat. Genet. 38, 228–233. Chow, K.-L., Schwartz, R.J., 1990. A combination of closely associated positive and negative cis-acting promoter elements regulates transcription of the skeletal -actin gene. Mol. Cell Biol. 10, 528–538. Chow, K.-L., Hogan, M.E., Schwartz, R.J., 1991. Multiple phased cisacting promoter elements interact at short distances to direct skeletal α-actin gene transcription. Proc. Natl. Acad. Sci. USA 88, 1301–1305. Cohen, S.M., Brennecke, J., Stark, A., 2006. Denoising feedback loops by thresholding – a new role for microRNAs. Genes Dev. 20, 2769–2772. Colas, J.F., Lawson, A., Schoenwolf, G.C., 2000. Evidence that translation of smooth muscle α-actin mRNA is delayed in the chick promyocardium until fusion of the bilateral heart-forming regions. Dev. Dyn. 218, 316–330. Costell, M., Gustafsson, E., Aszodi, A., Morgelin, M., Bloch, W., Hunziker, E., Addicks, K., Timple, R., Fassler, R., 1999. Perlecan maintains the integrity of cartilage and some basement membranes. J. Cell Biol. 147, 1109–1122. Croissant, J.D., Kim, J-H., Eichele, G., Goering, L., Lough, J., Prywes, R., Schwartz, R.J., 1996. Avian serum response factor expression restricted primarily to muscle cell lineages is required for -actin gene transcription. Dev. Biol. 177, 250–264. Davis, F.J., Gupta, M., Camoretti-Mercado, B., Schwartz, R.J., Gupta, M.P., 2003. Calcium/calmodulin-dependent protein kinase activates serum response factor transcription activity by its dissociation from histone deacetylase, HDAC4. Implications in cardiac muscle gene regulation during hypertrophy. J. Biol. Chem. 278, 20047–20058. Doyon, Y., Selleck, W., Lane, W.S., Tan, S., Cote, J., 2004. Structural and functional conservation of the NuA4 histone acetyltransferase complex from yeast to humans. Mol. Cell Biol. 24, 1884–1896. Du, K.L., Chen, M., Li, J., Lepore, J.J., Mericko, P., Parmacek, M.S., 2004. Megakaryoblastic leukemia factor-1 transduces cytoskeletal signals and induces smooth muscle cell differentiation from undifferentiated embryonic stem cells. J. Biol. Chem. 279, 17578–17586. Durocher, D., Chen, C.Y., Ardati, A., Schwartz, R.J., Nemer, M., 1996. The atrial natriuretic factor promoter is a downstream target for Nkx2-5 in the myocardium. Mol. Cell Biol. 16, 4648–4655. Durocher, D., Charron, F., Warren, R., Schwartz, R.J., Nemer, M., 1997. The cardiac transcription factors Nkx2-5 and GATA-4 are mutual cofactors. EMBO J. 16, 5687–5696. Escalante, R., Yamada, Y., Cotter, D., Sastre, L., Sameshima, M., 2004. The MADS-box transcription factor SrfA is required for actin cytoskeleton organization and spore coat stability during Dictyostelium sporulation. Mech. Dev. 121, 51–56. Fosset, N., Zhang, Q., Gajewski, K., Choi, C.Y., Kim, Y., Schulz, R.A., 2000. The multitype zinc-finger protein U-shaped functions in heart cell specification in the Drosophila embryo. Proc. Natl. Acad. Sci. USA 97, 7348–7353. Fukushige, T., Brodigan, T.M., Schriefer, L.A., Waterston, R.H., Krause, M., 2006. Defining the transcriptional redundancy of early bodywall muscle development in C. elegans, evidence for a unified theory of animal muscle development. Genes Dev. 20, 3395–3406.

Gajewski, K., Kim, Y., Lee, Y.M., Olson, E.N., Schultz, R.A., 1997. D-mef2 is a target for Tinman activation during Drosophila heart development. EMBO J. 16, 515–522. Gajewski, K., Fossett, N., Molkentin, J.D., Schulz, R.A., 1999. The zinc finger proteins Pannier and GATA4 function as cardiogenic factors in Drosophila. Development 126, 5679–5688. Galvagni, F., Cartocci, E., Oliviero, S., 1998. The dystrophin promoter is negatively regulated by YY1 in undifferentiated muscle cells. J. Biol. Chem. 273, 33708–33713. Ghosh, T., Packham, E., Bonser, A., Robinson, T., Cross, S., Brook, D., 2001. Characterization of the TBX5 binding site and analysis of mutations that cause Holt–Oram syndrome. Hum. Mol. Genet. 10, 1983–1994. Gliki, G., Wheeler-Jones, C., Zachary, I., 2002. Vascular endothelial growth factor induces protein kinase C (PKC)-dependent Akt/PKB activation and phosphatidylinositol 3-kinase-mediates PKC delta phosphorylation, role of PKC in angiogenesis. Cell Biol. Int. 26, 751–759. Goodman, R.H., Smolik, S., 2000. CBP/p300 in cell growth, transformation, and development. Genes Dev. 14, 1553–1577. Gottlieb, P.D., Pierce, S., Sims, R.J., Yamagishi, H., Weihe, E.K., Harris, J.V., Maika, S.D., Kuziel, W.A., King, H.L., Olson, E.N., Nakagawa, O., Srivastava, D., 2002. Bop encodes a muscle restricted protein containing MYND and SET domains and is essential for cardiac differentiation and morphogenesis. Nat. Genet. 31, 25–32. Gray, T.A., Azama, K., Whitmore, K., Min, A., Abe, S., Nicholls, R.D., 2001. Phylogenetic conservation of the Makorin-2 gene, encoding a multiple zinc-finger protein, antisense to the RAF1 proto-oncogene. Genomics 77, 119–126. Gualberto, A., LePage, D., Pons, G., Mader, S.L., Park, K., Atchison, M. L., Walsh, K., 1992. Functional antagonism between YY1 and the serum response factor. Mol. Cell Biol. 12, 4209–4214. Guillemin, K., Groppe, J., Ducker, K., Treisman, R., Hafen, E., Affolter, M., Krasnow, M.A., 1996. The pruned gene encodes the Drosophila serum response factor and regulates cytoplasmic outgrowth during terminal branching of the tracheal system. Development 122, 1353–1362. Habets, P., Moorman, A., Clout, D., van Roon, M., Lingbeek, M., van Lohuizen, M., Campione, M., Christoffels, V., 2002. Cooperative action of Tbx2 and Nkx2-5 inhibits ANF expression in the atrio­ ventricular canal, implications for cardiac chamber formation. Genes Dev. 16, 1234–1246. Hayashi, T., Seki, M., Maeda, D., Wang, W., Kawabe, Y., Seki, T., Saitoh, H., Fukagawa, T., Yagi, H., Enomoto, T., 2002. Ubc9 is essential for viability of higher eukaryotic cells. Exp. Cell Res. 280, 212–221. Henderson, J.R., Macalma, T., Brown, D., Richardson, J.A., Olson, E.N., Beckerle, M.C., 1999. The LIM protein, CRP1, is a smooth muscle marker. Dev. Dyn. 214, 229–238. Herskowitz, I., 1989. A regulatory hierarchy for cell specialization in yeast. Nature 342, 749–757. Hiroi, Y., Kudoh, S., Monzen, K., Ikeda, Y., Yazaki, Y., Nagai, R., Komuro, I., 2001. Tbx5 associates with Nkx2-5 and synergistically promotes cardiomyocyte differentiation. Nat. Genet. 28, 276–280. Ikura, T., Ogryzko, V.V., Grigoriev, M., Groisman, R., Wang, J., Horikoshi, M., Scully, R., Qin, J., Nakatani, Y., 2000. Involvement of the TIP60 histone acetylase complex in DNA repair and apoptosis. Cell 102, 463–473. Iyer, D., Belaguli, N., Fluck, M., Rowan, B.G., Wei, L., Weigel, N.L., Booth, F.W., Epstein, H.F., Schwartz, R.J., Balasubramanyam, A., 2003. A novel phosphorylation target in the serum response factor MADS box regulates α-actin transcription. Biochemistry 42, 7477–7486.

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

Iyer, D., Chang, D., Marx, J., Wei, L., Olson, E.O., Parmacek, M.S., Balasubramanyam, A., Schwartz, R.J., 2006. Serum response factor MADS box serine 162 phosphorylation switches proliferation and myogenic gene programs. Proc. Natl. Acad. Sci. USA 103, 4516–4521. Jain, M.K., Fujita, K.P., Hsieh, C.M., Endege, W.O., Sibinga, N.E., Yet, S.F., Kashiki, S., Lee, W.S., Perrella, M.A., Haber, E., Lee, M.E., 1996. Molecular cloning and characterization of SmLIM, a developmentally regulated LIM protein preferentially expressed in aortic smooth muscle cells. J. Biol. Chem. 271, 10194–10199. Kim, M.S., Merlo, X., Wilson, C., Lough, J., 2006. Co-activation of atrial natriuretic factor promoter by Tip60 and serum response factor. J. Biol. Chem. 281, 15082–15089. Komuro, I., Izumo, S., 1993. Csx, a murine homeobox-containing gene specifically expressed in the developing heart. Proc. Natl. Acad. Sci. USA 90, 8145–8149. Konrat, R., Krautler, B., Weiskirchen, R., Bister, K., 1998. Structure of cysteine- and glycine-rich protein CRP2. Backbone dynamics reveal motional freedom and independent spatial orientation of the lim domains. J. Biol. Chem. 273, 23233–23240. Lee, T.C., Chow, K.-L., Fang, P., Schwartz, R.J., 1991. Activation of skeletal -actin gene transcription, the cooperative formation of serum response factor-binding complexes over positive cis-acting promoter serum response elements displaces a negative-acting nuclear factor enriched in replicating myoblasts and nonmyogenic cells. Mol. Cell. Biol. 11, 5090–5100. Lee, T.-C., Shi, Y., Schwartz, R.J., 1992. Displacement of BrdU-induced YY1 by SRF activates skeletal α-actin gene transcription in embryonic myoblasts. Proc. Natl. Acad. Sci. USA 89, 9814–9818. Lee, T.-C., Zhang, Y., Schwartz, R.J., 1994. Bifunctional transcriptional properties of YY1 in regulating muscle actin and c-myc gene expression during myogenesis. Oncogene 9, 1047–1052. Li, S., Wang, D.Z., Wang, Z., Richardson, J.A., Olson, E.N., 2003. The serum response factor coactivator myocardin is required for vascular smooth muscle development. Proc. Natl. Acad. Sci. 100, 9366–9370. Ling, Y., Lakey, J.H., Roberts, C.E., Sharrocks, A.D., 1997. Molecular characterization of the B-box protein–protein interaction motif of the ETS-domain transcription factor Elk-1. EMBO J. 16, 2431–2440. Lints, T.J., Parsons, L.M., Hartley, L., Lyons, I., Harvey, R.P., 1993. Nkx2-5, a novel murine homeobox gene expressed in early heart progenitor cells and their myogenic descendants. Development 119, 419–431. Lough, J.W., 2002. Transient expression of TIP60 protein during early chick heart development. Dev. Dyn. 223, 419–425. Ma, L., Lu, M.F., Schwartz, R.J., Martin, J.F., 2005. Bmp2 is essential for cardiac cushion epithelial–mesenchymal transition and myocardial patterning. Development 132, 5601–5611. MacLellan, W.R., Lee, T.C., Schwartz, R.J., Schneider, M.D., 1994. Transforming growth factor-β response elements of the skeletal β-actin gene. Combinatorial action of serum response factor, YY1, and the SV40 enhancer-binding protein, TEF-1. J. Biol. Chem. 269, 16754–16760. Manabe, I., Owens, G.K., 2001. Recruitment of serum response factor and hyperacetylation of histones at smooth muscle-specific regulatory regions during differentiation of a novel P19-derived in vitro smooth muscle differentiation system. Circ. Res. 88, 1127–1134. Matsuzaki, K., Minami, T., Tojo, M., Honda, Y., Uchimura, Y., Saitoh, H., Yasuda, H., Nagahiro, S., Saya, H., Nakao, M., 2003. Serum response factor is modulated by the SUMO-1 conjugation system. Biochem. Biophys. Res. Commun. 306, 32–38.

647

McDonald, O.G., Wamhoff, B.R., Hoofnagle, M.H., Owens, G.K., 2006. Control of SRF binding to CArG box chromatin regulates smooth muscle gene expression in vivo. J. Clin. Invest. 116, 36–48. McKinsey, T.A., Zhang, C.L., Lu, J., Olson, E.N., 2000. Signal-dependent nuclear export of a histone deacetylase regulates muscle differentiation. Nature 408, 106–111. Mead, P.E., Deconinck, A.E., Huber, T.L., Orkin, S.H., Zon, L.I., 2001. Primitive erythropoiesis in the Xenopus embryo, the synergistic role of LMO-2, SCL and GATA-binding proteins. Development 128, 2301–2308. Mercher, T., Coniat, M.B., Monni, R., Mauchauffe, M., Khac, F.N., Gressin, L., Mugneret, F., Leblanc, T., Dastugue, N., Berger, R., Bernard, O.A., 2001. Involvement of a human gene related to the Drosophila spen gene in the recurrent t(1;22) translocation of acute megakaryocytic leukemia. Proc. Natl. Acad. Sci. 98, 5776–5779. Miano, J.M., Carlson, M.J., Spencer, J.A., Misra, R.P., 2000. Serum response factor-dependent regulation of the smooth muscle calponin gene. J. Biol. Chem. 275, 9814–9822. Miano J.M., Ramanan, N., Georger, M.A., de Mesy Bentley, K.L., Emerson, R.L., Balza, R.O., Jr, Xiao, Q., Weiler, H., Ginty, D. D., Misra, R.P., 2004. Restricted inactivation of serum response factor to the cardiovascular system. Proc. Natl. Acad. Sci. 101, 17132–17137. Miano, J.M., Long, X., Fujiwara, K., 2007. Serum response factor, master regulator of the actin cytoskeleton and contractile apparatus. Am. J. Physiol. Cell. Physiol. 292, C70–C81. Minchiotti, G., Parisi, S., Liguori, G.L., Andrea, D.D., Persico, M.G., 2002. Role of the EGF-CFC gene cripto in cell differentiation and embryo development. Gene (Amst.) 287, 33–37. Minty, A.J., Alonso, S., Caravatti, M., Buckingham, M.E., 1982. A fetal skeletal muscle actin mRNA in the mouse and its identity with cardiac actin mRNA. Cell 30, 185–192. Minty, A., Kedes, L., 1986. Upstream region of the human cardiac actin gene that modulates its transcription in muscle cells, presence of evolutionarily conserved repeated motif. Mol. Cell Biol. 6, 2125–2136. Miralles, F., Posern, G., Zaromytidou, A.I., Treisman, R., 2003. Actin dynamics control SRF activity by regulation of its coactivator MAL. Cell 113, 329–342. Nacerddine, K., Lehembre, F., Bhaumik, M., Artus, J., Cohen-Tannoudji, M., Babinet, C., Pandolfi, P.P., Dejean, A., 2005. The SUMO pathway is essential for nuclear integrity and chromosome segregation in mice. Dev. Cell 9, 769–779. Niu, Z., Yu, Y., Zhang, S.X., Barron, M., Belaguli, N.S., Schneider, M. D., Parmacek, M., Nordheim, A., Schwartz, R.J., 2005. Conditional mutagenesis of the murine serum response factor gene blocks cardiogenesis and the transcription of downstream gene targets. J. Biol. Chem. 280, 32531–32538. Niu, Z., Li, A., Zhang, S.X., Schwartz, R.J., 2007. Serum response factor micromanaging cardiogenesis. Curr. Opin. Cell Biol. 6, 618–627. Niu, Z., Iyer, D., Conway, S.J., Martin, J.F., Ivey, K., Srivastava, D., Nordheim, A., Schwartz, R.J., 2008. Serum response factor orchestrates nascent sarcomerogenesis and silences biomineralization in the heart. Proc. Natl. Acad. Sci. USA 105, 17824–17829. Nix, D.A., Beckerle, M.C., 1997. Nuclear-cytoplasmic shuttling of the focal contact protein, zyxin, a potential mechanism for communication between sites of cell adhesion and the nucleus. J. Cell Biol. 138, 1139–1147. Norman, C., Runswick, M., Pollock, R., Treisman, R., 1988. Isolation and properties of cDNA clones encoding SRF, and transcription factor that binds to the c-fos serum response element. Cell 55, 989–1003.

648

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Oh, J., Wang, Z., Wang, D.Z., Lien, C.L., Xing, W., Olson, E.N., 2004. Target gene-specific modulation of myocardin activity by GATA transcription factors. Mol. Cell Biol. 24, 8519–8528. Okagawa, H., Markwald, R.R., Sugi, Y., 2007. Functional BMP receptor in endocardial cells is required in atrioventricular cushion mesenchymal cell formation in chick. Dev. Biol. 306, 179–192. Parlakian, A., Tuil, D., Hamard, G., Tavernier, G., Hentzen, D., Concordet, J.P., Paulin, D., Li, Z., Daegelen, D., 2004. Targeted inactivation of serum response factor in the developing heart results in myocardial defects and embryonic lethality. Mol. Cell Biol. 24, 5281–5289. Parlakian, A., Charvet, C., Escoubet, B., Mericskay, M., Molkentin, J. D., Gary-Bobo, G., De Windt, L.J., Ludosky, M.-A., Paulin, D., Daegelen, D., Tuil, D., Li, Z., 2005. Temporally controlled onset of dilated cardiomyopathy through disruption of the SRF gene in adult heart. Circulation 112, 2930–2939. Paxton, C., Zhao, H., Chin, Y., Langner, K., Reecy, J., 2002. Murine Tbx2 contains domains that activate and repress gene transcription. Gene (Amst.) 283, 117–124. Pellegrini, L., Tan, S., Richmond, T.J., 1995. Structure of serum response factor core bound to DNA. Nature 376, 490–498. Pipes, G.C., Creemers, E.E., Olson, E.N., 2006. The myocardin family of transcriptional coactivators, versatile regulators of cell growth, migration, and myogenesis. Genes Dev. 20, 1545–1556. Posern, G., Treisman, R., 2006. Actin together, serum response factor, its cofactors and the link to signal transduction. Trends Cell Biol. 16, 588–596. Qiu, P., Li, L., 2002. Histone acetylation and recruitment of serum responsive factor and CREB-binding protein onto SM22 promoter during SM22 gene expression. Circ. Res. 90, 858–865. Riquelme, C., Barthel, K.K., Qin, F.X., Liu, X., 2006. Ubc9 expression is essential for myotube formation in C2C12. Exp. Cell Res. 312, 2132–2141. Roth, S.Y., Denu, J.M., Allis, C.D., 2001. Histone acetyltransferases. Annu. Rev. Biochem. 70, 81–120. Rottbauer, W., Saurin, A.J., Lickert, H., Shen, X., Burns, C.G., Wo, G.Z., Kemler, R., Kingston, R.K., Wu, C., Fishman, M., 2002. Reptin and pontin antagonistically regulate heart growth in zebrafish embryos. Cell 111, 661–672. Ruzicka, D.L., Schwartz, R.J., 1988. Sequential activation of -actin genes during avian cardiogenesis, vascular smooth muscle -actin gene transcripts mark the onset of cardiomyocyte differentiation. J. Cell. Biol. 107, 2575–2586. Santelli, E., Richmond, T.J., 2000. Crystal structure of MEF2A core bound to DNA at 1.5 A resolution. J. Mol. Biol. 297, 437–449. Schratt, G., Philippar, U., Hockemeyer, D., Schwarz, H., Alberti, S., Nordheim, A., 2004. SRF regulates Bcl-2 expression and promotes cell survival during murine embryonic development. EMBO J. 23, 1834–1844. Schwartz, R.J., Rothblum, K.N., 1981. Gene switching in myogenesis, differential expression of the chicken actin multigene family. Biochemistry 20, 4122–4129. Sepulveda, J.L., Belaguli, N., Nigam, V., Chen, C-Y., Nemer, M., Schwartz, R.J., 1998. GATA-4 and Nkx-2 DNA binding targets: Role for regulating early cardiac gene expression. Mol. Cell Biol. 18, 3405–3415. Sepulveda, J.L., Vlahopoulos, S., Iyer, D., Belaguli, N., Schwartz, R. J., 2002. Combinatorial expression of GATA4, Nkx2-5 and serum response factor directs early cardiac gene activity. J. Biol. Chem. 277, 25775–25782.

Shin, C.H., Liu, Z.P., Passier, R., Zhang, C.L., Wang, D.Z., Harris, T. M., Yamagishi, H., Richardson, J.A., Childs, G., Olson, E.N., 2002. Modulation of cardiac growth and development by HOP, an unusual homeodomain protein. Cell 110, 725–735. Shore, P., Sharrocks, A.D., 1994. The transcription factors Elk-1 and serum response factor interact by direct protein–protein contacts mediated by a short region of Elk-1. Mol. Cell Biol. 14, 3283–3291. Soh, J.-W., Lee, E.H., Prywes, R., Weinstein, I.B., 1999. Novel roles of specific isoforms of protein kinase C in activation of the c-fos serum response element. Mol. Cell Biol. 19, 1313–1324. Solloway, M.J., Harvey, R.P., 2003. Molecular pathways in myocardial development, a stem cell perspective. Cardiovasc Res. 58, 264–277. Sommer, H., Beltran, J.P., Huijser, P., Pape, H., Lonnig, W.E., Saedler, H., Schwarz-Sommer, Z., 1990. Deficiens, a homeotic gene involved in the control of flower morphogenesis in Antirrhinum-Majus – the protein shows homology to transcription factors. EMBO J. 9, 605–613. Spencer, J.A., Misra, R.P., 1996. Expression of the serum response factor gene is regulated by serum response factor binding sites. J. Biol. Chem. 271, 16535–16543. Spencer, J.A., Major, M.L., Misra, R.P., 1999. Basic fibroblast growth factor activates serum response factor gene expression by multiple distinct signaling mechanisms. Mol. Cell Biol. 19, 3977–3988. Sugi, Y., Lough, J., 1992. Onset of expression and regional deposition of alpha-smooth and sarcomeric actin during avian heart development. Dev. Dyn. 193, 116–124. Sun Jr, Q., Chen, G., Streb, J.W., Long, X., Yang, Y., Stoeckert, C.J., Miano, J.M., 2006. Defining the mammalian CArGome. Genome. Res. 16, 197–207. Tada, M., Smith, J.C., 2001. T-targets: Clues to understanding the functions of T-box proteins. Dev. Growth Differ. 43, 1–11. Tevosian, S.G., Deconinck, A.E., Cantor, A.B., Rieff, H.I., Fujiwara, Y., Corfas, G., Orkin, S.H., 1999. FOG-2, a novel GATA-family cofactor related to multitype zinc-finger proteins Friend of GATA-1 and U-shaped. Proc. Natl. Acad. Sci. USA 96, 950–955. Treisman, R., 1986. Identification of a protein-binding site that mediates transcription response of the c-fos gene to serum factors. Cell 46, 567–574. Treisman, R., 1994. Ternary complex factors, growth factor regulated transcriptional activators. Curr. Opin. Genet. Dev. 4, 96–101. Treisman, R., 1995. Journey to the surface of the cell: Fos regulation and the SRE. EMBO J. 14, 4905–4913. Treisman, R., Ammerer, G., 1992. The SRF and MCM1 transcription factors. Curr. Opin. Genet. Dev. 2, 221–226. van Rooij, E., Sutherland, L.B., Qi, X., Richardson, J.A., Hill, J., Olson, E.N., 2007. Control of stress-dependent cardiac growth and gene expression by a microRNA. Science 316, 575–579. Wadman, I.A., Osada, H., Grutz, G.G., Agulnick, A.D., Westphal, H., Forster, A., Rabbitts, T.H., 1997. The LIM-only protein Lmo2 is a bridging molecule assembling an erythroid, DNA-binding complex which includes the TAL1, E47, GATA-1 and Ldb1/NLI proteins. EMBO J. 16, 3145–3157. Wang, D., Chang, P.S., Wang, Z., Sutherland, L., Richardson, J.A., Small, E., Krieg, P.A., Olson, E.N., 2001. Activation of cardiac gene expression by myocardin, a transcriptional cofactor for serum response factor. Cell 105, 851–862. Wang, D.Z., Li, S., Hockemeyer, D., Sutherland, L., Wang, Z., Schratt, G., Richardson, J.A., Nordheim, A., Olson, E.N., 2002. Potentiation of serum response factor activity by a family of myocardin-related transcription factors. Proc. Natl. Acad. Sci. 99, 14855–14860.

Chapter  |  9.3  Serum Response Factor and Co-Factors, Roles in Cardiac Development

Wang, J., Feng, X.H., Schwartz, R.J., 2004. SUMO-1 modification activated GATA4 dependent cardiogenic gene activity. J. Biol. Chem. 279, 49091–49098. Wang, J., Li, A., Wang, Z.G., Feng, X.H., Olson, E.N., Schwartz, R.J., 2007. Myocardin sumoylation transactivate cardiogenic genes in pluripotent 10T1/2 fibroblasts. Mol. Cell. Biol. 27, 622–632. Wang, Z., Wang, D.Z., Pipes, G.C., Olson, E.N., 2003. Myocardin is a master regulator of smooth muscle gene expression. Proc. Natl. Acad. Sci. USA 100, 7129–7134. Wang, Z., Wang, D.Z., Hockemeyer, D., McAnally, J., Nordheim, A., Olson, E.N., 2004. Myocardin and ternary complex factors compete for SRF to control smooth muscle gene expression. Nature 428, 185–189. Weiskirchen, R., Pino, J.D., Macalma, T., Bister, K., Beckerle, M.C., 1995. The cysteine-rich protein family of highly related LIM domain proteins. J. Biol. Chem. 270, 28946–28954. Witte, S., Villaba, M., Bi, K., Liu, Y., Isakov, N., Altman, A., 2000. Inhibition of the c-Jun N-terminal Kinase/AP-1 and NF-kappa B pathways by PICOT, a novel protein kinase c-interacting protein with a thioredoxin homology domain. J. Biol. Chem. 275, 1902–1909. Xu, W., Edmondson, D.G., Evrard, Y.A., Wakamiya, M., Behringer, R. R., Roth, S.Y., 2000. Loss of Gcn5l2 leads to increased apoptosis and mesodermal defects during mouse development. Nat. Genet. 26, 229–232. Xue, Y., Wang, X., Li, Z., Gotoh, N., Chapman, D., Skolnik, E.Y., 2001. Mesodermal patterning defect in mice lacking the Ste20 NCK interacting kinase (NIK). Development 128, 1559–1572. Yamauchi, T., Yamauchi, J., Kuwata, T., Tamura, T., Yamashita, T., Bae, N., Westphal, H., Ozato, K., Nakatani, Y., 2000. Distinct but overlapping roles of histone acetylase PCAF and of the closely related PCAF-B/GCN5 in mouse embryogenesis. Proc. Natl. Acad. Sci. USA 97, 11303–11306.

649

Yao, T.P., Oh, S.P., Fuchs, M., Zhou, N.D., Ch’ng, L.E., Newsome, D., Bronson, R.T., Li, E., Livingston, D.M., Eckner, R., 1998. Gene dosagedependent embryonic development and proliferation defects in mice lacking the transcriptional integrator p300. Cell 93, 361–372. Yin, F., Herring, B.P., 2005. GATA-6 can act as a positive or negative regulator of smooth muscle-specific gene expression. J. Biol. Chem. 280, 4745–4752. Zhang, S.X., Gras, E.G., Wycuff, D.R., Marriot, S.J., Kadeer, N., Yu, W., Olson, E.N., Garry, D.J., Parmacek, M.S., Schwartz, R.J., 2005. Identification of direct serum response factor gene targets during DMSO induced P19 cardiac cell differentiation. J. Biol. Chem. 280, 19115–19126. Zhang, X., Azhar, G., Chai, J., Sheridan, P., Nagano, K., Brown, T., Yang, J., Khrapko, K., Borras, A.M., Lawitts, J., Misra, R.P., Wei, J. Y., 2001. Cardiomyopathy in transgenic mice with cardiac-specific overexpression of serum response factor. Am. J. Physiol. Heart. Circ. Physiol. 280, H1782–H1792. Zhao, Y., Samal, E., Srivastava, D., 2005. Serum response factor regulates a muscle-specific mircroRNA that targets Hand2 during cardiogenesis. Nature 436, 214–220. Zhao, Y., Ransom, J.F., Li, A., Vedantham, V., von Drehle, M., Muth, A. N., Tsuchihashi, T., McManus, M.T., Schwartz, R.J., Srivastava, D., 2007. Dysregulation of cardiogenesis, cardiac conduction, and cell cycle in mice lacking miRNA-1-2. Cell 129, 303–317. Zhou, J., Hu, G., Herring, B.P., 2005. Smooth muscle-specific genes are differentially sensitive to inhibition by Elk-1. Mol. Cell Biol. 25, 9874–9885. Zhu, Y.Y., Schwartz, R.J., Crow, M.T., 1991. Phorbol esters selectively downregulate contractile protein gene expression in terminally differentiated myotubes through transcriptional repression and message destabilization. J. Cell Biol. 115, 745–754.

Chapter 9.4

T-Box Factors Frank L. Conlon1 and Katherine E. Yutzey2 1

Department of Genetics, Fordham Hall, University of North Carolina-Chapel Hill, Chapel Hill, NC, USA  Division of Molecular Cardiovascular Biology, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, USA

2

I.  Introduction The T-box genes are a large family of transcription factors required for early cell fate decisions, differentiation and organogenesis during development. Members of the T-box gene family comprise approximately 0.1% of most metazoan genomes, and have been identified in genomes of organisms as diverse as nematodes and humans. All members of this class of proteins appear to function as classic transcription factors in that they localize to the nucleus, bind DNA in a sequence-specific manner, and directly regulate the transcriptional level of downstream target genes. The significance of the T-box genes in early development is exemplified by the observation that when T-box factors are mutated or absent, dramatic phenotypes occur in mouse, dog, zebrafish and frog, and such changes in these genes have been implicated in a number of human congenital malformations. A series of clinical studies have provided direct evidence of a role for T-box genes in heart development, and congenital heart disease (CHD) has been associated with mutations in TBX5, TBX1 and TBX20 (Papaioannou and Silver, 1998; Papaioannou, 2001; Prall et al., 2002; Packham and Brook, 2003; Ryan and Chin, 2003; Baldini, 2004; Showell et al., 2004; Mandel et al., 2005; Plageman and Yutzey, 2005; Stennard and Harvey, 2005; Kirk et al., 2007; Hammer et al., 2008; Liu et al., 2008; Qian et al., 2008).

II.  Brachyury and the t-box family of proteins Positional cloning and sequencing of the genes defective in the mouse gastrulation mutant Brachyury, also known as T, and of the Drosophila behavioral mutant optomotorblind (omb), show extensive sequence similarity between Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

the amino-terminal regions of the two encoded proteins (Herrmann et al., 1990; Pflugfelder et al., 1992); the region of similarity contains a unique sequencespecific DNA-binding domain. Since these initial observations were made, over 50 proteins with sequence similarity to the DNA-binding domain of Brachyury and omb have been identified. The T-box has been defined as the minimal region within the T-box protein that is necessary and sufficient for sequence-specific DNA binding (Pflugfelder et al., 1992; Kispert and Hermann, 1993; Smith, 1999). Although crystallographic analysis of T-box proteins has been achieved only for a truncated version of the Xenopus homolog of Brachyury (Xbra) and a truncated version of Tbx3, the results clearly demonstrate that the T-box is unlike any other DNA-binding domain (Muller and Herrmann, 1997; Coll et al., 2002). The T-box is a relatively large DNA-binding domain, generally comprising about a third of the entire protein (17–26 kDa), and individual T-box gene family members show varying degrees of homology across the domain. This observation has provided the basis for subdivision of the family (Fig. 1) (Agulnik et al., 1996, 1997). It has recently been demonstrated that the specificity of several T-box proteins for their target sites lies mainly within the T-box. But specificity alone does not appear to be directly related to binding affinity (Conlon et al., 2001), suggesting that the T-box domain may perform other functions, such as providing sites required for protein–protein interactions (see below). Members of the T-box family are expressed in, and required for, the development of multiple cell types in diverse organisms; this has been demonstrated by genetic studies in flies, worms, fish, mice, dogs and humans (Papaioannou and Silver, 1998; Haworth et al., 2001; Papaioannou, 2001; Prall et al., 2002; Wilson and Conlon, 2002; Packham and Brook, 2003; Ryan and Chin, 2003; 651

652

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Figure 1  Phylogenetic tree of related T-box genes expressed in the heart. A phylogenetic tree of genes in the Tbx1 and Tbx2 subfamilies was constructed based on sequence alignments of T-box DNA-binding domains of human, mouse, chick, Xenopus, zebrafish and Drosophila protein sequences. Plageman and Yutzey (2005).

Baldini, 2004; Showell et al., 2004; Mandel et al., 2005; Plageman and Yutzey, 2005). For many of these genes, such as Brachyury, there are clear orthologs (direct homologs) which show a high degree of similarity in sequence, expression pattern and function among a variety of vertebrates, including fish, frogs, dogs and mice. Other T-box genes appear to be unique to a particular species; for example, VegT, a T-box gene thought to be required for endoderm formation in Xenopus, has no apparent ortholog in the mouse or human. Genomic analyses of human and mouse have shown that there are 18 members of the T-box family in mammals, and representatives have been identified in a wide range of animals. Collectively, studies employing expression analysis, misexpression, overexpression, genetic mutation, or protein depletion of the members of the T-box gene family and clinical studies strongly suggest a fundamental role for a large subset of T-box-containing proteins including Tbx1, Tbx2, Tbx3, Tbx5, Tbx18, Tbx20 and Eomes in heart development (Table 1; Fig. 2) (Papaioannou and Silver, 1998; Haworth et al., 2001; Papaioannou, 2001; Prall et al., 2002; Packham and Brook, 2003; Ryan and Chin, 2003; Baldini, 2004; Showell et al., 2004; Mandel et al., 2005; Plageman and Yutzey, 2005).

II.A.  Tbx1 Tbx1 was initially cloned in mouse using PCR and degenerate flanking primers to amplify the DNA-binding domains

of Brachyury and the related Drosophila gene omb. The degenerate primers were used to screen embryonic day (E)12.5 mouse cDNA, resulting in the identification of three new members of the T-box gene family including Tbx1 (Bollag et al., 1994). Since this initial landmark screen, Tbx1 has been identified in a variety of model systems including amphioxus, lamprey, zebrafish, Xenopus, chick, mouse and human (Law et al., 1995; Gibson-Brown et al., 1996; He et al., 1999; Ruvinsky et al., 2000; Garg et al., 2001; Ataliotis et al., 2005; Showell et al., 2006; Tiecke et al., 2007). Evidence for a role for Tbx1 in heart development came from the observation that TBX1 is located in the region of human chromosome 22 (22q11.2) that is often deleted in one allele of patients with DiGeorge syndrome (see below). Congenital heart defects are the most common feature of DiGeorge syndrome, or del22q11 deletion syndrome, and may include tetralogy of Fallot, interruption of the aortic arch, ventricular septal defects (VSD), pulmonary atresia, or persistent truncus arteriosus (Yamagishi and Srivastava, 2003; Baldini, 2004). Mice deficient for the syntenic region of human 22q11.2, or those having null mutations in Tbx1, show a phenotype that is highly similar to, but less severe than, DiGeorge patients in the patterning of the pharyngeal endoderm and the aortic arches. The phenotype also includes reduced proliferation of cells within the second heart field (SHF), which directly or indirectly leads to defects in SHF-derived tissues, including the cardiac outflow tract (OFT) and the

653

Chapter  |  9.4  T-Box Factors

Table 1  Expression and Function of Vertebrate T-box Genes in the Developing Hearta Expressed

Cardiovascular-null phenotypes

pharynx heart, forelimb heart, somites lateral mesoderm, heart

aortic arch anomalies thin hypomorphic heart not reported not reported

primitive heart, atria heart, motor neurons

hypomorphic heart (morphant) hypomorphic heart (morphant)

heart, OFTb, pharynx heart, AVC, atria, CS heart, CS heart, atria, LV, IVS, CS epicardium, sinus venosus primitive heart, cushions

aortic arch anomalies, PTA AV canal and OFT anomalies SAN anomalies hypomorphic posterior heart tube inflow tract anomalies hypomorphic myocardium

Zebrafish van gogh (tbx1) heartstrings (tbx5) tbx18 hrT (tbx20)

Xenopus Tbx5 Tbx20

Mouse Tbx1 Tbx2 Tbx3 Tbx5 Tbx18 Tbx20

Human TBX1 TBX3 TBX5 TBX20

Mutant phenotypes not reported not reported atria, forelimbs not reported

DiGeorge (22q11del) syndrome (IAA, Oft anomalies) Ulnar mammary syndrome (no reported heart defects) Holt-Oram syndrome (ASD, conduction disease) Valvuloseptal defects

a

References are within the text. Abbreviations: outflow tract: OFT; atrioventricular: AV; persistent truncus arteriosus: PTA; atrioventricular canal: AVC; conduction system: CS; left ventricle: LV; sino-atrial node: SAN; interrupted aortic arch: IAA; atrial septal defect: ASD.

b

Figure 2  Expression of T-box genes in the developing four-chambered heart. Cardiac structures that express Tbx1, Tbx2, Tbx3, Tbx5, Tbx18 or Tbx20 are indicated based on studies of chick and mouse embryos. Atria and left ventricle are red, right ventricle is blue, atrioventricular canal is blue and green together, interventricular septum is orange and the outflow tract and great arteries are yellow. Tbx18 expression in the epicardium (black) is indicated by a circle. See text for details and references.

coronary ­arteries (Jerome and Papaioannou, 2001; Lindsay et al., 2001; Merscher et al., 2001; Vitelli et al., 2002a; Yamagishi and Srivastava, 2003; Xu et al., 2004; Paylor et al., 2006; Kelly and Papaioannou, 2007; Liao et al., 2008; Theveniau-Ruissy et al., 2008). A role for Tbx1 in the second heart field is further supported by genetic lineage analysis in the mouse, which showed that cells in the second heart field that express Tbx1 contribute to the outflow tract, endocardium and the mesenchymal cushions (Brown et al., 2004; Xu et al., 2004; Maeda et al., 2006). A functional role for Tbx1 in the second heart field emerged from studies showing that mesodermal-specific deletion of Tbx1 recapitulates the defects seen in Tbx1-null ­animals and that, importantly, defects in the outflow tract can be rescued by Tbx1 expression in the second heart field exclusively (Zhang et al., 2006). Within the second heart field, Tbx1 transcription has been shown to be regulated by the integration of several signal transduction pathways, including the sonic hedgehog (Shh) signaling pathway via the Foxc1 and Foxc2 transcription factors, and by retinoic acid (Yamagishi et al., 2003; Yamagishi and Srivastava, 2003; Roberts et al., 2005; Guris et al., 2006; Maeda et al., 2006). Collectively, these data suggest that Tbx1 functions exclusively within

654

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

the second heart field, but ablation of Tbx1 within the primary heart field by an Nkx2.5-Cre-driver also recapitulates the outflow tract phenotype of Tbx1-null animals (Xu et al., 2004). However, since the Nkx2.5-Cre-driver used in these studies is expressed in the pharyngeal mesoderm as well as in the primary heart field, and since ablation of Tbx1 by the pharyngeal endoderm driver Foxg1-Cre also recapitulates the Tbx1 outflow tract-null phenotype, the function of Tbx1 in the primary heart field remains to be determined (Xu et al., 2004; Zhang et al., 2005).

II.B.  Tbx2 In the same homology-based screen that identified Tbx1, two additional members of the T-box family, Tbx2 and Tbx3, were identified; as was done for Tbx1, full-length Tbx2 orthologs were identified and cloned from a wide variety of vertebrates including zebrafish, Xenopus, chick, mouse and human (Law et al., 1995; Gibson-Brown et al., 1996; He et al., 1999; Ruvinsky et al., 2000; Showell et al., 2006). Collectively, sequence analysis of Tbx2 strongly suggests that Tbx2 and Tbx3 are the orthologs of Drosophila omb (Agulnik et al., 1996, 1997). Indications of a role for Tbx2 in heart development came from in situ analysis showing expression of Tbx2 in the myocardium of the atrioventricular canal (AVC), inflow tract (IFT), and outflow tract (OFT) in mouse, chick and human (Chapman et al., 1996; Habets et al., 2002; Christoffels et al., 2004). Importantly, in mouse and human, expression of Tbx2 RNA and protein and expression of chamber-specific genes appear to be mutually exclusive (Christoffels et al., 2004). The ability of Tbx2 to repress expression of the chamber-specific genes natriuretic peptide precursor A (Nppa, also called Atrial Natriuretic Factor (ANF)), gap junction protein alpha 5 (Gja5, also called connexin 40) and gap junction protein alpha 1 (Gja1, also called connexin 43) in tissue culture and in transgenic mouse models, led to the proposal that Tbx2 functions in cardiac development to repress the gene programs associated with chamber formation and differentiation (Christoffels et al., 2004; Moorman et al., 2004) (see Chapter 3.2). This model is supported by the finding that uniform ectopic expression of Tbx2 throughout the developing heart of transgenic mice, via -myosin heavy chain (-MyHC) regulatory regions, leads to a failure of expression of the chamber-specific gene Nppa and defects in cardiac looping. Additional experimental support for a role for Tbx2 in repressing chamber identity was provided by studies of mice that genetically lack Tbx2. These mice die by E14.5 due to cardiac abnormalities, including grossly abnormal outflow tract septation and misalignment of the aorta and pulmonary trunk with the appropriate ventricles. Consistent with these abnormalities, Tbx2-homozygous mutant mice display the associated downregulation of the cardiac chamber genes Nppa, Gja5,

cited1 and small muscle protein X-linked (Smpx, also called chisel), while showing little or no alteration in the expression of ­ cardiac genes involved in anteroposterior patterning, such as myosin light chain (MLC)2v and HAND1. Moreover, the cardiac defects in Tbx2 mutants appear to be direct, and are independent of the role of Tbx2 in cellcycle progression (Harrelson et al., 2004) (see below). Taken together, the data strongly support a role for Tbx2 in restricting initiation of chamber formation to the prospective atria and ventricles and, therefore, functioning indirectly to position the developing conduction system.

II.C.  Tbx3 Similar to Tbx2, Tbx3 was cloned in an initial screen with the aim of identifying vertebrate T-box-containing proteins (Bollag et al., 1994). In mouse, Tbx3 is first expressed in the inner cell mass, and later in the endodermal and meso­ dermal components of the yolk sac, as well as in other extraembryonic tissues including amnion, chorion and allantois. Later in development, Tbx3 is expressed in the primordia of many organs including the limbs, mammary buds, liver, spleen, pituitary gland, lung, kidney and heart (Chapman et al., 1996; Gibson-Brown et al., 1996, 1998a; Yamada et al., 2000). Shortly after the identification of Tbx3, it was reported that mutations in this gene are associated with the human autosomal dominant disorder ulnar-mammary syndrome (UMS). This disease is characterized by malformation of the upper limb derived from the posterior arrays and hypoplasia of apocrine and/or mammary tissue (Bamshad et al., 1995, 1997, 1999). However, cardiac anomalies have not been reported for this patient population. Mice heterozygous for mutations in Tbx3 appear normal; however, homozygous Tbx3 mutant mice display defects very similar to those observed in patients with UMS (Davenport et al., 2003). Taken together, these results provide compelling evidence that the function of Tbx3 is evolutionarilyconserved; the data further imply that limb and mammary tissues are exquisitely sensitive to the relative levels of Tbx3. However, the precise cellular and molecular pathways by which Tbx3 functions remain to be determined. Despite its role in limb and mammary development, genetic analysis of mice deficient in Tbx3 initially revealed no overt cardiac abnormalities (Davenport et al., 2003). However, more detailed analysis of mouse embryos, including inactivation of the Tbx3 locus by insertion of Cre, demonstrated that diversification of the sinoatrial node from the atrial myocardium is Tbx3-dependent. Conversely, increased expression of Tbx3 in the atria leads to formation of ectopic pacemaker sites (see Chapter 2.3). Together, these studies show that Tbx3 is both necessary and sufficient for sinoatrial node development and establishment

Chapter  |  9.4  T-Box Factors

of pacemaker function in the atria (Hoogaars et al., 2007). In addition to having defects in the sinoatrial node, mice homozygous for mutations in Tbx3 also display abnormal extension of the arterial pole of the heart, leading either to the formation of a double outlet for the right ventricle or a transposition of the great arteries (see below). Given that Tbx3 is expressed in both the pharyngeal endoderm and the neural crest cells, which give rise to outflow tract, it remains to be determined which population of cells requires Tbx3, and whether there is a direct requirement for Tbx3 in the outflow tract (Mesbah et al., 2008). While most T-box proteins show sequence conservation between family members mainly within the T-box domain, Tbx2 and Tbx3 display extensive regions of hom­ ology across the entire protein, and together are thought to represent the products of an ancient gene duplication of omb (Agulnik et al., 1996, 1997; Habets et al., 2002; Christoffels et al., 2004). Consistent with this hypothesis, the genes show similar patterns of expression in the developing cardiac conduction system (Sinha et al., 2000; Carlson et al., 2001). However, their expression is tempor­ ally nonoverlapping and, as Tbx2 begins to be downregulated in the cardiac conduction system, Tbx3 begins to be upregulated in precisely the same tissue types. The observation that both Tbx2 and Tbx3 function as transcriptional repressors and the finding that they both repress the same target genes, including Nppa and Gja5, led Christoffels and colleagues to propose that Tbx2 represses the expression of chamber-specific genes during the early stages of cardiogenesis, while Tbx3 functions to maintain the boundaries between the chambers and conduction system at later stages of development and possibly into adulthood (Habets et al., 2002; Christoffels et al., 2004; Hoogaars et al., 2007). With the recent availability of genetic Tbx2 and Tbx3 mutant lines, new information on the functional redundancy of these proteins and the significance for chamber-specific gene expression and conduction system development should be forthcoming.

II.D.  Tbx5 Tbx5 was identified in a screen of an E8.5 mouse cDNA library using the T-box region of Tbx2 as a probe (Agulnik et al., 1996). Homologs were later identified in a wide variety of vertebrate species including chick (Gibson-Brown et al., 1998b; Ohuchi et al., 1998), zebrafish (Begemann and Ingham, 2000; Ruvinsky et al., 2000), Xenopus (Horb and Thomsen, 1999; Showell et al., 2006) and human (Basson et al., 1997; Li et al., 1997). Sequence comparison of the vertebrate homologs show that the Tbx5 amino acid sequence has an extremely high degree of conservation throughout evolution; 99% in the T-box domain, and 57% overall between the Xenopus and human proteins (Horb and Thomsen, 1999).

655

Similar to other individual T-box-encoding genes, Tbx5 homologs share not only sequence similarity, but also a high degree of conservation in their temporal and spatial patterns of expression. Initial studies in the mouse show Tbx5 expression is restricted to a subset of cells in the developing heart, eye and limb, and expression persists in each structure throughout later stages of development, with the exception of the dorsal eye (Chapman et al., 1996). Tbx5 homologs in Xenopus, zebrafish and avian species display very similar patterns of expression (Li et al., 1997; Gibson-Brown et al., 1998b; Horb and Thomsen, 1999; Begemann and Ingham, 2000). In the developing heart, Tbx5 is initially expressed throughout the heart field with relatively high levels in the inflow tract, atrium and left ventricle, and with low-to-undetect able levels in the right ventricle. Expression is not observed in the outflow tract or major arteries but, in some species, Tbx5 has been reported to be expressed at relatively high levels in the veins, including the common cardinal vein and the anterior aspects of the hepatic vein (Brown et al., 2005; Showell et al., 2006). In the atrium, Tbx5 is expressed both in the endocardial and myocardial layers. Thus, Tbx5 is expressed in a posterior-to-anterior gradient and, with Nkx2.5 and Tbx20, is among the first genes expressed in cardiogenic precursor cells (Tonissen et al., 1994; Chapman et al., 1996; Gibson-Brown et al., 1998b; Bruneau et al., 1999; Horb and Thomsen, 1999; Begemann and Ingham, 2000; Griffin et al., 2000; Brown et al., 2005; Showell et al., 2006; Bimber et al., 2007). Implications of a function for Tbx5 in heart development initially came from its association with the human congenital heart disease Holt-Oram syndrome (HOS), a relatively rare, highly-penetrant autosomal dominant condition that is associated with forelimb and cardiac anomalies (see below) (Basson et al., 1997; Li et al., 1997). The role of Tbx5 in heart development and in Holt-Oram syndrome is supported by gene-targeting experiments in mouse, which demonstrated that loss of one Tbx5 allele leads to many of the phenotypic abnormalities observed in Holt-Oram syndrome patients. Specifically, Tbx5-heterozygous mice display atrial septal defects (ASDs), including secundum and primum ASDs, defects in ventricular relaxation, and conduction system abnormalities, such as atrioventricular delay (Bruneau et al., 2001; Moskowitz et al., 2004; Zhu et al., 2008). The complete loss of Tbx5 results in embryonic lethality, with failure of the atria to develop and dramatic reduction in the expression of the chamber-specific genes Nppa, Gja5, MLC2v, Irx4 and Hey2, as well as two markers associated with early cardiac commitment, Nkx2.5 and Gata4 (Bruneau et al., 2001). Together with the analysis of Tbx2 and Tbx3, these data suggest that Tbx5 functions to activate a chamber differentiation program, and that Tbx2 and Tbx3 limit the ­expression of the program to the prospective ventricle and atrium while repressing it in prospective conductive tissue,

656

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

most likely through the regulation of the same set of target genes, including Nppa. Distinct from a role in chamber specification and differentiation, Tbx5 is also required for correct development of the cardiac conduction system. In newborn mice, Tbx5 is expressed in the atrioventricular bundle and in the left and right bundle branches, where it continues to be expressed to adulthood. In Tbx5 heterozygous mutant newborn mice, Tbx5 expression is initiated at the proper time in the tissue of the conduction system, but the mice display conductivity patterning defects in the bundle branches and in the atrioventricular bundle, with failure to undergo the morphological changes observed in the conduction system in wild-type mice. These defects in conductivity maturation are manifested in altered electrocardiograms (ECGs), with Tbx5 heterozygous mutant and wild-type mice having indistinguishable ECGs at birth; however, as the mice mature, the Tbx5 heterozygous mutants fail to undergo a shortening of the time required for electrical propagation from the sinoatrial node to the ventricular myocardium, leading to a prolonged PQ interval. Thus, it appears that Tbx5 is required for maturation of the atrioventricular node (AVN) or the atrioventricular bundle, and that Tbx5 is required for the normal function of the ventricular conduction system (Moskowitz et al., 2004). Experiments carried out in tissue culture and in vivo in mouse, chick and Xenopus are consistent with an evolutionarily-conserved requirement for Tbx5 in multiple aspects of vertebrate heart development (Horb and Thomsen, 1999; Liberatore et al., 2000; Hatcher et al., 2001; Hiroi et al., 2001; Garrity et al., 2002; Fijnvandraat et al., 2003; Takeuchi et al., 2003; Plageman and Yutzey, 2004, 2006; Brown et al., 2005; Goetz et al., 2006; Mori et al., 2006). For example, overexpression of Tbx5 in embryonic carcinoma cells leads to an increase in cardiac markers normally expressed in these cells (Hiroi et al., 2001). Similarly, when Tbx5 is misexpressed early in the heart of chick, mouse or Xenopus embryos, primitive heart tube formation is abnormal and the cardiac tissue fails to develop (Liberatore et al., 2000; Hatcher et al., 2001; Takeuchi et al., 2003; Goetz et al., 2006). At later stages of development, Tbx5 misexpression throughout the heart in avian embryos prevents the development and correct placement of the interventricular septum (Takeuchi et al., 2003). Additional studies in chick embryos have shown that Tbx5 is expressed in the coronary vasculature, and that ectopic expression of Tbx5 in the proepicardium prevents migration and establishment of epicardial progenitor populations (Hatcher et al., 2004). A potential role for Tbx5 in cardiac cell-cycle regulation was shown in Xenopus and zebrafish embryos, which exhibit thinning of the myocardium and loss of circulation, with reduced Tbx5 function (Garrity et al., 2002; Brown et al., 2005). In Xenopus these defects occur concomitantly with a decrease in cardiac cell number resulting from a G1/S-phase delay or

arrest (see below) (Brown et al., 2005; Goetz et al., 2006). Collectively, these studies have implicated Tbx5 in cardiomyogenic differentiation, heart chamber maturation, conduction system specialization, epicardial development, interventricular septum formation and cardiomyocyte cell proliferation.

II.E.  Tbx18 TBX18 was identified by a data homology search of the human genome with assembled expressed sequence tag (EST) contigs. Sequence analysis of one of the contigs showed that the ESTs code for a novel member of the Tbx1/8/14/15 T-box subclass termed Tbx18 (Yi et al., 1999). Subsequently, TBX18 homologs were identified in mouse, zebrafish, Xenopus and chick (Kraus et al., 2001; Begemann et al., 2002; Haenig and Kispert, 2004; Jahr et al., 2008). Expression studies in these species show that Tbx18 is expressed in a variety of developing tissues, including the heart. Specifically, in mouse and chick, Tbx18 is expressed in the proepicardium and its descendents, as they undergo epithelial–mesenchymal transition, migrate, adhere and colonize the myocardial and endocardial layers of the heart (Kraus et al., 2001; Schulte et al., 2007; Jahr et al., 2008). The expression of Tbx18 within the proepicardium appears to be in response to low levels of bone morphogenic protein 2 (BMP2), as elevated or reduced levels of BMP2 abolish expression of Tbx18 in the proepicardium and its derivatives (Schlueter, 2006). Although it remains unclear if zebrafish have a structure analogous to the proepicardium, Tbx18 has been shown to be expressed in the septum transversum, the tissue that gives rise to the epicardium in mammals and avian species, and to the tissues that are substrates for the migrating epicardial cells including the sinus venosus, atrium and ventricle (Begemann et al., 2002; Schulte et al., 2007). In contrast to mammals, axolotls and zebrafish have the ability to repair damaged adult heart tissue (Oberpriller and Oberpriller, 1974; Poss et al., 2002). A functional role for Tbx18 in the epicardium and in injury repair was established in zebrafish by observations from the Poss laboratory which showed that, on ventricular damage, FGF signals from the epicardium lead to a dramatic upregulation of Tbx18 in the epicardium surrounding the site of injury and, crucially, in the subepicardial tissue within the wound and regenerating ventricle (see Chapter 12.2). These findings are consistent with earlier studies by the same group, which showed strong and sustained expression of Tbx18 for 30 days or more after ventricular damage (Poss et al., 2002, 2003; Lepilina et al., 2006). In an analogous set of studies using genetic lineage tracing in mice, it was shown that a group of Tbx18expressing cells in the proepicardium can migrate to the heart and give rise to myocytes in the ventricular septum and walls of the atria and ventricle (Cai et al., 2008). These results

Chapter  |  9.4  T-Box Factors

are broadly consistent with an independent study using the expression of the transcription factor WT1 as a lin­eage marker (Zhou et al., 2008). Assuming that Tbx18 is not expressed in the primary or second heart field-derived cardiomyocytes, these studies suggest a role for Tbx18 in the recruitment of epicardial-derived cardiac progenitor cells during embryogenesis (Cai et al., 2008) (see Chapter 5.1). In addition to its expression in the proepicardium during the later stages of mouse heart development, Tbx18 is expressed in the sinus horns, which give rise to cells that comprise the inflow tract. At the later stages of development, Tbx18 is also upregulated in the left ventricle and on the left side of the developing interventricular septum, which will divide the ventricular chambers into the right and left sides (Christoffels et al., 2006; Franco et al., 2006). Genetic analysis of Tbx18 mutant mice suggest that Tbx18 is required in the same tissues of the heart in which it is expressed, with homozygous Tbx18-null embryos displaying defects in the maturation of the cardiac inflow tract. Although initially normal, the inflow tracts of Tbx18 mutant mice have common cardinal veins that remain embedded in the surrounding mesenchyme, and fail to express molecular markers associated with myocardial tissue (Christoffels et al., 2006). Although these studies establish a role for Tbx18 in the heart, the function of Tbx18 in the proepicardium remains unclear, because mice homozygous for mutations in Tbx18 still form a proepicardium, although it is not clear if each of the epicardial derivatives is formed and if cells in these lineages are present in correct numbers in the absence of Tbx18.

II.F.  Tbx20 Tbx20 was independently identified by three separate groups. Meins and colleagues searched the human genome for sequences homologous to the Drosophila T-box gene H15 (Meins et al., 2000); Ahn et al. conducted low-stringency screens of zebrafish cDNA libraries with a mixture of Tbx1, Tbx3 and Tbx5 mouse cDNAs (Ahn et al., 2000; Ruvinsky et al., 2000); and Griffin et al. used a PCR-based approach (Griffin et al., 2000). These studies resulted in the identification of human, mouse and zebrafish orthologs of Tbx20, and demonstrated an enrichment of Tbx20 transcripts in developing heart tissues. Shortly thereafter, orthologs of Tbx20 were identified in chick and Xenopus, where they were also shown to be expressed within the developing heart (Griffin et al., 2000; Iio et al., 2001; Brown et al., 2003). In all species examined, Tbx20expression begins in the anterior lateral plate mesoderm and gradually becomes restricted to the Nkx2.5-expressing cardiac primordia prior to ventral migration and convergence at the midline. Expression is maintained in the primary heart field throughout the processes of migration, looping and cardiac chamber formation. Thus, Tbx20

657

is expressed at the same time, and in many of the same regions of the primary heart field, as the cardiac markers Tbx5, Nkx2.5 and Gata4. Similar to Tbx5 and other T-box gene family members, Tbx20 is localized to the nucleus, binds to DNA in a sequence-specific fashion, and modulates transcription of downstream target genes (Stennard et al., 2003; Brown et al., 2005; Cai et al., 2005; Singh et al., 2005; Stennard et al., 2005; Takeuchi et al., 2005). Results of genetic analysis and protein depletion studies are consistent with a role for Tbx20 during the early stages of vertebrate heart development; hearts lacking Tbx20 show a progressive loss of cardiomyocytes, a failure of the heart to undergo looping and chamber formation, and defects in cardiomyocyte maturation (Brown et al., 2005; Cai et al., 2005; Singh et al., 2005; Stennard et al., 2005; Takeuchi et al., 2005). Despite the early cardiac expression of Tbx20, and the downregulation of early heart markers such as Nkx2.5 and Tbx5, embryos lacking Tbx20 still form a primary heart tube which appears to be correctly patterned along the anterior–posterior axis, since polarized expression of numerous cardiac molecular markers, including -MyHC, -MyHC, Tbx5 and Gata4 along the anterior–posterior axis of the developing heart is unaltered in Tbx20 mutants. Rather, the cardiac abnormalities in Tbx20-depleted and mutant embryos appear to be the consequence of a primary requirement for Tbx20 in the proliferation and maturation of cardiomyocytes mediated, at least in part, by the direct regulation of the Tbx20 target gene, Tbx2 (Brown et al., 2005; Cai et al., 2005; Singh et al., 2005; Stennard et al., 2005; Takeuchi et al., 2005). The inhibition of Tbx2 expression by Tbx20 is thought to relieve the direct repression of N-myc by Tbx2 to collectively regulate the embryonic cardiac cell-cycle (Cai et al., 2005) (see below). These cellular defects lead to a decrease in the number of terminally-differentiated cardiomyocytes, and inability of the heart to undergo proper looping and chamber maturation, as indicated by the absence of expression of Gja5, Hey2, Smpx and Nppa (Brown et al., 2005; Cai et al., 2005; Singh et al., 2005; Stennard et al., 2005; Takeuchi et al., 2005). Consistent with detailed expression analysis, the requirement for Tbx20 appears to be restricted to the cardiomyocytes derived from the primary heart field; there does not appear to be a requirement for Tbx20 in the growth of the heart due to the recruitment of cardiomyocytes from the anterior or second heart field. This theory is supported by the unaltered expression pattern of second heart field markers such as Fgf8, Fgf10, Foxh1 and Isl1 in Tbx20 mutants (Brown et al., 2005; Cai et al., 2005; Singh et al., 2005; Stennard et al., 2005; Takeuchi et al., 2005). More recent studies have demonstrated that human patients with dilated cardiomyopathy (DCM), ASD, or mitral valve disease carry mutations in TBX20, while upregulation of TBX20 gene expression has been reported in patients with tetralogy of Fallot. Of the 12 mutant forms of TBX20 associated with congenital heart disease, eight

658

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

are missense mutations within the T-box domain, two of which are the identical mutation in unrelated individuals. Of the remaining four TBX20 mutations, one is a missense mutation leading to truncation of TBX20 in the T-box domain, and the other three are missense mutations mapping to two different regions of unknown function (Kirk et al., 2007; Hammer et al., 2008; Liu et al., 2008; Qian et al., 2008). Additional studies are necessary to determine the precise molecular mechanisms whereby altered TBX20 function could lead to congenital or adult cardiovascular disease. The role of Tbx20 in heart formation appears to be evolutionarily ancient, with Drosophila having two Tbx20 orthologs, neuromancer1 (nmr1) and neuromancer2 (nmr2), which are also referred to as H15 and midline, respectively (see Chapter 1.2). Like Tbx20, this pair of genes is required for proper development of the dorsal vessel, a structure thought to be homologous to the vertebrate heart (Miskolczi-McCallum et al., 2005; Qian et al., 2005; Reim et al., 2005). Neuromancer1 is one of the earliest markers of the cardioblast lineage; nmr2 is subsequently expressed in the same cells at later developmental stages (Miskolczi-McCallum et al., 2005; Qian et al., 2005; Reim et al., 2005). Embryos lacking nmr1 have normal or mild dorsal vessel defects, while those lacking nmr1 and/or nmr2 have a relatively weak dorsal vessel phenotype. However, expression of the Nkx2.5 ortholog tinman (tin) in the double mutants is almost completely abolished in cardiomyocytes, while expression of the cardioblast markers Dorsocross (Doc), which is homologous to Tbx2, and evenskipped (eve), is expanded at the expense of the pericardial cell fate (Qian et al., 2005; Reim et al., 2005). Moreover, this effect of nmr1 and nmr2 on the cardioblast lineage is dependent on the function of the GATA factor pannier (Qian et al., 2005). Consistent with the roles of nmr1 and nmr2 in embryogenesis and with studies of humans with congenital heart disease-associated TBX20 mutations, adult flies which lack nmr1 or have reduced nmr2 show compromised cardiac performance, displaying an increase in the rate of pacing-induced arrest which increases with age (Qian et al., 2008). Taken together, these studies suggest that the Tbx20 orthologs nmr1 and nmr2 are involved, along with tin and pannier, in the specification, functional diversification and maturation of Drosophila cardioblasts. Like Tbx5, Tbx20 appears to have distinct tissue-specific roles in subsets of cells during vertebrate cardiac development. In addition to its role in cardiomyocyte maturation, Tbx20 is also involved in the proliferation, migration and differentiation of heart valve precursor cells of the endocardial cushions (Shelton and Yutzey, 2007, 2008). Overexpression of Tbx20 in avian endocardial cushion cell cultures leads to an increase in the proliferation and cell migration of the cushion mesenchyme, with a concomitant increase in the expression of the matrix metalloproteinases

(MMP) mmp9 and mmp13 and decreased expression of the chondroitin sulfate proteoglycans (CSPG) aggrecan and versican. Depletion of Tbx20 has the opposite effect on the proliferation, migration and expression of these genes, leading to a decrease in mmp9 and mmp13 expression and an increase in the CSPGs. These alterations in gene expression may be mediated by the same Tbx20 downstream pathways that are involved in cardiomyocyte proliferation and maturation, since like the function of Tbx20 earlier in development they are associated with alterations in the levels of N-myc (Shelton and Yutzey, 2007, 2008). In zebrafish, in addition to its role in cardiac development Tbx20 is also required in the developing vasculature, with the zebrafish ortholog of Tbx20, HrT, being expressed and functioning in the dorsal aorta (Ahn et al., 2000). In the absence of Tbx20, the endothelial cells are specified but fail to fuse correctly along the midline. In addition, the embryos fail to undergo sprouting of the intersegmental vessels. Interestingly, this effect occurs only in the trunk of the embryo (Szeto et al., 2002), suggesting that there are redundant pathways in the more anterior aspects of the embryo.

III.  T-Box genes and the cardiac   cell-cycle Although it is not possible to draw general conclusions about the functions of all members of a single large protein family such as the T-box family, mounting evidence suggests that many members of this family, including Tbx1, Tbx2, Tbx3, Tbx5 and Tbx20 play a critical role in the regulation of the cell-cycle regulation (Fig. 3). Direct evidence for a role for T-box genes in cell-cycle regulation was first demonstrated in studies of human TBX2 and TBX3, which showed that TBX2 maps to a chromosomal region that is frequently mutated in ovarian carcinoma (Campbell et al., 1995; Law et al., 1995); this region is also amplified in pancreatic cancer cells (Mahlamaki et al., 2002) and is overexpressed in BRCA1 and BRCA2 breast tumors (Sinclair et al., 2002). Shortly thereafter it was shown that TBX3 is also overexpressed in breast tumor cell lines (Fan et al., 2004), and can interact with Ras and Myc to abrogate apoptosis (Carlson et al., 2002). Ectopic expression of Tbx2 leads to polyploidy (Davis et al., 2008), and Tbx2 can also directly interact with mitotic chromatin and the histone H3 terminal tail (Demay et al., 2007). Furthermore, it has been demonstrated that both Tbx2 and Tbx3 can inhibit senescence by direct transcriptional regulation of the cell-cycle inhibitors p19ARF and p21cip1, suggesting a role for Tbx2 and Tbx3 in regulation of late S or early G2-phase (Jacobs et al., 2000; Brummelkamp et al., 2002; Lingbeek et al., 2002; Prince et al., 2004; Vance et al., 2005). With regard to heart development, zebrafish that

659

Chapter  |  9.4  T-Box Factors

Cardiac growth factors

TBX20

Cyclin Cdk 2 E2

Cytoplasm

TBX5

Cyclin D1 Cdk 4/6

Nmyc

TBX2

P

Nucleus P RB P

P

E2F

DP

P

P

P

P

p21/p27

RB

P

P

RB

P

P E2F DP

G1

S

Figure 3  Model of the predicted role for Tbx2, Tbx5 and Tbx20 in the regulation of the embryonic cardiac cell-cycle. The T-box containing proteins Tbx2, Tbx5 and Tbx20 are proposed to function by the direct and indirect regulation of the G1-to-S phase of the cardiac cell-cycle. These effects are mediated, at least in part, through cross-talk and an interaction between each of these proteins and N-myc, cyclin D:Cdk2 and/or cyclin E:Cdk4/6 genes. See text for references and details.

lack Tbx2 or Tbx3 show alterations in cardiac proliferation (Ribeiro et al., 2007). However, the possibility of a role for Tbx2 in regulating the cardiac cell-cycle remains in question, since mice genetically lacking Tbx2 do not show altered levels of p19ARF, p21 cip1, p16INK, or p15 INK RNA in vivo (Harrelson et al., 2004). Although Tbx3 homozygous mutant mice do not display cardiac structural malformations, one possible explanation to reconcile these apparent differences is that, in the absence of Tbx2, Tbx3 is upregulated in the tissues that normally express Tbx2, thereby maintaining normal levels of this subset of cell-cycle inhibitors. Thus, it will be interesting to observe levels of Tbx3 in Tbx2 mutant mice, and to see whether more severe cardiac phenotypes occur in Tbx2;Tbx3 double mutants.

In addition to Tbx2 and Tbx3, there is also growing evidence that Tbx5 can regulate the cardiomyocyte cellcycle and that Tbx1 can regulate proliferation of a subset of cells in the second heart field (Hatcher et al., 2001; Xu et al., 2004; Brown et al., 2005; Goetz et al., 2006; Zhang et al., 2006; Liao et al., 2008). It has been shown that in early heart development Tbx5 can both promote and inhibit cardiomyocyte proliferation. Depletion of Tbx5 in Xenopus leads to a G1/S-phase arrest associated with the upregulation of cardiac cell-cycle-associated S-phase proteins, including CDC6, cyclin E2, SLBP and PCNA. In contrast, overexpression of Tbx5 results in an increase in the embryonic cardiac mitotic index, collectively suggesting that Tbx5 is both necessary and sufficient to control the G1-to-S transition, at least at early stages of heart

660

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

development (Brown et al., 2005; Goetz et al., 2006; Goetz and Conlon, 2007). In contrast, once the heart has ­undergone chamber formation, Tbx5 appears to have the opposite role of inhibiting cardiomyocyte proliferation in quail cardiomyocytes and in embryonic chicken hearts (Hatcher et al., 2001). This discrepancy may be explained in part by studies demonstrating that Tbx5 exists as two alternate splice forms in the cardiac cell lines TC13 and HL1. The splice variants differ in their carboxyl termini, with the previously identified Tbx5 being 518 amino acids (AA) in length and a C-terminal-truncated variant, containing the N-terminal 251AA of the long variant and an additional four residues at the C-terminus, resulting in a length of 255AA. The shorter form contains the DNA-binding domain, but lacks the transcriptional activation domain. Intriguingly, expression of the long isoform correlates with growth stimulation, while the short isoform appears to be associated with growth arrest and cell death (Georges et al., 2008). Thus, Tbx5 appears to directly or indirectly control a central component of the cell-cycle pathway that functions in a temporally-distinct fashion depending on the state of cardiomyocyte differentiation. Studies have demonstrated that Tbx5 and Tbx20 can genetically, physically and functionally interact (Brown et al., 2005). Consistent with this finding, phenotypic analysis of mouse mutants and studies involving depletion of Tbx20 in Xenopus and zebrafish are all suggestive of a role for Tbx20 in controlling cardiac cell proliferation. Studies using the mouse Tbx20 mutants reveal that one function of Tbx20 appears to be repression of Tbx2 expression; Tbx2 is capable of repressing expression of N-myc, a regulator of cellular proliferation (Cai et al., 2005; Stennard et al., 2005). Thus, loss of Tbx20 would result in overexpression of Tbx2, repression of N-myc and corresponding repression of cellular proliferation (Cai et al., 2005; Singh et al., 2005; Stennard et al., 2005; Takeuchi et al., 2005). Although these data strongly imply a role for T-box genes in controlling cardiac cell-cycle progression, defining a precise role for any one member of the family awaits the identification of its downstream targets.

IV.  T-Box regulation of cardiac gene expression IV.A.  T-Box Proteins Act as Repressors and Activators T-box genes are defined by the presence of the T-box, which encodes the T-domain DNA-binding domain of approximately 180 amino acids (Bollag et al., 1994). Molecular analyses of mouse Brachyury demonstrated DNA-binding of the T-domain to a palindromic sequence with a core of AGGTG, and Brachyury activates transcription of a concatamerized consensus reporter construct

in transfected cells (Kispert and Hermann, 1993; Kispert et al., 1995). Multiple activation and repression domains were identified, and complex regulation of nuclear localization was observed with analysis of Brachyury-GAL4 fusion proteins (Kispert et al., 1995; Conlon et al., 2001). Crystallographic studies showed that the Brachyury T-domain binds DNA as a homodimer, and identified amino acid residues that contact DNA in both major and minor grooves (Muller and Herrmann, 1997). These initial studies of Brachyury protein structure, DNA-binding activity and transcriptional function have proven to be representative of properties of the T-box gene family as a whole. Tbx5, like Brachyury, binds a core sequence of GGTGT, which has been defined as a T-box-binding element (TBE) (Bruneau et al., 2001; Ghosh et al., 2001). Similarly, site selection studies of Tbx3 have defined a binding consensus site of nGTGnnAn (Coll et al., 2002). Unlike Brachyury, Tbx5 and Tbx3 generally bind a single consensus site that corresponds to half of the palindromic Brachyury binding consensus, and these sites are often adjacent to binding sites for other cardiac transcription factors, such as GATA4 or Nkx2.5 (Bruneau et al., 2001; Stennard et al., 2003; Plageman and Yutzey, 2004). TBEcontaining regulatory elements have been identified in a limited number of genes, including Nppa, Gja5 and Gja1, which are expressed preferentially in maturing chamber myocardium or conduction system lineages (Fig. 4). These T-box target genes contain multiple TBEs, several of which are adjacent to Nkx2.5- and GATA4-binding sequences (Bruneau et al., 2001; Chen et al., 2004; Hoogaars et al., 2004; Plageman and Yutzey, 2004). Tbx2, Tbx3, Tbx5, Tbx18 and Tbx20 have all been shown to bind and differentially affect transcription of Nppa regulatory sequences, thereby confirming modulatory regulatory functions for T-box proteins that are co-expressed in different compartments of the developing heart (Bruneau et al., 2001; Habets et al., 2002; Stennard et al., 2003; Farin et al., 2007). T-box proteins expressed in the heart include transcriptional activators such as Tbx5 and Tbx1 (Bruneau et al., 2001; Plageman and Yutzey, 2004; Xu et al., 2004), as well as the repressors, Tbx2 and Tbx3 (Carreira et al., 1998; Sinha et al., 2000; Carlson et al., 2001; Habets et al., 2002; Hoogaars et al., 2004). Tbx20 has weaker transcriptional regulatory activity and has been reported to have either activator or repressor function, depending on the cell type or molecular context (Stennard et al., 2003; Plageman and Yutzey, 2004; Cai et al., 2005; Takeuchi et al., 2005). In contrast, Tbx18 has been shown to be a potent transcriptional repressor. Consistent with these observations, structure–function analysis of Tbx18 identified an engrailed homology 1 domain that is necessary and sufficient for direct interaction of Tbx18 with members of the Groucho subfamily of protein co-repressors (Farin et al., 2007). Although these studies offer insight into the molecular mechanism of Tbx18 function, relatively little is

Chapter  |  9.4  T-Box Factors

661

Figure 4  Transcriptional regulatory domains and downstream target genes of T-box proteins expressed in the developing heart. Protein domains and their relative positions in T-box transcription factors are represented with the conserved T-box DNA-binding domain shown in black and strong transcriptional activation (A) and repression (R) domains indicated. Transcriptional targets activated or repressed by indicated T-box proteins are also shown. See text for details and references.

known of the transcriptional regulatory function or downstream targets of the Tbx18 protein. GAL4 fusion experiments were used to localize a strong transactivation domain to a 40-amino acid sequence in the C-terminus of Tbx5 (Plageman and Yutzey, 2004; Zaragoza et al., 2004). Tbx1 has comparable activator function to Tbx5, and transactivation by Xenopus Tbx1 also requires the C-terminal region of the protein (Sinha et al., 2000; Xu et al., 2004; Ataliotis et al., 2005). Transcriptional-repression domains have been identified in both the N- and C-termini of Tbx2 and Tbx3; however, these proteins also contain activation domains that may confer added complexity on gene regulation (He et al., 1999; Paxton et al., 2002). Likewise, both activator and repressor domains have been identified in Tbx20 protein sequences (Stennard et al., 2003). Overall, there is accumulating evidence that T-box genes expressed in the heart can have activator or repressor function depending on the promoter context or the presence of other transcriptional co-factors.

IV.B.  T-Box Protein Transcriptional Partners The Nppa promoter has been used extensively to examine synergistic regulatory functions of T-box proteins with other cardiac transcription factors (Bruneau et al., 2001; Hiroi et al., 2001; Habets et al., 2002; Garg et al., 2003; Stennard et al., 2003; Plageman and Yutzey, 2004). The rat Nppa proximal 5 regulatory sequence (-288) contains three TBEs in close proximity to two Nkx-binding elements (NKEs) and two GATA consensus sequences. Tbx5 and Tbx18 bind to and synergize with Nkx2.5 or GATA4 in cooperative regulation of this element (Bruneau et al.,

2001; Hiroi et al., 2001; Garg et al., 2003; Stennard et al., 2003; Plageman and Yutzey, 2004; Ching et al., 2005; Farin et al., 2008). Tbx2 also binds and represses Nppa regulatory sequences, and has been proposed to be a competitor of Tbx5 for regulation of Nkx2.5 interactions (Habets et al., 2002; Christoffels et al., 2004). Similarly, TBE, NKE and GATA consensus sites are all present in proximal Gja5 sequences, but Tbx5 appears to antagonize Nkx2.5/GATA4-mediated activation of gene expression in this molecular context (Linhares et al., 2004). Therefore, T-box regulation of regionalized gene expression in the heart is likely to involve titration of levels of protein family members having different transcriptional regulatory functions and partner affinities. T-box proteins interact with a variety of co-factors that further modulate their transcriptional functions and subcellular localization. Several co-activators of Tbx5 have been identified, including the WW-domain protein TAZ, which interacts with histone acetyltransferases; Baf60c, which is part of Swi/Snf chromatin remodeling complexes; and Tip60, a histone acetylase (Lickert et al., 2004; Barron et al., 2005; Murakami et al., 2005) (see Chapters 10.1 and 10.2). Moreover, histone deacetylase 1 (HDAC1) is a corepressor of Tbx2 in melanoma cells (Vance et al., 2005). Moreover, Tbx5, and possibly other T-box containing proteins, may be regulated by their cellular localization through interactions with the nuclear export protein CRM1 (Kulisz and Simon, 2008). Tbx5 transcriptional activity is further regulated by interaction with the PDZ-LIM-domain protein LMP4, which affects Tbx5 subcellular localization in subpopulations of developing cardiomyocytes, epicardial precursors and endocardial cushion cells (Krause et al., 2004; Camarata et al., 2006; Bimber et al., 2007). In the cytoplasm, LMP4 localizes Tbx5 to the actin cytoskeleton,

662

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

but the regulatory implications of this localization have not been fully determined (Camarata et al., 2006). LMP4 does not interact with Tbx2 or Tbx3, and it remains to be seen if similar complex regulatory interactions govern the activities of other T-box proteins expressed in the developing heart.

IV.C.  T-Box Protein Downstream Target Genes Studies in mice and cultured cells have indicated that maturation of the heart chamber myocardium is regulated through differential expression of Tbx5, Tbx2 and Tbx20 (Habets et al., 2002; Stennard et al., 2003; Plageman and Yutzey, 2004). Direct regulation of Nppa and Gja5 by Tbx5 activator activity and Tbx2 repressor activity has been demonstrated (Bruneau et al., 2001; Habets et al., 2002; Christoffels et al., 2004). Recent reports describing targeted mutation of Tbx20 have revealed additional direct targets of T-box proteins in the developing heart (Cai et al., 2005; Singh et al., 2005; Stennard et al., 2005; Takeuchi et al., 2005). Increased Tbx2 expression was observed in these mice, and TBEs present in Tbx2 flanking sequences mediate repression by Tbx20 in transfection studies (Cai et al., 2005). Multiple TBEs sensitive to repression by Tbx2 or Tbx20 were also identified in N-myc flanking sequences and decreased expression of N-myc in Tbx20-null mice was associated with decreased cell proliferation (Cai et al., 2005). In transfection assays, Tbx20 also affects expression of Isl1, Nkx2.5, Pitx2, Fgf10, -MyHC and Mef2c regulatory elements that contain conserved TBE consensus sequences (Cai et al., 2005; Takeuchi et al., 2005). Additional evidence for regulation of cardiac transcription factor gene expression by T-box proteins includes regulation of serum reponse factor (SRF) by Tbx2 and of zebrafish Gata4 by Tbx5 (Heicklen-Klein and Evans, 2004; Barron et al., 2005). Taken together, T-box target genes identified in the heart to date provide strong supporting evidence for critical functions for T-box proteins in cardiac patterning, differentiation and growth. Several additional genes are affected by T-box function in mutant embryos or culture experiments, but it is not known if regulation is by direct or indirect mechan­ isms. Expression of Sall4, a zinc-finger transcription factor, is reduced in Tbx5-null embryos, and Sall4 enhances Tbx5 transactivation of Nppa, Gja5 and Fgf10 promoters in co-transfection assays (Koshiba-Takeuchi et al., 2006). Expression of the hairy-related transcription factor Hey2 is also reduced in Tbx5-null embryos and is increased with ectopic expression of Tbx5 in cultured cardiac cells; but direct regulation through a TBE in Hey2 regulatory sequences has not yet been established (Bruneau et al., 2001; Plageman and Yutzey, 2006). Microarray studies of mouse embryos with varying levels of Tbx5 expression or of a cardiac cell line with ectopic expression of

Tbx5 have identified a large number of Tbx5-reponsive genes having various functions in cell differentiation, gene ­regulation, proliferation, adhesion, signal transduction and metabolism (Mori et al., 2006; Plageman and Yutzey, 2006). Further studies are necessary to determine if these genes are directly regulated by Tbx5, or if the observed changes in transcription occur through indirect mechanisms. Microarray analysis has also been used to identify candidate Tbx1 target genes, and Tbx20 gain- and lossof-function analyses in valve progenitor cells have shown altered expression of extracellular matrix structural proteins and remodeling enzymes (Ivins et al., 2005; Shelton and Yutzey, 2007). These candidate target genes form the basis for elucidation of extensive networks of genes that are directly and indirectly regulated by T-box proteins in the developing heart.

V.  T-Box regulatory networks V.A.  Upstream Regulatory Pathways That Control T-Box Gene Expression During heart development, T-box gene expression is subject to regulation by a variety of growth factors and signaling molecules. In many cases, the T-box proteins in turn regulate the expression of these same activating, growth factor genes representing feedback or feedforward regulatory mechanisms. In the primary heart field, BMP2 induces expression of Tbx2, Tbx3 and Tbx20, but does not induce Tbx5 (Yamada et al., 2000; Plageman and Yutzey, 2004). Similarly, BMP signaling is required for Tbx2 expression in the AVC myocardium, and BMP2 treatment induces Tbx20 expression in AVC endocardial cushion cells (Ma et al., 2005; Rivera-Feliciano and Tabin, 2006; Shelton and Yutzey, 2007). The expression of Tbx20 in valve progenitor cells is affected by altered function of the bHLH transcription factor Twist1, which is also induced by BMP2 signaling (Ma et al., 2005; Shelton and Yutzey, 2008). Cardiac expression of Bmp2, Bmp4 and Bmp5 genes is altered in Tbx20-null embryos, demonstrating feedback regulation of BMP signaling and Tbx20 function (Cai et al., 2005; Singh et al., 2005; Stennard et al., 2005; Takeuchi et al., 2005). Altered Bmp2 expression has not been reported with loss of Tbx2, but Tbx2 misexpression inhibits hey1 and hey2 expression in the AVC, which can in turn repress expression of Bmp2 (Harrelson et al., 2004; Rutenberg et al., 2006). Tbx18 expression in proepicardial cells is also sensitive to BMP2 signaling, but downstream regulation of signaling pathways by Tbx18 has not been reported (Schlueter et al., 2006). Tbx1 and Tbx5 participate in regulatory networks with a variety of signaling molecules and patterning genes. In the pharyngeal arches, Tbx1 gene expression is regulated by sonic hedgehog acting through forkhead transcription

663

Chapter  |  9.4  T-Box Factors

factors, and Tbx1 regulates expression of Fgf8 and Fgf10 (Garg et al., 2001; Vitelli et al., 2002b; Yamagishi et al., 2003; Hu et al., 2004; Maeda et al., 2006; Vitelli et al., 2006). While Tbx1 and FGF8 appear to function together in aortic arch development, Tbx1 expression in the pharyngeal arches is not dependent on intact Fgf8 alleles (AbuIssa et al., 2002; Vitelli et al., 2002b). Similarly, Tbx5 is required for Fgf10 expression in the developing limb, but Tbx5 regulation of Fgf gene expression in the developing heart has not been established (Agarwal et al., 2003). Retinoic acid (RA) can repress Tbx1 expression in the pharyngeal arches, and retinoid-metabolizing enzymes are decreased in Tbx1-null mice, demonstrating a feedback regulatory interaction (Roberts et al., 2005, 2006). In contrast, Tbx5 expression in the posterior primitive heart tube is reduced in aldh1a2 (also called raldh2)-null mice, which lack retinoid signaling, and expression is increased in RA-treated chick embryos (Liberatore et al., 2000; Niederreither et al., 2001). Tbx5 does not appear to be in a feedback regulatory relationship with RA signaling, since changes in expression of associated signaling molecules are not predominant in microarray screens of cells or embryos with altered Tbx5 function (Mori et al., 2006; Plageman and Yutzey, 2006). Together, these signaling interactions involving Tbx1 and Tbx5 are important for regulation of proliferation, survival, differentiation and patterning of local cell populations in the developing heart and associated structures.

V.B.  Cross-Talk Among T-Box Gene Family Members T-box regulatory networks often include regulatory interactions among different T-box genes and proteins. The most extensively characterized interaction is Tbx20 repression of Tbx2 expression in the primitive heart tube, which has been described as a hierarchy of repression (Cai et al., 2005; Singh et al., 2005; Stennard et al., 2005; Stennard and Harvey, 2005). In zebrafish embryos gain- and lossof-function studies support a similar repressor function of Tbx20 (hrT) on Tbx5 gene expression in the primitive heart tube, but this interaction was not observed in Tbx20-null mice (Szeto et al., 2002). In Xenopus embryos, expression of Tbx20 and Tbx5 are mutually-independent, but Tbx5 can bind Tbx20 protein, which modulates its transcriptional activity on Nppa regulatory sequences and affects heart morphogenesis (Brown et al., 2005); Tbx18 has been shown to repress the same transcriptional targets activated by Tbx5 (Farin et al., 2008). Microarray analyses of Tbx5 mutant mice demonstrate that levels of Tbx3 expression are sensitive to the dosage of Tbx5, and Tbx5 has also been reported to regulate its own expression through a TBE located in proximal promoter regulatory sequences (Sun et al., 2004; Mori et al., 2006). Based on

the evidence available, complex regulatory relationships at the level of coordinate gene regulation and protein–protein interactions are important features of regulation of heart development by T-box proteins.

VI.  T-Box factors and congenital heart malformations Several human disorders have been linked to mutations in T-box genes (Packham and Brook, 2003). Cardiac developmental anomalies are characteristic of Holt-Oram syndrome, caused by mutations in TBX5, and DiGeorge syndrome, often caused by microdeletions of regions of chromosome 22 that include TBX1. Mutations in TBX20 are associated with a spectrum of valvuloseptal anomalies, and increased cardiac expression of TBX20 was observed in patients with tetralogy of Fallot (Kirk et al., 2007; Hammer et al., 2008). While several other T-box genes are important for cardiac development, as demonstrated in animal model systems, it is surprising that mutations in these genes have not been reported in patients with congenital heart disease. Mutations in TBX3 cause UMS, but cardiac defects are not characteristic of this patient population (Meneghini et al., 2006). Currently there are no reports of mutations in TBX2 or TBX18 associated with cardiac anomalies, but these genes are certainly strong candidates in ongoing clinical studies of familial heart disease.

VI.A.  Holt-Oram Syndrome The first indication of the importance of T-box genes in heart development and congenital heart malformations was the identification of human TBX5 as the gene that causes HoltOram syndrome (HOS) (Basson et al., 1997; Li et al., 1997). Holt-Oram syndrome has an incidence of 1 per 100,000 live births and includes shortened forelimbs and heart defects such as ASD, VSD and conduction system anomalies (Basson et al., 1994; Newbury-Ecob et al., 1996; Sletten and Pierpont, 1996; Mori and Bruneau, 2004). Holt-Oram syndrome is an autosomal-dominant genetic condition and the original TBX5 mutant alleles identified encode truncated TBX5 proteins or proteins with amino acid substitutions (Basson et al., 1997; Li et al., 1997). Additional TBX5 alleles have been identified in Holt-Oram syndrome patients, and at least 37 different TBX5 mutations have been associated with the syndrome (Fig. 5) (Ispording et al., 2004; Clark et al., 2006). These mutations are clustered in or near the T-box DNA-binding domain, and many result in truncated or nonfunctional proteins. Additional mutations include chromosomal deletions which encompass all or part of the TBX5 coding region, including a mutation which extends the coding region of TBX5, as well as one putative gain-of-function mutation (Akrami et al., 2001; Le Meur et al., 2005; Borozdin et al., 2006a,b; Bohm et al., 2008; Postma et al., 2008).

664

* G169R * E190X * G195A * S196X * K197fs ins * T223M * R237Q & R237W * D241fs del

243 325 340

379

518

TA R420_S432del*

Q265fs del* A268fs del* R279X* E316X*

NLS S251I* S261C*

T-box

Q49K*

Int2–2C-->A* Int2+1G-->C*

G33fs del ∗∗ G33fs ins

NLS

Int4 + 2T -->C*

78 90

A381_S408del* P386fs ins*

* G133fs ins * Y136X * H138fs del * S141fs del * Q151fs del

* W121 G

53

TBX5

* I54T * W64X * E69X* * G80R * R82fs ins

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Figure 5  Mutations in human TBX5 associated with Holt-Oram syndrome (HOS). The structural domains of TBX5 include the T-box, nuclear localization signals (NLS) and transactivation domain (TA). Human mutations in TBX5 coding sequence identified in patients with Holt-Oram syndrome are indicated (Purple: missense; Red: nonsense; Black: deletions; Green: insertions; Blue: intronic mutations). Clark et al. (2006). See text for details and references.

The cardiac defects observed with Holt-Oram syndrome are consistent with embryonic expression of TBX5 in the developing atrial and interventricular septa, as well as the specialized conduction system, as has been demonstrated in animal model systems (Bruneau et al., 1999; Liberatore et al., 2000; Moskowitz et al., 2004). Mice with heterozygous loss of Tbx5 exhibit ASDs and conduction system anomalies characteristic of Holt-Oram syndrome (see above), and have proven to be a valuable model for investigation of disease mechanisms (Bruneau et al., 2001; Moskowitz et al., 2004). Mutations in human NKX2.5 and GATA4, which encode transcription factors that interact with TBX5 to regulate gene expression, are also associated with ASDs and conduction system anomalies (Schott et al., 1998; Basson et al., 1999; Garg et al., 2003). Similarly, mutations in human SALL4, which cooperates with TBX5, have been identified in patients initially diagnosed with Holt-Oram syndrome (Koshiba-Takeuchi et al., 2006). Together, these human genetic studies, with supporting work in animal models, demonstrate that TBX5 deficiency or deficiency of TBX5-interacting factors can lead to congenital heart malformations. Although much progress has been made in elucidating TBX5 function in heart development, the precise molecular mechanisms and developmental events that lead to Holt-Oram syndrome-related congenital heart disease are not completely understood. TBX5 Holt-Oram syndrome mutations have been located throughout the coding sequence and include deletions, insertions and rearrangements, as well as nonsense and missense mutations (Mori and Bruneau, 2004; Clark et al., 2006). It is hypothesized that most TBX5 Holt-Oram syndrome mutations result in haploinsufficiency, which is supported by demonstration of ASDs and conduction system anomalies in Tbx5-heterozygous mice (Bruneau et al., 2001). However, TBX5 proteins derived from HoltOram syndrome alleles exhibit compromised DNA binding, nuclear localization and interactions with NKX2.5 or

GATA4, which may represent dominant-negative or hypomorphic allele mechanisms (Ghosh et al., 2001; Hiroi et al., 2001; Collavoli et al., 2003; Fan et al., 2003a,b; Garg et al., 2003; Plageman and Yutzey, 2004). In addition, the dosesensitivity of heart development to Tbx5 is supported by the observation that congenital heart defects in mice heterozygous for a hypomorphic Tbx5 allele were less severe than those observed in Tbx5 haploinsufficient mice (Mori et al., 2006). A striking observation from extensive clinical studies of human Holt-Oram syndrome patients is the lack of genotype–phenotype correlation for specific HoltOram syndrome mutations and resulting congenital heart defects (Brassington et al., 2003; Mori and Bruneau, 2004; Clark et al., 2006). Targeted mutation of Tbx5 to introduce Holt-Oram syndrome alleles in mice may be informative of genotype–phenotype associations in a more controlled genetic background, but these studies have not yet been carried out.

VI.B.  DiGeorge Syndrome Microdeletions in chromosome 22 are associated with cardiac and craniofacial abnormalities characteristic of DiGeorge/CATCH22/velocardiofacial syndromes (Baldini, 2004). In most cases, the heterozygous deletion eliminates approximately 3 Mbp of the long arm of chromosome 22, resulting in the loss of an estimated 30 genes (Baldini, 2004). The spectrum of associated phenotypes includes aortic arch abnormalities, conotruncal defects, characteristic facies, ear defects, hypoplasia of the thymus and para­ thyroid glands, and neurobehavioral conditions. Deletion studies in mice were used to define a DiGeorge critical region, syntenic to human chromosome 22, that is assoc­ iated with aortic arch defects and cardiac anomalies characteristic of the syndrome (Lindsay et al., 1999). Further candidate gene analysis and mutagenesis studies in mice

665

Chapter  |  9.4  T-Box Factors

implicated Tbx1 as a critical gene for DiGeorge syndrome phenotypes (Jerome and Papaioannou, 2001; Lindsay et al., 2001; Merscher et al., 2001; Hu et al., 2004). Homozygous loss of Tbx1 leads to cardiac and craniofacial phenotypes representative of the syndrome, including aortic arch anomalies, cardiac conotruncal defects, cleft palate, and hypoplasia of the thymus and parathyroid glands. Therefore, Tbx1 function is required for development of structures affected in DiGeorge/CATCH22/velocardiofacial syndromes. DiGeorge syndrome results from a heterozygous deletion of regions of chromosome 22, which is consistent with haploinsufficiency of candidate genes causing the characteristic developmental anomalies. However, heterozygous loss of Tbx1 in mice only partially recapitulates the phenotype with minor aortic arch malformations (Jerome and Papaioannou, 2001; Lindsay et al., 2001; Merscher et al., 2001; Hu et al., 2004). Therefore, haploinsufficiency of modifier genes in the region, such as Fgf8 or Crckl, in addition to loss of a Tbx1 allele, has been proposed as contributing to the full spectrum of defects associated with DiGeorge syndrome (Vitelli et al., 2002b; Moon et al., 2006). Attempts to further associate TBX1 with DiGeorge syndrome have led to searches for mutations in this gene in patients lacking the typical chromosomal deletion. To this end, five patients have been identified as carrying only a TBX1 gene mutation (Yagi et al., 2003). Although these individuals do not exhibit all characteristics of DiGeorge syndrome, this demonstrates that mutations in human TBX1 are capable of causing many of the defects associated with del22q11. Almost all of the identified TBX1 point mutations associated with the syndromes are in the C-terminal transactivation–repression domain coding sequence, and an additional mutation results in deletion of a novel nuclear localization signal (Ispording et al., 2004; Stoller and Epstein, 2005). The implications of these mutations for TBX1 function in cardiac or craniofacial development have not been reported. While modifier genes may contri­ bute to DiGeorge syndrome anomalies, the identification of TBX1 coding sequence point mutations in patients with the phenotype supports TBX1 as the causative gene.

VII.  Summary and future   directions Since the initial association of TBX5 with human HoltOram syndrome was reported more than a decade ago, much progress has been made in elucidating the functions of T-box genes in normal and abnormal heart formation. T-box genes expressed in the heart have important functions in the development of most, if not all, cardiac cell lineages and anatomical structures. Studies in a variety of animal systems, including Drosophila, Xenopus, zebrafish, chick and mouse, as well as human genetic studies, have shown

that T-box proteins are critical for cell differentiation, patterning, proliferation and migration in the developing heart. Members of the T-box gene family are dynamically expressed in distinct and overlapping patterns in myocardial, endothelial and epicardial cell lineages, as well as in valvuloseptal, venous and arterial structures of the heart. Individual T-box family members have distinct transcriptional functions and dosage effects, which supports coordinated regulation of target genes by the balance of multiple T-box proteins where they are co-expressed. T-box proteins function with other cardiac transcription factors, including Nkx2.5 and GATA4, in normal heart development to regulate target gene expression. Likewise, human genetic studies provide evidence that compromised functions of any of the proteins in these regulatory complexes can lead to similar spectra of congenital heart malformations. There are many questions remaining regarding how multiple T-box proteins function in complex regulatory networks to orchestrate essentially all features of cardiac organogenesis. A surprisingly limited number of T-box target genes have been identified, but genomic approaches have provided an abundance of candidate genes for further analysis. The study of the regulatory elements of these target genes should provide novel insights into T-box specificity and dosage effects, as well as co-factor functions. While mutant mice lacking specific T-box proteins have been valuable in identifying critical functions in differentiation and patterning of the primitive heart tube, less is known of T-box functions in the later events of valvuloseptal development and chamber maturation. More targeted conditional manipulation of T-box function in specific regions or cell lineages of the heart later in development should provide new insights in this area. These studies are critical to the understanding of how compromised function of T-box proteins or their interacting factors can lead to a wide array of congenital structural malformations or functional anomalies in the developing heart.

Acknowledgments We thank Timothy Plageman, Elaine Shelton and Woody Benson for help with the preparation of figures, and Erika Paden for help with proofreading. Work from our laboratories cited in this article was supported by the American Heart Association and the National Institute of Health.

References Abu-Issa, R., Smyth, G., Smoak, I., Yamamura, K., Meyers, E.N., 2002. Fgf8 is required for pharyngeal arch and cardiovascular development in the mouse. Development 129, 4613–4625. Agarwal, P., Wylie, J.N., Galceran, J., Arkhitko, O., Li, C., Deng, C., Grosschedl, R., Bruneau, B.G., 2003. Tbx5 is essential for forelimb bud initiation following patterning of the limb field in the mouse embryo. Development 130, 623–633.

666

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Agulnik, S.I., Garvey, N., Hancock, S., Ruvinsky, I., Chapman, D. L., Agulnik, I., Bollag, R., Papaioannou, V., Silver, L.M., 1996. Evolution of mouse T-box genes by tandem duplication and cluster dispersion. Genetics 144, 249–254. Agulnik, S.I., Ruvinsky, I., Silver, L.M., 1997. Three novel T-box genes in Caenorhabditis elegans. Genome 40, 458–464. Ahn, D., Ruvinsky, I., Oates, A.C., Silver, L.M., Ho, R.K., 2000. Tbx20, a new vertebrate T-box gene expressed in the cranial motor neurons and developing cardiovascular structures. Mech. Dev. 95, 253–258. Akrami, S.M., Winter, R.M., Brook, J.D., Armour, J.A., 2001. Detection of a large TBX5 deletion in a family with Holt-Oram syndrome. J. Med. Genet. 38, E44. Ataliotis, P., Ivins, S., Mohun, T.J., Scambler, P.J., 2005. XTbx1 is a transcriptional activator involved in head and pharyngeal arch development in Xenopus laevis. Dev. Dyn. 232, 979–991. Baldini, A., 2004. DiGeorge syndrome: an update. Curr. Op. Cardiol. 19, 201–204. Bamshad, M., Krakowiak, P.A., Watkins, W.S., Root, S., Carey, J.C., Jorde, L.B., 1995. A gene for ulnar-mammary syndrome maps to 12q23–q24.1 Hum. Mol. Genet. 4, 1973–1977. Bamshad, M., Lin, R.C., Law, D.J., Watkins, W.C., Krakowiak, P.A., Moore, M.E., Franceschini, P., Lala, R., Holmes, L.B., Gebuhr, T.C., Bruneau, B.G., Schinzel, A., Seidman, J.G., Seidman, C.E., Jorde, L. B., 1997. Mutations in human TBX3 alter limb, apocrine and genital development in ulnar-mammary syndrome. Nat. Genet. 16, 311–315. Bamshad, M., Le, T., Watkins, W.S., Dixon, M.E., Kramer, B.E., Roeder, A.D., Carey, J.C., Root, S., Schinzel, A., Van Maldergem, L., Gardner, R.J., Lin, R.C., Seidman, C.E., Seidman, J.G., Wallerstein, R., Moran, E., Sutphen, R., Campbell, C.E., Jorde, L.B., 1999. The spectrum of mutations in TBX3: Genotype/phenotype relationship in ulnar-mammary syndrome. Am. J. Hum. Genet. 64, 1550–1562. Barron, M.R., Belaguli, N.S., Zhang, S.X., Trinh, M., Iyer, D., Merlo, X., Lough, J.W., Parmacek, M.S., Bruneau, B.G., Schwartz, R.J., 2005. Serum response factor, an enriched cardiac mesoderm obligatory factor, is a downstream target for Tbx genes. J. Biol. Chem. 280, 11816–11828. Basson, C.T., Cowley, G.S., Solomon, S.D., Weissman, B., Poznanski, A.K., Traill, T.A., Seidman, J.G., Seidman, C.E., 1994. The clinical and genetic spectrum of the Holt-Oram syndrome (Heart–Hand syndrome). New Engl. J. Med. 330, 885–891. Basson, C.T., Bachinsky, D.R., Lin, R.C., Levi, T., Elkins, J.A., Soults, J., Grayzel, D., Kroumpouzou, E., Traill, T.A., Leblanc-Straceski, J., Renault, B., Kucherlapati, R., Seidman, J.G., Seidman, C.E., 1997. Mutations in human TBX5 cause limb and cardiac malformation in Holt-Oram syndrome. Nat. Genet. 15, 30–35. Basson, C.T., Huang, T., Lin, R.C., Bachinsky, D.R., Weremowicz, S., Vaglio, A., Bruzzone, R., Quadrelli, R., Lerone, M., Romeo, G., Silengo, M., Pereira, A., Krieger, J., Mesquita, S.F., Kamisago, M., Morton, C.C., Pierpont, M.E., Müller, C.W., Seidman, J.G., Seidman, C.E., 1999. Different TBX5 interactions in heart and limb defined by Holt-Oram syndrome mutations. Proc. Natl. Acad. Sci. USA 96, 2919–2924. Begemann, G., Ingham, P.W., 2000. Developmental regulation of Tbx5 in zebrafish embryogenesis. Mech. Dev. 90, 299–304. Begemann, G., Gibert, Y., Meyer, A., Ingham, P.W., 2002. Cloning zebrafish T-box genes tbx15 and tbx18 and their expression during embryonic development. Mech. Dev. 114, 137–141. Bimber, B., Dettman, R.W., Simon, H.G., 2007. Differential regulation of Tbx5 protein expression and sub-cellular localization during heart development. Dev. Biol. 302, 230–242.

Bohm, J., Heinritz, W., Craig, A., Vujic, M., Ekman-Joelsson, B.M., Kohlhase, J., Froster, U., 2008. Functional analysis of the novel TBX5 c.1333delC mutation resulting in an extended TBX5 protein. BMC Med. Genet. 9, 88. Bollag, R.J., Siegfried, Z., Cebra-Thomas, J.A., Garvey, N., Davison, E.M., Silver, L.M., 1994. An ancient family of embryonically expressed mouse genes sharing a conserved protein motif with the T locus. Nat. Genet. 7, 383–389. Borozdin, W., Bravo-Ferrer Acosta, A.M., Seemanova, E., Leipoldt, M., Bamshad, M.J., Unger, S., Kohlhase, J., 2006a. Contiguous hemizygous deletion of TBX5, TBX3, and RBM19 resulting in a combined phenotype of Holt-Oram and ulnar-mammary syndromes. Am. J. Med. Genet. A 140A, 1880–1886. Borozdin, W., Bravo Ferrer Acosta, A.M., Bamshad, M.J., Botzenhart, E.M., Froster, U.G., Lemke, J., Schinzel, A., Spranger, S., McGaughran, J., Wand, D., Chrzanowska, K.H., Kohlhase, J., 2006b. Expanding the spectrum of TBX5 mutations in Holt-Oram syndrome: detection of two intragenic deletions by quantitative real time PCR, and report of eight novel point mutations. Hum. Mutat. 27, 975–976. Brassington, A.E., Sung, S.S., Toydemir, R.M., Le, T., Roeder, A.D., Rutherford, A.E., Whitby, F.G., Jorde, L.B., Bamshad, M.J., 2003. Expressivity of Holt-Oram syndrome is not predicted by TBX5 genotype. Am. J. Hum. Genet. 73, 74–85. Brown, C.B., Wenning, J.M., Lu, M.M., Epstein, D.J., Meyers, E.N., Epstein, J.A., 2004. Cre-mediated excision of Fgf8 in the Tbx1 expression domain reveals a critical role for Fgf8 in cardiovascular development in the mouse. Dev. Biol. 267, 190–202. Brown, D.D., Binder, O., Pagratis, M., Parr, B.A., Conlon, F.L., 2003. Developmental expression of the Xenopus laevis Tbx20 orthologue. Dev. Genes Evol. 212, 604–607. Brown, D.D., Martz, S.N., Binder, O., Goetz, S.C., Price, B.M., Smith, J. C., Conlon, F.L., 2005. Tbx5 and Tbx20 act synergistically to control vertebrate heart morphogenesis. Development 132, 553–563. Brummelkamp, T.R., Kortlever, R.M., Lingbeek, M., Trettel, F., MacDonald, M.E., van Lohuizen, M., Bernards, R., 2002. TBX3, the gene mutated in Ulnar-Mammary Syndrome, is a negative regulator of p19ARF and inhibits senescence. J. Biol. Chem. 277, 6567–6572. Bruneau, B.G., Logan, M., Davis, N., Levi, T., Tabin, C.J., Seidman, J.G., Seidman, C.E., 1999. Chamber-specific cardiac expression of Tbx5 and heart defects in Holt-Oram syndrome. Dev. Biol. 211, 100–108. Bruneau, B.G., Nemer, G., Schmitt, J.P., Charron, F., Robitaille, L., Caron, S., Conner, D.A., Gessler, M., Nemer, M., Seidman, C. E., Seidman, J.G., 2001. A murine model of Holt-Oram syndrome defines roles of the T-box transcription factor Tbx5 in cardiogenesis and disease. Cell 106, 709–721. Cai, C.L., Zhou, W., Yang, L., Bu, L., Qyang, Y., Zhang, X., Li, X., Rosenfeld, M.G., Chen, J., Evans, S., 2005. T-box genes coordinate regional rates of proliferation and regional specification during cardiogenesis. Development 132, 2475–2487. Cai, C.L., Martin, J.C., Sun, Y., Cui, L., Wang, L., Ouyang, K., Yang, L., Bu, L., Liang, X., Zhang, X., Stallcup, W.B., Denton, C.P., McCulloch, A., Chen, J., Evans, S.M., 2008. A myocardial lineage derives from Tbx18 epicardial cells. Nature 454, 104–108. Camarata, T., Bimber, B., Kulisz, A., Chew, T.-L., Yeung, J., Simon, H. G., 2006. LMP4 regulates Tbx5 protein subcellular localization and activity. J. Cell. Biol. 174, 339–348. Campbell, C., Goodrich, K., Casey, G., Beatty, B., 1995. Cloning and mapping of a human gene (TBX2) sharing a highly conserved protein motif with the Drosophila omb gene. Genomics 28, 255–260.

Chapter  |  9.4  T-Box Factors

Carlson, H., Ota, S., Campbell, C.E., Hurlin, P.J., 2001. A dominant repression domain in Tbx3 mediates transcriptional repression and cell immortalization: Relevance to mutations in Tbx3 that cause ulnar-mammary syndrome. Hum. Mol. Genet. 10, 2403–2413. Carlson, H., Ota, S., Song, Y., Chen, Y., Hurlin, P.J., 2002. Tbx3 impinges on the p53 pathway to suppress apoptosis, facilitate cell transformation and block myogenic differentiation. Oncogene 21, 3827–3835. Carreira, S., Dexter, T.J., Yavuzer, U., Easty, D.J., Goding, C.R., 1998. Brachyury-related transcription factor Tbx2 and repression of the melanocyte-specific TRP-1 promoter. Mol. Cell Biol. 18, 5099–5108. Chapman, D.L., Garvey, N., Hancock, S., Alexiou, M., Agulnik, S. I., Gibson-Brown, J., Cebra-Thomas, J., Bollag, R., Silver, L.M., Papaioannou, V., 1996. Expression of the T-box family genes, Tbx1Tbx5, during early mouse development. Dev. Dyn. 206, 379–390. Chen, J.R., Chatterjee, B., Meyer, R., Yu, J.C., Borke, J.L., Isales, C.M., Kirby, M.L., Lo, C.W., Bollag, R.J., 2004. Tbx2 represses expression of connexin43 in osteoblastic-like cells. Calcif. Tissue Int. 74, 561–573. Ching, Y.H., Ghosh, T.K., Cross, S.J., Packham, E.A., Honeyman, L., Loughna, S., Robinson, T.E., Dearlove, A.M., Ribas, G., Bonser, A. J., Thomas, N.R., Scotter, A.J., Caves, L.S., Tyrrell, G.P., NewburyEcob, R.A., Munnich, A., Bonnet, D., Brook, J.D., 2005. Mutation in myosin heavy chain 6 causes atrial septal defect. Nat. Genet. 37, 423–428. Christoffels, V.M., Hoogaars, W.M., Tessari, A., Clout, D.E., Moorman, A.F., Campione, M., 2004. T-box transcription factor Tbx2 represses differentiation and formation of the cardiac chambers. Dev. Dyn. 229, 763–770. Christoffels, V.M., Mommersteeg, M.T., Trowe, M.O., Prall, O.W., de Gier-de Vries, C., Soufan, A.T., Bussen, M., Schuster-Gossler, K., Harvey, R.P., Moorman, A.F., Kispert, A., 2006. Formation of the venous pole of the heart from an Nkx2-5-negative precursor population requires Tbx18. Circ. Res. 98, 1555–1563. Clark, K.L., Yutzey, K.E., Benson, D.W., 2006. Transcription factors and congenital heart defects. Annu. Rev. Physiol. 68, 97–121. Coll, M., Seidman, J.G., Muller, C.W., 2002. Structure of the DNA-bound T-box domain of human TBX3, a transcription factor responsible for Ulnar-Mammary syndrome. Structure 10, 343–356. Collavoli, A., Hatcher, C.J., He, J., Okin, D., Deo, R., Basson, C.T., 2003. Tbx5 nuclear localization is mediated by dual cooperative intra­ molecular signals. J. Mol. Cell Cardiol. 35, 1191–1195. Conlon, F.L., Fairclough, L., Price, B.M., Casey, E.S., Smith, J.C., 2001. Determinants of T box protein specificity. Development 128, 3749–3758. Davenport, T.G., Jerome-Majewska, L.A., Papaioannou, V.E., 2003. Mammary gland, limb and yolk sac defects in mice lacking Tbx3, the gene mutated in human ulnar mammary syndrome. Development 130, 2263–2273. Davis, E., Teng, H., Bilican, B., Parker, M.I., Liu, B., Carriera, S., Goding, C.R., Prince, S., 2008. Ectopic Tbx2 expression results in polyploidy and cisplatin resistance. Oncogene 27, 976–984. Demay, F., Bilican, B., Rodriguez, M., Carreira, S., Pontecorvi, M., Ling, Y., Goding, C.R., 2007. T-box factors: targeting to chromatin and interaction with the histone H3 N-terminal tail. Pigment. Cell Res. 20, 279–287. Fan, C., Duhagon, M.A., Oberti, C., Chen, S., Hiroi, Y., Komuro, I., Duhagon, P.I., Canessa, R., Wang, Q., 2003a. Novel TBX5 mutations and molecular mechanism for Holt-Oram syndrome. J. Med. Genet. 40, e29.

667

Fan, C., Liu, M., Wang, Q., 2003b. Functional analysis of TBX5 missense mutations associated with Holt-Oram syndrome. J. Biol. Chem. 278, 8780–8785. Fan, W., Huang, X., Chen, C., Gray, J., Huang, T., 2004. TBX3 and its isoform TBX32a are functionally distinctive in inhibition of senescence and are overexpressed in a subset of breast cancer cell lines. Cancer Res. 64, 5132–5139. Farin, H.F., Bussen, M., Schmidt, M.K., Schuster-Gossler, K., Kispert, A., 2007. Transcriptional repression by the T-box proteins Tbx18 and Tbx15 depends on Groucho corepressors. J. Biol. Chem. 282, 25748–25759. Farin, H.F., Mansouri, A., Petry, M., Kispert, A., 2008. T-box protein Tbx18 interacts with the paired box protein Pax3 in the development of the paraxial mesoderm. J. Biol. Chem. 283, 25372–25380. Fijnvandraat, A.C., Lekanne Deprez, R.H., Christoffels, V.M., Ruijter, J. M., Moorman, A.F., 2003. TBX5 overexpression stimulates differentiation of chamber myocardium in P19Cl6 embryonic carcinoma cells. J. Muscle. Res. Cell Motil. 24, 211–218. Franco, D., Meilhac, S.M., Christoffels, V.M., Kispert, A., Buckingham, M.E., Kelly, R.G., 2006. Left and right ventricular contributions to the formation of the ventricular septum in the mouse heart. Dev. Biol. 294, 366–375. Garg, V., Yamagishi, C., Hu, T., Kathiriya, I.S., Yamagishi, H., Srivastava, D., 2001. Tbx1, a DiGeorge syndrome candidate gene, is regulated by sonic hedgehog during pharyngeal arch development. Dev. Biol. 235, 62–73. Garg, V., Kathiriya, I.S., Barnes, R., Schluterman, M.K., King, I.N., Butler, C.A., Rothrock, C.R., Eapen, R.S., Hirayama-Yamada, K., Joo, K., Matsuoka, R., Cohen, J.C., Srivastava, D., 2003. GATA4 mutations cause human congenital heart defects and reveal an interaction with TBX5. Nature 424, 443–447. Garrity, D.M., Childs, S., Fishman, M.C., 2002. The heartstrings mutation in zebrafish causes heart/fin Tbx5 deficiency syndrome. Development 129, 4635–4645. Georges, R., Nemer, G., Morin, M., Lefebvre, C., Nemer, M., 2008. Distinct expression and function of alternatively spliced Tbx5 isoforms in cell growth and differentiation. Mol. Cell Biol. 28, 4052–4067. Ghosh, T.K., Packham, E.A., Bonser, A.J., Robinson, T.E., Cross, S.J., Brook, J.D., 2001. Characterization of the TBX5 binding site and analysis of mutations that cause Holt-Oram syndrome. Hum. Mol. Genet. 10, 1983–1994. Gibson-Brown, J., Agulnik, S.I., Chapman, D.L., Alexiou, M., Garvey, N., Silver, L.M., Papaioannou, V., 1996. Evidence for a role for Tbox genes in the evolution of limb morphogenesis and the specification of forelimb/hindlimb identity. Mech. Dev. 56, 93–101. Gibson-Brown, J.J., Agulnik, S.I., Silver, L.M., Niswander, L., Papaioannou, V.E., 1998a. Involvement of T-box genes Tbx2-Tbx5 in vertebrate limb specification and development. Development 125, 2499–2509. Gibson-Brown, J.J., Agulnik, S.I., Silver, L.M., Papaioannou, V.E., 1998b. Expression of T-box genes Tbx2-Tbx5 during chick organogenesis. Mech. Dev. 74, 165–169. Goetz, S.C., Brown, D.D., Conlon, F.L., 2006. TBX5 is required for embryonic cardiac cell cycle progression. Development 133, 2575–2584. Goetz, S.C., Conlon, F.L., 2007. Cardiac progenitors and the embryonic cell cycle. Cell Cycle 6, 1974–1981. Griffin, K.J., Stoller, J., Gibson, M., Chen, S., Yelon, D., Stainier, D. Y., Kimelman, D., 2000. A conserved role for H15-related T-box

668

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

­transcription factors in zebrafish and Drosophila heart formation. Dev. Biol. 218, 235–247. Guris, D.L., Duester, G., Papaioannou, V.E., Imamoto, A., 2006. Dosedependent interaction of Tbx1 and Crkl and locally aberrant RA signaling in a model of del22q11 syndrome. Dev. Cell 10, 81–92. Habets, P.E., Moorman, A.F., Clout, D.E., van Roon, M.A., Lingbeek, M., van Lohuizen, M., Campione, M., Christoffels, V.M., 2002. Cooperative action of Tbx2 and Nkx2.5 inhibits ANF expression in the atrioventricular canal: implications for cardiac chamber formation. Genes Dev. 16, 1234–1246. Haenig, B., Kispert, A., 2004. Analysis of TBX18 expression in chick embryos. Dev. Genes Evol. 214, 407–411. Hammer, S., Toenjes, M., Lange, M., Fischer, J.J., Dunkel, I., Mebus, S., Grimm, C.H., Hetzer, R., Berger, F., Sperling, S., 2008. Characterization of TBX20 in human hearts and its regulation by TFAP2. J. Cell Biochem. 104, 1022–1033. Harrelson, Z., Kelly, R.G., Goldin, S.N., Gibson-Brown, J.J., Bollag, R.J., Silver, L.M., Papaioannou, V.E., 2004. Tbx2 is essential for patterning the atrioventricular canal and for morphogenesis of the outflow tract during heart development. Development 131, 5041–5052. Hatcher, C.J., Kim, M.S., Mah, C.S., Goldstein, M.M., Wong, B., Mikawa, T., Basson, C.T., 2001. TBX5 transcription factor regulates cell proliferation during cardiogenesis. Dev. Biol. 230, 177–188. Hatcher, C.J., Diman, N.Y., Kim, M.S., Pennisi, D., Song, Y., Goldstein, M.M., Mikawa, T., Basson, C.T., 2004. A role for Tbx5 in proepicardial cell migration during cardiogenesis. Physiol. Genom. 18, 129–140. Haworth, K., Putt, W., Cattanach, B., Breen, M., Binns, M., Lingaas, F., Edwards, Y.H., 2001. Canine homolog of the T-box transcription factor T; failure of the protein to bind to its DNA target leads to a shorttail phenotype. Mamm. Genome 12, 212–218. He, M., Wen, L., Campbell, C.E., Wu, J.Y., Rao, Y., 1999. Transcription repression by Xenopus ET and its human ortholog TBX3, a gene involved in ulnar-mammary syndrome. Proc. Natl. Acad. Sci. USA 96, 10212–10217. Heicklen-Klein, A., Evans, T., 2004. T-box binding sites are required for activity of a cardiac GATA-4 enhancer. Dev. Biol. 267, 490–504. Herrmann, B.G., Labiet, S., Poustka, A., King, T., Lehrach, H., 1990. Cloning of the T gene required in mesoderm formation in the mouse. Nature 343, 617–622. Hiroi, Y., Kudoh, S., Monzen, K., Ikeda, Y., Yazaki, Y., Nagai, R., Komuro, I., 2001. Tbx5 associates with Nkx2-5 and synergistically promotes cardiomyocyte differentiation. Nat. Genet. 28, 276–280. Hoogaars, W.M., Tessari, A., Moorman, A.F., de Boer, P.A., Hagoort, J., Soufan, A.T., Campione, M., Christoffels, V.M., 2004. The transcriptional repressor Tbx3 delineates the developing central conduction system of the heart. Cardiovasc. Res. 62, 489–499. Hoogaars, W.M., Engel, A., Brons, J.F., Verkerk, A.O., de Lange, F. J., Wong, L.Y., Bakker, M.L., Clout, D.E., Wakker, V., Barnett, P., Ravesloot, J.H., Moorman, A.F., Verheijck, E.E., Christoffels, V.M., 2007. Tbx3 controls the sinoatrial node gene program and imposes pacemaker function on the atria. Genes Dev. 21, 1098–1112. Horb, M.E., Thomsen, G.H., 1999. Tbx5 is essential for heart development. Development 126, 1739–1751. Hu, T., Yamagishi, H., Maeda, J., McAnally, J., Yamagishi, C., Srivastava, D., 2004. Tbx1 regulates fibroblast growth factor in the anterior heart field through a reinforcing regulatory autoregulatory loop involving forkhead transcription factors. Development 131, 5491–5502. Iio, A., Koide, M., Hidaka, K., Morisaki, T., 2001. Expression pattern of novel chick T-box gene, Tbx20. Dev. Genes Evol. 211, 559–562.

Ispording, I.D., Leylek, A.M., Yeung, J., Mischel, A., Simon, H.G., 2004. T-box genes and congenital heart/limb malformations. Clin. Genet. 66, 253–264. Ivins, S., Lammerts van Beuren, K., Roberts, C., James, C., Lindsay, E., Baldini, A., Ataliotis, P., Scambler, P.J., 2005. Microarray analysis detects differentially expressed genes in the pharyngeal region of mice lacking Tbx1. Dev. Biol. 285, 554–569. Jacobs, J.J., Keblusek, P., Robanus-Maandag, E., Kristel, P., Lingbeek, M., Nederlof, P.M., van Welsem, T., van de Vijver, M.J., Koh, E. Y., Daley, G.Q., van Lohuizen, M., 2000. Senescence bypass screen identifies TBX2, which represses cdkn2a (p19(ARF)) and is amplified in a subset of human breast cancers. Nat. Genet. 26, 291–299. Jahr, M., Schlueter, J., Brand, T., Manner, J., 2008. Development of the proepicardium in Xenopus laevis. Dev. Dyn. 237, 3088–3096. Jerome, L.A., Papaioannou, V.E., 2001. DiGeorge syndrome phenotype in mice mutant for the T-box gene, Tbx1. Nat. Genet. 27, 286–291. Kelly, R.G., Papaioannou, V.E., 2007. Visualization of outflow tract development in the absence of Tbx1 using an FgF10 enhancer trap transgene. Dev. Dyn. 236, 821–828. Kirk, E.P., Sunde, M., Costa, M.W., Rankin, S.A., Wolstein, O., Castro, M.L., Butler, T.L., Hyun, C., Guo, G., Otway, R., Mackay, J.P., Waddell, L.B., Cole, A.D., Hayward, C., Keogh, A., Macdonald, P., Griffiths, L., Fatkin, D., Sholler, G.F., Zorn, A.M., Feneley, M.P., Winlaw, D.S., Harvey, R.P., 2007. Mutations in cardiac T-box factor gene TBX20 are associated with diverse cardiac pathologies, including defects of septation and valvulogenesis and cardiomyopathy. Am. J. Hum. Genet. 81, 280–291. Kispert, A., Hermann, B.G., 1993. The Brachyury gene encodes a novel DNA binding protein. EMBO J. 12, 4898–4899. Kispert, A., Koschorz, B., Herrmann, B.G., 1995. The T protein encoded by Brachyury is a tissue-specific transcription factor. EMBO J. 14, 4763–4772. Koshiba-Takeuchi, K., Takeuchi, J.K., Arruda, E.P., Kathiriya, I.S., Mo, R., Hui, C.C., Srivastava, D., Bruneau, B.G., 2006. Cooperative and antagonistic interactions between Sall4 and Tbx5 pattern the mouse limb and heart. Nat. Genet. 38, 175–183. Kraus, F., Haenig, B., Kispert, A., 2001. Cloning and expression analysis of the mouse T-box gene Tbx18. Mech. Dev. 100, 83–86. Krause, A., Zacharias, W., Camarata, T., Linkhart, B., Law, E., Lischke, A., Miljan, E., Simon, H.G., 2004. Tbx5 and Tbx4 transcription factors interact with a new chicken PDZ-LIM protein in limb and heart development. Dev. Biol. 273, 106–120. Kulisz, A., Simon, H.G., 2008. An evolutionarily conserved nuclear export signal facilitates cytoplasmic localization of the Tbx5 transcription factor. Mol. Cell Biol. 28, 1553–1564. Law, D.J., Gebuhr, T., Garvey, N., Agulnik, S.I., Silver, L.M., 1995. Identification, characterization, and localization to chromosome 17q21-22 of the human TBX2 homolog, member of a conserved developmental gene family. Mamm. Genome 6, 793–797. Le Meur, N., Goldenberg, A., Michel-Adde, C., Drouin-Garraud, V., Blaysat, G., Marret, S., Amara, S.A., Moirot, H., Joly-Helas, G., Mace, B., Kleinfinger, P., Saugier-Veber, P., Frébourg, T., Rossi, A., 2005. Molecular characterization of a 14q deletion in a boy with features of Holt-Oram syndrome. Am. J. Med. Genet. A 134, 439–442. Lepilina, A., Coon, A.N., Kikuchi, K., Holdway, J.E., Roberts, R.W., Burns, C.G., Poss, K.D., 2006. A dynamic epicardial injury response supports progenitor cell activity during zebrafish heart regeneration. Cell 127, 607–619. Li, Q.Y., Newbury-Ecob, R.A., Terrett, J.A., Wilson, D.I., Curtis, A.R., Yi, C.H., Gebuhr, T., Bullen, P.J., Robson, S.C., Strachan, T., Bonnet, D.,

Chapter  |  9.4  T-Box Factors

Lyonnet, S., Young, I.D., Raeburn, J.A., Buckler, A.J., Law, D.J., Brook, J.D., 1997. Holt-Oram syndrome is caused by mutations in TBX5, a member of the Brachyury (T) gene family. Nat. Genet. 15, 21–29. Liao, J., Aggarwal, V.S., Nowotschin, S., Bondarev, A., Lipner, S., Morrow, B.E., 2008. Identification of downstream genetic pathways of Tbx1 in the second heart field. Dev. Biol. 316, 524–537. Liberatore, C.M., Searcy-Schrick, R.D., Yutzey, K.E., 2000. Ventricular expression of tbx5 inhibits normal heart chamber development. Dev. Biol. 223, 169–180. Lickert, H., Takeuchi, J.K., Von Both, I., Walls, J.R., McAuliffe, F., Adamson, S.L., Henkelman, R.M., Wrana, J.L., Rossant, J., Bruneau, B.G., 2004. Baf60c is essential for function of BAF chromatin remodelling complexes in heart development. Nature 432, 107–112. Lindsay, E.A., Botta, A., Jurecic, V., Carattini-Rivera, S., Cheah, Y.C., Rosenblatt, H.M., Bradley, A., Baldini, A., 1999. Congenital heart disease in mice deficient for the DiGeorge syndrome region. Nature 401, 379–383. Lindsay, E.A., Vitelli, F., Su, H., Morishima, M., Huynh, T., Pramparo, T., Jurecic, V., Ogunrinu, G., Sutherland, H.F., Scambler, P., Bradley, A., Baldini, A., 2001. Tbx1 haploinsufficiency in the DiGeorge syndrome region causes aortic arch defects in mice. Nature 410, 97–101. Lingbeek, M.E., Jacobs, J.J., van Lohuizen, M., 2002. The T-box repressors TBX2 and TBX3 specifically regulate the tumor suppressor gene p14ARF via a variant T-site in the initiator. J. Biol. Chem. 277, 26120–26127. Linhares, V.L., Almeida, N.A., Menezes, D.C., Elliott, D.A., Lai, D., Beyer, E.C., Campos de Carvalho, A.C., Costa, M.W., 2004. Transcriptional regulation of the murine Connexin40 promoter by cardiac factors Nkx2-5, GATA4 and Tbx5. Cardiovasc. Res. 64, 402–411. Liu, C., Shen, A., Li, X., Jiao, W., Zhang, X., Li, Z., 2008. T-box transcription factor TBX20 mutations in Chinese patients with congenital heart disease. Eur. J. Med. Genet. 51, 580–587. Ma, L., Lu, M.F., Schwartz, R.J., Martin, J.F., 2005. Bmp2 is essential for cardiac cushion epithelial–mesenchymal transition and myocardial patterning. Development 132, 5601–5611. Maeda, J., Yamagishi, H., McAnally, J., Yamagishi, C., Srivastava, D., 2006. Tbx1 is regulated by forkhead proteins in the secondary heart field. Dev. Dyn. 235, 701–710. Mahlamaki, E.H., Barlund, M., Tanner, M., Gorunova, L., Hoglund, M., Karhu, R., Kallioniemi, A., 2002. Frequent amplification of 8q24, 11q, 17q, and 20q-specific genes in pancreatic cancer. Genes Chromosomes Cancer 35, 353–358. Mandel, E.M., Callis, T.E., Wang, D.Z., Conlon, F.L., 2005. Transcriptional mechanisms of congenital heart disease. Drug. Disc. Today: Dis. Mech. 2, 33–38. Meins, M., Henderson, D.J., Bhattacharya, S.S., Sowden, J.C., 2000. Characterization of the human TBX20 gene, a new member of the T-box gene family closely related to the Drosophila H15 gene. Genomics 67, 317–332. Meneghini, V., Odent, S., Platonova, N., Egeo, A., Merlo, G.R., 2006. Novel TBX3 mutation data in families with Ulnar-Mammary syndrome indicate a genotype–phenotype relationship: mutations that do not disrupt the T-domain are associated with less severe limb defects. Eur. J. Med. Genet. 49, 151–158. Merscher, S., Funke, B., Epstein, J.A., Heyer, J., Puech, A., Lu, M. M., Xavier, R.J., Demay, M.B., Russell, R.G., Factor, S., Tokooya, K., Jore, B.S., Lopez, M., Pandita, R.K., Lia, M., Carrion, D., Xu, H., Schorle, H., Kobler, J.B., Scambler, P., Wynshaw-Boris, A.,

669

Skoultchi, A.I., Morrow, B.E., Kucherlapati, R., 2001. TBX1 is responsible for cardiovascular defects in velo-cardio-facial/DiGeorge syndrome. Cell 104, 619–629. Mesbah, K., Harrelson, Z., Theveniau-Ruissy, M., Papaioannou, V.E., Kelly, R.G., 2008. Tbx3 is required for outflow tract development. Circ. Res. 103, 743–750. Miskolczi-McCallum, C.M., Scavetta, R.J., Svendsen, P.C., Soanes, K. H., Brook, W.J., 2005. The Drosophila melanogaster T-box genes midline and H15 are conserved regulators of heart development. Dev. Biol. 278, 459–472. Moon, A.M., Guris, D.L., Seo, J.H., Li, L., Hammond, J., Talbot, A., Imamoto, A., 2006. Crk1 deficiency disrupts Fgf8 signaling in a mouse model of 22q11 deletion syndromes. Dev. Cell 10, 71–80. Moorman, A.F., Soufan, A.T., Hagoort, J., de Boer, P.A., Christoffels, V. M., 2004. Development of the building plan of the heart. Ann. N.Y. Acad. Sci. 1015, 171–181. Mori, A.D., Bruneau, B.G., 2004. TBX5 mutations and congenital heart disease: Holt-Oram syndrome revealed. Curr. Opin. Cardiol. 19, 211–215. Mori, A.D., Zhu, Y., Vahora, I., Nieman, B., Koshiba-Takeuchi, K., Davidson, L., Pizard, A., Seidman, J.G., Seidman, C.E., Chen, A.J., Henkelman, R.M., Bruneau, B.G., 2006. Tbx5-dependent rheostat control of cardiac gene expression and morphogenesis. Dev. Biol. 297, 566–586. Moskowitz, I.P., Pizard, A., Patel, V.V., Bruneau, B.G., Kim, J.B., Kupershmidt, S., Roden, D., Berul, C.I., Seidman, C.E., Seidman, J.G., 2004. The T-box transcription factor Tbx5 is required for the patterning and maturation of the murine cardiac conduction system. Development 131, 4107–4116. Muller, C.W., Herrmann, B.G., 1997. Crystallographic structure of the T domain-DNA complex of the Brachyury transcription factor. Nature 389, 884–888. Murakami, M., Nakagawa, M., Olson, E.N., Nakagawa, O., 2005. A WW domain protein TAZ is a critical coactivator for TBX5, a transcription factor implicated in Holt-Oram syndrome. Proc. Natl. Acad. Sci. USA 102, 18034–18039. Newbury-Ecob, R.A., Leanage, R., Raeburn, J.A., Young, I.D., 1996. Holt-Oram syndrome: a clinical genetic study. J. Med. Genet. 33, 300–307. Niederreither, K., Vermot, J., Messaddeq, N., Schuhbaur, B., Chambon, P., Dolle, P., 2001. Embryonic retinoic acid synthesis is essential for heart morphogenesis in the mouse. Development 128, 1019–1031. Oberpriller, J.O., Oberpriller, J.C., 1974. Response of the adult newt ventricle to injury. J. Exp. Zool. 187, 249–253. Oberpriller, J.O., Oberpriller, J.C., Matz, D.G., Soonpaa, M.H., 1995. Stimulation of proliferative events in the adult amphibian cardiac myocyte. Ann. N.Y. Acad. Sci. 752, 30–46. Ohuchi, H., Takeuchi, J., Yoshioka, H., Ishimaru, Y., Ogura, K., Takahashi, N., Ogura, T., Noji, S., 1998. Correlation of wing-leg identity in ectopic FGF-induced chimeric limbs with the differential expression of chick Tbx5 and Tbx4. Development 125, 51–60. Packham, E.A., Brook, J.D., 2003. T-box genes in human disorders. Hum. Mol. Genet. 12, R37–R44. Papaioannou, V.E., 2001. T-box genes in development: from hydra to humans. Int. Rev. Cytol. 207, 1–70. Papaioannou, V.E., Silver, L.M., 1998. The T-box gene family. Bioessays 20, 9–19. Paxton, C., Zhao, H., Chin, Y., Langner, K., Reecy, J., 2002. Murine Tbx2 contains domains that activate and repress gene transcription. Gene 283, 117–124.

670

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Paylor, R., Glaser, B., Mupo, A., Ataliotis, P., Spencer, C., Sobotka, A., Sparks, C., Choi, C.H., Oghalai, J., Curran, S., Murphy, K.C., Monks, S., Williams, N., O’Donovan, M.C., Owen, M.J., Scambler, P.J., Lindsay, E., 2006. Tbx1 haploinsufficiency is linked to behavioral disorders in mice and humans: implications for 22q11 deletion syndrome. Proc. Natl. Acad. Sci. USA 103, 7729–7734. Pflugfelder, G.O., Roth, H., Poeck, B., 1992. A homology domain shared between Drosophila optomotor-blind and mouse Brachyury is involved in DNA binding. Biochem. Biophys. Res. Commun. 186, 918–925. Plageman Jr., T.F., Yutzey, K.E., 2004. Differential expression and function of Tbx5 and Tbx20 in cardiac development. J. Biol. Chem. 279, 19026–19034. Plageman Jr., T.F., Yutzey, K.E., 2005. T-box genes and heart development: Putting the “T” in hearT. Dev. Dyn. 232, 11–20. Plageman Jr., T.F., Yutzey, K.E., 2006. Microarray analysis of Tbx5induced genes expressed in the developing heart. Dev. Dyn. 235, 2868–2880. Poss, K.D., Wilson, L.G., Keating, M.T., 2002. Heart regeneration in zebrafish. Science 298, 2188–2190. Poss, K.D., Keating, M.T., Nechiporuk, A., 2003. Tales of regeneration in zebrafish. Dev. Dyn. 226, 202–210. Postma, A.V., van de Meerakker, J.B., Mathijssen, I.B., Barnett, P., Christoffels, V.M., Ilgun, A., Lam, J., Wilde, A.A., Lekanne Deprez, R.H., Moorman, A.F., 2008. A gain-of-function TBX5 mutation is associated with atypical Holt-Oram syndrome and paroxysmal atrial fibrillation. Circ. Res. 102, 1433–1442. Prall, O.W., Elliot, D.A., Harvey, R.P., 2002. Developmental paradigms in heart disease: insights from tinman. Ann. Med. 34, 148–156. Prince, S., Carreira, S., Vance, K.W., Abrahams, A., Goding, C.R., 2004. Tbx2 directly represses the expression of the p21(WAF1) cyclindependent kinase inhibitor. Cancer Res. 64, 1669–1674. Qian, L., Liu, J., Bodmer, R., 2005. Neuromancer Tbx20-related genes (H15/midline) promote cell fate specification and morphogenesis of the Drosophila heart. Dev. Biol. 279, 509–524. Qian, L., Mohapatra, B., Akasaka, T., Liu, J., Ocorr, K., Towbin, J.A., Bodmer, R., 2008. Transcription factor neuromancer/TBX20 is required for cardiac function in Drosophila with implications for human heart disease. Proc. Natl. Acad. Sci. USA 105, 19833–19838. Reim, I., Mohler, J.P., Frasch, M., 2005. Tbx20-related genes mid and H15 are required for tinman expression, proper patterning, and normal differentiation of cardioblasts in Drosophila. Mech. Dev. 122, 1056–1069. Ribeiro, I., Kawakami, Y., Buscher, D., Raya, A., Rodriguez-Leon, J., Morita, M., Rodriguez Esteban, C., Izpisua Belmonte, J.C, 2007. Tbx2 and Tbx3 regulate the dynamics of cell proliferation during heart remodeling. PLoS ONE 2, e398. Rivera-Feliciano, J., Tabin, C.J., 2006. Bmp2 instructs cardiac progenitors to form the heart-valve-inducing field. Dev. Biol. 295, 580–588. Roberts, C., Ivins, S.M., James, C.T., Scambler, P.J., 2005. Retinoic acid down-regulates Tbx1 expression in vivo and in vitro. Dev. Dyn. 232, 928–938. Roberts, C., Ivins, S., Cook, A.C., Baldini, A., Scambler, P.J., 2006. Cyp26 genes a1, b1, and c1 are down-regulated in Tbx1 null mice and inhibition of Cyp26 enzyme function produces a phenocopy of DiGeorge Syndrome in the chick. Hum. Mol. Genet. 15, 3394–3410. Rutenberg, J.B., Fischer, A., Jia, H., Gessler, M., Zhong, T.P., Mercola, M., 2006. Developmental patterning of the cardiac atrioventricular canal by Notch and Hairy-related transcription factors. Development 133, 4381–4390.

Ruvinsky, I., Oates, A.C., Silver, L.M., Ho, R.K., 2000. The evolution of paired appendages in vertebrates: T-box genes in the zebrafish. Dev. Genes Evol. 210, 82–91. Ryan, K., Chin, A.J., 2003. T-box genes and cardiac development. Birth Defects Res. Part C Embryo. Today 69, 25–37. Schlueter, J., Manner, J., Brand, T., 2006. BMP is an important regulator of proepicardial identity in the chick embryo. Dev. Biol. 295, 546–558. Schott, J.J., Benson, D.W., Basson, C.T., Pease, W., Silberbach, G. M., Moak, J.P., Maron, B.J., Seidman, C.E., Seidman, J.G., 1998. Congenital heart disease caused by mutations in the transcription factor NKX2-5. Science 281, 108–111. Schulte, I., Schlueter, J., Abu-Issa, R., Brand, T., Manner, J., 2007. Morphological and molecular left–right asymmetries in the development of the proepicardium: a comparative analysis on mouse and chick embryos. Dev. Dyn. 236, 684–695. Shelton, E.L., Yutzey, K.E., 2007. Tbx20 regulation of endocardial cushion cell proliferation and extracellular matrix gene expression. Dev. Biol. 302, 376–388. Shelton, E.L., Yutzey, K.E., 2008. Twist1 function in endocardial cell proliferation, migration and differentiation during heart valve development. Dev. Biol. 317, 282–295. Showell, C., Binder, O., Conlon, F.L., 2004. T-box genes in early embryo­ genesis. Dev. Dyn. 229, 201–218. Showell, C., Christine, K.S., Mandel, E.M., Conlon, F.L., 2006. Developmental expression patterns of Tbx1, Tbx2, Tbx5 and Tbx20 in Xenopus tropicalis. Dev. Dyn. 235, 1623–1630. Sinclair, C.S., Adem, C., Naderi, A., Soderberg, C.L., Johnson, M., Wu, K., Wadum, L., Couch, V.L., Sellers, T.A., Schaid, D., Slezak, J., Fredericksen, Z., Ingle, J.N., Hartmann, L., Jenkins, R.B., Couch, F. J., 2002. TBX2 is preferentially amplified in BRCA1- and BRCA2related breast tumors. Cancer Res. 62, 3587–3591. Singh, M.K., Christoffels, V.M., Dias, J.M., Trowe, M.O., Petry, M., Schuster-Gossler, K., Burger, A., Ericson, J., Kispert, A., 2005. Tbx20 is essential for cardiac chamber differentiation and repression of Tbx2. Development 132, 2697–2707. Sinha, S., Abraham, S., Gronostajski, R.M., Campbell, C.E., 2000. Differential DNA binding and transcriptional modulation by three Tbox proteins, T, TBX1 and Tbx2. Gene 258, 15–29. Sletten, L.J., Pierpont, M.E., 1996. Variation in severity of cardiac disease in Holt-Oram syndrome. Am. J. Med. Genet. 65, 128–132. Smith, J., 1999. T-box genes: what they do and how they do it. Trends. Genet. 15, 154–158. Stennard, F.A., Costa, M.W., Elliot, D.A., Rankin, S., Haast, S.J., Lai, D., McDonald, L.P., Niederreither, K., Dolle, P., Bruneau, B.G., Zorn, A. M., Harvey, R.P., 2003. Cardiac T-box factor Tbx20 directly interacts with Nkx2-5, GATA4 and GATA5 in regulation of gene expression in the developing heart. Dev. Biol. 262, 206–224. Stennard, F.A., Harvey, R.P., 2005. T-box transcription factors and their roles in regulatory hierarchies in the developing heart. Development 132, 4897–4910. Stennard, F.A., Costa, M.W., Lai, D., Biben, C., Furtado, M.B., Solloway, M.J., McCulley, D.J., Leimana, C., Preis, J.I., Dunwoodie, S.L., Elliott, D.E., Prall, O.W., Black, B.L., Fatkin, D., Harvey, R.P., 2005. Murine T-box transcription factor Tbx20 acts as a repressor during heart development, and is essential for adult heart integrity, function and adaptation. Development 132, 2451–2462. Stoller, J.Z., Epstein, J.A., 2005. Identification of a novel nuclear loca­ lization signal in Tbx1 that is deleted in DiGeorge syndrome patients harboring the 1223delC mutation. Hum. Mol. Genet. 14, 885–892.

Chapter  |  9.4  T-Box Factors

Sun, G., Lewis, L.E., Huang, X., Nguyen, Q., Price, C., Huang, T., 2004. TBX5, a gene mutated in Holt-Oram syndrome, is regulated through a GC box and T-box binding elements (TBEs). J. Cell Biochem. 92, 189–199. Szeto, D.P., Griffin, K.J., Kimelman, D., 2002. HrT is required for cardiovascular development in zebrafish. Development 129, 5093–5101. Takeuchi, J.K., Ohgi, M., Koshiba-Takeuchi, K., Shiratori, H., Sakaki, I., Ogura, K., Saijoh, Y., Ogura, T., 2003. Tbx5 specifies the left/right ventricles and ventricular septum position during cardiogenesis. Development 130, 5953–5964. Takeuchi, J.K., Mileikovskaia, M., Koshiba-Takeuchi, K., Heidt, A.B., Mori, A.D., Arruda, E.P., Gertsenstein, M., Georges, R., Davidson, L., Mo, R., Hui, C.C., Henkelman, R.M., Nemer, M., Black, B.L., Nagy, A., Bruneau, B.G., 2005. Tbx20 dose-dependently regulates transcription factor networks required for mouse heart and motor neuron development. Development 132, 2463–2474. Theveniau-Ruissy, M., Dandonneau, M., Mesbah, K., Ghez, O., Mattei, M.G., Miquerol, L., Kelly, R.G., 2008. The del22q11.2 candidate gene Tbx1 controls regional outflow tract identity and coronary artery patterning. Circ. Res. 103, 142–148. Tiecke, E., Matsuura, M., Kokubo, N., Kuraku, S., Kusakabe, R., Kuratani, S., Tanaka, M., 2007. Identification and developmental expression of two Tbx1/10-related genes in the agnathan Lethenteron japonicum. Dev. Genes Evol. 217, 691–697. Tonissen, K.F., Drysdale, T.A., Lints, T.J., Harvey, R.P., Krieg, P.A., 1994. XNkx-2.5, a Xenopus gene related to Nkx-2.5 and tinman: evidence for a conserved role in cardiac development. Dev. Biol. 162, 325–328. Vance, K.W., Carreira, S., Brosch, G., Goding, C.R., 2005. Tbx2 is overexpressed and plays an important role in maintaining proliferation and suppression of senescence in melanomas. Cancer Res. 65, 2260–2268. Vitelli, F., Morishima, M., Taddei, I., Lindsay, E.A., Baldini, A., 2002a. Tbx1 mutation causes multiple cardiovascular defects and disrupts neural crest and cranial nerve migratory pathways. Hum. Mol. Genet. 11, 915–922. Vitelli, F., Taddei, I., Morishima, M., Meyers, E.N., Lindsay, E.A., Baldini, A., 2002b. A genetic link between Tbx1 and fibroblast growth factor signaling. Development 129, 4605–4611. Vitelli, F., Zhang, Z., Huynh, T., Sobotka, A., Mupo, A., Baldini, A., 2006. Fgf8 expression in the Tbx1 domain causes skeletal abnormalities and modifies the aortic arch but not the outflow tract phenotype of Tbx1 mutants. Dev. Biol. 295, 559–570.

671

Wilson, V., Conlon, F.L., 2002. The T-box family. Genome Biol. 3, 3008. Xu, H., Morishima, M., Wylie, J.N., Schwartz, R.J., Bruneau, B.G., Lindsay, E.A., Baldini, A., 2004. Tbx1 has a dual role in the morphogenesis of the cardiac outflow tract. Development 131, 3217–3227. Yagi, H., Furutani, Y., Hamada, H., Sasaki, T., Asakawa, S., Minoshima, S., Ichida, F., Joo, K., Kimura, M., Imamura, S., Kamatani, N., Momma, K., Takao, A., Nakazawa, M., Shimizu, N., Matsuoka, R., 2003. Role of TBX1 in human del22q11.2 syndrome. Lancet 362, 1366–1373. Yamada, M., Revelli, J.P., Eichele, G., Barron, M., Schwartz, R.J., 2000. Expression of chick Tbx-2, Tbx-3, and Tbx-5 genes during early heart development: evidence for BMP2 induction of Tbx2. Dev. Biol. 228, 95–105. Yamagishi, H., Srivastava, D., 2003. Unraveling the genetic and developmental mysteries of 22q11 deletion syndrome. Trends. Mol. Med. 9, 383–389. Yamagishi, H., Maeda, J., Hu, T., McAnally, J., Conway, S.J., Kume, T., Meyers, E.N., Yamagishi, C., Srivastava, D., 2003. Tbx1 is regulated by tissue-specific forkhead proteins through a common Sonic hedgehog-responsive enhancer. Genes Dev. 17, 269–281. Yi, C.H., Terrett, J.A., Li, Q.Y., Ellingham, K., Packham, E.A., Armstrong-Buisseret, L., McClure, P., Slingsby, T., Brook, J.D., 1999. Identification, mapping, and phylogenomic analysis of four new human members of the T-box gene family: EOMES, TBX6, TBX18, and TBX19. Genomics 55, 10–20. Zaragoza, M.V., Lewis, L.E., Sun, G., Wang, E., Li, L., Said-Salman, I., Feucht, L., Huang, T., 2004. Identification of the TBX5 transactivating domain and the nuclear localization signal. Gene 330, 9–18. Zhang, Z., Cerrato, F., Xu, H., Vitelli, F., Morishima, M., Vincentz, J., Furuta, Y., Ma, L., Martin, J.F., Baldini, A., Lindsay, E., 2005. Tbx1 expression in pharyngeal epithelia is necessary for pharyngeal arch artery development. Development 132, 5307–5315. Zhang, Z., Huynh, T., Baldini, A., 2006. Mesodermal expression of Tbx1 is necessary and sufficient for pharyngeal arch and cardiac outflow tract development. Development 133, 3587–3595. Zhou, B., Ma, Q., Rajagopal, S., Wu, S.M., Domian, I., Rivera-Feliciano, J., Jiang, D., von Gise, A., Ikeda, S., Chien, K.R., Pu, W.T., 2008. Epicardial progenitors contribute to the cardiomyocyte lineage in the developing heart. Nature 454, 109–113. Zhu, Y., Gramolini, A.O., Walsh, M.A., Zhou, Y.Q., Slorach, C., Friedberg, M.K., Takeuchi, J.K., Sun, H., Henkelman, R.M., Backx, P.H., Redington, A.N., Maclennan, D.H., Bruneau, B.G., 2008. Tbx5-dependent pathway regulating diastolic function in congenital heart disease. Proc. Natl. Acad. Sci. USA 105, 5519–5524.

Chapter 9.5

Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease Brian L. Black1 and Richard M. Cripps2 1

Cardiovascular Research Institute and Department of Biochemistry and Biophysics, University of California, San Francisco, CA, USA Department of Biology, University of New Mexico, Albuquerque, NM, USA

2

I.  Introduction Members of the myocyte enhancer factor 2 (MEF2) family of transcription factors are important regulators of gene expression in numerous tissues, including the heart, where MEF2 plays important roles in development and in postnatal adaptation to a wide array of physiological and pathological signals. MEF2 functions as a transcriptional switch, by potently activating or repressing transcription through interaction with a variety of co-factors which serve as positive and negative regulators of transcription. The interaction of MEF2 with its co-factors is controlled by a multitude of signaling pathways that result in posttranslational modification of MEF2, and in the subsequent MEF2-dependent repression or activation of target gene transcription. This allows MEF2 to link the extracellular environment to distinct and highly-regulated transcriptional outputs through intracellular signaling cascades and co-factor interactions. In the heart, MEF2 is essential for development and plays fundamental roles in myocyte differentiation and gene activation. MEF2 is also crucial in the postnatal heart for integrating the transcriptional response to numerous environmental cues, and regulating normal physiological and pathological growth and adaptation of the heart. In this chapter, we review what is known about the general regulation and function of MEF2 transcription factors, with a focus on their role in the heart. We discuss the many co-factors of MEF2 with a particular attention to the interaction of MEF2 with class II histone deacetylases (HDACs) (see Chapter 10.2). MEF2-HDAC interactions

Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

are highly-regulated by signaling cascades that control MEF2’s function as a transcriptional switch. We highlight the many signaling pathways and kinases that regulate MEF2-HDAC interactions, MEF2 post-translational modification, and how these pathways influence MEF2 activity. We also discuss the genetic function of Mef2 genes in flies, fish and mice. These studies demonstrate the essential function for MEF2 proteins in heart development, as well as the development of numerous other lineages, and reflect the general conservation of MEF2 function throughout much of metazoan evolution. Many genes have been identified as direct transcriptional targets of MEF2 through direct MEF2 binding to their promoter and enhancer elements, and we summarize the known direct transcriptional targets of MEF2 in the heart. The Mef2 genes themselves are regulated at the transcriptional level by the activity of multiple, independent modular enhancers. These discrete enhancer modules control Mef2 expression in a restricted subset of the gene’s complete expression pattern. We review the transcriptional regulation of the single Mef2 gene in Drosophila and the mouse Mef2c gene, which has been the best-characterized vertebrate Mef2 gene in terms of transcriptional regulation. These studies have uncovered many new roles for Mef2 genes in the heart and other tissues by identifying unexpected expression patterns and regulatory interactions upstream of MEF2. Finally, in this chapter, we highlight several important areas for future investigation regarding the role of MEF2 transcription factors, how they are regulated and function in the developing and postnatal heart, and their possible involvement in human cardiovascular disease.

673

674

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

II.  The mef2 family of transcription factors II.A.  Discovery of MEF2 Transcription Factors In the late-1980s and early-1990s, numerous transcription factors involved in skeletal muscle development were identified. These studies were highlighted by the seminal observations of Davis et al. who showed that the myogenic basic helix-loop-helix (bHLH) transcription factor MyoD alone could initiate skeletal myogenesis in a broad range of cell types in culture (Davis et al., 1987). MEF2 proteins were also identified during this time as essential regulators of skeletal muscle transcription, and as partners for MyoD and other myogenic bHLH proteins. In addition, it soon became apparent that MEF2 was a critical regulator of muscle development in all muscle lineages, including the heart (Black and Olson, 1998). MEF2 transcription factors were discovered independently using two different approaches. On the one hand, several groups had determined that muscle cells contained different DNA-binding activities that could interact with muscle structural gene promoter sequences in vitro and in vivo. MEF2-binding activity interacted strongly and specifically with an AT-rich sequence that was found in the promoters of numerous muscle-specific genes, and the integrity of these sites was required for full musclespecific gene activation (Gossett et al., 1989; Horlick and Benfield, 1989; Mueller and Wold, 1989). Meanwhile, Treisman and colleagues were studying the function of the serum response factor (SRF) protein. Serum response factor had been shown to be a potent activator of genes important to proliferation, as well as to differentiation of vascular muscle lineages, interacting with DNA via a conserved binding domain (Treisman, 1990). By screening for cDNAs similar to that of serum response factor, Pollock and Treisman identified factors named Related to SRF4 (RSRF4, now called MEF2D) and RSRFR2 (MEF2B) (Pollock and Treisman, 1991). These investigators went on to demonstrate that RSRF proteins probably corresponded to the muscle-specific binding activities previously defined as MEF2. Using a similar approach, Chambers et al. (1992) identified Xenopus SL-1 (MEF2D) and SL-2 (MEF2A). A more direct connection between MEF2 binding activity and the RSRF proteins was achieved by Yu et al. (1992), who used a concatamerized MEF2 binding site from the muscle creatine kinase (MCK) gene to screen a cDNA expression library for factors that bound to the MEF2 sequence. This resulted in the identification of human MEF2A. The final MEF2 family member, MEF2C, was cloned based on its similarity to existing MEF2 factors, and was the first to show significant enrichment in muscle tissues in the developing

embryo (Martin et al., 1993; Edmondson et al., 1994). It is now well-established that most vertebrate genomes contain at least four MEF2-encoding genes, whereas simpler animals such as Drosophila melanogaster, Caenorhabditis elegans and Ciona intestinalis contain only a single Mef2 gene each (Olson et al., 1995; Davidson, 2007).

II.B.  The MEF2 Family in the Context of the MADS Domain Superfamily MEF2 proteins share, with several other factors, an Nterminal 57-amino acid sequence termed the MADS domain, which is responsible for protein dimerization and sequence-specific DNA-binding (Shore and Sharrocks, 1995). The MADS domain is an acronym for the earliest-described members of the protein family: the yeast mating type regulator MCM1; the plant floral determinants Agamous and Deficiens/Apetala3; and the animal protein Serum response factor (SRF) (Black and Olson, 1998). Phylogenetic analyses have concluded that the MADS protein domain is ancient and originated in a common ancestor of prokaryotes and eukaryotes, based on similarities in primary structure between eukaryotic MADS proteins and the Escherichia coli universal stress protein UspA (Mushegian and Koonin, 1996). Current evolutionary models propose that an ancestral MADS-box gene was duplicated prior to the divergence of the plant and animal kingdoms, and these duplicates formed the founders of the two major classes of eukaryotic MADS domain proteins found today: type I and type II MADS domain proteins (Alvarez-Buylla et al., 2000). In higher animals, type I proteins are represented by serum response factor, which contains a conserved SAM domain immediately C-terminal to the MADS domain. The SAM domain functions in homoand heterodimerization (Ling et al., 1998). Interestingly, despite the ancient evolution of serum response factor proteins, sequenced animal genomes contain only one serum response factor gene member per haploid genome. By contrast, type II MADS domain proteins have diverged significantly to generate additional family members, particularly in plants (Theissen et al., 1996; Becker and Theissen, 2003). In addition to the MADS domain, plant type II proteins also contain a conserved K domain for protein– protein interaction (Yang and Jack, 2004), whereas animal type II proteins acquired a 29-amino acid sequence which was termed the MEF2 domain, based on its inclusion in animal MEF2 proteins (Theissen et al., 1996; Black and Olson, 1998). The diversity of organisms encoding type I and type II MADS domain proteins indicates that MEF2 proteins arose early in the evolution of life on Earth, and have been retained in the genomes of organisms since that time to

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

fulfill functions essential to the development and survival of the organism. Indeed, it appears that all eukaryotic genomes contain at least one member of each of the type I and type II MADS domain families.

II.C.  Structure of MEF2 Proteins A striking feature of the MEF2 family is the retention of the MADS domain and the adjacent MEF2 domain at the N-terminus of all known MEF2 proteins. This is clearly apparent in comparisons of the overall domain structures of mammalian MEF2 proteins with that of Drosophila (Fig. 1A). In serum response factor (see Chapter 9.3), by contrast, the MADS domain begins at amino acid 141. Structural studies comparing MEF2 to serum response factor indicate that the presence of the MADS domain at the N-terminus affects DNA-binding and likely accounts for the differential binding sites preferred by MEF2 compared to serum response factor (West et al., 1997; Santelli and Richmond, 2000). The MADS and MEF2 domains have been deeply conserved throughout evolution. For example, within the combined MADS and MEF2 domains, Drosophila MEF2 differs from mouse MEF2D at only 9 of 86 amino acid residues (Fig. 1B).

675

As might be expected from the high sequence conservation within N-termini of MEF2 proteins, critical functions are imparted by the MADS and MEF2 domains. Deletion analyses confirmed that the MADS domain was required, although not sufficient, for DNA-binding. Full DNA-binding by MEF2 also required the presence of the MEF2 domain (Huang et al., 2000). Furthermore, MEF2 factors also dimerize via the MADS and MEF2 domains (Pollock and Treisman, 1991). To define the residues within MEF2 that contribute to dimerization and DNA-binding, extensive mutagenesis studies were conducted by Molkentin and colleagues. These authors generated a series of 22 point mutants of MEF2C, containing alterations in the sequence of the MADS and MEF2 domains, and assayed each mutant protein for dimerization, DNA-binding and transcriptional activation potential (Molkentin et al., 1996a). These studies identified three critical regions within the MADS domain and one region in the MEF2 domain that were required for DNA binding: amino acids 3–5; 23–24; 30–31; and 68–72 (Molkentin et al., 1996a). In addition, residues 35–50 of the MADS domain were critical for dimerization. Consistent with the in vitro mutagenesis studies, randomlyinduced point mutants of the Drosophila Mef2 gene have

Figure 1  Structure of MEF2 factors. (A) Domain structure of the four mammalian MEF2 factors MEF2A–D, and Drosophila MEF2. Note that each protein comprises N-terminally located MADS and MEF2 domains (shaded boxes) which function in dimerization and DNA-binding. The C-terminal regions (open boxes) are highly variable. In mammals, variability usually centers around three main regions termed ,  and  (indicated on MEF2C). (B) Sequence conservation among the MADS and MEF2 domains of murine and Drosophila MEF2 proteins. (C) Structure of a human MEF2A dimer complexed with DNA. Amino acids 1–85 are shown. DNA strands are shown in two shades of purple, and the two MEF2A polypeptides are shown in two shades of blue. Amino acid A39 is shown for orientation purposes. Image was created using Protein Explorer and the structure coordinates contained in Accession #1C7U. Huang et al. (2000).

676

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

also established the importance of the MADS domain for MEF2 function in vivo (Nguyen et al., 2002). Three point mutants of Drosophila MEF2 that affect conserved MADS domain residues ablated MEF2 DNA-binding (Nguyen et al., 2002). Notably, one of the residues affected in these mutants, R24, mimics the biochemical defects of the orthologous mutant mammalian protein (Molkentin et al., 1996a; Nguyen et al., 2002). More recently, a rigorous evaluation of the MEF2C mutagenesis data has been possible, with the determination by two separate groups of the crystal structure of the N-terminus of MEF2A bound to DNA (Fig. 1C) (Huang et al., 2000; Santelli and Richmond, 2000). The two structures are in close agreement with each other, and also with the published mutagenesis studies from Molkentin and colleagues. The structural analyses show that the Nterminus of MEF2A (amino acids 1–10) forms an extension that contacts the minor groove of DNA, presumably to stabilize protein–DNA interactions, while amino acids 13–36 form an -helix that contacts DNA at several sites in the major groove (Huang et al., 2000; Santelli and Richmond, 2000). Critical interacting residues are dispersed in the region encompassing amino acids 13–36, but also include amino acids 23–24 and 30–31, which were defined by functional studies. Following the -helical region, the MADS domain forms two antiparallel -sheets, comprising approximately amino acids 40–60. In the crystal structure, these motifs are critical contact points for dimerization of MEF2 (Huang et al., 2000; Santelli and Richmond, 2000). These observations are also in agreement with the mutagenesis studies, which showed that residues 35–50 were critical for protein–protein interactions within the dimer (Molkentin et al., 1996a). Finally, the MEF2 domain forms a short -helical region (amino acids 63–73), which also appears to function in dimer­ ization, since the location of the helix is remote from the DNA, and the structure predicts a contact point between dimerized MEF2A polypeptides. Here, the structural data diverge slightly from those predicted by mutagenesis studies, which suggested that the MEF2 domain mutations affected DNA-binding but not dimerization (Molkentin et al., 1996a). These differences might be explained if the function of the MEF2 domain -helix residues was to stabilize the dimer after it had formed. In vitro dimerization studies might not be sensitive enough to detect a mild destabilization, but this could be reflected in attenuated DNA-binding. There is also compelling evidence that the MADS and MEF2 domains function critically in the activation of target gene expression. Again, the first evidence in support of this came from the mutagenesis studies, which showed that individual mutation of several different residues scattered throughout the first 86 amino acids did not affect DNAbinding, but had severe effects on transcriptional activation ability (Molkentin et al., 1996a). These observations

are consistent with the large number of co-factors that interact with MEF2 through the N-terminal domains, as well as the observation that the phosphorylation of a serine at the junction of the MADS and MEF2 domains is an important post-translational mechanism for MEF2 regulation (Molkentin et al., 1996a; Cox et al., 2003). Phosphorylation and other post-translational modifications of MEF2 will be discussed in detail in Section III of this chapter. In contrast to the MADS and MEF2 domains, the Cterminal regions of MEF2 proteins are highly-divergent and also highly-variable within a single gene, as a result of regulated RNA splicing. MEF2 primary transcripts are subjected to alternative splicing, skip splicing and cryptic splice site selection, which generates a large number of potential MEF2 isoforms (Black and Olson, 1998; Zhu and Gulick, 2004; Zhu et al., 2005). For Mef2d, the alternative splicing of the exon immediately C-terminal to the MEF2 domain is regulated in a tissue-specific manner to give rise to a muscle-specific isoform (Fig. 1A) (Breitbart et al., 1993; Martin et al., 1994), and a similar pattern of alternative splicing of an equivalent domain has been observed for Mef2a and Mef2c transcripts (Martin et al., 1993; McDermott et al., 1993; Zhu and Gulick, 2004). The Drosophila Mef2 primary transcript is also subject to regulated splicing (Taylor et al., 1995). Recent studies from Gulick and colleagues have categorized the different protein domains resulting from regulated splicing of mammalian Mef2c transcripts as ,  and  (Fig. 1A) (Zhu and Gulick, 2004). Mef2a and Mef2d transcripts also show alternative splicing of the  and  regions, and constitutively include sequences encoding the  domain (Zhu et al., 2005). The  domain enhances transcriptional activation by the parent MEF2 molecule, and is also preferentially-included in brain and muscle transcripts. In contrast, the  domain is alternatively spliced in Mef2c, and acts as a phosphorylation-dependent transcriptional repressor (Zhu et al., 2005). These studies indicate that MEF2 function is modulated by alternative RNA splicing. Given the recent observation that alternate splicing of Mef2b transcripts is altered in Mef2c-null hearts (Vong et al., 2006), pathological conditions might significantly affect the patterns of Mef2 transcript splicing, and thus alter MEF2 function. In addition to transcriptional activation functions, the C-termini of MEF2A, MEF2C and MEF2D each contain nuclear localization signals (Fig. 1A), which are critical to in vivo function (Yu, 1996; Borghi et al., 2001). Furthermore, a Drosophila Mef2 mutant allele encodes a C-terminally truncated isoform that does not localize to the nucleus, suggesting that the location of nuclear localization signals is generally conserved across MEF2 proteins (Ranganayakulu et al., 1995). Furthermore, and as discussed in the next section, the C-terminal regions of MEF2 proteins contain potent transcriptional activation domains,

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

and are also important targets for phosphorylation and other post-translational modifications that regulate MEF2 function.

III.  Regulation of mef2 activity by post-translational modification III.A.  MEF2 Functions as a Transcriptional Co-Factor MEF2 transcription factors interact with a diverse array of co-factors that modulate MEF2 activity. MEF2 can function either as an activator or as a repressor, depending on co-factor interactions, and several MEF2 co-factors facilitate the ability of MEF2 to respond to intracellular signaling. MEF2 function as a transcriptional co-factor was first described from studies in skeletal muscle, where MEF2 proteins were shown to function as essential co-factors for the myogenic bHLH proteins, including MyoD and myogenin (Molkentin et al., 1995; Ornatsky et al., 1997; Black et al., 1998). Myogenic bHLH proteins have the remarkable ability to convert nonmuscle cells to muscle cells in culture, and it was observed that this activity was dependent on interaction with MEF2 (Molkentin et al., 1995; Ornatsky et al., 1997; Black et al., 1998). MEF2 and MyoD physically associate, and their binding sites are frequently coordinately positioned in the enhancers and promoters of muscle-specific genes (Molkentin et al., 1995; Fickett, 1996; Black et al., 1998). The interaction of MyoD and MEF2 occurs through the DNA-binding motifs of each factor, the bHLH domain on MyoD and the MADS and MEF2 domains of MEF2, raising the possibility that MEF2 factors may interact with a wide array of bHLH proteins. Indeed, MEF2 proteins interact with several other bHLH proteins in diverse contexts. MEF2 forms a complex and potently activates transcription of target genes in cooperation with the neural bHLH protein mammalian achaete-scute homolog 1 (MASH1), and this interaction is also dependent on the MADS and MEF2 domains (Black et al., 1996; Mao and Nadal-Ginard, 1996). In the heart, MEF2C cooperatively activates transcription of the Nppa gene with the bHLH proteins HAND1 and HAND2, which like MEF2, are essential regulators of cardiac development (Srivastava et al., 1995; Zang et al., 2004; Morin et al., 2005). GATA4 is another cardiac-enriched transcription factor that interacts with MEF2 to activate the Nppa promoter (Morin et al., 2000) (see Chapter 9.2). Given the broad overlap in the expression of GATA and MEF2 transcription factors and the prevalence of GATA and MEF2 sites in cardiac promoters, these members of these two families of transcription factors may participate in the co-activation of numerous other genes in the heart (Vanpoucke et al.,

677

2004). The homeodomain protein Pitx2 also interacts with MEF2A to activate the Nppa promoter synergistically (Toro et al., 2004). This interaction requires MEF2-binding to the promoter, suggesting that MEF2 may serve as a platform for Pitx2-binding, and that direct interaction of Pitx2 with a cis-acting element in the promoter may not be required (Toro et al., 2004). The majority of MEF2-interacting transcription factors discussed above contact MEF2 through the MADS and MEF2 domain. The transcription factor TEF-1 is an example of a MEF2 co-factor that does not bind to the MADS domain. Rather, TEF-1 interacts with motifs present near the C-terminus of MEF2C (Maeda et al., 2002). TEF-1 usually binds to MCAT elements, which are present in the promoters and enhancers of numerous skeletal and cardiac muscle genes (Mar and Ordahl, 1990). Many of these gene promoters also contain conserved MEF2 sites, which suggest the possibility that MEF2C and TEF-1 may co-regulate multiple genes involved in cardiac development and differentiation. In addition, it has been observed that TEF-1 can bind directly to MEF2 sites or to MEF2-like AT-rich elements, suggesting a further interplay between MEF2 and TEF-1 during muscle development, possibly through competition for shared binding elements (Karasseva et al., 2003). Thyroid hormone receptor (TR) is another MEF2 co-factor that interacts with MEF2 in the heart, and this interaction is facilitated by p300/CBP, which is thought to bridge the two factors and promote transcriptional activation (De Luca et al., 2003). TR and MEF2 interaction has been shown to be important for the activation of the -MHC gene via closely-positioned binding sites for the two factors in the proximal promoter region (Lee et al., 1997). The interaction of MEF2 and MyoD in skeletal muscle is also facilitated by p300 (Sartorelli et al., 1997). In addition to acetylation of histones and subsequent chromatin relaxation, p300 also directly acetylates MEF2, which promotes MEF2 transactivation, probably through negative regulation of sumoylation (Ma et al., 2005; Zhao et al., 2005a), which will be discussed later in this chapter. The cooperative transcriptional activation of the -MHC promoter by TR and MEF2 is attenuated by the action of Jumonji/Jarid2, which is also a direct MEF2 partner (Kim et al., 2005). Jumonji, a histone demethylase, directly interacts with the MADS domain of MEF2A to negatively regulate MEF2-dependent transcription (Kim et al., 2005). Jumonji interaction with MEF2A probably blocks TR interaction with the MADS domain, and represses transcription via a direct effect on histones. In addition to MEF2 co-factors that interact with DNA and MEF2, there are a number of tissue-specific and ubi­ quitous proteins that modulate MEF2 activity solely through protein–protein interactions. One such group of MEF2-interacting factors includes the SAP domain proteins myocardin and MASTR. Myocardin was first identified as a potent transcriptional activation partner for

678

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

serum response factor (see Chapter 9.3), a MADS domain protein that is closely related to MEF2, in cardiac and smooth muscle (Wang et al., 2003b; Yoshida et al., 2003). Myocardin itself does not bind DNA, but uses DNA-bound serum response factor as a platform to interact with chromatin through protein–protein interactions. Once bound to serum response factor, myocardin very potently activates transcription (Wang et al., 2003b). One molecule of myocardin interacts with each serum response factor dimer bound to a CArG box, and myocardin itself probably dimerizes to bridge two CArG elements (Wang et al., 2003b; Yoshida et al., 2003). A longer form of myocardin, generated from a distinct splicing event can interact with either serum response factor or MEF2 (Creemers et al., 2006). The MEF2 interaction domain in the myocardin long-form comprises a short amino acid sequence at the N-terminus that is distinct from the N-terminus of the short-form of myocardin, which facilitates serum response factor interaction (Creemers et al., 2006). The unique N-terminus of the myocardin long-form directly interacts with MEF2. Interestingly, however, the myocardin longform still retains the sequences necessary to facilitate interaction with serum response factor, such that this isoform of myocardin can interact with either serum response factor or MEF2 (Creemers et al., 2006). Olson and colleagues used the unique N-terminal sequence of the myocardin long-form to identify another SAP domain-containing protein that also contains the MEF2 interacting sequence (Creemers et al., 2006). This myocardin homolog was named MASTR (MEF2-associated SAP domain transcriptional regulator). Unlike the myocardin long-form, MASTR only contains sequences sufficient to interact with MEF2 and not with serum response factor (Creemers et al., 2006). Thus, the myocardin family includes: the myocardin short-form, which only interacts with serum response factor; MASTR, which only interacts with MEF2; and the myocardin long-form, which can interact with either MEF2 or serum response factor. The myocardin long- and short-forms contain sequences that should allow dimerization between the two isoforms, which raises the intriguing possibility that myocardin heterodimers, containing long- and short-forms, might bridge serum response factor and MEF2 binding sites. Since numerous promoters contain binding elements for each of these classes of MADS domain transcription factors, myocardin members may bridge the sites and allow synergistic activation mediated by MEF2 and serum response factor. This possible relationship between serum response factor and MEF2 through myocardin may be further facilitated by the presence of binding sites that can be bound by either MEF2 or serum response factor (L’Honore et al., 2007). These composite SRF/MEF2 cis-elements provide additional targets for myocardin and MASTR regulation of cardiac genes. To date, no specific genes have been identified as myocardin targets through the bridging of MEF2

and serum response factor bound to their respective cognate binding sites, but it seems likely that this mechanism will function in the heart and other tissues. If this notion is correct, it would provide an additional mode for MEF2 to serve as a transcriptional switch through the assembly of a multi-protein complex.

III.B.  Chromatin Remodeling by MEF2 through Interaction with Histone Deacetylases It is now becoming increasingly appreciated that binding sites for regulatory factors must be accessible in the context of the overall chromatin structure of the cell in order to be recognized, and for gene expression to be controlled. Along these lines, it now appears that a major function of MEF2 is to control the balance between chromatin acetylation and deacetylation, and thereby regulate the relative accessibility of promoters and enhancers to the transcriptional machinery (Fig. 2). Accordingly, MEF2 factors interact with multiple histone acetylases and deacetylases. Most notably, MEF2 forms a complex with class II histone deacetylases (HDACs), which include HDACs 4, 5, 6, 7 and 9 (McKinsey et al., 2001a) (see Chapter 10.2). Interaction with class II HDACs occurs through the MADS domain at the N-terminus of MEF2 (Lu et al., 2000b; Dressel et al., 2001; Zhang et al., 2001a). Similarly, a conserved N-terminal domain in HDAC dictates interaction with MEF2 (Wang et al., 1999; Lemercier et al., 2000; Dressel et al., 2001). MEF2-HDAC complexes repress transcription by deacetylating histones, resulting in chromatin condensation and reduced accessibility of core transcriptional machinery to promoter and enhancer regions of MEF2 target genes (Lu et al., 2000b; Kao et al., 2001; McKinsey et al., 2001a). MEF2 also interacts with several histone acetyltransferases, including p300/CBP and SIRT1, which likely serve to balance the repressive effects of HDAC on MEF2 and allow MEF2 to function as a transcriptional switch (Sartorelli et al., 1999; Ma et al., 2005; Zhao et al., 2005a; Stankovic-Valentin et al., 2007). Class II HDACs are important regulators of transcription in the developing and postnatal heart that help to regulate the hypertrophic response (Zhang et al., 2002; Olson et al., 2006). Normal growth of the myocardium requires large amounts of structural and other regulatory proteins to be synthesized as cells enlarge, but excessive enlargement of the heart can result in pathologic hypertrophy, which ultimately can lead to heart failure (Olson et al., 2006). Thus, HDACs serve as a kind of regulated braking mechanism, keeping the MEF2-dependent transcriptional response in check until signals that stimulate myocardial growth are received. Hypertrophic induction results in signal-dependent export of HDACs from the nucleus, which results in chromatin relaxation due to increased histone

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

679

Normal and Pathologic Stimulation catecholamine stimulation electrical activity exercise

P

CamK

P 14-3-3

HDAC MAPK

HDAC MEF2

P300

X

Initial activation signals received: MEF2 and HDAC are phosphorylated by MAPK and CamK signaling

GRIP

MEF2 Phosphorylation of HDAC results in association with 14-3-3 proteins and nuclear export. MEF2 associates with HATs (P300) and potent coactivator

Figure 2  MEF2 functions as a signal-dependent transcriptional switch. MEF2 functions as a repressor by recruiting class II HDACs to promoter and enhancer regions of target genes. In response to a variety of developmental and pathological signals, CamK and MAPK signaling pathways are activated. These signals result in the phosphorylation of MEF2 and HDACs. HDAC phosphorylation by CamK results in exposure of a nuclear export signal at the C-terminus, interaction with 14-3-3 proteins, and export from the nucleus. MEF2 phosphorylation and HDAC dissociation result in recruitment of HATs and co-activator molecules, such as GRIP-1, and the conversion of MEF2 to an activator complex.

acetylation and subsequent MEF2-dependent transcription (McKinsey et al., 2000a; Kao et al., 2001; McKinsey et al., 2001a,b; Olson et al., 2006). Intriguingly, the deacetylase activity of class II HDACs is not required for interaction with MEF2, nor is this activity required for transcriptional repression by MEF2-HDAC (Lemercier et al., 2000; Zhang et al., 2001b; Chan et al., 2003). The dispensability of deacetylase activity is consistent with the observations that MITR (MEF2-interacting transcriptional repressor), an HDAC homolog lacking the deacetylase domain, also interacts with MEF2 and facilitates strong transcriptional repression, as its name implies (Youn et al., 2000; Zhang et al., 2001b). This may occur because class II HDACs and MITR have the ability to recruit the potent co-repressor protein CtBP (Dressel et al., 2001; Zhang et al., 2001a). CtBP co-repressors repress transcription via recruitment of HDACs (Bertos et al., 2001). CtBP physically associates with the N-terminus of HDAC4, HDAC5 and the HDAC homolog MITR, which interact with MEF2 to repress its activity (Zhang et al., 2001a). The deacetylase activity-independent repression might also result from the observation that HDACs can multimerize, allowing deactylase-defective HDACs to recruit other HDACs that possess full enzymatic activity, although this may not explain the strong transcriptional repression conferred to MEF2 by MITR (Youn et al., 2000; Zhang et al., 2001b). Alternatively, a crucial function of class II HDACs may be to help control the balance of acetylation and sumoylation of MEF2. These two

post-translational modifications have mutually-exclusive and opposing functions in promoting MEF2-dependent activation and repression, respectively (Zhao et al., 2005a; Gregoire et al., 2006; Shalizi et al., 2006; StankovicValentin et al., 2007). In addition to its role in deacetyl­ ation of histones, HDAC4 also inhibits MEF2-dependent transcription by promoting sumoylation of MEF2 (Zhao et al., 2005a; Gregoire et al., 2006; Stankovic-Valentin et al., 2007). As noted above, it was recently discovered that MEF2 proteins are modified by sumoylation (Gregoire et al., 2006; Kang et al., 2006; Riquelme et al., 2006; Shalizi et al., 2006). Sumoylation is the process by which a protein moiety, SUMO, is covalently added to proteins by the activity of SUMO-conjugating enzymes (Gill, 2005). Addition of SUMO-1 to the C-terminus of MEF2 modifies MEF2 to a repressor form, which has been demonstrated for MEF2A, MEF2C and MEF2D (Gregoire and Yang, 2005; Zhao et al., 2005a; Kang et al., 2006; Shalizi et al., 2006). Interestingly, SUMO addition to MEF2 occurs at a lysine residue in the C-terminal activation domain and is controlled by MAPK phosphorylation at a nearby serine residue for MEF2A, MEF2C and MEF2D (Fig. 3) (Gregoire et al., 2006; Kang et al., 2006; Shalizi et al., 2006). In each case, dephosphorylation of MEF2 promotes a switch from acetylation to sumoylation at the neighboring lysine and the conversion of MEF2 from an activator to a repressor (Gregoire et al., 2006; Kang et al., 2006; Shalizi et al., 2006). The dephosphorylation event is

680

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

P38 BMK1 CKII

P38

P38

cdk5

CKII

ERK5 P38

BMK1 ERK5

P38

CnA

P38

?

MEF2A

MADS

P

P

P

P

S59

S98

S192

S255

MEF2

P

PP

S289 T312 T319

P S355

S U M Ac O

P

K403 S408

P S479

Figure 3  MEF2 proteins are extensively modified post-translationally. A schematic of human MEF2A shows sites of phosphorylation (red ovals), sumoylation (pink oval) and acetylation (yellow oval). Signaling pathways that are known or thought to modify MEF2 are noted above and the amino acid residues are noted below. The MADS and MEF2 domains are depicted. Differential modification of lysine 403 by acetylation and sumoylation is controlled by phosphorylation and dephosphorylation by p38 and calcineurin at the neighboring serine 408. Note that similar (but not identical) posttranslational modifications depicted here for hMEF2A also occur on other MEF2 isoforms.

controlled by calcineurin, which serves as a link between calcium-mediated hypertrophic signaling and posttranslational modification of MEF2 by sumoylation (Flavell et al., 2006; Shalizi et al., 2006). In contrast to sumoylation, direct acetylation of MEF2C at the C-terminus by CBP promotes MEF2 activity, as noted earlier in this chapter (Sartorelli et al., 1997). The balance between sumoylation and acetylation of MEF2 is also controlled directly by HDACs themselves. HDAC4 has been shown to promote SUMO addition to MEF2, providing a dual mechanism for repression of MEF2 activity by class II HDACs through deacetylation of chromatin, and further through addition of SUMO to MEF2 (Zhao et al., 2005a; Gregoire et al., 2006). This may partially explain why the enzymatic activity of HDACs is not required for repression of MEF2, since the catalytic domain of HDAC4 is not required for sumoyl­ ation of MEF2 (Zhao et al., 2005a). Interestingly, HDAC4 is also regulated by sumoylation. However, sumoylation of HDAC4 inhibits its ability to promote SUMO addition to MEF2, which provides a post-translational negativefeedback mechanism for control of MEF2 sumoylation and repression (Zhao et al., 2005a; Gregoire et al., 2006). It was previously believed that only class II HDACs interacted with MEF2, but several reports have now demonstrated that class I HDACs also interact with MEF2, and that these interactions play an important role in the heart (Montgomery et al., 2007; Trivedi et al., 2007). The class I HDAC, HDAC3, was shown recently to interact with and supress the transcriptional activity of MEF2 (Gregoire et al., 2007). The nature of the interaction between class I HDACs and MEF2 is not as well-characterized as the interaction with class II HDACs, so it remains unclear which residues within the MEF2 MADS domain make direct contact with the class I HDAC (Gregoire et al., 2007). Structural studies should resolve how the interactions

of these two broad classes of HDACs with MEF2 result in different biological outputs.

III.C.  MEF2 Functions as a Signal-Dependent Transcriptional Switch The interaction of MEF2 with HDACs underscores the function of MEF2 proteins as both positive and negative regulators of transcription. Prior to those groundbreaking observations, MEF2 proteins were generally thought of only as transcriptional activators that functioned through protein–protein interactions with other transcription factors containing more potent activation domains (Molkentin and Olson, 1996; Black and Olson, 1998). The notion that MEF2 functions as both a repressor and an activator, depending on the gene, cell type and cellular differentiation state led to the idea that MEF2 serves as a switch capable of interpreting distinct signals into opposing transcriptional outputs. Therefore, it has been important to define what dictates whether MEF2 functions as an activator or a repressor. Over the last decade, numerous intra­ cellular signaling pathways have been identified to interact with MEF2 and class II HDACs. Not surprisingly, MEF2 and HDAC proteins are each regulated by their phosphorylation state (McKinsey et al., 2001a, 2002). Class II HDACs are phosphorylated in response to a variety of extracellular signals, including electrical activity, pressure, adrenergic signaling and other normal developmental and postnatal cues. These signals result in an increase in the concentration of Ca2 in the cytoplasm, which activates the phosphatase calcineurin and stimulates the activity of calcium/calmodulin-dependent kinases (CaMK) I, II and VI (Lu et al., 2000a; Kao et al., 2001; McKinsey et al., 2001b; Little et al., 2007). Phosphorylation of class II HDACs by CaMK occurs on

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

two residues near the N-terminus of the HDAC protein (McKinsey et al., 2000a,b; Kao et al., 2001). CaMK phosphorylation of HDAC facilitates interaction with 14-3-3 proteins, which activates a nuclear export sequence at the C-terminus (McKinsey et al., 2000b; Choi et al., 2001; McKinsey et al., 2001b; Wang and Yang, 2001). HDAC nuclear export, in turn, results in MEF2 activation of gene expression in the heart and other MEF2-dependent tissues (Fig. 2). In addition, SIK1 kinase phosphorylates class II HDACs, which affects MEF2-dependent transcription through dissociation of MEF2 from HDAC (Berdeaux et al., 2007). MEF2 factors themselves are extensively phosphorylated in response to a host of intracellular and extracellular cues (Fig. 3). The p38, BMK1/ERK5 and ERK1 mitogen-activated protein kinases (MAPK) each play a role in MEF2 regulation through phosphorylation. BMK1/ERK5 signaling results in the phosphorylation of MEF2C at serine-387 in the C-terminal activation domain in response to a variety of extracellular signals, including adrenergic signaling and pressure overload in the myocardium (Kato et al., 1997; Yang et al., 1998; Yan et al., 1999; Nadruz et al., 2003). Each of these cues stimulates the nuclear localization of BMK1/ERK5, which leads to phosphorylation and activation of MEF2 (Kato et al., 2000; Yan et al., 2001). Although these studies investigated BMK1/ ERK5 signaling in the postnatal heart, it is also quite likely that BMK1/ERK5 signaling regulates MEF2 phosphorylation during cardiac development, since all of these signals lead to MEF2-dependent c-jun transcription, which functions in numerous developmental contexts (Kato et al., 1997; Marinissen et al., 1999; Nadruz et al., 2003). The p38 MAPK signaling pathway also plays a fundamentally important role in the post-translational modification of MEF2 in myocytes (Han and Molkentin, 2000). Signaling by p38 results in phosphorylation of MEF2A and MEF2C, but not MEF2B or MEF2D (Han et al., 1997; Ornatsky et al., 1999; Zhao et al., 1999; Chang et al., 2002). p38 phosphorylation promotes the role of MEF2 as a transcriptional activator in response to normal developmental and postnatal hypertrophic growth of the heart, as well as to pathologic hypertrophic cues (Han et al., 1997; Kolodziejczyk et al., 1999; Ornatsky et al., 1999; Zhao et al., 1999; Han and Molkentin, 2000; Cox et al., 2003). Other studies have shown that retinoic acid (RA) signaling during myocardial development results in the phosphor­ ylation of MEF2 via p38 MAPK (Ren et al., 2007), which potentially links normal growth and development of the myocardium via RA signaling to MEF2 (Tran and Sucov, 1998; Lavine et al., 2005). Casein kinase II (CKII) results in phosphorylation of a conserved serine (S59) found at the junction of the MADS and MEF2 domains in MEF2A and MEF2C (Molkentin et al., 1996c; Cox et al., 2003). Work from McDermott and colleagues showed that CKII also results in the direct phosphorylation of MEF2A at

681

serine 289 in response to p38 MAPK signaling (Fig. 3) (Cox et al., 2003). An interesting study of MEF2 function in dominant induction of skeletal myogenesis in culture showed that RAF kinase inhibits MyoD-induced conversion of fibro­ blasts into muscle cells by blocking MEF2 nuclear local­ ization (Winter and Arnold, 2000). These studies were the first to suggest a regulated nuclear localization of MEF2 proteins themselves, which may have important implications if MEF2 is shuttled between the nucleus and cytoplasm during the development of the heart and other tissues. Cyclins and cyclin-dependent kinase pathways have also been shown to regulate MEF2 activity and to provide a link between the cell-cycle and MEF2-dependent transcription. For example, cdk5 phosphorylates MEF2C in neurons, supporting the notion that a cell-cycle-dependent signaling event functions via post-translational modification of MEF2C in neurons (Gong et al., 2003; Tang et al., 2005; Smith et al., 2006). The role of cdk phosphorylation of MEF2 in the heart has not been examined, but it is likely that MEF2 plays a role in cell-cycle control along with GATA4, which is known to regulate cardiac cell-cycle, and to interact with MEF2 and other partners downstream of these signaling pathways (Morin et al., 2000; Vanpoucke et al., 2004; Zeisberg et al., 2005; Xin et al., 2006). MEF2 factors themselves have only weak inherent transcriptional activation potential (Molkentin et al., 1996a,b). This is also the case for several of the many MEF2 cofactors that have been described to date, such as GATA4. However, MEF2 proteins and their co-factors are sufficient to direct extremely robust activation of many genes and reporter genes, both in vivo and in reporter assays in cell culture. A likely mechanism for how MEF2 is able to drive strong transcriptional activation is through interaction with potent transcriptional co-activators, such as GRIP-1, which belongs to the p160 steroid receptor co-activator (SRC) family of transcriptional co-activator proteins (Chen et al., 2000; Leo and Chen, 2000; Lazaro et al., 2002; Xu and Li, 2003; Liu et al., 2004). Signaling downstream of D cyclins and cdk4 activity blocks muscle-differentiation by disrupting the interaction between MEF2C and GRIP-1 (Lazaro et al., 2002). Interactions with transcriptional co-activators and corepressors provide another mechanism for MEF2 to serve as a transcriptional switch, repressing transcription in some contexts, while activating it in others. In this regard, the interaction between MEF2 and GRIP-1 is also targeted by the TGF signaling pathway (Liu et al., 2004). TGF signaling is transmitted intracellularly by Smads transcription factors, and MEF2 interacts with several different members of the Smads family, including Smads 2, 3 and 4 (Quinn et al., 2001; Derynck and Zhang, 2003; Liu et al., 2004). Smads 2 and 4 interact with MEF2 to influence transcription positively in skeletal muscle by promoting

682

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

interaction between MEF2 and MyoD (Quinn et al., 2001). Presumably, TGF or BMP signaling influences MEF2 activity in the heart through interactions with Smads as well, although this has not been demonstrated. In contrast, Smad3 is an inhibitory Smad and functions as a negative regulator of transcription (Liu et al., 2004). Smad3 associates with MEF2 and disrupts MEF2 association with GRIP-1, thereby repressing transcription by targeting the interaction of MEF2 and one of its co-activators (Liu et al., 2004). MAML1 (mastermind like 1) is also a potent transcriptional co-activator that interacts with MEF2 (Shen et al., 2006). MAML factors belong to a family of co-activator proteins that are required for Notch signaling. Notch sig­ naling inhibits skeletal muscle differentiation and the muscle-inducing activity of myogenic bHLH proteins (Kopan et al., 1994; Kuroda et al., 1999; Wilson-Rawls et al., 1999; Shen et al., 2006). Activated Notch interacts directly with MEF2C via a specific splice isoform of MEF2C found in muscle lineages (Wilson-Rawls et al., 1999). Thus, it is clear that this pathway interacts with MEF2 via Notch itself and via the Notch co-activator MAML1 to inhibit skeletal myogenesis (Wilson-Rawls et al., 1999; Shen et al., 2006). The presence of the muscle-specific exon has not been examined in cardiac muscle, so it is not known whether Notch interacts directly with MEF2C in the heart. In addition to direct interactions with co-activator molecules, MEF2 also interacts with a number of co-repressor proteins. For example, CABIN1/CAIN is a potent transcriptional co-repressor protein that binds to calcineurin and interacts directly with MEF2 (Esau et al., 2001; Han et al., 2003). Interaction of CABIN1 with MEF2 blocks the activation of MEF2 by BMK1/ERK5 MAP kinase signaling (Kasler et al., 2000). CABIN1 further represses MEF2 by recruitment of mSIN3 and associated HDACs, and CABIN1 also recruits the histone methyltransferase SUV39H1 to the complex (Youn and Liu, 2000; Jang et al., 2007). Chen and colleagues solved the crystal structure of the MADS domain of MEF2B in a complex with CABIN1 (Han et al., 2003). The structure showed that CABIN1 contacts the MADS domain in a similar, but not identical, fashion to class II HDACs, raising the possibility that the binding of HDAC and CABIN1 to MEF2 might be mutually exclusive (Han et al., 2003, 2005). The discrete contacts of each of these MEF2 co-regulators suggest that they likely affect MEF2 conformation differently, but this remains to be determined. In general, it is clear that the major function of MEF2 is to respond to intracellular signaling pathways and to respond to those signaling cues through co-factor interactions (Figs 2; 3). In response to growth and differentiation signals, MEF2 proteins are modified post-translationally, which is primarily via phosphorylation but also includes acetylation and sumoylation (Fig. 3). These modifications influence whether MEF2 interacts with positive

transcriptional co-factors, such as SRC family co-activators, myocardin proteins and histone acetyltransferases, or co-factors that repress transcription, including class II HDACs, MAML, CABIN1 and Jumonji proteins (Czubryt and Olson, 2004; Backs and Olson, 2006; Liu and Olson, 2006). In addition, MEF2 interacts with multiple other transcription factors, which provides additional combinatorial complexity and allows for more precise co-factor recruitment. It is likely that additional MEF2 co-factors will continue to be identified, but the general role for MEF2 as a nodal point for balancing growth and differentiation signals through post-translational modification and co-factor interaction is now well-established.

IV.  MEF2 gene function in the heart and other tissues IV.A.  MEF2 Proteins are Expressed in Multiple Lineages During Development and in Adulthood Despite the initial identification of MEF2 proteins as factors that are bound to skeletal muscle gene promoters, early studies suggested that MEF2-binding activity was present in a broad tissue distribution (Olson et al., 1995). Indeed, Mef2d transcripts are detected in many adult tissues, although the muscle-specific splice variant of Mef2d is expressed only in adult muscles (Breitbart et al., 1993; Martin et al., 1994). Furthermore, Xenopus Mef2 genes are broadly-expressed in the adult frog (Chambers et al., 1992). By contrast, in situ hybridization studies performed on vertebrate embryos indicate that expression of Mef2 genes is indeed predominantly muscle-specific at earlier stages of development (Edmondson et al., 1994). Mef2 gene expression in frogs and fish is restricted to muscle lineages in the embryo (Chambers et al., 1992; Ticho et al., 1996), and in the mouse embryo, Mef2a, Mef2c and Mef2d expression is largely restricted to skeletal and cardiac muscle early in development (Edmondson et al., 1994). Similarly, mouse Mef2b is a marker of early myogenic lineages (Molkentin et al., 1996b). Nevertheless, expression of Mef2 genes was also detected in the developing nervous system of mice, suggesting that it might also play a role in neuronal development (Leifer et al., 1994; Ikeshima et al., 1995; Lyons et al., 1995; Lin et al., 1996b). In addition, specific functions for MEF2 have been described more recently in specialized cells of the blood and other tissues of mesodermal and ectodermal origin. MEF2 is an activator of cell death in T-lymphocytes via direct transcriptional activation of the steroid receptor gene Nur77 (Youn et al., 1999; Youn and Liu, 2000; Liu et al., 2001a). Similarly, immunoglobulin gene expression in B-cells is regulated by MEF2 function (Rao et al., 1998; Satyaraj and Storb, 1998; Wallin

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

et al., 1999). In addition, MEF2C is an essential regulator of B-cell proliferation in response to B-cell receptor sig­ naling, which results in direct phosphorylation of MEF2C via p38 MAPK (Khiem et al., 2008; Wilker et al., 2008). Furthermore, Mef2c and Mef2d are expressed in chondrocytes during bone formation in the embryo (Arnold et al., 2007), and Mef2c is highly-expressed in developing neural crest cells in the mouse and fish (Miller et al., 2007; Verzi et al., 2007). Based on these studies, it is clear that although MEF2 is not restricted to muscle lineages, muscle cells in the embryo are a major location of Mef2 gene expression. Later in vertebrate development, however, transcription of Mef2 genes is detected in a number of additional tissues and organs, although in many cases the precise cells that express MEF2 within these organs have yet to be defined. In Drosophila, embryonic expression of the single Mef2 gene is restricted to the developing mesodermal tissues, including the cardiac tube (Lilly et al., 1994; Nguyen et al., 1994). Later in development, there is accumulation of MEF2 protein in portions of the Drosophila central nervous system (Schulz et al., 1996). In summary, although they were initially identified as muscle-enriched binding factors, it is now clear that MEF2 proteins function in many tissues during development and in adulthood, including all muscle lineages, neuronal cells, blood cells, neural crest, bone, chondrocytes and vascular endothelium. Direct transcriptional targets of MEF2 have been identified in each of these tissues and, as discussed below, gene knockout strategies and targeted ablation of MEF2 function have begun to define the roles of Mef2 in many of these cell types more precisely.

683

expression of several of these genes have been described (Bour et al., 1995; Lilly et al., 1995; Ranganayakulu et al., 1995; Kelly et al., 2002; Marin et al., 2004). Interestingly, expression of Tropomyosin 2 in Mef2 mutants was reduced in skeletal and visceral muscles, but was relatively normal in the heart (Lin et al., 1996a). Similarly, expression of the -tubulin gene Tub60D was strongly reduced in the skeletal muscles of Mef2 mutants, but was expressed normally in the cardiac tube (Damm et al., 1998). These two examples emphasize that transcriptional programs can differ significantly between different muscle types in the embryo. Nevertheless, the Drosophila studies represent the first example of whole-organism inactivation of MEF2 function and its developmental consequences. These studies remain the clearest examples of MEF2 function and its requirement for the differentiation of muscle lineages. To define the function of vertebrate Mef2 genes in cardiac development, a variety of approaches have been employed. In mice, global knockouts of Mef2c and Mef2a have been performed and have shown that mice lacking these genes have severe defects in cardiac development and function (Lin et al., 1997; Naya et al., 2002). By contrast, a recent publication indicated that homozygotes for a mutation in Mef2d are phenotypically normal (Arnold et al., 2007). Targeted inactivation of Mef2c in mice results in death by embryonic day (E) 10.5, and mutant mice exhibit severe defects in cardiac development (Lin et al., 1997). Affected embryos display pericardial edema, a

IV.B.  Genetic Analyses of Mef2 Gene Function To determine the function of MEF2 in tissues where it is expressed, numerous genetic studies have been performed in flies, mice and fish. The fact that vertebrate genomes have four distinct Mef2 genes makes defining the precise function of MEF2 complicated in higher animals, due to issues of redundancy and overlapping expression. As noted earlier in this chapter, Drosophila has only a single Mef2 gene, which has made genetic analyses of Mef2 function highly tractable in that organism. Inactivation of the Drosophila Mef2 gene results in a severe failure in muscle differentiation for cardiac, skeletal and visceral mesoderm, even though precursors for each of these muscle lineages are normally specified in mutant embryos (Fig. 4) (Bour et al., 1995; Lilly et al., 1995; Ranganayakulu et al., 1995). In Mef2 mutants, the expression of a large number of muscle structural genes was essentially absent. These potential target genes include the single Myosin heavy-chain gene, Actin57B, Myosin light chain-2, Myosin light chain-alkali and Troponin I, and MEF2 binding sites critical for the

Figure 4  Requirement for MEF2 function for cardiac development in Drosophila. Muscle tissue can be visualized in mature embryos via immunohistochemical staining for myosin heavy-chain (MHC) protein. Left column: In wild-type embryos, cardiac muscle (Ca) forms at the dorsal midline; skeletal muscles (Sk) are apparent in dorsal and lateral views; and visceral muscle (Vi) showing MHC accumulation can be observed in a ventral view. Right column: in Mef2-null mutants, there is no detectable MHC accumulation in cardiac muscle, only a handful of skeletal muscle cells accumulate MHC, and the visceral muscle accumulates low levels of MHC. In addition, morphogenesis of the mutant gut is abnormal, with the gut predominantly comprising a single lumen.

684

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

failure of cardiac looping, a reduced heart rate and the lack of a well-defined right ventricle (Fig. 5) (Lin et al., 1997). These morphological defects can now be reinterpreted in the context of the second heart field, which has been shown to contribute predominantly to the outflow tract and right ventricle (Kelly and Buckingham, 2002; Black, 2007). Clearly, Mef2c mutants show profound defects in structures derived from the second heart field (see Chapter 2.2), although second heart field cells do contribute to the common ventricular chamber in Mef2c-null mice (Fig. 5C,D) (Verzi et al., 2005). In addition to the morphological defects, Mef2c mutants displayed a downregulation of a number of cardiac muscle structural genes, including Nppa, cardiac -actin, -myosin heavy chain, myosin light chain (MLC) 1 A, angiopoietin 1 and Vegf (Lin et al., 1997; Bi et al., 1999). Note that relatively few of the downregulated genes were those previously demonstrated to be MEF2 targets (Molkentin and Markham, 1993; Lee et al., 1997). This may result from indirect effects of MEF2C on target genes, either acting via an intermediate transcriptional regulator or by modulating the functions of known cardiac transcription factors. Support for the latter explanation was provided by Morin and colleagues, who showed that MEF2 was able to strongly potentiate Nppa activation by GATA4, as discussed in Section IIIA of this chapter (Morin et al., 2000). However, Mef2c mutant embryos still have normal expression of a number of other cardiac markers, including Mlc2v and Mlc2a (Lin et al., 1997). Since cardiac Mlc2v expression is MEF2-dependent (Navankasattusas et al., 1992), this finding suggested that other MEF2 proteins might compensate for the absence of MEF2C in the mutants. In support of this conclusion, there was a significant upregulation in the expression of Mef2b in Mef2c mutant embryos (Lin et al., 1997). The reason why other MEF2 isoforms can compensate for the lack of MEF2C on some cardiac promoters and not others is still not clear, but is strongly supported by observations that different MEF2 isoforms have distinct co-factor interactions. The early lethality associated with cardiac defects in Mef2c mutants has necessitated more complex approaches to defining MEF2 requirements in later stages of vertebrate development. Recently, Schwarz and colleagues described a floxed conditional allele of Mef2c (Vong et al., 2005). General inactivation of this Mef2c conditional allele resulted in cardiac defects that were essentially identical to those described previously for the conventional Mef2c knockout (Lin et al., 1997; Vong et al., 2005). By contrast, later removal of Mef2c function in the heart using Mlc2vCre or -MHC-Cre resulted in the development of viable offspring (Vong et al., 2005). Thus, it seems that the most critical requirement for MEF2C in the heart is early during cardiac development. It is possible that an initial wave of Mef2c expression is sufficient to initiate the expression of other Mef2 genes, which may result in compensation and

(A)

(B)

hrt hrt

Mef2c +/+

Mef2c -/-

(C)

(D)

PM

PM OFT

RV

OFT V

LV Mef2c +/+

Mef2c -/-

Figure 5  Mef2c is required for cardiac development in mice. Targeted inactivation of Mef2c in mice results in embryonic lethality at E10 due to severe cardiac morphogenetic defects. (A, B) E9.5 wild-type and knockout embryos are shown. Note the failed looping in the mutant animal shown in (B). (C, D) Transverse sections through wild-type and mutant animals showing the single ventricular chamber and a failure of looping morphogenesis in mutants. All embryos express -galactosidase from the Mef2c-AHF-lacZ transgene, which shows the contribution of second heart field-derived cells to both wild-type and mutant hearts; second heart field contribution to the hearts of Mef2c mutant embryos is abnormal (compare staining in C and D) (hrt: heart; LV: left ventricle; OFT: outflow tract; PM: pharyngeal mesoderm; RV: right ventricle).

account for the lack of a cardiac phenotype in conditional knockout embryos with later Mef2c deletion. Alternatively, there may be an unappreciated role for Mef2c earlier in development or in another tissue, such as the vasculature, which then has a secondary effect on cardiac development in the global Mef2c knockout (Lin et al., 1998; Bi et al., 1999). Additional gene inactivation studies should resolve these issues in the future. The Mef2c conditional allele was also utilized recently to define the function of Mef2c in the neural crest and its derivatives. Combination of the floxed Mef2c allele with Wnt1-Cre resulted in animals showing severe craniofacial abnormalities, ultimately causing neonatal mortality due to an occluded airway. Subsequent studies indicated that defective neural crest cell development arose due to a failure of Mef2c to activate the target genes Dlx5, Dlx6 and Hand2 in craniofacial mesenchyme (Verzi et al., 2007). Other recent studies using the zebrafish hoover

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

(hoo) mutant, which has disrupted Mef2c function, also showed a requirement of Mef2c for craniofacial development upstream of the Dlx genes in an endothelin signaling-dependent pathway (Miller et al., 2007). Another role for Mef2c, in the formation of skeletal elements, was also identified using a conditional Mef2c allele in mice. Inactivation of Mef2c function in chondrocytes using a Col2-Cre transgene, or combined reduction of both Mef2c and Mef2d gene dosages, resulted in a failure of normal bone development (Arnold et al., 2007). Targeted inactivation of Mef2a in mice also results in lethality due to cardiac defects, but homozygous Mef2a mutants survive embryogenesis and then the majority of mutants die shortly after birth (Naya et al., 2002). Affected animals showed severe right ventricular dilation and the activation of genes characteristic of cardiac hypertrophy and failure. Mef2a-null adult mice display significant dilated cardiomyopathy, without associated cardiac hypertrophy and have disorganized sarcomeres (Naya et al., 2002). Mutants that survived the neonatal period lived to adulthood, but showed defects in mitochondrial biogenesis, reflected by a three-fold reduction in mitochondrial gene copy number as assayed in the whole heart. An analysis of gene expression comparing Mef2anull to control mice identified several genes with aberrant expression, including myospryn and myomaxin, which each encode costameric proteins (Durham et al., 2006; Huang et al., 2006). These observations suggest the possibility that MEF2A coordinately regulates the expression of intermediate filament proteins, which are essential for cytoskeletal architecture. Disrupted costameric protein expression probably affects the stoichiometry of the components of the sarcomeres, and likely accounts for the disrupted cytoskeleton and cardiomyopathy in Mef2a knockout mice (Naya et al., 2002). Consistent with this idea, overexpression of either MEF2A or MEF2C in the adult myocardium results in dilated cardiomyopathy without significant hypertrophy (Xu et al., 2006). MEF2 overexpression in the heart resulted in significant sarcomeric disorganization, consistent with a role for MEF2 in costameric gene regulation and with the notion that alterations in costameric protein stoichiometry lead to dilated cardiomyopathy (Durham et al., 2006; Huang et al., 2006; Xu et al., 2006). The observation that strikingly different phenotypes result from inactivation of Mef2a versus Mef2c or Mef2d underlines the distinct functions that each isoform must perform in the developing animal. A mouse knockout for Mef2b has yet to be described, but it will be interesting to determine if the phenotype of this mutant, once it is generated, is similar to the phenotypes of existing Mef2 gene mutants. Furthermore, combination of individual knockouts into multiple knockout genotypes will provide more complete information on the roles of MEF2 factors in cardiac development. For example, given the strong upregulation of Mef2b transcription in Mef2c mutants

685

(Lin et al., 1997; Vong et al., 2006), the phenotype of a Mef2b/;Mef2c/ double mutant should be most informative. A number of other approaches have also been employed to define the role of MEF2 in muscle development. McDermott and colleagues generated a truncated form of MEF2A comprising the MADS and MEF2 domains, but lacking the majority of sequences C-terminal to those domains (Ornatsky et al., 1997). This molecule was predicted to bind to MEF2 target genes, but since it does not contain the transactivation domains, it is predicted to compete for activation of direct MEF2 target genes. Indeed, when this molecule was expressed in C2C12 skeletal myo­ blasts, there was an overall inhibition of muscle differentiation (Ornatsky et al., 1997). A similar approach was utilized by Karamboulas et al. who generated a MEF2C fusion with the transcriptional repressor domain of the Engrailed protein (EnR) from Drosophila (Karamboulas et al., 2006). When expressed in P19 embryonal carcinoma cells in a tissue culture cardiomyogenesis system, the chimeric MEF2 molecule inhibited the appearance of cardiomyocytes. When MEF2CEnR was expressed in the developing hearts of transiently transgenic mouse embryos using the Nkx2-5 promoter, some embryos showed defects in cardiac development (Karamboulas et al., 2006). The effects of this treatment ranged from an almost complete absence of differentiated myocardium to relatively normal embryos, presumably a reflection of different levels of MEF2C-EnR expression in the transiently transgenic embryos. An advantage to the approaches utilizing dominantnegative or repressor versions of MEF2 are that such constructs probably interfere with the functions of all isoforms of MEF2 in the target cell, obviating the issues of genetic redundancy among MEF2 family members. Furthermore, utilization of tissue-specific promoters increases the specificity of the effects within the embryo. On the other hand, generation of MEF2-EnR chimeras is likely to show greater defects in cells than simply loss of MEF2 function. This is because MEF2 collaborates with a number of other factors on muscle-specific promoters. In addition, MEF2 often functions as a repressor, in which case MEF2-EnR may serve a gain-of-function role by hyper-repressing target genes.

IV.C.  Direct Transcriptional Targets of MEF2 in the Heart Since the initial identification of MEF2 factors as DNAbinding activities for muscle-specific promoters (Gossett et al., 1989), it is now apparent that MEF2 factors are present in a wide range of different cell types. As such, significant effort has been invested in defining direct transcriptional targets of MEF2. A current review of the literature indicates that there are over 80 known genes for

686

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Table 1  A Subset of the Genes Identified as Direct MEF2 Targets in the Heart or Other Cardiovascular and Related Tissues (Columns Show the Name, Species, and Citations for the Indicated Genes) Gene

Species

References

Actin57B

Drosophila

Kelly et al., 2002

Aldolase A

Human

Hidaka et al., 1993; Salminen et al., 1995

Alpha cardiac actin

Rat, Mouse

Morin et al., 2000; Molinari et al., 2004;

Alpha cardiac MHC

Mouse

Molkentin and Markham, 1993, 1994; Adolph et al., 1993

Lemonnier and Buckinham, 2004 Alpha MHC

Rat

Morin et al., 2000

Alpha T catenin

Human

Vanpoucke et al., 2004

Angiotensin II type Ia

Rat

Beason et al., 1999

Atrial natriuretic factor (Nppa)

Rat

Morin et al., 2000

BOP

Mouse

Phan et al., 2005

Brain natriuretic peptide (BNP)

Rat

Morin et al., 2000

CHAMP

Mouse

Liu et al., 2001

c-jun

Human, Rat

Han and Prywes, 1995; Nadruz et al., 2003

CPTI 

Rat

Baldan et al., 2004

Cytochrome c oxidase VIA

Mouse

Wan and Moreadith, 1995

Desmin

Mouse

Kuisk et al., 1996

Dlx5/6

Mouse

Verzi et al., 2007

Dystrophin

Human

Klamut et al., 1997

Gax

Mouse

Andres et al., 1995

GLUT4

Human

Thai et al., 1998; Mora and Pessin, 2000

HDAC9

Mouse

Haberland et al., 2007

HRC

Mouse

Anderson et al., 2004

Matrix Metalloproteinase 10 (MMP10)

Mouse

Chang et al., 2006

Mef2

Drosophila

Cripps et al., 2004

MicroRNA-1 (Mir1)

Drosophila

Sokol and Ambros, 2005

MLC2

Xenopus

Chambers et al., 1994

MLC2v (cardiac)

Chick, Rat

Goswami and Siddiqui, 1995; Navankasattusas et al., 1993

Muscle creatine kinase (MCK)

Mouse

Amacher et al., 1993; Cserjesi et al., 1994; Qin et al., 1997

Muscle LIM protein 60D

Drosophila

Stronach et al., 1999

Muscle LIM protein 84B

Drosophila

Stronach et al., 1999

Myocardin

Mouse

Creemers et al., 2006

Myomaxin

Mouse

Huang et al., 2006

Myospryn

Mouse

Durham et al., 2006

PGC1 alpha

Mouse

Czubryt et al., 2003; Handschin et al., 2003

Phosphoglycerate mutase

Human

Nakatsuji et al., 1992

Phosphoglycerate mutase

Rat

Ruizlozano et al., 1994; Nakatsuji et al., 1992

SERCA2

Rat

Moriscot et al., 1997

SRPK3

Mouse

Nakagawa et al., 2005

Troponin C (slow, cardiac)

Human, Mouse

Parmacek et al., 1994; Christensen and Kedes, 1999

Troponin I (cardiac)

Human, Rat

Di Lisi et al., 1998; Bhausar et al., 2000

Troponin T (cardiac)

Rat

Wang et al., 1994

Troponin I

Drosophila

Marin et al., 2004

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

which promoter analyses have implicated MEF2 as a direct transcriptional activator. Table 1 contains a partial list of bona fide MEF2 target genes in the heart. Recent genomic analyses have also begun to identify more globally the genes that are regulated by MEF2. Subtractive cloning assays were first used to define genes that were downregulated in Mef2c mutants, and were successful in identifying cardiac-restricted transcripts (Liu et al., 2001b). More recent microarray assays have been informative in delineating the physiological effects of MEF2 loss-of-function, and in defining potential MEF2 targets (Naya et al., 2002; Nakagawa et al., 2005; Sandmann et al., 2006). ChIP-on-Chip assays using antiMEF2 antibodies have determined that there are as many as 670 nonoverlapping genomic regions bound by MEF2 in the developing Drosophila embryo (Sandmann et al., 2006, 2007). This large number of potential targets highlights the pervasive role that MEF2 plays in animal development. In summary, the requirement for Mef2 for muscle differentiation in Drosophila is well-established. In the absence of Mef2 function, cardiac myocytes in the fly are properly specified, but fail to differentiate (Bour et al., 1995; Lilly et al., 1995). It is clear that MEF2 is also important for heart development and function in vertebrates, but the absolute requirement for myocyte differentiation is less clear. As discussed here, loss of Mef2c in mouse results in embryonic lethality due to severe morphogenetic defects in the heart, including a failure to undergo rightward looping (Lin et al., 1997). However, cardiac myocytes clearly form in the absence of Mef2c, and in general, myocytes are normally differentiated in those mutant mice (Lin et al., 1997, 1998; Bi et al., 1999; Vong et al., 2005, 2006). Similarly, loss of Mef2a function in mice causes perinatal lethality in the majority of mice due to cardiac defects, but again, myocytes differentiate properly in Mef2a-null animals (Naya et al., 2002). Zebrafish that lack the function of one or two Mef2 genes have craniofacial and skeletal muscle defects, but no obvious cardiac defects, and muscle differentiation in the cardiac and skeletal muscle lineages appear normal (Hinits and Hughes, 2007; Miller et al., 2007). Given the many commonalities in MEF2-dependent transcriptional pathways that have been conserved between flies and vertebrates, it is reasonable to expect that MEF2 is also required for differentiation of myocytes in vertebrates. A likely explanation for the less severe differentiation phenotypes in mouse and fish Mef2 gene knockouts/knockdowns is the partial redundancy in Mef2 gene function due to the presence of four Mef2 genes in vertebrates compared to invertebrates. Thus, the question remains as to whether MEF2 is really required for myocyte differentiation in vertebrates. As noted earlier in this chapter, studies using dominant-negative forms of MEF2 in C2C12 skeletal myoblasts in culture have strongly implicated a requirement for MEF2 in myoblast differentiation in that system (Ornatsky et al., 1997), and studies

687

using a repressor form of MEF2 in P19 cells also suggest an important role for MEF2 in cardiac muscle differentiation (Karamboulas et al., 2006). Ultimate genetic testing of the requirement of MEF2 for cardiomyocyte differentiation in vertebrates will require additional compound mouse knockouts. It will be important to use multiple approaches to understand the function of Mef2 genes in vivo, including gene disruption, dominant-negative and siRNA approaches. This combined strategy should ultimately elucidate the complex function of MEF2 in the many tissues where Mef2 genes are expressed in vertebrates.

V.  Regulation of mef2 gene transcription V.A.  Mef2 Gene Regulation as a Paradigm for Modular Transcriptional Control Mef2 transcripts show remarkably broad expression in muscle lineages during embryogenesis. While we are now aware that MEF2 expression and function are

(A)

(B)

(C)

(D)

(E)

(F)

Figure 6  Cis-regulatory elements of the Drosophila Mef2 gene. Top: Schematic of the upstream region of Mef2 in Drosophila. The locations of specific enhancers discussed in the text are shown, as well as the transcription factors that regulate these enhancers. (A–F) Activity of Mef2 enhancers visualized by staining for -galactosidase accumulation in transgenic embryos containing Mef2-lacZ fusions. (A) At stage 9 the earliest enhancer is active throughout the invaginating mesoderm. (B) At stage 10 a dorsal mesoderm enhancer, responsive to Dpp, is active. (C) At stage 12, the Tin-dependent enhancer is active in precursors of cardiac muscle (ca), visceral muscle (vm) and a subset of skeletal muscles (sk). (D) At stage 14, a late cardiac enhancer is active, predominantly in Tin cells of the dorsal vessel. (E) An enhancer active in the svp-expression cells is active at stage 14. (F) The Mef2 autoregulatory enhancer functions during larval development in the cardiac tube. All images show anterior to the left and dorsal side uppermost.

688

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

extensively regulated post-transcriptionally, early RNA in situ hybridization studies indicated that a major component of Mef2 gene regulation occurs at the level of transcription. As has been observed for other regulatory genes that have essential and broad functions during embryonic development, Mef2 cis-regulatory regions are composed of a large number of dispersed and independently-functioning enhancers. In Drosophila, the Mef2 transcribed region spans 15 kb, and most of the Drosophila Mef2 enhancers that have been described are contained within a region of 16 kb upstream of the gene (Fig. 6) (Schulz et al., 1996; Nguyen and Xu, 1998). Murine Mef2c regulatory elements appear to be encompassed by a region spanning approximately 200 kb, which also encompasses the transcribed portion of the gene (Fig. 7) (Dodou et al., 2004). Thus, the dynamic pattern of Mef2 gene expression observed in animals arises from the combined actions of a number of independently-functioning regulatory modules that together provide the spatial and temporal signals for Mef2 transcription. Given the fact that Mef2 expression is strongly detected during the specification of multiple lin­ eages, it is likely that that transcription factors that regulate each distinct Mef2 enhancer are themselves critical regulators of cell fate. For this reason, significant effort has focused on the identification and characterization of Mef2 enhancer sequences.

V.B.  Regulation of Mef2 Transcription in the Drosophila Heart In Drosophila embryos, cells contributing to the mature cardiac tube arise as a result of a series of tightly-regulated transcriptional, cell signaling and morphological events (Cripps and Olson, 2002). Presumptive mesodermal cells are specified on the ventral surface of the blastoderm embryo by the influence of the maternal gene product Dorsal, and the activation of the mesodermal determinant Twist. The mesodermal cells subsequently invaginate and spread laterally and dorsally along the ectoderm. Mesodermal cells on each side of the embryo are defined as dorsal mesoderm when they are in proximity with the dorsal ectoderm and receive the TGF molecule Decapentaplegic (Dpp) (Frasch, 1995). Within the dorsal mesoderm, expression of the NK homeodomain transcription factor Tinman (Tin) is responsible for the specification of cardiac mesoderm. Subsequently, a number of inductive processes subdivide the cardiac mesoderm into specific cell types (Frasch, 1995, 1999). A major decision at this point is the designation of cardial cells that either continue to express tin, or activate expression of the orphan nuclear receptor gene seven-up (svp) and downregulate tin transcription (Gajewski et al., 2000; Lo and Frasch, 2001; Lovato et al., 2002). Following these important cell fate decisions, the

MADS

MEF2

* SkM (A) (A)

SHF (B)

Endo (C) (B)

(C)

Figure 7  Cis-regulatory elements of the mouse Mef2c gene. The top shows a schematic of the upstream region of the mouse Mef2c gene. The locations of specific enhancers discussed in the text are shown as colored rectangles (Red: skeletal muscle (SkM); Green: Isl1-dependent SHF enhancer; Blue: Nkx2-5/FoxH1-dependent SHF enhancer; Purple: endothelial (Endo) enhancer). Photographs of X-gal stained transgenic embryos with the SkM, SHF and Endo enhancers directing the expression of -galactosidase are shown in (A), (B) and (C), respectively. An E11.5 embryo is shown in (A); E9.5 embryos are shown in (B) and (C). Known transcriptional starry sites are depicted on the schematic as arrows; exons are depicted as vertical black lines.

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

two rows of cardial cells on either side of the embryo migrate dorsally and meet at the midline to form a linear cardiac tube (Cripps and Olson, 2002). Studies over the past 10 years have identified Mef2 enhancers with activities that overlap temporally and spatially with each of the important morphological and regulatory events contributing to cardiac tube formation, and their activities are shown in Fig. 6. The earliest-acting Mef2 enhancer is responsible for uniform expression throughout the newly-specified mesoderm (Lilly et al., 1995). Genetic studies indicate that the earliest expression of Mef2 might be dependent on the mesodermal determinant Twist, since both gain and loss of Twist activity results in concomitant alterations in the pattern of early Mef2 expression (Lilly et al., 1994; Nguyen et al., 1994; Taylor, 1995). The early expression of Mef2 in the mesoderm is controlled by a proximal regulatory region approximately 2 kb upstream of the transcriptional start site (Fig. 6A) (Lilly et al., 1995; Nguyen and Xu, 1998). Sequences within this region show strong binding to Twist protein in vitro, and ectopic Twist activates the enhancer (Cripps et al., 1998; Nguyen and Xu, 1998). Adjacent to the Twist-responsive enhancer, and possibly overlapping with it, is a Mef2 regulatory region that directs expression to the dorsal mesoderm (Fig. 6B). Dorsal mesoderm is specified in the Drosophila embryo by ectodermal Dpp signaling (Frasch, 1995, 1999). Not surprisingly, the Mef2 dorsal mesoderm enhancer is highly responsive both to Dpp levels and to levels of the Drosophila Smads orthologs, Medea and Mad (Nguyen and Xu, 1998). Medea and Mad bind robustly to sequences within the Mef2 dorsal mesoderm enhancer, but it is not known if the Medea binding sequences are required for enhancer activity (Nguyen and Xu, 1998). The Mef2 dorsal mesoderm enhancer also contains the Twist sites required for early mesoderm expression of Mef2, although it is not clear if these Twist sites contribute to the function of this enhancer (Nguyen and Xu, 1998). As cells within the dorsal mesoderm become specified to a cardiac fate, a third Mef2 enhancer is active in the precardiac cells, beginning around stage 11 (Fig. 6C). This enhancer is located further upstream, at approximately 5.8 kb, and can direct reporter gene activity in precardiac cells of the fly embryo, precursors of the trunk visceral mesoderm and a subset of skeletal muscles (Gajewski et al., 1997, 1998; Cripps et al., 1999). The enhancer contains two binding sites for the NK homeodomain factor Tin, and mutation of either or both Tin sites completely ablates the activity of this enhancer in all three of the muscle lineages (Gajewski et al., 1997, 1998; Cripps et al., 1999). At stage 14, another Mef2 enhancer predominantly active in the Tin cardiac cells, has been identified (Fig. 6D) (Nguyen and Xu, 1998). This enhancer appears to support strong reporter gene activity in the cardiac tissue through the end of embryogenesis. Factors regulating the activity

689

of this enhancer have yet to be defined, although the region does contain a consensus-binding site for the Tin protein. From stage 13 onwards, the maturing dorsal vessel comprises mutually exclusive tin and svp-expressing cardial cells (Gajewski et al., 2000; Ward and Skeath, 2000; Lo and Frasch, 2003). Therefore, it was reasonable to postulate that a Mef2 enhancer active in the svp-expressing cells of the dorsal vessel must exist. Early data suggested that this was a relatively distal enhancer (Schulz et al., 1996), and sequences responsible for this activity have now been identified (Fig. 6E) (Gajewski et al., 2000). Deletional analyses indicate that the regulation of this region is complex, since several discrete elements are required for enhancer activity, and transcriptional regulators of this sequence have yet to be identified (Gajewski et al., 2000). A final Mef2 enhancer that is active in cardiac tissue in Drosophila shows a relatively broad pattern of activity in the mature muscles of the developing animal. This late muscle enhancer is located at approximately 9 kb relative to the transcriptional start site, and becomes active in visceral and skeletal muscle during embryogenesis (Nguyen and Xu, 1998: Cripps et al., 2004). Cardiac activity of this enhancer is not detected until larval life (Fig. 6F). The function of this late heart enhancer depends specifically on the integrity of a binding site for MEF2 itself, indicating that the Mef2 gene is subject to positive autoregulation (Cripps et al., 2004). This enhancer probably functions as a late-maintenance element for mature muscles. The activities of the six Drosophila Mef2 enhancers described above provide a detailed and discrete description of the fate of cardiac cells, from the time when the cells are first specified as mesoderm in the embryo to the time when they power circulation in the developing larva. As indicated above, regulators of Mef2 transcription are critical regulators of mesodermal cell fate decisions, underlining the importance of MEF2 in regulating mesodermal development. For most of these Drosophila enhancers, only single regulators have been identified to date. This suggests that additional critical muscle-specific activators of Mef2 are still to be identified, since the majority of well-characterized enhancers are activated combinatorially. Future studies should resolve how the activities of these many distinct enhancers are integrated to give the complex and precise pattern of Mef2 expression in Drosophila.

V.C.  Regulation of Mef2c Transcription in the Mammalian Heart The large and complex organization of mammalian Mef2 genes has hindered the identification of regulators of Mef2 transcription in higher animals, although recent studies are providing critical insight into the regulatory elements of the mouse Mef2c gene. Mouse Mef2c spans almost 200 kb,

690

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

and comprises at least 14 exons. Alternative promoters are used to generate transcripts that are restricted to skeletal muscle, heart or brain (Fig. 7). Each of these alternate first exons are spliced to common exons encoding the MADS and MEF2 domains (Wang et al., 2001; Zhu and Gulick, 2004; Zhu et al., 2005). The first Mef2c enhancer that was described was identified based on its proximity to a skeletal muscle-specific exon (Wang et al., 2001). Accordingly, this enhancer is active in all skeletal muscles and their precursors, beginning at E9.0 (Fig. 7A). Deletion and mutagenesis studies of this sequence revealed a critical requirement for a MyoD E box (Wang et al., 2001; Dodou et al., 2003). In addition, an AT-rich sequence, which might comprise a MEF2 binding site, was also required for maximal activation of the enhancer (Wang et al., 2001; Dodou et al., 2003). While this enhancer is not activated in cardiac tissue, these studies importantly demonstrated that mammalian Mef2 genes are likely to be regulated by dispersed modular enhancers, similar to the situation observed for Drosophila. Mef2c-regulatory regions that show activity in cardiovascular tissues have more recently been identified. One element, located adjacent to the heart-specific exon identified by Wang et al. is responsible for activation of Mef2c transcription in the anterior, or second, heart field (Fig. 7B) (Wang et al., 2001; Dodou et al., 2004) (see Chapter 2.2). This enhancer, termed Mef2c AHF, becomes activated very early during cardiac cell commitment at E7.5, and its activity is restricted to cells of the second heart field and its descendants (Dodou et al., 2004). The function of this enhancer is absolutely dependent on the integrity of binding sites for GATA factors and the LIM-homeodomain transcription factor Islet-1 (Dodou et al., 2004), which has been shown recently to function at or near the top of a transcriptional network for second heart field development (Cai et al., 2003; Black, 2007). The isolation and characterization of the Mef2c AHF enhancer has enabled detailed studies of the second heart field and its contributions to the heart. By combining transgenic mice in which the Mef2c enhancer is fused to Cre-recombinase, with mice carrying floxed alleles of a variety of genes, studies ranging from lineage tracing analyses (Verzi et al., 2005) to tissue-specific knockouts (Ai et al., 2006, 2007; Park et al., 2006; Goddeeris et al., 2007) have been achieved. Identification of other Mef2 enhancers showing specificity to other important cell lineages will likely provide similar essential reagents for studying cardiac development. A third Mef2c enhancer, located just upstream of exon 4, activates reporter gene expression broadly in the vascular endothelial cells of transgenic animals, but not in vascular smooth muscle cells (De Val et al., 2004). In vivo, the enhancer first becomes active in blood islands of the yolk sac at E7.5. Subsequently, the Mef2c F7 endothelial enhancer becomes active throughout the vascular

endothelium during embryogenesis, and in most adult vascular endothelial cells (Fig. 7C). Mutational analyses of the enhancer revealed a requirement for Ets transcription factor binding sites for enhancer function in vivo (De Val et al., 2004). Given the broad requirement for Ets proteins in the formation of vascular endothelial cells (Parmacek, 2001), and given the requirement for Mef2c function in vascular development (Lin et al., 1998; Bi et al., 1999), the identification of Mef2c as a target of Ets factors defines an important component of the gene regulatory network for the specification and differentiation of vascular endothelial cells. More recently, an additional endothelial-specific enhancer from the mouse Mef2c gene was identified (De Val et al., 2008). This enhancer, termed Mef2c F10, is activated exclusively in endothelial cells and their progenitors, beginning at the blood island stage (E7.5), but init­ iates prior to Mef2c F7 (De Val et al., 2004, 2008). Mef2c F10 contains a deeply conserved 44 bp core element that is both necessary and sufficient for enhancer function in vivo. This small region contains a composite-binding motif for Forkhead and Ets transcription factors, and this motif is simultaneously bound and synergistically activated by the Forkhead protein FoxC2 and the Ets protein Etv2 (De Val et al., 2008). Interestingly, the FOX:ETS motif, first discovered in Mef2c, is prevalent throughout the human genome, and is strongly-associated with endothelial genes and enhancers. All FOX:ETS-containing enhancers tested to date are synergistically activated by FoxC2 and Etv2, and these two transcription factors are sufficient to induce ectopic vasculogenesis in Xenopus embryos (De Val et al., 2008), suggesting that combinatorial activation of Mef2c by FoxC2 and Etv2 is part of a much broader cooperative mechanism for endothelial gene activation. By investigating the role of Foxh1 in the formation of the second heart field, von Both et al. showed that Foxh1/ embryos displayed cardiac defects highly similar to those observed for Mef2c mutants, including failure to form the right ventricle and outflow tract (von Both et al., 2004). To determine if this result arose from a direct transcriptional interaction between FoxH1 and the Mef2c gene, the authors searched for candidate FoxH1-binding regions in Mef2c. These analyses identified an additional putative enhancer responsive to FoxH1 and Nkx2-5 (Fig. 7) (von Both et al., 2004). These observations are potentially interesting, since mutations that affect two different genes show similar phenotypes, and this similarity may be due to a direct transcriptional regulatory interaction between the two genes. Second, the identification of two distinct second heart field enhancers from the Mef2c gene raises the possibility that there might be redundant enhancers functioning for this cell type. Alternatively, the expression patterns of the two enhancers might, on further investigation, show subtle yet important functional differences. Third, the activation of Mef2c by Nkx2-5 is highly reminiscent of the activation of Drosophila

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

Mef2 in cardiac precursors by the Drosophila NK homeodomain factor Tinman, underlining the ancient evolution of the transcriptional network that functions during cardiac development (see Chapter 1.2). It will be interesting to identify additional cardiac enhancers from Mef2c in order to determine how all of these elements are coordinated and integrated to govern Mef2c expression in the heart. The Mef2 gene transcriptional regulatory modules described in this chapter have been studied in isolation, following fusion of a wild-type or mutant enhancer to a reporter gene and analysis of reporter gene expression in transgenic animals. Whether there is any interplay among discrete enhancers that show partially-overlapping temporal or spatial activities, or whether the activation of one enhancer sequence specifically affects the activation of another has yet to be determined. Furthermore, the nature of global chromatin organization at the Mef2 gene loci has yet to be defined in detail. Why should Mef2 genes be subject to such complex transcriptional regulation? Given the broad expression of Mef2, it might seem more efficient if Mef2 transcription were activated and then subjected to positive autoregulation to maintain its expression. However, this mechanism might be too rigid, since there are a number of instances where the expression of specific Mef2 genes is lost from a particular tissue or lineage. For example, all cells of the Drosophila dorsal vessel arise from the MEF2-positive dorsal mesoderm; however, the accessory pericardial cells downregulate Mef2 expression and are MEF2-negative in mature embryos (Cripps and Olson, 2002). Therefore, the complexity of Mef2 gene regulation probably allows for greater modulation of levels of gene transcription as new cell types are specified and subsequently differentiate. How the regulation of the many distinct enhancers from all of the various Mef2 genes are integrated to give rise to the endogenous pattern of Mef2 transcription are important questions that remain to be resolved by future studies.

VI.  Future directions MEF2 proteins function as a transcriptional switch in many diverse lineages during development and in adulthood. To accomplish this, MEF2 has evolved numerous regulatory interactions. As discussed in this chapter, MEF2 interacts with a wide array of co-factors, including both positive and negative regulators of transcription. MEF2 serves as an essential binding platform for enzymes that influence chromatin structure, including histone acetylases and deacetylases. Similarly, MEF2 serves as a platform for potent transcriptional regulators that do not directly contact DNA themselves, such as myocardin and MASTR. How does MEF2 interact with such a diverse group of cofactors and accomplish seemingly opposing influences on transcription, in some cases acting as a repressor and in

691

others as an activator? In part, the answer to this question appears to lie in the post-translational regulation of MEF2 by many diverse signaling pathways. Numerous different signaling pathways that influence MEF2 function and protein–protein interactions via phosphorylation were discussed in this chapter. However, many of these pathways function in the same cell types and in response to the same signals. Thus, it still remains unclear how MEF2 integrates and interprets these distinct and often opposing inputs to control transcription into precise developmental or physio­ logical outputs. Future investigation in this area will certainly focus on defining how various post-translational modifications of MEF2 interact with and influence each other. In turn, it will be important to determine how posttranslational modifications of MEF2 affect its structure and protein–protein interactions. Given the fundamental role of MEF2 proteins in so many biological processes, it is not surprising that MEF2 expression and function is tightly-regulated, and that this regulation is complex and occurs at many levels. Much has been learned over the last decade about the posttranslational regulation of MEF2 function, and we are also beginning to decipher the complex transcriptional regulation of Mef2 genes. However, much remains to be defined about the regulation of this transcription factor family. As noted above, it is still not clear how the many post-translational modifications of MEF2 are integrated. It is also not clear how the many modular enhancers of Mef2c work in the context of the genome, and nothing is known currently about the transcriptional regulation of other vertebrate Mef2 genes. Recent studies have uncovered a fundamental regulatory role for microRNAs in muscle gene regulation by targeting 3 untranslated regions (UTRs) of muscle genes (Zhao et al., 2005b; Chen et al., 2006; van Rooij et al., 2006; Callis et al., 2007; van Rooij and Olson, 2007; van Rooij et al., 2007). Interestingly, the 3UTRs of the vertebrate Mef2 genes are evolutionarily conserved, suggesting a possible biological function. Indeed, the 3UTR of the Mef2a transcript has been shown to be a cis-acting translational repressor, although the mechanism is not clear (Black et al., 1997). It remains to be determined whether the Mef2a 3UTR or the UTRs of other Mef2 genes are regulated by microRNAs, although it seems likely since fine-tuned control of Mef2 gene expression and function appears to occur at almost every other level of regulation. Mutations that result in haploinsufficiency of any of several transcription factor genes are known to cause or contribute to human disease. These include genes for several transcription factors that function in the core cardiac network, including TBX5, GATA4 and NKX2-5 (Basson et al., 1997; Schott et al., 1998; Pehlivan et al., 1999; Giglio et al., 2000; Garg et al., 2003; Mori and Bruneau, 2004; Schluterman et al., 2007; Pabst et al., 2008). Interestingly, MEF2 family members also function in the core network, and as noted in this chapter, MEF2C has been shown to

692

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

interact transcriptionally or physically with Nkx2-5, GATA4 and HAND2 (Skerjanc et al., 1998; Morin et al., 2000; Vanpoucke et al., 2004; Zang et al., 2004). Together, these observations suggest that mutations in human MEF2 genes may also influence heart development or function. Furthermore, the role of MEF2 transcription factors in hypertrophic responsiveness and cardiomyopathy in response to pathologic stimuli in mice is well-established (Backs and Olson, 2006; Liu and Olson, 2006; Olson et al., 2006). Thus, it is reasonable to expect that MEF2 genes may be involved in human heart failure as well. Indeed, some studies have suggested a role for MEF2A in inherited coronary artery disease in humans where a 21 bp deletion in the MEF2A gene may be correlated with late-onset disease (Wang et al., 2003a; Bhagavatula et al., 2004). However, other studies have suggested that the 21 bp deletion, which results in an in-frame loss of 7 amino acids in the C-terminus of MEF2A, is a rare but normal allelic variant (Altshuler and Hirschhorn, 2005; Kajimoto et al., 2005; Wang et al., 2005; Weng et al., 2005; Horan et al., 2006). Therefore, it still remains unclear what role, if any, MEF2 genes may play in human cardiovascular disease. This will be an important area of future investigation, and human genetic studies may identify even more previously unappreciated roles for MEF2 transcription factors in development and disease.

Acknowledgments We are grateful to Mike Verzi, John Schwarz and Will Schachterle for providing embryo pictures shown in Figs 5 and 7, and we thank Cheryl Sensibaugh for providing Fig. 1C. The Black and Cripps laboratories are supported by grants from the NIH.

References Ai, D., Liu, W., Ma, L., Dong, F., Lu, M.F., Wang, D., Verzi, M.P., Cai, C., Gage, P.J., Evans, S., Black, B.L., Brown, N.A., Martin, J.F., 2006. Pitx2 regulates cardiac left–right asymmetry by patterning second cardiac lineage-derived myocardium. Dev. Biol. 296, 437–449. Ai, D., Fu, X., Wang, J., Lu, M.F., Chen, L., Baldini, A., Klein, W.H., Martin, J.F., 2007. Canonical Wnt signaling functions in second heart field to promote right ventricular growth. Proc. Natl. Acad. Sci. USA 104, 9319–9324. Altshuler, D., Hirschhorn, J.N., 2005. MEF2A sequence variants and coronary artery disease: a change of heart? J. Clin. Invest. 115, 831–833. Alvarez-Buylla, E.R., Pelaz, S., Liljegren, S.J., Gold, S.E., Burgeff, C., Ditta, G.S., Ribas de Pouplana, L., Martinez-Castilla, L., Yanofsky, L., 2000. An ancestral MADS-box gene duplication occurred before the divergence of plants and animals. Proc. Natl. Acad. Sci. USA 97, 5328–5333. Arnold, M.A., Kim, Y., Czubryt, M.P., Phan, D., McAnally, J., Qi, X., Shelton, J.M., Richardson, J.A., Bassel-Duby, R., Olson, E.N., 2007. MEF2C transcription factor controls chondrocyte hypertrophy and bone development. Dev. Cell 12, 377–389.

Backs, J., Olson, E.N., 2006. Control of cardiac growth by histone acetylation/deacetylation. Circ. Res. 98, 15–24. Basson, C.T., Bachinsky, D.R., Lin, R.C., Levi, T., Elkins, J.A., Soults, J., Grayzel, D., Kroumpouzou, E., Traill, T.A., Leblanc-Straceski, J., Renault, B., Kucherlapati, R., Seidman, J.G., Seidman, C.E., 1997. Mutations in human TBX5 [corrected] cause limb and cardiac malformation in Holt-Oram syndrome. Nat. Genet. 15, 30–35. Becker, A., Theissen, G., 2003. The major clades of MADS-box genes and their role in the development and evolution of flowering plants. Mol. Phylogenet. Evol. 29, 464–489. Berdeaux, R., Goebel, N., Banaszynski, L., Takemori, H., Wandless, T., Shelton, G.D., Montminy, M., 2007. SIK1 is a class II HDAC kinase that promotes survival of skeletal myocytes. Nat. Med. 13, 597–603. Bertos, N.R., Wang, A.H., Yang, X.J., 2001. Class II histone deacetyl­ ases: structure, function, and regulation. Biochem. Cell Biol. 79, 243–252. Bhagavatula, M.R., Fan, C., Shen, G.Q., Cassano, J., Plow, E.F., Topol, E.J., Wang, Q., 2004. Transcription factor MEF2A mutations in patients with coronary artery disease. Hum. Mol. Genet. 13, 3181–3188. Bi, W., Drake, C.J., Schwarz, J.J., 1999. The transcription factor MEF2Cnull mouse exhibits complex vascular malformations and reduced cardiac expression of angiopoietin 1 and VEGF. Dev. Biol. 211, 255–267. Black, B.L., 2007. Transcriptional pathways in second heart field development. Semin. Cell Dev. Biol. 18, 67–76. Black, B.L., Ligon, K.L., Zhang, Y., Olson, E.N., 1996. Cooperative transcriptional activation by the neurogenic basic helix-loop-helix protein MASH1 and members of the myocyte enhancer factor-2 (MEF2) family. J. Biol. Chem. 271, 26659–26663. Black, B.L., Lu, J., Olson, E.N., 1997. The MEF2A 3 untranslated region functions as a cis-acting translational repressor. Mol. Cell Biol. 17, 2756–2763. Black, B.L., Olson, E.N., 1998. Transcriptional control of muscle development by myocyte enhancer factor-2 (MEF2) proteins. Annu. Rev. Cell Dev. Biol. 14, 167–196. Black, B.L., Molkentin, J.D., Olson, E.N., 1998. Multiple roles for the MyoD basic region in transmission of transcriptional activation signals and interaction with MEF2. Mol. Cell Biol. 18, 69–77. Borghi, S., Molinari, S., Razzini, G., Parise, F., Battini, R., Ferrari, S., 2001. The nuclear localization domain of the MEF2 family of transcription factors shows member-specific features and mediates the nuclear import of histone deacetylase 4. J. Cell Sci. 114, 4477–4483. Bour, B.A., O’Brien, M.A., Lockwood, W.L., Goldstein, E.S., Bodmer, R., Taghert, P.H., Abmayr, S.M., Nguyen, H.T., 1995. Drosophila MEF2, a transcription factor that is essential for myogenesis. Genes. Dev. 9, 730–741. Breitbart, R.E., Liang, C.S., Smoot, L.B., Laheru, D.A., Mahdavi, V., Nadal-Ginard, B., 1993. A fourth human MEF2 transcription factor, hMEF2D, is an early marker of the myogenic lineage. Development 118, 1095–1106. Cai, C.L., Liang, X., Shi, Y., Chu, P.H., Pfaff, S.L., Chen, J., Evans, S., 2003. Isl1 identifies a cardiac progenitor population that proliferates prior to differentiation and contributes a majority of cells to the heart. Dev. Cell 5, 877–889. Callis, T.E., Chen, J.F., Wang, D.Z., 2007. MicroRNAs in skeletal and cardiac muscle development. DNA Cell Biol. 26, 219–225. Chambers, A.E., Kotecha, S., Towers, N., Mohun, T.J., 1992. Musclespecific expression of SRF-related genes in the early embryo of Xenopus laevis. Embo. J. 11, 4981–4991.

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

Chan, J.K., Sun, L., Yang, X.J., Zhu, G., Wu, Z., 2003. Functional characterization of an amino-terminal region of HDAC4 that possesses MEF2 binding and transcriptional repressive activity. J. Biol. Chem. 278, 23515–23521. Chang, C.I., Xu, B.E., Akella, R., Cobb, M.H., Goldsmith, E.J., 2002. Crystal structures of MAP kinase p38 complexed to the docking sites on its nuclear substrate MEF2A and activator MKK3b. Mol. Cell 9, 1241–12459. Chen, J.F., Mandel, E.M., Thomson, J.M., Wu, Q., Callis, T.E., Hammond, S.M., Conlon, F.L., Wang, D.Z., 2006. The role of microRNA-1 and microRNA-133 in skeletal muscle proliferation and differentiation. Nat. Genet. 38, 228–233. Chen, S.L., Dowhan, D.H., Hosking, B.M., Muscat, G.E., 2000. The steroid receptor coactivator, GRIP-1, is necessary for MEF-2Cdependent gene expression and skeletal muscle differentiation. Genes. Dev. 14, 1209–1228. Choi, S.J., Park, S.Y., Han, T.H., 2001. 14-3-3tau associates with and activates the MEF2D transcription factor during muscle cell differentiation. Nucleic. Acids. Res. 29, 2836–2842. Cox, D.M., Du, M., Marback, M., Yang, E.C., Chan, J., Siu, K.W., McDermott, J.C., 2003. Phosphorylation motifs regulating the stability and function of myocyte enhancer factor 2A. J. Biol. Chem. 278, 15297–15303. Creemers, E.E., Sutherland, L.B., Oh, J., Barbosa, A.C., Olson, E.N., 2006. Coactivation of MEF2 by the SAP domain proteins myocardin and MASTR. Mol. Cell 23, 83–96. Cripps, R.M., Black, B.L., Zhao, B., Lien, C.L., Schulz, R.A., Olson, E.N., 1998. The myogenic regulatory gene Mef2 is a direct target for transcriptional activation by Twist during Drosophila myogenesis. Genes. Dev. 12, 422–434. Cripps, R.M., Zhao, B., Olson, E.N., 1999. Transcription of the myogenic regulatory gene Mef2 in cardiac, somatic, and visceral muscle cell lineages is regulated by a Tinman-dependent core enhancer. Dev. Biol. 215, 420–430. Cripps, R.M., Olson, E.N., 2002. Control of cardiac development by an evolutionarily conserved transcriptional network. Dev. Biol. 246, 14–28. Cripps, R.M., Lovato, T.L., Olson, E.N., 2004. Positive autoregulation of the Myocyte enhancer factor-2 myogenic control gene during somatic muscle development in Drosophila. Dev. Biol. 267, 536–547. Czubryt, M.P., Olson, E.N., 2004. Balancing contractility and energy production: the role of myocyte enhancer factor 2 (MEF2) in cardiac hypertrophy. Recent Prog. Horm. Res. 59, 105–124. Damm, C., Wolk, A., Buttgereit, D., Loher, K., Wagner, E., Lilly, B., Olson, E.N., Hasenpusch-Theil, K., Renkawitz-Pohl, R., 1998. Independent regulatory elements in the upstream region of the Drosophila beta 3 tubulin gene (beta Tub60D) guide expression in the dorsal vessel and the somatic muscles. Dev. Biol. 199, 138–149. Davidson, B., 2007. Ciona intestinalis as a model for cardiac development. Semin. Cell Dev. Biol. 18, 16–26. Davis, R.L., Weintraub, H., Lassar, A.B., 1987. Expression of a single transfected cDNA converts fibroblasts to myoblasts. Cell 51, 987–1000. De Luca, A., Severino, A., De Paolis, P., Cottone, G., De Luca, L., De Falco, M., Porcellini, A., Volpe, M., Condorelli, G., 2003. p300/ cAMP-response-element-binding-protein (“CREB”)-binding protein (CBP) modulates co-operation between myocyte enhancer factor 2A (MEF2A) and thyroid hormone receptor-retinoid X receptor. Biochem. J. 369, 477–484.

693

De Val, S., Anderson, J.P., Heidt, A.B., Khiem, D., Xu, S.M., Black, B.L., 2004. Mef2c is activated directly by Ets transcription factors through an evolutionarily conserved endothelial cell-specific enhancer. Dev. Biol. 275, 424–434. De Val, S., Chi, N.C., Meadows, S.M., Minovitsky, S., Anderson, J.P., Harris, I.S., Ehlers, M.L., Agarwal, P., Visel, A., Xu, S.M., Pennacchio, L.A., Dubchak, I., Krieg, P.A., Stainier, D.Y., Black, B.L., 2008. Combinatorial regulation of endothelial gene expression by Ets and Forkhead transcription factors. Cell 135, 1053–1064. Derynck, R., Zhang, Y.E., 2003. Smad-dependent and Smad-independent pathways in TGF-beta family signalling. Nature 425, 577–584. Dodou, E., Xu, S.M., Black, B.L., 2003. Mef2c is activated directly by myogenic basic helix-loop-helix proteins during skeletal muscle development in vivo. Mech. Dev. 120, 1021–1032. Dodou, E., Verzi, M.P., Anderson, J.P., Xu, S.M., Black, B.L., 2004. Mef2c is a direct transcriptional target of ISL1 and GATA factors in the anterior heart field during mouse embryonic development. Development 131, 3931–3942. Dressel, U., Bailey, P.J., Wang, S.C., Downes, M., Evans, R.M., Muscat, G.E., 2001. A dynamic role for HDAC7 in MEF2-mediated muscle differentiation. J. Biol. Chem. 276, 17007–17013. Durham, J.T., Brand, O.M., Arnold, M., Reynolds, J.G., Muthukumar, L., Weiler, H., Richardson, J.A., Naya, F.J., 2006. Myospryn is a direct transcriptional target for MEF2A that encodes a striated muscle, alpha-actinin-interacting, costamere-localized protein. J. Biol. Chem. 281, 6841–6849. Edmondson, D.G., Lyons, G.E., Martin, J.F., Olson, E.N., 1994. Mef2 gene expression marks the cardiac and skeletal muscle lineages during mouse embryogenesis. Development 120, 1251–1263. Esau, C., Boes, M., Youn, H.D., Tatterson, L., Liu, J.O., Chen, J., 2001. Deletion of calcineurin and myocyte enhancer factor 2 (MEF2) binding domain of Cabin1 results in enhanced cytokine gene expression in T cells. J. Exp. Med. 194, 1449–1459. Fickett, J.W., 1996. Coordinate positioning of MEF2 and myogenin binding sites. Gene 172, GC19–GC32. Flavell, S.W., Cowan, C.W., Kim, T.K., Greer, P.L., Lin, Y., Paradis, S., Griffith, E.C., Hu, L.S., Chen, C., Greenberg, M.E., 2006. Activitydependent regulation of MEF2 transcription factors suppresses excitatory synapse number. Science 311, 1008–1012. Frasch, M., 1995. Induction of visceral and cardiac mesoderm by ectodermal Dpp in the early Drosophila embryo. Nature 374, 464–467. Frasch, M., 1999. Intersecting signalling and transcriptional pathways in Drosophila heart specification. Semin. Cell Dev. Biol. 10, 61–71. Gajewski, K., Kim, Y., Lee, Y.M., Olson, E.N., Schulz, R.A., 1997. Dmef2 is a target for Tinman activation during Drosophila heart development. Embo. J. 16, 515–522. Gajewski, K., Kim, Y., Choi, C.Y., Schulz, R.A., 1998. Combinatorial control of Drosophila mef2 gene expression in cardiac and somatic muscle cell lineages. Dev. Genes Evol. 208, 382–392. Gajewski, K., Choi, C.Y., Kim, Y., Schulz, R.A., 2000. Genetically distinct cardial cells within the Drosophila heart. Genesis 28, 36–43. Garg, V., Kathiriya, I.S., Barnes, R., Schluterman, M.K., King, I.N., Butler, C.A., Rothrock, C.R., Eapen, R.S., Hirayama-Yamada, K., Joo, K., Matsuoka, R., Cohen, J.C., Srivastava, D., 2003. GATA4 mutations cause human congenital heart defects and reveal an interaction with TBX5. Nature 424, 443–447. Giglio, S., Graw, S.L., Gimelli, G., Pirola, B., Varone, P., Voullaire, L., Lerzo, F., Rossi, E., Dellavecchia, C., Bonaglia, M.C., Digilio, M.C., Giannotti, A., Marino, B., Carrozzo, R., Korenberg, J.R., Danesino, C., Sujansky, E., Dallapiccola, B., Zuffardi, O., 2000. Deletion of a 5-cM

694

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

region at chromosome 8p23 is associated with a spectrum of congenital heart defects. Circulation 102, 432–437. Gill, G., 2005. Something about SUMO inhibits transcription. Curr. Opin. Genet. Dev. 15, 536–541. Goddeeris, M.M., Schwartz, R., Klingensmith, J., Meyers, E.N., 2007. Independent requirements for Hedgehog signaling by both the anterior heart field and neural crest cells for outflow tract development. Development 134, 1593–1604. Gong, X., Tang, X., Wiedmann, M., Wang, X., Peng, J., Zheng, D., Blair, L.A., Marshall, J., Mao, Z., 2003. Cdk5-mediated inhibition of the protective effects of transcription factor MEF2 in neurotoxicityinduced apoptosis. Neuron 38, 33–46. Gossett, L.A., Kelvin, D.J., Sternberg, E.A., Olson, E.N., 1989. A new myocyte-specific enhancer-binding factor that recognizes a conserved element associated with multiple muscle-specific genes. Mol. Cell Biol. 9, 5022–5033. Gregoire, S., Yang, X.J., 2005. Association with class IIa histone deacetylases upregulates the sumoylation of MEF2 transcription factors. Mol. Cell Biol. 25, 2273–2287. Gregoire, S., Tremblay, A.M., Xiao, L., Yang, Q., Ma, K., Nie, J., Mao, Z., Wu, Z., Giguere, V., Yang, X.J., 2006. Control of MEF2 transcriptional activity by coordinated phosphorylation and sumoylation. J. Biol. Chem. 281, 4423–4433. Gregoire, S., Xiao, L., Nie, J., Zhang, X., Xu, M., Li, J., Wong, J., Seto, E., Yang, X.J., 2007. Histone deacetylase 3 interacts with and deacetylates myocyte enhancer factor 2. Mol. Cell Biol. 27, 1280–1295. Han, A., Pan, F., Stroud, J.C., Youn, H.D., Liu, J.O., Chen, L., 2003. Sequence-specific recruitment of transcriptional co-repressor Cabin1 by myocyte enhancer factor-2. Nature 422, 730–734. Han, A., He, J., Wu, Y., Liu, J.O., Chen, L., 2005. Mechanism of recruitment of class II histone deacetylases by myocyte enhancer factor-2. J. Mol. Biol. 345, 91–102. Han, J., Jiang, Y., Li, Z., Kravchenko, V.V., Ulevitch, R.J., 1997. Activation of the transcription factor MEF2C by the MAP kinase p38 in inflammation. Nature 386, 296–299. Han, J., Molkentin, J.D., 2000. Regulation of MEF2 by p38 MAPK and its implication in cardiomyocyte biology. Trends Cardiovasc. Med. 10, 19–22. Hinits, Y., Hughes, S.M., 2007. Mef2s are required for thick filament formation in nascent muscle fibres. Development 134, 2511–2519. Horan, P.G., Allen, A.R., Hughes, A.E., Patterson, C.C., Spence, M., McGlinchey, P.G., Belton, C., Jardine, T.C., McKeown, P.P., 2006. Lack of MEF2A Delta7aa mutation in Irish families with early onset ischaemic heart disease, a family based study. BMC Med. Genet. 7, 65. Horlick, R.A., Benfield, P.A., 1989. The upstream muscle-specific enhancer of the rat muscle creatine kinase gene is composed of multiple elements. Mol. Cell Biol. 9, 2396–2413. Huang, H.T., Brand, O.M., Mathew, M., Ignatiou, C., Ewen, E.P., McCalmon, S.A., Naya, F.J., 2006. Myomaxin is a novel transcriptional target of MEF2A that encodes a Xin-related alphaactinin-interacting protein. J. Biol. Chem. 281, 39370–39379. Huang, K., Louis, J.M., Donaldson, L., Lim, F.L., Sharrocks, A.D., Clore, G.M., 2000. Solution structure of the MEF2A-DNA complex: structural basis for the modulation of DNA bending and specificity by MADS-box transcription factors. Embo. J. 19, 2615–2628. Ikeshima, H., Imai, S., Shimoda, K., Hata, J., Takano, T., 1995. Expression of a MADS box gene, MEF2D, in neurons of the mouse central nervous system: implication of its binary function in myogenic and neurogenic cell lineages. Neurosci. Lett. 200, 117–120.

Jang, H., Choi, D.E., Kim, H., Cho, E.J., Youn, H.D., 2007. Cabin1 represses MEF2 transcriptional activity by association with a methyltransferase, SUV39H1. J. Biol. Chem. 282, 11172–11179. Kajimoto, K., Shioji, K., Tago, N., Tomoike, H., Nonogi, H., Goto, Y., Iwai, N., 2005. Assessment of MEF2A mutations in myocardial infarction in Japanese patients. Circ. J. 69, 1192–1195. Kang, J., Gocke, C.B., Yu, H., 2006. Phosphorylation-facilitated sumoylation of MEF2C negatively regulates its transcriptional activity. BMC Biochem. 7, 5. Kao, H.Y., Verdel, A., Tsai, C.C., Simon, C., Juguilon, H., Khochbin, S., 2001. Mechanism for nucleocytoplasmic shuttling of histone deacetyl­ ase 7. J. Biol. Chem. 276, 47496–47507. Karamboulas, C., Dakubo, G.D., Liu, J., De Repentigny, Y., Yutzey, K., Wallace, V.A., Kothary, R., Skerjanc, I.S., 2006. Disruption of MEF2 activity in cardiomyoblasts inhibits cardiomyogenesis. J. Cell Sci. 119, 4315–4321. Karasseva, N., Tsika, G., Ji, J., Zhang, A., Mao, X., Tsika, R., 2003. Transcription enhancer factor 1 binds multiple muscle MEF2 and A/ T-rich elements during fast-to-slow skeletal muscle fiber type transitions. Mol. Cell Biol. 23, 5143–5164. Kasler, H.G., Victoria, J., Duramad, O., Winoto, A., 2000. ERK5 is a novel type of mitogen-activated protein kinase containing a transcriptional activation domain. Mol. Cell Biol. 20, 8382–8389. Kato, Y., Kravchenko, V.V., Tapping, R.I., Han, J., Ulevitch, R.J., Lee, J.D., 1997. BMK1/ERK5 regulates serum-induced early gene expression through transcription factor MEF2C. Embo. J. 16, 7054–7066. Kato, Y., Zhao, M., Morikawa, A., Sugiyama, T., Chakravortty, D., Koide, N., Yoshida, T., Tapping, R.I., Yang, Y., Yokochi, T., Lee, J.D., 2000. Big mitogen-activated kinase regulates multiple members of the MEF2 protein family. J. Biol. Chem. 275, 18534–18540. Kelly, K.K., Meadows, S.M., Cripps, R.M., 2002. Drosophila MEF2 is a direct regulator of Actin57B transcription in cardiac, skeletal, and visceral muscle lineages. Mech. Dev. 110, 39–50. Kelly, R.G., Buckingham, M.E., 2002. The anterior heart-forming field: voyage to the arterial pole of the heart. Trends Genet. 18, 210–216. Khiem, D., Cyster, J.G., Schwarz, J.J., Black, B.L., 2008. A p38 MAPKMEF2C pathway regulates B-cell proliferation. Proc. Natl. Acad. Sci. USA 105, 17067–17072. Kim, T.G., Jung, J., Mysliwiec, M.R., Kang, S., Lee, Y., 2005. Jumonji represses alpha-cardiac myosin heavy chain expression via inhibiting MEF2 activity. Biochem. Biophys. Res. Commun. 329, 544–553. Kolodziejczyk, S.M., Wang, L., Balazsi, K., DeRepentigny, Y., Kothary, R., Megeney, L.A., 1999. MEF2 is upregulated during cardiac hypertrophy and is required for normal post-natal growth of the myocardium. Curr. Biol. 9, 1203–1206. Kopan, R., Nye, J.S., Weintraub, H., 1994. The intracellular domain of mouse Notch: a constitutively activated repressor of myogenesis directed at the basic helix-loop-helix region of MyoD. Development 120, 2385–2396. Kuroda, K., Tani, S., Tamura, K., Minoguchi, S., Kurooka, H., Honjo, T., 1999. Delta-induced Notch signaling mediated by RBP-J inhibits MyoD expression and myogenesis. J. Biol. Chem. 274, 7238–7244. Lavine, K.J., Yu, K., White, A.C., Zhang, X., Smith, C., Partanen, J., Ornitz, D.M., 2005. Endocardial and epicardial derived FGF signals regulate myocardial proliferation and differentiation in vivo. Dev. Cell 8, 85–95. Lazaro, J.B., Bailey, P.J., Lassar, A.B., 2002. Cyclin D-cdk4 activity modulates the subnuclear localization and interaction of MEF2 with SRC-family coactivators during skeletal muscle differentiation. Genes. Dev. 16, 1792–1805.

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

Lee, Y., Nadal-Ginard, B., Mahdavi, V., Izumo, S., 1997. Myocyte-specific enhancer factor 2 and thyroid hormone receptor associate and synergistically activate the alpha-cardiac myosin heavy-chain gene. Mol. Cell Biol. 17, 2745–2755. Leifer, D., Golden, J., Kowall, N.W., 1994. Myocyte-specific enhancer binding factor 2C expression in human brain development. Neuroscience 63, 1067–1079. Lemercier, C., Verdel, A., Galloo, B., Curtet, S., Brocard, M.P., Khochbin, S., 2000. MHDA1/HDAC5 histone deacetylase interacts with and represses MEF2A transcriptional activity. J. Biol. Chem. 275, 15594–15599. Leo, C., Chen, J.D., 2000. The SRC family of nuclear receptor coactivators. Gene 245, 1–11. L’Honore, A., Rana, V., Arsic, N., Franckhauser, C., Lamb, N.J., Fernandez, A., 2007. Identification of a new hybrid serum response factor and myocyte enhancer factor 2-binding element in MyoD enhancer required for MyoD expression during myogenesis. Mol. Biol. Cell 18, 1992–2001. Lilly, B., Galewsky, S., Firulli, A.B., Schulz, R.A., Olson, E.N., 1994. D-MEF2: a MADS box transcription factor expressed in differentiating mesoderm and muscle cell lineages during Drosophila embryogenesis. Proc. Natl. Acad. Sci. USA 91, 5662–5666. Lilly, B., Zhao, B., Ranganayakulu, G., Paterson, B.M., Schulz, R.A., Olson, E.N., 1995. Requirement of MADS domain transcription factor D-MEF2 for muscle formation in Drosophila. Science 267, 688–693. Lin, M.H., Nguyen, H.T., Dybala, C., Storti, R.V., 1996a. Myocyte-specific enhancer factor 2 acts cooperatively with a muscle activator region to regulate Drosophila tropomyosin gene muscle expression. Proc. Natl. Acad. Sci. USA 93, 4623–4628. Lin, Q., Lu, J., Yanagisawa, H., Webb, R., Lyons, G.E., Richardson, J.A., Olson, E.N., 1998. Requirement of the MADS-box transcription factor MEF2C for vascular development. Development 125, 4565–4574. Lin, Q., Schwarz, J., Bucana, C., Olson, E.N., 1997. Control of mouse cardiac morphogenesis and myogenesis by transcription factor MEF2C. Science 276, 1404–1407. Lin, X., Shah, S., Bulleit, R.F., 1996b. The expression of MEF2 genes is implicated in CNS neuronal differentiation. Brain Res. Mol. Brain Res. 42, 307–316. Ling, Y., West, A.G., Roberts, E.C., Lakey, J.H., Sharrocks, A.D., 1998. Interaction of transcription factors with serum response factor. Identification of the Elk-1 binding surface. J. Biol. Chem. 273, 10506–10514. Little, G.H., Bai, Y., Williams, T., Poizat, C., 2007. Nuclear calcium/ calmodulin-dependent protein kinase IIdelta preferentially transmits signals to histone deacetylase 4 in cardiac cells. J. Biol. Chem. 282, 7219–7231. Liu, D., Kang, J.S., Derynck, R., 2004. TGF-beta-activated Smad3 represses MEF2-dependent transcription in myogenic differentiation. Embo. J. 23, 1557–1566. Liu, N., Olson, E.N., 2006. Coactivator control of cardiovascular growth and remodeling. Curr. Opin. Cell Biol. 18, 715–722. Liu, W., Youn, H.D., Liu, J.O., 2001a. Thapsigargin-induced apoptosis involves Cabin1-MEF2-mediated induction of Nur77. Eur. J. Immunol. 31, 1757–1764. Liu, Z.P., Nakagawa, O., Nakagawa, M., Yanagisawa, H., Passier, R., Richardson, J.A., Srivastava, D., Olson, E.N., 2001b. CHAMP, a novel cardiac-specific helicase regulated by MEF2C. Dev. Biol. 234, 497–509.

695

Lo, P.C., Frasch, M., 2001. A role for the COUP-TF-related gene sevenup in the diversification of cardioblast identities in the dorsal vessel of Drosophila. Mech. Dev. 104, 49–60. Lo, P.C., Frasch, M., 2003. Establishing A–P polarity in the embryonic heart tube: a conserved function of Hox genes in Drosophila and vertebrates? Trends Cardiovasc. Med. 13, 182–187. Lovato, T.L., Nguyen, T.P., Molina, M.R., Cripps, R.M., 2002. The Hox gene abdominal-A specifies heart cell fate in the Drosophila dorsal vessel. Development 129, 5019–5027. Lu, J., McKinsey, T.A., Nicol, R.L., Olson, E.N., 2000a. Signal-dependent activation of the MEF2 transcription factor by dissociation from histone deacetylases. Proc. Natl. Acad. Sci. USA 97, 4070–4075. Lu, J., McKinsey, T.A., Zhang, C.L., Olson, E.N., 2000b. Regulation of skeletal myogenesis by association of the MEF2 transcription factor with class II histone deacetylases. Mol. Cell 6, 233–244. Lyons, G.E., Micales, B.K., Schwarz, J., Martin, J.F., Olson, E.N., 1995. Expression of mef2 genes in the mouse central nervous system suggests a role in neuronal maturation. J. Neurosci. 15, 5727–5738. Ma, K., Chan, J.K., Zhu, G., Wu, Z., 2005. Myocyte enhancer factor 2 acetylation by p300 enhances its DNA binding activity, transcriptional activity, and myogenic differentiation. Mol. Cell Biol. 25, 3575–3582. Maeda, T., Gupta, M.P., Stewart, A.F., 2002. TEF-1 and MEF2 transcription factors interact to regulate muscle-specific promoters. Biochem. Biophys. Res. Commun. 294, 791–797. Mao, Z., Nadal-Ginard, B., 1996. Functional and physical interactions between mammalian achaete-scute homolog 1 and myocyte enhancer factor 2A. J. Biol. Chem. 271, 14371–14375. Mar, J.H., Ordahl, C.P., 1990. M-CAT binding factor, a novel trans-acting factor governing muscle-specific transcription. Mol. Cell Biol. 10, 4271–4283. Marin, M.C., Rodriguez, J.R., Ferrus, A., 2004. Transcription of Drosophila troponin I gene is regulated by two conserved, functionally identical, synergistic elements. Mol. Biol. Cell. 15, 1185–1196. Marinissen, M.J., Chiariello, M., Pallante, M., Gutkind, J.S., 1999. A network of mitogen-activated protein kinases links G proteincoupled receptors to the c-jun promoter: a role for c-Jun NH2terminal kinase, p38s, and extracellular signal-regulated kinase 5. Mol. Cell Biol. 19, 4289–4301. Martin, J.F., Schwarz, J.J., Olson, E.N., 1993. Myocyte enhancer factor (MEF) 2C: a tissue-restricted member of the MEF-2 family of transcription factors. Proc. Natl. Acad. Sci. USA 90, 5282–5286. Martin, J.F., Miano, J.M., Hustad, C.M., Copeland, N.G., Jenkins, N.A., Olson, E.N., 1994. A Mef2 gene that generates a muscle-specific isoform via alternative mRNA splicing. Mol. Cell Biol. 14, 1647–1656. McDermott, J.C., Cardoso, M.C., Yu, Y.T., Andres, V., Leifer, D., Krainc, D., Lipton, S.A., Nadal-Ginard, B., 1993. HMEF2C gene encodes skeletal muscle- and brain-specific transcription factors. Mol. Cell Biol. 13, 2564–2577. McKinsey, T.A., Zhang, C.L., Lu, J., Olson, E.N., 2000a. Signal-dependent nuclear export of a histone deacetylase regulates muscle differentiation. Nature 408, 106–111. McKinsey, T.A., Zhang, C.L., Olson, E.N., 2000b. Activation of the myocyte enhancer factor-2 transcription factor by calcium/calmodulin-dependent protein kinase-stimulated binding of 14-3-3 to histone deacetylase 5. Proc. Natl. Acad. Sci. USA 97, 14400–14405. McKinsey, T.A., Zhang, C.L., Olson, E.N., 2001a. Control of muscle development by dueling HATs and HDACs. Curr. Opin. Genet. Dev. 11, 497–504.

696

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

McKinsey, T.A., Zhang, C.L., Olson, E.N., 2001b. Identification of a signal-responsive nuclear export sequence in class II histone deacetylases. Mol. Cell Biol. 21, 6312–6321. McKinsey, T.A., Zhang, C.L., Olson, E.N., 2002. MEF2: a calciumdependent regulator of cell division, differentiation and death. Trends Biochem. Sci. 27, 40–47. Miller, C.T., Swartz, M.E., Khuu, P.A., Walker, M.B., Eberhart, J.K., Kimmel, C.B., 2007. Mef2ca is required in cranial neural crest to effect Endothelin1 signaling in zebrafish. Dev. Biol. 308, 144–157. Molkentin, J.D., Markham, B.E., 1993. Myocyte-specific enhancer-binding factor (MEF-2) regulates alpha-cardiac myosin heavy chain gene expression in vitro and in vivo. J. Biol. Chem. 268, 19512–19520. Molkentin, J.D., Black, B.L., Martin, J.F., Olson, E.N., 1995. Cooperative activation of muscle gene expression by MEF2 and myogenic bHLH proteins. Cell 83, 1125–1136. Molkentin, J.D., Olson, E.N., 1996. Combinatorial control of muscle development by basic helix-loop-helix and MADS-box transcription factors. Proc. Natl. Acad. Sci. USA 93, 9366–9373. Molkentin, J.D., Black, B.L., Martin, J.F., Olson, E.N., 1996a. Mutational analysis of the DNA binding, dimerization, and transcriptional activation domains of MEF2C. Mol. Cell Biol. 16, 2627–2636. Molkentin, J.D., Firulli, A.B., Black, B.L., Martin, J.F., Hustad, C.M., Copeland, N., Jenkins, N., Lyons, G., Olson, E.N., 1996b. MEF2B is a potent transactivator expressed in early myogenic lineages. Mol. Cell Biol. 16, 3814–3824. Molkentin, J.D., Li, L., Olson, E.N., 1996c. Phosphorylation of the MADS-Box transcription factor MEF2C enhances its DNA binding activity. J. Biol. Chem. 271, 17199–17204. Montgomery, R.L., Davis, C.A., Potthoff, M.J., Haberland, M., Fielitz, J., Qi, X., Hill, J.A., Richardson, J.A., Olson, E.N., 2007. Histone deacetylases 1 and 2 redundantly regulate cardiac morphogenesis, growth, and contractility. Genes Dev. 21, 1790–1802. Mori, A.D., Bruneau, B.G., 2004. TBX5 mutations and congenital heart disease: Holt-Oram syndrome revealed. Curr. Opin. Cardiol. 19, 211–215. Morin, S., Charron, F., Robitaille, L., Nemer, M., 2000. GATA-dependent recruitment of MEF2 proteins to target promoters. Embo. J. 19, 2046–2055. Morin, S., Pozzulo, G., Robitaille, L., Cross, J., Nemer, M., 2005. MEF2dependent recruitment of the HAND1 transcription factor results in synergistic activation of target promoters. J. Biol. Chem. 280, 32272–32278. Mueller, P.R., Wold, B., 1989. In vivo footprinting of a muscle specific enhancer by ligation mediated PCR. Science 246, 780–786. Mushegian, A.R., Koonin, E.V., 1996. Sequence analysis of eukaryotic developmental proteins: ancient and novel domains. Genetics 144, 817–828. Nadruz Jr., W., Kobarg, C.B., Constancio, S.S., Corat, P.D., Franchini, K.G., 2003. Load-induced transcriptional activation of c-jun in rat myocardium: regulation by myocyte enhancer factor 2. Circ. Res. 92, 243–251. Nakagawa, O., Arnold, M., Nakagawa, M., Hamada, H., Shelton, J.M., Kusano, H., Harris, T.M., Childs, G., Campbell, K.P., Richardson, J.A., Nishino, I., Olson, E.N., 2005. Centronuclear myopathy in mice lacking a novel muscle-specific protein kinase transcriptionally regulated by MEF2. Genes Dev. 19, 2066–2077. Navankasattusas, S., Zhu, H., Garcia, A.V., Evans, S.M., Chien, K.R., 1992. A ubiquitous factor (HF-1a) and a distinct muscle factor (HF1b/MEF-2) form an E-box-independent pathway for cardiac muscle gene expression. Mol. Cell Biol. 12, 1469–1479.

Naya, F.J., Black, B.L., Wu, H., Bassel-Duby, R., Richardson, J.A., Hill, J.A., Olson, E.N., 2002. Mitochondrial deficiency and cardiac sudden death in mice lacking the MEF2A transcription factor. Nat. Med. 8, 1303–1309. Nguyen, H.T., Bodmer, R., Abmayr, S.M., McDermott, J.C., Spoerel, N.A., 1994. D-mef2: a Drosophila mesoderm-specific MADS box-containing gene with a biphasic expression profile during embryogenesis. Proc. Natl. Acad. Sci. USA 91, 7520–7524. Nguyen, H.T., Xu, X., 1998. Drosophila mef2 expression during mesoderm development is controlled by a complex array of cis-acting regulatory modules. Dev. Biol. 204, 550–566. Nguyen, T., Wang, J., Schulz, R.A., 2002. Mutations within the conserved MADS box of the D-MEF2 muscle differentiation factor result in a loss of DNA binding ability and lethality in Drosophila. Differentiation 70, 438–446. Olson, E.N., Perry, M., Schulz, R.A., 1995. Regulation of muscle differentiation by the MEF2 family of MADS box transcription factors. Dev. Biol. 172, 2–14. Olson, E.N., Backs, J., McKinsey, T.A., 2006. Control of cardiac hypertrophy and heart failure by histone acetylation/deacetylation. Novartis Found Symp. 274, 3–12; discussion 13–19, 152–155, 272–276. Ornatsky, O.I., Andreucci, J.J., McDermott, J.C., 1997. A dominantnegative form of transcription factor MEF2 inhibits myogenesis. J. Biol. Chem. 272, 33271–33278. Ornatsky, O.I., Cox, D.M., Tangirala, P., Andreucci, J.J., Quinn, Z.A., Wrana, J.L., Prywes, R., Yu, Y.T., McDermott, J.C., 1999. Posttranslational control of the MEF2A transcriptional regulatory protein. Nucleic Acids Res. 27, 2646–2654. Pabst, S., Wollnik, B., Rohmann, E., Hintz, Y., Glanzer, K., Vetter, H., Nickenig, G., Grohe, C., 2008. A novel stop mutation truncating critical regions of the cardiac transcription factor NKX2-5 in a large family with autosomal-dominant inherited congenital heart disease. Clin. Res. Cardiol. 97, 39–42. Park, E.J., Ogden, L.A., Talbot, A., Evans, S., Cai, C.L., Black, B.L., Frank, D.U., Moon, A.M., 2006. Required, tissue-specific roles for Fgf8 in outflow tract formation and remodeling. Development 133, 2419–2433. Parmacek, M.S., 2001. Transcriptional programs regulating vascular smooth muscle cell development and differentiation. Curr. Top Dev. Biol. 51, 69–89. Pehlivan, T., Pober, B.R., Brueckner, M., Garrett, S., Slaugh, R., Van Rheeden, R., Wilson, D.B., Watson, M.S., Hing, A.V., 1999. GATA4 haploinsufficiency in patients with interstitial deletion of chromosome region 8p23.1 and congenital heart disease. Am. J. Med. Genet. 83, 201–206. Pollock, R., Treisman, R., 1991. Human SRF-related proteins: DNAbinding properties and potential regulatory targets. Genes Dev. 5, 2327–2341. Quinn, Z.A., Yang, C.C., Wrana, J.L., McDermott, J.C., 2001. Smad proteins function as co-modulators for MEF2 transcriptional regulatory proteins. Nucleic Acids Res. 29, 732–742. Ranganayakulu, G., Zhao, B., Dokidis, A., Molkentin, J.D., Olson, E.N., Schulz, R.A., 1995. A series of mutations in the D-MEF2 transcription factor reveal multiple functions in larval and adult myogenesis in Drosophila. Dev. Biol. 171, 169–181. Rao, S., Karray, S., Gackstetter, E.R., Koshland, M.E., 1998. Myocyte enhancer factor-related B-MEF2 is developmentally expressed in B cells and regulates the immunoglobulin J chain promoter. J. Biol. Chem. 273, 26123–26129.

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

Ren, X., Li, Y., Ma, X., Zheng, L., Xu, Y., Wang, J., 2007. Activation of p38/MEF2C pathway by all-trans retinoic acid in cardiac myoblasts. Life Sci. 81, 89–96. Riquelme, C., Barthel, K.K., Liu, X., 2006. SUMO-1 modification of MEF2A regulates its transcriptional activity. J. Cell Mol. Med. 10, 132–144. Sandmann, T., Jensen, L.J., Jakobsen, J.S., Karzynski, M.M., Eichenlaub, M.P., Bork, P., Furlong, E.E., 2006. A temporal map of transcription factor activity: mef2 directly regulates target genes at all stages of muscle development. Dev. Cell 10, 797–807. Sandmann, T., Girardot, C., Brehme, M., Tongprasit, W., Stolc, V., Furlong, E.E., 2007. A core transcriptional network for early mesoderm development in Drosophila melanogaster. Genes Dev. 21, 436–449. Santelli, E., Richmond, T.J., 2000. Crystal structure of MEF2A core bound to DNA at 1.5 A resolution. J. Mol. Biol. 297, 437–449. Sartorelli, V., Huang, J., Hamamori, Y., Kedes, L., 1997. Molecular mechanisms of myogenic coactivation by p300: direct interaction with the activation domain of MyoD and with the MADS box of MEF2C. Mol. Cell Biol. 17, 1010–1026. Sartorelli, V., Puri, P.L., Hamamori, Y., Ogryzko, V., Chung, G., Nakatani, Y., Wang, J.Y., Kedes, L., 1999. Acetylation of MyoD directed by PCAF is necessary for the execution of the muscle program. Mol. Cell 4, 725–734. Satyaraj, E., Storb, U., 1998. Mef2 proteins, required for muscle differentiation, bind an essential site in the Ig lambda enhancer. J. Immunol. 161, 4795–4802. Schluterman, M.K., Krysiak, A.E., Kathiriya, I.S., Abate, N., Chandalia, M., Srivastava, D., Garg, V., 2007. Screening and biochemical analysis of GATA4 sequence variations identified in patients with congenital heart disease. Am. J. Med. Genet. A 143, 817–823. Schott, J.J., Benson, D.W., Basson, C.T., Pease, W., Silberbach, G.M., Moak, J.P., Maron, B.J., Seidman, C.E., Seidman, J.G., 1998. Congenital heart disease caused by mutations in the transcription factor NKX2-5. Science 281, 108–111. Schulz, R.A., Chromey, C., Lu, M.F., Zhao, B., Olson, E.N., 1996. Expression of the D-MEF2 transcription in the Drosophila brain suggests a role in neuronal cell differentiation. Oncogene 12, 1827–1831. Shalizi, A., Gaudilliere, B., Yuan, Z., Stegmuller, J., Shirogane, T., Ge, Q., Tan, Y., Schulman, B., Harper, J.W., Bonni, A., 2006. A calciumregulated MEF2 sumoylation switch controls postsynaptic differentiation. Science 311, 1012–1017. Shen, H., McElhinny, A.S., Cao, Y., Gao, P., Liu, J., Bronson, R., Griffin, J.D., Wu, L., 2006. The Notch coactivator, MAML1, functions as a novel coactivator for MEF2C-mediated transcription and is required for normal myogenesis. Genes Dev. 20, 675–688. Shore, P., Sharrocks, A.D., 1995. The MADS-box family of transcription factors. Eur. J. Biochem. 229, 1–13. Skerjanc, I.S., Petropoulos, H., Ridgeway, A.G., Wilton, S., 1998. Myocyte enhancer factor 2C and Nkx2-5 up-regulate each other’s expression and initiate cardiomyogenesis in P19 cells. J. Biol. Chem. 273, 34904–34910. Smith, P.D., Mount, M.P., Shree, R., Callaghan, S., Slack, R.S., Anisman, H., Vincent, I., Wang, X., Mao, Z., Park, D.S., 2006. Calpain-regulated p35/cdk5 plays a central role in dopaminergic neuron death through modulation of the transcription factor myocyte enhancer factor 2. J. Neurosci. 26, 440–447. Srivastava, D., Cserjesi, P., Olson, E.N., 1995. A subclass of bHLH proteins required for cardiac morphogenesis. Science 270, 1995–1999.

697

Stankovic-Valentin, N., Deltour, S., Seeler, J., Pinte, S., Vergoten, G., Guerardel, C., Dejean, A., Leprince, D., 2007. An acetylation/ deacetylation-SUMOylation switch through a phylogenetically conserved psiKXEP motif in the tumor suppressor HIC1 regulates transcriptional repression activity. Mol. Cell Biol. 27, 2661–2675. Tang, X., Wang, X., Gong, X., Tong, M., Park, D., Xia, Z., Mao, Z., 2005. Cyclin-dependent kinase 5 mediates neurotoxin-induced degradation of the transcription factor myocyte enhancer factor 2. J. Neurosci. 25, 4823–4834. Taylor, M.V., 1995. Muscle development. Making Drosophila muscle. Curr. Biol. 5, 740–742. Taylor, M.V., Beatty, K.E., Hunter, H.K., Baylies, M.K., 1995. Drosophila MEF2 is regulated by twist and is expressed in both the primordia and differentiated cells of the embryonic somatic, visceral and heart musculature. Mech. Dev. 50, 29–41. Theissen, G., Kim, J.T., Saedler, H., 1996. Classification and phylogeny of the MADS-box multigene family suggest defined roles of MADSbox gene subfamilies in the morphological evolution of eukaryotes. J. Mol. Evol. 43, 484–516. Ticho, B.S., Stainier, D.Y., Fishman, M.C., Breitbart, R.E., 1996. Three zebrafish MEF2 genes delineate somitic and cardiac muscle development in wild-type and mutant embryos. Mech. Dev. 59, 205–218. Toro, R., Saadi, I., Kuburas, A., Nemer, M., Russo, A.F., 2004. Cellspecific activation of the atrial natriuretic factor promoter by PITX2 and MEF2A. J. Biol. Chem. 279, 52087–52094. Tran, C.M., Sucov, H.M., 1998. The RXRalpha gene functions in a non-cell-autonomous manner during mouse cardiac morphogenesis. Development 125, 1951–1956. Treisman, R., 1990. The SRE: a growth factor responsive transcriptional regulator. Semin. Cancer Biol. 1, 47–58. Trivedi, C.M., Luo, Y., Yin, Z., Zhang, M., Zhu, W., Wang, T., Floss, T., Goettlicher, M., Noppinger, P.R., Wurst, W., Ferrari, V.A., Abrams, C.S., Gruber, P.J., Epstein, J.A., 2007. Hdac2 regulates the cardiac hypertrophic response by modulating Gsk3 beta activity. Nat. Med. 13, 324–331. van Rooij, E., Sutherland, L.B., Liu, N., Williams, A.H., McAnally, J., Gerard, R.D., Richardson, J.A., Olson, E.N., 2006. A signature pattern of stress-responsive microRNAs that can evoke cardiac hypertrophy and heart failure. Proc. Natl. Acad. Sci. USA 103, 18255– 18260. van Rooij, E., Olson, E.N., 2007. MicroRNAs put their signatures on the heart. Physiol. Genomics 31, 365–366. van Rooij, E., Sutherland, L.B., Qi, X., Richardson, J.A., Hill, J., Olson, E.N., 2007. Control of stress-dependent cardiac growth and gene expression by a microRNA. Science 316, 575–579. Vanpoucke, G., Goossens, S., De Craene, B., Gilbert, B., van Roy, F., Berx, G., 2004. GATA-4 and MEF2C transcription factors control the tissue-specific expression of the alphaT-catenin gene CTNNA3. Nucleic Acids Res. 32, 4155–4165. Verzi, M.P., McCulley, D.J., De Val, S., Dodou, E., Black, B.L., 2005. The right ventricle, outflow tract, and ventricular septum comprise a restricted expression domain within the secondary/anterior heart field. Dev. Biol. 287, 134–145. Verzi, M.P., Agarwal, P., Brown, C., McCulley, D.J., Schwarz, J.J., Black, B.L., 2007. The transcription factor MEF2C is required for craniofacial development. Dev. Cell 12, 645–652. von Both, I., Silvestri, C., Erdemir, T., Lickert, H., Walls, J.R., Henkelman, R.M., Rossant, J., Harvey, R.P., Attisano, L., Wrana, J.L., 2004. Foxh1 is essential for development of the anterior heart field. Dev. Cell 7, 331–345.

698

PART  |  9  Transcriptional Circuits in Cardiac Development and Disease

Vong, L., Bi, W., O’Connor-Halligan, K.E., Li, C., Cserjesi, P., Schwarz, J.J., 2006. MEF2C is required for the normal allocation of cells between the ventricular and sinoatrial precursors of the primary heart field. Dev. Dyn. 235, 1809–1821. Vong, L.H., Ragusa, M.J., Schwarz, J.J., 2005. Generation of conditional Mef2cloxP/loxP mice for temporal- and tissue-specific analyses. Genesis 43, 43–48. Wallin, J.J., Rinkenberger, J.L., Rao, S., Gackstetter, E.R., Koshland, M.E., Zwollo, P., 1999. B cell-specific activator protein prevents two activator factors from binding to the immunoglobulin J chain promoter until the antigen-driven stages of B cell development. J. Biol. Chem. 274, 15959–15965. Wang, A.H., Bertos, N.R., Vezmar, M., Pelletier, N., Crosato, M., Heng, H.H., Th’ng, J., Han, J., Yang, X.J., 1999. HDAC4, a human histone deacetylase related to yeast HDA1, is a transcriptional corepressor. Mol. Cell Biol. 19, 7816–7827. Wang, A.H., Yang, X.J., 2001. Histone deacetylase 4 possesses intrinsic nuclear import and export signals. Mol. Cell Biol. 21, 5992–6005. Wang, D.Z., Valdez, M.R., McAnally, J., Richardson, J., Olson, E.N., 2001. The Mef2c gene is a direct transcriptional target of myogenic bHLH and MEF2 proteins during skeletal muscle development. Development 128, 4623–4633. Wang, L., Fan, C., Topol, S.E., Topol, E.J., Wang, Q., 2003a. Mutation of MEF2A in an inherited disorder with features of coronary artery disease. Science 302, 1578–1581. Wang, Q., Rao, S., Topol, E.J., 2005. Miscues on the “lack of MEF2A mutations” in coronary artery disease. J. Clin. Invest. 115, 1399– 1400 author reply 1400–1401. Wang, Z., Wang, D.Z., Pipes, G.C., Olson, E.N., 2003b. Myocardin is a master regulator of smooth muscle gene expression. Proc. Natl. Acad. Sci. USA 100, 7129–7134. Ward, E.J., Skeath, J.B., 2000. Characterization of a novel subset of cardiac cells and their progenitors in the Drosophila embryo. Development 127, 4959–4969. Weng, L., Kavaslar, N., Ustaszewska, A., Doelle, H., Schackwitz, W., Hebert, S., Cohen, J.C., McPherson, R., Pennacchio, L.A., 2005. Lack of MEF2A mutations in coronary artery disease. J. Clin. Invest. 115, 1016–1020. West, A.G., Shore, P., Sharrocks, A.D., 1997. DNA binding by MADSbox transcription factors: a molecular mechanism for differential DNA bending. Mol. Cell Biol. 17, 2876–2887. Wilker, P.R., Kohyama, M., Sandau, M.M., Albring, J.C., Nakagawa, O., Schwarz, J.J., Murphy, K.M., 2008. Transcription factor Mef2c is required for B cell proliferation and survival after antigen receptor stimulation. Nat. Immunol. 9, 603–612. Wilson-Rawls, J., Molkentin, J.D., Black, B.L., Olson, E.N., 1999. Activated notch inhibits myogenic activity of the MADS-Box transcription factor myocyte enhancer factor 2C. Mol. Cell Biol. 19, 2853–2862. Winter, B., Arnold, H.H., 2000. Activated raf kinase inhibits muscle cell differentiation through a MEF2-dependent mechanism. J. Cell Sci. 113 (Pt 23), 4211–4220. Xin, M., Davis, C.A., Molkentin, J.D., Lien, C.L., Duncan, S.A., Richardson, J.A., Olson, E.N., 2006. A threshold of GATA4 and GATA6 expression is required for cardiovascular development. Proc. Natl. Acad. Sci. USA 103, 11189–11194. Xu, J., Li, Q., 2003. Review of the in vivo functions of the p160 steroid receptor coactivator family. Mol. Endocrinol. 17, 1681–1692.

Xu, J., Gong, N.L., Bodi, I., Aronow, B.J., Backx, P.H., Molkentin, J.D., 2006. Myocyte enhancer factors 2A and 2C induce dilated cardiomyopathy in transgenic mice. J. Biol. Chem. 281, 9152–9162. Yan, C., Takahashi, M., Okuda, M., Lee, J.D., Berk, B.C., 1999. Fluid shear stress stimulates big mitogen-activated protein kinase 1 (BMK1) activity in endothelial cells. Dependence on tyrosine kinases and intracellular calcium. J. Biol. Chem. 274, 143–150. Yan, C., Luo, H., Lee, J.D., Abe, J., Berk, B.C., 2001. Molecular cloning of mouse ERK5/BMK1 splice variants and characterization of ERK5 functional domains. J. Biol. Chem. 276, 10870–10878. Yang, C.C., Ornatsky, O.I., McDermott, J.C., Cruz, T.F., Prody, C.A., 1998. Interaction of myocyte enhancer factor 2 (MEF2) with a mitogen-activated protein kinase, ERK5/BMK1. Nucleic. Acids Res. 26, 4771–4777. Yang, Y., Jack, T., 2004. Defining subdomains of the K domain important for protein–protein interactions of plant MADS proteins. Plant Mol. Biol. 55, 45–59. Yoshida, T., Sinha, S., Dandre, F., Wamhoff, B.R., Hoofnagle, M.H., Kremer, B.E., Wang, D.Z., Olson, E.N., Owens, G.K., 2003. Myocardin is a key regulator of CArG-dependent transcription of multiple smooth muscle marker genes. Circ. Res. 92, 856–864. Youn, H.D., Sun, L., Prywes, R., Liu, J.O., 1999. Apoptosis of T cells mediated by Ca2-induced release of the transcription factor MEF2. Science 286, 790–793. Youn, H.D., Liu, J.O., 2000. Cabin1 represses MEF2-dependent Nur77 expression and T cell apoptosis by controlling association of histone deacetylases and acetylases with MEF2. Immunity 13, 85–94. Youn, H.D., Grozinger, C.M., Liu, J.O., 2000. Calcium regulates transcriptional repression of myocyte enhancer factor 2 by histone deacetylase 4. J. Biol. Chem. 275, 22563–22567. Yu, Y.T., 1996. Distinct domains of myocyte enhancer binding factor-2A determining nuclear localization and cell type-specific transcriptional activity. J. Biol. Chem. 271, 24675–24683. Yu, Y.T., Breitbart, R.E., Smoot, L.B., Lee, Y., Mahdavi, V., Nadal-Ginard, B., 1992. Human myocyte-specific enhancer factor 2 comprises a group of tissue-restricted MADS box transcription factors. Genes Dev. 6, 1783–1798. Zang, M.X., Li, Y., Xue, L.X., Jia, H.T., Jing, H., 2004. Cooperative activation of atrial naturetic peptide promoter by dHAND and MEF2C. J. Cell Biochem. 93, 1255–1266. Zeisberg, E.M., Ma, Q., Juraszek, A.L., Moses, K., Schwartz, R.J., Izumo, S., Pu, W.T., 2005. Morphogenesis of the right ventricle requires myocardial expression of Gata4. J. Clin. Invest. 115, 1522–1531. Zhang, C.L., McKinsey, T.A., Lu, J.R., Olson, E.N., 2001a. Association of COOH-terminal-binding protein (CtBP) and MEF2-interacting transcription repressor (MITR) contributes to transcriptional repression of the MEF2 transcription factor. J. Biol. Chem. 276, 35–39. Zhang, C.L., McKinsey, T.A., Olson, E.N., 2001b. The transcriptional corepressor MITR is a signal-responsive inhibitor of myogenesis. Proc. Natl. Acad. Sci. USA 98, 7354–7359. Zhang, C.L., McKinsey, T.A., Chang, S., Antos, C.L., Hill, J.A., Olson, E.N., 2002. Class II histone deacetylases act as signal-responsive repressors of cardiac hypertrophy. Cell 110, 479–488. Zhao, M., New, L., Kravchenko, V.V., Kato, Y., Gram, H., di Padova, F., Olson, E.N., Ulevitch, R.J., Han, J., 1999. Regulation of the MEF2 family of transcription factors by p38. Mol. Cell Biol. 19, 21–30.

Chapter  |  9.5  Myocyte Enhancer Factor 2 Transcription Factors in Heart Development and Disease

Zhao, X., Sternsdorf, T., Bolger, T.A., Evans, R.M., Yao, T.P., 2005a. Regulation of MEF2 by histone deacetylase 4- and SIRT1 deacetylase-mediated lysine modifications. Mol. Cell Biol. 25, 8456–8464. Zhao, Y., Samal, E., Srivastava, D., 2005b. Serum response factor regulates a muscle-specific microRNA that targets Hand2 during cardiogenesis. Nature 436, 214–220.

699

Zhu, B., Gulick, T., 2004. Phosphorylation and alternative pre-mRNA splicing converge to regulate myocyte enhancer factor 2C activity. Mol. Cell Biol. 24, 8264–8275. Zhu, B., Ramachandran, B., Gulick, T., 2005. Alternative pre-mRNA splicing governs expression of a conserved acidic transactivation domain in myocyte enhancer factor 2 factors of striated muscle and brain. J. Biol. Chem. 280, 28749–28760.

Chapter 10.1

Chromatin Modification and Remodeling in Heart Development Benoit G. Bruneau Gladstone Institute of Cardiovascular Disease and Department of Pediatrics, University of California, San Francisco, CA, USA

I.  Introduction A complex network of transcription factors regulates cardiac cell fate and morphogenesis. Chromatin modifications and remodeling are also key regulatory elements of gene expression, and these processes are relevant to heart development. This takes place largely via the interaction of cardiac transcription factors, with proteins that chemicallymodify or structurally-remodel chromatin. Some of these components are themselves cardiac-restricted, adding another dimension to cardiac-specific regulation of gene expression.

II.  Chromatin modification and remodeling: general concepts and key players Transcription factors are important regulators of morphogenesis and lineage decisions. In the heart, for example, DNA-binding transcription factors such as Tbx5 and Nkx2-5 bind to target genes to activate or repress them (Fig. 1), thereby initiating lineage decisions and setting morphogenetic events in motion (Bruneau, 2002; Srivastava, 2006) (see also Chapters 9.1 and 9.4). However, they must function in the context of DNA that is packed into the dense structure that is chromatin (Fig. 2). This poses a daunting topological problem for gene activation, as dense chromatin must be “loosened” to allow gene activation. Chromatin structure is regulated by two key modulators: chemical modification of histones – the structural components of nucleosomes – and the actions of ATP-dependent Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

chromatin-remodeling complexes. DNA–­histone interactions provide a high degree of stability to nucleosomes, while allowing them to facilitate significant structural changes without becoming unstable. The dynamic interplay between stability and plasticity is the most critical quality of histones, because it allows them to fulfill both structural and regulatory roles. The actions at both regulatory levels result in various degrees of “openness”. Genes are activated through two primary mechanisms: (1) chromatin-remodeling and increased recruitment of transcriptional activators; and (2) derepression, which involves the removal of repressive chromatin marks or repressor complexes. Both mechanisms have a high degree of regulatory potential, and the reverse action, active repression, is probably also important in all cell types, including those of the developing heart.

II.A.  Histone-Modifying Proteins Several modifications of histones affect the structure of chromatin, rendering it more or less active for transcription or other processes. These modifications, whose output is referred to as the histone code (Jenuwein and Allis, 2001), include acetylation and mono-, di- and trimethylation of lysine, phosphorylation of serine, mono- and dimethylation of arginine, and ubiquitination of various amino acid residues on histones H2A, H2B, H3 and H4 (Khorasanizadeh, 2004) (Fig. 3). Several of these modifications, such as acetylation of lysine 9 in histone H3 (H3K9Ac), are considered to be marks of “active” chromatin, while others, such as trimethylation of lysine 27 in histone H3 (H3K27Me3),

703

704

PART  |  10  Epigenetic Modifiers of Cardiac Development

Figure 1  Simplistic view of transcriptional activation by cardiac transcription factors. In this example, Tbx5 binds to its cognate DNA-binding sites adjacent to a target gene, resulting in the recruitment of the transcriptional apparatus, which activates target genes such as ANF or Fgf10.

30nm fiber

Linker histones

Long range fiber-fiber interactions

TF

Chromatin modifications

Nucleus

Nucleosome

DNA

Core histone tall domain

Figure 2  Structural aspects of DNA packaging in chromatin. Transcription factors must access DNA that is tightly-packaged in nucleosomes that are densely packed as chromatin. Histone modifications and chromatin-remodeling factors alter this structure to allow activation or maintain repression.

705

Chapter  |  10.1  Chromatin Modification and Remodeling in Heart Development

(A)

- KKTESHHKAKGK-COOH

H3

H2A

NH2-ARTKQTARKSTGGKAPRKQLATKAARKSAPATGGVKKPHR-

K79 in H3 core

H4 Ac-SGRGKGGKGLGKGGAKRHRKVLRD-

H2B NH2-PEPSKSAPAPKKGSKKAITKAQKKDGKKRKRSRKH2B

SK-COOH

H2B

NH2-SGRGKQGGKARAKAK-

(B) H3C O

H

H

N CH3 H RHN

COR

H

RHN

+

COR

RHN

H COR

ε-,N-monomethyl ε-N,N-dimethyl lysine lysine

HN

+

H3C N CH3 CH3

N CH3 CH3 H

H

ε-N-Acetyl lysine

+

N H CH3

RHN

COR

ε-N,N,N-trimethyl lysine

RHN

O COR

O-Phospho serine

H3C 2

H3C N

HN

PO4 H

+ NH

H RHN

+ NH

COR

H H3C N

HN H

ω-N-methyl arginine

2

RHN

CH3

+ NH

Ubiquition

2

HN H

COR

ω-N,N-dimethyl arginine

RHN

COR

ω-N,N-dimethyl arginine, symmetric

Figure 3  Several histone modifications. (A) Histone modification shown with a nucleosome structure. (B) Legend for the different modifications shown. Adapted from Khorasanizadeh (2004).

are thought to correspond to a “repressive” code. The interpretation of histone ­ modifications is complex, as some trimethyl histone marks act as repressive marks (e.g., H3K27Me3 and H3K9Me3), while others function as ­activator marks (e.g., H3K4Me2 and H3K36Me3); in some cases, this is context-dependent, as H3K9Me3 has also been associated with transcriptionally-active loci, but only when located in coding regions. Thus, complex ­ histone modifications are critical for fine-tuning the chromatin ­status of genes. These chemical modifications are carried out by a wide variety of proteins, including histone acetyltransferases (HATs), histone deacetylases (HDACs; see Chapter 10.2, Volume II), histone methyltransferases and histone demethylases. Histone methyltrans­ferases, such as Ezh2, are often part of complexes, such as the Polycomb (Pc) repressor complexes PRC2, one of two Pc complexes. Other complexes, such as PRC1, contain proteins that recognize the specific methyl marks laid down by histone-modifying

proteins (Sparmann and van Lohuizen, 2006). These proteins include those with PHD domains, such as those found in the ISWI NURF complex protein BPTF (bromodomain and PHD-­finger transcription factor), and WD40 domains such as those found in WDR5, which bind H3K4Me3 marks (Li et al., 2006; Pena et al., 2006; Wysocka et al., 2006; Ruthenburg et al., 2007). Several of the recently-identified histone demethylases are part of the large family of Jumonji (Jmj) domain proteins (Chen et al., 2006; Klose et al., 2006; Trojer and Reinberg, 2006; Whetstine et al., 2006; Shi and Whetstine, 2007). Since some demethylases remove methyl marks from both types of methylated histone residues, their roles in activation and/or repression are likely to be very complex. Indeed, this has been further complicated by the proposed role for Lid2, which was initially characterized as a histone demethylase, but which also appears to function to physically recruit various other specific demethylases to particular loci, depending on the chromatin context (Li et al., 2008).

706

II.B.  Chromatin-Remodeling Complexes Several ATP-dependent chromatin-remodeling complexes have been identified in eukaryotes. The prototype of these complexes is the yeast Swi/Snf complex (originally identified in yeast as “mating type switching and sucrose nonfermenting” mutants) and related complexes such as RSC and ISWI (Martens and Winston, 2003; de la Serna et al., 2006). In mammals, the Swi/Snf complex is represented by several related polymorphic complexes, referred to as Brg1/Brm-associated factor (BAF) complexes (Wang et al., 1996; Olave et al., 2002; Martens and Winston, 2003; de la Serna et al., 2006). Other complexes with ­ similar functions in mammalian cells are the ISWI-related CHRAC and NURD complexes. The mammalian BAF complex and other mammalian chromatin-remodeling complexes are illustrated in Fig. 4. BAF complexes are large (2 MDa) and contain one of two ATPases, and at least 11 variable subunits which are highly-conserved from yeast to humans. The two major BAF complex ATPases are Brahma-related gene 1 (Brg1) and Brm (derived from Brahma, the ATPase of the Drosophila Swi2/Snf2 complex) (Tamkun et al., 1992). BAF complexes are thought to be mostly important for transcriptional activation, which they facilitate by unwinding nucleosomes and allowing RNA polymerase II to access transcriptional initiation sites. However, they are also involved in repressive complexes by virtue of their association with repressor proteins such as mSin3a, HDACs and methylases (Martens and Winston, 2003; de la Serna et al., 2006).

Figure 4  Mammalian ATP-dependent chromatin-remodeling factors. Shown are the Swi/Snf-like BAF and PBAF complexes, the ISWI-like hCHRAC and the mammalian NURD complex. ATPases are red; other components of the complexes are in yellow.

PART  |  10  Epigenetic Modifiers of Cardiac Development

Studies of BAF complex subunits in mice have not been particularly informative about the roles of ATPdependent chromatin-remodeling complexes in organogenesis or cellular differentiation. The reason is that knockout mice of several BAF complex subunit genes (e.g., Brg1, Baf155, Snf5 and the BAF complex-related ISWI complex ATPase gene Snf2h) do not develop past the implantation stage (Bultman et al., 2000; Klochendler-Yeivin et al., 2000; Guidi et al., 2001; Kim et al., 2001; Stopka and Skoultchi, 2003). Conversely, the loss of Brm results in viable mice that have slight defects in cell proliferation (Reyes et al., 1998). In the case of the BAF complexes, it appears that during organogenesis Brg1 may be the primary effector of BAF complex function and that Brm is dispensable for embryonic development, but in some tissues such as the skin Brg1 and Brm function cooperatively (Indra et al., 2005). In the immune system, BAF complexes have a specific role in lineage switching (Chi et al., 2002, 2003; Gebuhr et al., 2003), indicating that they may participate in the specification of tissue-specific cell fate in other organs.

III.  Histone-modifying enzymes in heart development III.A.  Histone Acetyl Transferases The potential role of histone acetylation in cardiac development is highlighted largely by work done on the HATs p300 and CBP and their interactions with cardiac trans­ criptional regulators. Mice lacking p300 have multiple defects in embryogenesis, including cardiac defects such as reduced ventricular trabeculation and impaired expression of cardiac genes (Yao et al., 1998). The interaction of p300 with cardiac transcription factors, such as MEF2D, regulates the expression of the cardiac alpha-actin gene (Actc) through a distal enhancer (Molinari et al., 2004). A p300/CBP complex also interacts with Cited2, a TFAP2 co-activator (Bhattacharya et al., 1999) which is essential for cardiac development (Bamforth et al., 2001). Also, p300 is a transcriptional co-activator of the GATA family of transcription factors, which are critical regulators of cardiac development (Kakita et al., 1999; Dai and Markham, 2001). Finally, CBP promotes the transcription of target genes by the combined actions of Hand1 and Mef2-family transcription factors (Morin et al., 2005). Thus, p300-­mediated histone acetylation is likely to be a widespread mechanism for the co-activation of cardiac genes. However, its roles are not restricted to the developing heart; p300 can also directly modify GATA factors by acetylating specific amino acid residues (Kawamura et al., 2005). Whether this affects the interaction of GATA factors with chromatin remains to be seen, but similar modifications of the myogenic transcription factor MyoD by PCAF,

Chapter  |  10.1  Chromatin Modification and Remodeling in Heart Development

another HAT, result in increased affinity for its DNA target (Sartorelli et al., 1999). Tbx5, a T-box transcription factor, is critical for early cardiac morphogenesis and gene expression (Bruneau et al., 2001; Stennard and Harvey, 2005). Tbx5 interacts with TAZ, a WW-domain-containing transcriptional regulator, which recruits the HATs p300 and PCAF, thereby enhancing Tbx5-dependent transactivation of the Nppa promoter (Murakami et al., 2005). This cooperation is prevented by mutations in TBX5 in patients with Holt-Oram syndrome, in which congenital heart defects are prevalent (Mori and Bruneau, 2004). Thus, defective Tbx5-mediated histone acetylation may be a key mechanism for the congenital heart defects associated with the syndrome. Tbx5 and other T-box transcription factors, as well as serum response factor (SRF), also interact with Tip60, a MYSTfamily HAT, to potently activate target genes, which include serum response factor itself (Barron et al., 2005).

III.B.  Histone Deacetylases The importance of histone modifications in the heart is exemplified by histone deacetylases (HDACs), which repress gene expression (see Chapter 10.2, Volume II). HDACs have been largely characterized from mouse knockouts as important for adaptive gene regulation in the adult heart (McKinsey and Olson, 2004; Backs and Olson, 2006), but their roles in the developing heart are emerging. It should be noted that the function of HDACs is not limited to general deacetylation of histones, as they have specific mechanisms of action that involve direct interactions with DNA-binding transcription factors. For example, HDACs prevent MyoDdependent myogenesis by interactions with Mef2, thus preventing differentiation (Lu et al., 2000). Furthermore, these interactions are highly-regulated by cellular signaling processes (Lu et al., 2000; McKinsey et al., 2000). Thus, in the developing heart HDACs may have complex functions that depend on interactions with DNA-binding transcription f actors, as well as growth factor signaling pathways. Type II HDACs have specific expression patterns during embryogenesis, and in some cases these include the developing cardiovascular system (Zhang et al., 2001; Chang et al., 2004, 2006; Haberland et al., 2007). HDAC7, for example, is expressed in the endocardial lining of the heart and in the endothelium of the vasculature (Chang et al., 2006). HDAC7 mutant mice have defective heart formation, but this may be secondary to abnormal vascular structure and function, or may be related to paracrine signals that depend on HDAC7 function in the endocardium (Chang et al., 2006). One of these paracrine signals could be matrix metalloproteinase 10, a secreted protein that is important for the regulation of the extracellular matrix (Chang et al., 2006). Loss of other individual HDAC genes results in mice with no detectable embryonic cardiac

707

defects, but mice lacking both HDAC5 and HDAC9 die in utero from cardiac defects that include thinned myocardium and ventricular septal defects (Chang et al., 2004). Class I HDACs are generally ubiquitously expressed, but may have specific roles in discrete tissue types. Indeed, loss of HDAC2 results in myocardial hyperplasia, and the combined deletion in the heart of the class I HDACs, HDAC1 and HDAC2 leads to dilated hearts, arrhythmias, and aberrant expression of skeletal muscle-specific trans­cripts (Montgomery et al., 2007). While the HDAC loss-of-function phenotypes might be ascribed to a number of possible and as yet undefined pathways, HDAC activity is clearly important for processes that regulate heart development. Indeed, two transcrip­ tion factors that are essential for heart development – Smyd1 (also known as mBop) and homeodomain-only protein – both function in part through their associations with HDACs (Chen et al., 2002; Gottlieb et al., 2002; Shin et al., 2002). Serum response factor, which is important for activation of several cardiac genes, interacts with HDACs through homeodomain-only protein (HOP) (Chen et al., 2002; Shin et al., 2002) or more directly, as in adult myocytes (Davis et al., 2003). The co-repressor protein atrophin2 interacts with HDACs, and mice lacking this protein have defects in heart looping (Zoltewicz et al., 2004). It is not clear, however, which transcriptional processes are affected by disruption of this interaction. A more defined interaction between the bicoid homeo­ domain transcription factor Pitx2 and HDACs on the CyclinD2 promoter is critical for outflow tract formation. Pitx2 normally represses CyclinD2 by interacting with HDACs, but Wnt-dependent activation of -catenin relieves this transcriptional repression by preventing Pitx2 from associating with HDAC1, thereby activating CyclinD2 (Kioussi et al., 2002). Thus, through their many associations with a variety of transcription factors, HDACs are important for several aspects of heart development.

III.C.  Histone Methylation/Demethylation The potential roles of histone methylation in mammalian development are unclear, since mice lacking key histone methyltransferases (e.g., Suv39h, G9a and Ezh2) die at peri-implantation stages (O’Carroll et al., 2001; Peters et al., 2001; Tachibana et al., 2002). Tissue-specific deletion of Ezh2 has shown that this histone methyltransferase has important roles in B-cell development (Su et al., 2003), suggesting that it may also be important for other developmental processes. In particular, a role in myogenesis has been ascribed to Ezh2; activation of muscle-specific genes correlates with a loss of Ezh2 at their promoters and a concomitant recruitment of serum response factor and MyoD (Caretti et al., 2004). Ezh2 is thought to be recruited initially to muscle genes by YYI, which binds to CarG-box

708

DNA motifs. Several cardiac genes, such as Mlc2, Nppb and Actc, are also regulated by YYI (Chen and Schwartz, 1997; Bhalla et al., 2001; Latinkic et al., 2004), suggesting that negative histone methylation marks could contribute to the establishment of their expression patterns during cardiogenesis. Despite this suggestive evidence, the function of histone methyltransferases in the embryonic heart is unknown. A link between Tbx5 (and other T-box factors) and histone demethylases has recently been made (Miller et al., 2008). It was determined that T-box proteins, including Tbx5, recruit H3K27 demethylases, such as Jmjd3, and H3K4 methyltransferases, such as RbBp5, to shift a locus from harboring repressive histone marks to active histone marks (Miller et al., 2008). The impact of this function on heart development is not known, but mutations that are found in inherited disease interfere with this process (Miller et al., 2008), and thus it is likely that this is an important component of T-box factor function.

III.D.  Smyd1: A Versatile Histone-Modifying Protein Smyd1 (mBop), a protein that contains SET and MYND domains, is essential for cardiac development (Gottlieb et al., 2002). Mice lacking Smyd1 have severe defects in chamber formation, including the apparent loss of the right ventricle, thinned myocardium, defective trabeculation and excessive production of cardiac jelly. Expressed in cardiac and skeletal muscle precursors, it regulates the transcription of Hand2, possibly through a mechanism that is dependent on HDAC activity (Gottlieb et al., 2002). The MYND domain, a conserved motif that is important for recruitment of HDACs, is necessary for Smyd1’s recruitment of HDACs, which confers on it the ability to repress transcription. The recruitment of HDACs by Smyd1 is a key component of the function of the homeodomain transcription factor Irx5; Irx5 is expressed in an endocardial– epicardial gradient in the developing and adult heart, and is required for the establishment of potassium channel gradients in the heart by repressing the expression of the Kv4.2 ion channel protein, thus forming an ion channel gradient opposite that of Irx5 (Costantini et al., 2005). The repressive ability of Irx5 is conferred on it by its interaction with Smyd1, and the recruitment by Smyd1 of HDACs. Therefore, Smyd1 can act as a co-factor for DNA-binding transcription factors, and thus help shift the balance from an activator transcriptional complex to a more repressive complex. The SET domain in Smyd1 is thought to confer histone methyltransferase activity, and indeed biochemical evidence suggests that zebrafish Smyd1 can act as a histone methyltransferase (Tan et al., 2006). However, independent biochemical assays have not shown that mammalian Smyd1

PART  |  10  Epigenetic Modifiers of Cardiac Development

also possesses this ability (R. Sengupta, personal communication). Other Smyd proteins have also been shown to function as histone methyltransferases (Hamamoto et al., 2004; Brown et al., 2006; Huang et al., 2006), but intriguingly, Smyd2 has also been shown to be able to methylate other proteins, and in particular its methylation of p53 gives it an important role in regulating the cell-cycle (Huang et al., 2006). Smyd2 is also highly-enriched in heart, and it interacts with HDACs, in particular the Sin3A HDAC complex (Brown et al., 2006). Because they can function as both HDAC recruiters and as histone methyltransferases, Smyd1 and Smyd2 are versatile muscle-restricted histonemodifying proteins. It remains to be determined if both histone- and protein-modifying functions act simultaneously or independently.

III.E.  Jumonji: A Cardiac Histone Demethylase? The recent discovery that the JmjC domain is a signature motif conserved in several histone demethylases suggests that Jumonji, a member of the JmjC-containing family of proteins, is a cardiac histone demethylase (Tsukada et al., 2006; Shi and Whetstine, 2007). Mice lacking Jumonji have a variety of cardiac defects, including hyperproliferation of cardiomyocytes due to deregulation of CyclinD1 (Lee et al., 2000; Toyoda et al., 2003). Jumonji acts primarily as a transcriptional repressor (Toyoda et al., 2003; Kim et al., 2004), a function that requires the JmjC domain. This implies that Jumonji’s putative histone demethylase activity could be important for transcriptional repression (Kim et al., 2004). However, although it has yet to be demonstrated, it is highly likely that Jumonji functions as a histone demethylase during heart development. It will be interesting to ascertain what type of histone demethylation Jumonji performs, and what genetic program these modifications regulate in the developing heart.

IV.  Pc complexes and the establishment of cardiac identity IV.A.  Pc Complexes in Stem Cells: Poising Genes for Lineage Activation? Histone methylation and the recognition of its marks by Pc complexes are important in undifferentiated embryonic stem cells to repress developmentally-important transcription factors, including several that are important for heart development (Bernstein et al., 2006; Boyer et al., 2006; Lee et al., 2006). In particular, binding locations for a PRC2 component and a repressive histone methylation mark have been found to be associated with promoters for developmentally-important genes, such as Hox

Chapter  |  10.1  Chromatin Modification and Remodeling in Heart Development

clusters, and genes relevant to heart development, such as Tbx5, Gata4, Hand1 and Irx (Bernstein et al., 2006; Boyer et al., 2006; Lee et al., 2006). Some genes were characterized by the presence of both the repressive Pc proteins and RNA Pol II, or by methylation of histone H3 Lys4 (a mark of active chromatin) and Lys27 (a mark of rep­ ressed chromatin), suggesting that these genes were repressed but “poised” for rapid activation (Bernstein et al., 2006; Boyer et al., 2006; Lee et al., 2006). Indeed, several genes bound by Pc complexes were activated during differentiation (Boyer et al., 2006; Lee et al., 2006). Thus, it was proposed that regulation of these modifications is key to the transition from a pluripotent cell to a committed undifferentiated precursor, which subsequently allows the finely-directed differentiation of multipotent progenitors into distinct cardiovascular cell types. However, it remains to be seen whether this is actually the case, and whether Pc occupancy at developmentally-important genes is relevant to normal mammalian development.

IV.B.  A Role for Pc Complexes in Heart Development? At least one Pc protein, rae28, has been implicated in heart development through its regulation of Nkx2-5 (Shirai et al., 2002). Mice lacking rae28 have defects in heart looping, accompanied by changes in gene expression that include decreased levels of the basic helix-loop-helix transcription factor Hand1 and impaired expression of Nkx2-5. This finding was rather unexpected, as PcG proteins are thought mainly to keep target genes in a repressed state (Sparmann and van Lohuizen, 2006). Rae28-dependent activation of Nkx2-5 may be indirect, or this particular Pc protein may positively regulate Nkx2-5. It is quite likely that other Pc proteins are important for several aspects of heart development, but to date these remain undiscovered.

V.  Chromatin-remodeling complexes in heart development V.A.  Swi/Snf (BAF) Complexes: Baf60c and Heart Development Baf60c is one isoform of three 60 kDa BAF complex subunits that are conserved in yeast and fly Swi/Snf complexes (Fig. 4). Baf60a, Baf60b and Baf60c are encoded by three distinct genes, Smarcd1, Smarcd2 and Smarcd3. Members of this family appear to function as links between DNAbinding factors and BAF complexes. For example, Baf60a can link the glucocorticoid receptor to BAF complexes (Hsiao et al., 2003), is recruited to PPARa-binding sites in hepatic cells via interactions with PGC1a (Li et al., 2008), and may also be important for transcriptional activity of

709

c-fos/c-jun (Ito et al., 2001). In mice, Baf60a is not expressed until embryonic day (E) 9.5, and Baf60b is expressed in the developing gut and limb buds (Lickert et al., 2004). Baf60c is expressed very early in precardiac mesoderm at late gastrulation and prenodal plate stages; at later stages, it is specifically expressed in the heart and in somites and the developing central nervous system (Debril et al., 2004; Lickert et al., 2004). Therefore, Baf60c is a tissue-specific subunit of the BAF complex, perhaps serving to remodel chromatin at specific loci during heart development. RNAi-mediated knockdown of Baf60c in mouse embryos results in specific defects in heart and somite formation (Lickert et al., 2004). Embryos with a 95% knockdown of Baf60c have severely malformed hearts and aberrant gene expression, most notably loss of trabecular markers and several outflow tract marker genes. Expression of Baf60c at around 40% of wild-type levels results in a less severe, but more specific, defect in outflow tract formation that is reminiscent of human congenital heart defects, such as persistent truncus arteriosus. The expression of other cardiac transcription factors is not disrupted in Baf60c-deficient embryos, which suggests that loss of Baf60c impairs these factors’ abilities to activate their target genes. This indicates that Baf60c might potentiate transcription factor action. Indeed, in co-immunoprecipitation experiments, Baf60c acted as a bridge, creating or reinforcing molecular interactions between cardiac transcription factors (Gata4, Nkx2-5 and Tbx5) and Brg1, thus bringing the BAF complex to target genes bound by Gata4, Nkx25 or Tbx5 (Fig. 5) (Lickert et al., 2004). Baf60c is also important for Notch-mediated transcriptional activation, as exemplified by its role in activating the Notch-dependent gene Nodal, which is required for the initiation of left– right asymmetry (Takeuchi et al., 2007). Indeed, embryos with a knockdown of Baf60c have defective establishment of the left–right cascade due to loss of Nodal expression around the node. In this case, Baf60c is important for stabilizing the interaction between cleaved intracellular Notch and its DNA-binding partner protein, RBPj, while simultaneously bringing the BAF complex to Notch target genes (Takeuchi et al., 2007). Why would a tissue-specific chromatin-remodeling complex subunit be needed? Since DNA-binding trans­cription factors recognize a particularly loosely defined DNA sequence, proteins such as Baf60c might coordinate specific chromatin-remodeling by interacting with DNAbinding transcription factors at specific loci that require tight regulation. The critical importance of Baf60c has been identified in recent studies showing that Baf60c is a required component of a trio of cardiogenic transcription factors that include GATA4 and Tbx5, which together can activate the entire cardiac program de novo in noncardiac mesoderm and fully reprogram noncardiac mesoderm to contractile cardiac myocytes (Takeuchi and Bruneau,

710

PART  |  10  Epigenetic Modifiers of Cardiac Development

Figure 5  A model for gene activation by cardiac transcription factors and their interactions with Baf60c. Baf60c provides a link between the BAF complexes and DNA-binding factors such as Tbx5 and Nkx2-5, thus promoting chromatin-remodeling at target gene loci and recruitment of the Pol II complex.

2009). Unlike other contexts where DNA-binding factors are sufficient to enact reprograming (Davis et al., 1987; Xie et al., 2004; Takahashi and Yamanaka, 2006; Zhou et al., 2008), Baf60c is required to potentiate the function of GATA4 and Tbx5. The function of Baf60c in this case is to allow the initial binding of GATA4 to target loci, indicating that its function is absolutely essential for the cardiogenic activity of GATA4 and Tbx5.

V.B.  Swi/Snf (BAF) Complexes: Brg1 and Heart Development As mentioned above, Brg1-null mice die before implantation. However, zebrafish that harbor a null allele of Brg1, the zebrafish young mutant, survive until late stages of embryogenesis, perhaps due to maternal deposition of the protein, and have defective differentiation of the retina cell, as well as defects in heart formation (Link et al., 2000; Gregg et al., 2003). On closer examination, young hearts had very specific functional and structural defects; underdeveloped chambers and striking arrhythmias (Scott, personal communication). Deletion of Brg1 in the differentiating mouse heart also deregulates gene expression, leading to specific defects in chamber morphogenesis, including a lack of distinction between the two ventricular chambers (Takeuchi and Bruneau, unpublished data). Also striking is the dose-dependence of heart development on Brg1. Some heterozygous Brg1-null mice have congenital heart defects, and compound heterozygosity of a Brg1-null allele with individual null alleles for a number of cardiac transcription factor genes results in synthetic lethality due to severe cardiac malformations (Takeuchi and Bruneau, unpublished data). This finding suggests that the balance of transcriptional activation is stoichiometrically regulated

by the interaction between BAF complexes and DNAbinding transcription factors. Deletion of Brg1 in the developing endocardium has led to the surprising finding that BAF complexes are required for very specific functions in establishing endocardial– myocardial signaling (Stankunas et al., 2008). Loss of Brg1 in endocardium results in abnormal trabeculation of the myocardium, but generally normal gene expression in both endocardial and myocardial layers. A notable exception is a striking upregulation of the matrix metalloprotease ADAMTS1 in Brg1-null endocardium, leading to abnormal establishment of the cardiac jelly and, as a secondary consequence, defective trabeculation. Thus, as in T-cells, BAF complexes in the endocardium are important for setting up transcriptional programs, partly via repression of target genes.

V.C.  BAF Complexes: Baf250a and Heart Development Baf250, a key component of BAF complexes not found in the related PBAF complex (Wang et al., 1996; Lemon et al., 2001), appears to be critically important in a dosesensitive manner for several aspects of embryogenesis, particularly heart development. BAF250a heterozygous embryos die around midgestation in chimera-derived embryos, and a high percentage had neural tube ­ closure defects and thin myocardial wall syndrome (Zhong Wang, personal communication). Studies of a conditional BAF250a knockout showed that BAF250a is clearly involved in the proper development of the cardiac cell lineages derived from both primary and second heart fields (Gao et al., 2008; and Zhong Wang, personal communication). Why certain BAF complex subunits are so critical for the stoichiometric function of the complexes while

Chapter  |  10.1  Chromatin Modification and Remodeling in Heart Development

others are not is unclear. Nevertheless, tightly-regulated amounts of assembled complexes are certainly important for fine regulation of cardiac lineage specification and organogenesis.

V.D.  PBAF Complexes: Baf180 and Heart Development Further insight into the role of chromatin-remodeling complexes during heart development came from studies of knockout mice lacking the Polybromo protein Baf180 (Wang et al., 2004). Baf180, a subunit of PBAF complexes, is not found in the related but distinct BAF complexes (Lemon et al., 2001; de la Serna et al., 2006). PBAF complexes are thought to be mainly involved in potentiating transcriptional activation by nuclear receptors (e.g., RXR, VDR and PPAR). Loss of Baf180 did not lead to early embryonic lethality. Instead, Baf180-null embryos had a very thin cardiac wall and fewer trabeculations (Wang et al., 2004) – a phenotype similar to that of RXRa knockout mice (Sucov et al., 1994) that reveals a specific role for the PBAF complex in late aspects of cardiac chamber maturation. Indeed, similar to retinoic acid signaling, Baf180 is critically required for coronary artery development and epicardial differentiation (Huang et al., 2008). These results suggest that the PBAF complex has critical roles in embryogenesis, especially in the late stages of heart formation, and different roles from Baf60c during cardiac development. Thus, diversity of transcriptional regulation by two related but distinct chromatin-remodeling complexes, BAF and PBAF, may regulate distinct pathways in heart development.

V.E.  Baf45c/DPF3 and Recognition of Histone Modifications The coordination of chromatin-remodeling complexes, DNA-binding transcription factors and histone modifications is not well-understood. It might occur through interactions with proteins that recognize histone modifications such as chromodomain, PHD-domain, or WD40-domain proteins (Ruthenburg et al., 2007). A subunit of the BAF complex, Baf45c (also known as DPF3), may provide some answers. Baf45c is predominantly expressed in the developing heart and recognizes specific histone modifications via two PHD domains, suggesting that it may enhance the specificity of target-gene recognition (Lange et al., 2008). Along with Baf60c, a combination of tissuerestricted BAF complex subunits may be required to enact a specific transcriptional program. The other Baf45 family members, Baf45a and Baf45b, have been implicated in a shift between neural progenitor- and neuron-specific transcription (Lessard et al., 2007), and thus the particular combination of these subunits may dictate recognition of specific histone modifications during development.

711

VI.  Conclusions Transcriptional regulation of heart development is tightlylinked to chromatin-remodeling. This is especially apparent with respect to complexes that alter the structure of chromatin, such as the BAF complexes. It is also readily apparent that enzymatic modification of histones is a powerful means of affecting lineage decisions, and is likely to be a critical mediator of several aspects of heart development. The integration of signaling inputs and transcriptional activity by chromatin-modifying and chromatinremodeling complexes adds a layer of complexity and fine-tuning to the myriad events that drive heart development, from early lineage determination, throughout cardiac organogenesis and including postnatal regulation of patterned gene expression.

References Backs, J., Olson, E.N., 2006. Control of cardiac growth by histone acetylation/deacetylation. Circ. Res. 98, 15–24. Bamforth, S.D., Bragança, J., Eloranta, J.J., Murdoch, J.N., Marques, F.I., Kranc, K.R., Farza, H., Henderson, D.J., Hurst, H.C., Bhattacharya, S., 2001. Cardiac malformations, adrenal agenesis, neural crest defects and exencephaly in mice lacking Cited2, a new Tfap2 co-activator. Nat. Genet. 29, 469–474 doi: 10.1038/ng768. Barron, M.R., Belaguli, N.S., Zhang, S.X., Trinh, M., Iyer, D., Merlo, X., Lough, J.W., Parmacek, M.S., Bruneau, B.G., Schwartz, R.J., 2005. Serum response factor, an enriched cardiac mesoderm obligatory factor, is a downstream gene target for Tbx genes. J. Biol. Chem. 280, 11816–11828. Bernstein, B.E., Mikkelsen, T.S., Xie, X., Kamal, M., Huebert, D.J., Cuff, J., Fry, B., Meissner, A., Wernig, M., Plath, K., Jaenisch, R., Wagschal, A., Feil, R., Schreiber, S.L., Lander, E.S., 2006. A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell 125, 315–326. Bhalla, S.S., Robitaille, L., Nemer, M., 2001. Cooperative activation by GATA-4 and YY1 of the cardiac B-type natriuretic peptide promoter. J. Biol. Chem. 276, 11439–11445. Bhattacharya, S., Michels, C.L., Leung, M.K., Arany, Z.P., Kung, A.L., Livingston, D.M., 1999. Functional role of p35srj, a novel p300/CBP binding protein, during transactivation by HIF-1. Genes Dev. 13, 64–75. Boyer, L.A., Plath, K., Zeitlinger, J., Brambrink, T., Medeiros, L.A., Lee, T.I., Levine, S.S., Wernig, M., Tajonar, A., Ray, M.K., Bell, G.W., Otte, A.P., Vidal, M., Gifford, D.K., Young, R.A., Jaenisch, R., 2006. Polycomb complexes repress developmental regulators in murine embryonic stem cells. Nature 441, 349–353. Brown, M.A., Sims, R.J. III, Gottlieb, P.D., Tucker, P.W., 2006. Identification and characterization of Smyd2, a split SET/MYND domain-containing histone H3 lysine 36-specific methyltransferase that interacts with the Sin3 histone deacetylase complex. Mol. Cancer 5, 26. Bruneau, B.G., 2002. Transcriptional regulation of vertebrate cardiac morphogenesis. Circ. Res. 90, 509–519. Bruneau, B.G., Nemer, G., Schmitt, J.P., Charron, F., Robitaille, L., Caron, S., Conner, D., Gessler, M., Nemer, M., Seidman, C.E., Seidman, J.G., 2001. A murine model of Holt-Oram syndrome

712

defines roles of the T-box transcription factor Tbx5 in cardiogenesis and disease. Cell 106, 709–721. Bultman, S., Gebuhr, T., Yee, D., La Mantia, C., Nicholson, J., Gilliam, A., Randazzo, F., Metzger, D., Chambon, P., Crabtree, G., Magnuson, T., 2000. A Brg1 null mutation in the mouse reveals functional differences among mammalian SWI/SNF complexes. Mol. Cell 6, 1287–1295. Caretti, G., Di Padova, M., Micales, B., Lyons, G.E., Sartorelli, V., 2004. The Polycomb Ezh2 methyltransferase regulates muscle gene expression and skeletal muscle differentiation. Genes Dev. 18, 2627–2638. Chang, S., McKinsey, T.A., Zhang, C.L., Richardson, J.A., Hill, J.A., Olson, E.N., 2004. Histone deacetylases 5 and 9 govern responsiveness of the heart to a subset of stress signals and play redundant roles in heart development. Mol. Cell Biol. 24, 8467–8476. Chang, S., Young, B.D., Li, S., Qi, X., Richardson, J.A., Olson, E.N., 2006. Histone deacetylase 7 maintains vascular integrity by repressing matrix metalloproteinase 10. Cell 126, 321–334. Chen, C.Y., Schwartz, R.J., 1997. Competition between negative acting YY1 versus positive acting serum response factor and tinman homologue Nkx-2.5 regulates cardiac alpha-actin promoter activity. Mol. Endocrinol. 11, 812–822. Chen, F., Kook, H., Milewski, R., Gitler, A.D., Lu, M.M., Li, J., Nazarian, R., Schnepp, R., Jen, K., Biben, C., Runke, G., Mackay, J. P., Novotny, J., Schwartz, R.J., Harvey, R.P., Mullins, M.C., Epstein, J.A., 2002. Hop is an unusual homeobox gene that modulates cardiac development. Cell 110, 713–723. Chen, Z., Zang, J., Whetstine, J., Hong, X., Davrazou, F., Kutateladze, T.G., Simpson, M., Mao, Q., Pan, C.H., Dai, S., Hagman, J., Hansen, K., Shi, Y., Zhang, G., 2006. Structural insights into histone demethylation by JMJD2 family members. Cell 125, 691–702. Chi, T.H., Wan, M., Zhao, K., Taniuchi, I., Chen, L., Littman, D.R., Crabtree, G.R., 2002. Reciprocal regulation of CD4/CD8 expression by SWI/SNF-like BAF complexes. Nature 418, 195–199. Chi, T.H., Wan, M., Lee, P.P., Akashi, K., Metzger, D., Chambon, P., Wilson, C.B., Crabtree, G.R., 2003. Sequential roles of Brg, the ATPase subunit of BAF chromatin remodeling complexes, in thymocyte development. Immunity 19, 169–182. Costantini, D.L., Arruda, E.P., Agarwal, P., Kim, K.H., Zhu, Y., Zhu, W., Lebel, M., Cheng, C.W., Park, C.Y., Pierce, S.A., Guerchicoff, A., Pollevick, G.D., Chan, T.Y., Kabir, M.G., Cheng, S.H., Husain, M., Antzelevitch, C., Srivastava, D., Gross, G.J., Hui, C.C., Backx, P.H., Bruneau, B.G., 2005. The homeodomain transcription factor Irx5 establishes the mouse cardiac ventricular repolarization gradient. Cell 123, 347–358. Dai, Y.S., Markham, B.E., 2001. P300 functions as a coactivator of transcription factor gata-4. J. Biol. Chem. 276, 37178–37185. Davis, F.J., Gupta, M., Camoretti-Mercado, B., Schwartz, R.J., Gupta, M.P., 2003. Calcium/calmodulin-dependent protein kinase activates serum response factor transcription activity by its dissociation from histone deacetylase, HDAC4. Implications in cardiac muscle gene regulation during hypertrophy. J. Biol. Chem. 278, 20047–20058. Davis, R.L., Weintraub, H., Lassar, A.B., 1987. Expression of a single transfected cDNA converts fibroblasts to myoblasts. Cell 51, 987–1000. Debril, M.B., Gelman, L., Fayard, E., Annicotte, J.S., Rocchi, S., Auwerx, J., 2004. Transcription factors and nuclear receptors interact with the SWI/SNF complex through the BAF60c subunit. J. Biol. Chem. 279, 16677–16686. de la Serna, I.L., Ohkawa, Y., Imbalzano, A.N., 2006. Chromatin remodelling in mammalian differentiation, lessons from ATP-dependent remodellers. Nat. Rev. Genet. 7, 461–473.

PART  |  10  Epigenetic Modifiers of Cardiac Development

Gao, X., Tate, P., Hu, P., Tjian, R., Skarnes, W.C., Wang, Z., 2008. ES cell pluripotency and germ-layer formation requires the SWI/SNF chromatin remodeling component BAF250a. Proc. Natl. Acad. Sci. U.S.A. 105, 6656–6661. Gebuhr, T.C., Kovalev, G.I., Bultman, S., Godfrey, V., Su, L., Magnuson, T., 2003. The role of Brg1, a catalytic subunit of mammalian chromatin-remodeling complexes, in T cell development. J. Exp. Med. 198, 1937–1949. Gottlieb, P.D., Pierce, S.A., Sims, R.J., Yamagishi, H., Weihe, E.K., Harriss, J.V., Maika, S.D., Kuziel, W.A., King, H.L., Olson, E.N., Nakagawa, O., Srivastava, D., 2002. Bop encodes a muscle-restricted protein containing MYND and SET domains and is essential for cardiac differentiation and morphogenesis. Nat. Genet. 31, 25–32. Gregg, R.G., Willer, G.B., Fadool, J.M., Dowling, J.E., Link, B.A., 2003. Positional cloning of the young mutation identifies an essential role for the Brahma chromatin remodeling complex in mediating retinal cell differentiation. Proc. Natl. Acad. Sci. U.S.A. 100, 6535–6540. Guidi, C.J., Sands, A.T., Zambrowicz, B.P., Turner, T.K., Demers, D.A., Webster, W., Smith, T.W., Imbalzano, A.N., Jones, S.N., 2001. Disruption of Ini1 leads to peri-implantation lethality and tumorigenesis in mice. Mol. Cell Biol. 21, 3598–3603. Haberland, M., Arnold, M.A., McAnally, J., Phan, D., Kim, Y., Olson, E.N., 2007. Regulation of HDAC9 gene expression by MEF2 establishes a negative-feedback loop in the transcriptional circuitry of muscle differentiation. Mol. Cell Biol. 27, 518–525. Hamamoto, R., Furukawa, Y., Morita, M., Iimura, Y., Silva, F.P., Li, M., Yagyu, R., Nakamura, Y., 2004. SMYD3 encodes a histone methyltransferase involved in the proliferation of cancer cells. Nat. Cell Biol. 6, 731–740. Hsiao, P.W., Fryer, C.J., Trotter, K.W., Wang, W., Archer, T.K., 2003. BAF60a mediates critical interactions between nuclear receptors and the BRG1 chromatin-remodeling complex for transactivation. Mol. Cell Biol. 23, 6210–6220. Huang, J., Perez-Burgos, L., Placek, B.J., Sengupta, R., Richter, M., Dorsey, J.A., Kubicek, S., Opravil, S., Jenuwein, T., Berger, S.L., 2006. Repression of p53 activity by Smyd2-mediated methylation. Nature 444, 629–632. Huang, X., Gao, X., Diaz-Trelles, R., Ruiz-Lozano, P., Wang, Z., 2008. Coronary development is regulated by ATP-dependent SWI/SNF chromatin remodeling component BAF180. Dev. Biol. 319, 258–266. Indra, A.K., Dupe, V., Bornert, J.M., Messaddeq, N., Yaniv, M., Mark, M., Chambon, P., Metzger, D., 2005. Temporally controlled targeted somatic mutagenesis in embryonic surface ectoderm and fetal epidermal keratinocytes unveils two distinct developmental functions of BRG1 in limb morphogenesis and skin barrier formation. Development 132, 4533–4544. Ito, T., Yamauchi, M., Nishina, M., Yamamichi, N., Mizutani, T., Ui, M., Murakami, M., Iba, H., 2001. Identification of SWI.SNF complex subunit BAF60a as a determinant of the transactivation potential of Fos/Jun dimers. J. Biol. Chem. 276, 2852–2857. Jenuwein, T., Allis, C.D., 2001. Translating the histone code. Science 293, 1074–1080. Kawamura, T., Ono, K., Morimoto, T., Wada, H., Hirai, M., Hidaka, K., Morisaki, T., Heike, T., Nakahata, T., Kita, T., Hasegawa, K., 2005. Acetylation of GATA-4 is involved in the differentiation of embryonic stem cells into cardiac myocytes. J. Biol. Chem. 280, 19682–19688. Khorasanizadeh, S., 2004. The nucleosome: From genomic organization to genomic regulation. Cell 116, 259–272. Kim, J.K., Huh, S.O., Choi, H., Lee, K.S., Shin, D., Lee, C., Nam, J.S., Kim, H., Chung, H., Lee, H.W., Park, S.D., Seong, R.H., 2001. Srg3,

Chapter  |  10.1  Chromatin Modification and Remodeling in Heart Development

a mouse homolog of yeast SWI3, is essential for early embryogenesis and involved in brain development. Mol. Cell Biol. 21, 7787–7795. Kim, T.G., Chen, J., Sadoshima, J., Lee, Y., 2004. Jumonji represses atrial natriuretic factor gene expression by inhibiting transcriptional activities of cardiac transcription factors. Mol. Cell Biol. 24, 10151–10160. Kioussi, C., Briata, P., Baek, S.H., Rose, D.W., Hamblet, N.S., Herman, T., Ohgi, K.A., Lin, C., Gleiberman, A., Wang, J., Brault, V., RuizLozano, P., Nguyen, H.D., Kemler, R., Glass, C.K., Wynshaw-Boris, A., Rosenfeld, M.G., 2002. Identification of a Wnt/Dvl/beta-Catenin -- Pitx2 pathway mediating cell-type-specific proliferation during development. Cell 111, 673–685. Klochendler-Yeivin, A., Fiette, L., Barra, J., Muchardt, C., Babinet, C., Yaniv, M., 2000. The murine SNF5/INI1 chromatin remodeling factor is essential for embryonic development and tumor suppression. EMBO Rep. 1, 500–506. Klose, R.J., Yamane, K., Bae, Y., Zhang, D., Erdjument-Bromage, H., Tempst, P., Wong, J., Zhang, Y., 2006. The transcriptional repressor JHDM3A demethylates trimethyl histone H3 lysine 9 and lysine 36. Nature 442, 312–316. Lange, M., Kaynak, B., Forster, U.B., Toenjes, M., Fisher, J.J., Grimm, C., Schlesinger, J., Just, S., Dunkel, I., Krueger, T., Mebus, S., Lehrach, H., Lurz, R., Gobom, J., Rottbauer, W., Abdelilah-Seyfried, S., Sperling, S., 2008. Regulation of muscle development by DPF3, a novel histone acetylation and methylation reader of the BAF chromatin remodeling complex. Genes Dev. 22 (17), 2370–2384. Latinkic, B.V., Cooper, B., Smith, S., Kotecha, S., Towers, N., Sparrow, D., Mohun, T.J., 2004. Transcriptional regulation of the cardiac-specific MLC2 gene during Xenopus embryonic development. Development 131, 669–679. Lee, T.I., Jenner, R.G., Boyer, L.A., Guenther, M.G., Levine, S.S., Kumar, R.M., Chevalier, B.E., Johnstone, S.E., Cole, M.F., Isono, K., Koseki, H., Fuchikami, T., Abe, K., Murray, H.L., Zucker, J.P., Yuan, B., Bell, G.W., Herbolsheimer, E., Hannett, N.M., Sun, K., Odom, D. T., Otte, A.P., Volkert, T.L., Bartel, D.P., Melton, D.A., Gifford, D.K., Jaenisch, R., Young, R.A., 2006. Control of developmental regulators by Polycomb in human embryonic stem cells. Cell 125, 301–313. Lee, Y., Song, A.J., Baker, R., Micales, B., Conway, S.J., Lyons, G.E., 2000. Jumonji, a nuclear protein that is necessary for normal heart development. Circ. Res. 86, 932–938. Lemon, B., Inouye, C., King, D.S., Tjian, R., 2001. Selectivity of chromatin-remodelling cofactors for ligand-activated transcription. Nature 414, 924–928. Lessard, J., Wu, J.I., Ranish, J.A., Wan, M., Winslow, M.M., Staahl, B.T., Wu, H., Aebersold, R., Graef, I.A., Crabtree, G.R., 2007. An essential switch in subunit composition of a chromatin remodeling complex during neural development. Neuron 55, 201–215. Li, F., Huarte, M., Zaratiegui, M., Vaughn, M.W., Shi, Y., Martienssen, R., Cande, W.Z., 2008. Lid2 is required for coordinating H3K4 and H3K9 methylation of heterochromatin and euchromatin. Cell 135, 272–283. Li, H., Ilin, S., Wang, W., Duncan, E.M., Wysocka, J., Allis, C.D., Patel, D.J., 2006. Molecular basis for site-specific read-out of histone H3K4me3 by the BPTF PHD finger of NURF. Nature 442, 91–95. Li, S., Liu, C., Li, N., Hao, T., Han, T., Hill, D.E., Vidal, M., Lin, J.D., 2008. Genome-wide coactivation analysis of PGC-1alpha identifies BAF60a as a regulator of hepatic lipid metabolism. Cell Metab. 8, 105–117. Lickert, H., Takeuchi, J.K., von Both, I., Walls, J.R., McAuliffe, F., Adamson, S.L., Henkelman, R.M., Wrana, J.L., Rossant, J., Bruneau, B.G., 2004. Baf60c is essential for function of BAF chromatin remodelling complexes in heart development. Nature 432, 107–112.

713

Link, B.A., Fadool, J.M., Malicki, J., Dowling, J.E., 2000. The zebrafish young mutation acts non-cell-autonomously to uncouple differentiation from specification for all retinal cells. Development 127, 2177–2188. Lu, J., McKinsey, T.A., Zhang, C.L., Olson, E.N., 2000. Regulation of skeletal myogenesis by association of the MEF2 transcription factor with class II histone deacetylases. Mol. Cell 6, 233–244. Martens, J.A., Winston, F., 2003. Recent advances in understanding chromatin remodeling by Swi/Snf complexes. Curr. Opin. Genet. Dev. 13, 136–142. McKinsey, T.A., Zhang, C.L., Lu, J., Olson, E.N., 2000. Signaldependent nuclear export of a histone deacetylase regulates muscle differentiation. Nature 408, 106–111. McKinsey, T.A., Olson, E.N., 2004. Cardiac histone acetylation – therapeutic opportunities abound. Trends Genet. 20, 206–213. Miller, S.A., Huang, A.C., Miazgowicz, M.M., Brassil, M.M., Weinmann, A.S., 2008. Coordinated but physically separable interaction with H3K27demethylase and H3K4-methyltransferase activities are required for T-box protein-mediated activation of developmental gene expression. Genes Dev. 22, 2980–2993. Molinari, S., Relaix, F., Lemonnier, M., Kirschbaum, B., Schafer, B., Buckingham, M.E., 2004. A novel complex regulates cardiac actin gene expression through interaction of Emb, a class VI POU domain protein, MEF2D, and the histone transacetylase p300. Mol. Cell Biol. 24, 2944–2957. Montgomery, R.L., Davis, C.A., Potthoff, M.J., Haberland, M., Fielitz, J., Qi, X., Hill, J.A., Richardson, J.A., Olson, E.N., 2007. Histone deacetylases 1 and 2 redundantly regulate cardiac morphogenesis, growth, and contractility. Genes Dev. 21, 1790–1802. Mori, A.D., Bruneau, B.G., 2004. TBX5 mutations and congenital heart disease: Holt-Oram syndrome revealed. Curr. Opin. Cardiol. 19, 211–215. Morin, S., Pozzulo, G., Robitaille, L., Cross, J., Nemer, M., 2005. MEF2dependent recruitment of the HAND1 transcription factor results in synergistic activation of target promoters. J. Biol. Chem. 280, 32272–32278. Murakami, M., Nakagawa, M., Olson, E.N., Nakagawa, O., 2005. A WW domain protein TAZ is a critical coactivator for TBX5, a transcription factor implicated in Holt-Oram syndrome. Proc. Natl. Acad. Sci. U.S.A. 102, 18034–18039. O’Carroll, D., Erhardt, S., Pagani, M., Barton, S.C., Surani, M.A., Jenuwein, T., 2001. The polycomb-group gene Ezh2 is required for early mouse development. Mol. Cell Biol. 21, 4330–4336. Olave, I.A., Reck-Peterson, S.L., Crabtree, G.R., 2002. Nuclear actin and actin-related proteins in chromatin remodeling. Annu. Rev. Biochem. 71, 755–781. Pena, P.V., Davrazou, F., Shi, X., Walter, K.L., Verkhusha, V.V., Gozani, O., Zhao, R., Kutateladze, T.G., 2006. Molecular mechanism of histone H3K4me3 recognition by plant homeodomain of ING2. Nature 442, 100–103. Peters, A.H., O’Carroll, D., Scherthan, H., Mechtler, K., Sauer, S., Schofer, C., Weipoltshammer, K., Pagani, M., Lachner, M., Kohlmaier, A., Opravil, S., Doyle, M., Sibilia, M., Jenuwein, T., 2001. Loss of the Suv39h histone methyltransferases impairs mammalian heterochromatin and genome stability. Cell 107, 323–337. Reyes, J.C., Barra, J., Muchardt, C., Camus, A., Babinet, C., Yaniv, M., 1998. Altered control of cellular proliferation in the absence of mammalian brahma (SNF2alpha). Embo. J. 17, 6979–6991. Ruthenburg, A.J., Allis, C.D., Wysocka, J., 2007. Methylation of lysine 4 on histone h3: Intricacy of writing and reading a single epigenetic mark. Mol. Cell 25, 15–30.

714

Sartorelli, V., Puri, P.L., Hamamori, Y., Ogryzko, V., Chung, G., Nakatani, Y., Wang, J.Y., Kedes, L., 1999. Acetylation of MyoD directed by PCAF is necessary for the execution of the muscle program. Mol. Cell 4, 725–734. Shi, Y., Whetstine, J.R., 2007. Dynamic regulation of histone lysine methylation by demethylases. Mol. Cell 25, 1–14. Shin, C.H., Liu, Z.P., Passier, R., Zhang, C.L., Wang, D.Z., Harris, T.M., Yamagishi, H., Richardson, J.A., Childs, G., Olson, E.N., 2002. Modulation of cardiac growth and development by HOP, an unusual homeodomain protein. Cell 110, 725–735. Shirai, M., Osugi, T., Koga, H., Kaji, Y., Takimoto, E., Komuro, I., Hara, J., Miwa, T., Yamauchi-Takihara, K., Takihara, Y., 2002. The Polycomb-group gene Rae28 sustains Nkx2.5/Csx expression and is essential for cardiac morphogenesis. J. Clin. Invest. 110, 177–184. Sparmann, A., van Lohuizen, M., 2006. Polycomb silencers control cell fate, development and cancer. Nat. Rev. Cancer 6, 846–856. Srivastava, D., 2006. Making or breaking the heart: From lineage determination to morphogenesis. Cell 126, 1037–1048. Stankunas, K., Hang, C.T., Tsun, Z.Y., Chen, H., Lee, N.V., Wu, J.I., Shang, C., Bayle, J.H., Shou, W., Iruela-Arispe, M.L., Chang, C.P., 2008. Endocardial Brg1 represses ADAMTS1 to maintain the microenvironment for myocardial morphogenesis. Dev. Cell 14, 298–311. Stennard, F.A., Harvey, R.P., 2005. T-box transcription factors and their roles in regulatory hierarchies in the developing heart. Development 132, 4897–4910. Stopka, T., Skoultchi, A.I., 2003. The ISWI ATPase Snf2h is required for early mouse development. Proc. Natl. Acad. Sci. U.S.A. 100, 14097–14102. Su, I.H., Basavaraj, A., Krutchinsky, A.N., Hobert, O., Ullrich, A., Chait, B.T., Tarakhovsky, A., 2003. Ezh2 controls B cell development through histone H3 methylation and Igh rearrangement. Nat. Immunol. 4, 124–131. Sucov, H.M., Dyson, E., Gumeringer, C.L., Price, J., Chien, K.R., Evans, R.M., 1994. RXR alpha mutant mice establish a genetic basis for vitamin A signaling in heart morphogenesis. Genes Dev. 8, 1007–1018. Tachibana, M., Sugimoto, K., Nozaki, M., Ueda, J., Ohta, T., Ohki, M., Fukuda, M., Takeda, N., Niida, H., Kato, H., Shinkai, Y., 2002. G9a histone methyltransferase plays a dominant role in euchromatic histone H3 lysine 9 methylation and is essential for early embryogenesis. Genes Dev. 16, 1779–1791. Takahashi, K., Yamanaka, S., 2006. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 126, 663–676. Takeuchi, J.K., Lickert, H., Bisgrove, B.W., Sun, X., Yamamoto, M., Chawengsaksophak, K., Hamada, H., Yost, H.J., Rossant, J., Bruneau, B.G., 2007. Baf60c is a nuclear Notch signaling component required for the establishment of left–right asymmetry. Proc. Natl. Acad. Sci. U.S.A. 104, 846–851. Takeuchi, J. K., Bruneau, B. G., 2009. Reprogramming, of mouse mesoderm to heart tissue by defined factors (submitted).

PART  |  10  Epigenetic Modifiers of Cardiac Development

Tamkun, J.W., Deuring, R., Scott, M.P., Kissinger, M., Pattatucci, A.M., Kaufman, T.C., Kennison, J.A., 1992. Brahma, a regulator of Drosophila homeotic genes structurally related to the yeast transcriptional activator SNF2/SWI2. Cell 68, 561–572. Tan, X., Rotllant, J., Li, H., De Deyne, P., Du, S.J., 2006. SmyD1, a histone methyltransferase, is required for myofibril organization and muscle contraction in zebrafish embryos. Proc. Natl. Acad. Sci. U.S.A. 103 (20), 7935. Toyoda, M., Shirato, H., Nakajima, K., Kojima, M., Takahashi, M., Kubota, M., Suzuki-Migishima, R., Motegi, Y., Yokoyama, M., Takeuchi, T., 2003. Jumonji downregulates cardiac cell proliferation by repressing cyclin D1 expression. Dev. Cell 5, 85–97. Trojer, P., Reinberg, D., 2006. Histone lysine demethylases and their impact on epigenetics. Cell 125, 213–217. Tsukada, Y., Fang, J., Erdjument-Bromage, H., Warren, M.E., Borchers, C.H., Tempst, P., Zhang, Y., 2006. Histone demethylation by a family of JmjC domain-containing proteins. Nature 439, 811–816. Wang, W., Cote, J., Xue, Y., Zhou, S., Khavari, P.A., Biggar, S.R., Muchardt, C., Kalpana, G.V., Goff, S.P., Yaniv, M., Workman, J.L., Crabtree, G.R., 1996. Purification and biochemical heterogeneity of the mammalian SWI-SNF complex. Embo J. 15, 5370–5382. Wang, W., Xue, Y., Zhou, S., Kuo, A., Cairns, B.R., Crabtree, G.R., 1996. Diversity and specialization of mammalian SWI/SNF complexes. Genes Dev. 10, 2117–2130. Wang, Z., Zhai, W., Richardson, J.A., Olson, E.N., Meneses, J.J., Firpo, M.T., Kang, C., Skarnes, W.C., Tjian, R., 2004. Polybromo protein BAF180 functions in mammalian cardiac chamber maturation. Genes Dev. 18, 3106–3116. Whetstine, J.R., Nottke, A., Lan, F., Huarte, M., Smolikov, S., Chen, Z., Spooner, E., Li, E., Zhang, G., Colaiacovo, M., Shi, Y., 2006. Reversal of histone lysine trimethylation by the JMJD2 family of histone demethylases. Cell 125, 467–481. Wysocka, J., Swigut, T., Xiao, H., Milne, T.A., Kwon, S.Y., Landry, J., Kauer, M., Tackett, A.J., Chait, B.T., Badenhorst, P., Wu, C., Allis, C.D., 2006. A PHD finger of NURF couples histone H3 lysine 4 trimethylation with chromatin remodelling. Nature 442, 86–90. Xie, H., Ye, M., Feng, R., Graf, T., 2004. Stepwise reprogramming of B cells into macrophages. Cell 117, 663–676. Yao, T.P., Oh, S.P., Fuchs, M., Zhou, N.D., Ch’ng, L.E., Newsome, D., Bronson, R.T., Li, E., Livingston, D.M., Eckner, R., 1998. Gene dosage-dependent embryonic development and proliferation defects in mice lacking the transcriptional integrator p300. Cell 93, 361–372. Zhang, C.L., McKinsey, T.A., Olson, E.N., 2001. The transcriptional corepressor MITR is a signal-responsive inhibitor of myogenesis. Proc. Natl. Acad. Sci. U.S.A. 98, 7354–7359. Zhou, Q., Brown, J., Kanarek, A., Rajagopal, J., Melton, D.A., 2008. In vivo reprogramming of adult pancreatic exocrine cells to betacells. Nature 455 (7213), 627–632. Zoltewicz, J.S., Stewart, N.J., Leung, R., Peterson, A.S., 2004. Atrophin 2 recruits histone deacetylase and is required for the function of multiple signaling centers during mouse embryogenesis. Development 131, 3–14.

Chapter 10.2

Histone Deacetylases in Cardiovascular Development and Disease Bryan D. Young and Eric N. Olson Department of Molecular Biology, University of Texas Southwestern Medical Center at Dallas, Dallas, TX, USA

The last decade has seen rapid progress toward defining the mechanisms that control cardiac gene expression during development. Chromatin-modifying enzymes have emerged as central regulators of the gene programs that drive cardiogenesis. Some of these enzymes, particularly the histone acetyl transferases (HATs) and histone deacetylases (HDACs), act at the interface between cellular signaling pathways and DNA-bound transcriptional co-activators. Here, we discuss the signal-responsive regulation that HDACs provide to the cardiogenic and hypertrophic transcriptional pathways.

I.  Histone acetyl transferases and histone deacetylases Both the development and stress-induced growth of the heart are controlled by combinatorial interactions of transcription factors, including myocyte enhancer factor 2 (MEF2), serum response factor (SRF), nuclear factor of activated T-cells (NFAT) and the GATA family zinc-finger proteins (Zhang et al., 2001c; Braz et al., 2003; Pikkarainen et al., 2004; see Chapters 9.2, 9.3, 9.5). The ability of these factors to activate gene expression is due in part to their ability to recruit

histone acetyltransferases (HATs) (Fig. 1). HATs provide these co-activators with the ability to modulate the “histone code” – a diverse array of post-translational modifications which together determine the higher-order structure of chromatin (Strahl and Allis 2000). Specifically, HATs transfer acetyl groups from acetyl coenzyme A to conserved lysine residues on histone tails. This modification results in a neutralization of the positive charge of the residue, leading to the local relaxation of intranucleosome and internucleosome interactions. In this less condensed state, chromatin is more accessible to transcriptional machinery, favoring gene expression. The most extensively-studied HATs in muscle are the closely-related co-­activators p300 and CREB-binding protein (CBP). The activity of p300 and CBP is increased in response to hypertrophic signaling pathways in the cardiac myocyte (Gusterson et al., 2002; Miyamoto et al., 2006). Also in the cardiac myocyte, a p300-responsive element in the skeletal -actin promoter was mapped to an MEF2-binding site, and a ternary complex containing this DNA element, MEF2, and p300 were demonstrated (Slepak et al., 2001). Moreover, ectopic overexpression of p300 and CBP stimulates, while dominant-negative mutants of p300 block, agonist-mediated cardiac growth in adult mice (Gusterson et al., 2002; Gusterson

Figure 1  Model of histone acetylation and deacetylation. HATs and HDACs are recruited to DNA by association with transcription factors (TFs). Acetylation of histone tails promotes chromatin relaxation and transcriptional activation. Deacetylation promotes chromatin condensation and transcriptional repression. Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

715

716

et al., 2003; Yanazume et al., 2003). The importance of p300 in cardiac development is illustrated by the phenotype of p300 knockout mice, which die between days 9 and 11.5 of gestation. These mice show reduced expression of muscle structural proteins such as alpha-MHC and alpha-actinin, as well as cardiac structural defects (Yao et al., 1998). The stimulatory effect of HATs on gene expression is countered by the HDACs. The HDACs promote chromatin compaction, and thereby repress gene expression. In the myocyte, HATs and HDACs coordinately act as links between signal transduction pathways and the DNA-­binding transcription factors that ultimately drive the gene programs of myogenic growth and differentiation. Eighteen human HDACs have been identified to date (Fig. 2). They fall into three classes based on their homology, with three structurally- and biochemically-distinct yeast HDACs. Class I HDACs (1, 2, 3, 8 and 11) are related to yeast RPD3, class II HDACs (4, 5, 6, 7, 9 and 10) to yeast HDA1, and class III HDACs (Sirt1–7) to yeast Sir2. Class II HDACs are further divided into two subclasses, IIa (HDACs 4, 5, 7 and 9) and IIb (HDACs 6 and 10). In the remainder of this chapter, the term “class II HDACs” will refer specifically to the class IIa HDACs. Members of this subset (HDACs 4, 5, 7 and 9) have been shown to repress MEF2-mediated gene expression and myocyte growth (McKinsey et al., 2002; Zhang et al., 2002b). In contrast, increasing evidence suggests class I HDACs may be involved in promoting cellular growth (Antos et al., 2003; Kook et al., 2003). The class III HDACs, which act as nicotinamide adenine dinucleotide (NAD)-dependent deacetylases, have been implicated in the inhibition of cardiac hypertrophy and enhanced cardiomyocyte survival (Alcendor et al., 2004). Class I HDACs are expressed ubiquitously, and are composed mainly of a catalytic domain (Grozinger and Schreiber, 2002). In contrast, class II HDACs show more restricted expression patterns, being enriched in heart, skeletal muscle and brain. Class II HDACs contain an amino acid N-­terminal extension of approximately 500 which mediates interactions with other transcriptional co-factors and confers responsiveness to calcium-dependent signaling (Grozinger et al., 1999; Miska et al., 1999; Lu et al., 2000a).

II.  Histone deacetylase–MEF2 interaction II.A.  Histone Deacetylases as Repressors of MEF2-Mediated Transcription A series of studies demonstrated that the class II HDACs control muscle growth and differentiation specifically through their associations with MEF2 and MyoD. The MEF2–HDAC interaction was initially identified by yeast two hybrid screens for MEF2-interacting proteins (Sparrow et al., 1999; Lu et al., 2000a). It is now clear that class II HDACs associate with MEF2 and act as potent

PART  |  10  Epigenetic Modifiers of Cardiac Development

Figure 2  Schematic representations of histone deacetylases. (A) HDACs are grouped into three classes – I, II and III – on the basis of their homology with three structurally- and biochemically-distinct yeast HDACs, Rpd3p, Hda1p and Sir2, respectively. Class IIa HDACs are expressed in a tissue-restricted manner (H: heart; B: brain; VE: vascular endothelium and endocardium; Ki: kidney; Skm: skeletal muscle; SmM: smooth muscle; Th: thymus; Pl: placenta; Lu: lung; Sp: spleen; Pa: pancreas; Ub: ubiquitous). (B) Schematic showing functional domains, phosphorylation sites and co-factor binding regions of human HDAC5 (NLS: nuclear localization signal; NES: nuclear export signal; HP1: heterochromatin protein 1; CtBP: C-terminal binding protein 1). Adapted from McKinsey et al. (2002).

inhibitors of MEF2-dependent transcription (Miska et al., 1999; Wang, 1999; Lemercier et al., 2000; Haberland et al., 2007; Chapter 9.5). Binding of these HDACs to MEF2 is mediated by 18 conserved amino acids in their amino­terminal extensions, whereas class I HDACs lack this domain and thus do not directly associate with MEF2. The class II HDACs bind sequences in MEF2 at the junction of the MADS–MEF2 domains, which mediate DNA-binding and dimerization (Lu et al., 2000b). However, association of HDAC with MEF2 does not appear to alter these properties

Chapter  |  10.2  Histone Deacetylases in Cardiovascular Development and Disease

significantly (Lu et al., 2000b), and the crystal structure of MEF2 bound to DNA is consistent with the formation of a ternary complex of MEF2, HDAC and MEF2-target genes (Santelli and Richmond, 2000). Further, the interaction of MEF2 with HATs and HDACs is mutually-exclusive (Sartorelli et al., 1997; McKinsey et al., 2001a). Thus, MEF2 acts as a platform to respond to positive or negative transcriptional signals by exchanging HATs and class II HDACs. Interestingly, the class II HDACs possess at least two separable repression domains, located at the amino- and ­carboxy-termini of the protein. The HDAC catalytic domain mediates the repressive activity of the carboxy-terminal region (Kao et al., 2000; Lu et al., 2000a). There is evidence to suggest that full-length class II HDACs lack intrinsic catalytic activity, instead deriving their deacetylase activity from the recruitment of class I HDACs (Fischle et al., 2002). The amino-terminal domain of class II HDACs does not possess catalytic activity, but instead represses transcription by recruiting other HDACs and co-repressors (Grozinger et al., 1999; Sparrow et al., 1999; Dressel et al., 2001; Zhang et al., 2001a or b). While the amino- and carboxy-terminal regions function independently as transcriptional repressors in transient promoter–reporter assays, the repressive activity of both domains appears to be required to inhibit endogenous MEF2 target genes (Lu et al., 2000b). MEF2-interacting transcription repressor (MITR), an HDAC9 splice variant, lacks a catalytic domain. Yet it is able to repress MEF2-dependent transcription and cardiac hypertrophy through the recruitment of other co-repressors (Sparrow et al., 1999; Zhou et al., 2000a; Zhang et al., 2001a or b, 2002a or b). As the class II HDACs do not bind DNA, they adopt target gene specificity by recruitment through MEF2 factors to MEF2 recognition sites. However, MEF2 is also capable of interacting with GATA and NFAT factors (Blaeser et al., 2000; Morin et al., 2000). Thus, class II HDACs may also repress elements of the developmental and hypertrophic cardiac gene program via recruitment by MEF2 to complexes assembled at GATA and NFAT response elements. Abnormal cardiac growth in HDAC-knockout animals correlates with superactivation of the MEF2 transcription factor (Zhang et al., 2002b), suggesting a relationship between the MEF2-HDAC interaction and the control of the hypertrophic response. These knockout animals are discussed in further detail in Section V.

II.B.  MEF2-Independent Functions for Class II Histone Deacetylases Class II HDACs are capable of associating with nuclear receptor co-repressor (N-CoR) and silencing mediator for retinoid and thyroid receptors (SMRT), suggesting they have MEF2-independent roles (Huang et al., 2000; Kao et al., 2000). Indeed, SMRT and N-CoR repress transcription by association with a vast array of transcription factors. The binding site for SMRT/N-CoR on class II HDACs overlaps

717

with their nuclear export sequence (Huang et al., 2000; Kao et al., 2000; McKinsey et al., 2001b). It has been shown that SMRT can drive HDAC4 from the cytoplasm to the nucleus, further suggesting a role for this interaction in the control of HDAC subcellular localization (Wu et al., 2001) (see also Section IV). SMRT has also been shown to undergo nuclear export in response to MEK-1 signaling (Hong and Privalsky, 2000), suggesting that CaMKindependent pathways may control class II HDAC activity indirectly via SMRT. In this regard, it has also been found that the ERK1/2 MAP kinases phosphorylate HDAC4, similarly promoting nuclear export (Zhou et al., 2000b). Thus, it is likely that class II HDACs are regulated both directly and indirectly by multiple signaling networks. Other MEF2-independent functions for class II HDACs include the control of cardiac gene expression through indirect interactions with stress-responsive transcription factors. These indirect interactions include HDAC5 and Nkx2.5 via calmodulin binding transcription activator 2 (CAMTA2) (Song et al., 2006), HDAC5 and SRF via myocardin (Cao et al., 2005; Xing et al., 2006), and HDAC4 and NFAT via mammalian relative of DnaJ (Mrj) (Dai et al., 2005).

III.  Class ii histone deacetylases as regulators of cardiac remodeling In response to stress signals that arise from a variety of cardiovascular disorders, including myocardial infarction and hypertension, the adult heart typically becomes enlarged due to cardiomyocyte hypertrophy. During this hypertrophic response, the individual myocytes increase in size without dividing and assemble additional sarcomeres to maximize force generation. While this response may provide initial compensatory advantages, such as the normalization of wall tension, prolonged hypertrophy in response to pathological signals is associated with increased morbidity and mortality due to both systolic and diastolic dysfunction (Levy et al., 1990). These observations are mirrored in animal models, in which prevention of the hypertrophic response ultimately maintains cardiac performance and enhances survival in conditions of hemodynamic stress (Frey and Olson, 2003).

III.A.  The Development–Hypertrophy Connection At the cell surface, humoral factors such as isoproterenol, angiotensin II and endothelin-1 trigger cardiac hypertrophy and remodeling by activating diverse downstream signaling pathways. These include pathways involving the calcium– calmodulin-dependent phosphatase calcineurin, CaMK and MAPKs (Zou et al., 2002; Molkentin, 2004). Many of these pathways converge on members of the MEF2, GATA and NFAT families of transcription factors that together ­control

718

fetal cardiac gene expression. Through these pathways, stress signaling in the myocardium results in upregulation of genes encoding embryonic isoforms of proteins that govern contractility, calcium handling and energetics, with a concomitant downregulation of adult isoforms. For example, stress signals enhance the expression of embryonic beta-myosin heavy chain (b-MyHC) and reduce expression of adult alpha-MyHC. This isoform switch results in diminished myofibrillar ATPase activity and impaired cardiac contractility (Mercadier et al., 1981; Whalen et al., 1981). This thick filament isoform switch is best-characterized in rodent models, but there is strong evidence that it also contributes to the progression of heart disease in humans. Specifically, phenotypic improvement in patients receiving beta-adrenergic antagonists (beta-blockers) directly correlates with an increase in adult alpha-MyHC and a decrease in fetal beta-MyHC (Abraham et al., 2002). The potential involvement of chromatin-remodeling in the regulation of MyHC expression was first suggested by the demonstration that nuclease hypersensitive sites appear in the a- and b-MyHC promoters in a spatio–temporal pattern that correlates with expression of each isoform (Huang and Liew, 1998). Enhancer mapping of the beta-MyHC gene revealed muscle-specific regulatory elements, including recognition sites for MEF2 and MyoD (Huang et al., 1997), which further suggests a potential involvement of HATs and HDACs in the regulation of contractile protein isoforms. The genes encoding atrial and brain natriuretic peptides (ANP and BNP) are also components of the fetal gene program that are upregulated in hypertrophic and failing hearts (Cameron and Ellmers, 2003). ANP and BNP bind to the natriuretic peptide receptor-A (NPR-A), which possesses intrinsic guanylyl cyclase activity and produces the second messenger cyclic guanosine monophosphate (cGMP) following peptide binding. NPR-A knockout mice spontaneously develop cardiac hypertrophy (Knowles et al., 2001; Holtwick et al., 2003), suggesting that natriuretic peptides might serve counter-regulatory functions to negatively control pathological cardiac signaling, perhaps through the activation of cGMP-dependent protein kinase (PKG) (Fiedler et al., 2002). ANP and BNP gene expression is repressed by neuron-restrictive silencer factor (NRSF), a repressor that recruits both class I and class II HDACs to repress the fetal cardiac gene program (Nakagawa et al., 2006). This study further demonstrated decreased interaction between NRSF and class II HDACs in both in vitro and in vivo cardiac hypertrophy models.

IV.  Signal-dependent regulation of class ii histone deacetylases Class II HDACs levels do not appear to change in the stressed myocardium (Zhang et al., 2002b; Chang et al., 2004). Instead, these HDACs are regulated by shuttling

PART  |  10  Epigenetic Modifiers of Cardiac Development

from the nucleus to the cytoplasm in response to stress signals, providing a post-translational mechanism to override HDAC-mediated repression of cardiac growth (Bush et al., 2004; Harrison et al., 2004; Vega et al., 2004a). This redistribution of HDACs enables MEF2 and other transcriptional activators and co-activators to associate with HATs, resulting in increased local histone acetylation and activation of downstream genes that promote cellular growth (Youn et al., 2000; Han et al., 2003). The nuclear-cytoplasmic translocation of class II HDACs is induced by phosphorylation of two conserved, serine-containing motifs found in the aminoterminal extensions (Grozinger and Schreiber, 2000; McKinsey et al., 2000a and b). When phosphorylated, these motifs associate with a chaperone protein, 14-3-3, which results in masking of the nuclear localization sequence located between the phosphorylation sites, and induces a conformational change that unmasks a nuclear export sequence at the C-terminus of the HDAC (Fig. 3) (McKinsey et al., 2001b; Wang and Wang, 2001). The nuclear export sequence is subsequently bound by the CRM1 nuclear export receptor, which facilitates translocation from the nucleus to the cytoplasm (Harrison et al., 2004). Induction of cardiac hypertrophy includes the post-translational activation of MEF2, which occurs in part as a consequence of the dissociation and nuclear export of class II HDACs (Lu et al., 2000a). Thus, signal-dependent HDAC export couples cellular signaling pathways to the MEF2-dependent activation of the fetal cardiac gene program (Fig. 4). Further insight was provided by the generation of ­signal-resistant class II HDACs. By mutating the conserved serines to alanines, HDAC phosphorylation is blocked, thereby inhibiting 14-3-3 docking and nuclear export. Such a signal-responsive, constitutively nuclear HDAC5 was shown to block cardiomyocyte hypertrophy induced by known hypertrophic agents (Zhang et al., 2002b).

V.  Histone deacetylase kinases Investigations thus far have identified three families of HDAC kinases, all of which belong to the Ca2–calmodulindependent protein kinase (CaMK) superfamily. The families include CaMKII, protein kinase D (PKD) and microtubule affinity-regulating kinase (MARK).

V.A.  Protein Kinase D The PKD family includes three highly-homologous PKD isoforms, all of which are able to phosphorylate all four class II HDACs (Chang et al., 2005; Dequiedt et al., 2005; Parra et al., 2005), indicating that the PKD family may exert redundant control over the class II HDACs. PKD is activated and translocates to the nucleus in response to G-proteincoupled receptor agonists, which include the hypertrophic agonists phenylephrine, angiotensin II and endothelin-1

Chapter  |  10.2  Histone Deacetylases in Cardiovascular Development and Disease

719

Figure 3  Model of the regulated nuclear–cytoplasmic shuttling of class II HDACs (HDAC5 shown). In the unphosphorylated state, HDAC5 is localized to the nucleus and associates with MEF2, resulting in repression of MEF2 target genes. Phosphorylation of conserved serines that flank the NLS of HDAC5 results in recruitment of 14-3-3 and dissociation from MEF2. Binding of 14-3-3 masks the nuclear localization signal and activates a cryptic nuclear export signal at the carboxyl terminus of HDAC5, resulting in nuclear export. This process is blocked by leptomycin B, which inhibits the CRM1 exportin protein. Re-entry of HDAC5 into the nucleus requires the action of a protein phosphatase (PPase). Adapted from McKinsey et al. (2001a or b).

(Rey et al., 2001; Auer et al., 2005; Harrison et al., 2006). Some of these agonists, such as phenylephrine, activate protein kinase C (PKC), which directly phosphorylates PKD. Others, such as endothelin-1, activate PKD through PKCindependent mechanisms (Wood et al., 2005; Harrison et al., 2006). In the heart, PKD activation occurs not only in response to the agonists previously mentioned, but also in response to chronic hypertension and pressure overload, such as that caused by aortic constriction. Activation of PKD in cultured cardiomyocytes causes the activation of the fetal cardiac genes, including ANF, BNP and alpha-skeletal actin (Vega et al., 2004a).

V.B.  CaMKII CaMKII is the most highly expressed CaMK family member in the heart (Edman and Schulman, 1994). CaMKII activity is low at basal states and is activated by Ca2–calmodulin complexes. Overexpression of calmodulin in mouse hearts causes pathological remodeling (Gruver et al., 1993), implicating CaMKII in the transduction of cardiac stress signals. In vitro assays, based on the expression of constitutively active CaMKI and IV in fibroblasts, clearly identify this kinase family as strong class II HDAC kinases (McKinsey et al., 2000a). However, studies in cultured cardiac myocytes

fail to clearly-establish a connection between endogenous CaMKI and IV activity and HDAC export. Moreover, ­pressure-overload hypertrophy selectively upregulates CaMKII, further implicating this CaMK family member in cardiac stress signaling (Colomer et al., 2003). Suprisingly, overexpression of a constitutively-active CaMKII is unable to drive HDAC5 nuclear export efficiently (Backs et al., 2006), and calcium signaling in myocytes that leads to CaMKII activation drives HDAC4, but not HDAC5, nuclear export (Liu et al., 2005). The inability of CaMKII to induce HDAC5 export is explained by the finding that HDAC4 contains a unique CaMKII docking site that is absent in HDACs 5 and 9 (Backs et al., 2006). This domain confers HDAC4 responsiveness to Ca2-­signaling relayed through CaMKII. In support of this, adrenergic agonist-dependent export of HDAC4, but not HDAC5, is sensitive to CaMK inhibitors. Together, these findings indicate that the class II HDACs can be differentially-exported in response to different signals, and may regulate different sets of genes (McKinsey, 2007).

V.C.  Mark Kinases Of the MARK kinase family members (MARK1, 2, 3 and 4), MARK1 and 2 are most abundant in the heart (Drewes, 2004; Tassan and Le Goff, 2004). MARK2 phosphorylates

720

PART  |  10  Epigenetic Modifiers of Cardiac Development

Figure 4  Model depicting kinase signaling to class II HDACs. Diverse agonists act through G-protein-coupled receptors to activate the PKC–PKD axis, CaMK, Rho and other effectors leading to phosphorylation and nuclear export of class II HDACs. Release of MEF2 from class II HDACs allows p300 to dock on MEF2 and stimulate genes involved in cardiac growth and remodeling. Adapted from Backs and Olson (2006).

Adrenergic Agonists Endothelin Angiotensin 5-Hydroxytryptamine Others

GPCRs

Other effectors PKC

14-3-3 14-3-3

CaMK

P

P

Class II HDACs PKD

Phosphatase

Class II HDACs CRM1 P

P

Class II HDACs 14-3-3

p300

Class II HDACs

Nucleus

MEF2

Other factors

Cardiac growth and remodeling genes

the same sites on HDAC5 that are targeted by PKD, and these sites are conserved among the other class II HDACs (Drewes et al., 1997). MARK3 is able to phosphorylate HDAC7 (Dequiedt et al., 2006), and appears to preferentially phosphorylate the amino-terminal 14-3-3-binding site. Despite these findings, there is no evidence yet that the MARK kinases mediate HDAC phosphorylation and export in cardiomyocytes (McKinsey et al., 2000a; Vega et al., 2004a). Thus, more studies are needed to determine if MARK kinases have a role in cardiac development or remodeling.

VI.  Histone deacetylase   knockout mice VI.A.  HDAC9 and HDAC5 Knockout Mice HDAC9 is the most abundantly-expressed HDAC in the myocardium, and the primary product of the HDAC9 locus, MITR, is a highly-effective suppressor of hypertrophy in vitro (Zhang et al., 2001a and b). The HDAC9 gene was inactivated in mice by homologous recombination in embryonic stem cells (Zhang et al., 2002b). Mice

Chapter  |  10.2  Histone Deacetylases in Cardiovascular Development and Disease

Cn-Tg x HDAC9–/–

Cn-Tg

HDAC9–/–

WT (A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

(I)

(J)

(K)

(L)

homozygous for the null allele are obtained in predicted Mendelian ratios, and show no pathology in early life. However, they display a dramatically-increased hypertrophic response to thoracic aortic banding, a model for cardiac pressure overload. HDAC9-nulls also show an exaggerated hypertrophic response to a known inducer of cardiac growth, the heart-specific calcineurin transgene (Fig. 5). This increased hypertrophy in response to calcineurin includes an exaggerated upregulation of ANF, BNP and beta-MyHC. Finally, even in the absence of such cardiovascular stresses, HDAC9-nulls develop cardiac hypertrophy by eight months of age. A transgenic mouse bearing an MEF2 consensus-binding site-driven LacZ cassette (Naya et al., 1999) revealed that HDAC9-nulls also displayed an exaggerated elevation in MEF2 activity in response to the calcineurin transgene. Together, these findings identify HDAC9 as a participant in hypertrophic signaling pathways in vivo. Targeted deletion of HDAC5 resulted in similar findings (Chang et al., 2004). HDAC5-nulls are viable, fertile and show no abnormalities early in life. Expression of LacZ from the targeted allele revealed that HDAC5 is strongly-expressed in the looping heart tube at embryonic

721

Figure 5  Hypersensitivity to calcineurinmediated hypertrophy in HDAC9 mutant mice. HDAC9 mutant mice were bred with mice harboring the alpha-MHC-calcineurin transgene (Cn-Tg). Hearts from one-month-old mice of the indicated genotype were isolated (A–D), sectioned, and stained with H & E (E–H). (I–L) Wild-type or HDAC9 mutant mice were crossed to transgenic mice harboring a MEF2dependent LacZ reporter gene (Naya et al., 1999). Offspring were bred with Cn-Tg mice. Cardiac MEF2 activity was detected by staining for beta-galactosidase activity in hearts from eight-week-old littermates of the indicated geno­types. Adapted from Zhang et al. (2002).

day 9.5 (E9.5). Expression becomes broader during later embryogenesis, appearing in the spinal cord and skeletal muscle. The adult expression pattern is broader, with staining present in the heart, lung, brain, skeletal muscle, liver and other tissues. They show similar age-dependent cardiac hypertrophy, and exaggerated responses to calcineurin signaling and pressure overload due to aortic constriction. Mice lacking both HDAC5 and HDAC9 show some embryonic or early perinatal lethality. Double-nulls that survive show cardiac hypertrophy by one month of age. Analysis of double-null embryos at E15.5 reveals ventricular septal defects, multifocal hemorrhages and thinning of the ventricular myocardium. Similar defects were also seen in a subset of the mice that survived to birth. Thus, HDAC5 and 9 both appear to share a redundant function in normal cardiovascular development.

VI.B.  HDAC7 Knockout Mouse Targeted deletion of HDAC7 results in embryonic lethality due to vascular dilation and rupture (Fig. 6) (Chang et al., 2006). This phenotype is totally penetrant, with all

722

PART  |  10  Epigenetic Modifiers of Cardiac Development

(D)

(G)

(H)

HDAC7loxp/–; tie2-cre

HDAC7�/�

(F)

HDAC7�/�

(E)

(C) HDAC7loxp/�

(B)

WT

(A)

Figure 6  Expression pattern of HDAC7 and vascular phenotype of HDAC7 mutant mice. (A, E) Wild-type and HDAC7 mutant embryos at E11.0. The HDAC7 mutant shows dramatically-enlarged dorsal aortas, pericardial effusion and hemorrhage. (B, F) Histological sections of HDAC7/ embryos at E11.0 prestained for lacZ. Shown are (B) the endocardium and (F) a small vessel in the head. LacZ staining is restricted to endothelial cells. (C, D, G, H) Conditional deletion of HDAC7. HDAC7loxP/ (C, D) and HDAC7flox/KO; tie2-cre (G, H) embryos at E11.5. Boxes show regions of the heart expanded in panels (D) and (H). Adapted from Chang et al. (2006).

embryos dying by E11.5 and showing the same defects. Consistent with the phenotype, HDAC7 is expressed specifically in vascular and endocardial endothelial cells. Gross examination of LacZ-staining shows expression in many organs, especially in heart, skeletal muscle, major vessels and lungs. Histological examination reveals that this broad expression is highly-specific for the vascular endothelial cells within all of these tissues, being excluded in vascular smooth muscle, cardiomyocytes and other organ parenchyma. Expression is absent in the yolk sac vasculature. In wild-type embryos, electron microscopy reveals normal tight junctions between adjacent endothelial cells of major vessels. However, such junctions are often missing in mutant embryos just prior to death. Endothelial cells appear to extend processes that fail to establish the tight endothelial cell–cell interactions which are essential for maintaining vascular integrity. In addition to the endothelial cell defects, mutants have a reduced number of smooth muscle cells surrounding their major vessels. The dorsal aortas show extreme dilation, and the myocardium is significantly thinner in both the atria and ventricles. Thus, the HDAC7-null phenotype includes nonendothelial cell developmental abnormalities. Likely explanations for the phenotypic features that are not restricted to the HDAC7-expressing cell type include aberrant paracrine signaling between endothelial cells and adjacent cell layers, or the misregulation of some other secreted factor or factors that influence adjacent cells or the surrounding matrix. The necessity of HDAC7 gene expression specifically in the endothelium is further confirmed by targeted, conditional gene deletion. The absence of HDAC7 in the endothelium recapitulates with complete penetrance the same phenotypic features as seen in the null mice. These embryos also die just prior to E11.5. In contrast, conditional deletion of HDAC7 in smooth muscle cells, ­cardiomyocytes

and skeletal muscle produces animals that are viable, fertile and without obvious pathology. Subsequent analysis reveals that loss of HDAC7 leads to an MEF2-dependent upregulation in matrix metalloproteinase 10 (MMP10) and downregulation of tissue inhibitor of metalloproteinase 1 (TIMP1) in vivo and in vitro. These findings suggest that the misregulation of secreted matrix remodeling enzymes is contributing to the loss of normal vascular development and maintenance of vascular integrity. In sum, HDAC7 is essential for normal cardiovascular development. Its critical role, at least in part, is to modulate MEF2 activity in the endocardium and in the major vessels. It remains to be determined if these observations are indicative of a role for HDAC7 in pathological processes in the adult vasculature.

VI.C.  HDAC4 Knockout Mouse HDAC4 is expressed in prehypertrophic chondrocytes of the developing skeleton, and determines the timing and extent of endochondral bone formation (Vega et al., 2004b). Mice homozygous for an HDAC4 mutation exhibit lethal ossification of endochondral cartilage due to ectopic hypertrophy of chondrocytes, whereas ectopic expression of HDAC4 in chondrocytes inhibits hypertrophic growth and differentiation. HDAC4 directly interacts with, and represses, runt-related transcription factor 2 (Runx2), a factor known to drive chondrocyte hypertrophy. Hypertrophic growth of different cell types depends on different cellular stimuli and different sets of transcription factors. The HDAC4-knockout phenotype suggests that even within these diverse cellular contexts, class II HDACs maintain their role as repressors of pathological or developmental hypertrophic gene programs.

Chapter  |  10.2  Histone Deacetylases in Cardiovascular Development and Disease

VI.D.  HDAC1 Knockout Phenotype Deletion of HDAC1 results in embryonic lethality by E10.5, due to severe developmental defects involving the head and allantois (Lagger et al., 2002). These morphological defects appear to result from impaired cellular proliferation. Analysis of embryonic stem cells from mutant embryos reveals increased levels of the cyclin-dependent kinase inhibitors p21WAF1/CIP1 and p27KIP1 secondary to the hyperacetylation of their promoters. These findings are consistent with the observation that the activation of tumor suppressors by HDAC inhibitors is integral to their anti-tumorigenetic activity (Kramer et al., 2001). HDAC1 is not expressed in the heart prior to E10.5, but it is expressed in the adult heart (Yang et al., 1997). Interestingly, the cyclin-dependent kinase inhibitor p21 is upregulated in human cells treated with HDAC inhibitors (Sowa et al., 1999; Richon et al., 2000), and several studies have implicated p21 as a repressor of pressure- and angiotensin-induced cardiac hypertrophy (Li and Brooks, 1997; Nozato et al., 2000). Therefore, HDAC1 inhibition in cardiac myocytes may upregulate p21, and thereby blocks agonist-induced hypertrophy (Backs et al., 2006). These findings may in part explain the observation that HDAC inhibitors prevent stress-induced cardiac hypertrophy.

VII.  Histone deacetylase inhibitors and therapeutics VII.A.  Histone Deacetylase Inhibitors Given the role of class II HDACs as repressors of the hypertrophic gene program, it would follow that pharmacological inhibitors of HDACs should stimulate cardiac hypertrophy. Paradoxically, treatment of cardiomyocytes with general antagonists of both class I and class II HDACs, such as trichostatin A and sodium butyrate, represses the increases in cell size and fetal gene expression that are normally evoked by hypertrophic agonists (Antos et al., 2003; McKinsey et al., 2004). It was also demonstrated that trichostatin A upregulates alpha-MHC expression in cultured cardiac myocytes, as well as in an in vivo model of hypothyroid rats, while it downregulates the expression of alpha- and beta-tubulins, and prevents their induction in response to angiotensin II (Davis et al., 2005). HDAC inhibitors, therefore, have the capacity to antagonize pathological gene expression that leads to the impairment of contractility (Kong et al., 2006). There are several possible explanations for the apparent conflict between the antigrowth effects of HDAC inhibitors and the role of class II HDACs in repressing prohypertrophy factors such as MEF2. First, MITR, the splicing variant of HDAC9 that lacks the catalytic domain, can repress MEF2 as efficiently as the full-length HDAC9 protein (Zhang et al., 2002b). This indicates that the deacetylase

723

domain is dispensable for the antigrowth effect of class II HDACs, and therefore might not be affected by HDAC inhibitors. Also, class II HDACs appear to lack intrinsic catalytic activity (Fischle et al., 2002) and the general HDAC inhibitors described here do not affect NAD-dependent class III HDACs (Imai et al., 2000). Thus, it has been proposed that distinct HDACs play positive or negative roles in the control of cardiac growth by regulating opposing sets of target genes via interactions with different sets of transcription factors, and it is the class I HDACs that are required for the repression of prohypertrophic genes (McKinsey et al., 2004). The simplest interpretation may be that class I HDAC activity is necessary to repress genes that encode antihypertrophic factors. Alternatively, class I HDACs may stimulate expression of progrowth genes. While HDACs are typically associated with gene repression, there are several studies in which HDACs have been linked to gene induction (Chang et al., 2004; Zupkovitz et al., 2006). Finally, class I HDACs may act through the deacetylation of nonhistone targets (Westermann and Weber, 2003; Yanazume et al., 2003), altering signal transduction or the cytoskeleton. HDAC inhibitors might then oppose hypertrophic signaling in part by affecting the acetylation state of these nonhistone targets.

VII.B.  Perspectives on Therapeutics Many of the same transcription factors that are crucial for embryonic cardiac gene expression and development, including MEF2, SRF and GATA, act as endpoints for stressresponsive hypertrophic signaling pathways. Activation of the fetal gene program with repression of corresponding adult cardiac genes ultimately leads to cardiac failure (Lowes et al., 2002). Thus, tremendous effort has been devoted to the identification of the mechanisms coupling cardiac stress signaling pathways to the fetal gene program. Studies have demonstrated that both genetic and pharmacological blockade of these signaling pathways in the context of chronic cardiac stress preserves cardiac function (Koch et al., 1995; Rothermel et al., 2001; Antos et al., 2002). General inhibitors of HDACs are tolerated well in vivo, and are currently being tested in clinical trials as anticancer drugs (Garcia-Manero and Issa, 2005). Gain- and loss-of-function experiments in cardiomyocytes and mouse models will help in focusing efforts to develop isoformspecific HDAC inhibitors. New high-throughput screens and medicinal chemistry strategies are beginning to identify small molecules with narrow target specificities (Arts et al., 2003; Haggarty et al., 2003; Hu et al., 2003). Such discoveries should lead to novel therapeutic strategies that enable the treatment of pathological cardiac remodeling, while minimizing unpredicted clinical side-effects. The identification of signaling pathways that ­ control cardiac gene transcription has provided new ­ opportunities

724

for drug discovery (Bush et al., 2004; Wu et al., 2004). Understanding the mechanisms and patterns of gene regulation in the developing heart has proven to be a key to identifying molecular signals that cause pathological remodeling in the stressed heart. Acting at the interface between diverse signaling pathways and the genome, the HDACs occupy a unique position in the control of both pathological and developmental gene expression. Further insight into the biological roles of the HDACs will continue to reveal new possibilities for therapeutic intervention.

References Abraham, W.T., Gilbert, E.M., Lowes, B.D., Minobe, W.A., Larrabee, P., Roden, R.L., Dutcher, D., Sederberg, J., Lindenfeld, J.A., Wolfel, E.E., Shakar, S.F., Ferguson, D., Volkman, K., Linseman, J.V., Quaife, R.A., Robertson, A.D., Bristow, M.R., 2002. Coordinate changes in myosin heavy chain isoform gene expression are selectively associated with alterations in dilated cardiomyopathy phenotype. Mol. Med. 8 (11), 750–760. Alcendor, R.R., Kirshenbaum, L.A., Imai, S., Vatner, S.F., Sadoshima, J., 2004. Silent information regulator 2alpha, a longevity factor and class III histone deacetylase, is an essential endogenous apoptosis inhibitor in cardiac myocytes. Circ. Res. 95 (10), 971–980. Antos, C.L., McKinsey, T.A., Frey, N., Kutschke, W., McAnally, J., Shelton, J.M., Richardson, J.A., Hill, J.A., Olson, E.N., 2002. Activated glycogen synthase-3 beta suppresses cardiac hypertrophy in vivo. Proc. Natl. Acad. Sci. USA 99 (2), 907–912. Antos, C.L., McKinsey, T.A., Dreitz, M., Hollingsworth, L.M., Zhang, C.L., Schreiber, K., Rindt, H., Gorczynski, R.J., Olson, E.N., 2003. Dose-dependent blockade to cardiomyocyte hypertrophy by histone deacetylase inhibitors. J. Biol. Chem. 278 (31), 28930–28937. Arts, J., de Schepper, S., Van Emelen, K., 2003. Histone deacetylase inhibitors, from chromatin remodeling to experimental cancer therapeutics. Curr. Med. Chem. 10 (22), 2343–2350. Auer, A., von Blume, J., Sturany, S., von Wichert, G., Van Lint, J., Vandenheede, J., Adler, G., Seufferlein, T., 2005. Role of the regulatory domain of protein kinase D2 in phorbol ester binding, catalytic activity, and nucleocytoplasmic shuttling. Mol. Biol. Cell 16 (9), 4375–4385. Backs, J., Olson, E.N., 2006. Control of cardiac growth by histone acetylation/deacetylation. Circ. Res. 98 (1), 15–24. Backs, J., Song, K., Bezprozvannaya, S., Chang, S., Olson, E.N., 2006. CaM kinase II selectively signals to histone deacetylase 4 during cardiomyocyte hypertrophy. J. Clin. Invest. 116 (7), 1853–1864. Blaeser, F., Ho, N., Prywes, R., Chatila, T.A., 2000. Ca(2)-dependent gene expression mediated by MEF2 transcription factors. J. Biol. Chem. 275 (1), 197–209. Braz, J.C., Bueno, O.F., Liang, Q., Wilkins, B.J., Dai, Y.S., Parsons, S., Braunwart, J., Glascock, B.J., Klevitsky, R., Kimball, T.F., Hewett, T.E., Molkentin, J.D., 2003. Targeted inhibition of p38 MAPK promotes hypertrophic cardiomyopathy through upregulation of calcineurin-NFAT signaling. J. Clin. Invest. 111 (10), 1475–1486. Bush, E., Fielitz, J., Melvin, L., Martinez-Arnold, M., McKinsey, T.A., Plichta, R., Olson, E.N., 2004. A small molecular activator of cardiac hypertrophy uncovered in a chemical screen for modifiers of the calcineurin signaling pathway. Proc. Natl. Acad. Sci. USA 101 (9), 2870–2875.

PART  |  10  Epigenetic Modifiers of Cardiac Development

Cameron, V.A., Ellmers, L.J., 2003. Mini review: Natriuretic peptides during development of the fetal heart and circulation. Endocrinology 144 (6), 2191–2194. Cao, D., Wang, Z., Zhang, C.L., Oh, J., Xing, W., Li, S., Richardson, J.A., Wang, D.Z., Olson, E.N., 2005. Modulation of smooth muscle gene expression by association of histone acetyltransferases and deacetylases with myocardin. Mol. Cell Biol. 25 (1), 364–376. Chang, S., McKinsey, T.A., Zhang, C.L., Richardson, J.A., Hill, J.A., Olson, E.N., 2004. Histone deacetylases 5 and 9 govern responsiveness of the heart to a subset of stress signals and play redundant roles in heart development. Mol. Cell Biol. 24 (19), 8467–8476. Chang, S., Bezprozvannaya, S., Li, S., Olson, E.N., 2005. An expression screen reveals modulators of class II histone deacetylase phosphorylation. Proc. Natl. Acad. Sci. USA 102 (23), 8120–8125. Chang, S., Young, B.D., Li, S., Qi, X., Richardson, J.A., Olson, E.N., 2006. Histone deacetylase 7 maintains vascular integrity by repressing matrix metalloproteinase 10. Cell 126 (2), 321–334. Colomer, J.M., Mao, L., Rockman, H.A., Means, A.R., 2003. Pressure overload selectively up-regulates Ca2/calmodulin-dependent protein kinase II in vivo. Mol. Endocrinol. 17 (2), 183–192. Dai, Y.S., Xu, J., Molkentin, J.D., 2005. The DnaJ-related factor Mrj interacts with nuclear factor of activated T cells c3 and mediates transcriptional repression through class II histone deacetylase recruitment. Mol. Cell Biol. 25 (22), 9936–9948. Davis, F.J., Pillai, J.B., Gupta, M., Gupta, M.P., 2005. Concurrent opposite effects of trichostatin A, an inhibitor of histone deacetylases, on expression of alpha-MHC and cardiac tubulins, implication for gain in cardiac muscle contractility. Am. J. Physiol. Heart Circ. Physiol. 288 (3), H1477–H1490. Dequiedt, F., Van Lint, J., Lecomte, E., Van Duppen, V., Seufferlein, T., Vandenheede, J.R., Wattiez, R., Kettmann, R., 2005. Phosphorylation of histone deacetylase 7 by protein kinase D mediates T cell receptorinduced Nur77 expression and apoptosis. J. Exp. Med. 201 (5), 793–804. Dressel, U., Bailey, P.J., Wang, S.C., Downes, M., Evans, R.M., Muscat, G.E., 2001. A dynamic role for HDAC7 in MEF2-mediated muscle differentiation. J. Biol. Chem. 276 (20), 17007–17013. Drewes, G., 2004. MARKing tau for tangles and toxicity. Trends Biochem Sci. 29 (10), 548–555. Drewes, G., Ebneth, A., Preuss, U., Mandelkow, E.M., Mandelkow, E., 1997. MARK, a novel family of protein kinases that phosphorylate microtubule-associated proteins and trigger microtubule disruption. Cell 89 (2), 297–308. Edman, C.F., Schulman, H., 1994. Identification and characterization of delta B-CaM kinase and delta C-CaM kinase from rat heart, two new multifunctional Ca2/calmodulin-dependent protein kinase isoforms. Biochim. Biophys. Acta. 1221 (1), 89–101. Fiedler, B., Lohmann, S.M., Smolenski, A., Linnemuller, S., Pieske, B., Schroder, F., Molkentin, J.D., Drexler, H., Wollert, K.C., 2002. Inhibition of calcineurin-NFAT hypertrophy signaling by cGMP-dependent protein kinase type I in cardiac myocytes. Proc. Natl. Acad. Sci. USA 99 (17), 11363–11368. Fischle, W., Dequiedt, F., Hendzel, M.J., Guenther, M.G., Lazar, M.A., Voelter, W., Verdin, E., 2002. Enzymatic activity associated with class II HDACs is dependent on a multiprotein complex containing HDAC3 and SMRT/N-CoR. Mol. Cell 9 (1), 45–57. Frey, N., Olson, E.N., 2003. Cardiac hypertrophy, the good, the bad, and the ugly. Annu. Rev. Physiol. 65, 45–79. Garcia-Manero, G., Issa, J.P., 2005. Histone deacetylase inhibitors, a review of their clinical status as antineoplastic agents. Cancer Invest. 23 (7), 635–642.

Chapter  |  10.2  Histone Deacetylases in Cardiovascular Development and Disease

Grozinger, C.M., Hassig, C.A., Schreiber, S.L., 1999. Three proteins define a class of human histone deacetylases related to yeast Hda1p. Proc. Natl. Acad. Sci. USA 96 (9), 4868–4873. Grozinger, C.M., Schreiber, S.L., 2000. Regulation of histone deacetyl­ ase 4 and 5 and transcriptional activity by 14-3-3-dependent cellular localization. Proc. Natl. Acad. Sci. USA 97 (14), 7835–7840. Grozinger, C.M., Schreiber, S.L., 2002. Deacetylase enzymes, biological functions and the use of small-molecule inhibitors. Chem. Biol. 9 (1), 3–16. Gruver, C.L., DeMayo, F., Goldstein, M.A., Means, A.R., 1993. Targeted developmental overexpression of calmodulin induces proliferative and hypertrophic growth of cardiomyocytes in transgenic mice. Endocrinology 133 (1), 376–388. Gusterson, R., Brar, B., Faulkes, D., Giordano, A., Chrivia, J., Latchman, D., 2002. The transcriptional co-activators CBP and p300 are activated via phenylephrine through the p42/p44 MAPK cascade. J. Biol. Chem. 277 (4), 2517–2524. Gusterson, R.J., Jazrawi, E., Adcock, I.M., Latchman, D.S., 2003. The transcriptional co-activators CREB-binding protein (CBP) and p300 play a critical role in cardiac hypertrophy that is dependent on their histone acetyltransferase activity. J. Biol. Chem. 278 (9), 6838–6847. Haberland, M., Arnold, M.A., McAnally, J., Phan, D., Kim, Y., Olson, E.N., 2007. Regulation of HDAC9 gene expression by MEF2 establishes a negative-feedback loop in the transcriptional circuitry of muscle differentiation. Mol. Cell Biol. 27 (2), 518–525. Haggarty, S.J., Koeller, K.M., Wong, J.C., Butcher, R.A., Schreiber, S.L., 2003. Multidimensional chemical genetic analysis of diversityoriented synthesis-derived deacetylase inhibitors using cell-based assays. Chem. Biol. 10 (5), 383–396. Han, A., Pan, F., Stroud, J.C., Youn, H.D., Liu, J.O., Chen, L., 2003. Sequence-specific recruitment of transcriptional co-repressor Cabin1 by myocyte enhancer factor-2. Nature 422 (6933), 730–734. Harrison, B.C., Roberts, C.R., Hood, D.B., Sweeney, M., Gould, J.M., Bush, E.W., McKinsey, T.A., 2004. The CRM1 nuclear export receptor controls pathological cardiac gene expression. Mol. Cell Biol. 24 (24), 10636–10649. Harrison, B.C., Kim, M.S., van Rooij, E., Plato, C.F., Papst, P.J., Vega, R.B., McAnally, J.A., Richardson, J.A., Bassel-Duby, R., Olson, E.N., McKinsey, T.A., 2006. Regulation of cardiac stress signaling by protein kinase d1. Mol. Cell Biol. 26 (10), 3875–3888. Holtwick, R., van Eickels, M., Skryabin, B.V., Baba, H.A., Bubikat, A., Begrow, F., Schneider, M.D., Garbers, D.L., Kuhn, M., 2003. Pressureindependent cardiac hypertrophy in mice with cardiomyocyterestricted inactivation of the atrial natriuretic peptide receptor guanylyl cyclase-A. J. Clin. Invest. 111 (9), 1399–1407. Hong, S.H., Privalsky, M.L., 2000. The SMRT corepressor is regulated by a MEK-1 kinase pathway, inhibition of corepressor function is associated with SMRT phosphorylation and nuclear export. Mol. Cell Biol. 20 (17), 6612–6625. Hu, E., Dul, E., Sung, C.M., Chen, Z., Kirkpatrick, R., Zhang, G.F., Johanson, K., Liu, R., Lago, A., Hofmann, G., Macarron, R., de los Frailes, M., Perez, P., Krawiec, J., Winkler, J., Jaye, M., 2003. Identification of novel isoform-selective inhibitors within class I histone deacetylases. J. Pharmacol. Exp. Ther. 307 (2), 720–728. Huang, E.Y., Zhang, J., Miska, E.A., Guenther, M.G., Kouzarides, T., Lazar, M.A., 2000. Nuclear receptor corepressors partner with class II histone deacetylases in a Sin3-independent repression pathway. Genes Dev. 14 (1), 45–54. Huang, W.Y., Chen, J.J., Shih, N., Liew, C.C., 1997. Multiple musclespecific regulatory elements are associated with a DNase I

725

h­ ypersensitive site of the cardiac beta-myosin heavy-chain gene. Biochem. J. 327 (Pt 2), 507–512. Huang, W.Y., Liew, C.C., 1998. Chromatin remodelling of the cardiac beta-myosin heavy chain gene. Biochem. J. 330 (Pt 2), 871–876. Imai, S., Armstrong, C.M., Kaeberlein, M., Guarente, L., 2000. Tran­s­criptional silencing and longevity protein Sir2 is an NAD-dependent histone deacetylase. Nature 403 (6771), 795–800. Kao, H.Y., Downes, M., Ordentlich, P., Evans, R.M., 2000. Isolation of a novel histone deacetylase reveals that class I and class II deacetylases promote SMRT-mediated repression. Genes Dev. 14 (1), 55–66. Knowles, J.W., Esposito, G., Mao, L., Hagaman, J.R., Fox, J.E., Smithies, O., Rockman, H.A., Maeda, N., 2001. Pressure-independent enhancement of cardiac hypertrophy in natriuretic peptide receptor A-­deficient mice. J. Clin. Invest. 107 (8), 975–984. Koch, W.J., Rockman, H.A., Samama, P., Hamilton, R.A., Bond, R.A., Milano, C.A., Lefkowitz, R.J., 1995. Cardiac function in mice overexpressing the beta-adrenergic receptor kinase or a beta ARK inhibitor. Science 268 (5215), 1350–1353. Kong, Y., Tannous, P., Lu, G., Berenji, K., Rothermel, B.A., Olson, E.N., Hill, J.A., 2006. Suppression of class I and II histone deacetylases blunts pressure-overload cardiac hypertrophy. Circulation 113 (22), 2579–2588. Kook, H., Lepore, J.J., Gitler, A.D., Lu, M.M., Wing-Man Yung, W., Mackay, J., Zhou, R., Ferrari, V., Gruber, P., Epstein, J.A., 2003. Cardiac hypertrophy and histone deacetylase-dependent transcriptional repression mediated by the atypical homeodomain protein Hop. J. Clin. Invest. 112 (6), 863–871. Kramer, O.H., Gottlicher, M., Heinzel, T., 2001. Histone deacetylase as a therapeutic target. Trends. Endocrinol. Metab. 12 (7), 294–300. Lagger, G., O’Carroll, D., Rembold, M., Khier, H., Tischler, J., Weitzer, G., Schuettengruber, B., Hauser, C., Brunmeir, R., Jenuwein, T., Seiser, C., 2002. Essential function of histone deacetylase 1 in proliferation control and CDK inhibitor repression. Embo. J. 21 (11), 2672–2681. Lemercier, C., Verdel, A., Galloo, B., Curtet, S., Brocard, M.P., Khochbin, S., 2000. MHDA1/HDAC5 histone deacetylase interacts with and represses MEF2A transcriptional activity. J. Biol. Chem. 275 (20), 15594–15599. Levy, D., Garrison, R.J., Savage, D.D., Kannel, W.B., Castelli, W.P., 1990. Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N. Engl. J. Med. 322 (22), 1561–1566. Li, J.M., Brooks, G., 1997. Downregulation of cyclin-dependent kinase inhibitors p21 and p27 in pressure-overload hypertrophy. Am. J. Physiol. 273 (Pt 2), H1358–H1367. Liu, Y., Randall, W.R., Schneider, M.F., 2005. Activity-dependent and independent nuclear fluxes of HDAC4 mediated by different kinases in adult skeletal muscle. J. Cell Biol. 168 (6), 887–897. Lowes, B.D., Gilbert, E.M., Abraham, W.T., Minobe, W.A., Larrabee, P., Ferguson, D., Wolfel, E.E., Lindenfeld, J., Tsvetkova, T., Robertson, A.D., Quaife, R.A., Bristow, M.R., 2002. Myocardial gene expression in dilated cardiomyopathy treated with beta-blocking agents. N. Engl. J. Med. 346 (18), 1357–1365. Lu, J., McKinsey, T.A., Nicol, R.L., Olson, E.N., 2000a. Signal­dependent activation of the MEF2 transcription factor by dissociation from histone deacetylases. Proc. Natl. Acad. Sci. USA 97 (8), 4070–4075. Lu, J., McKinsey, T.A., Zhang, C.L., Olson, E.N., 2000b. Regulation of skeletal myogenesis by association of the MEF2 transcription factor with class II histone deacetylases. Mol. Cell 6 (2), 233–244.

726

McKinsey, T.A., 2007. Derepression of pathological cardiac genes by members of the CaM kinase superfamily. Cardiovasc. Res. 73 (4), 667–677. McKinsey, T.A., Zhang, C.L., Lu, J., Olson, E.N., 2000a. Signal­dependent nuclear export of a histone deacetylase regulates muscle differentiation. Nature 408 (6808), 106–111. McKinsey, T.A., Zhang, C.L., Olson, E.N., 2000b. Activation of the myocyte enhancer factor-2 transcription factor by calcium/calmodulin-dependent protein kinase-stimulated binding of 14-3-3 to histone deacetylase 5. Proc. Natl. Acad. Sci. USA 97 (26), 14400–14405. McKinsey, T.A., Zhang, C.L., Olson, E.N., 2001a. Control of muscle development by dueling HATs and HDACs. Curr. Opin. Genet. Dev. 11 (5), 497–504. McKinsey, T.A., Zhang, C.L., Olson, E.N., 2001b. Identification of a ­signal-responsive nuclear export sequence in class II histone deacetyl­ ases. Mol. Cell. Biol. 21 (18), 6312–6321. McKinsey, T.A., Zhang, C.L., Olson, E.N., 2002. Signaling chromatin to make muscle. Curr. Opin. Cell Biol. 14 (6), 763–772. McKinsey, T.A., Olson, E.N., 2004. Cardiac histone acetylation – ­therapeutic opportunities abound. Trends. Genet. 20 (4), 206–213. Mercadier, J.J., Lompre, A.M., Wisnewsky, C., Samuel, J.L., Bercovici, J., Swynghedauw, B., Schwartz, K., 1981. Myosin isoenzyme changes in several models of rat cardiac hypertrophy. Circ. Res. 49 (2), 525–532. Miska, E.A., Karlsson, C., Langley, E., Nielsen, S.J., Pines, J., Kouzarides, T., 1999. HDAC4 deacetylase associates with and represses the MEF2 transcription factor. Embo. J. 18 (18), 5099–5107. Miyamoto, S., Kawamura, T., Morimoto, T., Ono, K., Wada, H., Kawase, Y., Matsumori, A., Nishio, R., Kita, T., Hasegawa, K., 2006. Histone acetyltransferase activity of p300 is required for the promotion of left ventricular remodeling after myocardial infarction in adult mice in vivo. Circulation 113 (5), 679–690. Molkentin, J.D., 2004. Calcineurin-NFAT signaling regulates the cardiac hypertrophic response in coordination with the MAPKs. Cardiovasc. Res. 63 (3), 467–475. Morin, S., Charron, F., Robitaille, L., Nemer, M., 2000. GATA-dependent recruitment of MEF2 proteins to target promoters. Embo. J. 19 (9), 2046–2055. Nakagawa, Y., Kuwahara, K., Harada, M., Takahashi, N., Yasuno, S., Adachi, Y., Kawakami, R., Nakanishi, M., Tanimoto, K., Usami, S., Kinoshita, H., Saito, Y., Nakao, K., 2006. Class II HDACs mediate CaMK-dependent signaling to NRSF in ventricular myocytes. J. Mol. Cell Cardiol. 41 (6), 1010–1022. Naya, F.J., Wu, C., Richardson, J.A., Overbeek, P., Olson, E.N., 1999. Transcriptional activity of MEF2 during mouse embryogenesis monitored with a MEF2-dependent transgene. Development 126 (10), 2045–2052. Nozato, T., Ito, H., Tamamori, M., Adachi, S., Abe, S., Marumo, F., Hiroe, M., 2000. G1 cyclins are involved in the mechanism of cardiac myocyte hypertrophy induced by angiotensin II. Jpn. Circ. J. 64 (8), 595–601. Parra, M., Kasler, H., McKinsey, T.A., Olson, E.N., Verdin, E., 2005. Protein kinase D1 phosphorylates HDAC7 and induces its nuclear export after T-cell receptor activation. J. Biol. Chem. 280 (14), 13762–13770. Pikkarainen, S., Tokola, H., Kerkela, R., Ruskoaho, H., 2004. GATA transcription factors in the developing and adult heart. Cardiovasc. Res. 63 (2), 196–207. Rey, O., Young, S.H., Cantrell, D., Rozengurt, E., 2001. Rapid protein kinase D translocation in response to G protein-coupled receptor

PART  |  10  Epigenetic Modifiers of Cardiac Development

activation. Dependence on protein kinase C. J. Biol. Chem. 276 (35), 32616–32626. Richon, V.M., Sandhoff, T.W., Rifkind, R.A., Marks, P.A., 2000. Histone deacetylase inhibitor selectively induces p21WAF1 expression and gene-associated histone acetylation. Proc. Natl. Acad. Sci. USA 97 (18), 10014–10019. Rothermel, B.A., McKinsey, T.A., Vega, R.B., Nicol, R.L., Mammen, P., Yang, J., Antos, C.L., Shelton, J.M., Bassel-Duby, R., Olson, E.N., Williams, R.S., 2001. Myocyte-enriched calcineurin-interacting protein, MCIP1, inhibits cardiac hypertrophy in vivo. Proc. Natl. Acad. Sci. USA 98 (6), 3328–3333. Santelli, E., Richmond, T.J., 2000. Crystal structure of MEF2A core bound to DNA at 1.5 A resolution. J. Mol. Biol. 297 (2), 437–449. Sartorelli, V., Huang, J., Hamamori, Y., Kedes, L., 1997. Molecular mechanisms of myogenic coactivation by p300, direct interaction with the activation domain of MyoD and with the MADS box of MEF2C. Mol. Cell. Biol. 17 (2), 1010–1026. Slepak, T.I., Webster, K.A., Zang, J., Prentice, H., O’Dowd, A., Hicks, M.N., Bishopric, N.H., 2001. Control of cardiac-specific transcription by p300 through myocyte enhancer factor-2D. J. Biol. Chem. 276 (10), 7575–7585. Song, K., Backs, J., McAnally, J., Qi, X., Gerard, R.D., Richardson, J.A., Hill, J.A., Bassel-Duby, R., Olson, E.N., 2006. The transcriptional coactivator CAMTA2 stimulates cardiac growth by opposing class II histone deacetylases. Cell 125 (3), 453–466. Sowa, Y., Orita, T., Hiranabe-Minamikawa, S., Nakano, K., Mizuno, T., Nomura, H., Sakai, T., 1999. Histone deacetylase inhibitor activates the p21/WAF1/Cip1 gene promoter through the Sp1 sites. Ann. N. Y. Acad. Sci. 886, 195–199. Sparrow, D.B., Miska, E.A., Langley, E., Reynaud-Deonauth, S., Kotecha, S., Towers, N., Spohr, G., Kouzarides, T., Mohun, T.J., 1999. MEF-2 function is modified by a novel co-repressor, MITR. Embo. J. 18 (18), 5085–5098. Strahl, B.D., Allis, C.D., 2000. The language of covalent histone modifications. Nature 403 (6765), 41–45. Tassan, J.P., Le Goff, X., 2004. An overview of the KIN1/PAR-1/MARK kinase family. Biol. Cell 96 (3), 193–199. Vega, R.B., Harrison, B.C., Meadows, E., Roberts, C.R., Papst, P.J., Olson, E.N., McKinsey, T.A., 2004a. Protein kinases C and D mediate agonist-dependent cardiac hypertrophy through nuclear export of histone deacetylase 5. Mol. Cell Biol. 24 (19), 8374–8385. Vega, R.B., Matsuda, K., Oh, J., Barbosa, A.C., Yang, X., Meadows, E., McAnally, J., Pomajzl, C., Shelton, J.M., Richardson, J.A., Karsenty, G., Olson, E.N., 2004b. Histone deacetylase 4 controls chondrocyte hypertrophy during skeletogenesis. Cell 119 (4), 555–566. Wang, A.H., Bertos, N.R., Vezmar, M., Pelletier, N., Crosato, M., Heng, H.H., Th’ng, J., Han, J., Yang, X.J., 1999. HDAC4, a human histone deacetylase related to yeast HDA1, is a transcriptional corepressor. Mol. Cell Biol. 19 (11), 7816–7827. Wang, A.H., Yang, X.J., 2001. Histone deacetylase 4 possesses intrinsic nuclear import and export signals. Mol. Cell Biol. 21 (17), 5992–6005. Westermann, S., Weber, K., 2003. Post-translational modifications regulate microtubule function. Nat. Rev. Mol. Cell Biol. 4 (12), 938–947. Whalen, R.G., Sell, S.M., Butler-Browne, G.S., Schwartz, K., Bouveret, P., Pinset-Harstom, I., 1981. Three myosin heavy-chain isozymes appear sequentially in rat muscle development. Nature 292 (5826), 805–809. Wood, C.D., Marklund, U., Cantrell, D.A., 2005. Dual phospholipase C/diacylglycerol requirement for protein kinase D1 activation in ­lymphocytes. J. Biol. Chem. 280 (7), 6245–6251.

Chapter  |  10.2  Histone Deacetylases in Cardiovascular Development and Disease

Wu, X., Li, H., Park, E.J., Chen, J.D., 2001. SMRTE inhibits MEF2C transcriptional activation by targeting HDAC4 and 5 to nuclear domains. J. Biol. Chem. 276 (26), 24177–24185. Wu, X., Ding, S., Ding, Q., Gray, N.S., Schultz, P.G., 2004. Small molecules that induce cardiomyogenesis in embryonic stem cells. J. Am. Chem. Soc. 126 (6), 1590–1591. Xing, W., Zhang, T.C., Cao, D., Wang, Z., Antos, C.L., Li, S., Wang, Y., Olson, E.N., Wang, D.Z., 2006. Myocardin induces cardiomyocyte hypertrophy. Circ. Res. 98 (8), 1089–1097. Yanazume, T., Hasegawa, K., Morimoto, T., Kawamura, T., Wada, H., Matsumori, A., Kawase, Y., Hirai, M., Kita, T., 2003. Cardiac p300 is involved in myocyte growth with decompensated heart failure. Mol. Cell Biol. 23 (10), 3593–3606. Yang, W.M., Yao, Y.L., Sun, J.M., Davie, J.R., Seto, E., 1997. Isolation and characterization of cDNAs corresponding to an additional member of the human histone deacetylase gene family. J. Biol. Chem. 272 (44), 28001–28007. Yao, T.P., Oh, S.P., Fuchs, M., Zhou, N.D., Ch’ng, L.E., Newsome, D., Bronson, R.T., Li, E., Livingston, D.M., Eckner, R., 1998. Gene dosage-dependent embryonic development and proliferation defects in mice lacking the transcriptional integrator p300. Cell 93 (3), 361–372. Youn, H.D., Grozinger, C.M., Liu, J.O., 2000. Calcium regulates transcriptional repression of myocyte enhancer factor 2 by histone deacetylase 4. J. Biol. Chem. 275 (29), 22563–22567. Zhang, C.L., McKinsey, T.A., Lu, J.R., Olson, E.N., 2001a. Association of COOH-terminal-binding protein (CtBP) and MEF2-interacting transcription repressor (MITR) contributes to transcriptional repression of the MEF2 transcription factor. J. Biol. Chem. 276 (1), 35–39.

727

Zhang, C.L., McKinsey, T.A., Olson, E.N., 2001b. The transcriptional corepressor MITR is a signal-responsive inhibitor of myogenesis. Proc. Natl. Acad. Sci. USA 98 (13), 7354–7359. Zhang, C.L., McKinsey, T.A., Olson, E.N., 2002a. Association of class II histone deacetylases with heterochromatin protein 1, potential role for histone methylation in control of muscle differentiation. Mol. Cell Biol. 22 (20), 7302–7312. Zhang, C.L., McKinsey, T.A., Chang, S., Antos, C.L., Hill, J.A., Olson, E.N., 2002b. Class II histone deacetylases act as signalresponsive repressors of cardiac hypertrophy. Cell 110 (4), 479–488. Zhang, X., Azhar, G., Chai, J., Sheridan, P., Nagano, K., Brown, T., Yang, J., Khrapko, K., Borras, A.M., Lawitts, J., Misra, R.P., Wei, J.Y., 2001c. Cardiomyopathy in transgenic mice with cardiacspecific overexpression of serum response factor. Am. J. Physiol. Heart Circ. Physiol. 280 (4), H1782–H1792. Zhou, X., Richon, V.M., Rifkind, R.A., Marks, P.A., 2000a. Identification of a transcriptional repressor related to the noncatalytic domain of histone deacetylases 4 and 5. Proc. Natl. Acad Sci. USA 97 (3), 1056–1061. Zhou, X., Richon, V.M., Wang, A.H., Yang, X.J., Rifkind, R.A., Marks, P.A., 2000b. Histone deacetylase 4 associates with extracellular signalregulated kinases 1 and 2, and its cellular localization is regulated by oncogenic Ras. Proc. Natl. Acad Sci. USA 97 (26), 14329–14333. Zou, Y., Takano, H., Akazawa, H., Nagai, T., Mizukami, M., Komuro, I., 2002. Molecular and cellular mechanisms of mechanical stressinduced cardiac hypertrophy. Endocr. J. 49 (1), 1–13. Zupkovitz, G., Tischler, J., Posch, M., Sadzak, I., Ramsauer, K., Egger, G., Grausenburger, R., Schweifer, N., Chiocca, S., Decker, T., Seiser, C., 2006. Negative and positive regulation of gene expression by mouse histone deacetylase 1. Mol. Cell Biol. 26 (21), 7913–7928.

Chapter 10.3

MicroRNA Regulation of Cardiac Development and Disease Kimberly R. Cordes and Deepak Srivastava Gladstone Institute of Cardiovascular Disease and Departments of Pediatrics and Biochemistry & Biophysics, University of California, San Francisco, CA, USA

I.  Introduction Over the last decade, animal studies and advances in human genetics have highlighted the need for precise regulation of the expression of nodal pathway genes during embryonic development. This is particularly true for the cardiovascular system, where haploinsufficiency is the most common known cause of human disease. The dosage of cardiogenic pathway genes can be controlled at numerous levels, some of which have been well studied. In particular, the transcriptional regulation of cardiomyocyte differentiation and cardiac morphogenesis is highly-conserved across species, and heterozygous mutations of transcription factors have frequently been implicated in human cardiac malformations. Protein dosage can also be controlled at the level of protein stability; however, despite advances in our understanding of ubiquitin-mediated degradation pathways, there is limited evidence that this mechanism is important in cardiogenesis. Over the last few years, a novel mechanism involving post-transcriptional regulation by small noncoding RNAs, such as microRNAs (miRNAs), has emerged as a central regulator of many cardiogenic processes. miRNAs are a large class of evolutionarily-conserved, small, noncoding RNAs, typically 20–26 nucleotides (nt) in length, that primarily function post-transcriptionally by interacting with the 3 untranslated region (UTR) of specific target mRNAs in a sequence-specific manner. Over 500 miRNAs are encoded in the human genome, and each is thought to target from 50 to more than 100 mRNAs, resulting in mRNA degradation or translational inhibition. Interactions between miRNAs and mRNAs are thought to require sequence homology in the 5 end of the miRNA, but significant variance in the degree of complementation in the remaining sequence allows a single miRNA to target a wide Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

range of mRNAs, often regulating multiple genes within a common pathway. The first described animal miRNA, lin-4, was cloned in a forward genetic screen and characterized as a translational repressor of developmental timing in Caenorhabditis elegans, targeting the 3UTR of lin-14 mRNA (Lee et al., 1993). However, due to the lack of homology of lin-4 in other species, it was considered a genetic peculiarity specific to C. elegans. Years later a second miRNA, let-7, was cloned. It targeted the 3UTR of the highly conserved mRNA lin-41, another heterochronic gene. However, let7 was highly conserved across species, providing the first indication that miRNAs might be widely used across species to titrate protein expression (Pasquinelli et al., 2000; Reinhart et al., 2000). Through small RNA cloning efforts, it soon became clear that miRNAs were widespread in the genomes of all eukaryotes (Lee and Ambros, 2001). Over one-third of mRNAs in the mammalian genome are thought to be regulated by one or more miRNAs (Chaudhuri and Chatterjee, 2007). Despite advances in miRNA discovery, the role of miRNAs in physiological and pathophysiological processes is just emerging. It has become clear that miRNAs play diverse roles in fundamental biological processes, such as cell proliferation, differentiation, apoptosis, stress response and tumorigenesis. In many cancers, miRNAs are dysregulated and may act as tumor suppressors; in fact, the tumor suppressor gene p53 regulates the miR-34 family (He et al., 2007), and let-7 represses a prevalent oncogene found in a variety of tumors (Lee and Dutta, 2007). Identification of miRNAs expressed in specific cardiac cell types has led to the discovery of important regulatory roles for these small RNAs during cardiomyocyte differentiation, cell-cycle, conduction and during stages of cardiac 729

730

PART  |  10  Epigenetic Modifiers of Cardiac Development

hypertrophy in the adult, indicating that miRNAs may be as important as transcription factors in controlling cardiac gene expression. Here we review the basic mechanisms by which miRNAs function, with a focus on the role of miRNAs during development and maintenance of the heart and vessels. Ultimately, knowledge of the function and regulation of specific miRNAs and their mRNA targets will further our understanding of the mechanisms that regulate gene expression, and may also lead to new and distinct therapeutic targets for heart disease.

II.  Biogenesis, organization and target recognition of miRNA miRNAs regulate gene expression at the post-transcriptional level by mRNA deadenylation, translation repression, or miRNA-mediated mRNA decay (Fig. 1). Mature miRNAs are formed in a multi-step biological process involving critical endonucleases. miRNAs are initially

TF (e.g. SRF)

transcribed from the genome into long (several kilobases) 5-capped polyadenylated (poly(A)) primary transcripts (pri-miRNAs) by RNA polymerase II (Cai et al., 2004). Some miRNAs interspersed among repetitive DNA elements, such as Alu repeats (5 AG/CT 3), can also be transcribed by RNA polymerase III (Borchert et al., 2006). The miRNA-encoding portion of the pri-miRNA forms a hairpin structure that is recognized and cleaved in the nucleus by a microprocessor complex. This complex consists of the double-stranded RNA-specific nuclease DROSHA and its co-factor, DiGeorge syndrome-critical region 8 (DGCR8) (Landthaler et al., 2004). The resulting 70 nt hairpin precursor miRNA (pre-miRNA) is exported to the cytoplasm by the RAN-GTP-dependent nuclear transport receptor, exportin-5, which acts by recognizing a 2–3 basepair (bp) overhang of the pre-miRNA stem-loop structure (Bohnsack et al., 2004; Zeng and Cullen, 2004). The pre-miRNA is further processed by a complex of the RNAse III-like ribonuclease Dicer and the transactivator RNA-binding protein, which cleaves the pre-miRNA to release the mature miRNA duplex.

Figure 1  Schematic representation of miRNA biogenesis and function. Transcription of miRNA genes is typically mediated by RNA polymerase II (pol II). The initial miRNA-containing transcript, termed primary miRNAs (pri-miRNAs), can range from a few hundred nucleotides (nt) to several kilobases long. Inside the nucleus, the pri-miRNA has a characteristic stem-loop structure that can be recognized and cleaved by the ribonuclease III (RNase III) endonuclease Drosha, along with its partner DGCR8 (DiGeorge syndrome-critical region 8 gene, also known as Pasha). The cleavage product, a 70 nt stem-loop pre-miRNA, is exported from the nucleus by Exportin 5. In the cytoplasm, another RNase III enzyme, Dicer, further cleaves the pre-miRNA into a double-stranded mature miRNA (21 nt), which is incorporated into the RNA-induced silencing complex (RISC) allowing preferential strand-separation of the mature miRNA to repress mRNA translation or destabilize mRNA transcripts through cleavage or deadenylation (SRF: serum response factor; TF: transcription factor). Adapted from Zhao and Srivastava (2007).

miR

Transcription by RNA Pol II Nucleus Pri-miRNA 7mGpppG

PolyA

Microprocessing by Drosha-DGCR8

Pre-miR

Exportin 5 Pre-miRNA Dicer

RISC

P-Body

5′p

3′

An

Translation

Translational repression

AAAAAA

Degradation

AA

AAAA

Deadenylation

Chapter  |  10.3  MicroRNA Regulation of Cardiac Development and Disease

731

An asymmetry in the relative thermodynamic stability of the 5 ends of the miRNA duplex results in preferential loading of the less stable 22-nt strand into the RNA-induced silencing complex (RISC); the other strand is degraded, although in some cases both strands are incorporated into the RISC (Khvorova et al., 2003; Schwarz et al., 2003; Ro et al., 2007). The RISC helps mediate miRNA–mRNA interactions and subsequent mRNA repression or destabilization (Gregory et al., 2005). miRNAs typically bind to the 3UTRs of their mRNA targets with im­precise complementarity. Typically, the degree of Watson-Crick base-pairing between bases 2 and 7 (the “seed region”) at the 5 end of the miRNA is critical for binding mRNA targets (Stark et al., 2005; Rajewsky, 2006) and causing repression. RISC-bound miRNAs may also be sequestered away from translational machinery in processing bodies (P-bodies) that act by recruiting poly(A) nucleases to help modulate deadenylatation of mRNA and thereby prevent translation (Liu et al., 2005; Kim et al., 2006b; Ezzeddine et al., 2007). miRNAs can be found in exons or introns of noncoding transcripts with independent enhancer regulation, and in the introns and 39UTRs of protein-coding transcripts. They can also overlap with either an exon or an intron, depending on the alternative splicing pattern. In flies and worms, some miRNAs in intronic regions bypass Drosha processing and enter the miRNA biogenesis pathway as pre-miRNAs (Ruby et al., 2007). In many cases, miRNAs are clustered near other miRNAs, suggesting they may be co-regulated transcriptionally and share cooperative regulatory roles. Among the hundreds of miRNAs identified so far, only a limited number have been assigned target mRNAs. Several algorithmic databases have been designed for miRNA target prediction that rely, for the most part, on the following criteria: (1) conservation across species; (2) complementarity of the 5 miRNA “seed match” to the 39UTR (~7 nt) (Lewis et al., 2005; Rajewsky, 2006; Zhao et al., 2007); (3) G:U wobbles in the seed (Brennecke et al., 2005); (4) the thermodynamic context of target mRNA-binding sites (i.e., mRNA targets located in regions of high free energy and unstable secondary structure are favored) (Zhao et al., 2005, 2007); and (5) multiple miRNAbinding sites in 3UTR (Doench and Sharp, 2004). These computational programs are continuously updated to integrate these criteria with knowledge from newly-validated miRNA–mRNA interactions. Until more miRNA targets are validated, the precise mechanism of what makes one predicted target mRNA-binding site more desirable than the next remains to be determined.

hybridization, real-time PCR and microarray profiling. The development of miRNA microarray platforms for highthroughput miRNA expression profiling has allowed identification of tissue-specific miRNAs, cancer-related miRNAs, disease-associated miRNAs and ubiquitously-expressed miRNAs. Array analyses reveal miRNAs that may be the most critical in modulating gene expression in a “tissue of interest”. Ultimately, to identify the mRNA targets of miRNAs, it will be necessary to understand their biological relevance. An approach to study the comprehensive requirements of miRNAs during vertebrate development has been to create mutations in Dicer, the enzyme required to process miRNAs into their active mature forms. Dicer is encoded by a single locus in vertebrates. Zebrafish lacking maternal and endogenous Dicer die from defects in gastrulation, brain morphogenesis, somitogenesis and heart development (Wienholds et al., 2003; Giraldez et al., 2005). In mice, targeted deletion of Dicer causes lethality at embryonic day 7.5 (E7.5), before body axis formation (Bernstein et al., 2003). Tissue-specific approaches to delete Dicer have shown that Dicer activity is essential for morphogenesis of the mouse limb, lung, brain and heart (Harfe et al., 2005; Harris et al., 2006; Schaefer et al., 2007; Zhao et al., 2007). To create a heart-specific deletion of Dicer, Cre-recombinase was expressed under the control of the endogenous Nkx2.5 regulatory elements in the presence of a conditional Dicer allele. The Nkx2.5-Cre is active from E8.5 onward, during heart patterning and differentiation (Moses et al., 2001). The cardiac deletion of Dicer results in embryoniclethality caused from cardiac dysfunction at E12.5 (Zhao et al., 2007). Numerous miRNAs were affected by deletion of Dicer and may be important in regulating aspects of cardiogenesis. Dicer activity is also required for normal function of the mature heart as adult mice lacking Dicer in the myocardium have a high incidence of sudden death, cardiac hypertrophy, and reactivation of the fetal cardiac gene program (da Costa Martins et al., 2008). It will be important to determine if Dicer is required for earlier stages of cardiogenesis (before E8.5), such as cardiac lineage specification, since Dicer is required for embryonic stem cell differentiation (Kanellopoulou et al., 2005; Murchison et al., 2005). Collectively, these studies demonstrate the importance of miRNAs during development. In the future, additional cardiac conditional knockouts of Dicer, miRNA gene-targeting approaches and conditional targeted deletions of miRNAs in different heart populations will reveal the importance of miRNAs during the different stages of heart development (e.g., cardiomyocyte commitment, chamber morphogenesis, outflow tract remodeling).

III.  The function of mirnas during cardiogenesis

IV.  Cardiac- and muscle-specific miRNAs

Tools similar to those used to study mRNA expression are also used to study miRNA expression, such as in situ

Several miRNAs that are specific to or abundant in muscle have been identified, including miR-1, miR-133,

732

miR-138, miR-181, miR-206 and miR-208 (Zhao et al., 2005; Chen et al., 2006; Kim et al., 2006a; Naguibneva et al., 2006; van Rooij et al., 2007). miRNAs that function during heart morphogenesis and postnatal heart maintenance, and their roles in skeletal muscle, are described below.

IV.A.  Organization and Regulation of miR-1 and miR-133 Two widely-conserved miRNAs that display cardiac- and skeletal muscle-specific expression during development and in the adult are miR-1 and miR-133, which are derived from a common precursor transcript (bicistronic) (Zhao et al., 2005; Chen et al., 2006). Multiple loci encode the mature miR-1 and miR-133 transcripts; miR-1-1 is encoded on mouse chromosome 2 (human chromosome 20) and miR-1-2 on mouse chromosome 18 (human chromosome 18). Mouse chromosomes 18, 2 and 1, encode miR-133a-1, miR-133a2 and miR-133b, respectively. The mature forms of miR-1 derived from the distinct loci are identical, as are the miR133a forms. The miR-1 and miR-133 paralogs are present in the human, mouse, chick and fish genomes; a single ortholog of miR-1 and miR-133 exists in fly and worm (Kwon et al., 2005). The related miR-1 family member, miR-206, shares extensive sequence homology to miR-1, but is expressed exclusively in skeletal muscle with the co-transcribed miR133b (Rao et al., 2006). The miR-1-1/miRmiR-133a-2 cluster is located in an intergenic region, whereas the miR-1-2/miR133a-1 reads in an antisense orientation between exons 12 and 13 of the Mindbomb 1 (Mib1) gene, which is involved in Delta-mediated Notch signaling (Fig. 2A) (Itoh et al., 2003). Transcription of the miR-1/miR-133 bicistronic precursors are directly regulated by the major myogenic differentiation factors, MyoD, myocyte enhancer factor-2 (Mef2) and serum response factor (SRF) (Zhao et al., 2005; Chapters 9.3, 9.4). MyoD functions exclusively in skeletal muscle, while Mef2 and SRF regulate gene expression in cardiac, skeletal and smooth muscle development (Fig. 2B) (Edmondson et al., 1994; Miano, 2003). SRF binds to CArG motifs in promoters and enhancers of musclespecific genes that regulate differentiation, cell-cycle progression and tissue-specific gene expression (Catala et al., 1995). In the heart, SRF binds and activates the enhancer regions of miR-1/miR-133 in vitro and in vivo through a serum response element conserved from fly to human (Zhao et al., 2005). Similarly, SRF regulates the cardiac expression of miR-1 in flies, and the bHLH transcription factor Twist and Mef2 regulate somatic muscle expression (Kwon et al., 2005; Sokol and Ambros, 2005). Mef2 can also activate transcription of the bicistronic miR-1/miR133 transcript via an intragenic muscle-specific enhancer, which provides cooperative temporo-spatial regulation of miRNA expression (Liu et al., 2007). In contrast to the upstream miR-1/133 enhancer, which directs expression

PART  |  10  Epigenetic Modifiers of Cardiac Development

within the ventricular chambers (Zhao et al., 2005), the intragenic miR-1/133 enhancer is active in the atrial and ventricular chambers (Liu et al., 2007). This suggests there may be differential regulation of miR-1 and miR-133 in order to modulate their regulation on target mRNAs in the muscle differentiation pathway downstream of Mef2. Concordant with their common cis- and transregulation, both miR-1 and miR-133 are co-expressed in cardiac and skeletal muscle throughout mouse development, and are robustly expressed in the adult (Fig. 2B–C) (Zhao et al., 2005; Chen et al., 2006).

IV.B.  Function of miR-1 during Cardiogenesis The expression of miR-1 directed by the enhancers described above commences at approximately E8.5 in mouse and increases throughout development. However, in Drosophila, miR-1 transcripts are detectable during early mesoderm formation, before the onset of mef2 expression. This may also be the case in mouse, through as yet undescribed enhancers. Overexpression of miR-1 under the control of the -MHC promoter diminishes the pool of proliferating ventricular myocytes by causing a premature exit from the cell-cycle. This negatively-regulates cardiac growth, in part by inhibiting translation of the heart and neural crest derivative-2, Hand2 (Zhao et al., 2005), a basic helix-loop-helix protein involved in ventricular myocyte expansion. In mice, Hand2 is initially expressed throughout the linear heart tube, and then becomes restricted to the developing atrial and ventricular myocardium, with highest expression in the right ventricle. Mice that lack Hand2 die at E10.5 from right ventricular hypoplasia and decreased trabeculation in the left ventricle (Srivastava et al., 1997; Thomas et al., 1998; Yamagishi et al., 2001). In mice overexpressing miR-1, trabeculation is also decreased, consistent with the Hand2 mutant phenotype, corroborating Hand2 as a direct target of miR-1 (Zhao et al., 2005). Mice lacking miR-1 have an increase in Hand2 protein, providing further evidence of Hand2 as a direct target of miR-1. In Drosophila, miR-1 functions to pattern the dorsal vessel (i.e., aorta/heart tube). Moreover, the deletion of the single miR-1 gene (dmiR-1) results in a muscle-differentiation defect (Kwon et al., 2005; Sokol and Ambros, 2005). In a subset of dmiR-1-null flies, muscle progenitors are arrested in a proliferative state and accumulate ectopically. Drosophila hand does not seem to be a target of miR-1, since the fly hand ortholog lacks miR-1 binding sites in its 3UTR, suggesting that miRNA–mRNA interactions may differ somewhat between species. Instead, dmiR-1 targets transcripts encoding the Notch ligand, Delta, which regulates the expansion of cardiac and muscle progenitor cells (Kwon et al., 2005), suggesting that miR-1 promotes muscle differentiation through downregulation of the Notch signaling pathway. This is consistent with the known function

733

Chapter  |  10.3  MicroRNA Regulation of Cardiac Development and Disease

(A) miR-1-1

Chr 2

miR-133a-2

(intergenic)

9.2kb miR-1-2

Chr 18 Exon13

miR-133a-1 2.6kb

Mib Exon12

(B)

SRF Hand2

Cardiac miRNA Regulation miR-1

SRF MEF2

miR-133

Irx5

miR-133

MYOD

miR-1

Delta

HDAC4

Skeletal Muscle miRNA Regulation MEF2/ MYOD

(C)

miR-1-1/ miR-133a-2

miR-1-2/ miR-133a-1

h ht

somites Figure 2  Summary of miR-1 and miR-133 genomic organization, regulation and expression during mouse cardiogenesis. (A) Chromosome locations of miR-1 and miR-133a orthologs. The miR-1-2 and miR-133a-1 cluster is intragenic, and the miR-1-1 and miR-133a-2 cluster is intergenic. The miR-1/133a clusters are transcribed as a bicistronic transcript. (B) Cardiac (red) and muscle (green)-specific expression of miR-1 and miR-133 clusters is regulated by SRF and myogenic transcription factors, Mef2c and Myod. Targets of miR-1 and miR-133 that regulate cardiac or skeletal muscle are shown. (C) LacZ directed by an upstream enhancer of the miR-1-2/miR-133a-2 cluster and the miR-1-1/miR-133a-1 cluster, respectively, shows expression in the heart (ht) and somites (arrowhead) at mouse embryonic day 11.5.

of the Notch–Delta signaling pathway during developmental cell fate decisions, including those involving cardiac specification (Artavanis-Tsakonas et al., 1999). In cultured myoblasts, miR-1 promotes myoblast differentiation, whereas miR-133 stimulates myoblast proliferation (Chen et al., 2006). Chen and colleagues showed, in

myoblast culture, that miR-1 targets the histone deacetylase 4 (HDAC4) mRNA, a transcriptional repressor of Mef2cdependent activation of muscle-specific gene expression, suggesting that translational repression of HDAC4 by miR-1 enhances gene activation of Mef2-dependent promoters. They also showed that miR-133 targets SRF which is

734

PART  |  10  Epigenetic Modifiers of Cardiac Development

miR-1

ES Cell

Dll-1 miR-1 miR-133

miR-1

Endodermal Precursor

Mesodermal Precursor

miR-1

Gut Cell

miR-133

miR-133

Muscle Cell

Ectodermal Precursor

miR-1 miR-133

Neural Cell

Figure 3  Model of miR-1/miR-133 effects during embryonic stem cell differentiation. Promotion of mesoderm and inhibition of endoderm and ectoderm differentiation by miR-1 and miR-133 at specific stages are indicated. Opposing effects of the two miRNAs in later steps of muscle differentiation are also shown. Inhibition of Delta-like 1(Dll-1) translation by miR-1, along with yet unknown targets, likely contributes to the observed effects of miR-1. From Ivey et al. (2008) Cell Stem Cell.

important in muscle proliferation, differentiation and activation of the miR-1/miR-133 transcript, and thus creates a negative feedback loop of regulation (Fig. 2B). When rat ventricular cells are subjected to oxidative stress, miR-1 and miR-133 have opposing effects on apoptosis. Antiapoptotic heat shock proteins HSP60 and HSP70 are targeted by miR-1, which is proaptotic, whereas miR-133 represses caspase-9, a regulator of mitochondria-mediated apoptosis (Xu et al., 2007), and is anti-apoptotic. During early cell fate decisions of mouse and human embryonic stem (ES) cells, miR-1 and miR-133 function in concert to promote mesoderm induction, while suppressing differentiation into the ectodermal or endodermal lineages (Fig. 3) (Ivey et al., 2008). However, miR-1 and miR-133 have antagonistic effects on further adoption of muscle lineages; miR-1 promotes differentiation of mouse and human ES cells toward a cardiac fate, while miR-133 inhibits differentiation into cardiac muscle. In part, miR-1 appears to exert this effect by translationally repressing the mammalian ortholog of delta, Delta-like-1 (Dll-1), similar to the repression seen in the fly (Ivey et al., 2008). Thus, the bicistronic miR-1/miR-133 transcript encodes distinct mature miRNAs that likely share common targets, yet complement each other by balancing the differentiation and proliferation of cardiac and skeletal muscle lineages. In contrast to in vitro data showing that miR-133 promotes proliferation in cultured myoblasts and cardiac progenitors (Chen et al., 2006; Ivey et al., 2008), mice lacking miR133a-1 and miR-133a-2 had excessive cardiac proliferation

(Liu et al., 2008). In addition, compound mutants had partial embryonic lethality due to large ventricular septal defects, similar to miR-1-2 knockout mice ((Zhao et al., 2007), discussed in section IV.C.). Dysregulation of cell cycle control genes and aberrant activation of the smooth muscle gene program were observed in double-mutant mice, which may be due to the upregulation of the miR133a mRNA targets cyclinD2 and SRF, respectively.

IV.C.  Targeted Deletion of Mouse miR-1-2 Targeted deletion in mice will be invaluable for investigating the functional role of individual miRNAs. Recently, three distinct miRNAs (miR-1-2, miR-208 and miR-155) were deleted. Surprisingly, disruption of just one of the two miR-1 family members, miR-1-2, results in a range of abnormalities, including cell-cycle dysregulation, heart malformations and postnatal electrophysiological defects. Deletion of both miR-1-2 and miR-1-1 will likely cause more profound cardiac defects. Loss of the heart-specific miR-208 had less severe consequences, but impaired the heart’s ability to respond to stress postnatally. These deletions have significantly advanced our understanding of miRNA function and are described in some detail below.

IV.C.i.  miR-1 and Heart Morphogenesis Heterozygous miR-1-2-null mice survive to reproduce, but 50% of miR-1-2-homozygous-null mice die between E15.5

735

Chapter  |  10.3  MicroRNA Regulation of Cardiac Development and Disease

(A)

LV

LV

RV

RV

WT

E15.5

miR-1-2

–/–

E15.5

(B)

PR 20 msec

R′

miR-1-2–/–

WT

PR QRS

20 msec

QRS

Figure 4  Cardiac morphologic and electrophysiologic defects in the miR-1-2 mutants. (A) Transverse sections of wild-type (wt) or miR-1-2/ hearts at E15.5 showing ventricular septal defect (arrowhead). (B) Representative diagrams of electrocardiograms indicate the location of PR and QRS intervals. The second peak in the QRS complex (R) was observed in the majority of mutant mice representing delay of electrical conduction (RV: right ventricle; LV: left ventricle; msec: milliseconds).

and just after birth, due to apparent ventricular septal defects (and cardiac dysfunction) (Fig. 4A–B). These defects can occur from dysregulation of a multitude of events during cardiogenesis, and it is likely that miR-1-2 regulates numerous genes during this process. Precise dosage of Hand2 is crucial for normal cardiomyocyte proliferation and development, and elevated levels of Hand2 may contribute to the ventricular septal defects and cardiac death (Zhao et al., 2007).

IV.C.ii.  The Postnatal Heart and miR-1: Cardiac Electrophysiology and Cell-Cycle Mice that are miR-1-2-null that survive until birth often suffer sudden death (Zhao et al., 2007). Electrophysiological testing revealed a spectrum of cardiac arrhythmias in mutant mice. The heart’s electrical activity begins in the sinoatrial node and propagates impulses to the ventricles, resulting in depolarization, ventricular contraction and subsequent repolarization of the heart to initiate cardiac relaxation. The miR-1-2 mutants have a shortened P–R interval (the time from the beginning of atrial excitation to the beginning of ventricular excitation) and a prolonged QRS complex (the duration of ventricular depolarization) (Fig. 4B). Ventricular depolarization occurs by rapid conduction through the atrioventricular bundle, bundle branches and Purkinje fibers. A prolonged QRS often corresponds to a bundle-branch block, and can increase the risk of sudden death in humans (Turrini et al., 2001; Desai

et al., 2006). The cardiac arrhythmias in miR-1-2 mutants may be caused, in part, by elevated levels of the transcription factor Iroquois homeobox 5 (Irx5). Irx5 regulates the cardiac ventricular repolarization gradient by negatively-regulating the expression of potassium channel genes, such as Kcnd2 (Costantini et al., 2005). The 3UTR of Irx5 contains a wellconserved miR-1-binding site, and was shown by Zhao and colleagues to be a direct target of miR-1. In the miR-1-2-null hearts, Irx5 transcripts were upregulated, and its target gene, Kcnd2, was correspondingly downregulated. Additional studies showed that cardiac electrophysiology is sensitive to miR-1 and miR-133 dosage. In humans with coronary heart disease, miR-1 expression is elevated, and in normal rats or rats subjected to myocardial infarction, overexpression of miR-1 increased the occurrence of arrhythmias (Yang et al., 2007). Arrhythmias are common after a heart attack, and delivery of an antisense oligonucleotide to decrease miR-1 in the rat infarct model reverses the slow conduction and depolarization of the heart, consistent with the knockout studies (Yang et al., 2007). Furthermore, Yang and colleagues showed that miR-1 targets the 3UTRs of two ion channels prevalent in the adult heart, GJA1 and KCNJ2, which encode the cardiac gap junction connexin 43 (Cx43) and the potassium channel subunit Kir2.1, respectively. GJA1 is responsible for intercellular conductance in the ventricles, while KCNJ2 is responsible for setting and maintaining the cardiac resting membrane potential.

736

Interestingly, miR-133 also participates in regulating cardiac electrophysiology, as it translationally represses the ion channels KCNQ1 and KCNH2, which encode potassium channel subunits and are responsible for producing the cardiac repolarization current (Luo et al., 2007; Xiao et al., 2007). Luo and colleagues showed that miR-133 gain-of-function animals displayed a prolonged QT interval, with increased ventricular depolarization and repolarization times. Prolonged QT interval is characteristic of the congenital long-QT syndrome, which is associated with an increased risk of fatal ventricular tachyarrhythmias. Thus, miR-1 and miR-133 appear to regulate numerous genes that control proper cardiac conduction, providing an example of miRNA’s role as a “master” regulator, by virtue of the multiple mRNAs it can target. As such, manipulation of these miRNAs may have therapeutic value in the prevention of cardiac arrhythmias, particularly during the high-risk period immediately following myocardial infarction. Postnatal mouse cardiomyocytes terminally exit the cellcycle after the first 10 days of life. However, miR-1-2-null adult mice have an increase in mitotic cardiac myocytes, along with cardiac hyperplasia. These abnormalities could reflect the effect of miR-1 on Notch signaling and the derepression of Hand2, which promotes myocyte expansion. In addition, genome-wide profiling of miR-1-2 mutant adult hearts suggests a broad upregulation of positive regulators of the cell-cycle and downregulation of tumor suppressors, indicating a shift in the “threshold” of cells to re-enter the cell-cycle (Zhao et al., 2007). Whether this change promotes cardiac repair after injury remains to be determined.

IV.D.  miR-138 Regulation of Cardiac Patterning Intricate transcriptional networks establish chamberspecific gene expression, and these patterning events are highly conserved across species from zebrafish to human (Srivastava, 2006). Zebrafish are useful models to study cardiac patterning events because of their simple twochambered heart consisting of a single atrium and ventricle separated by the atrioventricular canal (AVC). The atrial and ventricular chambers express unique myosin genes, whereas the AVC expresses distinct genes such as cspg2, encoding versican, notch1b, and tbx2 (Rutenberg et al., 2006; Chi et al., 2008). miR-138 is a highly conserved miRNA found in many parts of the embryo, but within the zebrafish heart is specifically expressed in the ventricular chamber (Morton et al., 2008). Disruption of miR-138 function led to expansion of AVC gene expression into the ventricle and failure of ventricular cardiomyoctyes to fully mature. miR-138 normally restricts AVC gene expression by directly repressing cspg2 in the ventricle. This event is reinforced by ventricular repression of retinoic acid dehydrogenase, resulting in decreased retinoic acid, which is a positive regulator

PART  |  10  Epigenetic Modifiers of Cardiac Development

of cspg2 (Morton et al., 2008). It is likely that other regionspecific miRNAs will reinforce known signaling and transcriptional networks that establish patterns of gene expression throughout the developing heart tube.

IV.E.  Targeted Deletion of miR-208 The miRNA miR-208 is found within the intron of the gene encoding the highly-abundant -myosin heavy chain (-MHC), the predominant MHC isoform in the postnatal mouse heart. The -MHC protein allows for fast and efficient contractions, while the -MHC protein is less abundant and produces slower, less efficient contraction. In mice with cardiac hypertrophy, -MHC decreases while -MHC, the fetal isoform, increases – a signature contributing to cardiac death. This single miRNA was found to regulate stress-dependent cardiomyocyte growth and gene expression in vitro and in vivo. Van Rooij and colleagues showed that targeted deletion of miR-208 protects mice against cardiac hypertrophy and myocardial fibrosis, and results in ectopic activation of fast skeletal muscle gene expression (Fig. 5). Moreover, during cardiac stress, the absence of miR-208 resulted in failure of -MHC upregulation, which may in part be due to an increase in thyroid hormone receptor signaling. The thyroid receptor functions through a negative thyroid receptor response element to repress -MHC expression in the adult heart (Morkin, 2000). Translationally, miR-208 inhibits production of the thyroid receptor-associated protein, THRAP1, a transcriptional co-regulator of the thyroid receptor. Thus, loss of miR-208 results in increased thyroid hormone signaling, and thereby may regulate myosin isoform switching.

IV.F.  Function of miR-206 and miR-181 Cardiac muscle arises from lateral plate mesoderm and skeletal muscle from the paraxial mesoderm of somites. However, they have common mesodermal precursors and a similar sarcomeric organization for contractility. Thus, they may also share common modes of translational control by the miR-1 family. The miRNA miR-206 is a close homolog of miR-1, as its mature form differs from miR-1 by only three nucleotides that lie outside the seed region. Like miR-1, the primary miR-206 transcript is activated by MyoD and highly-expressed in skeletal muscle, but it is not found in cardiac muscle. It lies 3.8 kb upstream of, and is co-expressed with, miR-133b, suggesting that it is part of the same primary transcript, similar to the other miR-1/ miR-133 family members. In culture, miR-206, like miR-1, promotes the differentiation of skeletal myoblasts by repressing genes such as GJA1 (Cx43) and other antagonists of myogenesis (Anderson et al., 2006; Kim et al., 2006; Rosenberg et al., 2006). In myoblasts, miR-206 and miR-1 regulate DNA synthesis and cell-cycle withdrawal, a function

737

Chapter  |  10.3  MicroRNA Regulation of Cardiac Development and Disease

(A)

(B) TAB

miR-208–/–

WT Sham

TAB

Sham

TAB

Figure 5  Requirement of miR-208 for cardiomyocyte hypertrophy and fibrosis. (A) Schematic diagram of a heart following thoracic aortic banding (TAB). (B) Sections of hearts of approximately three-month-old wild-type and miR-208/ mice are shown following sham operation of TAB for 21 days. High-magnification views of the ventricular wall are shown at the bottom. Trichrome staining identifies fibrosis in blue. Note that hypertrophy and fibrosis are diminished in miR-208/ mice compared with wild-type following TAB. Scale bars: 2 mm (top); 20 mm (bottom). From van Rooij et al. (2007), © Science.

similar to miR-1’s cell-cycle regulatory role in the heart (Nakajima et al., 2006; Rosenberg et al., 2006). It is likely that miR-206 shares some common targets with miR-1, but it also has unique targets reflected by the sequence variance. Another miRNA important during mammalian myo­ blast differentiation is miR-181. It is upregulated during muscle differentiation and regeneration, but is barely detectable in terminally-differentiated muscle in adult mice. miR-181 acts upstream of MyoD by targeting the homebox protein HoxA11, a MyoD repressor (Yamamoto and Kuroiwa, 2003; Naguibneva et al., 2006). Depletion of miR-181 in myoblast culture inhibits MyoD expression and its downstream target myogenin, preventing myo­blast differentiation (Naguibneva et al., 2006). Expression of miR-181 is not exclusive to muscle, as it also promotes B-lymphocyte differentiation lineage in mouse hemapoietic tissues (Chen et al., 2004). Targeted deletion loss-of-function studies have not been performed.

V.  Cardiac stress-responsive miRNAs Heart disease is the leading cause of death of adults in developed countries (Thom et al., 2006). Cardiac hypertrophy often occurs in conjunction with or as a prelude to heart failure, and results in a global change in gene expression. Pathological stresses, including hemodynamic stress associated with myocardial infarction, hypertension, aortic stenosis and valvular incompetence result in increased protein synthesis, and ultimately activate a fetal program of gene expression leading to heart failure. For example, the genes encoding atrial natriuretic factor and -MHC are readily-expressed during embryonic heart development

and are downregulated postnatally (Fatkin et al., 2000). Although there have been major advances in mechanistically-coupling stress signals to fetal gene reprograming, the underlying mechanisms are only partly-understood and suggest additional regulatory mechanisms, perhaps by miRNAs. Recent reports show that miRNAs are indeed regulated during cardiac hypertrophy and dilated cardiomyopathy. This discovery may help elucidate the mechanisms of disease progression leading to heart failure. Several independent research groups have analyzed miRNA expression in the hearts of cardiac hypertrophy mouse models and idiopathic end-stage failing human hearts, and detected a dysregulation of miRNAs with some overlap. These studies suggest that aberrant miRNA expression may contribute to cardiac disease. Van Rooij and colleagues showed that overexpression of some dysregulated miRNAs was sufficient to elicit a hypertrophic response in rat primary cardiomyoctes, and to provoke dilated cardiomyopathy in transgenic mice. For example, overexpression of miR-195 resulted in cardiac hypertrophy in mice, as revealed by increase in cardiac growth, disorganization of cardiomyoctes, and activation of the fetal genes encoding atrial natriuretic factor and -MHC. Expression profiles of miRNA in human failing hearts are concordant with miRNA expression profiles in fetal human hearts, suggesting that activation of the fetal gene program during heart failure includes activation of fetal miRNAs (Thum et al., 2007). Notably, miR-1 and miR-133 are downregulated on induction of cardiac stress in mice and humans (van Rooij et al., 2006; Care et al., 2007; Sayed et al., 2007). Inhibiting miR-133 in wild-type mice by injecting a chemicallymodified synthetic antisense RNA oligonucleotide, called an

738

antagomir (Krutzfeldt et al., 2005), recapitulated hypertrophic parameters (Care et al., 2007). Conversely, a miR133 gain-of-function in an AKT-Tg hypertrophic mouse model resulted in a significant reduction of ventricular cardiomyocyte size and a significant decrease in fetal gene expression (Care et al., 2007). Three genes involved in cardiac hypertrophy are targeted by miR-133: Rhoa and Cdc42 (which are members of the RhoA subfamily of GTP-binding proteins); and Whsc2 (a negative regulator of RNA pol II), further elucidating the pathogenesis of disease. The bicistronic expression of miR-133 and miR-1 and their cooperativeness in skeletal myoblast differentiation and proliferation, respectively, suggest that they may regulate cardiac hypertrophy in a cooperative manner. To support this notion, in vitro functional assays performed by Sayed and colleagues showed that miR-1 behaves similarly to miR-133 in response to hypertrophic stimuli.

PART  |  10  Epigenetic Modifiers of Cardiac Development

promote new blood vessel formation in ischemic conditions as well to inhibit angiogenesis during tumor growth.

VII.  Summary The miRNAs play fascinating roles in the heart, both pre- and postnatally. Through their ability to post-transcriptionally regulate mRNA levels, and thus manage protein dosage, miRNAs provide finer regulation within the complex molecular networks that regulate cardiogenesis. The importance of this fine regulation is highlighted by the recognition that most known genetic causes of heart malformations in humans result from haploinsufficiency or heterozygous point mutations. The field of miRNA biology is growing rapidly, and new tools and mechanisms are becoming available. With further characterization, elucidating the function of cardiac-enriched miRNAs may provide us with new diagnostic, prognostic and therapeutic targets for many forms of cardiovascular disease.

VI.  miRNA Function during Angiogenesis

References

In addition to miRNA regulation of cardiomyocytes, recent reports illustrate what will likely be a broader function of tissue-specific miRNAs during vascular development (reviewed in (Fish and Srivastava, 2008; Fish et al., 2008; Wang et al., 2008). In particular, miR-126, which is located in the intron of an endothelial-specific gene, Egfl7, is the most highly enriched miRNA in endothelial cells derived from embryo­ nic stem cells or developing embryos (Fish et al., 2008). miR-126 is a key positive regulator of andiogenic signaling in endothelial cells and of vascular integrity in vivo (Fish et al., 2008; Wang et al., 2008). Knock-down of miR-126 during zebrafish embryogenesis or deletion of miR-126 in mice resulted in defects in vascular development. For example, collapsed blood vessels and cranial hemorrhages occurred in zebrafish with reduced levels of miR-126 (Fish et al., 2008), and delayed angiogenic sprouting, widespread hemorrhaging and partially embryonic lethality were observed in mice deficient in miR-126 (Wang et al., 2008; Kuhnert et al., 2008). MiR-126 mutant mice that successfully completed embryogenesis displayed diminished angiogenesis and increased mortality after coronary ligation, a model for myocardial infarction (Wang et al., 2008). Molecular analysis revealed that miR-126-deficient endothelial cells failed to respond to angiogenic factors, including VEGF, epidermal growth factor (EGF) and bFGF (Fish et al., 2008; Wang et al., 2008) and the regulatory subunit of PI3K, p85 (also known as PIK3R2) (Fish et al., 2008). Because Spred1 and PIK3R2 are negative regulators of cellular signaling cascades, affecting the MAPK and PI3K signaling pathways, respectively, miR-126 promotes VEGF and other growth factor signaling. By targeting multiple signaling pathways, miR-126 may fine-tune angiogenic responses. Because of miR-126’s central role in vascular development, miR-126-mediated regulation of angiogenesis may be a valuable therapeutic target to

Anderson, C., Catoe, H., Werner, R., 2006. MIR-206 regulates connexin43 expression during skeletal muscle development. Nucleic. Acids Res. 34, 5863–5871. Artavanis-Tsakonas, S., Rand, M.D., Lake, R.J., 1999. Notch signaling: Cell fate control and signal integration in development. Science 284, 770–776. Bernstein, E., Kim, S.Y., Carmell, M.A., Murchison, E.P., Alcorn, H., Li, M.Z., Mills, A.A., Elledge, S.J., Anderson, K.V., Hannon, G.J., 2003. Dicer is essential for mouse development. Nat. Genet. 35, 215–217. Bohnsack, M.T., Czaplinski, K., Gorlich, D., 2004. Exportin 5 is a RanGTP-dependent dsRNA-binding protein that mediates nuclear export of pre-miRNAs. RNA 10, 185–191. Borchert, G.M., Lanier, W., Davidson, B.L., 2006. RNA polymerase III transcribes human microRNAs. Nat. Struct. Mol. Biol. 13, 1097–1101. Brennecke, J., Stark, A., Russell, R.B., Cohen, S.M., 2005. Principles of microRNA-target recognition. PLoS Biol. 3, e85. Cai, X., Hagedorn, C.H., Cullen, B.R., 2004. Human microRNAs are processed from capped, polyadenylated transcripts that can also function as mRNAs. RNA 10, 1957–1966. Care A., Catalucci, D., Felicetti, F., Bonci, D., Addario, A., Gallo, P., Bang, M.L., Segnalini, P., Gu, Y., Dalton, N.D., Elia, L., Latronico, M.V., Hoydal, M., Autore, C., Russo, M.A., Dorn, G.W. 2nd, Ellingsen, O., Ruiz-Lozano, P., Peterson, K.L., Croce, C.M., Peschle, C., Condorelli, G., 2007. MicroRNA-133 controls cardiac hypertrophy. Nat. Med. 13, 613–618. Catala, F., Wanner, R., Barton, P., Cohen, A., Wright, W., Buckingham, M.E., 1995. A skeletal muscle-specific enhancer regulated by factors binding to E and CArG boxes is present in the promoter of the mouse myosin light-chain 1A gene. Mol. Cell Biol. 15, 4585–4596. Chaudhuri, K., Chatterjee, R., 2007. MicroRNA detection and target prediction: Integration of computational and experimental approaches. DNA Cell Biol. 26, 321–337. Chen, C.Z., Li, L., Lodish, H.F., Bartel, D.P., 2004. MicroRNAs modulate hematopoietic lineage differentiation. Science 303, 83–86. Chen, J.F., Mandel, E.M., Thomson, J.M., Wu, Q., Callis, T.E., Hammond, S.M., Conlon, F.L., Wang, D.Z., 2006. The role of

Chapter  |  10.3  MicroRNA Regulation of Cardiac Development and Disease

microRNA-1 and microRNA-133 in skeletal muscle proliferation and differentiation. Nat. Genet. 38, 228–233. Chi, N.C., Shaw, R.M., De Val, S., Kang, G., Jan, L.Y., Black, B.L., Stainier, D.Y., 2008. Foxn4 directly regulates tbx2b expression and atrioventricular canal formation. Genes Dev. 22, 734–739. Costantini, D.L., Arruda, E.P., Agarwal, P., Kim, K.H., Zhu, Y., Zhu, W., Lebel, M., Cheng, C.W., Park, C.Y., Pierce, S.A., Guerchicoff, A., Pollevick, G.D., Chan, T.Y., Kabir, M.G., Cheng, S.H., Husain, M., Antzelevitch, C., Srivastava, D., Gross, G.J., Hui, C.C., Backx, P.H., Bruneau, B.G., 2005. The homeodomain transcription factor Irx5 establishes the mouse cardiac ventricular repolarization gradient. Cell 123, 347–358. da Costa Martins, P.A., Bourajjaj, M., Gladka, M., Kortland, M., van Oort, R.J., Pinto, Y.M., Molkentin, J.D., De Windt, L.J., 2008. Conditional dicer gene deletion in the postnatal myocardium provokes spontaneous cardiac remodeling. Circulation 118, 1567–1576. Desai, A.D., Yaw, T.S., Yamazaki, T., Kaykha, A., Chun, S., Froelicher, V.F., 2006. Prognostic significance of quantitative QRS duration. Am. J. Med. 119, 600–606. Doench, J.G., Sharp, P.A., 2004. Specificity of microRNA target selection in translational repression. Genes Dev. 18, 504–511. Edmondson, D.G., Lyons, G.E., Martin, J.F., Olson, E.N., 1994. Mef2 gene expression marks the cardiac and skeletal muscle lineages during mouse embryogenesis. Development 120, 1251–1263. Ezzeddine, N., Chang, T.C., Zhu, W., Yamashita, A., Chen, C.Y., Zhong, Z., Yamashita, Y., Zheng, D., Shyu, A.B., 2007. Human TOB, an antiproliferative transcription factor, is a poly(A)-binding proteindependent positive regulator of cytoplasmic mRNA deadenylation. Mol. Cell Biol. 27, 7791–7801. Fatkin, D., McConnell, B.K., Mudd, J.O., Semsarian, C., Moskowitz, I.G., Schoen, F.J., Giewat, M., Seidman, C.E., Seidman, J.G., 2000. An abnormal Ca(2) response in mutant sarcomere protein-mediated familial hypertrophic cardiomyopathy. J. Clin. Invest. 106, 1351–1359. Fish, J.E., Santoro, M.M., Morton, S.U., Yu, S., Yeh, R.F., Wythe, J.D., Ivey, K.N., Bruneau, B.G., Stainier, D.Y., Srivastava, D., 2008. miR-126 regulates angiogenic signaling and vascular integrity. Dev. Cell 15, 272–284. Fish, J.E., Srivastava, D., 2008. MicroRNAs: opening a new vein in angiogenesis research. Science Signaling 1, pe1. Giraldez, A.J., Cinalli, R.M., Glasner, M.E., Enright, A.J., Thomson, J.M., Baskerville, S., Hammond, S.M., Bartel, D.P., Schier, A.F., 2005. MicroRNAs regulate brain morphogenesis in zebrafish. Science 308, 833–888. Gregory, R.I., Chendrimada, T.P., Cooch, N., Shiekhattar, R., 2005. Human RISC couples microRNA biogenesis and posttranscriptional gene silencing. Cell 123, 631–640. Harfe, B.D., McManus, M.T., Mansfield, J.H., Hornstein, E., Tabin, C.J., 2005. The RNaseIII enzyme Dicer is required for morphogenesis but not patterning of the vertebrate limb. Proc. Natl. Acad. Sci. USA 102, 10898–10903. Harris, K.S., Zhang, Z., McManus, M.T., Harfe, B.D., Sun, X., 2006. Dicer function is essential for lung epithelium morphogenesis. Proc. Natl. Acad. Sci. USA 103, 2208–2213. He, L., He, X., Lim, L.P., de Stanchina, E., Xuan, Z., Liang, Y., Xue, W., Zender, L., Magnus, J., Ridzon, D., Jackson, A.L., Linsley, P.S., Chen, C., Lowe, S.W., Cleary, M.A., Hannon, G.J., 2007. A microRNA component of the p53 tumour suppressor network. Nature 447, 1130–1134. Itoh, M., Kim, C.H., Palardy, G., Oda, T., Jiang, Y.J., Maust, D., Yeo, S.Y., Lorick, K., Wright, G.J., Ariza-McNaughton, L., Weissman, A.M., Lewis, J., Chandrasekharappa, S.C., Chitnis, A.B., 2003. Mind bomb is a ubiquitin ligase that is essential for efficient activation of Notch signaling by Delta. Dev. Cell 4, 67–82.

739

Ivey, K.N., Muth, A., Arnold, J., King, F.W., Yeh, R., Fish, J.E., Hsiao, E.C., Schwartz, R.J., Conklin, B.R., Bernstein, H.S., Srivastava, D., 2008. MicroRNA regulation of cell lineages in mouse and human embryonic stem cells. Cell Stem Cell, 2, 219–229. Kanellopoulou, C., Muljo, S.A., Kung, A.L., Ganesan, S., Drapkin, R., Jenuwein, T., Livingston, D.M., Rajewsky, K., 2005. Dicer-deficient mouse embryonic stem cells are defective in differentiation and centromeric silencing. Genes. Dev. 19, 489–501. Khvorova, A., Reynolds, A., Jayasena, S.D., 2003. Functional siRNAs and miRNAs exhibit strand bias. Cell 115, 209–216. Kim, H.K., Lee, Y.S., Sivaprasad, U., Malhotra, A., Dutta, A., 2006a. Muscle-specific microRNA miR-206 promotes muscle differentiation. J. Cell Biol. 174, 677–687. Kim, S.K., Nam, J.W., Rhee, J.K., Lee, W.J., Zhang, B.T., 2006b. miTarget: microRNA target gene prediction using a support vector machine. BMC Bioinformatics 7, 411. Krutzfeldt, J., Rajewsky, N., Braich, R., Rajeev, K.G., Tuschl, T., Manoharan, M., Stoffel, M., 2005. Silencing of microRNAs in vivo with “antagomirs”. Nature 438, 685–689. Kuhnert, F., Mancuso, M.R., Hampton, J., Stankunas, K., Asano, T., Chen, C.Z., Kuo, C.J., 2008. Attribution of vascular phenotypes of the murine Egfl7 locus to the microRNA miR-126. Dev. 135, 3989–3993. Kwon, C., Han, Z., Olson, E.N., Srivastava, D., 2005. MicroRNA1 influences cardiac differentiation in Drosophila and regulates Notch sig­ naling. Proc. Natl. Acad. Sci. USA 102, 18986–18991. Landthaler, M., Yalcin, A., Tuschl, T., 2004. The human DiGeorge syndrome critical region gene 8 and its D. melanogaster homolog are required for miRNA biogenesis. Curr. Biol. 14, 2162–2167. Lee, R.C., Feinbaum, R.L., Ambros, V., 1993. The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14. Cell 75, 843–854. Lee, R.C., Ambros, V., 2001. An extensive class of small RNAs in Caenorhabditis elegans. Science 294, 862–864. Lee, Y.S., Dutta, A., 2007. The tumor suppressor microRNA let-7 represses the HMGA2 oncogene. Genes Dev. 21, 1025–1030. Lewis, B.P., Burge, C.B., Bartel, D.P., 2005. Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell 120, 15–20. Liu, J., Valencia-Sanchez, M.A., Hannon, G.J., Parker, R., 2005. MicroRNA-dependent localization of targeted mRNAs to mammalian P-bodies. Nat. Cell Biol. 7, 719–723. Luo, X., Xiao, J., Lin, H., Li, B., Lu, Y., Yang, B., Wang, Z., 2007. Transcriptional activation by stimulating protein 1 and post-transcriptional repression by muscle-specific microRNAs of IKs-encoding genes and potential implications in regional heterogeneity of their expressions. J. Cell Physiol. 212, 358–367. Miano, J.M., 2003. Serum response factor: toggling between disparate programs of gene expression. J. Mol. Cell Cardiol. 35, 577–593. Morkin, E., 2000. Control of cardiac myosin heavy chain gene expression. Microsc. Res. Tech. 50, 522–531. Morton, S.U., Scherz, P.J., Cordes, K.R., Ivey, K.N., Stainier, D.Y., Srivastava, D., 2008. microRNA-138 modulates cardiac patterning during embryonic development. Proc. Natl. Acad. Sci. USA 105, 17830–17835. Moses, K.A., DeMayo, F., Braun, R.M., Reecy, J.L., Schwartz, R.J., 2001. Embryonic expression of an Nkx2-5/Cre gene using ROSA26 reporter mice. Genesis 31, 176–180. Murchison, E.P., Partridge, J.F., Tam, O.H., Cheloufi, S., Hannon, G.J., 2005. Characterization of Dicer-deficient murine embryonic stem cells. Proc. Natl. Acad. Sci. USA 102, 12135–12140.

740

Naguibneva, I., Ameyar-Zazoua, M., Polesskaya, A., Ait-Si-Ali, S., Groisman, R., Souidi, M., Cuvellier, S., Harel-Bellan, A., 2006. The microRNA miR-181 targets the homeobox protein Hox-A11 during mammalian myoblast differentiation. Nat. Cell Biol. 8, 278–284. Nakajima, N., Takahashi, T., Kitamura, R., Isodono, K., Asada, S., Ueyama, T., Matsubara, H., Oh, H., 2006. MicroRNA-1 facilitates skeletal myogenic differentiation without affecting osteoblastic and adipogenic differentiation. Biochem. Biophys. Res. Commun. 350, 1006–1012. Pasquinelli, A.E., Reinhart, B.J., Slack, F., Martindale, M.Q., Kuroda, M.I., Maller, B., Hayward, D.C., Ball, E.E., Degnan, B., Muller, P., Spring, J., Srinivasan, A., Fishman, M., Finnerty, J., Corbo, J., Levine, M., Leahy, P., Davidson, E., Ruvkun, G., 2000. Conservation of the sequence and temporal expression of let-7 heterochronic regulatory RNA. Nature 408, 86–89. Rajewsky, N., 2006. MicroRNA target predictions in animals. Nat. Genet. 38 (Suppl), S8–S13. Rao, P.K., Kumar, R.M., Farkhondeh, M., Baskerville, S., Lodish, H.F., 2006. Myogenic factors that regulate expression of muscle-specific microRNAs. Proc. Natl. Acad. Sci. USA 103, 8721–8726. Reinhart, B.J., Slack, F.J., Basson, M., Pasquinelli, A.E., Bettinger, J.C., Rougvie, A.E., Horvitz, H.R., Ruvkun, G., 2000. The 21-nucleotide let-7 RNA regulates developmental timing in Caenorhabditis elegans. Nature 403, 901–906. Ro, S., Park, C., Young, D., Sanders, K.M., Yan, W., 2007. Tissuedependent paired expression of miRNAs. Nucleic. Acids Res. 35, 5944–5953. Rosenberg, M.I., Georges, S.A., Asawachaicharn, A., Analau, E., Tapscott, S.J., 2006. MyoD inhibits Fstl1 and Utrn expression by inducing transcription of miR-206. J. Cell Biol. 175, 77–85. Ruby, J.G., Jan, C.H., Bartel, D.P., 2007. Intronic microRNA precursors that bypass Drosha processing. Nature 448, 83–86. Sayed, D., Hong, C., Chen, I.Y., Lypowy, J., Abdellatif, M., 2007. MicroRNAs play an essential role in the development of cardiac hypertrophy. Circ. Res. 100, 416–424. Schaefer, A., O’Carroll, D., Tan, C.L., Hillman, D., Sugimori, M., Llinas, R., Greengard, P., 2007. Cerebellar neurodegeneration in the absence of microRNAs. J. Exp. Med. 204, 1553–1558. Schwarz, D.S., Hutvagner, G., Du, T., Xu, Z., Aronin, N., Zamore, P.D., 2003. Asymmetry in the assembly of the RNAi enzyme complex. Cell 115, 199–208. Sokol, N.S., Ambros, V., 2005. Mesodermally expressed Drosophila microRNA-1 is regulated by Twist and is required in muscles during larval growth. Genes Dev. 19, 2343–2354. Srivastava, D., Thomas, T., Lin, Q., Kirby, M.L., Brown, D., Olson, E.N., 1997. Regulation of cardiac mesodermal and neural crest development by the bHLH transcription factor, dHAND. Nat. Genet. 16, 154–160. Srivastava, D., 2006. Making or breaking the heart: from lineage determination to morphogenesis. Cell 126, 1037–1048. Stark, A., Brennecke, J., Bushati, N., Russell, R.B., Cohen, S.M., 2005. Animal MicroRNAs confer robustness to gene expression and have a significant impact on 3’UTR evolution. Cell 123, 1133–1146. Thom, T., Haase, N., Rosamond, W., Howard, V.J., Rumsfeld, J., Manolio, T., Zheng, Z.J., Flegal, K., O’Donnell, C., Kittner, S., Lloyd-Jones, D., Goff, D.C., Hong, Y., Adams, R., Friday, G., Furie, K., Gorelick, P., Kissela, B., Marler, J., Meigs, J., Roger, V., Sidney, S., Sorlie, P., Steinberger, J., Wasserthiel-Smoller, S., Wilson, M., Wolf, P., 2006. Heart disease and stroke statistics – 2006 update: A report from the American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Circulation 113, e85–e151.

PART  |  10  Epigenetic Modifiers of Cardiac Development

Thomas, T., Yamagishi, H., Overbeek, P.A., Olson, E.N., Srivastava, D., 1998. The bHLH factors, dHAND and eHAND, specify pulmonary and systemic cardiac ventricles independent of left–right sidedness. Dev. Biol. 196, 228–236. Thum, T., Galuppo, P., Wolf, C., Fiedler, J., Kneitz, S., van Laake, L.W., Doevendans, P.A., Mummery, C.L., Borlak, J., Haverich, A., Gross, C., Engelhardt, S., Ertl, G., Bauersachs, J., 2007. MicroRNAs in the human heart: a clue to fetal gene reprogramming in heart failure. Circulation 116, 258–267. Turrini, P., Corrado, D., Basso, C., Nava, A., Bauce, B., Thiene, G., 2001. Dispersion of ventricular depolarization-repolarization: a noninvasive marker for risk stratification in arrhythmogenic right ventricular cardiomyopathy. Circulation 103, 3075–3080. van Rooij, E., Sutherland, L.B., Liu, N., Williams, A.H., McAnally, J., Gerard, R.D., Richardson, J.A., Olson, E.N., 2006. A signature pattern of stress-responsive microRNAs that can evoke cardiac hypertrophy and heart failure. Proc. Natl. Acad. Sci. USA 103, 18255–18260. van Rooij, E., Sutherland, L.B., Qi, X., Richardson, J.A., Hill, J., Olson, E.N., 2007. Control of stress-dependent cardiac growth and gene expression by a microRNA. Science 316, 575–579. Wang, S., Aurora, A.B., Johnson, B.A., Qi, X., McAnally, J., Hill, J.A., Richardson, J.A., Bassel-Duby, R., Olson, E.N., 2008. The endothelial-specific microRNA miR-126 governs vascular integrity and angiogenesis. Dev. Cell 15, 261–271. Wienholds, E., Koudijs, M.J., van Eeden, F.J., Cuppen, E., Plasterk, R. H., 2003. The microRNA-producing enzyme Dicer1 is essential for zebrafish development. Nat. Genet. 35, 217–218. Xiao, J., Luo, X., Lin, H., Zhang, Y., Lu, Y., Wang, N., Yang, B., Wang, Z., 2007. MicroRNA miR-133 represses HERG K channel expression contributing to QT prolongation in diabetic hearts. J. Biol. Chem. 282, 12363–12367. Xu, C., Lu, Y., Pan, Z., Chu, W., Luo, X., Lin, H., Xiao, J., Shan, H., Wang, Z., Yang, B., 2007. The muscle-specific microRNAs miR-1 and miR-133 produce opposing effects on apoptosis by targeting HSP60, HSP70 and caspase-9 in cardiomyocytes. J. Cell Sci. 120, 3045–3052. Yamagishi, H., Yamagishi, C., Nakagawa, O., Harvey, R.P., Olson, E.N., Srivastava, D., 2001. The combinatorial activities of Nkx2.5 and dHAND are essential for cardiac ventricle formation. Dev. Biol. 239, 190–203. Yamamoto, M., Kuroiwa, A., 2003. Hoxa-11 and Hoxa-13 are involved in repression of MyoD during limb muscle development. Dev. Growth Differ. 45, 485–498. Yang, B., Lin, H., Xiao, J., Lu, Y., Luo, X., Li, B., Zhang, Y., Xu, C., Bai, Y., Wang, H., Chen, G., Wang, Z., 2007. The muscle-specific microRNA miR-1 regulates cardiac arrhythmogenic potential by targeting GJA1 and KCNJ2. Nat. Med. 13, 486–491. Zeng, Y., Cullen, B.R., 2004. Structural requirements for pre-microRNA binding and nuclear export by Exportin 5. Nucleic. Acids Res. 32, 4776–4785. Zhao, Y., Samal, E., Srivastava, D., 2005. Serum response factor regulates a muscle-specific microRNA that targets Hand2 during cardiogenesis. Nature 436, 214–220. Zhao, Y., Ransom, J.F., Li, A., Vedantham, V., von Drehle, M., Muth, A.N., Tsuchihashi, T., McManus, M.T., Schwartz, R.J., Srivastava, D., 2007. Dysregulation of cardiogenesis, cardiac conduction, and cell cycle in mice lacking miRNA-1-2. Cell 129, 303–317.

Chapter 11.1

Genomic Analyses in the Developing and Diseased Heart Shuaib Latif1 and Daniel J. Garry2,3 1

Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, TX, USA Lillehei Heart Institute, University of Minnesota, Minneapolis, MN, USA 3 Department of Internal Medicine, University of Minnesota, Minneapolis, MN, USA 2

I.  Introduction Genomics is the study of the entire genome of the organism of interest. Initially, the study of gene expression was limited to single genes. Emerging technological advancements now provide new opportunities to study the entire genome of organisms. Following the sequence analysis of the human genome, the biological tools used to probe the development of the normal and pathological state of the cardiovascular system have undergone a revolution. Rather than probing the response of a small group of candidate transcripts, these technological advancements have facilitated the examination of the entire transcriptome (the set of all messenger RNA or transcripts of an organism) of a cell population in response to various experimental conditions. These technologies have unveiled new pathways or networks that direct cell fate decisions, growth and differentiation of tissues during development. However, the sea of data generated by these experiments is relatively new in the biological sciences. A systematic approach to experimental design, execution and analysis are important for successful interpretation of the studies. In this chapter, we will discuss different approaches to cardiovascular genomics and their implications for this field.

et al., 1980; Okayama et al., 1982; Gubler et al., 1983; Jandreski et al., 1987) were the initial strategies used to pursue a broader analysis of gene expression and promote the discovery of novel tissue-restricted genes. New techniques emerged such as “differential display” and serial analysis of gene expression (SAGE), which were valuable tools for cardiovascular gene discovery. The successful sequencing of the human genome in 2001 revealed that the three billion nucleotide bases comprised approximately 20,000–25,000 genes (Venter et al., 2001; IHGSC, 2004) (Fig. 1). The Human Genome Project also generated expressed sequence tags (ESTs), which are short doublestranded fragments usually hundreds of bases in length that were derived from tissue-specific cDNA libraries and used for gene discovery. The sequencing of the human genome, the need for efficient whole-genome screening strategies and engineering advances all led to the development of the cDNA chip in the early-1990s. The first scientific report using a cDNA platform was published in 1992 (Liang and Pardee, 1992). Since that time, studies in many species utilizing transcriptome analysis (i.e., DNA chips) to interrogate transcript expression have increased exponentially, and now provide a voluminous amount of data to the scientific community (Fig. 1).

II.  Genomic profiling strategies

II.A.  Conventional Methods for Cardiovascular Gene Discovery

Gene expression profiling on a large scale first became possible with the advent of tools to manipulate DNA and RNA. Subtraction hybridization (Hedrick et al., 1984) and the generation of tissue-specific cDNA libraries (Woods

Differential display is a global, unbiased technique that can be utilized to study gene expression in two or more samples using conventional molecular biology techniques (Liang et al., 1992; Zimmerman et al., 1994). While

Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

743

744

PART  |  11  Cardiomics

several variations exist, the overall approach is the same. Initially, RNA is isolated from the samples of interest and then subjected to reverse transcription of the polyaden­ ylated population of RNA. The cDNA is then used in a polymerase chain reaction (PCR) assay; the PCR products are gel-isolated, reamplified and sequenced. While simple in design, the technique allows simultaneous analysis of both up- and downregulated transcripts between samples and has been used to analyze gene expression in the developing and diseased heart. This strategy has been successful for the identification of cardiac restricted genes, such as Ankrd1 (Jeyaseelan et al., 1997), and has been utilized to profile gene expression in cardiac transplant rejection (Utans et al., 1994). This early genomic technique has provided significant insight into cardiac development and disease. Serial analysis of gene expression (SAGE) also allows the study of global gene expression and was first used in 1995 (Velculescu et al., 1995). In this technique, total cellular RNA is isolated from samples of interest and cDNA is synthesized using a biotinylated primer. The cDNA is cleaved, oligonucleotide linkers with a specific recognition site for a tagging enzyme are fused to the cDNA ends, and the tagging enzyme cuts the DNA at a constant number of bases 3 to the recognition site, releasing a part of the linker and a unique fragment of cDNA. These fragments are ligated together and cloned, creating a library of oligomerized cDNAs that can be sequenced (Ye et al., 2002). While SAGE can be a technically-challenging technique and is labor intensive, it is useful in the identification of novel genes (Schwartz et al., 2004). Furthermore, this technique can be used more broadly in the analysis of the cardiovascular systems of model organisms whose genomes have not been well-characterized, such as the urodele and the anuran Xenopus. SAGE has been successfully applied to examine differential gene expression in the adult mouse heart and embryonic cells that have been induced to form cardiomyocytes (Anisimov et al., 2002).

Number of manuscripts

Sequences (106)

(B) 800 60 56 700 52 48 600 44 40 500 36 32 400 28 24 300 20 16 200 12 8 100 4 0 0 1985 1990 1995 2000 2005 1997 1999 2001 2003 2005 Figure 1  Genomic sequence analysis provides a foundation for emerging technologies. (A) Sequence analysis of the human genome performed each year. (B) The number of manuscripts that principally use microarray analysis increase each year, as identified using the Medline database. Base pairs Sequence

Base pairs of DNA (109)

(A) 56 52 48 44 40 36 32 28 24 20 16 12 8 4 0

III.  Microarray technologies The ideal technology to evaluate gene expression is one that is sensitive (i.e., utilizing limited amounts of total cellular RNA), highly-reproducible and able to provide a global analysis of transcript expression. Several methods have been developed to achieve these goals. However, the microarray chip is perhaps the most versatile and reproducible technology to evaluate gene expression (i.e., mRNA expression). The amount of total cellular RNA required for microarray technologies has progressively decreased over the past decade and is significantly less than other global technologies such as SAGE. The two most common platforms used today for transcriptome analysis are cDNA microarray and high-density oligonucleotide chips.

III.A.  Complementary DNA (cDNA) Microarrays Complementary DNA (cDNA) microarrays were the first platform used to rapidly analyze a large number of transcripts from Arabidopsis thaliana (Schena et al., 1995). Since this initial report in 1995, cDNA microarrays have been used with exponentially increasing frequency in studies ranging across a number of species (Fig. 1B). Typically, this platform utilizes PCR to amplify cDNA libraries and the individual amplicons are spotted onto a glass slide using a robotic system to minimize artifacts and efficiently print a number of chips at one time (Fig. 2). The spots containing the PCR products (cDNAs) are approximately 200 microns in diameter, thus allowing 20–40,000 cDNA spots to be placed on one glass microscope slide. The utilization of cDNA libraries allows for the customization of the arrays, which can be produced from engineered cell populations, genetic mouse models or diseased tissue (hearts with congenital or acquired defects) (Hwang et al., 2002; Peng et al., 2002; Zhao et al., 2002;

Chapter  |  11.1  Genomic Analyses in the Developing and Diseased Heart

cDNA

Label with fluorophores

Oligonucleotide

Label with biotin

Combine samples Hybridize to slide

Hybridize to chip

Scan and analyze

Scan and analyze

Figure 2  Strategies using cDNA and oligonucleotide arrays for transcriptome analysis. The basic steps using cDNA or oligonucleotide arrays are highlighted to compare the transcript expression in two samples. Note that the total cellular RNA isolated from two separate samples is hybridized to the same chip for cDNA arrays. The samples competitively hybridize to the chip and result in green, red and yellow (no change) signals. In contrast, two samples can be independently processed and hybridized to two separate chips for the oligonucleotide arrays.

Tabibiazar et al., 2003). Typically, two samples (consisting of total cellular RNA isolated from different cell populations or tissues) are used to generate first strand cDNA, and each sample is then labeled with one or the other of two unique fluorescent dyes (e.g., Cy3 or Cy5). Equal amounts of the Cy3- and Cy5-labeled samples are hybridized to the same chip, thus allowing competitive binding of the fluorescently-labeled cDNAs to the array (Fig. 2). The hybridized chip is then washed, scanned (using a highresolution laser system) and analyzed for the intensity of the respective fluorescent labels for each spot (Schinke et al., 2003) (Fig. 2). There are a number of advantages of the cDNA array platform (Table 1). The principle advantage is the ability for the investigator to customize the array and directly compare the transcript expression of two separate samples. The cost-per-slide is affordable and, in contrast to other technologies (i.e., differential display and SAGE), is laborsaving (i.e., results can be obtained within days). The limitations of the technology include the amount of RNA that is required, and that only two samples can be compared in any one experiment (Fig. 2; Table 1). Since 1995, the cDNA arrays have been broadly applied for the analysis of gene expression and have been an important technology for studies directed towards an enhanced understanding of cardiac morphogenesis and cardiac repair.

745

III.B.  Oligonucleotide Arrays High-density oligonucleotide arrays differ from cDNA microarrays in that synthesized oligomers (often 25–50 mers) are “printed” directly onto a substrate (Fig. 2). Utilizing photolithographic chip technology (Fodor et al., 1991), the density of oligomers or features per chip continues to increase. The Affymetrix GeneChip (GeneChip Expression Analysis, 2004) was the first oligonucleotide array platform, and is still commonly used. The HG-U133 Plus 2.0 Affymetrix chip contains 1.3 million oligonucleotides, which covers over 47,000 transcript variants on a single chip. Expression of each gene is represented by eleven pairs of overlapping but staggered oligonucleotide probes (each probe is 25 base pairs in length). The signal of all the probe pairs representing the expression of a single gene is integrated and normalized based on the expression of a preselected group of housekeeping genes. Five-to-ten micrograms of total cellular RNA from the sample of interest is used to generate first and second strand cDNA; this is followed by an in vitro transcription reaction that generates biotin-labeled cRNA (GenChip Expression Analysis, 2004). The biotin-cRNA is fragmented, hybridized to the chip and scanned (Fig. 2). Several important differences are apparent between the cDNA and oligonucleotide arrays (Table 2). In contrast to the cDNA arrays where two samples are hybridized to the same chip, the oligonucleotide arrays are designed such that one sample is hybridized to one chip and the signal obtained is an absolute quantity (Fig. 2). Because only one sample is applied to each chip and generates an absolute quantification of expression level, data from oligonucleo­tide chips are comparable across experiments. Further advantages for the oligonucleotide array platforms include standardized chip production and processing which decreases the variability, the limited RNA quantity that is required for each reaction, and the availability of software programs for sample analysis. Current limitations of oligonucleotide arrays include the cost and the difficulty when applying this technology to gene discovery, as the arrays lack probe sets for transcripts not represented as documented genes or ESTs (Table 1). This latter hurdle will be addressed with the production of future generations of chips as they begin to include additional transcripts highlighted by sequence data. Thus, a single experiment will allow the interrogation of the entire genome of an organism.

III.C.  RNA Amplification The estimated total amount of RNA contained within each cell is around 10 pg (Sambrook et al., 2001). Therefore, tissue samples yielding limited RNA quantities such as the embryonic heart, a cardiac biopsy specimen (or the postinjured/regenerating murine heart) require RNA amplification. RNA amplification requires the technique

746

PART  |  11  Cardiomics

Table 1  Transcriptional Expression Analysis Methods

Sample comparisons Amount of starting RNA Cost Complexity of technique Reproducibility Detection of novel genes Global genetic profiling Detection of differential splicing Customization

Differential display

SAGE

cDNA microarray

Oligonucleotide microarray

Any 5 g total Low Low Medium Yes No Yes

Any 5 g poly-A Medium High Medium Yes Yes No

Two 1 g poly-A High High Medium Possible Yes Possible

Any 1–8 g Medium to high Medium High No Yes Possible

Yes

Yes

Yes

No

Table 2  cDNA versus Oligonucleotide Microarray

Number of genes Probe size Feature size Starting material

cDNA microarray

Oligonucleotide microarray

Variable (up to 19,000) 500 bp–1.5 kb 100– 400 m 5–10 g total RNA

35,000 25 mer–50 mer 11–100 m 1–10 g total RNA

to be reproducible and accurately reflect the unamplified RNA specimen. This latter requirement calls for linear amplification with an absence of skewing or bias of transcript expression. Amplification protocols have utilized a balanced PCR amplification of pooled samples, an exponential PCR amplification protocol, or a T7-based linear RNA amplification protocol (van Gelder et al., 1990; Livesey et al., 2003; Wang et al., 2003). Previous studies have demonstrated that samples having as few as 100 cells in combination with RNA amplification are sufficient for transcriptome analysis using an oligonucleotide array platform (Gallardo et al., 2003). Using a T7-based RNA amplification protocol, approximately 1000-fold amplification was possible with each round of in vitro transcription. While the T7-based amplification protocol was highlyreproducible, there was some skewing of transcripts in the 100 cell sample that may be due to increased G–C content of certain transcripts, increased transcript length, or secondary structure formation of the transcript (Gallardo et al., 2003). Further improvement in the fidelity of the RNA amplification protocol may ultimately allow investigators to examine global transcript expression at the single cell level (Chiang et al., 2003). Collectively, these technical strategies have enhanced our ability to apply these genome-wide technologies to assess global gene expression in the developing and regenerating heart.

Experimental Strategies Reductionist

Open-Ended

All Knowledge

Hypothesis

Hypothesis

Test, Study, Improve

Candidate

Themes/Knowledge

Test, Study, Improve

Test All

All Possibilities

All Possibilities

Figure 3  Conventional and transcriptome analyses result in complementary strategies. Conventional experimental strategies pursue a topdown or reductionist approach, where a hypothesis leads to the design of an experiment which is tested and improved. In contrast, an open-ended approach (i.e., transcriptome analysis) is an example of a bottom-up approach, which examines all possibilities to direct the selection of key experiments and ultimately to arrive at a hypothesis.

IV.  Data analysis and bioinformatics In contrast to a conventional hypothesis-directed approach to experimentation, transcriptome analysis allows the investigator to broadly examine a number of variables (i.e., transcript expression) to ultimately formulate a hypothesis (Fig. 3). Consequently, these emerging technologies are revolutionizing the scale of scientific information, as well as our experimental design and strategy. While microarray experiments have allowed the scientist to capture a significant portion of the transcriptome at a moment in time, analyzing the results of these experiments is perhaps the most daunting aspect of using microarray technology. The output of microarray experiments

Chapter  |  11.1  Genomic Analyses in the Developing and Diseased Heart

produces thousands-upon-thousands of transcript levels. Analyzing this amount of data from a single experiment is a relatively new phenomenon in the biological sciences and requires a systematic approach. First, the scanned image should be inspected to assess sample quality and chip quality. The scanned image should appear homogeneous without defects, holes, or speckles. The data from each image are processed, often using the software provided by the chip manufacturer, resulting in a signal for each transcript. Chipmakers will often add several controls for different aspects of the experiment. For example, the Affymetrix GeneChip includes the following controls: hybridization controls (which consist of four biotin-labeled transcripts in the hybridization cocktail at varying amounts to assess sample hybridization); a poly-A control (which consists of four poly-A transcripts used in the labeling reaction to assess the efficiency of the labeling reaction); a normalization control set (a set of 100 transcripts that span the expression spectrum); and housekeeping genes (such as Gapdh with probes spanning the transcript used to assess degradation of RNA) (GeneChip Mouse Genome Arrays Data Sheet, 2007). Collectively, these controls allow the researcher to assess the quality of the experimental sample and hybridization. After initial inspection and signal acquisition, data are typically normalized to adjust the hybridization intensities. The goal of normalization is to attempt to remove nonbiological variables that affect the signal, such as differing quantities of starting RNA, differences in labeling efficiencies and biases in measuring the signal. There are many approaches to normalizing expression levels, and numerous sources discuss the background and attributes for different methods with the overall goal being the same (Quackenbush, 2002; Smyth and Speed, 2003; Mocellin and Rossi, 2007). Once normalization has been completed, data are typically presented as either a ratio (in the case of cDNA microarrays), or as a signal intensity (oligonucleotide arrays). These data are then subjected to comparative analysis to assess differential gene expression. Data analysis in microarray experiments may be as simple as searching for a “significant” fold-change in transcript expression, or may involve the application of an increasing range of statistical tools and software packages for data management (e.g., dChip, MAS, GCOS). Finally, new approaches to data analysis subject the individual genes to further grouping or “clustering,” using the hypothesis that groups of expressed transcripts that move in a specific direction may add more meaning than searching for individual genes (Belacel et al., 2006; Gollub and Sherlock, 2006). Mathematical algorithms for the analysis of gene chip data sets continue to evolve rapidly, and will potentially make the enormous data sets generated from these experiments more versatile and manageable.

747

IV.A.  Applications Utilizing Transcriptome Analysis IV.A.i.  Global Gene Expression in the Developing Heart Elegant studies using morphological and gene disruption strategies have identified numerous transcription factors and signaling molecules that have an essential role in cardiac development in mice, fly and zebrafish (Buckingham et al., 2005; Garry and Olson, 2006). Using these genetic technologies, embryonic cardiac defects have been defined for transcription factors including homeodomain proteins (Nkx2-5, Irx4 and Hop), GATA factors (Gata4, Gata5 and Gata6), bHLH proteins (Hand1, Hand2, Hey), T-box proteins (Tbx5, Tbx1) and members of the Mef2 family (Mef2a, Mef2c, Mef2d) (reviewed in Buckingham et al., 2005; Srivastava, 2006) (see also Section 9). Genetic mutant mouse models display cardiac defects associated with hypoplasia (Nkx2-5 (Lyons et al., 1995)), defects in the second heart field (Foxh1 (von Both et al., 2004), Isl1 (Cai et al., 2003)), absence of the right ventricle (Mef2c (Lin et al., 1997), Hand2 (Srivastava et al., 1997), Foxh1) and vascular defects (Mef2c (Lin et al., 1998), Myocd (Li et al., 2003)). These transcription factors function in concert to regulate cardiac-specific molecular programs that promote heart development (Garry and Olson, 2006; Srivastava, 2006). However, none of the transcription factors mentioned is sufficient alone to confer cardiac, endothelial, or smooth muscle identity on nonmyogenic cells. Therefore, it is reasonable to conclude that specification of cardiac, endothelial, or smooth muscle fates requires combinations of regulatory factors, rather than a single “master” gene, as is the case for skeletal muscle, in which any member of the MyoD family of bHLH proteins is sufficient to convert a wide range of nonmuscle cell types to skeletal muscle (Davis et al., 1987; Braun et al., 1989; Rhodes et al., 1989; Wright et al., 1989). Therefore, the use of geneticallymodified mouse models in combination with transcriptome analysis has been particularly useful in uncovering transcriptional networks for cardiac morphogenesis. Considerable overlap of gene expression is observed between smooth muscle, endothelial and cardiac muscle lineages (SM22, SMA, Mef2c, Myocd, Flk1, etc.). These overlapping expression patterns occur in lineages that will ultimately contribute to heart formation (i.e., endothelial, cardiac and smooth muscle lineages), and suggest either common regulatory networks or a developmental relationship between these lineages. Previous studies have utilized a genetic mouse model harboring an Nkx2-5 enhancer-eYFP transgene to tag cardiac progenitors that reside in the cardiac crescent (Fig. 4). These geneticallytagged progenitors/cells were isolated from single embryos at distinct embryonic stages (cardiac crescent, heart tube and looped heart) using flow cytometry, and definition of

748

PART  |  11  Cardiomics

(A)

E7.75

E8.5

E9.5

(B) E7.75 (213)

24

123

284

E9.5 (606)

35 31

263 338 E8.5 (667)

Figure 4  Cardiac progenitors have a distinct and common molecular signature. Using transgenic technologies and an Nkx2-5 upstream fragment fused to the EYFP reporter, the cardiac progenitors/cells were identified at E7.75 (cardiac crescent), E8.5 (heart tube stage), and E9.5 (looped heart stage). Using flow cytometry, the EYFP cardiac progenitors/cells were isolated from individual embryos and transcriptome analysis was undertaken to define the gene expression of the cardiac progenitors/cells at distinct developmental stages. (B) Venn diagram of significantly upregulated transcripts in cardiac progenitors/cells (compared to EYFP-negative cells isolated from the same embryo) at specific developmental stages identifies distinct and common molecular programs. Figure adapted from Masino et al. (2004).

the molecular signature was performed using transcriptome analysis (Masino et al., 2004). Using this strategy, stage-specific gene expression profiles revealed that E7.75 cardiac progenitors (isolated from the cardiac crescent) had a distinct signature compared to the cardiac cells isolated from either the heart tube or looped heart (Figs 4; 5). More than 50% of the transcripts that were significantly enriched in the cardiac progenitors isolated from the crescent were not enriched at later developmental stages (i.e., heart tube or looped heart) (Masino et al., 2004). Moreover, approximately half of the transcripts isolated from the E7.75 cardiac progenitors were absent from the cardiac cells isolated from the heart tube or looped heart stages, suggesting that these molecular programs function in the specification of the progenitors for the cardiac lin­ eage (Masino et al., 2004). The cardiac progenitors isolated from the crescent stage embryos were enriched in transcripts commonly expressed in other mesodermal lineages, suggesting that a common ancestor may generate mesodermal lineages such as cardiac, endothelial, and hematopoietic lineages (Masino et al., 2004). Similar strategies that combine transgenic technologies and microarray analysis

have been useful in the global definition of transcript expression at distinct stages of heart development (Latif et al., 2006) (Fig. 5). Transcriptome analysis may also be used to identify downstream target genes for cardiac specific transcription factors early in development. This strategy typically compares the global transcript expression of wild-type hearts with mutant littermate hearts (generated using a gene disruption strategy) at distinct developmental stages. Transcriptome analysis of Nkx2-5-heterozygous and Nkx2-5-null embryonic hearts identified an Nkx2-5/ Bmp2/Smad1 negative feedback loop which subsequently regulates cardiac progenitor specification and proliferation (Prall et al., 2007) (see Chapter 9.1). Moreover, the use of transcriptome analysis allows investigators to pursue an open-ended strategy to “cast a net” for multiple factors or pathways that may be directly or indirectly regulated by early cardiac transcription factors such as Nkx2-5. Comparable strategies using transcriptome analysis will increasingly be utilized to identify networks that direct stem/ progenitor cells to acquire a cardiac fate in geneticallymodified mouse embryos. These applications will further enhance our understanding of the transcriptional regulatory events that direct cardiac morphogenesis. The use of transcriptome analysis has further elucidated pathways regulated by Nkx2-5 in the postnatal heart, as in a global analysis undertaken in a ventricular-restricted Nkx2-5 knockout. The conditional knockout of Nkx2-5 in mice results in complete heart block and trabecular muscle overgrowth (Pashmforoush et al., 2004). Transcriptome analysis of the postnatal wild-type and mutant heart identified a profile of gene expression which was dysregulated in the absence of Nkx2-5. In these experiments, RNA was isolated from pooled wild-type or Nkx2-5-deficient hearts and used for microarray analysis. Each experiment was repeated three times to minimize “noise”. One transcript that was upregulated in the absence of Nkx2-5 encoded Bmp10 (Pashmforoush et al., 2004). Previous studies have demonstrated that Bmp10 dosage is an important modulator of myocardial growth and differentiation (Chen et al., 2004). Therefore, this global analysis was important in revealing a cardiac pathway that includes Nkx2-5 as a negative regulator of Bmp10 expression.

IV.B.  Global Gene Expression in the   Post-Injured Heart Previous studies have utilized transcriptome analysis to identify pathways that direct repair and regeneration of adult mammalian tissues. These include array studies undertaken in post-injured skeletal muscle (Bakay et al., 2002), liver (Reilly et al., 2001), lung (Satomi et al., 2006) and brain (Marciano et al., 2002). Recent studies have performed transcriptome analysis using Affymetrix

Chapter  |  11.1  Genomic Analyses in the Developing and Diseased Heart

(A)

749

(B)

Figure 5  Cardiac morphogenesis has stage-specific molecular signatures. Using flow cytometry, cardiac progenitors/cells (as defined using an Nkx2-5 enhancer fused to EYFP and transgenic technologies) were collected from stage specific embryos (E7.75 and E9.5) or the adult heart (ACM) and their molecular signature was examined using microarray analysis (Masino et al., 2004). Expression was based on absence/presence call (•) or by significant fold change (1.4-fold change) (•) (cardiac progenitor cells: CPC (isolated from the cardiac crescent at E7.75); fetal cardiomyocytes: FCM (isolated from the E9.5 heart); adult cardiomyocytes: ACM). Figure adapted from Latif et al. (2006).

Zebrafish Genechips to examine global gene expression in the regenerating zebrafish heart (Lien et al., 2006). Prior studies have established that the zebrafish has the capacity for myocardial regeneration following amputation of the apical region of the ventricle (approximately 20% ventricular resection) (Poss et al., 2002; Raya et al., 2003). Gene expression profiling was undertaken at defined periods post-amputation (days 3, 7 and 14) and signatures of gene expression were identified using hierarchical cluster analysis. Inflammatory markers, secreted molecules, matrix metalloproteinases, platelet-derived growth factors and migratory factors were induced in response to repair/ regeneration. Two factors have consistently been identified using transcriptome analysis in regenerating tissues, and were also upregulated in the regenerating zebrafish heart – thymosin4 and tenascin C. Previous studies have identified thymosin family members as being strongly induced following injury in regenerating tissues. Thymosin4, a G-actin-sequesting peptide, has been shown to promote migration and survival of cardiomyocytes (BockMarquette et al., 2004). Thymosin4 was upregulated three

days and seven days post-amputation in the regenerating zebrafish heart. These temporal and spatial results further support recent knockout studies in mice where thymosin 4-deficient mice had impaired cellular migration and perturbed vasculogenesis following injury (Smart et al., 2007). Microarray analysis further identified tenascin C, which was upregulated in the regenerating zebrafish heart. Tenascin C has been shown to be upregulated in the blastema of the regenerating newt (Onda et al., 1991), the regenerating ear of the MRL mouse (Heber-Katz, 1999) and regenerating skeletal muscle in the mouse (Goetsch et al., 2003). Tenascin C has been proposed to limit the fibroproliferative response (i.e., scar formation) by inhibiting the activation of T-lymphocytes and the secretion of proinflammatory cytokines such as IL-2 (Harty et al., 2003). In addition, this protein has been further shown to maintain an undifferentiated state through the modulation of the pro-adhesive effects of laminin, fibronectin and collagens (Naseem et al., 2007). Future studies will be necessary to examine whether administration of tenascin C during the acute period following myocardial injury, as shown with

750

thymosin4, may limit the fibro-proliferative response and promote an increased regenerative response. These studies further emphasize the utility of transcriptome analysis to uncover factors that may have essential roles during regeneration.

IV.C.  Global Gene Expression in the Hypertrophic and Failing Heart Global expression profiling strategies using microarray technologies have enhanced our understanding of cardiac hypertrophy and failure in the adult heart. Using a custom microarray platform, gene expression was evaluated following the administration of angiotensin II or isopro­ terenol by osmotic minipumps in mice, which resulted in approximately 30% increase in heart weight (Friddle et al., 2000). Comparison of the hypertrophic response of the left ventricle following either of these agents resulted in the induction of a common program of gene expression that included transcripts encoding secreted factors (Anf, Bnp, Serpine1), signaling pathways (Ece-2, Comt, Pkc-bp) and structural proteins (Alpha-sm, Acta1, Actc), as well as novel proteins (Friddle et al., 2000). A number of these genes were also induced in the failing mouse heart (Blaxall et al., 2003). Utilizing the Affymetrix oligonucleotide microarray platform, global gene expression was examined in hearts isolated from a wild-type and a genetic mouse model that develops progressive heart failure (i.e., muscle LIM protein also known as Mlp-deficient mice). The results of this study supported the hypothesis that early and late stages of heart failure share similar gene expression profiles, and that the heart failure program appears to be established during the early stages of heart failure progression (Blaxall et al., 2003). This heart failure program (Anf, Bnp, Mhc, Des, Acta1, Myl1, Postn, Fhl1, Serpine1 and Actc) included, but was not limited to, the transcripts that were also induced in response to hypertrophic stimuli (Blaxall et al., 2003). Future studies will be needed to examine other genetic mouse models of heart failure to further define distinct and common molecular programs of heart failure. Microarray analysis has also been useful to define further the molecular response to myocardial unloading following the implementation of mechanical support using a ventricular-assist device in patients with advanced heart failure. Comparing ventricular myocardium before and after mechanical support reveals profiles of gene expression that support the conclusion that the unloaded myopathic heart has a decrease in the fibro-proliferative response and an increase in vasculoneogenesis (Blaxall et al., 2003; Chen et al., 2003; Hall et al., 2004; Kittleson et al., 2005). This strategy emphasizes the utility of a global genomic analysis in patients following treatment with a pharmacological or mechanical intervention. Global analysis removes the bias

PART  |  11  Cardiomics

which is obtained with a candidate gene assay, and broadly examines the response to an intervention (Fig. 3). In summary, microarray analysis has revolutionized our studies of developing, regenerating and diseased mammalian tissues. In the myocardial lineage, transcriptome analysis has uncovered networks that are essential for the growth and differentiation of the heart. Refinement of the technology, the addition of controls and the inclusion of thousands of features on a chip now allows the interrogation of the entire genome in a single experiment. Transcriptome analysis has progressed from being the only technique in a study to one that complements or extends a study. While this global analysis has been extensively used in the characterization of genetic models, it is increasingly being utilized in the clinical setting. Applications in the clinic include the microarray analysis of endomyocardial biopsy specimens or serum specimens in an attempt to define a signature expression of transcripts that serve as a predictor of outcomes (Barth et al., 2006; Morgun et al., 2006). Clinical applications for these technologies will continue to emerge, and will enhance our understanding of congenital and acquired cardiovascular disease and the molecular response to therapies.

References Anisimov, S.V., Tarasov, K.V., Stem, M.D., Lakatta, E.G., Boheler, K.R., 2002. A quantitative and validated SAGE transcriptome reference for adult mouse heart. Genomics 80, 213–222. Anisimov, S.V., Tarasov, K.V., Riordon, D., Wobus, A.M., Boheler, K.R., 2002. SAGE identification of differentiation responsive genes in P19 embryonic cells induced to form cardiomyocytes in vitro. Mech. Dev. 117, 25–74. Bakay, M., Zhao, P., Chen, J., Hoffman, E.P., 2002. A web-accessible complete transcriptome of normal human and DMD muscle. Neuromuscul. Disord. 12 (Suppl. 1), S125–S141. Barth, A.S., Hare, J.M., 2006. The potential for the transcriptome to serve as a clinical biomarker for cardiovascular diseases. Circ. Res. 98, 1459–1461. Belacel, N., Wang, Q., Cuperlovic-Culf, M., 2006. Clustering methods for microarray gene expression data. OMICS 10, 507–531. Blaxall, B.C., Sprang, R., Rockman, H.A., Koch, W.J., 2003. Differential myocardial gene expression in the development and rescue of murine heart failure. Physiol. Genom. 15, 105–114. Blaxall, B.C., Tschannen-Moran, B.M., Milano, C.A., Koch, W.J., 2003. Differential gene expression and genomic patient stratification following left ventricular assist device support. J. Am. Coll. Cardiol. 41, 1096–1106. Bock-Marquette, I., Saxena, A., White, M.D., Dimaio, J.M., Srivastava, D., 2004. Thymosin beta4 activates integrin-linked kinase and promotes cardiac cell migration, survival, and cardiac repair. Nature 432, 466–472. Braun, T., Buschhausen-Denker, G., Bober, E., Tannich, E., Arnold, H. H., 1989. A novel human muscle factor related to but distinct from MyoD1 induces myogenic conversion in 10T1/2fibroblasts. EMBO J. 8, 701–709.

Chapter  |  11.1  Genomic Analyses in the Developing and Diseased Heart

Buckingham, M.E., Meilhac, S., Zaffran, S., 2005. Building the mammalian heart from two sources of myocardial cells. Nat. Rev. Genet. 6, 826–835. Cai, C.L., Liang, X., Shi, Y., Chu, P.H., Pfaff, S.L., Chen, J., Evans, S., 2003. Isl1 identifies a cardiac progenitor population that proliferates prior to differentiation and contributes a majority of cells to the heart. Dev. Cell 5, 877–889. Chen, H., Shi, S., Acosta, L., Li, W., Lu, J., Bao, S., Chen, Z., Yang, Z., Schneider, M.D., Chien, K.R., Conway, S.J., Yoder, M.C., Haneline, L.S., Franco, D., Shou, W., 2004. BMP10 is essential for maintaining cardiac growth during murine cardiogenesis. Development 131, 2219–2231. Chen, Y., Park, S., Li, Y., Missov, E., Hou, M., Han, X., Hall, J.L., Miller, L.W., Bache, R.J., 2003. Alterations of gene expression in failing myocardium following left ventricular assist device support. Physiol. Genom. 14, 251–260. Chiang, M., Melton, D., 2003. Single-cell transcript analysis of pancreas development. Dev. Cell 4, 383–393. Davis, R.L., Weintraub, H., Lassar, A.B., 1987. Expression of a single transfected cDNA converts fibroblasts to myoblasts. Cell 51, 987–1000. Fodor, S.P., Read, J.L., Pirrung, M.C., Stryer, L., Lu, A.T., Solas, D., 1991. Light-directed, spatially addressable parallel chemical synthesis. Science 251, 767–773. Friddele, C.J., Koga, T., Rubin, E.M., Bristow, J., 2000. Expression profiling reveals distinct sets of genes altered during induction and regression of cardiac hypertrophy. Proc. Natl. Acad. Sci. USA 97 (12), 6745–6750. Gallardo, T.D., Hammer, R.E., Garry, D.J., 2003. RNA amplification and transcriptional profiling for analysis of stem cell populations. Genesis 37, 57–63. Garry, D.J., Olson, E.N., 2006. A common progenitor at the heart of development. Cell 127, 1101–1104. GeneChip Expression Analysis. (2004), Affymetrix, Inc., 30 March 2007. http://www.Affymetrix.com GeneChip Mouse Genome Arrays Data Sheet, (2007). Affymetrix, Inc., 30 March 2007. http://www.Affymetrix.com Goetsch, S.C., Hawke, T.J., Gallardo, T.D., Richardson, J.A., Garry, D.J., 2003. Transcriptional profiling and regulation of the extracellular matrix during muscle regeneration. Physiol. Genom. 14, 261–267. Gollub, J., Sherlock, G., 2006. Clustering microarray data. Methods Enzymol. 411, 194–213. Growth of GenBank, (2006). National Institutes of Health, 30 March 2007. http://www.ncbi.nlm.nih.gov/Genbank/genbankstats.html Gubler, U., Hoffman, B.J., 1983. A simple and very efficient method for generating cDNA libraries. Gene 25, 263–269. Hall, J.L., Grindle, S., Han, X., Fermin, D., Park, S., Chen, Y., Bache, R.J., Mariash, A., Guan, Z., Ormaza, S., Thompson, J., Graziano, J., de Sam Lazaro, S.E., Pan, S., Simari, R.D., Miller, L.W., 2004. Genomic profiling of the human heart before and after mechanical support with a ventricular assist device reveals alterations in vascular signaling networks. Physiol. Genom. 17, 283–291. Harty, M., Neff, A.W., King, M.W., Mescher, A.L., 2003. Regeneration or scarring, an immunologic perspective. Dev. Dyn. 226, 268–279. Heber-Katz, E., 1999. The regenerating mouse ear. Semin. Cell Dev. Biol. 4, 415–419. Hedrick, S.M., Cohen, I., Nielsen, E.A., Davis, M.M., 1984. Isolation of cDNA clones encoding T cell-specific membrane-associated proteins. Nature 308, 149–153.

751

Hwang, J., Allen, P.D., Tseng, G.C., Lam, C., Fananapazir, L., Dzau, V. J., Liew, C., 2002. Micorarray gene expression profiles in dilated and hypertrophic end-stage heart failure. Physiol. Genom. 10, 31–44. International Human Genome Sequencing Consortium, 2004. Finishing the euchromatic sequence of the human genome. Nature 431, 931–945. Jandreski, M.A., Liew, C.C., 1987. Construction of a human ventricular cDNA library and characterization of a beta myosin heavy chain cDNA clone. Hum. Genet. 76, 47–53. Jeyaseelan, R., Poizat, C., Baker, R.K., Abdishoo, S., Isterabadi, L., Lyons, G.E., Kedes, L., 1997. A novel cardiac-restricted target for doxorubicin. J. Biol. Chem. 272, 22800–22808. Kittleson, M.M., Hare, J.M., 2005. Molecular signature analysis, using the myocardial transcriptome as a biomarker in cardiovascular disease. Trends Cardiovasc. Med. 15, 130–138. Latif, S., Masino, A., Garry, D.J., 2006. Transcriptional pathways direct cardiac development and regeneration. Trends Cardiovasc. Med. 16, 234–240. Li, S., Wang, D.Z., Wang, Z., Richardson, J.A., Olson, E.N., 2003. The serum response factor coactivator myocardin is required for vascular smooth muscle development. Proc. Natl. Acad. Sci. USA 100, 9366–9370. Liang, P., Pardee, A.B., 1992. Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction. Science 257, 967–971. Lien, C.L., Schebesta, M., Makino, S., Weber, G.J., Keating, M.T., 2006. Gene expression analysis of zebrafish heart regeneration. PLoS. Biol. 4, e260. Lin, Q., Schwarz, J., Bucana, C., Olson, E.N., 1997. Control of mouse cardiac morphogenesis and myogenesis by transcription factor MEF2C. Science 276, 1404–1407. Lin, Q., Lu, Y., Yanagisawa, H., Webb, R., Lyons, G.E., Richardson, J. A., Olson, E.N., 1998. Requirement of the MADS-box transcription factor MEF2C for vascular development. Development 125, 4565–4574. Livesey, F.J., 2003. Strategies for microarray analysis of limiting amounts of RNA. Brief Funct. Genomic. Proteomic. 2, 31–36. Lyons, I., Parsons, L.M., Hartley, L., Li, R., Andrews, J.E., Robb, L., Harvey, R.P., 1995. Myogenic and morphogenetic defects in the heart tubes of murine embryos lacking the homeo box gene Nkx2-5. Genes Dev. 9, 1654–1666. Marciano, P.G., Eberwine, J.H., Ragupathi, R., Saatman, K.E., Meaney, D.F., McIntosh, T.K., 2002. Expression profiling following traumatic brain injury, a review. Neurochem. Res. 27, 1147–1155. Masino, A.M., Gallardo, T.D., Wilcox, C.A., Olson, E.N., Williams, R.S., Garry, D.J., 2004. Transciptional regulation of cardiac progenitor cell populations. Circ. Res. 95, 389–397. Mocellin, S., Rossi, C.R., 2007. Principles of gene microarray data analysis. Adv. Exp. Med. Biol. 593, 19–30. Morgun, A., Shulzhenko, N., Perez-Diez, A., Diniz, R.V., Sanson, G.F., Almeida, D.R., Matzinger, P., Gerbase-DeLima, M., 2006. Molecular profiling improves diagnosis of rejection and infection in transplanted organs. Circ. Res. 98, e74–e83. Naseem, R.H., Meeson, A.P., Dimaio, J.M., White, M.D., Kallhoff, J.B., Humphries, C.G., Goetsch, S.C., De Windt, L.J., Williams, M.A., Garry, M.G., Garry, D.J., 2007. Reparative myocardial mechanisms in adult C57BL/6 and MRL mice following injury. Physiol. Genom. 30, 44–52. Okayama, H., Berg, P., 1982. High-efficiency cloning of full-length cDNA. Mol. Cell Biol. 2, 161–170.

752

Onda, H., Poulin, M., Tassava, R.A., Chiu, I.M., 1991. Characterization of a newt tenascin cDNA and localization of tenascin mRNA during newt limb regeneration by in situ hybridization. Dev. Biol. 148, 219–232. Pashmforoush, M., Lu, J.T., Chen, H., Amand, T.S., Kondo, R., Pradervand, S., Evans, S.M., Clark, B., Feramisco, J.R., Giles, W., Ho, S.Y., Benson, D.W., Silberbach, M., Shou, W., Chien, K.R., 2004. Nkx2-5 pathways and congenital heart disease; loss of ventricular myocyte lineage specification lead to progressive cardiomyopathy and complete heart block. Cell 117, 373–386. Peng, C., Wei, Y., Levsky, J.M., McDonald, T.V., Child, G., Kitsis, R.N., 2002. Microarray analysis of global changes in gene expression during cardiac myocyte differentiation. Physiol. Genom. 9, 145–155. Poss, K.D., Wilson, L.G., Keating, M.T., 2002. Heart regeneration in zebrafish. Science 298, 2188–2190. Prall, O.W., Mennon, M.K., Solloway, M.J., Watanabe, Y., Zaffran, S., Bajolle, F., Biben, C., McBride, J.J., Robertson, B.R., Chaulet, H., Stennard, F.A., Wise, N., Schaft, D., Wolstein, O., Furtado, M. B., Shiratori, H., Chien, K.R., Hamada, H., Black, B.L., Saga, Y., Robertson, E.J., Buckingham, M.E., Harvey, R.P., 2007. An Nkx2-5/ Bmp2/Smad1 negative feedback loop controls heart progenitor specification and proliferation. Cell 128, 947–959. Quackenbush, J., 2002. Microarray data normalization and transformation. Nat. Genet. 32S, 496–501. Raya, A., Koth, C.M., Buscher, D., Kawakami, Y., Itoh, T., Raya, R.M., Sternik, G., Tsai, H.J., Rodriguez-Esteban, C., Izpisua-Belmonte, J.C., 2003. Activation of Notch signaling pathway precedes heart regeneration in zebrafish. Proc. Natl. Acad. Sci. USA 100S, 11889–11895. Reilly, T.P., Bourdi, M., Brady, J.N., Pise-Masison, C.A., Radonvich, M. F., George, J.W., Pohl, L.R., 2001. Expression profiling of acetaminophen liver toxicity in mice using microarray technology. Biochem. Biophys. Res. Commun. 282, 321–328. Rhodes, S.J., Konieczny, S.F., 1989. Identification of MRF4, a new member of the muscle regulatory factor gene family. Genes Dev. 3, 2050–2061. Sambrook, J., Russell, D.W., 2001. Molecular Cloning: A Laboratory Manual, third ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Satomi, Y., Tsuchiya, W., Miura, D., Kasahara, Y., Akahori, F., 2006. DNA microarray analysis of pulmonary fibrosis three months after exposure to paraquat in rats. J. Toxicol. Sci. 31, 345–355. Schena, M., Shalon, D., Davis, R.W., Brown, P.O., 1995. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270, 467–470. Schinke, M., Riggi, L., Butte, A.J., Izumo, S., 2003. Large scale expression profiling in cardiovascular disease using microarrays: Prospects and pitfalls. In: Van Eyk, J.E., Dunn, M.J. (Eds.), Proteomic and Genomic Analysis of Cardiovascular Disease. Wiley-VCH, Weinheim, Germany, pp. 3–60. Schwartz, F., Duka, A., Duka, I., Cui, J., Gavras, H., 2004. Novel targets of ANG II regulation in mouse heart identified by serial analysis

PART  |  11  Cardiomics

of gene expression. Am. J. Physiol. Heart. Circ. Physiol. 287, H1957–H1966. Smart, N., Risebro, C.A., Melville, A.A., Moses, K., Schwartz, R.J., Chien, K.R., Riley, P.R., 2007. Thymosin beta4 induces adult epicardial progenitor mobilization and neovascularization. Nature 445, 177–182. Smyth, G.K., Speed, T.P., 2003. Normalization of cDNA microarray data. Methods 31, 265–273. Srivastava, D., 2006. Making or breaking the heart, from lineage determination to morphogenesis. Cell 126, 1037–1048. Srivastava, D., Thomas, T., Lin, Q., Kirby, M.L., Brown, D., Olson, E.N., 1997. Regulation of cardiac mesodermal and neural crest development by the bHLH transcription factor, dHAND. Nat. Genet. 16, 154–160. Tabibiazar, R., Wagner, R.A., Liao, A., Quertermous, T., 2003. Transcriptional profiling of the heart reveals chamber-specific gene expression patterns. Circ. Res. 93, 1193–1201. Utans, U., Liang, P., Wyner, L.R., Karnovsky, M.J., Russell, M.E., 1994. Chronic cardiac rejection: Identification of five upregulated genes in transplanted heart by differential mRNA display. Proc. Natl. Acad. Sci. USA 91, 6463–6467. Van Gelder, R.N., von Zastrow, M.E., Yool, A., Dement, W.C., Barchas, J.D., Eberwine, J.H., 1990. Amplified RNA synthesized from limited quantities of heterogenous cDNA. Proc. Natl. Acad. Sci. USA 87, 1663–1667. Velculescu, V.E., Zhang, L., Vogelstein, B., Kinzler, K.W., 1995. Serial analysis of gene expression. Science 270, 484–487. Venter, J.C., Adams, M.D., Myers, E.W., Li, P.W., Mural, R.J., Sutton, G. G., Smith, H.O., Yandell, M., Evans, C.A., et al., 2001. The sequence of the human genome. Science 291, 1304–1351. von Both, I., Silvestri, C., Erdemir, T., Lickert, H., Walls, J.R., Henkelman, R.M., Rossant, J., Harvey, R.P., Attisano, L., Wrana, J.L., 2004. Foxh1 is essential for development of the anterior heart field. Dev. Cell 7, 331–345. Wang, J., Hu, L., Hamilton, S.R., Coombes, K.R., Zhang, W., 2003. RNA amplification strategies for cDNA microarray experiments. Biotechniques 34, 394–400. Woods, D., Crampton, J., Clarke, B., Williamson, R., 1980. The construction of a recombinant cDNA library representative of the poly(A) mRNA population from normal human lymphocytes. Nucleic. Acids Res. 8, 3187–3198. Wright, W.E., Sassoon, D.A., Lin, V.K., 1989. Myogenin, a factor regulating myogenesis, has a domain homologous to MyoD. Cell 56, 607–617. Ye, S.Q., Usher, D.C., Zhang, L.Q., 2002. Gene expression profiling of human diseases by serial analysis of gene expression. J. Biomed. Sci. 9, 384–394. Zhao, X., Gallardo, T.D., Lin, L., Schageman, J.J., Shohet, R.V., 2002. Transcriptional mapping and genomic analysis of the cardiac atria and ventricles. Physiol. Genom. 12 (1), 53–60. Zimmerman, J.W., Schultz, R., 1994. Analysis of gene expression in the preimplantation mouse embryo, use of mRNA differential display. Proc. Natl. Acad. Sci. USA 91, 5456–5460.

Chapter 11.2

Exploring the Genetic Basis for Congenital Heart Disease with Mouse ENU Mutagenesis Cecilia W. Lo1, Qing Yu1, Yuan Shen1, Linda Leatherbury1, Richard Francis1, Xiao-Qing Zhao1, Zhen Zhang1, Andy Wessels2, Guo-Ying Huang3 and Bishwanath Chatterjee1 1

Laboratory of Developmental Biology, National Heart Lung Blood Institute, National Institutes of Health, Bethesda, MD, USA Deparment of Anatomy and Cell Biology, Medical University of South Carolina, Charleston, SC, USA 3 Pediatric Heart Center, Children’s Hospital of Fudan University, Shanghai, China 2

I.  Congenital heart disease Congenital heart disease is one of the most common birth defects, affecting up to 1% of live births (Hoffman, 1995a,b; Rosenthal, 1998; Hoffman and Kaplan, 2002; Hoffman et al., 2004). With rapid technical advances allowing early diagnosis and surgical repair of even the most complex congenital heart disease, most children with congenital heart disease are now surviving to adulthood. As a result, the number of adults with congenital heart disease has surpassed that of children, and is increasing at a rate of 5% per year (Brickner et al., 2000a,b; Williams et al., 2006). Adults surviving with congenital heart disease exhibit variable outcomes, with many continuing to have medical issues requiring lifelong care. To improve the long-term prognosis of children and adults with congenital heart disease, elucidating the etiology is essential, as only then will it be possible to stratify patients appropriately to optimize therapeutic strategies. The causes of congenital heart disease are complex (Strauss, 1998; Garg, 2006). A genetic contribution is clearly seen with chromosomal abnormalities such as in DiGeorge syndrome, where cardiac defects are found in conjunction with other birth defects due to haploinsufficiency associated with chromosome 22q11 microdeletion. Most human congenital heart diseases, however, are sporadic and nonsyndromic occurring in isolation from other birth defects. A genetic contribution is nevertheless indicated by the finding of increased recurrence risk in families with a history of congenital heart disease (Boughman et al., 1993). More recently, Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

human genetic studies have identified a number of genes contributing to sporadic and familial forms of congenital heart disease, several of which are transcription factors including TBX1, TBX5, TBX18, NKX2-5 and others (Gruber and Epstein, 2004; Garg, 2006; Srivastava, 2006; Andelfinger, 2008; Bruneau et al., 2008). A role for Notch signaling is also indicated with the finding of NOTCH1 mutations in patients with bicuspid aortic valves (Garg et al., 2005), and mutations in NOTCH2 and JAGGED1, the latter encoding a Notch ligand, in patients with Alagille syndrome (Li et al., 1997; Oda et al., 1997; McDaniell et al., 2006). The involvement of the Ras-MAP kinase pathway is indicated with the finding of mutations in SHP-2, ki-Ras, RAF1 and SOS1 in Noonan’s syndrome (Gelb and Tartaglia, 2006; Razzaque et al., 2007; Roberts et al., 2007). The genetics of human congenital heart disease are complex. Human congenital heart disease is often associated with variable penetrance, so that not all individuals with a disease-causing mutation will have the same structural heart disease. Variable expressivity is also observed, where the same mutation can present with different congenital heart disease in different patients. Finally, gene–environment interactions can affect disease risk, further complicating the analysis. The complex genetics associated with human congenital heart disease have been difficult to unravel, given the inherent genetic heterogeneity of the human population. This makes a compelling case for the use of geneticallyinbred animal models to pursue studies to elucidate the genetic etiology of congenital heart disease. 753

754

PART  |  11  Cardiomics

Drosophilia heart

Chick heart

Common carotid

Left subclavian artery

Right subclavian artery

Heart cell

Right brachiocephalic artery Right aortic arch

Pericardial cell

RA Fish heart

Left bracheocephalic artery

LA

Main pulmonary artery

Atrium Ascending aorta

Truncus

AV valve

Tricuspid valve

Ventricle

Figure 1  Structure of the heart in model organisms. The Drosophila heart is a one-chambered, cell-lined tube. The zebrafish heart has two chambers, with an atrioventricular (AV) valve between the single atrium and the pumping ventricle with a single outflow vessel, the truncus. Neither of the hearts have separate circulation for routing deoxygenated and oxygenated blood. In contrast, the chick heart is four-chambered with two distinct atrioventricular valves, two great arteries with semilunar valves, and an atrial and ventricular septum that separates circulation of deoxygenated (blue) and oxygenated (red) blood on the right versus the left side of the heart. It differs from the human heart in that the right ventricular pumping chamber is smaller with a more muscular valve, and the aortic arch is right-sided compared to the human and mouse hearts, which have a left-sided aortic arch.

II.  Modeling congenital heart disease in mice Many animal models have been used to study cardiovascular development (Fig. 1). Studies in Drosophila yielded the first clue to the importance of tinman, the Drosophila homolog of Nkx2-5, as an essential transcription factor in heart development. Embryological experiments with cell ablations in chick embryos gave the first evidence that outflow tract septation requires a migratory cell population derived from the dorsal hindbrain neural fold, coined the cardiac neural crest cells (Hutson and Kirby, 2007). More recently, studies in zebrafish embryos have provided direct evidence for the importance of flow in cardiac

­ orphogenesis, as the viability of zebrafish embryos does m not require a beating heart or intact circulatory system (Hove et al., 2003). For studying the genetic origin of congenital heart disease, it is important that the animal model fulfill several criteria: it must have a cardiovascular anatomy similar to humans; genetically-homogenous inbred strains must be available; the genome of the animal model must be well-characterized and show good conservation to the human genome. The mouse fulfills all of these criteria. Its genome is completely sequenced, and inbred mouse strains comprising genetically-identical individuals are a perfect backdrop for analyzing the genetic basis for congenital heart disease. Mice, like humans, have left–right asymmetric four-chambered hearts and separate systemic

Chapter  |  11.2  Exploring the Genetic Basis for Congenital Heart Disease with Mouse ENU Mutagenesis

Human heart Left aortic arch

Right superior vena cava

Right superior vena cava Coronary sinus

LA

Pulmonary artery

3 RA 1

2

755

Mouse heart Left aortic arch Ascending aorta

LA

Main pulmonary artery

Left superior vena cava

RA

4

LV RV

LV RV

Inferior vena cava

Right atrial appendage

Figure 2  Comparison of the human and mouse heart. The human heart has four chambers – the right atrium (RA), left atrium (LA), right ventricle (RV), and left ventricle (LV). The two ventricles give rise to two great arteries, the pulmonary artery from the right ventricle and aorta from the left ventricle. There are four heart valves: the tricuspid valve between the right atrium and right ventricle (1); the pulmonary valve at the base of the pulmonary outflow (2); the mitral valve (3) separating the left ventricle and left atrium; and the aortic valve (4) at the base of the aortic outflow. This fourchambered heart structure gives rise to two circulations connected in series so that deoxygenated (blue) blood returning from the body is pumped via the pulmonary artery to the lungs via the right ventricle, while oxygenated (red) blood is pumped from the left ventricle via the aorta systemically to the body. The mouse heart has an identical four-chambered heart with two great arteries, but unlike humans, which have only a right superior vena cava, the murine heart has an additional left superior vena cava that empties into the coronary sinus. This is only infrequently seen in humans. Also, the mouse atrial appendages are larger and do not have distinctive right versus left trabeculations and shapes as seen in humans. Both mice and humans have a left aortic arch; while chick exhibits a right aortic arch (see Fig. 1).

versus pulmonary circulation that are the main substrates for congenital heart disease (Fig. 2). Most of the serious congenital heart diseases causing significant morbidity and mortality involve defects in partitioning of the pulmonary and systemic circulation essential for the efficient oxygenation of blood (Figs 3; 4). Knockout mouse models have successfully phenocopied many of the congenital heart diseases observed clinically. In knockout mice, the function of one gene can be assayed in the absence of any other genetic variation. This allows for the systematic analysis of complex genetic interactions, such as that observed between genes encoding transcription factors essential for heart development. Thus, knockout mouse models of Nkx2-5, Gata4 and Tbx5, and intercrosses between them have yielded new insights into the synergistic interactions between these three transcription factors essential for heart development (Bruneau et al., 2001; Hiroi et al., 2001; Olson, 2006; Srivastava, 2006). Recent studies further suggest that Tbx5 genetically interacts with Gata6 (Maitra et al., 2009). However,

whereas mutations associated with congenital heart disease in patients are usually presented in heterozygosity, i.e., autosomal dominant, in mice congenital heart disease is usually only seen in homozygous-knockout animals, with heterozygous animals showing either no phenotype or only milder defects. This discrepancy may reflect an inherent difference in the genetics of mice versus humans. Alternatively, this may simply reflect ascertainment bias, as homozygosity may be rare in the genetically-diverse human population, and when present may cause prenatal or neonatal lethality. This would be consistent with the notion that diploid genomes are evolutionarily protective, with null alleles of most genes being heterozygous viable (Charlesworth, 1991; Kondrashov and Crow, 1991; Perrot et al., 1991). In fact, the incidences of congenital heart disease in human fetuses that die prenatally are reported to be as high as 39% (median 7.3%), indicating tracking congenital heart disease in live births only grossly underestimates the true incidence of congenital heart disease from the time of conception (Hoffman, 1995a,b).

756

PART  |  11  Cardiomics

Normal

Ventricular septal defect

Aorta

Pulmonary artery

LA

Atrial septum

Hole between chambers

RA LV RV

Double-outlet right ventricle

Atrioventricular canal

Common AV valve

VSD

Both great arteries arise from RV

RV

Holes between chambers

Tetralogy of Fallot

Pulmonary atresia Narrowed artery

Patent ductus arteriosis Underdeveloped main pulmonary artery

Underdeveloped artery

Narrow opening

Aorta just above hole

Closed valve

Hole between chambers

Ventricular septal defect

Enlarged

Figure 3  Congenital heart defects. Ventricular septal defect: The most common congenital heart defect that requires surgery is the ventricular septal defect (VSD). Blood “shunts” left-to-right through a hole in the septum, causing increased blood flow to the lungs and the mixing of oxygenated and deoxygenated blood. Atrioventricular canal: The atrioventricular canal, also known as an atrioventricular septal defect, has one inflow and a common atrioventricular valve with left-to-right shunt through the atrial and ventricular septal defects. Double-outlet right ventricle: Both the pulmonary artery and the aorta arise from the right ventricle. The “blue” and “red” blood mix in the right ventricle, and thus the aortic blood is not fully-oxygenated and the patient is cyanotic. Tetralogy of Fallot: This is the most common cyanotic congenital heart defect. The aorta overrides the ventricular septal defect and the pulmonary outflow is narrowed, referred to as pulmonary stenosis. In addition, there is right ventricular hypertrophy that is thought to arise secondary to the ventricular septal defect and pulmonary stenosis. The obstruction to pulmonary flow can occur at several levels – the infundibular muscle below the valve, the small valve annulus, the thick valve leaflets and the narrow pulmonary arteries. Due to the pulmonary obstruction, the blood shunts right-to-left through the ventricular septal defect and out the aorta, causing deoxygenated (blue) blood to flow to the body. Pulmonary atresia: This is a severe form of tetralogy of Fallot, with no opening in the pulmonary valve. For a newborn to survive, the patent ductus arteriosus (PDA) must be kept open with intravenous medications followed by palliative surgery (detailed description of CHD seen in Chapter 3.4).

Persistent truncus arteriosus Aorta

Persistant truncus arteriosis with interrupted aortic arch Interrupted aortic arch

Truncus overriding VSD

Patent ductus arteriosis

LA

LA

RA

RA

One valve

Ventricular septal defect

LV

LV

RV

RV

Transposition of great arteries

Corrected transposition of great arteries Aorta

Aorta Pulmonary artery

Pulmonary artery

LA LA

Coronary arteries

RA LV Atrial septal defect

RA

LV

RV

RV

Atrial septum

Dextrocardia

Mesocardia with superior–inferior ventricles Right superior vena cava

Superior vena cava Pulmonary artery

Right aortic arch

LA RA

Pulmonary artery LV

RA Pulmonary veins

RV Inferior vena cava

Inferior vena cava

LA

RV

LV Inferior vena cava

Figure 4  Congenital heart defects. Persistent truncus arteriosus: This is a cyanotic heart disease with a ventricular septal defect, one outflow valve and one great artery arising from the heart instead of two. This single “truncal” artery gives rise to all three circulations of the body – systemic to the body through the aorta off the truncus, pulmonary arteries which also arise from the truncal artery to the lungs, and the coronary arteries that emerge from the base of the truncal artery. This defect arises from the persistence of the single truncal outflow found early in embryonic development, which normally undergoes septation to form the two great arteries – the aorta and pulmonary artery. Persistent truncus arteriosus with interrupted aortic arch: Persistent truncus arteriosus is frequently complicated by interruption of the aorta where a persistent ductus arteriosus provides blood flow to the lower body and is required for neonatal survival. Transposition of the great arteries (TGA): this produces very severe cyanosis of a newborn. The two great arteries are switched and thus the systemic venous, deoxygenated blood flows from the right atrium and right ventricle back to the body, while the oxygenated blood flows from the lungs through the left atrium and left ventricle and back out to the pulmonary artery and to the lungs. The two circulations are in parallel, not in series as in the normal heart. For immediate survival a hole (atrial septal defect) is made between the right atrium and left atrium to allow the mixing of red and blue blood, and therefore a small amount of oxygenated blood flows to the body. In “corrected” transposition of the great arteries, the great arteries are not switched. The blood is physiologically flowing through the heart normally. The defect is that the ventricular morphology is switched such that the right ventricle is a smooth walled chamber and the left ventricle is a trabeculated chamber. Dextrocardia: This means that the heart is positioned in the right chest with its apex towards the right, instead of in the left chest with the apex towards the left. Frequently there is “mirror-image” dextrocardia with the entire heart structures reversed, but the blood is physiologically flowing normally. Mesocardia with superior–inferior ventricles: This is a rare and very complex heart defect, where the heart is midline in the chest and the ventricular septum is oriented horizontally such that the right ventricle is superior to the inferior left ventricle. The systemic venous deoxygenated (blue) blood comes to the right atrium, and the pulmonary venous oxygenated (red) blood comes to the left atrium and then all the blood mixes in the ventricles and exits the heart through a double outlet right ventricle to the aorta and pulmonary artery.

758

III.  Forward genetic screens To elucidate the genetic basis for congenital heart disease, we conducted a forward genetic screen in mice focused on the recovery of mutations causing congenital heart disease. The power of forward genetic screens is well-illustrated by the remarkable success of the Drosophila chemical mutagenesis screen of Nusslein-Volhard and Wieschaus (NussleinVolhard and Wieschaus 1980). They elegantly showed the efficacy of phenotype-driven mutagenesis screens for elucidating the genetic pathways regulating developmental patterning of the segmental body plan. Similar large-scale screens to analyze the genetic regulation of early development in vertebrates have utilized the zebrafish model system. Zebrafish is advantageous for developmental screens, given that its embryos develop externally, have optical clarity and can be obtained in large numbers. Large-scale zebrafish developmental screens have been successfully conducted using either insertional or chemical mutagenesis (Mullins et al., 1994; Driever et al., 1996; Haffter et al., 1996; Amsterdam et al., 2004; Sun et al., 2004). However, the zebrafish model system is not suitable for studying congenital heart disease, given that fish do not have the cardio­ pulmonary adaptations for gas exchange from air which are the major targets of congenital heart disease. Thus, fish have a two-chambered heart with a single inflow and outflow (Fig. 1). Also problematic is the significant maternal contribution in fish embryos, and the partial tetraploidy of its genome due to a genome duplication event (Woods et al., 2000). These factors likely explain why only 1,400 embryonic-lethal mutations are projected for zebrafish (Amsterdam

PART  |  11  Cardiomics

et al., 2004), as compared to 3,500–5,000 expected in mice (Wilson et al., 2005). The challenge with developmental screens in mice is the inaccessibility of the embryos and the small litter sizes – typically 6–8 embryos per litter for inbred strains. Nevertheless, with well-defined phenotyping protocols in focused screens, developmental mutants in mice have been successfully recovered (Garcia-Garcia et al., 2005; Zohn et al., 2005; Caspary and Anderson, 2006). Moreover, many embryonic-lethal mutations have been obtained in several elegant balancer chromosome screens (see for example Kile et al., 2003; Wilson et al., 2005). Typically, phenotypic screens for developmental mutants have involved harvesting the embryos following euthanasia of the mother. With balancer screens, prenatal lethality of homozygote animals have been imputed from missing progeny (with specific coat color) (Kile et al., 2003; Wilson et al., 2005). Our screen was conducted in the C57BL6/J (B6) inbred strain background, and was designed to recover recessive mutations (Yu et al., 2004). This entailed a two-generation backcross breeding scheme (Fig. 5), with G2 females backcrossed to their G1 fathers derived from ENU mutagenized G0 males. Therefore, recessive mutations induced in the G0 male may be presented in homozygosity in the G3 animals. Four pregnant G2 females from each family were ultrasound scanned to look for G3 fetuses with cardiovascular defects (Fig. 5). The mutations were tracked via the G1 male, with all offspring derived from a single G1 male considered a pedigree or family. The design of this breeding scheme would ensure at least 20 G3 fetuses would be generated from each G1 male, which would provide a 92% Figure 5  Recessive backcross breeding scheme. ENU-mutagenized G0 male was bred with female mice to produce G1 male offspring, and these are further bred to produce G2 females. The G2 females are then backcrossed with their G1 fathers to produce G3 offspring, some of which will be homozygous for ENUinduced mutations generated in the G0 male. Ultrasound screening is conducted by in utero scanning of the G3 fetuses. The mutations are tracked via the G1 male, such that all offspring from a single G1 male are designated as a family.

Chapter  |  11.2  Exploring the Genetic Basis for Congenital Heart Disease with Mouse ENU Mutagenesis

chance that a mutation in the G1 male would have been seen once, and 71% chance that it would be seen two or more times in the G3 progeny (Silver et al., 2007).

IV.  Mouse fetal echocardiography screening For cardiovascular phenotyping, ultrasound imaging can be used for noninvasive in utero interrogation of fetal cardiovascular structure and function by echocardiography. Fetal echocardiographic studies in humans utilizing twodimensional, M-mode and pulsed Doppler technology have become an efficient screening tool for detecting human fetuses with congenital heart disease. Using such a clinical ultrasound system and the same principles guiding the clinical diagnosis of congenital heart disease, we developed a noninvasive and very high-throughput cardiovascular phenotyping protocol (Yu et al., 2004; Shen et al., 2005). The same mouse fetuses are scanned many times over the course of development, allowing the tracking of disease progression and the detection of prenatal/neonatal death. In other mutagenesis screens for mouse developmental mutants, pregnant female mice are euthanized for harvesting the embryos and mutants are identified with visual screening for developmental malformations or changes in expression of marker gene expression patterns. A recessive screen is expected to be more productive and cost-effective, as studies in many diploid species have shown recessive mutations are more frequent, with some studies showing 90% of all mutations are recessive (Vogel and Motulsky, 1986; Jaenisch, 1988; Friedrich and Soriano, 1991; Wilkie, 1994). In addition, as serious congenital heart disease may be inviable in the absence of medical support or palliative surgery, heterozygous mutations causing serious congenital heart disease in mice will not be recoverable. Each litter was scanned three times, at gestation day 12.5 (E12.5), E14.5, E16.5 or at E13.5, E15.5 and E17.5 (E18.5–19.0 being term). The screening was initiated at E12.5, as this is the time when structural heart defects relevant to human congenital heart disease become evident, with the completion of aortic arch remodeling and ventricular and outflow septation. For litters where abnormal echo presentations were observed, additional scans were conducted on the interim days. The screening was conducted by ultrasound imaging using a clinical ultrasound system, the Acuson Sequoia. The noninvasive nature of the fetal ultrasound screen is advantageous, not only because it allows cardiovascular function to be assessed under physiological conditions, but also the mother and her surviving pups need not be sacrificed for the phenotyping. Screening prenatally also greatly improves the efficiency in recovering fetuses with heart defects, as the same fetuses can be scanned many times over the course of development, thereby allowing the tracking of disease progression and the detection of ­ prenatal/neonatal lethality. The screen was conducted

759

using the ultra-high-frequency 15 MHz L8 linear transducer, which provides excellent two-dimensional images (Fig. 6 A,B,H–K), color-flow (Fig. 6A,B), spectral Doppler (Fig. 6C,D) and M-mode images (Fig. 6F–G). Traditional clinical ultrasound imaging planes can be used such as apical 3, 4 and 5 chamber views and left ventricular short- and long-axis views (Fig. 7) (Yu et al., 2004, 2008). However, we found the use of orthogonal imaging planes defined relative to the fetus vertebral column and body axes more suitable for high-throughput mouse fetal cardiovascular imaging. The latter structures are easily visualized by ultrasound (Fig. 8) (Shen et al., 2005). This markedly improved the throughput and accuracy of our screen, allowing a litter of eight fetuses to be scanned in less than 10 minutes. All of the major structures of the cardiovascular system can be interrogated using these three imaging planes (Table 1).

IV.A.  Sagittal Views Sagittal imaging planes are obtained when the vertebral column of the fetus is captured in the long-axis view, showing a typical fetal position (Fig. 8B). The fetus is then interrogated in the sagittal plane by angulating the transducer in small increments, while focusing on the heart and great vessels. Two-dimensional imaging in this plane, together with color-flow Doppler mapping of blood flow, allows the visualization of the right inflow tract and outflow tract (Fig. 9). In this view, the crossing of the pulmonary trunk and ascending aorta also can be observed, which is often perturbed with misalignment of the great arteries or outflow tract septation defects. Although the valves cannot be visualized, color-flow Doppler nevertheless facilitates the detection of retrograde versus anterograde flow across the tricuspid and pulmonary valves. This imaging plane can also provide a longitudinal view of the ductus arteriosus, allowing color-flow mapping for viewing normal right-to-left anterograde flow or abnormal leftto-right retrograde flow in the ductal arch.

IV.B.  Frontal Views Frontal views are obtained when the fetus is seen in its long axis, and the body and head are symmetrical (Fig. 8D). In this view, the heart is sectioned in the frontal plane, and both ventricles and interventricular septum can be observed (Fig. 10A). Color-flow mapping allows delin­ eation of the left outflow tract and blood inflow into both ventricles and, together with spectral Doppler analysis, can detect regurgitant flows in the left outflow or inflow tract (see also Section V). A two-dimensional ultrasound image in the frontal plane is shown in Fig. 10A, which compares favorably with the histological section shown in Fig. 10B, showing long, narrow, right ventricular outflow tract and egg-shaped left ventricle. Color-flow mapping in

760

PART  |  11  Cardiomics

Figure 6  Ultrasound imaging modalities used for screen. (A–D). Blue inflow from left atrium (LA) to left ventricle (LV) was seen by color-flow imaging (A) in the long axis view. Red outflow from left ventricle to aorta (Ao) was seen by color-flow imaging (B) Spectral Doppler tracings guided by the color-flow imaging showed the typical pattern of inflow (E) and (A) waves, and outflow (C). Diagrammatic representation of spectral Doppler tracings (D) showed the position of inflow and outflow, and time intervals used for the calculation of the myocardial performance index (MPI), including isovolemic ventricular contraction time (IVCT), isovolemic ventricular relaxation time (IVRT) and ejection time (ET). (E–G). M-mode images showed position of ventricular wall and septum in diastole and systole through multiple cardiac cycles (E). Wall thickness and chamber dimensions in diastole (F) and systole (G) were obtained by measuring the distances between the waveforms in magnified M-mode images. (H–K). Two-dimensional images in left ventricle long-axis view in diastole (H) and systole (J). Ejection fraction and fractional area change were calculated from the left ventricle area tracing in diastole (I) and systole (K). Figure reproduced from Yu et al. (2008).

Ao LA RA LV

RV

Apical 5 chamber view

Ao

LA

RA

LV

RV

Apical 4 chamber view

LA

RA

RV Ao

LA

LV

DAo

Apical 3 chamber view

RV LV

DA

Great vessels short-axis view

Ao LA

Left ventricular long-axis view

RV LV

Ventricular short-axis view

Figure 7  Clinical ultrasound imaging planes used for cardiovascular assessments of fetal mice. Diagnostic ultrasound imaging planes similar to those used clinically for cardiovascular assessments are shown with two-dimensional ultrasound images of the fetal mouse heart and accompanying diagrams. These include apical 3, 4 and 5 chamber views, left ventricular long-axis view, ventricular short-axis and great vessels short-axis views (LA, left atrium; LV: left ventricle; Ao: aorta; RA: right atrium; RV: right ventricle; PA: pulmonary artery; DA: ductus arteriosus; DAo: descending aorta). Figure reproduced from Yu et al. (2008).

Chapter  |  11.2  Exploring the Genetic Basis for Congenital Heart Disease with Mouse ENU Mutagenesis

(A) 140

8

100 80

6

60

4

40

Heart area (mm2)

Fetus area (mm2)

120

Figure 8  Ultrasound imaging planes guided by the vertebral column and body axes of the fetus. (A) Two-dimensional ultrasound image in the sagittal or frontal plane was used to measure the crownto-rump length, fetus area and the heart area. Fetus area and heart area were highly correlated to the crown-to-rump length. (B, C, D) Two-dimensional ultrasound images obtained in utero showing an embryonic day 17.5 (E17.5) fetus in the sagittal (B) and E16.5 fetuses in transverse (C) and frontal planes (D). The vertebral column (V) can be clearly seen in the sagittal plane (B), while along the frontal plane (D) the ribs are symmetric­ ally positioned on either side of the heart (H). Distance between markers on x- or y-axis in (C) is 5 mm. Scale bar  2.5 mm. Part of the figure reproduced from Yu et al. (2008).

(B)

10

Fetus area, r=0.98 Heart area, r=0.94

2

20

0

0 5

10

15

20

Crown-to-rump length (mm)

(C)

761

(D)

Table 1  Structures Seen via Orthogonal Imaging Planes Sagittal plane

Frontal plane

Transverse plane

Inferior vena cava–right ventricle inflow tract

Right ventricle outflow tract and left ventricle

Bifurcation of pulmonary artery

Pulmonary artery trunk

Ascending aorta in long axis

Short axis of pulmonary artery/aorta

Ductus arteriosus

Left ventricle and right ventricle

Left ventricle and right ventricle

Aortic arch Left ventricle outflow tract

this imaging plane allows visualization of the left outflow tract or aorta (red) together with a cross-sectional view of the pulmonary artery (blue) (Fig. 10C). Also seen is blue color-flow into the ventricles during diastole (Fig. 10E). Corresponding histological sections are shown in Fig. 10D and F, with spectral Doppler tracings confirming the colorflow mapping (Fig. 10G–I). It is in this imaging plane that we manipulate the transducer to obtain M-mode images, an imaging modality that provides the most accurate quantitative measures of ventricular chamber dimensions and wall thicknesses. This is achieved by angulating the transducer from the initial frontal plane to position the sample volume perpendicular to the interventricular septum, while maximizing ventricular diameter.

IV.C.  Transverse Views A transverse imaging plane corresponds to a short-axis view of the fetus (Fig. 8C). This view is the least informative and most difficult to acquire, requiring a great deal of time for appropriate positioning of the transducer. Examples of two-dimensional images in the transverse view are shown from base to apex in color-flow images shown in Fig. 11A, C and E, and these are compared with corresponding histological views in Fig. 11B, D and F. It should be noted that color-flow is especially important for mapping heart structures in this view, with spectral Doppler analysis needed for further confirmation of the position of inflow versus outflow (Fig. 11G,H). In Fig. 11H, the spectral

762

PART  |  11  Cardiomics

(A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

(I)

(J)

Figure 9  Ultrasound images obtained using the sagittal imaging plane. Color-flow mapping with ultrasound imaging in the sagittal view from right to left side of the fetus is presented for an E17.5 fetus in (A–F), with gold, square-shaped symbols indicating transducer position. Comparable histological sections are also presented from right to left side of the fetus in (G–J). (A) Blood inflow during systole from the ductus venosus (DV; red) to the inferior vena cava (IVC; blue). The ascending aorta (AAO) is seen as red, as it flows cranially. (B) In diastole, IVC flow is shown in red, reflecting reverse blood flow from the right atrial (RA) contraction, with the right ventricular inflow delineated in blue from right atrial blood flow. (C) Long-axis of the pulmonary trunk (PA; blue) and the perpendicular crossing of the pulmonary artery and aorta (AO; red) are shown. (D) Longitudinal view of the ductus arteriosus arch resembles a hockey stick with flow into the descending aorta (DAO). Because the color-flow gate is wide compared with the ductal arch and is perpendicular to the arch, all of the color-flow appears red. (E) Longitudinal view of the aortic arch has a circular shape, resembling a candy cane. (F) A more oblique sagittal plane shows a long-axis view of the left ventricle. The color blooming blurs the interventricular septum, causing blood flow in the right ventricle (RV) and left ventricle (LV) to merge. (G, H) Histological sections delineate the connection of the inferior vena cava with the right atrium and the right atrium with the right ventricle. These sections can be compared with images (A) and (B), in which similar structures are profiled by color-flow. Also, (H) shows the anatomy and crossing relationship of the pulmonary artery and aorta, resembling the ultrasound image in C. (I) Longitudinal view of the ductus arteriosus arch, which is similar to the ultrasound image in (D). (J) A histological section in oblique sagittal plane that is similar to the ultrasound image in F (LA: left atrium; PA: pulmonary artery; RPA: right pulmonary artery; RA: right atrium). Figure reproduced from Shen et al. (2005).

Doppler suggests an inflow pattern for the color-flow seen in Fig. 11E. The spectral Doppler pattern for the outflow in Fig. 11C is shown in Fig. 11G. In this view, is not possible to obtain spectral Doppler tracings with maximal velocity, as the imaging plane is perpendicular to the great arteries.

V.  Ultrasound detection of cardiovascular defects We observed all of the major congenital heart diseases found clinically from scanning over 13,000 fetuses from nearly 500 mouse pedigrees in a recessive mutagenesis screen (Table 2). This corresponds approximately to a 50% genome

coverage. Prenatal death peaked between E13.5–16.5, with the majority of fetuses (74%) expiring with severe cardiovascular defects dying at E15.5–E17.5 (Fig. 12). Given the limited resolution of ultrasound imaging and the small size of the fetal mouse heart, the diagnosis of cardiovascular defects are made in broad categories that include arrhythmia, outflow regurgitation, outflow velocity increase (in the absence of regurgitant flow), heart failure, hypertrophy and ectopia cordis (heart outside chest cavity) (Table 3). Arrhythmia was easily observed by spectral Doppler analysis, and included premature systole, pause, bradycardia and tachycardia. However, bradycardia accounted for nearly half (45%) of the arrhythmias. This is frequently a manifestation of dying fetuses. Consistent with this, we note

Chapter  |  11.2  Exploring the Genetic Basis for Congenital Heart Disease with Mouse ENU Mutagenesis

(A)

(B)

(C)

(D)

(E)

(F)

(A)

(C)

763

(B)

(D)

(E) (F)

(G)

(H)

(I)

Figure 10  Ultrasound images using frontal views. Two-dimensional ultrasound imaging in frontal views shows the top of the LV, the interventricular septum, and the right ventricle outflow tract (RVOT). In the histology image in (B), the left ventricle can be seen in an egg-shaped configuration, while the RVOT is seen as a half moon. (C, D) Color-flow imaging (C) delineates the aorta in red and the cross-section of the pulmonary artery in blue. Similar anatomy is seen in the histological image (D). (E, F) Color-flow (E) in blue depicts two ventricles, and this same view is seen in a corresponding histological section (F). (G–I) Spectral Doppler analysis was used to confirm the position of the outflow and inflow. Correlating with the color-flow in (C) is the spectral Doppler in (G) that shows the maximal aortic velocity, while (H) shows a spectral Doppler for the PA. The inflow spectral Doppler in (I) was obtained from the blue color-flow in (E). The heart rate in this fetus was determined to be 230 beats per minute (bpm). Figure reproduced from Shen et al. (2005).

that 14% of fetuses with heart failure also exhibited bradycardia. Heart failure was discerned by dynamic two-dimensional imaging, and is characterized by poor contractile function, pericardial effusion and fetal hydrops. Contractile motion of the beating heart was assessed qualitatively using two-dimensional cine sequences, and quantitatively with

(G)

(H)

Figure 11  Ultrasound imaging using transverse views. Ultrasound images in the transverse planes (A, C, E) and comparable histological sections (B, D, F) are presented from rostral to caudal in an E16.5 fetus. (A, B) The pulmonary artery crossing over the aorta in cross-section and bifurcating into two PA branches. (C, D) The aorta posterior to the pulmonary artery. (E) Both ventricles by color-inflow mapping in diastole. (F) The corresponding histological image. Color-flow mapping with spectral Doppler confirmation of outflow and inflow waveforms is presented in (G) and (H). The heart rate for this fetus was determined to be 204 bpm (PV: pulmonary vein). Figure reproduced from Shen et al. (2005).

measurements of ejection fraction, fractional area change and shortening fraction. In many instances, heart failure was confirmed with examination of the expired fetus, which typically showed congested heart, lung and liver. Outflow

764

PART  |  11  Cardiomics

Table 2  Ultrasound Detection of Developmental Defects in ENU Mutagenized Mice # G1 Males (families)

# G2 females*

Total fetuses¥

Abnormal fetuses¥1

477

1,619

13,405

902

Prenatal lethal

188 (39%)

284 (17%)

2,039

405 (45%)

16%

Growth retardation†2

121 (26%)

158 (10%)

1,032

195 (22%)

19%

Cardiovascular defects

224 (47%)

372 (23%)

2,642

541 (60%)

20%

Craniofacial defects

22 (4.6%)

26 (1.6%)

181

30 (3.3%)

17%

Abdominal wall defects

9 (1.9%)

10 (0.6%)

78

16 (1.8%)

20%

Total screened †1

% in affected litters¥2

*Number G2 females with abnormal fetuses, except Total screened  total G2 females scanned. ¥ Total fetuses found in litters; ¥1Number in parenthesis  % relative to total abnormal fetuses; ¥2% relative to total # G3 fetuses in affected litters; †1 Prenatal lethality was found in 74 families without any known cardiovascular defects. †2 Growth retardation was seen in 20 families without any known cardiovascular defects.

60% 50 40 30 20 10 0 E12.5

E13.5–E14.5 E15.5–E16.5 Embryo stage

Prenatal lethality

E17.5

Prenatal lethality from CHD

Figure 12  Incidence of prenatal lethality detected by fetal ultrasound. Prenatal lethality is observed throughout the period of ultrasound scanning spanning E12.5 to E17.5, but prenatal lethality associated with cardiovascular defects is largely observed only between E13.5 to E17.5, with a peak at E15.5–E16.5. At E15.5–16.5, 38% of the total deaths are associated with severe cardiovascular defects.

regurgitation was identified first with qualitative assessments using color-flow, followed by quantitative analysis using spectral Doppler to assess for abnormal diastolic flow indicative of semilunar valve regurgitation. This was observed in 229 fetuses from 101 families, corresponding to 43% of fetuses identified with heart defects. All fetuses exhibiting outflow regurgitations and/or signs of heart failure eventually die, either prenatally or at birth. The rarest defect was ectopia cordis, a condition where the heart develops outside the chest cavity. Hypertrophy was detected using M-mode imaging and is defined as those fetuses with wall thickness more than two standard deviations from the mean. Due to difficulty in achieving the optimal imaging planes, M-mode data were obtained for less than 25% of the fetuses screened, and thus hypertrophic change is under-represented in the screen.

Nevertheless, 34% of fetuses identified with heart defects exhibited evidence of hypertrophy (Table 3). This was observed in 181 fetuses from 102 families. Such high incidence of hypertrophy was unexpected, since hypertrophy is usually thought of as an adult-acquired disease. However, early-onset hypertrophy is observed in Noonan’s syndrome and Barth syndrome (Shlame et al., 2006; Allanson, 2007). Most fetuses diagnosed with hypertrophy survive postnatally. Analysis of two mutant pedigrees exhibiting hypertrophy with neonatal lethality confirmed severe cardiac hypertrophy (Fig. 13). Although we largely did not pursue mutants exhibiting fetal hypertrophy, we undertook the analysis of one such mutant to ascertain its heritability to validate the ultrasound diagnosis of fetal hypertrophy (Leatherbury et al., 2008). Indeed, this mutation was observed to be heritable in successive generations of breeding.

VI.  Diagnosis of structural heart defects Necropsies and histopathology were performed to delin­ eate the structural heart malformations. EFIC imaging was developed to facilitate the analysis of the often complex cardiac anatomy associated with congenital heart disease (Rosenthal et al., 2003). For EFIC imaging, tissue is embedded in paraffin and cut with a sledge microtome. Tissue autofluorescence at the block face is captured and used to generate registered serial two-dimensional images of the specimen. The image resolution achieved is comparable to or better than MRI. Data obtained by EFIC imaging can be easily resectioned digitally or reconstructed in three dimensions to facilitate the analysis of complex morphological changes in the developing embryonic heart. In this manner, the structural heart defects in every fetus can be diagnosed in their entirety, without regard to the orientation of

765

Chapter  |  11.2  Exploring the Genetic Basis for Congenital Heart Disease with Mouse ENU Mutagenesis

Table 3  Ultrasound Diagnosis of Cardiovascular Anomalies Families

Mutant fetuses

%Affected/ litter

Growth retarded

Prenatal lethal

Bradycardia

All cardiovascular defects

224

541

20%

144 (27%)

88 (14%)

67 (12%)

Arrhythmia

95 (42%)

139 (27%)

21%

17 (12%)

14 (10%)

67 (45%)

Outflow Regurgitation

101 (43%)

229 (37%)

21%

85 (41%)

60 (23%)

29 (14%)

Outflow velocity increase

35 (17%)

44 (9%)

18%

11 (25%)

4 (9%)

2 (4%)

Heart failure

108 (47%)

227 (39%)

20%

113 (54%)

78 (33%)

30 (14%)

Hypertrophy

102 (47%)

181 (34%)

20%

57 (30%)

17 (8%)

4 (2.4%)

Ectopia cordis

4 (2%)

4 (0.8%)

16%

4 (100%)

3 (75%)

3 (75%)

embedding or the plane of sectioning. This is highly advantageous for assessing the consistency and heritability of the congenital heart disease phenotype. Such detailed histological analysis showed our screen has yielded most of the major congenital heart diseases observed clinically, including left-to-right shunt lesions and obstruction lesions, as well as other structural heart defects (Table 4). We observed a wide range of outflow anomalies, such as persistent truncus arteriosus, transposition of the great arteries, doubleoutlet right ventricle, pulmonary atresia and other outflow alignment defects (Fig. 14). Some mutants exhibited complex structural heart defects, such as single ventricle (Fig. 15) and superior–inferior ventricles (Fig. 16). Ventricular/ atrial septation defects were also commonly observed, as were aortic arch anomalies, coronary vascular defects, and valvular and myocardial/epicardial anomalies (Fig. 17). In addition to cardiac defects, some of our mutants exhibited extracardiac defects that provide syndromic presentations reminiscent of those seen clinically, including one mutant family that has PTA and cranio–facial defects that are reminiscent of DiGeorge syndrome (Fig. 18) (Yu et al., 2004). Another family exhibited forelimb defects together with ventricular septation defects (VSDs) that show similarities to Heart–Hand syndrome (Fig. 19). Yet other mutants showed eye, body wall closure defects and polycystic kidney disease which overlaps with Bardet-Biedl syndrome.

VII.  Noncardiac defects The percentage of fetuses with defects observed per litter in the ultrasound screen was approximately 20% (Table 2), close to the 25% expected for Mendelian inheritance of recessive mutations. We broadly grouped defects in categories defined as cardiovascular, prenatal lethal, growth retardation, cranio–facial/head anomaly and body wall defects

(Table 2). Of the 902 abnormal fetuses, 405 from 188 families died prenatally, detected as fetuses without a heartbeat (Table 2). Growth retardation is defined as fetuses exhibiting crown-to-rump length and fetal area measurements two standard deviations below the mean (Yu et al., 2004). This is best visualized via measurements made in the sagittal view (Fig. 8B), but it also can be seen in frontal and transverse views (Fig. 8C,D). We note that embryos in heart failure will exhibit hydrops, which is observed by ultrasound as a halo of echo transparency around the fetus, referred to as nuchal translucency – an indication of the accumulation of fluid around the body (Fig. 20A,B). Body wall defect is indicated when the liver, gut and/or heart are seen protruding outside the body wall (Fig. 20C,D). Encephalocele is indicated when part of the brain is seen protruding outside the cranium, and often this is associated with polyhydramnios (Fig. 20E,F). There was a markedly lower overall incidence of cranio–facial and abdominal wall defects versus cardiac anomalies (2–3% versus 60%) (Table 2). Despite this, extracardiac defects were observed in 37% of fetuses with cardiovascular anomalies. This compares favorably with clinical studies showing 17% to 62% of human fetuses with cardiac defects also have various extracardiac anomalies (Hoffman, 1995a,b). The lower incidence of cranio–facial and body wall closure defects could simply reflect the limited two-dimensional spatial resolution of the ultrasound scan. However, we note omphalocele and gastroschisis occur in the human population at an incidence of only 0.01% to 0.02% (Wilson et al., 2004), and are similar to the incidences reported for encephalocele and holoprosencephaly. In contrast, human congenital heart diseases occur at a much higher rate of up to 1% of live births (Hoffman, 1995a). These differences in the frequency of cardiac versus cranio–facial/abdominal wall defects in human compare surprisingly favorably to the incidences observed in our mouse fetal screen.

766

PART  |  11  Cardiomics

(A)

(B)

(C)

(D) (E) (J)

(F)

(G)

(K)

(H)

(I)

Figure 13  Ultrasound detection and recovery of mutants with fetal cardiac hypertrophy. (A–E) The heart of a stillborn pup showed an abnormal shape that indicated a hypoplastic left ventricle (A). M-mode analysis of the same fetus at E17.5 obtained from short-axis view is shown in (B, C). Measurements made in diastole (B) and systole (C) indicated a smaller left ventricle and increased wall thickness associated with both ventricles. M-mode calculations indicated good systolic function, with left ventricular fractional shortening of 43% and ejection fraction of 81%. Histological analysis showed small left ventricle with ventricular septal defects (D, E). The pulmonary artery arises normally from the right ventricle (D). The aortic outlet is positioned directly posterior to the pulmonary outlet (E) resulting in a double-outlet right ventricle, with both the aorta and pulmonary artery arising from the right ventricle. Note the relative thick subepicardial mesenchyme on ventricular as well as atrial surfaces (D, E). (F–K) Stillborn fetus from another ENU mutant line that presented with cardiac hypertrophy based on M-mode ultrasound measurements. The heart showed an unusual shape indicative of hypertrophy (F), which was confirmed by histology (G). The right ventricular outflow tract appears to be narrowed by hypertrophied trabeculae that projected into the right ventricular chamber. EFIC images showed hypertrophic heart in another stillborn mutant from the same line (I) when compared with normal C57BL6 neonate (H). The mother of this fetus, heterozygous for the mutation, died at 10.5 months with significant cardiac hypertrophy (J), which was later confirmed by histology (K). Parts of the figure were produced from Leatherbury et al. (2008).

VIII.  Mapping mutations and strain modifier effects To map the ENU-induced mutations, we interbred the G2 females which are in the C57BL6/J inbred strain background with C3H males to generate C3H/B6 hybrid offspring. These were intercrossed and the mutants were

genome scanned using B6/C3H polymorphic microsatellite markers to map the mutation via linkage to the B6 markers (Yu et al., 2004). Mutations were identified as regions that are exclusively B6 in homozygote mutant animals. We mapped 15 mutations to 11 different chromosomes. Fourteen of these had structural heart disease. One mutation with severe fetal hypertrophy was recovered and shown to

Chapter  |  11.2  Exploring the Genetic Basis for Congenital Heart Disease with Mouse ENU Mutagenesis

767

Table 4  Congenital Heart Defects Recovered in Mutagenesis Screen Congenital heart defect

Phenotype

Left to right shunt lesions

Atrial septal defect; ventricular septal defect; patent ductus arteriosus; coronary fistula

Obstructive lesions

Pulmonary valvular stenosis; conotruncal pouch; aortic stenosis; coarctation of the aorta; bicuspid aortic valve; subaortic stenosis; Ebstein’s anomaly

Cyanotic defects

Persistent tuncus arteriosus; transposition of the great arteries; double-outlet right ventricle (DORV); tetralogy of Fallot; tricuspid atresia; pulmonary atresia; atrioventricular septal defect; total anomalous pulmonary venous return

Hypoplastic heart defects

Right heart hypoplasia; left heart hypoplasia

(A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

(I)

(J)

(K)

(L)

Figure 14  Congenital heart defects in mouse mutants recovered from ENU mutagenesis screen. (A) Normal newborn mouse heart with levocardia, i.e., heart apex pointing to the left and with normal positioning of pulmonary artery anterior to the aorta. (B) Mouse mutant with overriding aorta, i.e., aorta straddling the right and left ventricle. This was confirmed by EFIC histology. (C) Mutant with D-transposition of the great arteries. The aorta is anterior and to the right of the pulmonary artery, with aorta connected to the right ventricle and pulmonary artery to the left ventricle (confirmed by histology). (D) Mutant with dextracardia, i.e., heart apex pointing to the right. The ascending aorta is dilated at the root, and exhibits a right aortic arch with enlarged descending aorta. (E) Mutant with single outflow, or persistent truncus arteriosus. Also note coronary artery tunneling in the wall of the aorta and its abnormal high take off. (F) Mutant with persistent truncus arteriosus with interrupted aortic arch with large patent ductus arising from the distal end of the main pulmonary artery. (G) Mutant with double outlet of the outflows, enlarged RV, and interrupted right-sided aortic arch. EFIC histology show atrial–ventricular canal, pulmonary stenosis and left pulmonary isomerism. (H) Mutant with aorta right and posterior to a very small main pulmonary artery, a phenotype consistent with tetrology of Fallot physiology. (I) Mutant with unusual variation of persistent truncus arteriosus. The ascending aorta is the very small caliber vertical vessel in front of the truncal root where the coronaries arise. A posterior small short main pulmonary artery is seen with pulmonary branches. EFIC histology show the truncus mainly arises from the right ventricle and there is an outflow VSD. (J) Mutant with very small main pulmonary artery (MPA), suggesting pulmonary stenosis. The MPA is connected to the descending aorta by a ligamentum arteriosus – closed patent ductus arteriosus. EFIC histology also showed an atrial–ventricular canal. (K) Mutant with mesocardia. The aorta is right of a very small main pulmonary artery. EFIC histology identifies double-outlet right ventricle with pulmonary atresia, perimembranous VSD, patent ductus arteriosus and primum ASD. (L) Mutant with double-outlet right ventricle associated with aorta that is rightward and posterior to the pulmonary artery. EFIC analysis shows hypertrophy of both ventricles.

768

PART  |  11  Cardiomics

(A)

(D)

(G)

(B)

(E)

(H)

(C)

(F)

(I)

Figure 15  Heart situs anomalies with single ventricle spectrum of complex structural heart defects in Megf8 mutant. (A–C) Mesocardia is seen with anterior positioning of the aorta (Ao) and a right-sided aortic arch (RAA). EFIC reconstructed three-dimensional images show the dominant chamber is a morphological left ventricle (mLV), as indicated by attachment of the papillary muscles (asterisk in (C)) to the free wall. Note atrial and ventricular septal defects (white and black arrows, respectively), resulting in a common atrioventricular canal (P: pulmonary outflow; RC: right carotid; LC: left carotid; Tr: trachea). Scale bars (A, B)  500 um and (C)  400 um. (D–F) Levocardia is seen with anterior positioning of the aorta and a left-sided aortic arch (LAA). EFIC reconstructions (E, F) show a single dominant ventricle of left ventricular morphology (mLV). Note papillary muscles attached to the free wall (asterisks in (F)). Scale bars (D, E)  500 um and (F)  400 um. (G–I) Dextracardia is seen with anterior positioning of the aorta and a right-sided aortic arch (RAA). Three-dimensional reconstruction of the mutant heart is shown in a posterior view (H). Total anomalous pulmonary venous connection (TAPVC) is indicated by insertion of the pulmonary vein (PV) confluence into the anatomical right atria (RA) near the midline (white arrow in (H)). The RSVC and LSVC (asterisks) are inserted symmetrically into the atria on the right and left, making the left-sided atrium a morphologic right atrium (mRA). There was no coronary sinus present. Three-dimensional reconstructed imaging of the same heart is also shown in an anterior view (I). Right atrial isomerism is indicated by insertion of the left and right superior vena cava (RSVC, LSVC) symmetrically into the anatomic right (RA) and left atria (mRA). This suggests the anatomical left atrium is morphologically a right atrium (mRA). TAPVC is indicated by insertion of the pulmonary vein (PV) confluence into the atrium on the right (RA). This image shows the aorta anterior and rightward of the pulmonary artery, which is consistent with the necropsy finding (G) (A: anterior; P: posterior; L: left; R: right). Scale bars (H, I)  400 um. Parts of the figure were reproduced from Aune et al. (2008).

be heritable with a semi-dominant inheritance (Leatherbury et al., 2008). We noted severe C3H strain modifier effects such that after 3–5 generations of C3H intercrosses the mutant phenotype was no longer recoverable. Strain modifier effects were observed in all the ENU mutant lines recovered. This was indicated by an apparent failure to recover the mutant phenotype in offspring derived from the third to the fifth generation C3H intercrosses. As the only means of identifying heterozygous carriers is by phenotyping the offspring from presumptive heterozygote crosses, maintenance of the line was critically dependent on the recovery of the C57BL6 background. However,

this must be achieved without reintroducing the C57BL6 chromosome on which the mutation of interest is situated. Fortunately, this can be achieved using commercially available C57BL6 chromosome substitution strains, also referred to as consomic mice, which were generated by Dr J. Nadeau (Nadeau et al., 2000). In consomic mice, the chromosome of one strain (donor) has been moved into the genetic background of another strain (recipient) through repeated backcrosses such that only a single chromosome is of the donor strain (Fig. 21). For example, for a mutant with the mutation mapped to chromosome11, we would use the C57BL/6JChr11A consomic mice carrying all B6 chromosomes except

Chapter  |  11.2  Exploring the Genetic Basis for Congenital Heart Disease with Mouse ENU Mutagenesis

769

(A)

(E)

(B)

(F)

(C)

(D)

(G)

Figure 16  Superior–inferior ventricles and AV canal defect with anomalous venous return in Dnah5 mutant. (A) Superior–inferior ventricles with congested superior ventricle and atrial appendages with the aorta and pulmonary artery side-by-side. Note this same heart is shown in different views in Fig. 4. (B) Three-dimensional reconstruction in an apical orientation looking anteriorly at two papillary muscles in the smaller, inferior mLV, as indicated by the two arrowheads. There is also a small, muscular VSD present near the apex. (C) Anterior view of three-dimensional EFIC reconstruction illustrates side-by-side, semilunar valves of similar height associated with double-outlet right ventricle. The bicuspid aortic valve (AV) is on the left, and the tricuspid pulmonic valve (PV) is on the right. An AV canal defect (AVC) is situated at the crux of the heart. (D) Frontal two-dimensional section showing the same AVC as in (E) entering both the mRV and the mLV, with a VSD denoted by the arrow. LA, left atrium. (E) Two inferior venous structures are seen in this mutant. (F) Three-dimensional reconstruction looking posteriorly at the left and right IVCs connecting to the base of the two morphological right atria suggests right atrial isomerism. The arrowheads highlight the border of the AV canal, from which blood empties into both the mLV and mRV. Side-by-side outflows are also present with the aorta anterior and on the left and the pulmonary artery on the right. (G) Original two-dimensional image looking anteriorly at the two atria. Note the presence of duplicated IVC entering into the base of both morphologic right atria (RA: right atrium). Scale bars: 500 μm. Parts of the figures were reproduced from Tan et al. (2007).

770

PART  |  11  Cardiomics

(A)

(B)

(C)

(D)

(E)

(F)

Figure 17  Myocardial and epicardial anomalies. The histological appearance of the ventricular myocardium and epicardium of hearts from mice identified as having cardiovascular defects from the ultrasound interrogations (and which died at birth) showed a range of myocardial and epicardial defects (A–E). Shown in (F) for comparison is a section from a control normal newborn mouse heart. In some of the mutant mice, the compact myocardium was thickened (A, B). In others (C), the compact myocardium was nearly absent. In all of these mutants, the myocardium was less organized than in the control. The epicardium was also variable, some showing a thin or nearly absent epicardium (A, E), while others (B, C, D) show significant expansion of the epicardial layer.

for chromosome 11, which is derived from the A/J mouse strain. In this way, the mutation on chromosome 11 could still be tracked using B6 markers, even as the C57BL6 background is reintroduced genome-wide. This strategy allowed the immediate recovery of the mutant phenotype from all of the mutant lines. These findings indicate mutations causing congenital heart disease are exquisitely sensitive to genetic modifier effects. The future mapping of these genetic modifiers may provide further insights into the complex genetics associated with human congenital heart disease.

IX.  Mutation identification Beyond initial mapping of the mutation to a large chromosome interval, genotyping of an additional 50 to over 100 meiotic recombinants is necessary to narrow the map interval to 2–5 Mb, at which point compelling candidate genes in the map interval can be sequenced. This entailed a

combination of genomic and cDNA sequencing. Overall, we have found a combination of three strategies to be effective for mutation identification. In a few mutant lines, we identified the mutation in the map interval with either candidate gene sequencing or systematic sequencing of genes in the map interval. A second strategy entailed gene expression profiling with microarray analysis with RNA obtained from the mutant embryonic heart, with the mutated gene identified via a marked reduction in transcript expression. The third strategy entailed BAC library construction and assembly of a BAC contig spanning the map interval, followed by massively parallel sequencing (Fig. 22). This strategy was used to recover the mutation in a mutant exhibiting heterotaxy with a single ventricle spectrum of complex congenital heart disease (Zhang et al., 2009). The bottleneck in mutation recovery has been the biggest stumbling block in ENU mutagenesis screen. The extensive breeding required to narrow the map interval is time-consuming and costly. However, with the recent

Chapter  |  11.2  Exploring the Genetic Basis for Congenital Heart Disease with Mouse ENU Mutagenesis

(A)

(B)

771

(C)

(D)

(E)

(H)

(I)

(F)

(G)

(J)

(K)

Figure 18  Mutant with DiGeorge syndrome-like anomalies. (A, B) Color-flow and Doppler analysis of a fetus at E18.5 from ENU family 26 in the apical three-chamber view. Color-flow analysis showed blue outflow with aliasing (arrow in (A)) indicative of increased velocity. Superimposed on the blue outflow is an orange inflow with aliasing (arrow in (B)) indicative of regurgitant flow with increased velocity. (C) Spectral Doppler analysis obtained with the sample volume of the transducer positioned over the blue outflow (see white bar over blue outflow in inset of (C)). They showed the outflow with an abnormal regurgitant flow. (D, E) Fetus with outflow regurgitation, found dead at birth, exhibited a shortened snout, low-set ears, rounded head and shortened neck (D). Shown in (E) is a normal C57BL6/J neonate. (F, G) Analysis of the skeleton of the dead neonate (F) compared to the control normal C57BL6/J neonate (G) showed in dead neonate, a shortened nasal bone (N), while the frontal bone (F) was expanded. In addition, the maxilla was decreased in length (M). (H, I) The heart of the dead pup (I) exhibited a single outflow vessel, a condition referred to as persistent truncus arteriosus (PTA). Shown in (H) is a normal C57BL6/J neonate with normal pulmonary (P) and aortic (Ao) arrangement. (J, K) Histological analysis of the heart from (H) showed single outflow positioned over the right ventricle (RV), with the right and left ventricles (LV) connected by a ventricular septal defects (VSD) in panel (J). Panel (K) showed a very short main pulmonary artery segment giving rise to the right and left pulmonary arteries (RPA and LPA) and a five leaflet single AV valve (Asterisk in (K)) (LCA: left coronary artery; RCA: right coronary artery). Figure reproduced from Yu et al. (2004).

emergence of sequence capture strategies, coupled with massively parallel sequencing technology, targeted genome sequencing may be feasible without the expense entailed in BAC library construction (Albert et al., 2007; Hodges et al., 2007). Moreover, with the advent of third-generation sequencing technology on the horizon, such as real time single molecule DNA sequencing (Eid et al., 2009), whole genome resequencing is on the horizon, making the recovery of ENU-induced mutations affordable and straightforward. This would make mutagenesis screening an attractive non-gene biased approach to genetically dissect a wide variety of biological processes. A surprisingly large number of the mutant lines recovered in our screen exhibited laterality defects, even though our ultrasound screen was not focused on the recovery of mutants with laterality defects. We note that the heart is the single most left–right asymmetric organ in the body. The left–right asymmetry associated with the cardiopulmonary anatomy is critical for establishing the systemic/pulmonary ­ circulation

that is, in turn, essential for the survival of air-breathing animals. It is interesting to note that the incidence and type of laterality defects are different among the different mutant lines. Below we provide more detailed presentations of two mutants recovered in our mutagenesis screen. Both exhibit randomization of laterality referred to as heterotaxy, but each mutation causes a very different spectrum of congenital heart disease, with different spectrum of laterality defects.

X.  Mutation in megf8 causes single ventricle spectrum of complex congenital heart disease One mutant line was recovered from the ENU screen with a recessive mutation causing a single ventricle spectrum of complex congenital heart disease (Aune et al., 2007). Situs abnormalities were associated with organs in both the thoracic and abdominal cavities. Mutants exhibited abnormal

772

PART  |  11  Cardiomics

(C)

(A)

(B)

(F)

(G)

(J)

(N)

(D)

(E)

(H)

(K)

(I)

(L)

(O)

(M)

(P)

Figure 19  Forelimb and heart defects (heart–hand syndrome) in Family 166 (A–D). Forelimbs of dead pups from family 166 are abnormally flexed towards the chest, with clubbed front paws. Typically three digits are seen (B), but in some pups there appears to be fusion of the third and fourth digits (C, D). (F, G) Whole mount view shows enlarged atria and abnormal heart shape, indicative of hypertrophy. An abnormal large peritruncal coronary vein is observed at the base of the outflows (G). (E, H–M) Histology shows biventricular hypertrophy (H, J, L), and abnormal coronaries around the outflow vessels (E, H). Also evident are muscular VSDs (see arrows in J–M), and disorganized myocardium (E, I, K, M), which frequently show gaps between the myofiber bundles (see arrowheads in (E)). (N–P) The heart shown in (G) was processed for episcopic fluorescence image capture imaging, and the image stacks obtained were resectioned at different pitch to visualize the VSDs (see white arrows). Sections in (E, H, I) are from heterozygous, while the rest are from homozygous animals. Figure reproduced from Yu et al. (2004).

lung lobation consisting of four lung lobes symmetrically placed on both sides of the chest cavity, a pattern which is normally associated with the anatomical right and thus is indicative of right pulmonary isomerism. In addition, heart situs can be at the midline, mesocardia, apex pointing to the right, dextracardiac, or apex pointing to the left, levocardia (Fig. 15). The stomach was found either on the right or left side of the body, and we also observed symmetrical bilobed liver. Together, these findings of randomized organ situs in the thoracic and abdominal cavities indicated heterotaxy. Analysis by EFIC imaging followed by threedimensional reconstructions showed all mutants have a dominant ventricle of left ventricular morphology (Fig. 15). Interestingly, the dominant chamber exhibited varied situs,

with half found on the anatomic right and half on the anatomic left (Fig. 15A–F). Those positioned on the anatomic right can be considered as inverted ventricles. The small chamber did not show the trabeculation typical for the right ventricle, nor did it have the trabecula septomarginalis, a band of muscle situated at the roof of the right ventricle that normally subdivides the inlet versus outlet portion of the right ventricular chamber. As, typically, both outflows emerge from this small chamber, we refer to this as the outflow chamber. Together these findings suggest this mutant has a single ventricle spectrum of heart defect that is characterized by a dominant chamber of left ventricular morphology that may be situated on the right or left side of the chest cavity.

Chapter  |  11.2  Exploring the Genetic Basis for Congenital Heart Disease with Mouse ENU Mutagenesis

(A)

(C)

(E)

(G)

(B)

(D)

(F)

(H)

Figure 20  Ultrasound detection of noncardiac developmental defects. (A, B) Two-dimensional ultrasound image (A) revealed severe edema forming a halo-like effect around the body (see arrowheads) at E14.5. Corresponding whole mount view of the fetus (B) confirmed hydrops as indicated by translucency of the skin. (C, D) Two-dimensional ultrasound image (C) showed a fetus at E14.5 with a large tissue mass exhibiting high echogenicity (arrowheads) outside the body cavity. Subsequent examination of the fetus by necropsy supports the ultrasound finding and revealed an omphalocele consisting of the liver and intestine in a membrane-enclosed sac (D). (E, F) Two-dimensional ultrasound image (E) shows a fetus with brain tissue that appeared to protrude outside the cranium (arrowheads). Polyhydramnios is also seen, as indicated by excess echolucency or “black” space around the fetus. Examination of this fetus by necropsy confirmed this fetus with an encephalocele (F). (G, H) A monochorionic, twin pregnancy was found in a litter at E 17.5 by ultrasound (G). Twin A (pump twin) was appropriate in size for gestational age and normal shape with cardiomegaly (G, H). Twin B (recipient twin) showed growth retardation and abnormal small head (G, H). Figure reproduced from Shen et al. (2005).

All mutants also exhibited misalignment of the great arteries with anterior positioning of the aorta or transposition of the great arteries, with both outflows connected to the small outflow chamber. All mutants also had atrial and ventricular septal defects that resulted in atrioventricular canals (Fig. 15C). Half of the mutants exhibited right atrial isomerism. This is indicated by the insertion of the left and

773

right superior vena cava into the anatomic left and right atria with no coronary sinus (Fig. 15H). This contrasts with normal insertion of the right superior vena cava into the right atria and the left superior vena cava into the coronary sinus, which then enters the right atria. Some mutants also had total anomalous pulmonary venous connection; this is indicated by the pulmonary vein confluence connecting with the right-sided atrium near the midline (Fig. 15H,I). Overall, mutants in this line showed a highly-consistent phenotype comprising a dominant chamber of left ventricular morphology and endocardial cushion defects, together with right pulmonary isomerism and transposed positioning of the great arteries. Mutants with right atrial isomerism also exhibited right pulmonary isomerism. Beyond this, the overall patterning of the aortic arch, atria, and heart and stomach situs were varied, with no correlation between situs in the thoracic and abdominal cavities. However, it is interesting to note that most of the animals with TAPVC and all animals with right-sided aortic arch exhibited inverted ventricles. The mutation was mapped to a 2.2 Mb interval, but given this region was extremely gene-dense, cDNA and exon sequencing was impractical. Thus, a BAC library was constructed from the mutant genome and a BAC contig assembled over the 2.2 Mb interval. Solexa sequencing of the entire 2.2 Mb interval identified a single C-to-T base change (C193R) causing a cysteine-to-arginine substitution in the coding region of Megf8. A second mutation was found in the intergenic region of the map interval (Fig. 22). These results are remarkably in-line with the previous estimate of one ENU-induced mutation per Mb, with one in 1.82 Mb expected to alter function (Quwailid et al., 2004). We note Megf8 was not yet annotated in the mouse genome when this mutation was first mapped to mouse chromosome 7. Thus, a candidate gene-sequencing strategy at the time would have failed to recover the mutation. The point mutation in the intergenic region resides in a repetitive element that is not conserved, even between mouse and rat. Hence, this intergenic mutation is unlikely to be deleterious. Megf8 is a novel gene encoding a well-conserved protein of 2,789 amino acids. Orthologs are found in many other species, including human, zebrafish and Drosophila (Zhang et al., 2009). Megf8 contains predicted EGF, EGF-like, calcium-binding EGF-like and laminin-type EGF-like repeats, kelch domains, plexin repeats, and a CUB and transmembrane domain (Fig. 22). The missense mutation eliminates an invariant cysteine situated in the second putative EGF domain, and could disrupt formation of a disulfide bond required for proper protein folding. Thus, this second mutation is a good candidate as the genetic lesion causing the defect phenotype in this mutant line. This was confirmed when Megf8 knockdown in zebrafish embryos recapitulated the heterotaxy phenotype of Megf8m/m mutants, with discordant heart and gut situs observed in 75% of zebrafish Megf8 morphants. The

774

PART  |  11  Cardiomics

Figure 21  Remediation of genetic modifier effects using consomic mice. ENU mutants obtained in the C57BL6 background are intercrossed with C3H mice to map the mutation. After several generations of C3H intercrosses, the offspring contain a mix of C57BL6 (red) and C3H (blue) chromosome segments (left panel). Modifier effects from the C3H genetic background will often cause an apparent loss of the mutant phenotype. To recover the C57BL6 background and yet still allow continued tracking of the mutation in the original ENU mutagenized C57BL6 chromosome, mating with consomic mice can be used. For example, if a mutation was mapped to chromosome 11, then mating is to be carried out with C57BL/6J-Ch11A consomic mice in which all chromosomes are derived from C57BL/6J except for chromosome 11, which is from A/J (green) (middle panel). After one generation of intercross with the consomic mice, the C57BL6 background is globally recovered, except for chromosome 11 (right panel).

(A)

(B)

Figure 22  Recovery of Megf8 Mutation. (A) A 15-BAC contig was obtained spanning the 2.2 Mb critical region containing the mutation in chromosome 7 (25,837,390-28,028,384 Mb, NCBI m37 assembly). Sequencing showed a T-to-C substitution, causing cysteine (C) to arginine (R) amino acid replacement in Megf8, and another C-to-A substitution in a noncoding intergenic region. (B) Megf8 encodes a protein containing EGF, EGF-calcium, EGF-like, EGF-laminin, kelch, plexin and CUB domains, with the C193R substitution situated in the second conserved EGF domain. The mutated cysteine is highlighted in the comparative sequence alignment. Reproduced from Zhang et al. (2009).

Chapter  |  11.2  Exploring the Genetic Basis for Congenital Heart Disease with Mouse ENU Mutagenesis

recapitulation of the mouse heterotaxy phenotype with Megf8 morpholino knockdown would suggest this C193R mutation is a loss-of-function mutation. Megf8 is expressed ubiquitously, and further studies indicate that it plays an essential role in left–right patterning through the regulation of Nodal signaling. Surprisingly, antibodies made to the N-terminus of Megf8 showed localization of Megf8 as punctate dots in the nucleus, with co-localization to Gfi1b, a protein known to play an important role in heterochromatin­ ization in erythrocytes (Saleque et al., 2002; Lim et al., 2006; Vassen et al., 2006) and Baf60C, a protein in the SwiSNF chromatin remodeling complex (Takeuchi et al., 2007).

XI.  DnaH5 mutation, heterotaxy and primary ciliary dyskinesia Another mutant line recovered with complex congenital heart disease and laterality defects was identified as harboring a mutation in the gene encoding the heavy chain dynein Dnah5. Mutations in the human DNAH5 ortholog are frequently found in patients with primary ciliary dyskinesia – a genetically-heterogeneous disorder known to cause sinopulmonary disease due to a defect in mucociliary clearance in the airway, and also Kartagener’s syndrome, a condition in which there is complete reversal of body situs, i.e., situs inversus totalis (Fliegauf et al., 2007). Indeed, we observed some of the Dnah5 mutant animals survived postnatally with situs inversus totalis. The laterality defects observed in primary ciliary dyskinesia patients reflect the essential role of the cilium in the specification of left–right patterning (Chapters 4.1 and 4.2). Primary ciliary dyskinesia was

(A)

(B) del593

775

thought to cause either no situs defects (situs solitus), or complete mirror reversal of situs (situs inversus totalis or Kartagener’s syndrome). However, a recent study of patients with primary ciliary dyskinesia showed a 6% incidence of heterotaxy, but it was not possible to ascertain whether this was elicited by the primary ciliary dyskinesia causing genetic lesions or some other lesions in the genetic make-up of the patients (Kennedy et al., 2007). Analysis of our Dnah5 mutant mouse model showed a surprisingly high incidence of heterotaxy – 40% with heterotaxy, 25% situs solitus and 35% situs inversus totalis (Fig. 23). Mutants with heterotaxy had complex structural heart defects that included discordant atrioventricular and ventricular–arterial situs, atrioventricular canals and atrial/pulmonary isomerisms. Variable combinations of a distinct set of cardiovascular anomalies were observed, including superior–inferior ventricles (Fig. 16), great artery alignment defects, duplicated inferior vena cava (Fig. 16) and interrupted inferior vena cava with azygos continuation. The surprisingly high incidence of heterotaxy led us to evaluate the diagnosis of primary ciliary dyskinesia, which we confirmed with transmission electron microscopy, which revealed missing outer dynein arms in the respiratory cilia, a typical finding in primary ciliary dyskinesia patients with DNAH5 mutations. In addition, ciliary dyskinesia was observed by videomicroscopy of airway epithelia from the trachea of homozygote Dnah5 mutant mice (Tan et al., 2007). Immunofluorescence staining showed Dnah5 is localized in the axoneme and basal body in wild-type tracheal epithelia, but in the homozygote Dnah5 mutants the mutant protein (with internal in frame deletion) was mislocalized to the cytoplasm. Videomicroscopy showed the cilia in the embryonic node were immotile.

(C)

(D)

Figure 23  Situs anomalies in Dnah5 mutant. (A) Situs solitus with levocardia. The right lung (R) has four lobes and the left lung (L) has one. Arrow indicates direction of heart (H) apex. The stomach (S) is on the left. (B) Situs inversus totalis with dextrocardia. Note four left (1, 2, 3, 4) and one right (R1) lung lobes, and stomach on the right. (C, D) Heterotaxy with levocardia. Note a right aortic arch (RAA), one lung lobe on each side (R1 and L1), and stomach on the right (C). After removal of the liver, azygos continuation (arrowhead) of the interrupted inferior vena cava from the kidney (K) can be observed (D) (H  heart; arrows indicate direction of heart apex). Scale bar for (A–D) in (D)  250 mm. Reproduced from Tan et al. (2007).

776

These findings show Dnah5 is required for left–right patterning, and suggest the primary ciliary dyskinesiacausing Dnah5 mutation can also cause heterotaxy. In contrast to the ambiguity created by the genetic heterogeneity of patients diagnosed with primary ciliary dyskinesia, our primary ciliary dyskinesia mouse model is inbred, and thus genetically identical except for the disease-causing mutation. Hence, these findings conclusively show Dnah5 mutation alone can cause both primary ciliary dyskinesia and complex structural heart defects in conjunction with heterotaxy. These findings are clinically significant, as the medical management of patients with primary ciliary dyskinesia may be substantially altered given heterotaxy. Conversely, heterotaxy patients who suffer from complex structural heart defects requiring high-risk surgical repairs may benefit from different clinical management strategies if possible complications from primary ciliary dyskinesia are appreciated a priori. Complications arising from primary ciliary dyskinesia could cause ventilator dependence, with a long clinical course and poor outcome. These findings suggest a compelling case for human studies to examine possible undiagnosed ciliary dysfunction in patients with complex congenital heart disease and heterotaxy. Such translational studies may suggest changes in the standard of care that could help improve outcomes for patients undergoing high-risk cardiac surgeries with high mortality rates.

XII.  Future prospects for saturation mutagenesis screens The genetic basis for congenital heart disease is not wellunderstood, as human genetic studies are confounded by genetic heterogeneity, variable penetrance and variable expressivity, and gene–environment interactions can also affect disease risk. Given the high throughput nature of cardiovascular phenotyping with noninvasive mouse fetal echocardiography, it should be possible to undertake a large-scale saturation mouse mutagenesis screen to recover the majority of genes likely to play a role in congenital heart disease. The feasibility of such a saturation screen is indicated not only by the efficacy of noninvasive mouse fetal echocardiography for rapid cardiovascular phenotyping, but also by the advent of third-generation whole-genome sequencing technology, such as real time single molecule sequencing (Eid et al., 2009). Whole genome sequencing would be relatively straightforward, given the mice are inbred and genetically-identical except for the ENU-induced mutations. This would obviate the need for the extensive breeding otherwise required to localize the genomic interval of the mutation and subsequent candidate gene sequencing. The identification of a core set of congenital heart disease genes may yield new insights into genetic pathways and disease mechanisms that underlie human congenital heart disease. A diagnostic chip could be made for future clinical

PART  |  11  Cardiomics

translational studies to establish a genotype–phenotypeoutcome matrix. Such studies may reveal combinations of gene sequence variants that may be predictive of disease and disease outcomes, and shed new light into the complex genetics of human congenital heart disease.

References Albert, T., Molla, M., Muzny, D., Nazareth, L., Wheeler, D., Song, X., Richmond, T., Middle, C., Rodesch, M., Packard, C., Weinstock, G.M., Gibbs, R.A., 2007. Direct selection of human genomic loci by microarray hybridization. Nat Methods 4, 903–905. Allanson, J., 2007. Noonan syndrome. Am. J. Med. Genet. C Semin. Med. Genet. 145C, 274–279. Amsterdam, A., Nissen, R., Sun, Z., Swindell, E., Farrington, S., Hopkins, N., 2004. Identification of 315 genes essential for early zebrafish development. Proc Natl Acad Sci USA 101, 12792–12797. Andelfinger, G., 2008. Genetic factors in congenital heart malformation. Clin. Genet. 73, 516–527. Aune, C., Chatterjee, B., Zhao, X., Francis, R., Bracero, L., Yu, Q., Rosenthal, J., Leatherbury, L., Lo, C., 2008. Mouse model of heterotaxy with single ventricle spectrum of cardiac anomalies. Pediatr. Res. 63, 9–14. Boughman, J.A., Neill, C.A., Ferencz, C., Loffredo, C.A., 1993. The Genetics of Congenital Heart Disease. Futura Publishing, Mount Kisco, NY. Brickner, M., Hillis, L., Lange, R., 2000a. Congenital heart disease in adults. First of two parts. N. Engl. J. Med. 342, 256–263. Brickner, M., Hillis, L., Lange, R., 2000b. Congenital heart disease in adults. Second of two parts. N. Engl. J. Med. 342, 334–342. Bruneau, B., 2008. The developmental genetics of congenital heart disease. Nature 451, 943–948. Bruneau, B., Nemer, G., Schmitt, J., Charron, F., Robitaille, L., Caron, S., Conner, D., Gessler, M., Nemer, M., Seidman, C., Seidman, J.G., 2001. A murine model of Holt-Oram syndrome defines roles of the T-box transcription factor Tbx5 in cardiogenesis and disease. Cell 106, 709–721. Caspary, T., Anderson, K., 2006. Uncovering the uncharacterized and unexpected: unbiased phenotype-driven screens in the mouse. Dev. Dyn. 235, 2412–2423. Charlesworth, B., 1991. Evolution. When to be diploid. Nature 351, 273–274. Driever, W., Solnica-Krezel, L., Schier, A., Neuhauss, S., Malicki, J., Stemple, D., Stainier, D., Zwartkruis, F., Abdelilah, S., Rangini, Z., Belak, J., Boggs, C., 1996. A genetic screen for mutations affecting embryogenesis in zebrafish. Development 123, 37–46. Eid, J., Fehr, A., Gray, J., Luong, K., Lyle, J., Otto, G., Peluso, P., Rank, D., Baybayan, P., Bettman, B., et al., 2009. Real-time DNA sequencing from single polymerase molecules. Science 323, 133–138. Fliegauf, M., Benzing, T., Omran, H., 2007. When cilia go bad: cilia defects and ciliopathies. Nat. Rev. Mol. Cell Biol. 8, 880–893. Friedrich, G., 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. García-García, M., Eggenschwiler, J., Caspary, T., Alcorn, H., Wyler, M., Huangfu, D., Rakeman, A., Lee, J., Feinberg, E., Timmer, J.R., Anderson, K.V., 2005. Analysis of mouse embryonic patterning and morphogenesis by forward genetics. Proc Natl Acad Sci USA 102, 5913–5919.

Chapter  |  11.2  Exploring the Genetic Basis for Congenital Heart Disease with Mouse ENU Mutagenesis

Garg, V., 2006. Insights into the genetic basis of congenital heart disease. Cell Mol. Life Sci. 63, 1141–1148. Garg, V., Muth, A., Ransom, J., Schluterman, M., Barnes, R., King, I., Grossfeld, P., Srivastava, D., 2005. Mutations in NOTCH1 cause aortic valve disease. Nature 437, 270–274. Gelb, B., Tartaglia, M., 2006. Noonan syndrome and related disorders: dysregulated RAS-mitogen activated protein kinase signal transduction. Hum. Mol. Genet. 15 (Spec No 2), R220–R226. Gruber, P., Epstein, J., 2004. Development gone awry: congenital heart disease. Circ. Res. 94, 273–283. Haffter, P., Granato, M., Brand, M., Mullins, M., Hammerschmidt, M., Kane, D., Odenthal, J., van Eeden, F., Jiang, Y., Heisenberg, C.P., Kelsh, R.N., Furutani-Seiki, M., Vogelsang, E., Beuchle, D., Schach, U., Fabian, C., Nüsslein-Volhard, C., 1996. The identification of genes with unique and essential functions in the development of the zebrafish, Danio rerio. Development 123, 1–36. Hiroi, Y., Kudoh, S., Monzen, K., Ikeda, Y., Yazaki, Y., Nagai, R., Komuro, I., 2001. Tbx5 associates with Nkx2-5 and synergistically promotes cardiomyocyte differentiation. Nat. Genet. 28, 276–280. Hodges, E., Xuan, Z., Balija, V., Kramer, M., Molla, M., Smith, S., Middle, C., Rodesch, M., Albert, T., Hannon, G., McCombie, W.R., 2007. Genome-wide in situ exon capture for selective resequencing. Nat. Genet. 39, 1522–1527. Hoffman, J., 1995. Incidence of congenital heart disease: I. Postnatal incidence. Pediatr. Cardiol. 16, 103–113. Hoffman, J., 1995. Incidence of congenital heart disease: II. Prenatal incidence. Pediatr. Cardiol. 16, 155–165. Hoffman, J., Kaplan, S., 2002. The incidence of congenital heart disease. J. Am. Coll. Cardiol. 39, 1890–1900. Hoffman, J., Kaplan, S., Liberthson, R., 2004. Prevalence of congenital heart disease. Am. Heart J. 147, 425–439. Hove, J., Köster, R., Forouhar, A., Acevedo-Bolton, G., Fraser, S., Gharib, M., 2003. Intracardiac fluid forces are an essential epigenetic factor for embryonic cardiogenesis. Nature 421, 172–177. Hutson, M., Kirby, M., 2007. Model systems for the study of heart development and disease. Cardiac neural crest and conotruncal malformations. Semin. Cell Dev. Biol. 18, 101–110. Jaenisch, R., 1988. Transgenic animals. Science 240, 1468–1474. Kennedy, M., Omran, H., Leigh, M., Dell, S., Morgan, L., Molina, P., Robinson, B., Minnix, S., Olbrich, H., Severin, T., Ahrens, P., Lange, L., Morillas, H.N., Noone, P.G., Zariwala, M.A., Knowles, M.R., 2007. Congenital heart disease and other heterotaxic defects in a large cohort of patients with primary ciliary dyskinesia. Circulation 115, 2814–2821. Kile, B., Hentges, K., Clark, A., Nakamura, H., Salinger, A., Liu, B., Box, N., Stockton, D., Johnson, R., Behringer, R., Bradley, A., Justice, M.J., 2003. Functional genetic analysis of mouse chromosome 11. Nature 425, 81–86. Kondrashov, A., Crow, J., 1991. Haploidy or diploidy: which is better? Nature 351, 314–315. Leatherbury, L., Yu, Q., Chatterjee, B., Walker, D., Yu, Z., Tian, X., Lo, C., 2008. A novel mouse model of X-linked cardiac hypertrophy. Am. J. Physiol. Heart Circ. Physiol. 294, H2701–H2711. Li, L., Krantz, I., Deng, Y., Genin, A., Banta, A., Collins, C., Qi, M., Trask, B., Kuo, W., Cochran, J., Costa, T., Pierpont, M.E., Rand, E.B., Piccoli, D.A., Hood, L., Spinner, N.B., 1997. Alagille syndrome is caused by mutations in human Jagged1, which encodes a ligand for Notch1. Nat. Genet. 16, 243–251. Lim, J., Hao, T., Shaw, C., Patel, A., Szabó, G., Rual, J., Fisk, C., Li, N., Smolyar, A., Hill, D.E., Barabási, A.L., Vidal, M., Zoghbi, H.Y., 2006.

777

A protein–protein interaction network for human inherited ataxias and disorders of Purkinje cell degeneration. Cell 125, 801–814. Maitra, M., Schluterman, M., Nichols, H., Richardson, J., Lo, C., Srivastava, D., Garg, V., 2009. Interaction of Gata4 and Gata6 with Tbx5 is critical for normal cardiac development. Dev. Biol. 326, 368–377. McDaniell, R., Warthen, D., Sanchez-Lara, P., Pai, A., Krantz, I., Piccoli, D., Spinner, N., 2006. NOTCH2 mutations cause Alagille syndrome, a heterogeneous disorder of the notch signaling pathway. Am. J. Hum. Genet. 79, 169–173. Mullins, M., Hammerschmidt, M., Haffter, P., Nüsslein-Volhard, C., 1994. Large-scale mutagenesis in the zebrafish: in search of genes controlling development in a vertebrate. Curr. Biol. 4, 189–202. Nadeau, J., Singer, J., Matin, A., Lander, E., 2000. Analysing complex genetic traits with chromosome substitution strains. Nat. Genet. 24, 221–225. Nüsslein-Volhard, C., Wieschaus, E., 1980. Mutations affecting segment number and polarity in Drosophila. Nature. 287, 795–801. Oda, T., Elkahloun, A.G., Pike, B.L., Okajima, K., Krantz, I.D., Genin, A., Piccoli, D.A., Meltzer, P.S., Spinner, N.B., Collins, F.S., Chandrasekharappa, S.C., 1997. Mutations in the human Jagged1 gene are responsible for Alagille syndrome. Nat. Genet. 16, 235–242. Olson, E., 2006. Gene regulatory networks in the evolution and development of the heart. Science 313, 1922–1927. Perrot, V., Richerd, S., Valéro, M., 1991. Transition from haploidy to diploidy. Nature 351, 315–317. 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., Rodger, S., Carey, J., Roberts, A., Glenister, P., Lalanne, Z., Parkinson, N., Coghill, E.L., McKeone, R., Cox, S., Willan, J., Greenfield, A., Keays, D., Brady, S., Spurr, N., Gray, I., Hunter, J., Brown, S.D., Cox, R.D., 2004. A gene-driven ENU-based approach to generating an allelic series in any gene. Mamm. Genome 15, 585–591. Razzaque, M., Nishizawa, T., Komoike, Y., Yagi, H., Furutani, M., Amo, R., Kamisago, M., Momma, K., Katayama, H., Nakagawa, M., Fujiwara, Y., Matsushima, M., Mizuno, K., Tokuyama, M., Hirota, H., Muneuchi, J., Higashinakagawa, T., Matsuoka, R., 2007. Germline gain-of-function mutations in RAF1 cause Noonan syndrome. Nat. Genet. 39, 1013–1017. Roberts, A., Araki, T., Swanson, K., Montgomery, K., Schiripo, T., Joshi, V., Li, L., Yassin, Y., Tamburino, A., Neel, B., Kucherlapati, R.S., 2007. Germline gain-of-function mutations in SOS1 cause Noonan syndrome. Nat. Genet. 39, 70–74. Saleque, S., Cameron, S., Orkin, S., 2002. The zinc-finger protooncogene Gfi-1b is essential for development of the erythroid and megakaryocytic lineages. Genes Dev. 16, 301–306. Shen, Y., Leatherbury, L., Rosenthal, J., Yu, Q., Pappas, M., Wessels, A., Lucas, J., Siegfried, B., Chatterjee, B., Svenson, K., Lo, C.W., 2005. Cardiovascular phenotyping of fetal mice by noninvasive high-frequency ultrasound facilitates recovery of ENU-induced mutations causing congenital cardiac and extracardiac defects. Physiol. Genomics. 24, 23–36. Silver, J., Hilton, D., Bahlo, M., Kile, B., 2007. Probabilistic analysis of recessive mutagenesis screen strategies. Mamm. Genome 18, 5–22. Srivastava, D., 2006. Genetic regulation of cardiogenesis and congenital heart disease. Annu. Rev. Pathol. 1, 199–213. Strauss, A., 1998. The molecular basis of congenital cardiac disease. Semin. Thorac. Cardiovasc. Surg. Pediatr. Card Surg. Annu. 1, 179–188. Sun, Z., Amsterdam, A., Pazour, G., Cole, D., Miller, M., Hopkins, N., 2004. A genetic screen in zebrafish identifies cilia genes as a principal cause of cystic kidney. Development 131, 4085–4093. Takeuchi, J.K., Lickert, H., Bisgrove, B.W., Sun, X., Yamamoto, M., Chawengsaksophak, K., Hamada, H., Yost, H.J., Rossant, J.,

778

Bruneau, B.G., 2007. Baf60c is a nuclear Notch signaling component required for the establishment of left-right asymmetry. Proc. Natl. Acad. Sci. USA. 104, 846–851. Tan, S., Rosenthal, J., Zhao, X., Francis, R., Chatterjee, B., Sabol, S., Linask, K., Bracero, L., Connelly, P., Daniels, M.P., Yu, Q., Omran, H., Leatherbury, L., Lo, C.W., 2007. Heterotaxy and complex structural heart defects in a mutant mouse model of primary ciliary dyskinesia. J. Clin. Invest. 117, 3742–3752. Vassen, L., Fiolka, K., Möröy, T., 2006. Gfi1b alters histone methylation at target gene promoters and sites of gamma-satellite containing heterochromatin. EMBO J. 25, 2409–2419. Vogel, F., Motulsky, A., 1986. Human Genetics: Problem and Approaches. Springer-Verlag, Berlin. Wilkie, A., 1994. The molecular basis of genetic dominance. J. Med. Genet. 31, 89–98. Williams, R., Pearson, G., Barst, R., Child, J., del Nido, P., Gersony, W., Kuehl, K., Landzberg, M., Myerson, M., Neish, S.R., Sahn, D.J., Verstappen, A., Warnes, C.A., Webb, C.L., 2006. Report of the National Heart, Lung, and Blood Institute Working Group on research in adult congenital heart disease. J. Am. Coll. Cardiol. 47, 701–707. Wilson, L., Ching, Y., Farias, M., Hartford, S., Howell, G., Shao, H., Bucan, M., Schimenti, J., 2005. Random mutagenesis of proximal

PART  |  11  Cardiomics

mouse chromosome 5 uncovers predominantly embryonic lethal mutations. Genome Res. 15, 1095–1105. Woods, I., Kelly, P., Chu, F., Ngo-Hazelett, P., Yan, Y., Huang, H., Postlethwait, J., Talbot, W., 2000. A comparative map of the zebrafish genome. Genome Res. 10, 1903–1914. Yu, Q., Shen, Y., Chatterjee, B., Siegfried, B., Leatherbury, L., Rosenthal, J., Lucas, J., Wessels, A., Spurney, C., Wu, Y., Kirby, M.L., Svenson, K., Lo, C.W., 2004. ENU induced mutations causing congenital cardiovascular anomalies. Development 131, 6211–6223. Yu, Q., Leatherbury, L., Tian, X., Lo, C., 2008. Cardiovascular assessment of fetal mice by in utero echocardiography. Ultrasound Med. Biol. 34, 741–752. Zhang, Z., Alpert, D., Francis, R., Chatterjee, B., Yu, Q., Tansey, T., Sabol, S., Cui, C., Bai, Y., Koriabine, M., Yoshinaga, Y., Cheng, J.F., Chen, F., Martin, J., Schackwitz, W., Gunn, T.M., Kramer, K.L., De Jong, P.J., Pennacchio, L.A., Lo, C.W., 2009. Massively parallel sequencing identifies the gene Megf8 with ENU-induced mutation causing heterotaxy. Proc. Natl. Acad. Sci. USA 106, 3219–3224. Zohn, I., Anderson, K., Niswander, L., 2005. Using genome-wide mutagenesis screens to identify the genes required for neural tube closure in the mouse. Birth Defects Res. A Clin. Mol. Teratol. 73, 583–590.

Chapter 11.3

Imaging Cardiac Developmental Malformations in the Mouse Embryo Timothy Mohun1, Wolfgang Weninger2 and Shoumo Bhattacharya3 1

Division of Developmental Biology, MRC National Institute for Medical Research, London, UK Integrative Morphology Group, Center for Anatomy & Cell Biology, Medical University Vienna, Austria 3 Wellcome Trust Centre for Human Genetics, University of Oxford, UK 2

I.  Introduction Congenital heart disease, the commonest birth defect (Burn and Goodship, 2002; Petersen et al., 2003; Pierpont et al., 2007) is defined as a gross structural abnormality of the heart or intrathoracic great vessels that is present at birth and is of functional significance (Hoffman and Kaplan, 2002; Petersen et al., 2003) (see Chapter 3.4). Human congenital heart disease is typically characterized by lesions including atrial, ventricular and atrioventricular septal defects (ASD, VSD, AVSD), outflow tract malformations (e.g., tetralogy of Fallot (TOF), common arterial trunk (CAT), transposition of great arteries (TGA), valvular defects (aortic and pulmonary stenosis – AS, PS) and aortic arch malformations (e.g., patent ductus arteriosus (PDA), interrupted arch and aortic coarctation) (Clark, 2001; Hoffman and Kaplan, 2002). Extensive data support the idea that congenital heart disease has an underlying genetic basis (Burn and Goodship, 2002). Environmental factors are also important. For instance, both maternal diabetes and obesity are risk factors for cardiovascular malformation (Schaefer-Graf et al., 2000; Loffredo et al., 2001a,b; Martinez-Frias, 2001; Kuehl and Loffredo, 2002; Watkins et al., 2003). Studies of human genetic architecture cannot provide a mechanistic understanding of gene function in cardiac development, and analysis of gene–gene and gene–environment interactions in humans remains a challenge. These considerations necessitate the use of animal models, the most appropriate and widely-used being the mouse. Like the human, the mouse has a four-chambered heart with a septated outflow tract, left-sided great arteries, and parallel pulmonary and systemic circulations. Common cardiac malformations such as septal, outflow tract and aortic arch and pulmonary trunk defects, can thus be identified in mouse embryos, and mouse mutations typically recapitulate Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

the cardiac malformations observed in patients with mutations in the same genes (reviewed in Schneider and Bhattacharya, 2004). Indeed, much of progress in our understanding of heart development over the past two decades is a result of studies using mouse models.

II.  The limits of histology The past decade, in particular, has brought an extraordinary proliferation in the number of genetically-modified mouse lines showing phenotypes affecting heart development and function (http://www.informatics.jax.org/). While the transgenic technologies used to produce these lines continue to improve, it is striking that the methods used to analyze the impact on heart structure have changed little. The nearuniversal approach for examining cardiac structure remains conventional histology, a procedure largely unaltered for over a century, following the invention of the microtome by Wilhelm His. The power of this approach is obviously the cellular resolution it provides, along with the ability to use a wide variety of histological stains to highlight different cellular or subcellular structures and components of the heart. (In fact, very few studies have taken advantage of diverse staining procedures, almost always confining themselves to a standard hematoxylin/eosin stain to reveal tissue morphology, occasionally supplemented by Masson’s trichrome for analysis of tissue fibrosis). There are, however, several obvious limitations inherent in histological analysis that can restrict its effectiveness in analyzing cardiac morphology. Most obviously, it provides two-dimensional images of three-dimensional structures, and three-dimensional information can only be reconstructed on the basis of a series of sections through the entire specimen. Systematic, serial acquisition of 779

780

histological sections is extremely time-consuming and labor intensive, requiring considerable skill to avoid any loss of data. In addition, the structural information presented in any single image critically depends on the plane of section, and this is extremely difficult to standardize solely through specimen orientation during embedding. As a result, detailed comparison of samples can be difficult. As a consequence, despite the enormous expansion of cardiac developmental biology, no comparable change has occurred in our ability to appreciate the impact of molecular changes on three-dimensional cardiac morphology, or in our ability to appreciate comparative aspects of cardiac structure. Cardiac morphology remains heavily-dependent on the accumulated histological knowledge and experience of relatively few cardiac anatomists. Of course, a number of other methods besides conventional histology have been used to analyze three-dimensional morphology of the developing heart and its vessels. For example, corrosion casting of plastic dye injections provides detailed images of vasculature associated with the heart in the early embryo (Kondo et al., 1993; Hiruma et al., 2002). Similarly, the combination of careful dissection and scanning electron microscopy can provide dramatic three-dimensional images of cardiac anatomy (for example Webb et al., 1997; King et al., 2002). However, both of these are demanding, specialist techniques, ill-suited for widespread adoption. Within clinical cardiology, magnetic resonance imaging (MRI) has transformed our ability to image the human heart, providing a level of tissue detail that far surpasses that provided by echocardiography. Applying the same technique to the mouse heart has posed serious challenges, both to achieve comparable resolution with such small organ sizes and, in the case of in vivo imaging, to adapt to the high heart rate of the mouse. Nevertheless, impressive improvements have been achieved using high magnetic field strengths and contrast agents (Johnson et al., 2002). These results, however, remain rather modest when set against the structural detail afforded by histology. The best isotropic three-dimensional spatial resolutions achieved with embryo samples so far reported fall within the range of 20–30 m, which is at least an order of magnitude lower than that comfortably achieved by light microscopy with histological sections. As well as limiting the detail that can be discerned in the monochrome magnetic resonance imaging images, the three-dimensional data set is only useful for three-dimensional modeling using isosurface rendering (of which more below).

III.  The promise of optical projection tomography A second, noninvasive imaging method that can be used for analyzing mouse heart morphology is optical projection tomography (OPT) (Sharpe et al., 2002). Since it utilizes optical microscopy for image capturing optical

PART  |  11  Cardiomics

projection tomography can, in principle, analyze a wide range of specimen sizes, encompassing hearts at a variety of developmental stages. Furthermore, because samples can be imaged using illumination at several wavelengths, optical projection tomography can be used to obtain data sets for tissue morphology, gene expression and antigen distribution (for example using GFP, visible light and red spectrum fluorochromes respectively). The great potential of this technique has been reviewed elsewhere and several examples of its successful use have now been published (Sarma et al., 2005; DeLaurier et al., 2006; Lee et al., 2006; Breckenridge et al., 2007; Bryson-Richardson et al., 2007; Ijpenberg et al., 2007; McGurk et al., 2007; Miller et al., 2007). However, there are a number of practical difficulties that have so far limited its utility in analyzing mouse embryo heart morphology and gene expression. The resolution of two-dimensional tissue images computed from optical projection tomography data remains relatively low (around 5–10 m), irrespective of increases in optical magnification or camera CCD size. As a result, cellular resolution is not generally possible in studies of tissue or organ morphology. It is therefore difficult, for example, to image the vasculature associated with the developing heart (such as the aortic arches) until relatively late in development (Yashiro et al., 2007). Furthermore, studies of the cardiovascular system face a particular challenge resulting from the variable presence of blood. Morphology is best captured in optical projection tomography using autofluorescence of the tissue, with GFP or RFP excitation and emission filters. Under these conditions, blood shows extremely intense fluorescence that can either mask or merge into surrounding tissue signal. Such signal foci also generate significant interference during data computation that can extend throughout the data set. While this problem can be ameliorated to some degree during sample fixation (for example, by using Dent’s fixative rather than formalin or paraformaladehyde), the presence of blood inevitably degrades the quality of optical projection tomography images (Fig. 1). The same problem bedevils optical projection tomography imaging of isolated neonatal or adult hearts, despite procedures for exsanguination prior to organ harvest. Optical projection tomography imaging of the heart within the pericardial cavity is also affected by the developmental stage. Up to mid-gestation, embryos (though small) can be reliably imaged. With these, optical projection tomography can provide good three-dimensional models of looped heart tube morphology, including the cardiac cushions and myocardial trabeculae (Fig. 1A,G). However, as the embryo grows the increasing density of tissue dorsal to the heart and the presence of the adjacent, blood-rich liver together seriously impairs the ability of optical projection tomography to image any detail of heart structure (Fig. 1B). (In our hands, attempts to image cardiac morphology in embryos older than E14.5 by optical projection

Chapter  |  11.3  Imaging Cardiac Developmental Malformations in the Mouse Embryo

(A)

(B)

(C)

(D)

(E)

(F)

(G)

Figure 1  Embryo heart morphology using optical projection tomography. (A, B) Transverse “virtual sections” through the heart of E10.5 and E14.5 embryos, images using GFP illumination. Note the loss of resolution in the older embryo, despite its increased size, and the intense signal produced by the presence of blood. (C) Autofluorescence image (GFP excitation/emission) of dissected thoracic organs removed from an E12.5 embryo, resulting in optical projection tomography data shown in (D–F). Note the presence of some blood (arrow) in the left ventricle. (G) Threedimensional model of the looped heart tube (E10.5) showing the formation of trabeculae within the prospective left ventricular chamber.

tomography are largely futile, and comparable difficulties are evident in studies of human embryos (http://www.ncl. ac.uk/ihg/EADHB/atlasbox.html).) One simple way to overcome these difficulties is to isolate the heart (or thoracic organs) from the embryo prior to imaging (Fig. 1C). While this is not feasible for studies of the associated vasculature, it has a dramatic impact on the quality of optical projection tomography imaging of internal cardiac architecture, and consequently on the three-dimensional models such data provides (Fig. 1D–F). Perhaps the most exciting promise of optical projection tomography is the potential to combine imaging of organ morphology with analysis of gene expression and protein distribution on the same sample. With appropriate resolution, this could transform our ability to analyze cardiac phenotypes. As yet, however, several serious technical obstacles need to be overcome. Since optical projection tomography imaging is nondestructive, staining for RNA or protein distribution must be performed in wholemount, but the density of heart muscle seriously restricts reagent penetration. As a result, staining can be disproportionate or exclusive for accessible surfaces, especially with antibodies. A method to improve tissue penetration and signal-to-noise ratios (using Dent’s fixative, bleaching and freeze–thaw cycles) for immunohistochemistry

781

has recently been published (Alanentalo et al., 2007), but its effectiveness with cardiac tissue has yet to be established. Finally, while gene expression patterns are detected using chromogenic substrates, these can only be detected by brightfield optical projection tomography imaging, the success of which is dependent on achieving an appropriate level of staining to allow detection, while avoiding entirely occluding light transmission through the tissue. Adoption of fluorescence in situ hybridization methods would overcome this difficulty, but reliable methods for use with mouse embryos have not yet been developed.

IV.  Episcopic imaging If alternative imaging methods cannot approach the resolution afforded by conventional histology, is it not best simply to reconstruct three-dimensional morphology using serial sections, following essentially the camera lucida approach pioneered by His and Ziegler a century and a half ago (Hopwood, 2002)? Especially with the option of computer-based interpolation to overcome gaps in histological data sets, this has proved a consistently beguiling possibility. Of course, while the resolution of individual section images can be extremely high, the effective threedimensional resolution of this approach is defined by the thickness of the individual sections. For reasons of technical convenience (not to mention the availability of the widest range of histological stains and compatibility with both immunohistochemistry and RNA in situ hybridization techniques), wax sections are usually the method of choice for embryo studies. With section thicknesses between 6 and 10 microns, models from this data are incapable of providing detailed morphological resolution, and if the sections in any series are used individually for several different purposes (e.g., alternate section for RNA in situ hybridization and antibody staining) the three-dimensional resolution of each individual model is correspondingly reduced. However, even when section thickness is minimized (for example by the use of plastic embedding resins), all attempts to retain the high-resolution of histological data in the three-dimensional reconstruction process have encountered the intractable problem of accurate image registration. In the absence of independent fiduciary markers, it has proved impossible to realign individual histological images while maintaining cellular and subcellular resolution. Indeed, variable nonlinear distortion of individual sections during histological procedures has compounded the problem. Where low-resolution models are adequate for addressing a particular biological question, it is possible to ignore the inherent difficulties of registration and by using a combination of isosurface rendering with data “smoothing” it is perfectly possible to generate useful three-dimensional models (Ruijter et al., 2004). Indeed, modeling of lineage

782

labeling, gene expression and cell proliferation patterns in this way has led to important insights in the process of heart morphogenesis (for example Soufan et al., 2003, 2004; de Lange et al., 2004; Hoogaars et al., 2004; Mommersteeg et al., 2006; Soufan et al., 2006). However, such models remain only gross and simplified approximations of the actual tissue or organ morphology. An alternative approach that can retain intricate morphological detail, but eliminate the need for section registration, is the use of episcopic imaging. Here, sequential images of the embedded specimen block face are captured during the sectioning process, rather than imaging individual sections that have been subject to histological staining. As a result, the images retain their alignment without the need for fiduciary markers, and the accuracy is determined entirely by the mechanical stability of the imaging apparatus. The technical challenge for episcopic imaging is therefore to devise methods to image embedded tissue in a convenient way that avoids repetitive surface staining and that eliminates (or at least minimizes) the contribution of tissues below the surface to the captured image (Weninger and Mohun, 2002; Rosenthal et al., 2004; Weninger et al., 2006). The initial procedure (episcopic fluorescence image capturing (EFIC)) relied on tissue autofluorescence for detection, using the incorporation of dyes within the wax embedding medium to block signal from below the block surface (Weninger and Mohun, 2002). However, autofluorescence has disadvantages as the basis for tissue detection. Signal strength and contrast obtained from different tissue and cell types varies, and cannot always provide useful representations of tissue architecture. Furthermore, the autofluorescence of early mouse embryos is rather weak, and EFIC proves most successful with embryos nearer to full-term.

PART  |  11  Cardiomics

1–2 m are therefore standard (i.e., 1–8 m3 per voxel). Unlike autofluorescence, the staining obtained in high resolution episcopic microscopy is not significantly attenuated in early embryos, and images more faithfully reflect the tissue structure and densities revealed by histology, albeit in monochrome (Fig. 2A–B). High resolution episcopic microscopy data can be analyzed either as two-dimensional image stacks or by three-dimensional rendering, and can reveal alterations in morphology (for example, AV junction defects) that are impossible to detect using lower-resolution techniques such as optical projection tomography and magnetic resonance imaging. Image alignment permits virtual resectioning of data stacks in any user-defined plane, greatly facilitating such analysis (Fig. 2C–E). A particularly useful feature of high resolution episcopic microscopy data is the broad gray-scale range of the

(A)

(B)

V.  High resolution episcopic microsocopy These limitations have been overcome in a more recent version of episcopic imaging, high resolution episcopic microscopy (HREM) (Weninger et al., 2006). In this method, samples are embedded in a methacrylate resin (JB4) containing fluorescent dyes (eosin and optionally, acridine orange) that serve both to stain the tissue and to provide a uniform fluorescent background that masks tissue below the block surface. Positive surface images are thereby obtained, and the method enjoys the high level of resolution afforded by plastic-embedded samples. High resolution episcopic microscopy is ideally suited to imaging mouse embryos, since the only practical constraints on image resolution are the optics used to visualize the block surface. As a result, the voxel resolution of three-dimensional models is limited only by the thickness of sections successively removed from the block surface. In practice, spatial resolutions of

(C)

(D)

(E)

Figure 2  High resolution episcopic microsocopy imaging of mouse embryo heart development. High resolution episcopic microsocopy images retain fine detail of tissue and organ structure, broadly compar­ able with conventional histology. (A, B) These are “virtual” orthogonal sections computed from high resolution episcopic microsocopy data obtained by transverse sectioning of an E10.5 embryo at 4 m thickness. (C–E) High resolution episcopic microsocopy imaging of thoracic organs from an E12.5 embryo with the imaged, four-chamber view (C) and orthogonal “virtual sections” at the level of the AV junction ((D), shortaxis) and in the plane of the interventricular septum, aortic and atrial venous valves (E).

Chapter  |  11.3  Imaging Cardiac Developmental Malformations in the Mouse Embryo

images that is ideally suited for three-dimensional visual­ ization by volume rendering. The advantages this offers are several. In lower-resolution imaging methods, reasonable three-dimensional models are only possible by reconstruction of object surfaces, (i.e., isosurface rendering), a method that, like its camera lucida ancestor, requires outlining (or “segmentation”) of the object on each successive image. Despite considerable advances in computer-based segmentation methods, these remain extremely laborious, almost always requiring considerable manual input or editing. Because of the inevitable imprecision in such editing, data “smoothing” (i.e., loss of data) is frequently employed to generate acceptable models. Modeling by surface rendering is also computationally intensive and slow on current desktop computers. In contrast, volume rendering is relatively rapid, and by adjusting the manner in which voxel gray values are mapped, it is possible to obtain a range of models varying in their effective transparency. Since it requires no user intervention to identify structures, the data is not “filtered” through the experimenter’s judgements on object continuity and structure. Instead, its success depends on the assumption that discrete anatomical structures yield reasonably consistent voxel gray values so that they appear as recognizable objects after rendering. This has largely proved to be the case (Fig. 3A). By taking advantage of software developments pioneered for medical imaging modalities such at CT and magnetic resonance, high resolution episcopic microscopy data can be manipulated and visualized in a variety of useful

783

ways in any of the available DICOM viewer software packages. Using these, customized color look-up tables can be applied to the data to visualize discrete tissue types; segmentation can be used to isolate discrete regions or structures for viewing or morphometry (e.g., valve dimensions, wall thickness and chamber or vessel volumes); selective removal of voxel data can be used for “virtual dissection” of the sample, allowing any feature or facet to be visualized. For example, the complex architecture of ventricular trabeculations, the arrangement of great vessel valve leaflets and the coronary ostia are all easily revealed through volume rendered high resolution episcopic microscopy data (Fig. 4A). Indeed, using software that combines segmentation by voxel gray-scale with connectivity (Wang and Smedby, 2007) it is possible to generate models of the major coronary arterial network in the embryonic heart (Fig. 4B,C). For some structures, for example those intimately associated with, or embedded within, adjacent tissue, volume rendering is of little use. In these cases, the resolution of high resolution episcopic microscopy data permits accurate surface rendering, although this usually requires manual outlining (Fig. 4D). High resolution episcopic microscopy imaging can be extended to encompass specific staining patterns generated using chromagenic substrates, and can in principle be used to examine gene expression patterns within the developing heart. This is achieved by imaging in red fluorescence wavelengths, conditions under which tissue imaging is minimal, but stain patterns remain clear (Weninger et al., 2006).

(A)

(B) Figure 3  Three-dimensional heart models from high resolution episcopic microsocopy data. Three-dimensional volume rendering of high resolution episcopic microsocopy data sets provides a simple way to study morphology of the developing heart. (A) E15.5 heart showing progressive erosion of data, first to reveal the pulmonary valve leaflets and then to the AV junction, revealing the mitral valve leaflets and the primary atrial septum. (B) E12.5 heart (with associated lungs, thymus and trachea) showing progressive removal of the atrial chambers. Note the intricate detail of atrial wall trabeculation, venous valve architecture and the primary atrial septum (arrows), which can be viewed from any orientation.

784

(A)

PART  |  11  Cardiomics

(B)

Little specialist apparatus is required for high resolution episcopic microscopy imaging, and by customizing standard microtome, fluorescent optics and imaging equipment it is possible to produce an entirely-automated system. Data analysis can be based entirely on open source software (for example ImageJ (http://rsb.info.nih.gov/ij/) and Osirix 3.0 DICOM viewer (http://www.osirix-viewer. com)), which in addition to economy (all software is free) offers the advantage that data sets and their analysis can be easily shared across the research community. Because high resolution episcopic microscopy imaging is dependent on physical sectioning of samples, its throughput is inevitably modest. Automated high resolution episcopic microscopy is currently capable of accumulating several thousand images per day, which approximates to 1–2 whole embryos or 2–4 hearts at E14.5 (assuming section thickness of 4 and 2 microns, respectively).

VI.  High-throughput phenotyping (C)

(D) Figure 4  Modeling the coronary vasculature. (A) Three-dimensional model showing the base of the aorta and the aortic valve leaflets of an E18.5 heart, eroded from the caudal side. The left and right coronary ostia (LO and RO) are clearly visible in the aorta wall. (B) Using data segmentation with connectivity, it is possible to volume-render the right and left coronary arteries (RCA and LCA) and their branching network. Note that in this embryo, the septal branch (arrow) originates from the left coronary artery, immediately adjacent to the coronary sinus. (C) The arrangement of the coronary arteries in the same heart is visualized by superimposing the heart model. The coronary arteries can be seen to originate from the aorta, lying behind the partially-eroded base of the pulmonary trunk. (D) Topology of the great arteries of a 14.5 dpc wild-type mouse embryo. Surface rendered three-dimensional model of the great arteries (red), trachea and principal bronchi (yellow) and thymus (semitransparent blue) (v: ventral view; l: view from the left; r: view from the right). Note the origin of the segmental arteries from the descending aorta.

However, as with optical projection tomography imaging, the sensitivity of this approach rests heavily on the success of whole-mount in situ hybridization techniques, with all the difficulties of tissue penetration that this entails. The high resolution episcopic microscopy procedure does generate a series of tissue sections during the imaging procedure, but the presence of plastic embedding medium has so far proved an insuperable barrier to using these for RNA in situ hybridization. They can, however, be used successfully for at least some histological staining procedures. (For example, hematoxylin and eosin work well, but no success has been reported with Masson’s trichrome.)

Like conventional histology, high resolution episcopic microscopy and optical projection tomography are techniques appropriate for detailed examination of individual specimens. However, there is also a need for alternative imaging methods that are better suited to high-throughput phenotyping. Consider, for example, the number of genes likely to be involved in mammalian heart development. Examination of the Mouse Genome Informatics database reveals that, of 4,373 genes with targeted mutations, 246 are associated with abnormal cardiac morphology. If extrapolated to 28,000 genes (the estimated number in the mouse genome), we expect that 1,500 genes are necessary for heart development. A major gap in our knowledge is what most of these cardiac developmental genes are, and how they genetically-interact to form a network. In theory, we should be able to predict the effect of genetic architecture on the gene-interaction network, and determine if a particular architecture is likely to result in network failure – i.e., in congenital heart disease. Identifying the genes within the cardiac genetic network systematically will require phenotypic analysis of all 28,000 mutants. It also becomes clear that to identify the interactions between the genes will require, at the very least, the analysis of 1,500*1,499/2 (i.e., 1.1 million) pair-wise genetic crosses and analysis of compound mutant embryos. These are challenging numbers. High-throughput mutagenesis has been successfully used in yeast and worm to deconstruct genetic interaction networks (Tong et al., 2004; Lehner et al., 2006) and in theory, the same type of approach could be used to identify the 1,500 genes and their interactions necessary for cardiac development in the mouse. This would be the first step necessary to build a cardiac genetic network. The mouse genome can be mutagenized in high-throughput using systematic, targeted gene-trapping (Friedel et al., 2005) or randomly, using

Chapter  |  11.3  Imaging Cardiac Developmental Malformations in the Mouse Embryo

gene-traps (Mitchell et al., 2001), transposon-based systems (Zagoraiou et al., 2001; Drabek et al., 2003) and genotoxic agents such as ethylnitrosourea (Justice et al., 1999; Herron et al., 2002) (see Chapter 11.2). However, the opacity of mouse embryos prevents easy identification of cardiac malformations without resort to specialist imaging techniques,

785

and this has hampered the use of genome-wide screens to study cardiogenesis in the mouse. One imaging method that has been used for phenotyping in mouse mutagenesis screens is ultrasound (see Chapter 11.2). Although this is an in vivo imaging technique capable of visualizing blood flow, it has very low resolution, and images

(A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

(I)

Figure 5  Cardiac malformations identified by magnetic resonance imaging. (A) Atrial septal defect primum. Transverse section through a cyclin E1/E2 doubly-deficient embryo showing left and right atria and ventricles (la, ra, lv, rv). An ostium primum-type of atrial septal defect (ASD) is present at the ventral margin of the primary atrial septum (pas). (B) Right atrial isomerism. Transverse section through a Pitx2c-deficient embryo showing a large atrial septal defect resulting in common atrium (A), with bilateral systemic venous sinuses (l-svs, r-svs) into which the left and right superior vena cava drain. (C) Ventricular septal defect. Section showing a ventricular septal defect (VSD) in the interventricular septum (ivs) in a Cited2-null embryo. Bamforth et al. (2001). (D, E) Right aortic arch and double-outlet right ventricle. Transverse section and three-dimensional reconstruction (right ventral oblique view) of a cyclinE1/E2-deficient embryo, showing a right-sided aortic arch (ao-a) passing to the right of the trachea (tr) and the esophagus (es) before continuing as the descending aorta (d-ao). The ascending aorta (a-ao) arises from the right ventricle (rv) as does the pulmonary artery (pa), resulting in a double-outlet right ventricle. (F, G) Aortic vascular ring. Transverse section and three-dimensional reconstruction (anterior view) respectively, showing bilateral aortic arches (ao-a) forming a vascular ring around the trachea (tr) and the esophagus (es). Also indicated is the thymus (th). (H, I) Three-dimensional reconstructions of a wild-type embryo showing right ventral oblique and ventral views, respectively. Note the aortic arch and pulmonary artery on the left of the trachea. Scale bars: 500 μm (axes: d: dorsal; v: ventral; r: right; l: left). Reproduced from Schneider and Bhattacharya (2004), with permission.

786

(A)

PART  |  11  Cardiomics

(B)

(C)

(E)

(F)

(D)

Figure 6  High-throughput, high-resolution magnetic resonance imaging. (A) Stack of 32 embryos embedded in an NMR tube. (B) Section through the long axis of the NMR tube showing embryos in eight layers. (C) Sagittal section through layer 8 showing the four embryos in this layer. (D–F) Transverse, sagittal, and coronal sections through individual embryos in layers 5, 1 and 4 respectively. The voxel size is 25.4  25.4  24.4 μm. Structures indicated are: spinal cord: sc; the right and left lungs, atria and ventricles: rl, ll, ra, la, rv, lv; primary atrial and interventricular septa: pas, ivs; mitral valve: mv; midbrain roof: mbr; midbrain: mb; mesencephalic vesicle: mes; thalamus: tha; hypothalamus: hy; pons: po; cerebellum: c; medulla oblongata: mo; pituitary: pit; tongue: t; thymus: th; left superior vena cava and main bronchus: lsvc, lmb; aorta: ao; liver: li; stomach: s; left adrenal and kidney: lad, lk; pancreas: pa; intestines: i; umbilical hernia: uh; aqueduct of Sylvius: aq; fourth ventricle: fv; inner ear: ie; larynx: lar; right ventricular outflow tract: rvot; spleen: sp; and testes: te). Scale bars  500 μm (axes: d: dorsal; v: ventral; r: right; l: left; a: anterior, p: posterior). Reproduced from Schneider et al. (2004), with permission.

individual structures rather poorly. For instance, in a recently published ENU mutagenesis screen using ultrasound (Yu et al., 2004), the predominant abnormalities identified were cardiac arrhythmia, outflow tract regurgitation, heart failure and hypertrophy. No malformations typical of congenital heart disease (septal, outflow tract, left–right patterning, or aortic arch defects) were identified directly. In addition, while ultrasound imaging is noninvasive, not all embryos in a litter can be imaged efficiently. Moreover, other embryonic structural abnormalities (e.g., lungs, thymus, adrenals, kidneys), which are frequently associated with congenital heart disease, are not efficiently identified using ultrasound. Other disadvantages include a requirement for operator expertise in performing and interpreting ultrasound, the necessity for live animal experimentation (i.e., imaging anesthetized female mice) and the requirement that either such imaging is established at the animal compound, or that each animal is transported to the ultrasound facility. The challenge has therefore been to develop alternative imaging approaches that deliver the high throughput necessary for mutagenesis screening programs, while providing sufficiently high-resolution images of mouse embryos to permit accurate phenotyping.

VII.  Magnetic resonance imaging Such an approach is provided by high-throughput magnetic resonance imaging (MRI). This provides a rapid way to simultaneously obtain images of multiple, fixed, mouse embryos at a level of resolution that allows identification of

many cardiac and other embryonic malformations (Figs 5; 6). It can be usefully applied to analyze late-developmental embryo stages (13.5–17.5 dpc) at reason­able cost in unattended overnight runs, employing magnetic resonance facilities which are normally used for daytime, live animal imaging. Thirty-two embryos (at 15.5 dpc) can currently be analyzed simultaneously in single overnight runs, at an experimental resolution of 46  46  38 um/voxel (Schneider et al., 2004). This resolution, although limited, has allowed detection of both simple and complex cardiovascular malformations (e.g., ASD, VSD, DORV, CAT, TGA, L/R patterning, or aortic arch defects), at time points ranging from 13.5 to 17.5 dpc, with high accuracy and throughput (Schneider et al., 2004). It has also allowed identification of structural and topological abnormalities in other organs (e.g., brain, palate, thymus, adrenals, kidneys, lungs and gut). The resulting image data can also be used to measure organ and embryo dimensions and volumes accurately accurately. This additional information is essential, as approximately 25% of congenital heart disease is associated with a noncardiac malformation (Boughman et al., 1993). Importantly for a high-throughput technique, sample preparation and set up for 32 embryos takes under two hours, while data analysis by skilled observers typically takes 30 minutes per embryo. All other steps are automated. Unlike ultrasound, all embryos in a litter are equally amenable to imaging, and data analysis can be performed off-line. Overall, high-throughput magnetic resonance imaging has the potential to image over 10,000

Chapter  |  11.3  Imaging Cardiac Developmental Malformations in the Mouse Embryo

embryos per year, the major bottleneck being data analysis, which requires an experienced observer. Furthermore, since it utilizes fixed embryos for imaging, there is no need for the imaging facility to be located at the site of the mutagenesis screen. As an example, in the last two years, more than 4,000 embryos have been imaged using the high-throughput magnetic resonance imaging facility at Oxford University (S. B., unpublished data), resulting in the identification or phenotypic characterization of several gene mutations (CyclinD, CyclinE, Ptdsr, Pinch1, Flna, Sox4, Cited2 and Pitx2) important for cardiac development (Geng et al., 2003; Schneider et al., 2003a,b,c; Bamforth et al., 2004; Kozar et al., 2004; Schneider and Bhattacharya, 2004; Schneider et al., 2004; Bogani et al., 2005; Hart et al., 2006; Liang et al., 2007). ENU mutagenesis screens conducted at another site (MRC Harwell) are now being screened by magnetic resonance imaging to identify cardiac malformations, with the goal of identifying the relevant genes and undertaking gene-environment interaction studies. In the current protocol for high-throughput multiembryo imaging, embryos (routinely at 15.5 dpc) are exsanguinated prior to imaging. This is essential, as blood creates a strong magnetic resonance contrast. The left forelimb and the yolk sac are removed and flash frozen for DNA extraction and genotyping. Embryos are then fixed in 4% paraformaldehyde containing 2 mM Gd-DTPA (a magnetic resonance contrast agent) for approximately one week. For older embryos, an incision in the skin of the dorsal lumbar region facilitates full penetration of fixative. Fixed embryos are then embedded in eight layers of four, in 1% agarose containing 2 mM Gd-DTPA. (Selective removal of hindlimbs and tail in unique combinations permits unequivocal identification of each within any layer.) The entire embryo stack is imaged in a Varian VNMRS magnetic resonance system comprising a 9.4 Tesla (400 MHz) horizontal magnet, a shielded gradient system with a maximal gradient strength of 1 Tesla/m, and a quadrature-driven birdcage-type coil with an inner diameter of 28 mm. A matrix size of 608  608  1408 at a field of view of 26  26  50 mm yields an experimental resolution of 43  43  36 m. The total experimental time is 12.3 hours. The raw magnetic resonance data is reconstructed into a stack of 2,048 (for 32 embryos) two-dimensional TIFF files using purpose-written software (Schneider et al., 2004). Individual image stacks are analyzed using the Amira software (www.amiravis.com/) package, and three-dimensional reconstructions made by isosurface rendering after image segmentation. Of course, a major factor limiting the use of ex vivo embryo magnetic resonance imaging is the very high equipment cost (£2 million). However, most major academic centers now have appropriate magnetic resonance imaging facilities which are typically used for physiological imaging during the day. Their overnight use for embryo

787

imaging would minimize magnet downtime and maximize returns on an expensive investment. Unlike optical projection tomography or high resolution episcopic microscopy, embryo magnetic resonance imaging cannot image gene expression patterns, but this is of little significance for high-throughput screening. More important is the current (A)

(B)

(C)

Figure 7  High resolution episcopic microsocopy analysis from magnetic resonance imaging prescreened embryos. Magnetic resonance imaging scans of a 15.5 dpc mouse embryo suggested the presence of a right retro-esophageal artery and subcutaneous edema. (A). Confirmation of right retro-esophageal artery (al) from high resolution episcopic microsocopy analysis of the same embryo. Note that the artery passes through a sympathetic ganglion (g). (B) Malconnection and aberrant course of the umbilical vein (uv), which directly drains into the right atrium. Note the very small inferior vena cava (ivc). (C) Small ventricular septal defect (asterisk). Neither of the malformations displayed in (B) and (C) were detected by magnetic resonance imaging (sc: spinal chord; br: brain; to: tongue; ie: inner ear; li: liver; l: lung; t: trachea; e: esophagus; ao: aorta; vn: vagus nerve; ga: stomach; aa: appendage of the atrium; pv: pulmonary valve; vs: ventricular septum; co: rib).

788

PART  |  11  Cardiomics

practical limit on imaging resolution. Although many of the types of lesions seen in congenital heart disease can be resolved in the current protocol, it is not suitable for identifying smaller structures such as coronary vasculature, nor can it reveal detailed myocardial architecture or valvular structure. It has not been possible to detect these small structures even using a higher-resolution magnetic resonance imaging approach (i.e., 25.4  25.4  26 um/ voxel) (Schneider et al., 2003a,c). In addition, imaging of younger embryos (prior to 13.5 dpc) is difficult, due to their small size and lack of magnetic resonance imaging contrast in relatively acellular areas. Within these constraints, however, magnetic resonance imaging offers the best available combination of resolution and throughput for phenotypic screening in mutagenesis studies. In practice, the most important limitation to throughput proves not to be the imaging itself, but rather the observer expertise and training necessary for rapid, visual screening of data. Furthermore, because image data segmentation remains time-consuming, it has not proved practical to perform quantitative analysis of the heart or other organs

routinely, despite the potential significance of this data. There is, therefore, a great need to develop automated analytical procedures such as mechanical deformation approaches for rapidly-extracting information from the image data sets.

VIII.  Conclusions: a phenotyping pipeline Mouse models are proving key to advancing our understanding of cardiac development and the etiology of congenital heart disease. Imaging such a complex and dynamic structure as the developing heart plays a crucial role in such studies, underlying accurate assessment of phenotype. New anatomical imaging techniques such as optical projection tomography, high resolution episcopic microscopy and magnetic resonance imaging provide ways to supplement conventional histology with comprehensive image data, yielding three-dimensional models that allow systematic, quantitative comparisons of structure. These imaging methods have complementary strengths, and by combining them it is

Harvesting and Fixation

Gd-DPTA impregnation and agarose embedding (32 embryos/tube)

MRI scanning 12 hours 2000 images voxel: 25.4 � 25.4 � 24.4 µm

Extraction of embryos from tube

Data analysis

Selection of embryos needing HREM analysis

Resin embedding (1 day) Section collection and staining (1 day)

HREM data generation 4–7 hours 1500–2500 images voxel: 1.07 � 1.07 � 2 µm

Data analysis and 3D modeling

Results

Figure 8  Flow chart of a phenotyping pipeline. A phenotyping pipeline for high-throughput screening is possible by combining magnetic resonance imaging with more detailed analysis provided by selective analysis of embryos by high resolution episcopic microsocopy. (As a nondestructive technique, optical projection tomography could also be incorporated into this pipeline.)

Chapter  |  11.3  Imaging Cardiac Developmental Malformations in the Mouse Embryo

possible to envisage a powerful cardiac phenotyping pipeline (Fig. 7). In mutagenesis programs, for example, highthroughput screening of 15.5 dpc embryos by magnetic resonance imaging can identify many cardiac abnormalities; subsequent optical projection tomography or high resolution episcopic microscopy imaging of selected embryos can then be used for high-resolution analysis to identify cardiovascular and extra-cardiac lesions not apparent by magnetic resonance imaging (Fig. 8). In pilot studies, these have included myocardial thinning, muscular VSD, persistent vitelline veins, absent ductus venosus, duplex inferior vena cava and thyroid anomalies (Pieles et al., 2007). Finally, where mutant phenotypes are identified, high-resolution methods can be targeted to studying earlier developmental stages, in order to investigate their developmental progression. In this way, the new imaging methods can be flexibly combined, providing powerful tools for studying mammalian heart development.

References Alanentalo, T., Asayesh, A., Morrison, H., Loren, C.E., Holmberg, D., Sharpe, J., Ahlgren, U., 2007. Tomographic molecular imaging and 3D quantification within adult mouse organs. Nat. Methods 4, 31–33. Bamforth, S.D., Braganca, J., Farthing, C.R., Schneider, J.E., Broadbent, C., Michell, A.C., Clarke, K., Neubauer, S., Norris, D., Brown, N.A., Anderson, R.H., Bhattacharya, S., 2004. Cited2 controls left–right patterning and heart development through a Nodal-Pitx2c pathway. Nat. Genet. 36, 1189–1196. Bogani, D., Willoughby, C., Davies, J., Kaur, K., Mirza, G., Paudyal, A., Haines, H., McKeone, R., Cadman, M., Pieles, G., Schneider, J.E., Bhattacharya, S., Hardy, A., Nolan, P.M., Tripodis, N., Depew, M.J., Chandrasekara, R., Duncan, G., Sharpe, P.T., Greenfield, A., Denny, P., Brown, S.D., Ragoussis, J., Arkell, R.M., 2005. Dissecting the genetic complexity of human 6p deletion syndromes by using a region-specific, phenotype-driven mouse screen. Proc. Natl. Acad. Sci. USA 102, 12477–12482. Boughman, J.A., Neill, C.A., Ferencz, C., Loffredo, C., 1993. The genetics of congenital heart disease. In: Ferencz, C., Rubin, J.D., Loffredo, C.A., Magee, C.M. (Eds.), Epidemiology of Congenital Heart Disease: The Baltimore-Washington Infant Study 1981–1989. Futura Publishing, Mount Kisco, NY, pp. 123–133. Breckenridge, R., Kotecha, S., Towers, N., Bennett, M., Mohun, T., 2007. Pan-myocardial expression of Cre recombinase throughout mouse development. Genesis 45, 135–144. Bryson-Richardson, R.J., Berger, S., Schilling, T.F., Hall, T.E., Cole, N.J., Gibson, A.J., Sharpe, J., Currie, P.D., 2007. FishNet: an online database of zebrafish anatomy. BMC Biol. 5, 34. Burn, J., Goodship, J., 2002. Congenital heart disease. In: Rimoin, D.L., Connor, J.M., Pyeritz, R.E., Korf, B.R. (Eds.), Principles and Practice of Medical Genetics. Churchill Livingstone, London, UK. Clark, E.B., 2001. Etiology of congenital cardiac malformations: epidemiology and genetics. In: Allen, H.D., Gutgessell, H.P., Clark, E.B., Driscoll, D.J. (Eds.), Moss and Adams’ Heart Disease in Infants, Children, and Adolescents. Lipincott Williams and Wilkins, Philadelphia, PA. de Lange, F.J., Moorman, A.F., Anderson, R.H., Manner, J., Soufan, A.T., de Gier-de Vries, C., Schneider, M.D., Webb, S., van den Hoff, M.J., Christoffels, V.M., 2004. Lineage and morphogenetic analysis of the cardiac valves. Circ. Res. 95, 645–654.

789

DeLaurier, A., Schweitzer, R., Logan, M., 2006. Pitx1 determines the morphology of muscle, tendon, and bones of the hindlimb. Dev. Biol. 299, 22–34. Drabek, D., Zagoraiou, L., deWit, T., Langeveld, A., Roumpaki, C., Mamalaki, C., Savakis, C., Grosveld, F., 2003. Transposition of the Drosophila hydei Minos transposon in the mouse germ line. Genomics 81, 108–111. Friedel, R.H., Plump, A., Lu, X., Spilker, K., Jolicoeur, C., Wong, K., Venkatesh, T.R., Yaron, A., Hynes, M., Chen, B., Okada, A., McConnell, S.K., Rayburn, H., Tessier-Lavigne, M., 2005. 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, 13188–13193. Geng, Y., Yu, Q., Sicinska, E., Das, M., Schneider, J.E., Bhattacharya, S., Rideout, W., Bronson, R.T., Gardner, H., Sicinski, P., 2003. Cyclin E ablation in the mouse. Cell 114, 431–443. Hart, A.W., Morgan, J.E., Schneider, J., West, K., McKie, L., Bhattacharya, S., Jackson, I.J., Cross, S.H., 2006. Cardiac malformations and midline skeletal defects in mice lacking filamin A. Hum. Mol. Genet. 15, 2457–2467. Herron, B.J., Lu, W., Rao, C., Liu, S., Peters, H., Bronson, R.T., Justice, M.J., McDonald, J.D., Beier, D.R., 2002. Efficient generation and mapping of recessive developmental mutations using ENU mutagenesis. Nat. Genet. 30, 185–189. Hiruma, T., Nakajima, Y., Nakamura, H., 2002. Development of pharyngeal arch arteries in early mouse embryo. J. Anat. 201, 15–29. Hoffman, J.I., Kaplan, S., 2002. The incidence of congenital heart disease. J. Am. Coll. Cardiol. 39, 1890–1900. Hoogaars, W.M., Tessari, A., Moorman, A.F., de Boer, P.A., Hagoort, J., Soufan, A.T., Campione, M., Christoffels, V.M., 2004. The transcriptional repressor Tbx3 delineates the developing central conduction system of the heart. Cardiovasc. Res. 62, 489–499. Hopwood, N., 2002. Embryos in wax: models from the Ziegler studio, with a reprint of “Embryological wax models” by Friedrich Ziegler. Ijpenberg, A., Perez-Pomares, J.M., Guadix, J.A., Carmona, R., PortilloSanchez, V., Macias, D., Hohenstein, P., Miles, C.M., Hastie, N.D., Munoz-Chapuli, R., 2007. Wt1 and retinoic acid signaling are essential for stellate cell development and liver morphogenesis. Dev. Biol. 312, 157–170. Johnson, G.A., Cofer, G.P., Fubara, B., Gewalt, S.L., Hedlund, L.W., Maronpot, R.R., 2002. Magnetic resonance histology for morphologic phenotyping. J. Magn. Reson. Imaging 16, 423–429. Justice, M.J., Noveroske, J.K., Weber, J.S., Zheng, B., Bradley, A., 1999. Mouse ENU mutagenesis. Hum. Mol. Genet. 8, 1955–1963. King, T., Bland, Y., Webb, S., Barton, S., Brown, N.A., 2002. Expression of Peg1 (Mest) in the developing mouse heart: involvement in trabeculation. Dev. Dyn. 225, 212–215. Kondo, S., Suzuki, R., Yamazaki, K., Aihara, K., 1993. Application of corrosion cast method for scanning electron microscopic observation of mouse embryo vasculature. J. Electron. Microsc. (Tokyo) 42, 14–23. Kozar, K., Ciemerych, M.A., Rebel, V.I., Shigematsu, H., Zagozdzon, A., Sicinska, E., Geng, Y., Yu, Q., Bhattacharya, S., Bronson, R.T., Akashi, K., Sicinski, P., 2004. Mouse development and cell proliferation in the absence of D-cyclins. Cell 118, 477–491. Kuehl, K.S., Loffredo, C.A., 2002. Risk factors for heart disease associated with abnormal sidedness. Teratology 66, 242–248. Lee, K., Avondo, J., Morrison, H., Blot, L., Stark, M., Sharpe, J., Bangham, A., Coen, E., 2006. Visualizing plant development and gene expression in three dimensions using optical projection tomography. Plant Cell 18, 2145–2156.

790

Lehner, B., Crombie, C., Tischler, J., Fortunato, A., Fraser, A.G., 2006. Systematic mapping of genetic interactions in Caenorhabditis elegans identifies common modifiers of diverse signaling pathways. Nat Genet. 38, 896–903. Liang, X., Sun, Y., Schneider, J., Ding, J.H., Cheng, H., Ye, M., Bhattacharya, S., Rearden, A., Evans, S., Chen, J., 2007. Pinch1 is required for normal development of cranial and cardiac neural crestderived structures. Circ. Res. 100, 527–535. Loffredo, C.A., Hirata, J., Wilson, P.D., Ferencz, C., Lurie, I.W., 2001a. Atrioventricular septal defects: possible etiologic differences between complete and partial defects. Teratology 63, 87–93. Loffredo, C.A., Wilson, P.D., Ferencz, C., 2001b. Maternal diabetes: an independent risk factor for major cardiovascular malformations with increased mortality of affected infants. Teratology 64, 98–106. Martinez-Frias, M.L., 2001. Heterotaxia as an outcome of maternal diabetes: an epidemiological study. Am. J. Med. Genet. 99, 142–146. McGurk, L., Morrison, H., Keegan, L.P., Sharpe, J., O’Connell, M.A., 2007. Three-dimensional imaging of Drosophila melanogaster. PLoS ONE 2, e834. Miller, S.F., Summerhurst, K., Runker, A.E., Kerjan, G., Friedel, R.H., Chedotal, A., Murphy, P., Mitchell, K.J., 2007. Expression of Plxdc2/ TEM7R in the developing nervous system of the mouse. Gene. Expr. Patterns 7, 635–644. 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., Pangilinan, E., Wakenight, P., Tessier-Lavigne, M., Skarnes, W.C., 2001. Functional analysis of secreted and transmembrane proteins critical to mouse development. Nat. Genet. 28, 241–249. Mommersteeg, M.T., Soufan, A.T., de Lange, F.J., van den Hoff, M.J., Anderson, R.H., Christoffels, V.M., Moorman, A.F., 2006. Two distinct pools of mesenchyme contribute to the development of the atrial septum. Circ. Res. 99, 351–353. Petersen, S., Peto, V., Rayner, M., 2003. Congenital heart disease statistics 2003. British Heart Foundation Statistics available at: www. heartstats.org. Pieles, G., Geyer, S.H., Szumska, D., Schneider, J., Neubauer, S., Clarke, K., Dorfmeister, K., Franklyn, A., Brown, S.D., Bhattacharya, S., Weninger, W.J., 2007. MicroMRI-HREM pipeline for high-throughput, high-resolution phenotyping of murine embryos. J. Anat. 211, 132–137. Pierpont Jr, M.E., Basson, C.T., Benson, D.W., Gelb, B.D., Giglia, T.M., Goldmuntz, E., McGee, G., Sable, C.A., Srivastava, D., Webb, C.L., 2007. Genetic basis for congenital heart defects: Current knowledge. A scientific statement from the American Heart Association Congenital Cardiac Defects Committee, Council on Cardiovascular Disease in the Young, endorsed by the American Academy of Pediatrics. Circulation 115, 3015–3038. Rosenthal, J., Mangal, V., Walker, D., Bennett, M., Mohun, T.J., Lo, C.W., 2004. Rapid high resolution three dimensional reconstruction of embryos with episcopic fluorescence image capture. Birth Defects Res. C Embryo. Today 72, 213–223. Ruijter, J.M., Soufan, A.T., Hagoort, J., Moorman, A.F., 2004. Molecular imaging of the embryonic heart: Fables and facts on 3D imaging of gene expression patterns. Birth Defects Res. C Embryo. Today 72, 224–240. Sarma, S., Kerwin, J., Puelles, L., Scott, M., Strachan, T., Feng, G., Sharpe, J., Davidson, D., Baldock, R., Lindsay, S., 2005. 3D modelling, gene expression mapping and post-mapping image analysis in the developing human brain. Brain Res. Bull. 66, 449–453.

PART  |  11  Cardiomics

Schaefer-Graf, U.M., Buchanan, T.A., Xiang, A., Songster, G., Montoro, M., Kjos, S.L., 2000. Patterns of congenital anomalies and relationship to initial maternal fasting glucose levels in pregnancies complicated by type 2 and gestational diabetes. Am. J. Obstet. Gynecol. 182, 313–320. Schneider, J.E., Bamforth, S.D., Farthing, C.R., Clarke, K., Neubauer, S., Bhattacharya, S., 2003a. High-resolution imaging of normal anatomy, and neural and adrenal malformations in mouse embryos using magnetic resonance microscopy. J. Anat. 202, 239–247. Schneider, J.E., Bamforth, S.D., Farthing, C.R., Clarke, K., Neubauer, S., Bhattacharya, S., 2003b. Rapid identification and 3D reconstruction of complex cardiac malformations in transgenic mouse embryos using fast gradient echo sequence magnetic resonance imaging. J. Mol. Cell Cardiol. 35, 217–222. Schneider, J.E., Bamforth, S.D., Grieve, S.M., Clarke, K., Bhattacharya, S., Neubauer, S., 2003c. High-resolution, high-throughput magnetic resonance imaging of mouse embryonic anatomy using a fast grad­ ient-echo sequence. Magma 16, 43–51. Schneider, J.E., Bhattacharya, S., 2004. Making the mouse embryo transparent: identifying developmental malformations using magnetic resonance imaging. Birth Defects Res. C Embryo. Today 72, 241–249. Schneider, J.E., Bose, J., Bamforth, S.D., Gruber, A.D., Broadbent, C., Clarke, K., Neubauer, S., Lengeling, A., Bhattacharya, S., 2004. Identification of cardiac malformations in mice lacking Ptdsr using a novel high-throughput magnetic resonance imaging technique. BMC Dev. Biol. 4, 16. Sharpe, J., Ahlgren, U., Perry, P., Hill, B., Ross, A., Hecksher-Sorensen, J., Baldock, R., Davidson, D., 2002. Optical projection tomography as a tool for 3D microscopy and gene expression studies. Science 296, 541–545. Soufan, A.T., Ruijter, J.M., van den Hoff, M.J., de Boer, P.A., Hagoort, J., Moorman, A.F., 2003. Three-dimensional reconstruction of gene expression patterns during cardiac development. Physiol. Genomics 13, 187–195. Soufan, A.T., van den Hoff, M.J., Ruijter, J.M., de Boer, P.A., Hagoort, J., Webb, S., Anderson, R.H., Moorman, A.F., 2004. Reconstruction of the patterns of gene expression in the developing mouse heart reveals an architectural arrangement that facilitates the understanding of atrial malformations and arrhythmias. Circ. Res. 95, 1207–1215. Soufan, A.T., van den Berg, G., Ruijter, J.M., de Boer, P.A., van den Hoff, M.J., Moorman, A.F., 2006. Regionalized sequence of myocardial cell growth and proliferation characterizes early chamber formation. Circ. Res. 99, 545–552. Tong, A.H., Lesage, G., Bader, G.D., Ding, H., Xu, H., Xin, X., Young, J., Berriz, G.F., Brost, R.L., Chang, M., et al., 2004. Global mapping of the yeast genetic interaction network. Science 303, 808–813. Wang, C., Smedby, O., 2007. Coronary artery segmentation and skeletonization based on competing fuzzy connectedness tree. Med. Image Comput. Comput. Assist. Interv. Int. Conf. 10, 311–318. Watkins, M.L., Rasmussen, S.A., Honein, M.A., Botto, L.D., Moore, C.A., 2003. Maternal obesity and risk for birth defects. Pediatrics 111, 1152–1158. Webb, S., Brown, N.A., Anderson, R.H., 1997. Cardiac morphology at late fetal stages in the mouse with trisomy 16: consequences for different formation of the atrioventricular junction when compared to humans with trisomy 21. Cardiovasc. Res. 34, 515–524. Weninger, W.J., Mohun, T., 2002. Phenotyping transgenic embryos: a rapid 3-D screening method based on episcopic fluorescence image capturing. Nat. Genet. 30, 59–65.

Chapter  |  11.3  Imaging Cardiac Developmental Malformations in the Mouse Embryo

Weninger, W.J., Geyer, S.H., Mohun, T.J., Rasskin-Gutman, D., Matsui, T., Ribeiro, I., Costa Lda, F., Izpisua-Belmonte, J.C., Muller, G.B., 2006. High-resolution episcopic microscopy: a rapid technique for high detailed 3D analysis of gene activity in the context of tissue architecture and morphology. Anat. Embryol. (Berlin) 211, 213–221. Yashiro, K., Shiratori, H., Hamada, H., 2007. Haemodynamics determined by a genetic programme govern asymmetric development of the aortic arch. Nature 450, 285–288.

791

Yu, Q., Shen, Y., Chatterjee, B., Siegfried, B.H., Leatherbury, L., Rosenthal, J., Lucas, J.F., Wessels, A., Spurney, C.F., Wu, Y.J., Kirby, M.L., Svenson, K., Lo, C.W., 2004. ENU induced mutations causing congenital cardiovascular anomalies. Development 131, 6211–6223. Zagoraiou, L., Drabek, D., Alexaki, S., Guy, J.A., Klinakis, A.G., Langeveld, A., Skavdis, G., Mamalaki, C., Grosveld, F., Savakis, C., 2001. In vivo transposition of Minos, a Drosophila mobile element, in mammalian tissues. Proc. Natl. Acad. Sci. USA 98, 11474–11478.

Chapter 11.4

Proteomic Strategies for Understanding Cardiac Function, Development, and Disease Charis Himeda and Steve Hauschka Department of Biochemistry, University of Washington, Seattle, WA, USA

I.  Introduction: the need for proteomics in cardiac analysis Sequencing the human genome has brought to light an abundance of novel genes with unknown functions, and parallel RNA and protein studies have shown that pre- and posttranslational modifications allow a single gene to encode multiple protein products. Because the functional complexity of an organism far exceeds the complexity of its genome, proteomics – analysis of the protein complement of tissues, cells and subcellular compartments – has emerged as a powerful tool to understand the dynamic behavior of cells. Proteomic strategies have been developed for the analysis of global changes in protein expression, as well as the identification of single proteins within complex mixtures, seminal achievements which underscore the power and versatility of the field. The integration of genomic and proteomic information should provide us with a better understanding of cardiac function in both health and disease. The potential applications of proteomics in cardiac analysis include: (1) detection of altered protein expression or modifications associated with cardiac dysfunction; (2) identification of cardiac-specific autoantigens in heart disease and post-transplantation; (3) elucidation of the proteins, signaling pathways, and mechanisms involved in normal and dysfunctional cardiac biology; (4) elucidation of signaling pathways involved in mobilizing cardiac stem cells for repair of the damaged myocardium; and (5) analysis of protein changes during cardiac development. This chapter will review some of the major contributions in the field, focusing particularly on the usefulness of proteomics as a tool for cardiac transcription factor identification. Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

II.  Proteomics and cardiac disease Identifying which proteins are up- or down-regulated in heart disease has critical implications for developing prognostic/diagnostic markers and identifying therapeutic targets. Thus far, the use of proteomics for cardiac disease analysis has focused primarily on measuring changes in protein abundance in dilated cardiomyopathy (DCM), a disorder characterized by progressive dilation of the left or both ventricles. Over 100 proteins have been observed to change abundance in DCM, mostly decreasing in the disease state (Pleissner et al., 1995, 1997; Thiede et al., 1996; Corbett et al., 1998; Heinke et al., 1998, 1999; Weekes et al., 1999). These proteins fall into three broad categories: (1) cytoskeletal and myofibrillar; (2) mitochondrial and energy-production; and (3) stress response. In the hearts of DCM patients, proteomics revealed an increase in levels of ubiquitin carboxyl-terminal hydrolase and overall protein ubiquitination, suggesting a role for proteolysis in the pathology of dilated cardiomyopathy (Weekes et al., 2003). Key protein alterations have also been observed in other cardiac diseases, including: atrial fibrillation, in which the ventricular form of myosin light chain 2 is increased (Lai et al., 2004); coronary atherosclerosis, in which ferritin light chain is increased (You et al., 2003); right ventricular hypertrophy, in which a shift from fatty acid to glucose metabolism is indicated (Faber et al., 2005); and ischemia, in which proteins associated with mitochondrial respiration, energy metabolism and stress response show differential abundance (White et al., 2005, 2006; Kim et al., 2006). In a rabbit model of the latter disease, complement inhibition

793

794

was shown to preserve the levels of 11 anti-inflammatory and structural proteins (Buerke et al., 2006). The effects of aging and alcohol/lead toxicity on the expression of cardiac proteins have also been investigated. In aging monkey hearts, gender-specific differences in glycolytic and mitochondrial metabolic proteins have been detected (Yan et al., 2004), and chronic alcohol feeding decreases the levels of contractile proteins, particularly myosin heavy chain, which may explain the reduced contractile function in clinical cases of alcohol-induced heart disease (Patel et al., 1997, 2000). However, a twodimensional gel electrophoresis analysis of myocardial proteins from rabbits exposed to lead indicated no significant changes among ~800 cardiac proteins (Toraason et al., 1997). Although not a primary cardiomyopathy, diabetes mellitus leads to cardiovascular complications, which are the major cause of death in diabetic patients. A proteomic analysis of glucose-perfused rat hearts demonstrated altered levels of specific metabolic proteins and reduced levels of many anti-apoptotic factors, suggesting a mechanism for heart damage in diabetic hyperglycemia (Warda et al., 2007). Pathways involved in normal cardiac biology have also been explored via proteomics; e.g., using a transgenic mouse model, 121 proteins were observed to be ubiquitinated in the mouse heart, indicating a major role for ubiquitination in cardiac function (Jeon et al., 2007). For the diseases under study, the development of diagnostic markers and drug targets awaits further characterization of altered factors and their role in disease progression. Progress has also been made toward identifying cardiac-specific antigens that elicit immune responses during cardiac disease and following heart transplantation. Using proteomic approaches, several cardiac antigens have been determined to react with autoantibodies in dilated cardiomyopathy (Latif et al., 1993; Pohlner et al., 1997) and myocarditis (Pankuweit et al., 1997), as well as posttransplantation (Latif et al., 1995; Wheeler et al., 1995). Additionally, novel markers of cardiac allograft rejection have been identified using proteomics (Borozdenkova et al., 2004). Although not examined at the protein level, chamber-specific gene expression profiles have been delineated in the mouse heart, as well as in the failing human heart, using microarray analysis (Tabibiazar et al., 2003; Ellinghaus et al., 2005). Correlations of these transcriptional changes with the accompanying proteomic changes should be highly informative. Similar studies have been carried out in skeletal muscle. Proteomics has been used to determine altered protein expression in a model of skeletal muscle wasting (Duan et al., 2006), to investigate mechanisms of protection from oxidative damage (Gelfi et al., 2004), and to delineate the different patterns of metabolic and structural protein expression in slow- versus fast-twitch muscle (Okumura et al., 2005). These experiments are important because the

PART  |  11  Cardiomics

success of any muscle therapy is critically dependent on understanding the pathways and mechanisms involved in disease causality and progression. Analysis of the cardiac proteome in health and disease has relied heavily on the use of two-dimensional gel electrophoresis (2D-E) coupled with mass spectrometry. In 2D-E, proteins are separated in one dimension by isoelectric point (pI), then in the other dimension by relative molecular mass (Mr). The ability of 2D-E to generate reproducible, high-resolution maps and to separate thousands of proteins for subsequent mass spectrometry character-ization has given it widespread usefulness in proteomic analysis. Unfortunately, the dynamic range of protein abundance in cells exceeds that of 2D-E, which often results in multiple proteins comigrating to the same spot on a two-dimensional gel. In addition, it remains difficult to resolve highly acidic/basic, hydrophobic, and low-abundance proteins by 2D-E. Multidimensional protein identification technology (MudPIT) is a nongel-based approach that was developed to circumvent these shortcomings (Washburn et al., 2001). In this system, peptides are loaded onto a microcapillary column with two separate chromatographic phases, followed directly by tandem mass spectrometry. Because proteins are cleaved to peptides prior to chromatography, this strategy facilitates protein identifications independent of pI and Mr. Washburn et al. pioneered this approach with the identification of 1,484 representative yeast proteins with less bias than traditional 2D-E. Importantly, proteins with extreme pI or Mr, as well as low-abundance and transmembrane proteins, were identified with equal sensitivity to other proteins (Washburn et al., 2001). An analogous strategy was used to identify factors associating with yeast TFIID (Sanders et al., 2002). More recently, Kislinger et al. used MudPIT technology to identify hundreds of tissue-specific proteins in healthy and diseased hearts, including transcriptional regulators and proteins linked to cardiac disease (Kislinger et al., 2005). These high-throughput methods may be useful for the comprehensive analysis of cardiac proteins; however, for the identification of specific factors, more sensitive techniques are required (discussed in Section III.C).

III.  Proteomic identification   of cardiac transcription   factors While significant advances have been made in detecting alterations in known proteins associated with cardiac disease, the usefulness of proteomics as a strategy for identifying novel proteins associated with normal cardiac development and physiology has been underutilized. This section will focus specifically on proteomic strategies for identifying cardiac transcription factors.

Chapter  |  11.4  Proteomic Strategies for Understanding Cardiac Function, Development, and Disease

After a cis-regulatory element is identified within a gene and determined to exhibit specific factor binding, its sequence is commonly searched against databases of known transcription factor binding sites. If one or more candidates are indicated, these can be further analyzed via gel-shift or chromatin immunoprecipitation (ChIP) studies. Because it detects protein–DNA interactions within living cells, ChIP analysis has the potential advantage, relative to proteomics, of identifying “true” in vivo molecular associations. However, since ChIP depends on the existence of high-affinity, specific antisera for immunoprecipitating the candidate factor and cross-linked DNA fragments containing the factor-binding site, the technique does not work when such antisera are unavailable. Furthermore, since specific DNA-binding factor candidates must be hypothesized in order to select the appropriate antisera, ChIP data is essentially confirmatory, and is unable to suggest alternative candidates. An additional caveat of the ChIP technique is that it only demonstrates occupancy within the region encompassed by PCR primers (and possibly outside of the primers, since chromatin is sheared to a ladder of 200–1,000 bp), not occupancy at a specific binding site. In contrast to ChIP analysis, proteomics has the potential to provide an unbiased identification of candidate DNA-binding factors. Furthermore, proteomic strategies supplant ChIP analysis for cases in which control elements either match the putative binding sites for multiple factors, or have no significant similarity to any known consensus motifs. The factor(s) binding such sequences are good candidates for proteomic identification. For example, while the Trex site in the Muscle creatine kinase (MCK) enhancer is similar to both TEF-1- and GATA-binding sites (Fig. 1A), we demonstrated through quantitative proteomics that the Trex-binding factor corresponded to Six4, a homeodomain protein of the Six/sine oculis family, in skeletal muscle, and Six5 in cardiac muscle (Himeda et al., 2004). Six proteins recognize MEF3 motifs in the

regulatory regions of their target genes; however, because the Trex site in the MCK enhancer deviates from the previously-established MEF3 motif in 2 out of 7 bp (Fig. 1A), this relationship was not identifiable by in silico screening against a transcription factor binding site database. Thus, without applying a quantitative proteomic strategy to this problem, a major transcriptional regulator of MCK would likely have remained unknown. In a related study, we found that a control element in the MCK promoter, which is critical for activity in both skeletal and cardiac myocytes, matches consensus motifs for a number of transcription factors, including the prodigious Sp/KLF family (Fig. 1B). Using a quantitative proteomic strategy, we were able to identify the relevant binding proteins, one of which was MAZ (Himeda et al., 2008). In subsequent studies, we confirmed the critical role of MAZ in the regulation of MCK and other muscle genes (Himeda et al., 2008). Interestingly, we found that MAZ recognizes a number of DNA sequences that diverge from its established binding motif (Fig. 1B), and these alternate sequences are abundant in skeletal and cardiac muscle genes (Himeda et al., 2008). This finding underscores an important corollary of transcription factor identification – that in order to understand the role(s) of a particular factor, it is critical to determine the full spectrum of sequences that it recognizes.

III.A.  Source Material Obtaining source material of sufficient purity and quantity is a significant challenge for many proteomic investigations. In attempting to purify novel cardiac transcription factors, two major obstacles are typically envisioned: (1) the need to purify cardiomyocytes away from other cell types; and (2) the need to purify the transcription factors that bind a specific DNA sequence away from other nuclear factors. Neither of these difficult challenges needs

Trex MCK Enh: GGACACCCGAGATGCCTGGTTA MEF3

m MCK Trex: GATA-2: TEF-1: m MCK Trex: MEF3 (rev): (A)

TEF-1 GATA-2

C G A G A T G C C T B V N G A T R G B B R C A T N C Y W C A C C C G A A A C C T G A

B V R Y W

= = = = =

795

C/G/T A/C/G A/G C/T A/T

MCK Prom:

MPEX GGGCCCCTCCCTGGGGACAGCC

Factor:

Known binding site:

Sp/KLF MAZ AP2 Olf-1

CCCTCCC (GA box) GGGGCGGGG (GC box) GGTGTGGGG (GT box) CCCTCCC (GA box) CCNNNGG CTCCN[T/A]GGGGA

(B)

Figure 1  Control elements in the MCK enhancer and promoter match multiple transcription factor binding motifs. (A) Sequence of the Trex site in the mouse MCK enhancer. Regions of the mouse Trex sequence with similarity to GATA-2 (Merika and Orkin, 1993), TEF-1 and MEF3 consensus motifs are shown, with differing bases indicated in red. Quantitative proteomics identified the Trex-binding factor in skeletal myocytes as Six4, which recognizes the MEF3 motif. (B) Sequence of the MPEX site in the mouse MCK promoter. Regions that match Sp/KLF, MAZ, AP2 and Olf-1 binding sites are indicated. Quantitative proteomics identified MAZ as one of the MPEX-binding factors in skeletal myocytes.

796

to be entirely overcome, since quantitative proteomic strategies allow identification of candidate factors whose abundance may be enriched by as little as two-fold over that of nonspecifically-bound factors. The functional relevance of these candidates is then investigated in subsequent studies. Currently, proteomic identification using mass spectrometry methods requires at least 1 pmol of the target factor. Considering that transcription factors represent only a small fraction of total cellular proteins and their enrichment entails significant loss, large amounts of source material are usually required. An additional technical challenge in obtaining transcription factors from striated muscle tissues is that intact nuclei are difficult to isolate directly from adult, or even newborn, muscle. Thus, cell cultures are a more amenable starting material, but these may raise additional problems in terms of cell purity and accurate representation of in vivo developmental and physiological states. We have successfully identified factors binding a control element in the MCK promoter (discussed above) using nuclear extracts from 1  109 cultured skeletal myocytes. Skeletal muscle cell lines such as MM14 and C2C12, which can be grown on gelatin-coated plates, or on gelatin-coated beads in suspension, are particularly useful for proteomic studies where large cell numbers are required. However, since no definitive cardiomyocyte cell lines exist, cardiac factor identification relies on the use of heart tissue or cultured cells which can be induced to differentiate into cardiomyocytes. The use of heart tissue as source material is complicated by the presence of other cell types, mainly fibroblasts, endothelial cells, and smooth muscle cells. Pre-plating of primary neonatal rat cardiomyocytes onto standard tissue culture dishes results in 90% pure cardiomyocyte preparations due to differential attachment of fibroblasts (Iwaki et al., 1990; Nguyen et al., 2003). However, the ventricles from each neonatal rat yield only 1.5  106 cardiomyocytes. Since current proteomic technologies for identifying a low-abundance transcription factor could require on the order of 1  109 cardiomyocytes, 700 neonatal hearts would be needed for such studies – an expensive and labor-intensive requirement. Alternative sources are cells such as the P19 embryonal carcinoma line, which can be induced to differentiate to cardiomyocytes under certain conditions (Angello et al., 2006; McBurney et al., 1982; Skerjanc et al., 1999). For instance, a subclonal P19 population has been isolated that preferentially activates early cardiac transcription factors within 48 hours of cell aggregration, and subsequently exhibits markers of cardiomyocyte terminal differentiation, including spontaneous contraction (Angello et al., 2006). Since only 2–5% of these cells differentiate into cardiomyocytes, it could be beneficial to identify and sort cardiac lineage cells to obtain semi-pure populations for subsequent analysis. However, as explained above, since quantitative proteomics requires only differential enrichment to identify candidate factors, strict cellular purity is not required as long as sufficient starting material is available for detection of at least 2-fold enrichment

PART  |  11  Cardiomics

over background noise. Similar strategies should be feasible with human ES or iPS cells that are induced to enter the cardiac lineage (LaFlamme et al., 2005; Zhang et al., 2009). As the pathways that control cardiac specification and differentiation are better understood, it should be feasible to acquire large numbers of relatively pure cardiomyocytes (or other cell types in the heart) for proteomic analysis. Since proteomic identifications rely on genomic databases and the in silico translation of open reading frames, an additional hurdle is encountered when the experimental system under investigation is not yet represented by a complete genomic sequence. For example, many functional assays for unknown mouse proteins were available prior to completion of the mouse genome project; e.g., we had proven the existence of a mouse MCK enhancer binding factor based on its association with a known control element in gel-shift assays (Fabre-Suver and Hauschka, 1996). However, we were unable to use proteomics to identify the factor due to the incomplete sequencing of the mouse genome. This problem was circumvented by demonstrating the existence of a HeLa cell nuclear factor with identical gel mobility and oligonucleotide sequence-specific binding properties. We were then able to use HeLa cell nuclear extracts in conjunction with a protein database formulated from the human genomic sequence to identify our candidate mouse MCK enhancer binding factor (Himeda et al., 2004). A series of functional studies were then used to verify the relevant activities of the corresponding mouse/rat skeletal and cardiac muscle transcription factors.

III.B.  Transcription Factor Enrichment Purifying a target factor to homogeneity is a challenging task, particularly when source material is difficult or expensive to obtain, and when significant amounts are lost during purification. The simplest and most successful enrichments have been achieved using specific DNA-affinity chromatography preceded (in some cases) by partial purification (Briggs et al., 1986; Jones et al., 1987; Lee et al., 1987; Masternak et al., 1998; Nordhoff et al., 1999; Merante et al., 2002; Schweppe et al., 2003; Yaneva and Tempst, 2006). Binding, washing, and elution of the target factor must be optimized for multiple conditions, including time, temperature, salt concentration, and amount of competitor DNA. Once target factor recovery is estimated (for example, by calculating densitometry of gel-shift bands), the required amount of source material can be determined. In our studies, we attempted to purify transcription factors using biotinylated oligonucleotides coupled to streptavidin-linked magnetic beads (Himeda et al., 2004, 2008). However, due to the large number of proteins co-purifying with our target factor, we decided to employ a selective enrichment strategy that would permit the exclusion of common contaminating proteins by a normalization step. In this

797

Chapter  |  11.4  Proteomic Strategies for Understanding Cardiac Function, Development, and Disease

strategy, equal numbers of beads were coupled to oligonucleotides containing either a wild-type or a mutant target site (Fig. 2A). These were incubated with equal volumes of the same nuclear extract, washed, and the bound proteins eluted at a predetermined salt concentration (Fig. 2A). This strategy resulted in two samples, one enriched for the target factor, and both containing equal amounts of nonspecific co-purifying proteins. The selectively-enriched factor(s) were then identified by quantitative proteomics (discussed below). If an unknown transcriptional co-factor or other interacting protein is the target, a slightly different strategy can be employed. In this case, an antibody specific to the known transcription factor bait can be cross-linked to protein Sepharose beads, and the beads incubated with nuclear extracts, washed, and bound proteins eluted (Fig. 2B). This strategy has been successfully employed in conjunction with mass spectrometry to identify novel factors associating with yeast TFIID (Sanders et al., 2002). If no antibody to the transcription factor bait exists, an antibody to a protein tag can be cross-linked to the beads, and nuclear extracts from cells overexpressing a tagged form

(A)

Nuclear extracts

+

wt

mt

of the transcription factor can be used (Fig. 2C). However, in this case, it is important to verify that incorporation of the tag does not interfere with any known protein–protein interactions. (For more information on purifying tagged protein complexes, see Chapter 11.5.) In either case, the selectively-enriched fraction would then be compared to the control fraction via quantitative proteomics.

III.C.  Transcription Factor Identification by Quantitative Proteomics Quantitative proteomics is a technology originally developed to quantify global changes in protein abundance, and it has proven equally useful in the identification of specific unknown factors. Major quantitative proteomic approaches include: (1) protein expression array analysis; (2) difference gel electrophoresis of proteins followed by mass spectrometry (MS) identification; and (3) stable isotopic labeling of proteins followed by liquid chromatography/ MS analysis.

(B)

(C)

1.

1.

1) Incubate nuclear extracts w/DNAcoupled mag. beads TF

2.

TF

CF

3.

TF

CF

2.

tag TF

CF

3.

tag TF

CF

2) Apply magnet 3) Remove unbound proteins 4) Wash beads TF

5) Elute bound proteins 4.

4.

Figure 2  Enrichment of target factors. (A) Selective enrichment of transcription factors. Equal amounts of nuclear extracts are incubated with beads coupled to either a wild-type transcription factor binding site (wt) or a mutant site that abolishes transcriptional activity in cells, and is unable to bind the target factor in gel-shift assays (mt). The target transcription factor is indicated in red and other nuclear proteins are indicated in blue. (B, C) Enrichment of co-factors. In (B), antibodies to the transcription factor bait are cross-linked to protein Sepharose beads (1). Nuclear extracts are incubated with the beads (2). In (C), antibodies to a protein tag are cross-linked to protein Sepharose beads (1). Nuclear extracts from cells overexpressing the transcription factor bait fused to the protein tag are incubated with the beads (2). In (B) and (C), beads are washed to remove nonspecifically attached proteins (3), and bound proteins are eluted (4). The transcription factor bait is indicated in red, the protein tag in green, target co-factor(s) in yellow, and other nuclear proteins in blue.

798

In protein expression array analysis, capture reagents are immobilized on a surface and incubated with protein samples, leading to specific binding of target factors which can then be detected using fluorophores, antibodies, or liquid chromatography/MS. This technology has been used primarily for expression profiling, but also for functional and protein interaction studies (Sydor and Nock, 2003; Sydor et al., 2003). In difference gel electrophoresis, protein samples for comparison are reacted with different CyDye fluors of equal mass and charge which covalently modify lysine residues (Patton, 2002). The same protein labeled with different CyDye fluors essentially comigrates to the same spot in a two-dimensional gel; thus, up to three protein samples (labeled with Cy2, Cy3, and Cy5) can be run on a single gel for subsequent spot matching, quantitation and MS identification. While this technology has proven useful, it suffers from the disadvantages inherent in 2D-E (as discussed earlier), and issues related to sensitivity of detection make it particularly inadequate for identifying low-abundance factors (Patton, 2002). Rapid advances have been made in the stable isotopic labeling of proteins for subsequent liquid chromatography/MS analysis. These advances include the development of isotope-coded affinity tag (ICAT) reagents, which specifically label cysteine residues, to identify and quantify proteins from complex mixtures (Gygi et al., 1999). Isotope-coded affinity tag reagents contain a thiol-specific, cysteine-reactive group for labeling, a biotin moiety for isolation, and an isotopic linker for quantitation (Fig. 3A). ICAT protocols have now become well-established, and have been applied in both specific factor identification and in global protein profiling (Tao and Aebersold, 2003). There are several major advantages of ICAT technology: 1. Differential incorporation of stable isotopes in two samples allows for determination of the relative abundance of proteins in the two samples. This permits the identification of proteins that are enriched in one sample, despite a high background of unenriched proteins. 2. The biotin moiety facilitates the chromatographic separation of cysteine-containing peptides from all other peptides, thereby decreasing the background during subsequent liquid chromatography/MS detection. 3. Selection of cysteine-containing peptides permits the peptide identification software program to focus exclusively on cysteine-containing tryptic peptides in the protein database. 4. Due to these advantages, and to the high resolution of microcapillary liquid chromatography/MS systems, extensive purification is not necessarily required to identify low-abundance proteins such as transcription factors.

PART  |  11  Cardiomics

Major caveats of this approach are: (1) target factors must contain at least one cysteine in order to be tagged and identified; and (2) at least 1 pmol of the target factor is required for identification. However, the majority of known proteins (for example, 92% of S. cerivisiae proteins) contain cysteine (Gygi et al., 1999). The quantitative component of the ICAT method depends on the isotopic linker moiety. Pairs of isotopically heavy- and light-tagged peptides are chemically identical; thus, they co-elute into the mass spectrometer, but display discrete mass differences. The first ICAT reagents to be developed contained 1H (hydrogen) in the light form of the tag, and 2H (deuterium) in the heavy form (Fig. 3B). This led to an 8 Da mass difference between identical peptides labeled with the heavy and light tags. Newer ICAT reagents incorporate 13C and 12C in the heavy and light forms, respectively (a 9 Da mass difference) (Fig. 3B). The advantages of the new tags are two-fold: (1) 13C- and 12Ctagged peptide pairs co-elute during HPLC separation, as opposed to 1H- and 2H-tagged peptides, in which the light peptides exhibit a slight delay in retention time. Use of the 13 C- and 12C-ICAT reagants consequently simplifies the assignment of peptide pairs (Yi et al., 2005). (2) Secondgeneration tags also contain an acid-cleavable bond to allow removal of the relatively heavy biotin moiety, which further improves peptide identification (Yi et al., 2005). In a typical experiment, proteins from two samples to be compared (in Fig. 3C, these are proteins recovered from wild-type versus mutant DNA-coupled beads) are denatured, reduced, and differentially labeled with the heavy and light forms of the ICAT reagent. The labeled protein samples are combined and digested with trypsin (which cleaves after arginine and lysine) and endoproteinase LysC (which cleaves after lysine). The use of both enzymes ensures more complete digestion than is obtained with trypsin alone. Sample complexity is then reduced by sequential chromatography. Peptides are initially fractionated by strong cation exchange chromatography, after which ICAT-labeled, cysteine-containing peptides are captured on an avidin affinity column. The tags are cleaved with strong acid, leaving the biotin moiety attached to the column, and the released peptides can be further resolved by liquid chromatography before MS analysis. Following microcapillary reversed-phase liquid chromatography that separates peptide pairs by hydrophobicity, the peptides are ionized and subjected to tandem MS. During this last step, both the sequence identity and relative abundance of the peptides are determined by automated multistage MS. Peptides are selected for collision-induced dissociation, resulting in a fragmentation spectrum. Fragment ions in the spectrum represent mainly single-event preferential cleavage of peptide bonds, resulting in sequence information from both the N and C termini of the peptide. Each collision-induced dissociation spectrum is used to search a protein database using an algorithm such as

Chapter  |  11.4  Proteomic Strategies for Understanding Cardiac Function, Development, and Disease

799

O NH

HN

O

O O N H

S Biotin

O

O

Heavy / light linker

N H Thiol-specific reactive group

(A) Biotin

Biotin

Acid-cleavable linker

8 X 1H/ 2H

thiol-specific reactive group

9 X 12C/13C

thiol-specific reactive group

(B) Proteins recovered from wt beads

Proteins recovered from mt beads

(Heavy ICAT)

(Light ICAT)

Enriched candidate C H

C

C

H

L

H

L

C H

H C H C C HH CC

C

H

L

C

C

Co-purifying protein

Combine

C

C C H H

L

L L

Trypsinize SCX chromatography

H C

Avidin affinity chromatography

C

H C

H C

C

H

L C

C L C

L C

HPLC- MS/MS LCQIFSDLNATYR

YVLDGMVDTVCEDLETDKK Candidate factor (3-fold enrichment in wt vs. mt sample)

• Determine peptide sequence and ID of parent protein • Calculate relative abundance of H:L-tagged peptide

Contaminating factor (no enrichment in wt vs. mt sample)

(C) Figure 3  Quantitative proteomics using ICAT reagents. (A) Structure of the ICAT reagent. The tag consists of a biotin moiety for isolation, an isotopic linker, and a thiol-specific reactive group. (B) Types of ICAT reagents. Schematic diagrams comparing the first-generation (1H/2H) ICAT reagent with the second-generation (12C/13C) ICAT reagent. Refer to text for details. (C) Identification of binding factors via quantitative proteomics. Proteins recovered from wt and mt beads are labeled with heavy and light forms of the ICAT reagent, respectively, then combined and trypsinized. Peptides are fractionated by strong cation exchange chromatography. Tagged, cysteine-containing peptides are isolated via avidin-affinity chromatography, then resolved by high-performance liquid chromatography, and analyzed via tandem mass spectrometry. In the example shown, the candidate factor is threefold enriched in the wt versus mt sample, whereas the non specific contaminating factor is not enriched. Refer to text for more details.

SEQUEST, which identifies the cysteine-containing tryptic peptide in the database that is the best match to the collision-induced dissociation spectrum. Single-ion chromatograms are reconstructed for the isotopically heavy and light peptide pair using software such as XPRESS, and the relative abundance is calculated after summing the signal intensities for each peptide over their respective

elution times. (For a review of MS instrumentation, see Chapter 11.5.) As an indication of its versatility, ICAT-based quantitative proteomics has been used to identify a new component of the RNA polymerase II preinitiation complex in yeast (Ranish et al., 2004), to identify partner switching of transcription factor MafK on erythroid differentiation

800

(Brand et al., 2004), to analyze global changes in chromatin-associated proteins (Shiio et al., 2003), and we have used it to identify a single transcription factor binding the Trex sequence in the MCK enhancer (from a background of ~900 co-enriched proteins) (Himeda et al., 2004), as well as several different transcription factors binding the MPEX sequence in the MCK promoter (Himeda et al., 2008). Alternative but related strategies for isotopic labeling have been developed. Mass-coded abundance tagging (MCAT) is similar to isotope-coded affinity tagging, but relies on the modification of lysine residues within tryptic peptides (Cagney and Emili, 2002). In solid-phase isotope tagging, glass beads are attached to an isotope tag with a cysteine-reactive group and a photocleavable linker (Zhou et al., 2004). In a proof-of-principle study, Zhou et al. found that the number of yeast proteins identified and quantified by this method was around two- to three-fold higher than that obtained using the ICAT method (Zhou et al., 2004). Tandem mass tagging and iTRAQ methods utilize a multiplexed set of isobaric reagents to chemically tag peptides, allowing quantification of peptides in multiple samples (Thompson et al., 2003; Ross et al., 2004; Aggarwal et al., 2006). In vivo labeling of proteins in cell culture with isotopically-labeled amino acids has also been used successfully for quantitative proteomic analysis (Martinovic et al., 2002; Ong et al., 2002; Ibarrola et al., 2003). An important limitation of these techniques is that they are best suited for the analysis of protein abundance, which does not necessarily correlate with function. To assess protein function more directly, libraries of activity-based proteomic probes can be used. This strategy was successfully employed to identify differences in liver enzyme activities in lean versus obese mice (Barglow and Cravatt, 2004). Such probes could be extremely useful for the functional profiling of enzymes during cardiac disease states. Multiplexed proteomics technology allows comparisons of protein expression levels in parallel with functional attributes of proteins, such as post-translational modifications or drug-binding and/or metabolizing capability (Patton, 2002). The use of proteomics as a tool to analyze post-translational modifications is addressed in Chapter 11.5, whose authors have used multiple-reactionmonitoring to identify sites of phosphorylation in MEF2A and -casein (Cox et al., 2005).

III.D.  Confirmation of Candidates It is important to emphasize that when quantitative proteomics is used as a means of identifying unknown proteins associated with a specific function, the identified proteins are only candidates for the factors serving that particular function. Subsequent studies are thus required to verify the functional relevance of a proteomic candidate. If a noncardiac cell type has been used as the source material, or is

PART  |  11  Cardiomics

present in a heterogeneous mixture of cells, it is essential to verify the relevance of the candidate(s) in cardiomyocytes. If antibodies for candidate transcription factors exist, they can be used in immunohistology studies to assess the presence of candidate factors in cardiomyocytes or cardiac lineage precursors. Additionally, antibodies can be used in gel supershift studies to confirm in vitro binding of the candidate to the target control element, and in ChIP studies to demonstrate in vivo binding to the target regulatory region. Likewise, candidate co-factor interactions can be validated via co-immunoprecipitation studies. Additional evidence for a candidate factor’s involvement can be obtained by overexpressing or repressing its expression via transgenic or cell culture transfection strategies, followed by assessment of effects on the candidate’s presumed function. However, if the endogenous factor is already saturating the level of target gene expression, overexpression may not result in increased gene expression. Additionally, if the factor belongs to a family with a high degree of functional overlap between members, depletion may result in compensation by another family member. If the candidate is a novel, uncharacterized factor, confirmation paves the way for a number of exciting functional and descriptive studies. However, even if the factor has been previously characterized, it is still likely that novel aspects of its function may be uncovered. For example, we found that the Trex-binding factor in the MCK enhancer corresponds to members of the Six/sine oculis family, which recognize MEF3 elements in the regulatory regions of their target genes. Since the MCK Trex sequence does not precisely conform to the established MEF3 consensus, this study allowed us to modify and expand the consensus binding motif for Six proteins, thereby facilitating the identification of previously unrecognized MEF3 elements in other muscle genes (Himeda et al., 2004). Additionally, once Six4 was identified as the relevant Trex/MEF3 binding factor in skeletal muscle cells, this led to our identification of Six5 as the relevant binding factor in cardiomyocytes (Himeda et al., 2004). Similarly, on identifying MAZ as one of the factors binding the MPEX site in the MCK promoter, we found that MAZ is enriched at the promoters of other skeletal and cardiac muscle genes, and that alternate MAZ motifs are abundant in these regions (Himeda et al., 2008). In this study, identification by proteomics served as the first step in uncovering a previously unknown role for a transcription factor in striated muscle gene regulation.

IV.  Future prospects The identification of transcription factors by proteomics will be greatly aided by improved technology for isolating pure cell populations from tissues. Using laser capture microdissection, De Souza et al. were able to isolate

Chapter  |  11.4  Proteomic Strategies for Understanding Cardiac Function, Development, and Disease

separate populations of cardiomyocytes and blood vessels from human heart samples for analysis by 2D-E (De Souza et al., 2004). Although this technique is laborious and best suited for the analysis of abundant cardiac proteins, the development of more high-throughput technology for specific cell isolation may allow transcription factor identification from in vivo cardiac tissue instead of cultured cells. Numerous studies have demonstrated that some transcription factor–DNA interactions are highly cooperative, or dependent on interactions with factors bound to neighboring or even distal elements (Raychaudhuri et al., 1990; Lee et al., 1991; Wang et al., 1994; Ribeiro et al., 1999; Kitayner et al., 2006). Thus, the identification of multifactor complexes binding to complete enhancers or promoters, rather than factors binding single control elements, would greatly benefit our understanding of gene regulation. Although it will be more challenging to delineate factor binding to large regions of DNA, and to control for nonspecific protein–DNA and protein–protein binding, quantitative proteomics could be used for such studies. Another disadvantage of current strategies for transcription factor identification is that the binding conditions are artificial; i.e., nuclear extracts are added to naked DNA under nonphysiological ionic conditions. Thus, it would be advantageous to identify factors and multiprotein complexes binding to a given promoter in the more native context of chromatin, a situation which is difficult to duplicate in vitro. Ito et al. have generated nucleosomal arrays in vitro using the chromatin assembly factor ACF and the core histone chaperone NAP-1 (Ito et al., 1999), and these arrays have been used successfully to establish an in vitro transcription system (Dilworth et al., 2004). In principle, one could assemble chromatinized enhancer/promoter templates on a solid-phase support in order to identify factors important for initiating cardiac gene expression (in the context of intact nucleosomes). Alternatively, with the addition of a chromatin remodeling factor, it may be possible to ask questions regarding the different regulatory assemblies involved in different physiological, developmental, or disease states. One might be able to identify factors binding to larger regulatory regions such as enhancers in vivo by transducing cardiomyocytes with a lentivirus carrying a target enhancer/promoter ligated to multiple unique binding sites for an artificial zinc-finger protein, as well as the zincfinger protein driven by a cardiac promoter. After selection, the cells could be subjected to standard ChIP analysis, using an antibody to the artificial protein to pull down the chromatin fragment containing the target regulatory region. After reversing cross-links and digesting DNA, the immunoprecipitated proteins could then be identified via proteomics. Such a system would undoubtedly be difficult to perfect, but it would allow identification of regulatory factors in the nearly native context of a complete regulatory region integrated into normal chromatin.

801

In conjunction with more sophisticated systems for purifying/assembling protein complexes, advances in proteomic technology will undoubtedly lead to increased sensitivity and coverage, allowing the identification of greater numbers of low-abundance proteins from increasingly complex mixtures. As the studies mentioned here indicate, proteomics is a powerful and still underutilized tool for understanding cardiac biology.

References Aggarwal, K., Choe, L.H., Lee, K.H., 2006. Shotgun proteomics using the iTRAQ isobaric tags. Brief Funct. Genomic Proteomic 5, 112–120. Angello, J.C., Kaestner, S., Welikson, R.E., Buskin, J.N., Hauschka, S.D., 2006. BMP induction of cardiogenesis in P19 cells requires prior cell–cell interaction(s). Dev. Dyn. 235, 2122–2133. Barglow, K.T., Cravatt, B.F., 2004. Discovering disease-associated enzymes by proteome reactivity profiling. Chem. Biol. 11, 1523–1531. Borozdenkova, S., Westbrook, J.A., Patel, V., Wait, R., Bolad, I., Burke, M.M., Bell, A.D., Banner, N.R., Dunn, M.J., Rose, M.L., 2004. Use of proteomics to discover novel markers of cardiac allograft rejection. J. Proteome Res. 3, 282–288. Brand, M., Ranish, J.A., Kummer, N.T., Hamilton, J., Igarashi, K., Francastel, C., Chi, T.H., Crabtree, G.R., Aebersold, R., Groudine, M., 2004. Dynamic changes in transcription factor complexes during erythroid differentiation revealed by quantitative proteomics. Nat. Struct. Mol. Biol. 11, 73–80. Briggs, M.R., Kadonaga, J.T., Bell, S.P., Tjian, R., 1986. Purification and biochemical characterization of the promoter-specific transcription factor. Sp1. Science 234, 47–52. Buerke, M., Schwertz, H., Längin, T., Buerke, U., Prondzinsky, R., Platsch, H., Richert, J., Bomm, S., Schmidt, M., Hillen, H., Lindemann, S., Blaschke, G., Müller-Werdan, U., Werdan, K., 2006. Proteome analysis of myocardial tissue following ischemia and reperfusion – effects of complement inhibition. Biochim. Biophys. Acta 1764, 1536–1545. Cagney, G., Emili, A., 2002. De novo peptide sequencing and quantitative profiling of complex protein mixtures using mass-coded abundance tagging. Nat. Biotechnol. 20, 163–170. Corbett, J.M., Why, H.J., Wheeler, C.H., Richardson, P.J., Archard, L.C., Yacoub, M.H., Dunn, M.J., 1998. Cardiac protein abnormalities in dilated cardiomyopathy detected by two-dimensional polyacrylamide gel electrophoresis. Electrophoresis 19, 2031–2042. Cox, D.M., Zhong, F., Du, M., Duchoslav, E., Sakuma, T., McDermott, J. C., 2005. Multiple reaction monitoring as a method for identifying protein posttranslational modifications. J. Biomol. Tech. 16, 83–90. De Souza, A.I., McGregor, E., Dunn, M.J., Rose, M.L., 2004. Preparation of human heart for laser microdissection and proteomics. Proteomics 4, 578–586. Dilworth, F.J., Seaver, K.J., Fishburn, A.L., Htet, S.L., Tapscott, S.J., 2004. In vitro transcription system delineates the distinct roles of the coactivators pCAF and p300 during MyoD/E47-dependent transactivation. Proc. Natl. Acad. Sci. 101, 11593–11598. Duan, X., Berthiaume, F., Yarmush, D., Yarmush, M.L., 2006. Proteomic analysis of altered protein expression in skeletal muscle of rats in a hypermetabolic state induced by burn sepsis. Biochem. J. 397, 149–158.

802

Ellinghaus, P., Scheubel, R.J., Dobrev, D., Ravens, U., Holtz, J., Huetter, J., Nielsch, U., Morawietz, H., 2005. Comparing the global mRNA expression profile of human atrial and ventricular myocardium with high-density oligonucleotide arrays. J. Thorac. Cardiovasc. Surg. 129, 1383–1390. Faber, M.J., Dalinghaus, M., Lankhuizen, I.M., Bezstarosti, K., Dekkers, D.H., Duncker, D.J., Helbing, W.A., Lamers, J.M., 2005. Proteomic changes in the pressure overloaded right ventricle after 6 weeks in young rats: correlations with the degree of hypertrophy. Proteomics 5, 2519–2530. Fabre-Suver, C., Hauschka, S.D., 1996. A novel site in the muscle creatine kinase enhancer is required for expression in skeletal but not cardiac muscle. J. Biol. Chem. 271, 4646–4652. Gelfi, C., De Palma, S., Ripamonti, M., Eberini, I., Wait, R., Bajracharya, A., Marconi, C., Schneider, A., Hoppeler, H., Cerretelli, P., 2004. New aspects of altitude adaptation in Tibetans: a proteomic approach. FASEB J. 18, 612–614. Gygi, S.P., Rist, B., Gerber, S.A., Turecek, F., Gelb, M.H., Aebersold, R., 1999. Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nat. Biotechnol. 17, 994–999. Heinke, M.Y., Wheeler, C.H., Chang, D., Einstein, R., Drake-Holland, A., Dunn, M.J., dos Remedios, C.G., 1998. Protein changes observed in pacing-induced heart failure using two-dimensional electrophoresis. Electrophoresis 19, 2021–2030. Heinke, M.Y., Wheeler, C.H., Yan, J.X., Amin, V., Chang, D., Einstein, R., Dunn, M.J., dos Remedios, C.G., 1999. Changes in myocardial protein expression in pacing-induced canine heart failure. Electrophoresis 20, 2086–2093. Himeda, C.L., Ranish, J.A., Angello, J.C., Maire, P., Aebersold, R., Hauschka, S.D., 2004. Quantitative proteomic identification of six4 as the trex-binding factor in the muscle creatine kinase enhancer. Mol. Cell Biol. 24, 2132–2143. Himeda, C.L., Ranish, J.A., Hauschka, S.D., 2008. Quantitative proteomic identification of MAZ as a transcriptional regulator of muscle-specific genes in skeletal and cardiac myocytes. Mol. Cell Biol. 28, 6521–6535. Ibarrola, N., Kalume, D.E., Gronborg, M., Iwahori, A., Pandey, A., 2003. A proteomic approach for quantitation of phosphorylation using stable isotope labeling in cell culture. Anal. Chem. 75, 6043–6049. Ito, T., Levenstein, M.E., Fyodorov, D.V., Kutach, A.K., Kobayashi, R., Kadonaga, J.T., 1999. ACF consists of two subunits, Acf1 and ISWI, that function cooperatively in the ATP-dependent catalysis of chromatin assembly. Genes Dev. 13, 1529–1539. Iwaki, K., Sukhatme, V.P., Shubeita, H.E., Chien, K.R., 1990. Alpha- and beta-adrenergic stimulation induces distinct patterns of immediate early gene expression in neonatal rat myocardial cells: fos/jun expression is associated with sarcomere assembly; Egr-1 induction is primarily an alpha 1-mediated response. J. Biol. Chem. 265, 13809–13817. Jeon, H.B., Choi, E.S., Yoon, J.H., Hwang, J.H., Chang, J.W., Lee, E.K., Choi, H.W., Park, Z.Y., Yoo, Y.J., 2007. A proteomics approach to identify the ubiquitinated proteins in mouse heart. Biochem. Biophys. Res. Commun. 357, 731–736. Jones, K.A., Kadonaga, J.T., Rosenfeld, P.J., Kelly, T.J., Tjian, R., 1987. A cellular DNA-binding protein that activates eukaryotic transcription and DNA replication. Cell 48, 79–89. Kim, N., Lee, Y., Kim, H., Joo, H., Youm, J.B., Park, W.S., Warda, M., Cuong, D.V., Han, J., 2006. Potential biomarkers for ischemic heart damage identified in mitochondrial proteins by comparative proteo­mics. Proteomics 6, 1237–1249.

PART  |  11  Cardiomics

Kislinger, T., Gramolini, A.O., MacLennan, D.H., Emili, A., 2005. Multidimensional protein identification technology (MudPIT): technical overview of a profiling method optimized for the comprehensive proteomic investigation of normal and diseased heart tissue. J. Am. Soc. Mass. Spectrom. 16, 1207–1220. Kitayner, M., Rozenberg, H., Kessler, N., Rabinovich, D., Shaulov, L., Haran, T.E., Shakked, Z., 2006. Structural basis of DNA recognition by p53 tetramers. Mol. Cell. 22, 741–753. Laflamme, M.A., Gold, J., Xu, C., Hassanipour, M., Rosler, E., Police, S., Muskheli, V., Murry, C.E., 2005. Formation of human myocardium in the rat heart from human embryonic stem cells. Am. J. Pathol. 167, 663–671. Lai, L.P., Lin, J.L., Lin, C.S., Yeh, H.M., Tsay, Y.G., Lee, C.F., Lee, H. H., Chang, Z.F., Hwang, J.J., Su, M.J., Tseng, Y.Z., Huang, S.K., 2004. Functional genomic study on atrial fibrillation using cDNA microarray and two-dimensional protein electrophoresis techniques and identification of the myosin regulatory light chain isoform reprogramming in atrial fibrillation. J. Cardiovasc. Electrophysiol. 15, 214–223. Latif, N., Baker, C.S., Dunn, M.J., Rose, M.L., Brady, P., Yacoub, M.H., 1993. Frequency and specificity of antiheart antibodies in patients with dilated cardiomyopathy detected using SDS-PAGE and western blotting. J. Am. Coll. Cardiol. 22, 1378–1384. Latif, N., Rose, M.L., Yacoub, M.H., Dunn, M.J., 1995. Association of pretransplantation antiheart antibodies with clinical course after heart transplantation. J. Heart. Lung Transplant. 14, 119–126. Lee, T.C., Chow, K.L., Fang, P., Schwartz, R.J., 1991. Activation of skeletal alpha-actin gene transcription: the cooperative formation of serum response factor-binding complexes over positive cis-acting promoter serum response elements displaces a negative-acting nuclear factor enriched in replicating myoblasts and nonmyogenic cells. Mol. Cell Biol. 11, 5090–5100. Lee, W., Mitchell, P., Tjian, R., 1987. Purified transcription factor AP-1 interacts with TPA-inducible enhancer elements. Cell 49, 741–752. Martinovic, S., Veenstra, T.D., Anderson, G.A., Pasa-Tolic, L., Smith, R.D., 2002. Selective incorporation of isotopically labeled amino acids for identification of intact proteins on a proteome-wide level. J. Mass. Spectrom. 37, 99–107. Masternak, K., Barras, E., Zufferey, M., Conrad, B., Corthals, G., Aebersold, R., Sanchez, J.C., Hochstrasser, D.F., Mach, B., Reith, W., 1998. A gene encoding a novel RFX-associated transactivator is mutated in the majority of MHC class II deficiency patients. Nat. Genet. 20, 273–277. McBurney, M.W., Jones-Villeneuve, E.M., Edwards, M.K., Anderson, P. J., 1982. Control of muscle and neuronal differentiation in a cultured embryonal carcinoma cell line. Nature 299, 165–167. Merante, F., Altamentova, S.M., Mickle, D.A., Weisel, R.D., Thatcher, B. J., Martin, B.M., Marshall, J.G., Tumiati, L.C., Cowan, D.B., Li, R. K., 2002. The characterization and purification of a human transcription factor modulating the glutathione peroxidase gene in response to oxygen tension. Mol. Cell Biochem. 229, 73–83. Merika, M., Orkin, S.H., 1993. DNA-binding specificity of GATA family transcription factors. Mol. Cell Biol. 13, 3999–4010. Nguyen, Q.G., Buskin, J.N., Himeda, C.L., Fabre-Suver, C., Hauschka, S.D., 2003. Transgenic and tissue culture analyses of the muscle creatine kinase enhancer Trex control element in skeletal and cardiac muscle indicate differences in gene expression between muscle types. Transgenic Res. 12, 337–349. Nordhoff, E., Krogsdam, A.M., Jorgensen, H.F., Kallipolitis, B.H., Clark, B.F., Roepstorff, P., Kristiansen, K., 1999. Rapid identification of

Chapter  |  11.4  Proteomic Strategies for Understanding Cardiac Function, Development, and Disease

DNA-binding proteins by mass spectrometry. Nat. Biotechnol. 17, 884–888. Okumura, N., Hashida-Okumura, A., Kita, K., Matsubae, M., Matsubara, T., Takao, T., Nagai, K., 2005. Proteomic analysis of slow- and fasttwitch skeletal muscles. Proteomics 5, 2896–2906. Ong, S.E., Blagoev, B., Kratchmarova, I., Kristensen, D.B., Steen, H., Pandey, A., Mann, M., 2002. Stable isotope labeling by amino acids in cell culture, SILAC, as a simple and accurate approach to expression proteomics. Mol. Cell Proteomics 1, 376–386. Pankuweit, S., Portig, I., Lottspeich, F., Maisch, B., 1997. Autoantibodies in sera of patients with myocarditis: characterization of the corresponding proteins by isoelectric focusing and N-terminal sequence analysis. J. Mol. Cell Cardiol. 29, 77–84. Patel, V.B., Corbett, J.M., Dunn, M.J., Winrow, V.R., Portmann, B., Richardson, P.J., Preedy, V.R., 1997. Protein profiling in cardiac tissue in response to the chronic effects of alcohol. Electrophoresis 18, 2788–2794. Patel, V.B., Sandhu, G., Corbett, J.M., Dunn, M.J., Rodrigues, L.M., Griffiths, J.R., Wassif, W., Sherwood, R.A., Richardson, P.J., Preedy, V.R., 2000. A comparative investigation into the effect of chronic alcohol feeding on the myocardium of normotensive and hypertensive rats: an electrophoretic and biochemical study. Electrophoresis 21, 2454–2462. Patton, W.F., 2002. Detection technologies in proteome analysis. J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci. 771, 3–31. Pleissner, K.P., Regitz-Zagrosek, V., Weise, C., Neuss, M., Krudewagen, B., Soding, P., Buchner, K., Hucho, F., Hildebrandt, A., Fleck, E., 1995. Chamber-specific expression of human myocardial proteins detected by two-dimensional gel electrophoresis. Electrophoresis 16, 841–850. Pleissner, K.P., Soding, P., Sander, S., Oswald, H., Neuss, M., RegitzZagrosek, V., Fleck, E., 1997. Dilated cardiomyopathy-associated proteins and their presentation in a WWW-accessible two-dimensional gel protein database. Electrophoresis 18, 802–808. Pohlner, K., Portig, I., Pankuweit, S., Lottspeich, F., Maisch, B., 1997. Identification of mitochondrial antigens recognized by antibodies in sera of patients with idiopathic dilated cardiomyopathy by twodimensional gel electrophoresis and protein sequencing. Am. J. Cardiol. 80, 1040–1045. Ranish, J.A., Hahn, S., Lu, Y., Yi, E.C., Li, X.J., Eng, J., Aebersold, R., 2004. Identification of TFB5, a new component of general transcription and DNA repair factor IIH. Nat. Genet. 36, 707–713. Raychaudhuri, P., Bagchi, S., Neill, S.D., Nevins, J.R., 1990. Activation of the E2F transcription factor in adenovirus-infected cells involves E1A-dependent stimulation of DNA-binding activity and induction of cooperative binding mediated by an E4 gene product. J. Virol. 64, 2702–2710. Ribeiro, A., Pastier, D., Kardassis, D., Chambaz, J., Cardot, P., 1999. Cooperative binding of upstream stimulatory factor and hepatic nuclear factor 4 drives the transcription of the human apolipoprotein A-II gene. J. Biol. Chem. 274, 1216–1225. Ross, P.L., Huang, Y.N., Marchese, J.N., Williamson, B., Parker, K., Hattan, S., Khainovski, N., Pillai, S., Dey, S., Daniels, S., Purkayastha, S., Juhasz, P., Martin, S., Bartlet-Jones, M., He, F., Jacobson, A., Pappin, D.J., 2004. Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents. Mol. Cell Proteomics 3, 1154–1169. Sanders, S.L., Jennings, J., Canutescu, A., Link, A.J., Weil, P.A., 2002. Proteomics of the eukaryotic transcription machinery: identification of proteins associated with components of yeast TFIID by multidimensional mass spectrometry. Mol. Cell Biol. 22, 4723–4738.

803

Schweppe, R.E., Melton, A.A., Brodsky, K.S., Aveline, L.D., Resing, K. A., Ahn, N.G., Gutierrez-Hartmann, A., 2003. Purification and mass spectrometric identification of GA-binding protein (GABP) as the functional pituitary Ets factor binding to the basal transcription element of the prolactin promoter. J. Biol. Chem. 278, 16863–16872. Shiio, Y., Eisenman, R.N., Yi, E.C., Donohoe, S., Goodlett, D.R., Aebersold, R., 2003. Quantitative proteomic analysis of chromatinassociated factors. J. Am. Soc. Mass. Spectrom. 14, 696–703. Skerjanc, I.S., 1999. Cardiac and skeletal muscle development in P19 embryonal carcinoma cells. Trends Cardiovasc. Med. 9, 139–143. Sydor, J.R., Nock, S., 2003. Protein expression profiling arrays: tools for the multiplexed high-throughput analysis of proteins. Proteome Sci. 1, 3. Sydor, J.R., Scalf, M., Sideris, S., Mao, G.D., Pandey, Y., Tan, M., Mariano, M., Moran, M.F., Nock, S., Wagner, P., 2003. Chip-based analysis of protein–protein interactions by fluorescence detection and on-chip immunoprecipitation combined with microLC-MS/MS analysis. Anal. Chem. 75, 6163–6170. Tabibiazar, R., Wagner, R.A., Liao, A., Quertermous, T., 2003. Transcriptional profiling of the heart reveals chamber-specific gene expression patterns. Circ. Res. 93, 1193–1201. Tao, W.A., Aebersold, R., 2003. Advances in quantitative proteomics via stable isotope tagging and mass spectrometry. Curr. Opin. Biotechnol. 14, 110–118. Thiede, B., Otto, A., Zimny-Arndt, U., Muller, E.C., Jungblut, P., 1996. Identification of human myocardial proteins separated by twodimensional electrophoresis with matrix-assisted laser desorption/ ionization mass spectrometry. Electrophoresis 17, 588–599. Thompson, A., Schaefer, J., Kuhn, K., Kienle, S., Schwartz, J., Schmidt, G., Neumann, T., Johnstone, R.A., Mohammed, A.K., Hamon, C., 2006. Tandem mass tags: A novel quantification strategy for comparative analysis of complex protein mixtures by MS/MS. Anal. Chem. 78, 4235. Toraason, M., Moorman, W., Mathias, P.I., Fultz, C., Witzmann, F., 1997. Two-dimensional electrophoretic analysis of myocardial proteins from lead-exposed rabbits. Electrophoresis 18, 2978–2982. Wang, C.Y., Bassuk, A.G., Boise, L.H., Thompson, C.B., Bravo, R., Leiden, J.M., 1994. Activation of the granulocyte-macrophage colony-stimulating factor promoter in T cells requires cooperative binding of Elf-1 and AP-1 transcription factors. Mol. Cell Biol. 14, 1153–1159. Warda, M., Kim, H.K., Kim, N., Youm, J.B., Kang, S.H., Park, W.S., Khoa, T.M., Kim, Y.H., Han, J., 2007. Simulated hyperglycemia in rat cardiomyocytes: A proteomics approach for improved analysis of cellular alterations. Proteomics 7, 2570–2590. Washburn, M.P., Wolters, D., Yates, J.R., 3rd, 2001. Large-scale analysis of the yeast proteome by multidimensional protein identification technology. Nat. Biotechnol. 19, 242–247. Weekes, J., Wheeler, C.H., Yan, J.X., Weil, J., Eschenhagen, T., Scholtysik, G., Dunn, M.J., 1999. Bovine dilated cardiomyopathy: proteomic analysis of an animal model of human dilated cardiomyopathy. Electrophoresis 20, 898–906. Weekes, J., Morrison, K., Mullen, A., Wait, R., Barton, P., Dunn, M.J., 2003. Hyperubiquitination of proteins in dilated cardiomyopathy. Proteomics 3, 208–216. Wheeler, C.H., Collins, A., Dunn, M.J., Crisp, S.J., Yacoub, M.H., Rose, M.L., 1995. Characterization of endothelial antigens associated with transplant-associated coronary artery disease. J. Heart Lung Transplant. 14, S188–S197. White, M.Y., Cordwell, S.J., McCarron, H.C., Prasan, A.M., Craft, G., Hambly, B.D., Jeremy, R.W., 2005. Proteomics of

804

ischemia/reperfusion injury in rabbit myocardium reveals alterations to proteins of essential functional systems. Proteomics 5, 1395–1410. White, M.Y., Tchen, A.S., McCarron, H.C., Hambly, B.D., Jeremy, R.W., Cordwell, S.J., 2006. Proteomics of ischemia and reperfusion injuries in rabbit myocardium with and without intervention by an oxygenfree radical scavenger. Proteomics 6, 6221–6233. Yan, L., Ge, H., Li, H., Lieber, S.C., Natividad, F., Resuello, R.R., Kim, S.J., Akeju, S., Sun, A., Loo, K., Peppas, A.P., Rossi, F., Lewandowski, E. D., Thomas, A.P., Vatner, S.F., Vatner, D.E., 2004. Gender-specific proteomic alterations in glycolytic and mitochondrial pathways in aging monkey hearts. J. Mol. Cell Cardiol. 37, 921–929. Yaneva, M., Tempst, P., 2006. Isolation and mass spectrometry of specific DNA binding proteins. Methods Mol. Biol. 338, 291–303. Yi, E.C., Li, X.J., Cooke, K., Lee, H., Raught, B., Page, A., Aneliunas, V., Hieter, P., Goodlett, D.R., Aebersold, R., 2005. Increased quantitative

PART  |  11  Cardiomics

proteome coverage with (13)C/(12)C-based, acid-cleavable isotope-coded affinity tag reagent and modified data acquisition scheme. Proteomics 5, 380–387. You, S.A., Archacki, S.R., Angheloiu, G., Moravec, C.S., Rao, S., Kinter, M., Topol, E.J., Wang, Q., 2003. Proteomic approach to coronary atherosclerosis shows ferritin light chain as a significant marker: evidence consistent with iron hypothesis in atherosclerosis. Physiol. Genomics 13, 25–30. Zhang, J., Wilson, G.F., Soerens, A.G., Koonce, C.H., Yu, J., Palecek, S.P., Thomson, J.A., Kamp, T.J., 2009. Functional cardiomyocytes derived from human induced pluripotent stem cells. Circ. Res. 104, e30–e41. Zhou, H., Boyle, R., Aebersold, R., 2004. Quantitative protein analysis by solid phase isotope tagging and mass spectrometry. Methods Mol. Biol. 261, 511–518.

Chapter 11.5

Proteomic Analysis of MEF2 PostTranslational Regulation in the Heart David M. Cox,2 Min Du1 and John C. McDermott1 1

Department of Biology, York University, Toronto, Ontario, Canada MDS Sciex, Concord, Ontario, Canada

2

I.  Introduction The network of transcriptional regulatory proteins in a cardiac myocyte is a fundamental determinant of its gene expression profile (Chien et al., 1993; Doevendans and van Bilsen, 1996; Olson and Srivastava, 1996; Chang et al., 2006; Srivastava, 2006), ensuingly dictating the form and function of the cell. Regulation of transcription factors has therefore been a predominant theme in understanding the molecular control of ontogeny, physiology and pathology of the myocardium. A number of transcriptional regulators have been identified as playing important roles in the heart, having proven critical for cardiac gene expression during development, and also in the postnatal vertebrate heart in contexts such as in maladaptive cardiac hypertrophy (see also Chapter 11.4 for a review of proteomic strategies to identify novel transcriptional regulators). The elucidation of transcriptional regulators involved in the control of gene expression in the heart has indeed been a major step forward in understanding the control of cardiac gene expression; however, the post-translational control of the network of transcriptional regulators has proved more elusive. Substantive evidence now exists indicating that the expression of a transcriptional regulatory protein, in itself, is not necessarily a profound indicator that it is active in promoting or repressing transcription of its target genes. Protein–protein interactions and post-translational mechanisms serve to modulate activity dynamically. This allows various permutations of cellular signaling to exert control over the regulators, and the cellular context to contribute to regulation by providing a tissue-specific complement of interacting proteins (Pawson and Nash, 2000; Scott and Pawson, 2000; Holmberg et al., 2002; Pawson and Scott, Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

2005). Thus, transcription factor activity and subsequently the orchestration of gene expression provide a level of epigenetic control that is tailored to the cellular context and signals emanating from the extracellular milieu. Extensive work concerning the control of cardiac specific gene expression by myocyte enhancer factor 2 (MEF2) transcriptional regulatory proteins (Molkentin and Markham, 1993; Kuisk et al., 1996; Di Lisi et al., 1998; Anderson et al., 2004) and loss-of-function analysis in gene-targeted mice (Lin et al., 1997; Naya et al., 2002), has positioned these factors at a key nexus in muscle control circuitry. The MEF2 family of transcription factors (encoded by four genes labeled as MEF2A–D) are initially expressed in the early mesodermal progenitor populations that give rise to the cardiac crescent at 7.5 days post-coitum (dpc) in the mouse (Edmondson et al., 1994; Anderson et al., 2004; Dodou et al., 2004) (see Chapter 9.5 for an extensive review).They are crucial in regulating cardiac (Edmondson et al., 1994; Lin et al., 1997), skeletal (Yun and Wold, 1996; Ornatsky et al., 1997; Naya and Olson, 1999) and smooth muscle differentiation (Lin et al., 1998; Anderson et al., 2004; Creemers et al., 2006a), neuronal survival (Mao et al., 1999) and T-cell activation (Youn et al., 1999). The MEF2s belong to the MADS (MCM1, Agamous, Deficiens and serum response factor) superfamily of DNA-binding proteins. The amino terminus of MEF2 proteins is conserved among all family members, and consists of a 57-amino acid MADS domain (Shore and Sharrocks, 1995) anda 29-amino acid MEF2 domain that together mediate MEF2 protein dimerization, co-factor interaction and binding to a cognate cis element with the consensus (T/C)TA(A/T)4TA(G/A) (Molkentin et al., 1996a). The MEF2 transactivation domain (TAD) is 805

806

PART  |  11  Cardiomics

located in the carboxy terminus, and the sequence in this region is more divergent between MEF2 family members (Black and Olson, 1998). The transactivation domain is subjected to extensive alternative splicing and posttranslational modification by phosphorylation, acetylation and sumoylation (Black and Olson, 1998; Gregoire and Yang, 2005; Ma et al., 2005). MEF2s have already been shown to be highly-responsive to several signal transduction cascades, and their post-translational regulation by covalent modification by PKC (Ornatsky et al., 1999), p38 MAPK (Han et al., 1997; Ornatsky et al., 1999; Cox et al., 2003), ERK5 (Kato et al., 1997; Yang et al., 1998), CDK5 (Gong et al., 2003) and CK2 (Molkentin et al., 1996b) has been well-documented. In addition, a further level of control of the MEF2s involves protein–protein interactions. MEF2 has previously been shown to physically interact with MyoD (Molkentin et al., 1995), Class II HDACs (histone deacetylases) (Miska et al., 1999; Sparrow et al., 1999; Lemercier et al., 2000; Lu et al., 2000), GATA4 (Morin et al., 2000), p300 (Sartorelli et al., 1997), Sp1 (Grayson et al., 1998; Park et al., 2002), Smad2 (Quinn et al., 2001), myocardin (Creemers et al., 2006b) and MASTR (Creemers et al., 2006b). Thus, combinations of post-translational modifications and interactions with regulatory protein partners create a unique integrated code for the regulation of MEF2 function, underlying the diverse roles that MEF2 presides over in development and postnatal physiology. A vivid example of the fundamental nature of posttranslational regulation of MEF2 in the control of cardiac

gene expression is shown in Fig. 1, in which two different MEF2A-expressing cultured primary neonatal cardiac myocyte preparations (as indicated by the cells stained green (D)) show dramatically different levels of MEF2 activity (as indicated by the blue staining for expression of a transgene comprised of a LacZ reporter gene controlled by MEF2 DNA-binding sites (E)). Evidence that MEF2 is heavily-regulated at the post-translational level in the postnatal heart in vivo is correspondingly illustrated by the dramatic activation of MEF2 activity observed in the hearts of transgenic mice carrying the same MEF2-LacZ transgene referred to above (Fig. 1F), after exposure to pressure overload-induced cardiac hypertrophy imposed by aortic banding (AB) when compared to a sham operated control (SH). This activation occurs without a corresponding increase in MEF2 protein levels. Thus, the question to be addressed is: how can cardiac myocytes expressing the MEF2 transcription factors exhibit such a different pattern of MEF2 activity depending on the physiological or developmental context? The answer is not simple. There are myriad types of post-translational control of protein activity such as phosphorylation, acetylation, sumoylation, ubiquitylation, carboxylation, methylation and many more, as well as the capacity for that protein to interact with and be modulated by a suite of other proteins within the cell. Thus, mechanistic understanding of the determinants of MEF2 transcription factor activity and ultimately epigenetic control of cardiac gene expression requires a systematic integration of knowledge of post-translational modifications and protein–protein interactions.

Embryonic 16 dpc heart cultures (A)

(B)

MEF2A

(C)

β - Gal

Heart Neonatal 1-2 dheart cultures (D)

MEF2A

(E)

Somite (F) MEF2

MEF2

LacZ

AB

SH

β - Gal

Figure 1  MEF2A expression (green stain (B) and (D)) and activity (blue stain (C) and (E)) in cultured cardiac myocytes from the MEF2-LacZ “sensor mouse” (A), see text for details). (F) Represents MEF2 activity (blue stain) in the hearts of adult MEF2-LacZ mice after pressure overload hypertrophy by aortic banding (AB) or sham operated control (SH).

Chapter  |  11.5  Proteomic Analysis of MEF2 Post-Translational Regulation in the Heart

Ongoing advances in proteomics technology are now allowing a systematic multi-tier analysis of these modes of regulation, with robust and extremely sensitive tools. Although the integration of multiple modifiers and protein interactions as predictors of transcriptional network activity and outcome is still in its infancy, this is a tract­able goal in the foreseeable future, given the current trajectory of technological innovation. In this overview we will consider the regulation of the MEF2 transcriptional regulatory proteins to illustrate a number of proteomic strategies aimed at a holistic understanding of post-translational regulation of transcription factor activity. Moreover, the proteomic strategies discussed could equally be applied to the analysis of other transcriptional regulators, or indeed, to many other cellular protein complexes.

II.  The central role of mass spectrometric analysis in proteomic analysis of protein– protein interactions The complex expression of genetic material, in the form of the protein complement of the cell, remains the primary and ultimate determinant of cellular and thus, whole organism, form and function (Pawson, 1995; Pawson and Nash, 2000; Seet et al., 2006). Knowledge of the protein machinery of the cell is indeed a problem of infinite complexity, requiring new technologies to dissect protein networks and their multifaceted levels of regulation. Study of the identification and characterization of proteins has spawned the field of “proteomics”. Already, technological advances in the application of mass spectrometry to biological samples have launched protein analysis to a new level of sophistication and utility. The latest generation of mass spectrometers allow for sensitive and high-resolution identification of proteins within biological matrices (Fenselau, 1997; Gatlin et al., 1998; Yates, 1998). Coupled with traditional purification techniques, this instrumentation allows for robust assignment of proteins to specific functional protein networks within the cell. Large-scale proteomics efforts to date have focused primarily on differential protein expression analysis (Mann et al., 2001; Naaby-Hansen et al., 2001). Typically, proteins from separately-treated samples are analyzed by twodimensional gel electrophoresis (O’Farrell, 1975), and the resulting expression profiles are compared for increases or decreases in protein expression. The differentially-regulated proteins are then identified by peptide mass fingerprinting or peptide fragmentation sequencing using mass spectrometry (see later discussion), thereby allowing the identification of proteins that are regulated by a specific treatment or physiological manipulation. As this type of experiment directly measures the amount of protein present under specific conditions, it is partially complementary to, and

807

perhaps more useful than, the information generated by mRNA level analysis by gene-chip technology, since the correlation between cellular mRNA and protein levels is modest (Arcellana-Panlilio and Robbins, 2002). As more and better molecular tools become available for purifying physiologically-relevant protein complexes, research focus has shifted to smaller-scale “functional proteomics”. Translation of a protein is only the first step towards its lifetime of activity. In many cases, either an underlying change in the protein complement of the cell, or a post-translational modification of the existing protein network within the cell, mediates developmental or physio­ logical adaptations. Analysis of post-translational control is therefore an extremely complex biological question to address, requiring information regarding post-translational modifications of proteins and the higher order assembly of proteins into functionally different multi-protein complexes. The challenge of acquiring this type of information can now be met using the array of mass spectrometrybased technologies currently available and in development. Thus, functional proteomics is aimed at defining the regulation and multiple roles of specific protein targets, and is ideally suited to dissecting the complex protein networks involved in specific cellular processes. Mass spectrometric techniques for analyzing proteins have reached a hitherto unprecedented level of suitability for use as tools in understanding the molecular control of protein modification. Central to understanding these techniques is an appreciation of the slightly mundane but critical issue of sample preparation, the portal representing the interface between the biological chemistry that has classically been applied to study proteins in solution, and analytical mass spectrometry where molecules are measured in the gas phase. Thus, the problem of dissecting protein–protein interactions can be operationally divided into two parts: (1) purification of relevant multiprotein complexes; and (2) the subsequent identification of the protein components.

II.A.  Purification of Multiprotein Complexes Traditional methods for studying protein–protein interactions were limited in sensitivity, restricting most studies to large-scale purifications, or the study of protein complexes that were easily purified. Analytical protein purification methods, including fractionating cellular constituents by sucrose density gradients, centrifugation, precipitation and differential solubilization, could successfully purify abundant or highly-compartmentalized protein complexes. However, for less abundant complexes or transient interactions, these methods alone have proven limited. The traditional method of gel electrophoresis does, however, remain one of the most robust and trusted methods for separating intact proteins. The separated proteins can be visualized in

808

the gel matrix using a variety of stains, transferred to membrane for further processing, or dried and stored for many years. Historically, once a protein was isolated, identification could be obtained by performing Edman degradation chemistry (Edman, 1970). In brief, a peptide is bound to a solid phase support and the N-terminal amino acid is reacted with phenylisothiocyanate (PTC). The N-terminal amino acid is cleaved off, chromatographically separated, and detected by UV absorbance. The reaction is then repeated, and limited sequence information of the peptide from the N-terminus is thus derived. Some of the limitations of this method include its limited sensitivity and the requirement for an extremely purified protein–peptide sample. The pace of discovery for research depending on protein identification was correspondingly slow. Often, a single protein identification was extremely labor-intensive and then led to many years of functional studies. The modern advancement of tools, such as mass spectrometry, has enabled the acceleration of research that depends on protein identification and characterization. Complex samples, even at low concentrations, can now be directly analyzed. Some of the improvements in technology that have enabled this include molecular engineering of protein tags and their expression in cultured mammalian cells, and sample preparation methods such as liquid chromatography coupled to mass spectrometry. Although a few questions can be answered by analyzing crude purifications, naturally, most require a more targeted approach. Molecular biology techniques allow fusion of peptide “affinity tags” to target proteins and expression of these proteins in mammalian cells. The affinity tag is used to purify the protein and proteins it interacts with on an affinity column, based on the chemical specificity between the tag and the affinity matrix. A number of different affinity tags are available, including 6x histidine, glutathione S-transferase (GST) and haemagglutinin (HA). However, the tandem-affinity tag (TAP) (Rigaut et al., 1999) and FLAG tag (Chubet and Brizzard, 1996; Figeys et al., 2001) stand out for their ability to purify intact protein complexes from mammalian cells. Fig. 2 illustrates purification schema for different affinity tags. The FLAG tag (Einhauer and Jungbauer, 2001) is a specific amino acid sequence recognized by a host of well-characterized monoclonal antibodies. Purification is achieved by using these antibodies specifically to trap the FLAG-tagged protein and the proteins it interacts with. The fidelity of the FLAG tag is superior to other methods using antibodies, because anti-FLAG binding can easily be competed away using a FLAG peptide (Einhauer and Jungbauer, 2001). This method releases the target protein without releasing nonspecific contaminating proteins. The TAP tag consists of a protein A IgG-binding domain followed by a calmodulin-binding peptide. A tobacco etch virus (TEV) protease cleavage site separates these two affinity modules. Purification is achieved under

PART  |  11  Cardiomics

nondenaturing conditions by first binding protein complexes to an IgG column. The tagged protein is released by digestion with TEV protease, followed by a second round of purification using calmodulin beads. Release of the target protein and its associated factors can then be achieved by chelating calcium, thus dissociating the protein complex from the beads. Nonspecific contaminating proteins are efficiently eliminated by using two very different affinity-purification tags in this procedure, while low stringency binding conditions retain intact protein complexes (Rigaut et al., 1999). Both the TAP- and FLAG-affinity tags have been used successfully to purify intact protein complexes from yeast (Caspary et al., 1999; Bouveret et al., 2000; Gavin et al., 2002) and recently from mammalian cells (Cox et al., 2002; Westermarck et al., 2002). These methods offer great possibilities for mapping protein–protein interactions. The ability to express a foreign DNA construct in a mammalian cell is not new; however, recent advances in this area have allowed researchers to use a variety of techniques to achieve transfection in a much wider range of cell types. Charged lipids (Felgner et al., 1987, 1994) or geneticallyengineered viral particles (Gilboa, 1986; Muzyczka, 1992; Naldini, 1998; Prince, 1998), enable expression of proteins in cell types, such as primary cardiac myocytes, that were previously difficult to transfect. The tools currently available in molecular biology allow a much broader scope of experimental question to be answered than was the case just a few years ago. One issue is that mass spectrometry-based protein analysis requires careful consideration of traditional biochemical purification methods. These methods often utilize substances, such as detergents, that are easily ionized and detected by mass spectrometry, resulting in spectra that are completely dominated by detergent peaks, swamping the signal from peptides (Bornsen, 2000; King et al., 2000). Likewise, salts and metals may complex with peptides inside the instrument, thereby spreading the signal out over a number of peaks. For these reasons and others, it is extremely important that samples for mass spectrometry are free of nonvolatile salts and detergents. This presents a challenge for the biologist, as most biological samples are prepared in a cocktail of salts, metals and detergents. However, SDS-PAGE offers a relatively simple method of trapping proteins within a polyacrylamide matrix. Once separated by gel electrophoresis, it is difficult to elute proteins from the gel. Therefore, most methods of extraction require detergents and electrical current. Conversely, if the goal is to leave the protein trapped within the gel, simply washing with water will remove most detergents and buffers. Gel pieces corresponding to specific protein bands can then be dissected manually, or by automation from the gel, washed with acetonitrile to remove Coomassie Blue dye, and then the gel piece can be dehydrated. The dried gel pieces, still containing the

809

Chapter  |  11.5  Proteomic Analysis of MEF2 Post-Translational Regulation in the Heart

TAP tag purification system : Target protein

TEV site Protein A

C

Calmodulin-binding peptide

C Associated proteins

Protein A

C

Contaminants

CBP

C C

TEV protease Target cleavage site protein

C C IgG beads

TEV protease cleavage Calmodulin beads Separate on SDS-PAGE

EGTA elution

FLAG FLAG

Target

HA Target

HA

GST GST

Target

G

Anti-FLAG

Anti-HA SDS

Glutathione SDS

SDS

FLAG peptide elution

SDS elution

6xHis 6xHis

Target

Ni++

Nickel

SDS SDS G SDS

SDS elution

EDTA elution

Figure 2  TAP tag protein purification, and other affinity tags. Purification of protein complexes has been greatly assisted by affinity tags. The TAP tag consist of a protein A tag (which binds to IgG beads), and a calmodulin-binding peptide tag (which binds calmodulin when calcium is present), separated by a TEV protease cleavage site. Other affinity tags and their corresponding purification strategies such as FLAG, HA, GST and 6xHis are also shown.

proteins, are then rehydrated with a solution of trypsin (or another protease) in a “volatile” buffer, such as ammonium bicarbonate, and the digestion is then allowed to proceed (Rosenfeld et al., 1992). Trypsin is typically used because it cleaves proteins very specifically at the C-terminal side of the amino acids lysine and arginine, and it very rarely misses a cleavage. The consequence of this high specificity and rigorous digestion is a set of peptides of varying lengths that are specific to the target protein. The resultant peptides are much smaller than the full-length protein, and are therefore much easier to extract from the gel for analysis by mass spectrometry. Tryptic digestion has proven to be a highly-reproducible method for “fingerprinting”

proteins. Even two similarly-sized but distinct proteins, once digested by trypsin, yield a substantively different peptide fingerprint. Identification of the protein is accomplished by comparing the peptide masses of the tryptic digest to a database of in silico digested known proteins (databases include UniProtKB/Swiss-Prot, NCBI and MSDB) (Yates et al., 1993). The other evolving aspect of this technology that is changing rapidly is the seamless interface of various protein separation technologies with mass spectrometry instrumentation. High-performance liquid chromatography (HPLC) is proving an excellent “front end” method for directly introducing complex protein mixtures to the mass

810

spectrometer. Liquid chromatography (LC) is based on taking advantage of the chemical properties of peptides to remove unwanted salts and detergents, and to separate each peptide in time to deliver a minimal peptide set to the mass spectrometer per unit of time (Regnier and Gooding, 1980). The peptide solution is passed over a column filled with solid particles, for example silica particles coated with a hydrophobic layer of molecules that contain a chain of 18 carbon atoms (C18). The peptides interact with the solid particles, reducing the speed at which they flow through the column, or causing them to bind strongly with the column matrix. For C18 columns (hydrophobic interaction) the peptides are released by increasing the amount of organic solvent flowing through the column. Salts do not bind to the column, and are washed away before any peptides elute. As the amount of organic solvent increases, peptides elute based roughly on their hydrophobicity (Wu et al., 1998; Issaq et al., 2002; Liu et al., 2002; Mitulovic and Mechtler, 2006). Under reversed phase conditions, the peptides or proteins are trapped by the column and then eluted by using a gradient of increasing organic solvent (Aguilar and Hearn, 1996; Mant and Hodges, 1996). This method serves the dual purposes of concentrating and separating peptides, so that the mass spectrometer can analyze one or two peptides at a time, rather than hundreds simultaneously. High-performance liquid chromatography instrumentation and methodology have also improved greatly over the last few years; it is now possible to inject L samples, preconcentrate the peptides, and elute them using flow rates as low as tens of nL per minute (Martin et al., 2000; Takahashi et al., 2002; Ihling et al., 2003; Wagner et al., 2003). The modern trend is to use highperformance liquid chromatography as an automated replacement for labor-intensive gel technology; however it can also be used to separate tryptic peptides of proteins that are unresolved by gel electrophoresis. Having discussed the production and purification of proteins compatible with mass spectrometric analysis, we will next consider the principles of protein identification and characterization by mass spectrometry.

III.  Identification of protein complex components by mass spectrometry Mass spectrometers in various forms have been around for almost a century (Thomson, 1913), but it was not until recently that instruments were capable of efficiently transferring proteins and peptides into the gas phase without causing severe degradation. Having discussed the rigors of sample preparation, the next question concerns how a protein is identified using mass spectrometry. Fig. 3 shows a mass spectrum of a peptide that has been fragmented inside a mass spectrometer. The fragments are pieces of the original

PART  |  11  Cardiomics

Y-ions from C to N terminus R1

O

C H

C

NH2

N H

R2

O

C H

C

N H

R3

O

C H

C

N H

R4

O

C H

C

OH

B-ions from N to C terminus does not form

b1

y3

b2

y2

b3

y1

VMPTKpSPPPPGGGNLGMoxNSR y14

y14 -64 608.32

1302.64

MH 2+

PTK

y15 -H 3PO 4

1435.68

1045.44

327.28

MH 2+ -H 3PO 4 996.48

626.32

y12

1172.56

y13

1269.52 309.20 378.24

y1

y2 b3y3

240.08

200

300

b4

400

y4 y5 b5 500

600

y6 b6 700

y9 y10

y7 800

y8 930.48

1205.68

y15 1349.76

1417.68

y11 900

1000 1100 1200 1300 1400 1500 1600

Figure 3  Collision-induced dissociation (CID) or fragmentation of peptides. When peptide ions collide with neutral gas molecules, the absorbed energy causes fragmentation of the weakest bonds. This results in a series of y and b ions, depending on whether the charge remains on the N- or C-terminus of the peptide ion. The amino acid sequence can be inferred from the resulting fragmentation spectra, or the spectra can be matched to a database of predicted fragmentation patterns from known peptide/protein sequences.

peptide and, just as in a puzzle each piece contributes some information regarding the structure of the original peptide. Since the most intense fragments of peptides are formed by breaking the amine bond holding each amino acid together, the fragments can be used to infer the amino acid sequence of the original peptide. Using the large amounts of known genomic and proteomic sequences, it takes very few of these short peptides to identify which protein and gene they derive from. Producing a fragmentation spectrum that can identify a peptide or protein is a primary output of a mass spectrometer; however, this alone is not enough to accelerate the pace of proteomics discovery. The advantage of mass spectrometry over more traditional approaches is its ability to handle complex mixtures and identify proteins in low abundance, even in a background of other highly-abundant proteins. To appreciate

Chapter  |  11.5  Proteomic Analysis of MEF2 Post-Translational Regulation in the Heart

Separation

Ionization

Mass analyzer Triple quadrapole

TOF

1D gel electrophoresis

nanospray/ electrospray

QqTOF

2D

liquid chromatography

MALDI

811

Figure 4  Modern mass spectrometry is a combination of separation, ionization and mass analyzer technologies. Gel separation and liquid chromatography are the most common methods of preparing samples for introduction into a mass spectrometer. The peptide sample must be transferred to the gas phase with a charge, a process called ionization (electrospray or MALDI). There is a wide array of technologies for mass analysis, including hybrid instruments that combine some of the best features of each technology.

Q TRAP

Ion Trap

FT-ICR

how a mass spectrometer can achieve this we will next consider the fundamental principles of peptide detection and analysis. Prior to detection, the peptide has to carry charge and, significantly, has to be introduced to the mass spectrometer in the gas phase. Since most biological peptide samples are in liquid, translocating them into the gas phase in a charged undamaged state can be a challenge. The discovery of electrospray ionization (ESI) and matrix assisted laser desorption/ionization (MALDI) methods for introducing peptides into a mass spectrometer was recognized as a quantum step in advancing mass spectrometry and proteomics, for which the Nobel Prize in chemistry was awarded in 2002. The basic problem addressed in this groundbreaking work concerned the transfer of peptide from the liquid into the gas phase with a charge on it. Most modern instruments use a variation of either electrospray ionization (Fenn et al., 1989) or matrix assisted laser desorption/ionization (Karas and Hillenkamp, 1988) to achieve this (Fig. 4). Electrospray ionization is performed at atmospheric pressure and is a gentle method of ionization (Fenn et al., 1989). A miniaturized version of electrospray ionization, nanospray (Wilm and Mann, 1996; Wilm et al., 1996b) involves using a gold-plated glass needle to spray 1 L of acidified peptide solution under a high electric field. Nanospray is particularly suited for proteomics, because of its low sample consumption. Nanospray results in a plume of positively-charged peptide-containing droplets, from which the liquid evaporates as they move towards the entrance of the instrument. Although details of the

mechanisms are unknown, eventually the solvent molecules are completely stripped, allowing a protonated (charged) peptide to proceed for detection. The neutral solvent molecules are either swept away by a stream of nitrogen gas or pumped away, while the charged peptides are propelled into the low-pressure region of the instrument using ion lenses. Electrospray ionization has a high propensity for producing multiply-charged peptide ions. The advantage of this is best exemplified during collision-induced dissociation (CID) fragmentation, where multiply-charged peptides fragment more efficiently (Tang and Boyd, 1992; Tang et al., 1993) making sequence interpretation easier. Another advantage of electrospray ionization is that it is readily coupled to on-line chromatographic techniques. By preseparating a complex mixture of peptides, it is possible to detect and fragment more of the peptides from the original sample. A potential drawback of electrospray ionization is that all analyses must be completed during sample introduction, because it is difficult to store the small amount of liquid for later analysis. In comparison, matrix assisted laser desorption/ioniz­ ation allows multiple samples to be spotted on a single sample storage plate. This plate can be stored for hours to weeks, and analyzed at the operator’s convenience. The technique involves coprecipitating/crystallizing a peptide solution with a matrix material, typically -cyano-4hydroxycinamic acid or dihydroxybenzoic acid (DHB). The solid matrix absorbs energy from the UV laser, which is eventually available for peptide ionization. In a complex series of consecutive and competitive reactions that are

812

poorly-understood, singly-protonated peptide ions are produced. These are then extracted from the source and typically mass-analyzed by differences in flight time (Karas and Hillenkamp, 1988). The propensity for producing singly-protonated peptides simplifies interpretation of the spectra (Sherman, 2000). Relative to electrospray ionization, matrix assisted laser desorption/ionization more readily tolerates “dirty” samples (Karas and Giessmann, 1991; Mock et al., 1992). Another potential advantage is that there are no time constraints for analyzing samples. Peptides suitable for collision-induced dissociation fragmentation can be selected at any time for analysis, or the plate can be stored for days or weeks before reanalyzing. As sample is only consumed when the laser is actually firing, it is easy to conserve precious samples. Mass spectrometers do not measure molecular mass directly, but actually measure the mass-to-charge ratio (m/z) of ions that enter the instrument. If the substance to be analyzed does not contain a charge, the instrument will not detect it. When a molecule contains a charge, it can be manipulated (i.e., moved around) and detected using electric (or magnetic) fields. All mass spectrometers use electric fields to move ions around within a vacuum, and in many cases to produce a mass spectrum. The same electric fields that can be used to move an ion around can also be used to induce fragmentation. Mass spectrometers are composed of components, or lenses, that are used to move, direct and focus ions. These lenses can be as simple as a metal plate with a small orifice, or a complex set of metal rods. In either case, the lenses are used to guide ions from one section of the mass spectrometer to another. For example, a positively-charged ion will be attracted towards a negatively-charged lens, and repelled from a positivelycharged lens. The potential difference between two such lenses determines the speed at which that ion will travel as it passes through them. Most parts of a mass spectrometer are maintained at very low pressure, so that moving ions do not collide with air molecules. These collisions would scatter the ions and cause molecular fragmentation. However, in some cases it is desirable to fragment the ion. To accomplish this, a small amount of an inert gas (typically nitrogen or argon) is introduced into a collision cell inside the mass spectrometer. As ions travel through this collision cell, they collide with gas molecules, imparting energy into the ion. This excess energy leads to the breaking of the weakest bonds within the ion. For a peptide, the weakest bonds are the amine bonds between amino acids, thus producing a series of fragmented ions that can be used to determine the peptide sequence. To produce clean fragmentation spectra for interpreting a peptide sequence, the mass spectrometer must have a single peptide present for fragmentation. In addition to sample preparation techniques for reducing the complexity of a sample, the mass spectrometer itself is quite capable of separating a mixture. Tandem mass spectrometers use

PART  |  11  Cardiomics

a first mass analyzer to filter, or isolate, a specific peptide ion at its mass-to-charge ratio value. This isolated peptide ion is then fragmented and introduced to the second mass analyzer.

IV.  Mass spectrometry instrumentation Proteomic analysis by mass spectrometry is sweeping forward because of the advances in the technology of instrumentation. The ideal instrument for peptide analysis is extremely sensitive, selective, accurate and capable of high resolution. Although great strides have been made in improving these characteristics across a wide range of mass spectrometers, there are still notable strengths to each type of instrument. Before describing these types, it is important to review the operating characteristics of sensitivity, speed, selectivity and resolution/mass accuracy. The sensitivity of mass spectrometers has improved by orders of magnitude over the last decade. Improvements in sensitivity are often the driving force behind adopting mass spectrometry as a replacement for an older method of detection. Sensitivity not only addresses the need to detect smaller amounts of a sample, but also allows the use of simpler sample preparation strategies. This not only saves in sample consumption, but also in the time it takes to perform an experiment, from sample preparation to data analysis. In addition to sensitivity and speed, it is important to consider selectivity. Peptides and proteins are often from very complex samples. There are numerous uninteresting background proteins, and nonprotein components that will be detected. Simply increasing the amount of all compounds is not enough to answer many biologically-targeted experiments. Some mass spectrometers have modes of operation (precursor ion, neutral loss, multiple reaction monitoring (MRM)) that can selectively enhance a biologically relevant peptide above the noise of the noninteresting components. The details of some of these modes of operation will be discussed below (Fig. 4). The final main operating characteristic of a mass spectrometer is the resolution and mass accuracy of the peptide ions detected. Resolution is the width of the massto-charge ratio peak, which can vary from low resolution, 0.7 Dalton (Da) (the mass of a hydrogen atom is 1Da) (quadrupole and ion trap instruments), to high resolution, 0.02–0.08 Da time-of-flight (TOF) and Fourier transform (FT) instruments. Thus, all of these instruments are capable of resolution to less than 1 Da. Resolution can separate compounds with very similar masses, which is useful, but improved mass accuracy is more beneficial in identifying a peptide. Mass accuracy is a measure of how close the measured mass of a compound is to its actual mass. Most high resolution instruments (TOF and FT) can also achieve

Chapter  |  11.5  Proteomic Analysis of MEF2 Post-Translational Regulation in the Heart

exceptional mass accuracy. This improved mass accuracy narrows the number of possible matches when searching a database of known proteins, leading to more confident identifications.

IV.A.  Operational Peptide Detection Modes A variety of ion detection modes are possible and we will discuss the basics of these methods.

IV.A.i.  Quadrupole Filter Scan The quadrupole mass filter (Dawson, 1986, 1997; Miller, 1986) uses four metal poles to create an electric field consisting of a direct current (DC) and a radio frequency (RF) component. Under specific direct current and radio frequency amplitudes, an ionized peptide with a specific mass-to-charge ratio will have a stable trajectory within the quadrupole, allowing it to exit and be detected. Other peptide ions that have different mass-to-charge ratios will not have stable trajectories, and will either collide with the metal rods or be ejected between the rods. An ion detector is placed at the exit of the quadrupole filter and the number of ions at this particular mass value can be counted over a period of time. After counting this massto-charge ratio, the direct current and radio frequency amplitudes are changed to allow, typically, ions that have 1 Da higher masses to have stable trajectories and be detected. A computer controls this process of changing the electric fields and counting the ions, and in this way a mass spectrum for a range of masses can be generated. The quadrupole ion trap operates under similar principles, and can be thought of as a three-dimensional device created by bending and joining the rods from end to end. A quadrupole instrument has high sensitivity, but it can only resolve masses to about 0.3 m/z unit or 0.3 Da for singlycharged ions. However, one of the most useful features of a quadrupole mass filter is that it is a filter. Thus, a single peptide out of a complex mixture can be selected for analysis while all other peptides are discarded. When coupled to a second mass spectrometric stage, this filtering allows a number of interesting experiments to be performed.

IV.A.ii.  Time of Flight Analysis Time of flight analyzers separate ions based on differences in the time that it takes the ions to travel down a length of tube (the flight tube) and be detected at the end of it (Wollnik, 1993; Mamyrin, 1994; Guilhaus, 1995). The principle is actually quite simple; a group of ions with various masses are given an initial push, the initial energy given to these ions is effectively 1 2 mass * velocity 2 . The ions then travel through a field-free region where no

813

change in the kinetic energy of the ion occurs. As all ions were given the same initial energy, those that have different masses will have different velocities. Ions with smaller masses will have higher velocities, and will reach the detector ahead of those that have larger masses (and therefore lower velocities). The flight times are measured and converted to mass (really m/z) after calibrating with ions of known masses. Time of flight analyzers have very high transmission (i.e., offer high sensitivity), have very high resolution and, when calibrated, have very high mass accuracy. For most peptides, it is possible to measure the mass to within 0.05 Da, much less than the mass of a hydrogen atom.

IV.A.iii.  Ion Trap Scanning The quadrupole ion trap (linear of 3D) uses a quadrupole electric field to confine ions to particular space (Todd, 1991; Ghosh, 1995; March, 2005). The mass-to-charge ratio of the ions is then measured by applying an auxiliary electric field that excites a specific mass. This excitation provides enough energy for that mass-to-charge ratio to escape from the trapping potentials. Once the ion has escaped, it is transferred to a detector (e.g., electron multiplier) where the ion is converted into a detected signal. A mass spectrum is built by scanning the auxiliary electric field that sequentially excites each mass-to-charge ratio. The advantage of an ion trap over a quadrupole filter is that the ion trap can be filled for a given amount of time and then closed, so that all trapped ions are utilized for detection. A quadrupole filter only measures the ion current for a specific mass-to-charge ratio, while all other mass-to-charge ratios are discarded. In general, an ion trap is more sensitive and can scan a mass range faster than a quadrupole instrument.

IV.A.iv.  Fourier Transform Finally, some ion trap instruments use a Fourier transform (FT) technique to measure the mass spectrum (Buchanan, 1987; Marshall, 1991; Koester et al., 1992). For example, by trapping ions within an electric field and a strong magnetic field (or in a specially designed electric field) the ions can be induced to oscillate. The rate at which they travel back and forth is related to their mass-to-charge ratio. As the ions travel back and forth, this moving charge will induce a current in a radio frequency detector. This frequency spectrum can be converted to a mass spectrum by performing a Fourier transform mathematical operation, using known masses to calibrate the spectrum. The longer the ions are allowed to oscillate, the higher the mass resolution that can be achieved. Fourier transform instruments can achieve the highest resolution of any mass spectrometer.

814

All of the above mass spectrometric detection techniques have been combined in hybrid instruments that utilize the advantages of each technique. Some examples include quadrupole/time of flight, ion trap/Fourier transform and quadrupole/ion trap. These instruments often combine the speed or selectivity of one technology with the sensitivity or resolution of another technology. The ability to measure the masses of a set of tryptic peptides is a very powerful tool, and is often easily capable of identifying an unknown protein (James et al., 1993; Yates et al., 1993). Those proteins that cannot be identified by mass fingerprinting alone can still be identified by knowing, in addition, the sequence of typically a few residues of the peptide. Furthermore, sequencing permits post-translational modifications to be detected and the sites located. To obtain sequence information, it is necessary to select one peptide, fragment it, and then measure the masses of the fragments. Tandem mass spectrometers are capable of selecting one peptide out of a complex mixture using the first mass analyzer, fragmenting the peptide in a collision cell, and analyzing the fragments using a second mass analyzer. The peptides typically fragment at the amide bonds. This type of analysis typically results in a ladder of peaks, each separated by the mass difference corresponding to the amino acid residue that was removed. This sequence information can be used to complement the mass fingerprint analysis in confirming the identity of a protein, or it can be used to search for a homologous protein sequence from an organism with a more complete sequence database (Johnson and Taylor, 2000; Shevchenko et al., 2000; Nesvizhskii, 2006). The quadrupole filter is one of the most common techniques used for isolating an ion with a specific massto-charge ratio. When a quadrupole filter is placed in front of another mass analyzer, it can be used to filter a specific mass-to-charge ratio ion for passage into the second stage of analysis. By placing a collision cell between the quadrupole filter and the second mass analyzer, it is also possible to fragment the selected ion. Similarly, an ion trap can be used to remove all mass-to-charge ratio ions except one with a specific value by filling the ion trap with a range of mass-to-charge ratio ions, and then adjusting the electric field so that only a specified mass-to-charge ratio will remain stable inside the ion trap. All other masses will be unstable and will fall out of the trap during this isolation time. A final type of mass selection is unique to time of flight instruments (Medzihradszky et al., 2000; Rejtar et al., 2002) where a timed ion selector is utilized. This is simply an ion gate that is either open (ions can pass through) or closed (no ions may pass). As the ions are separated in the first stage of time of flight, the gate remains closed until the time of arrival for the target massto-charge ratio. The gate is then opened to allow only this mass through to fragmentation and the second stage of mass analysis.

PART  |  11  Cardiomics

Thus, the biological researcher now has a suite of mass spectrometry-based detection modes that can be tailored to the specific requirements of the types of analysis to be carried out (e.g., post-translational modification or peptide identification).

V.  Identification of A mef2a interacting protein Having described the methods that can be applied to the identification of proteins by mass spectrometry, we will next consider the test case of identifying a MEF2-interacting protein using this methodology. In terms of purifying a MEF2 transcription complex, we constructed a MEF2A protein which contained a tandem affinity tag at its N-terminus, and cloned the chimeric molecule into an expression vector that could express the protein in mammalian cells. At this point we tested the capability of the fusion protein to bind to the MEF2 cis element by electrophoretic mobility shift assay. Having ascertained that the tandem affinity tag MEF2A could bind to DNA, we next assessed its ability to activate transcription in a reporter gene assay. This proved positive, and we therefore interpreted this to indicate that the tandem affinity tag MEF2A was functional in two key properties within mammalian cells, DNA-binding and transcriptional activity. Thus, we proceeded to purify the tandem affinity tag MEF2A protein using a protocol developed to isolate the protein and associated proteins based on the affinity modules of the tandem affinity tag. Using this approach, the MEF2 complex was purified and then resolved using SDS-PAGE. Several bands from the gel were excised and digested using trypsin. The samples were then analyzed using a hybrid quadrupole time of flight instrument equipped with a matrix assisted laser desorption/ionization source. Proteins were first identified using a high-resolution, high-mass accuracy mass spectrometry spectra of the intact peptides (Fig. 5) (MALDI-TOF from Applied Biosystems/ Voyager). This peptide fingerprint identified MEF2A from the most prominent band on the gel. To confirm the identity of this protein, fragmentation spectra of several peptides were obtained (MALDI-QqTOF from Applied Biosystems/ MDS Sciex). These fragmentation-spectra unequivocally identify each peptide, and therefore the protein that each peptide came from. In addition, the same techniques were used to identify a higher molecular weight protein, corresponding to a histone deacetylase (HDAC4) that co-precipitated in the TAP purification (see Chapter 10.2 for a review of histone deacetylases). Previous yeast two hybrid genetic screening strategies and biochemical analyses, from our group and others, had identified HDAC4 as a bona fide transcriptional co-repressor of MEF2 proteins (Miska et al., 1999), thus indicating the utility of this mass spectrometry based approach for identifying endogenous interacting proteins of physiological relevance.

815

Chapter  |  11.5  Proteomic Analysis of MEF2 Post-Translational Regulation in the Heart

Tryptic fingerprint of MEF2A

*

842 800

1424.61

100

Tryptic fingerprint of HDAC4 (Histone deacetylase 4)

870.54

750

90

700 80

650 600

Intensity, counts

% Intensity

70 60 50 40 30

*

1062.50

*

1330.51

*1453.69

*

2135.01

10 0 1030.0

2211.26

1626.8

*3911.93

2223.6

2820.4

3417.2

* 3927.95

300 899.53 250

0

Precursor Ion 1317.69

Keratin 1179.60

*

*

928.29

900

**

1246.64 1317.65

*

1226.67

*

**

1365.74

1324.66

1941.94

1000

1100

1200

1300

1400

1500 m/z, amu

1600

1700

1800

1900

2000

Precursor ion 1365.74

13.0

11.0

HDAC4 123aa-132aa: HQQELLEHQR

28

HDAC4 1004aa-1015aa: VLQQRPNANAVR

b11+H2O 1209.7716

10.0

26

9.0

22

Intensity, counts

24

Intensity, counts

Trypsin 1045.56

*

12.0

30

y3 440.25

20 18 16 14 Immonium

ion

12 110.07 y1 (LEHQ-H2O) 175.12 338.17 10

b2 266.12 y1-NH3 158.09 y2 303.19

b4 y3-NH3 523.23 b5 423.23 636.26

200

300

400

500

600

8.0 7.0

y1 PNAN 175.12 397.20 y1-NH3 158.10 y2-NH3 b3 257.14 341.20 5.0 6.0

4.0

b6 749.43 y6 795.42

2 0

350

50

32

4

400

100

34

6

450

150

4014.0

Mass (m/z)

8

500

200

20

* 1334.57

550

700 800 m/z, amu

b7 878.44 900

b8 1015.45

1000

Precursor Ion- H2O 1299.67

3.0

Precursor ion-NH3 1348.70

b4 469.27 b5 y7 625.38 741.42 a5 597.42

2.0 1.0

1100

1200

1300

0.0 100

200

300

400

500

600

700

800 m/z, amu

900

1000

1100

1200

1300

Figure 5  Mass spectrometric identification of MEF2A and HDAC4. The top panels show peptide mass fingerprint data that identifies MEF2A (left) and HDAC4 (right). The bottom panels show tandem mass spectrometry fragmentation data that confirms these identifications.

VI.  Proteomic analysis of reversible phosphorylation: a rheostatic control mechanism for transcription factor activity A key event in multiple signal transduction cascades is the reversible phosphorylation of tyrosine, serine and threonine (Marks, 1996; Hunter, 2000). Our understanding of the molecular mechanisms of signal transduction are thus dependent on our capability to identify and characterize phospho-acceptor sites on target proteins and their function in regulating protein activity. The use of affinity tags to purify a target protein from a mammalian source also permits the study of regulatory or constitutive post-translational modifications of that protein by mass spectrometry. For example, phosphorylation is known to be the archetypal post-translational regulator of protein function. Historically, detecting a radiolabeled phosphopeptide on a thin layer chromatographic (TLC) plate was used to give clues as to the identity of the phosphopeptide (Quadroni and James, 2000), but it was not enough

to identify the phosphorylation site convincingly. Proof of a phosphorylation site would require extensive biochemical data and site-specific mutagenesis experiments. A fragmentation pattern from a tandem mass spectrometry spectra, however, will often give definitive evidence that identifies a specific phosphopeptide. In many cases, there is also enough fragmentation evidence to distinguish which specific amino acid residue was phosphorylated within the peptide (Yan et al., 1998; Sickmann and Meyer, 2001; Shou et al., 2002; Areces et al., 2004). Biochemical studies can now be more targeted, performed faster, and return biologically-relevant data. Since phosphopeptides are often in such low abundance relative to other peptides, it can be very difficult to obtain tandem mass spectrometry on the desired phosphopeptides. If the mass spectrometer triggers tandem mass spectrometry based solely on the intensity of the peptide ion, very few phosphopeptides will be detected relative to nonphosphopeptides. It is, therefore, desirable to enrich a sample specifically for phosphopeptides. This can be accomplished using chemistry (antibodies that preferentially bind phosphorylated peptides), or by

816

PART  |  11  Cardiomics

using affinity columns such as metal or titanium dioxide (Pinkse et al., 2004; Feuerstein et al., 2005; Kange et al., 2005; Wolschin et al., 2005; Liang et al., 2006) or directly by using mass spectrometry to enrich for phosphopeptides. Mass spectrometric identification of phosphorylation can be achieved by a variety of methods. The peptide fingerprint using MALDI-TOF of an enzymatically-digested phosphorylated protein can be compared to the same dephosphorylated protein. Peptides whose mass differs by exactly 79.97 Da (HPO3) can be interpreted as being phosphorylated, since phosphorylation confers this exact mass change to a peptide.

value lower (by the mass-to-charge ratio ratio of a loss of a phosphate fragment; for example, loss of H3PO4 causes a mass-to-charge ratio shift of 49 for a doubly-charged peptide) than the first analyzer. If a peptide at the mass-tocharge ratio being scanned by the first mass analyzer loses a phosphate ion in the collision cell, it will now have a mass-to-charge ratio equal to that being scanned by the second analyzer, and a signal will be detected. Other peptides will not have a mass-to-charge ratio equal to the second analyzer, and will not trigger a tandem mass spectrometry scan, and thus the tandem mass spectrometry results for phosphopeptides are selectively enriched.

VI.A.  Precursor Ion and Neutral Loss Scanning

VI.B.  Multiple Reaction Monitoring

A more targeted approach includes using mass spectrometry scanning modes that selectively enrich for phosphor­ ylation. A precursor ion scan is one such mode possible on a triple quadrupole instrument (Carr et al., 1996; Wilm et al., 1996a; Steen et al., 2001) (Fig. 6). Peptides are scanned using the first mass analyzer and then fragmented. The second mass analyzer is used as a filter to allow only the phosphate ion through to the detector. As the first mass analyzer scans through a mass range, a signal will only be detected at any given mass if a peptide with that mass generates a phosphate fragment. When a phosphopeptide signal is detected, a tandem mass spectrometry scan is triggered. This tandem mass spectrometry scan can be used to confirm the identity of the peptide and find the site of phosphorylation. For phosphorylation, the precursor scan is performed on negatively-charged ions (the phosphate fragment is a negative ion, PO3, with a mass of 79 Da) while the tandem mass spectrometry is performed in positive mode. Thus, a complex mixture of peptides and phosphopeptides can be simplified to reveal only the phosphopeptides. In positive mode, fragmentation of a phosphopeptide often leads to the loss of the phosphate group. This leaves behind an intact positively-charged peptide, with no phosphate group. If tandem mass spectrometry spectra are acquired for every peptide in a given sample, then phosphopeptides can be identified by searching through all tandem mass spectrometry scans for spectra that show this loss of phosphate ion. In many cases, the phosphopeptide is in such low abundance that the mass spectrometer will not trigger a tandem mass spectrometry analysis. However, a neutral loss scan on a triple quadrupole instrument can filter out only the peptides that display this loss of phosphate, and thus trigger tandem mass spectrometry more selectively on these phosphopeptides (Hunter and Games, 1994; Schlosser et al., 2001). The neutral loss scan works by scanning both mass analyzers at the same time. The second mass analyzer is scanning at a mass-to-charge ratio

A more directed and sophisticated approach, in which a triple quadrupole instrument can selectively detect target modified peptides, is by using multiple reaction monitoring scans. In this mode, the first analyzer is set to the specific mass of a predicted phosphopeptide. The peptide is fragmented, and the second mass analyzer is set to the specific mass of a predicted fragment of this phosphopeptide. A signal is only detected if a peptide matches both the mass-to-charge ratio of the first analyzer, and has a fragment matching the mass-to-charge ratio of the second analyzer. It is unlikely that other peptides would have this combination mass-to-charge ratio of intact and fragmented ions. A single multiple reaction monitoring scan, by itself, would not be useful in detecting phosphopeptides, however, this single multiple reaction monitoring scan can be performed in less than 10 msec. Thus, many different multiple reaction monitoring scans can be looped together, so that hundreds can be performed every few seconds. If a signal is detected for any specific multiple reaction monitoring transition, the instrument can trigger a tandem mass spectrometry scan to be used as confirmation of the detected phosphopeptides. Multiple reaction monitoringdirected acquisition takes advantage of prior biological and mass spectrometry knowledge. The predicted phosphopeptide mass-to-charge ratios are chosen based on sequence information and known protein kinase consensus sites. The fragment mass-to-charge ratio ions are chosen based on common fragmentation rules for peptides. By selecting several hundred possible phosphopeptides for a protein of interest, it is possible to study a given protein thoroughly. Ultimately, these selective modes of detection result in the generation of tandem mass spectrometry fragmentation spectra for a modified peptide. This fragmentation data is used to further clarify exactly which residue is phosphor­ ylated (Pruvost et al., 2001; Cox et al., 2005; Ciccimaro et al., 2006). In addition to phosphorylation, proteins can contain a number of other possible modifications, such as glycosylation, methylation, acetylation and sumoylation. As modern techniques and chemistry improve, determining

817

Chapter  |  11.5  Proteomic Analysis of MEF2 Post-Translational Regulation in the Heart

Product ion scan MH2+ 1091.92

MH2+ -H3PO4

883.44

y15+2-H3PO4 y15+2 932.48

y4 518.32

y5 631.44

y6

scan

fragment

select

y3 y1 275.12 200

y2

760.48

404.32

303.20 300

y7

500

y8 1002.48

506.40 400

1043.04

873.52 923.52

600

700

800 900 m/z, amu

1135.44

1000

1100

y10

y9 1260.64 1233.60

y12 -H3PO4

1325.60

1458.72

y11

1200

1300

1400

1571.84

1500

y12 1600

= GMMPPLpSEEEELELNTQR

select

fragment

scan

Precursor ion scan

select

fragment

select

Multiple reaction monitoring MGRKKIQITRIMDERNR QVTFTKRKFGLMKKAY ELSVLCDCEIALIIFNSS NKLFQYASTDMDKVLL KYTEYNEPHESRTNSDI VEALNKKEHRGCDSPD PDTSYVLTPHTEEKYKK INEE

sequence

Q1 537.3 545.2 571.8 617.3

Q3 488.3 496.2 522.8 568.3

Sequence NRQVTFTK MRMDAWVT NFIAVSAANR SEPISPPRDR

MRM transitions

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Time, min

Mass Spec detection

Figure 6  Triple quadrupole/QTRAP modes of operation used for detecting phosphopeptides. To identify a phosphopeptide, the instrument must obtain a product ion (fragmentation) spectrum of the peptide. The peptide ion is selected, fragmented, and the resulting fragments are scanned out for detection. Knowing which mass-to-charge ratio to select is difficult because phosphopeptide ions are often hidden by the much larger nonmodified peptide ion signals. Precursor ion scanning is a mode of operation that eliminates the signal from nonmodified peptides, enriching for phosphopeptides. If the sequence of the target protein is known, predictions of phosphopeptide ions and fragments can be made. These MRM pairs (transitions) can be monitored to detect the phosphopeptide and trigger a product ion scan. MRM is the most selective and sensitive mode of operation.

818

the full panoply of chemical modifications to assess protein status will become a priority.

VI.C.  New Vistas: Quantitative Analysis of Peptides Identification of proteins and post-translational modifications is critical for biologically-relevant proteomics research. Moreover, questions of biological function are often inextricably linked to quantitative assessment. Identifying the protein or its modifications elucidates partial answers. However, quantitating protein expression levels or the stoichiometry of modified versus nonmodified states may provide more insight into biological function. Many technologies currently exist for quantitative analysis of proteins (SDS-PAGE and staining, Western Blotting, ELISA), but many techniques suffer from poor specificity, or require extensive time and labor in creating specific antibodies. Thus, there is a growing interest in using mass spectrometry to obtain quantitative information exped­ iently from proteomics samples, including stoichiometric quantitation of phosphorylation sites. There are methods for both relative and absolute quantitation. For relative quantitation, isotopically-encoded chemical tags can be used to label samples. For example, iTRAQ tags can be used to label peptides, so that when a peptide fragments it generates a signature fragment ion that can be used to quantitate how much of the peptide was present in the sample (Zieske, 2006). The tag comes in four forms, each one fragments to 114, 115, 116 or 117 Da. Four individual samples can therefore be labeled with one form of the tag, thus allowing relative quantitation between each sample. For example, this approach has been used to compare tyrosine phosphorylation of the epidermal growth factor (EGF) receptor at various time points after epidermal growth factor treatment (Zhang et al., 2005). If the fragmentation pattern of a peptide is known (i.e., from previous identification studies), multiple reaction monitoring can be designed to accurately quantitate the amount of that peptide in a complex mixture (Tuthill et al., 2000; Lee et al., 2003; Zhang et al., 2004). Multiple reaction monitoring measures a specific precursor mass and specific fragment mass that, when combined with the separation power of chromatography, gives very fast and accurate quantitation of multiple targets. This method is used extensively in pharmaceutical research for quantitating the amount of drug and drug metabolites in an organism that has been treated with the drug (Tretyakova et al., 1998; Yang et al., 2002; Fang et al., 2006). In addition to relative quantitation between samples, if a standard of known concentration is used, then the absolute quantity of the sample can be calculated. Quantitation of target peptides and modified peptides using this technique is now growing in acceptance, and offers the potential for higher

PART  |  11  Cardiomics

degrees of speed, multiplexing, specificity and accuracy when compared with traditional antibody-based techniques (Anderson and Hunter, 2006; Mayya et al., 2006). Also, coupling quantitative measurements to the identification of proteins and their modifications will yield new possibil­ ities for understanding complex biological systems.

VI.D.  Phosphopeptide Analysis of MEF2A A number of the modern techniques for phosphorylation analysis have been applied to the study of the MEF2 protein family, although initial analysis revolved around highly-purified bacterial fusion proteins coupled with traditional methods of phosphopeptide mapping. While this analysis was useful, it did not provide insight into whether these kinase-catalyzed modifications could occur in mammalian cells. Preliminary attempts at analyzing MEF2 proteins from a correct cellular context by analyzing immunoprecipitated proteins from mammalian cardiac myocyte or skeletal muscle cells were fraught with problems associated with the low abundance of these transcription factor complexes and noise from contaminating proteins. A key step forward involved the engineering and expression of tagged versions of MEF2A in mammalian cells, using either transfection or viral vectors (Cox et al., 2003). In addition to expression of MEF2A, specific kinase pathways were activated to observe their effects on MEF2A phosphorylation patterns. The MEF2A protein was purified using the engineered peptide tag, and visualized on an SDS-PAGE gel. The gel bands containing MEF2 were excised and digested with trypsin for subsequent mass spectrometry analysis (Cox et al., 2003). Several phosphorylation sites were identified in this MEF2 tryptic digest. The evidence used to identify these sites included: comparative analysis between tryptic mass fingerprints; characteristic 80 Da shifts in mass caused by phosphorylation; collision-induced dissociation fragmentation; and immobilized metal-affinity capture (IMAC) (Corthals et al., 2005) of phosphopeptides. The MALDI-TOF spectra for tryptic mass fingerprints of MEF2A expressed in mammalian cells were very similar to MEF2A expressed in bacterial cells, with one notable exception. A peak at mass 1437.69 Da, corresponding to the amino acids 255–269 (SPPPPGGGNLGMNSR), was significantly lower in mammalian samples when compared with MEF2A that was expressed and purified from bacteria. Phosphorylation of this peptide was a likely cause; however, no peak was detected at 80, 160, or 240 Da higher in mass. Careful inspection revealed a peak with a mass-to-charge ratio value of 1994.02, corresponding to a MEF2A peptide with one missed trypsin cleavage, amino acids 250–269 (VMPTKSPPPPGGGNLGMNSR) and a peak 79.94 Da higher in mass (2073.96). Equipped with this information, a targeted tandem mass spectrometry

Chapter  |  11.5  Proteomic Analysis of MEF2 Post-Translational Regulation in the Heart

was performed on the suspected phosphopeptide using a hybrid Qq-TOF mass spectrometer. Fragmentation confirmed that this peptide was MEF2A 250–269 and that it was phosphorylated at Ser-255. Functional analysis of Ser 255 phosphorylation has revealed that it is a highlyevolutionarily-conserved phospho-acceptor site that regulates MEF2A protein stability (Cox et al., 2003). Another approach to characterize the phosphorylation status of MEF2A was a directed approach in which the tryptic mass fingerprint of MEF2A, co-expressed with p38 MAPK and activated MKK6 (members of the p38 MAP kinase pathway for which MEF2A has consensus phospho-acceptor sites), was compared to the fingerprint from cells expressing MEF2A alone. Using several mass spectrometric techniques, a number of phospho-acceptor target sites were subsequently identified. Peptide mass fingerprinting provided evidence for phosphorylation of MEF2A 190–233, MEF2A 95–114, MEF2A 283–300, and MEF2A 475–498 peptides. In addition to the observed 80 Da mass shift, other chemical evidence was also used to confidently identify these phosphorylated peptides. The MEF2A 190 233 peptide (NVSPGAPQRPPSTGNAGGMLSTTDLTVP NGAGSSPVGNGFVNSR), showed evidence of a 16 Da mass shift on both the nonmodified and phosphorylated forms of the peptide. This mass shift is caused by oxidation of the methionine residue when the sample is exposed to air. Observing the shift on both the modified and nonmodified forms of the peptide provided additional evidence for identifying this phosphopeptide. The MEF2A 283–300 (GMMPPLSEEEELELNTQR) peptide was also observed with this characteristic phosphorylation/oxidation pattern, due to the multiple methionine residues contained within it. Additionally, this peptide was selectively-enriched by using copper immobilized metalaffinity capture (IMAC) techniques (Corthals et al., 2005), again providing further evidence that this peptide was phosphorylated in vivo. In addition, the peptide MEF2A 404–413 was identified by targeted tandem mass spectrometry experiments for suspected phosphopeptides. This fragmentation data conclusively identified phosphorylation of Ser-408. Following identification of these phosphorylation sites, new techniques and instrumentation became available. In particular, the hybrid combination of a triple quadrupole and ion trap instrument provided new methods for phosphorylation analysis. Application of this new technique to MEF2A resulted in the confirmation of previous sites, as well as the identification of an additional site. Since the protein sequence was known, 58 targeted multiple reaction monitoring scans were performed to trigger tandem mass spectrometry on predicted phosphopeptides. The multiple reaction monitoring scans were designed based on the calculated mass of the predicted phosphopeptide, and a likely fragment (e.g., the b2 ion). Each multiple reaction monitoring scan was monitored once during a two second acquisition time (30–50 msec for each multiple reaction

819

monitoring transition). If a signal was detected, it triggered a dependent tandem mass spectrometry scan (a process termed information-dependent acquisition or IDA). The triggered tandem mass spectrometry data was used to conclusively identify the phosphopeptides. A single analysis using this technique was able to confirm the phosphorylation status of peptides (with fragmentation data confirming the sequence) that had previously taken several experiments to decipher on different instruments. Compared with the traditional methods of phosphopeptide discovery (neutral loss and precursor ion scans), targeted multiple reaction monitoring-informationdependent acquisition was successful in identifying more phosphopeptides from MEF2A, and also from a standard phosphopeptide (-casein). It is also significant to note that targeted multiple reaction monitoring-informationdependent acquisition had fewer false positives (dependent tandem mass spectrometry scans triggered for nonphosphorylated peptides). This can be extremely important when attempting to identify a potential phosphopeptide from a complex mixture of nonphosphopeptides. Taken together, these studies of MEF2 phosphorylation status have given us many insights into how the MEF2 proteins are regulated in a cellular context by signaling pathways. The characterization of the pathways regulating these post-translational modifications of MEF2 are just beginning to be unraveled, but will ultimately provide vital information on signaling pathways that are controlling cardiac gene expression during development and in physiological and pathological postnatal physiology of the heart. Thus, the post-translational modifications leading to altered functional activity (referred to in Fig. 7) are now being systematically dissected and characterized.

VII.  A transition-state model of MEF2 regulation While we are still in the early phase of understanding the multi-tiered complexity of how MEF2 transcription factors function as receivers and nuclear effectors of cellular signaling, we can now attempt to model how these different modes of regulation can be integrated. In this preliminary attempt, we have used several mass spectrometric-based lines of evidence concerning post-translational modification and protein–protein interactions in concert with functional assays to posit a tripartite model of MEF2 activity “states”, as indicated in Fig. 8. The first state is a “repressed” state in which the MEF2 protein is associated with co-repressors, such as HDAC4, and modified by negatively-acting post-translational modifications. The second state we refer to as the “derepressed” state, which would be typified by release from co-repressors (such as HDAC4) and the reversal of inhibitory post-translational modifications. The third state is deemed the “activated

820

PART  |  11  Cardiomics

Figure 7  Summary of MEF2A phosphorylation. Phosphorylation sites of MEF2A identified by both traditional methods (bottom) and mass spectrometry (top).

Phosphorylation of MEF2A

MADS

MEF2

p38

CK2 P

S-255 p38 dock

P S-408

S-289

P

P S-453 -

S-494 -

282

S-192 or 223

S-98

266

p38 P

p38 P

p38

ERK? cdk5? GSK3? P

T-312 T-319 P P

S-59 P CK2

p38

Figure 8  Transition state model of MEF2 regulation. See text for details.

Transition state model of MEF2 regulation

R

M

Repressed

R

M

P’pase R

M

R

M

M

Activated

M

A

De-repressed

A

Kinase

state”, a hallmark of which would be association with coactivators and positively acting post-translational modifications. Thus, each state would be identified by a specific “code” of targeted modifications and protein interactions. Using this model, it is possible to predict and test the implications of single- and multiple-layers of post-translational regulation for transcriptional output, and to work toward an integrated view of MEF2 transcription factor regulation. Clearly, from our discussion so far, the consequence of the post-translational code underlying transcription factor activity is that it will largely dictate the activation of target genes and ultimately the proteome of the cell. Since MEF2 activity has been implicated in pathological gene expression in cardiac disease states, such as cardiac hypertrophy and dilated cardiomyopathy, a tractable way to efficiently sample the MEF2 “code” in cardiac myocytes and ultimately in the intact heart and predict phenotypic consequences (e.g., cardiac hypertrophy) is paramount if we are to be able to neutralize or reverse these states therapeutically. Combinatorial chemistry could then be utilized to develop therapeutics

that target specific protein modifications or interactions that are crucial determinants of abnormal cardiac gene expression. Continuing advancement of mass spectrometry-based proteomic analysis strategies does indeed proffer us this level of sophisticated protein analysis. To ultimately realize the full potential of this integrated approach will require a fusion of molecular and cell biology, analytical chemistry, combinatorial chemistry and computational approaches, systematically applied to the study of cardiac transcription factors and gene expression.

References Aguilar, M.I., Hearn, M.T., 1996. High-resolution reversed-phase highperformance liquid chromatography of peptides and proteins. Methods Enzymol. 270, 3–26. Anderson, J.P., Dodou, E., Heidt, A.B., De Val, S.J., Jaehnig, E.J., Greene, S.B., Olson, E.N., Black, B.L., 2004. HRC is a direct transcriptional target of MEF2 during cardiac, skeletal, and arterial smooth muscle development in vivo. Mol. Cell Biol. 24, 3757–3768.

Chapter  |  11.5  Proteomic Analysis of MEF2 Post-Translational Regulation in the Heart

Anderson, L., Hunter, C.L., 2006. Quantitative mass spectrometric multiple reaction monitoring assays for major plasma proteins. Mol. Cell Proteomics 5, 573–588. Arcellana-Panlilio, M., Robbins, S.M., 2002. Cutting-edge technology. I. Global gene expression profiling using DNA microarrays. Am. J. Physiol. Gastrointest. Liver Physiol. 282, G397–G402. Areces, L.B., Matafora, V., Bachi, A., 2004. Analysis of protein phosphor­ ylation by mass spectrometry. Eur. J. Mass. Spectrom. (Chichester, UK) 10, 383–392. Black, B.L., Olson, E.N., 1998. Transcriptional control of muscle development by myocyte enhancer factor-2 (MEF2) proteins. Annu. Rev. Cell Dev. Biol. 14, 167–196. Bornsen, K., 2000. Influence of salts, buffers, detergents, solvents, and matrices on MALDI-MS protein analysis in complex mixtures. Methods Mol. Biol. 146, 387–404. Bouveret, E., Rigaut, G., Shevchenko, A., Wilm, M., Seraphin, B., 2000. An Sm-like protein complex that participates in mRNA degradation. Embo. J. 19, 1661–1671. Buchanan, M., 1987. Fourier Transform Mass Spectrometry (Evolution, Innovation and Applications). American Chemical Society, Washington, DC. Carr, S.A., Huddleston, M.J., Annan, R.S., 1996. Selective detection and sequencing of phosphopeptides at the femtomole level by mass spectrometry. Anal. Biochem. 239, 180–192. Caspary, F., Shevchenko, A., Wilm, M., Seraphin, B., 1999. Partial purification of the yeast U2 snRNP reveals a novel yeast pre-mRNA splicing factor required for pre-spliceosome assembly. Embo. J. 18, 3463–3474. Chang, S., Young, B.D., Li, S., Qi, X., Richardson, J.A., Olson, E.N., 2006. Histone deacetylase 7 maintains vascular integrity by repressing matrix metalloproteinase 10. Cell 126, 321–334. Chien, K.R., Zhu, H., Knowlton, K.U., Miller-Hance, W., van-Bilsen, M., O’Brien, T.X., Evans, S.M., 1993. Transcriptional regulation during cardiac growth and development. Annu. Rev. Physiol. 55, 77–95. Chubet, R.G., Brizzard, B.L., 1996. Vectors for expression and secretion of FLAG epitope-tagged proteins in mammalian cells. Biotechniques 20, 136–141. Ciccimaro, E., Hevko, J., Blair, I.A., 2006. Analysis of phosphorylation sites on focal adhesion kinase using nanospray liquid chromatography/multiple reaction monitoring mass spectrometry. Rapid. Commun. Mass. Spectrom. 20, 3681–3692. Corthals, G.L., Aebersold, R., Goodlett, D.R., 2005. Identification of phosphorylation sites using microimmobilized metal affinity chromatography. Methods Enzymol. 405, 66–81. Cox, D.M., Du, M., Guo, X., Siu, K.W., McDermott, J.C., 2002. Tandem affinity purification of protein complexes from mammalian cells. Biotechniques 33, 267–268. Cox, D.M., Du, M., Marback, M., Yang, E.C., Chan, J., Siu, K.W., McDermott, J.C., 2003. Phosphorylation motifs regulating the stability and function of myocyte enhancer factor 2A. J. Biol. Chem. 278, 15297–15303. Cox, D.M., Zhong, F., Du, M., Duchoslav, E., Sakuma, T., McDermott, J.C., 2005. Multiple reaction monitoring as a method for identifying protein posttranslational modifications. J. Biomol. Tech. 16, 83–90. Creemers, E.E., Sutherland, L.B., McAnally, J., Richardson, J.A., Olson, E.N., 2006a. Myocardin is a direct transcriptional target of Mef2, Tead and Foxo proteins during cardiovascular development. Development 133, 4245–4256. Creemers, E.E., Sutherland, L.B., Oh, J., Barbosa, A.C., Olson, E.N., 2006b. Coactivation of MEF2 by the SAP domain proteins myocardin and MASTR. Mol. Cell 23, 83–96.

821

Dawson, P.H., 1986. Quadrupole mass analyzers. Mass. Spectrom. Rev. 5, 1–37. Dawson, P.H., 1997. Quadrupole Mass Spectrometry and Its Applications. American Institute of Physics Press, Woodbury, NY. Di Lisi, R., Millino, C., Calabria, E., Altruda, F., Schiaffino, S., Ausoni, S., 1998. Combinatorial cis-acting elements control tissue-specific activation of the cardiac troponin I gene in vitro and in vivo. J. Biol. Chem. 273, 25371–25380. Dodou, E., Verzi, M.P., Anderson, J.P., Xu, S.M., Black, B.L., 2004. Mef2c is a direct transcriptional target of ISL1 and GATA factors in the anterior heart field during mouse embryonic development. Development 131, 3931–3942. Doevendans, P.A., van Bilsen, M., 1996. Transcription factors and the cardiac gene programme. Int. J. Biochem. Cell Biol. 28, 387–403. Edman, P., 1970. Sequence determination. Mol. Biol. Biochem. Biophys. 8, 211–255. Edmondson, D.G., Lyons, G.E., Martin, J.F., Olson, E.N., 1994. Mef2 gene expression marks the cardiac and skeletal muscle lineages during mouse embryogenesis. Development 120, 1251–1263. Einhauer, A., Jungbauer, A., 2001. The FLAG peptide, a versatile fusion tag for the purification of recombinant proteins. J. Biochem. Biophys. Methods 49, 455–465. Fang, T., Wang, Y., Ma, Y., Su, W., Bai, Y., Zhao, P., 2006. A rapid LC/ MS/MS quantitation assay for naringin and its two metabolites in rats’ plasma. J. Pharm. Biomed. Anal. 40, 454–459. Felgner, J.H., Kumar, R., Sridhar, C.N., Wheeler, C.J., Tsai, Y.J., Border, R., Ramsey, P., Martin, M., Felgner, P.L., 1994. Enhanced gene delivery and mechanism studies with a novel series of cationic lipid formulations. J. Biol. Chem. 269, 2550–2561. Felgner, P.L., Gadek, T.R., Holm, M., Roman, R., Chan, H.W., Wenz, M., Northrop, J.P., Ringold, G.M., Danielsen, M., 1987. Lipofection: a highly efficient, lipid-mediated DNA-transfection procedure. Proc. Natl. Acad. Sci. USA 84, 7413–7417. Fenn, J.B., Mann, M., Meng, C.K., Wong, S.F., Whitehouse, C.M., 1989. Electrospray ionization for mass spectrometry of large biomolecules. Science 246, 64–71. Fenselau, C., 1997. MALDI MS and strategies for protein analysis. Anal. Chem. 69, 661A–665A. Feuerstein, I., Morandell, S., Stecher, G., Huck, C.W., Stasyk, T., Huang, H.L., Huber, L.A., Bonn, G.K., 2005. Phosphoproteomic analysis using immobilized metal ion affinity chromatography on the basis of cellulose powder. Proteomics 5, 46–54. Figeys, D., McBroom, L.D., Moran, M.F., 2001. Mass spectrometry for the study of protein–protein interactions. Methods 24, 230–239. Gatlin 3rd, C.L., Kleemann, G.R., Hays, L.G., Link, A.J., Yates, J.R., 1998. Protein identification at the low femtomole level from silverstained gels using a new fritless electrospray interface for liquid chromatography-microspray and nanospray mass spectrometry. Anal. Biochem. 263, 93–101. Gavin, A.C., Bosche, M., Krause, R., Grandi, P., Marzioch, M., Bauer, A., Schultz, J., Rick, J.M., Michon, A.M., Cruciat, C.M., et al., 2002. Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 415, 141–147. Ghosh, P., 1995. Ion Traps. Clarendon Press, Oxford, UK. Gilboa, E., 1986. Retrovirus vectors and their uses in molecular biology. Bioessays 5, 252–257. Gong, X., Tang, X., Wiedmann, M., Wang, X., Peng, J., Zheng, D., Blair, L.A., Marshall, J., Mao, Z., 2003. Cdk5-mediated inhibition of the protective effects of transcription factor MEF2 in neurotoxicityinduced apoptosis. Neuron 38, 33–46.

822

Grayson, J., Bassel-Duby, R., Williams, R.S., 1998. Collaborative interactions between MEF-2 and Sp1 in muscle-specific gene regulation. J. Cell Biochem. 70, 366–375. Gregoire, S., Yang, X.J., 2005. Association with class IIa histone deacetyl­ ases upregulates the sumoylation of MEF2 transcription factors. Mol. Cell Biol. 25, 2273–2287. Guilhaus, M., 1995. Principles and instrumentation for TOF-MS. J. Mass. Spectrometry 30, 1519–1532. Han, J., Jiang, Y., Li, Z., Kravchenko, V.V., Ulevitch, R.J., 1997. Activation of the transcription factor MEF2C by the MAP kinase p38 in inflammation. Nature 386, 296–299. Holmberg, C.I., Tran, S.E., Eriksson, J.E., Sistonen, L., 2002. Multisite phosphorylation provides sophisticated regulation of transcription factors. Trends Biochem. Sci. 27, 619–627. Hunter, A.P., Games, D.E., 1994. Chromatographic and mass spectrometric methods for the identification of phosphorylation sites in phosphoproteins. Rapid. Commun. Mass. Spectrom. 8, 559–570. Hunter, T., 2000. Signaling – 2000 and beyond. Cell 100, 113–127. Ihling, C., Berger, K., Hofliger, M.M., Fuhrer, D., Beck-Sickinger, A.G., Sinz, A., 2003. Nano-high-performance liquid chromatography in combination with nano-electrospray ionization Fourier transform ion-cyclotron resonance mass spectrometry for proteome analysis. Rapid. Commun. Mass. Spectrom. 17, 1240–1246. Issaq, H.J., Conrads, T.P., Janini, G.M., Veenstra, T.D., 2002. Methods for fractionation, separation and profiling of proteins and peptides. Electrophoresis 23, 3048–3061. James, P., Quadroni, M., Carafoli, E., Gonnet, G., 1993. Protein identification by mass profile fingerprinting. Biochem. Biophys. Res. Commun. 195, 58–64. Johnson, R.S., Taylor, J.A., 2000. Searching sequence databases via de novo peptide sequencing by tandem mass spectrometry. Methods Mol. Biol. 146, 41–61. Kange, R., Selditz, U., Granberg, M., Lindberg, U., Ekstrand, G., Ek, B., Gustafsson, M., 2005. Comparison of different IMAC techniques used for enrichment of phosphorylated peptides. J. Biomol. Tech. 16, 91–103. Karas, M., Hillenkamp, F., 1988. Laser desorption ionization of proteins with molecular masses exceeding 10,000 daltons. Anal. Chem. 60, 2299–2301. Karas, M., Giessmann, U., 1991. Matrix-assisted laser desorption ionization mass spectrometry. Mass. Spectrometry Reviews 10, 335–357. Kato, Y., Kravchenko, V.V., Tapping, R.I., Han, J., Ulevitch, R.J., Lee, J.D., 1997. BMK1/ERK5 regulates serum-induced early gene expression through transcription factor MEF2C. Embo. J. 16, 7054–7066. King, R., Bonfiglio, R., Fernandez-Metzler, C., Miller-Stein, C., Olah, T., 2000. Mechanistic investigation of ionization suppression in electrospray ionization. J. Am. Soc. Mass. Spectrom. 11, 942–950. Koester, M.K., Castoro, J.A., Wilkins, C.A., 1992. FT-MS. Mass. Spectrom. Rec. 11, 459–512. Kuisk, I.R., Li, H., Tran, D., Capetanaki, Y., 1996. A single MEF2 site governs desmin transcription in both heart and skeletal muscle during mouse embryogenesis. Dev. Biol. 174, 1–13. Lee, J., Son, J., Lee, M., Lee, K.T., Kim, D.H., 2003. Simultaneous quantitation of enalapril and enalaprilat in human plasma by 96-well solid-phase extraction and liquid chromatography/tandem mass spectrometry. Rapid. Commun. Mass. Spectrom. 17, 1157–1162. Lemercier, C., Verdel, A., Galloo, B., Curtet, S., Brocard, M.P., Khochbin, S., 2000. MHDA1/HDAC5 histone deacetylase interacts with and represses MEF2A transcriptional activity. J. Biol. Chem. 275, 15594–15599.

PART  |  11  Cardiomics

Liang, S.S., Makamba, H., Huang, S.Y., Chen, S.H., 2006. Nanotitanium dioxide composites for the enrichment of phosphopeptides. J. Chromatogr. A. 1116, 38–45. Lin, Q., Schwarz, J., Bucana, C., Olson, E.N., 1997. Control of mouse cardiac morphogenesis and myogenesis by transcription factor MEF2C. Science 276, 1404–1407. Lin, Q., Lu, J., Yanagisawa, H., Webb, R., Lyons, G.E., Richardson, J.A., Olson, E.N., 1998. Requirement of the MADS-box transcription factor MEF2C for vascular development. Development 125, 4565–4574. Liu 3rd, H., Lin, D., Yates, J.R., 2002. Multidimensional separations for protein/peptide analysis in the post-genomic era. Biotechniques 32, 898. 900, 902 passim. Lu, J., McKinsey, T.A., Zhang, C.L., Olson, E.N., 2000. Regulation of skeletal myogenesis by association of the MEF2 transcription factor with class II histone deacetylases. Mol. Cell 6, 233–244. Ma, K., Chan, J.K., Zhu, G., Wu, Z., 2005. Myocyte enhancer factor 2 acetylation by p300 enhances its DNA binding activity, transcriptional activity, and myogenic differentiation. Mol. Cell Biol. 25, 3575–3582. Mamyrin, B., 1994. Reflectron TOF-MS. Int. J. Mass. Spectrom. Ion. Proc. 131, 1–19. Mann, M., Hendrickson, R.C., Pandey, A., 2001. Analysis of proteins and proteomes by mass spectrometry. Annu. Rev. Biochem. 70, 437–473. Mant, C.T., Hodges, R.S., 1996. Analysis of peptides by high-performance liquid chromatography. Methods Enzymol. 271, 3–50. Mao, Z., Bonni, A., Xia, F., Nadal-Vicens, M., Greenberg, M.E., 1999. Neuronal activity-dependent cell survival mediated by transcription factor MEF2. Science 286, 785–790. March, R.E., 2005. Quadrupole Ion Trap Mass Spectrometry. WileyInterscience, New York, NY. Marks, F., 1996. Protein Phosphorylation. VCH, New York, NY. Marshall, P.G., 1991. FT-ICR-MS. Anal. Chem. 63, 215A–229A. Martin, S.E., Shabanowitz, J., Hunt, D.F., Marto, J.A., 2000. Subfemtomole MS and MS/MS peptide sequence analysis using nano-HPLC micro-ESI fourier transform ion cyclotron resonance mass spectrometry. Anal. Chem. 72, 4266–4274. Mayya, V., Rezual, K., Wu, L., Fong, M.B., Han, D.K., 2006. Absolute quantification of multisite phosphorylation by selective reaction monitoring mass spectrometry: determination of inhibitory phosphory­ lation status of cyclin-dependent kinases. Mol. Cell Proteomics 5, 1146–1157. Medzihradszky, K.F., Campbell, J.M., Baldwin, M.A., Falick, A.M., Juhasz, P., Vestal, M.L., Burlingame, A.L., 2000. The characteristics of peptide collision-induced dissociation using a high-performance MALDI-TOF/ TOF tandem mass spectrometer. Anal. Chem. 72, 552–558. Miller, P.E., 1986. Operating concepts of quadrupole MS. J. Chem. Educ. 63, 617–622. Miska, E.A., Karlsson, C., Langley, E., Nielsen, S.J., Pines, J., Kouzarides, T., 1999. HDAC4 deacetylase associates with and represses the MEF2 transcription factor. Embo. J. 18, 5099–5107. Mitulovic, G., Mechtler, K., 2006. HPLC techniques for proteomics analysis – a short overview of latest developments. Brief Funct. Genomic. Proteomic. 5, 249–260. Mock, K.K., Sutton, C.W., Cottrell, J.S., 1992. Sample immobilization protocols for matrix-assisted laser-desorption mass spectrometry. Rapid. Commun. Mass. Spectrom. 6, 233–238. Molkentin, J.D., Markham, B.E., 1993. Myocyte-specific enhancer-binding factor (MEF-2) regulates alpha-cardiac myosin heavy chain gene expression in vitro and in vivo. J. Biol. Chem. 268, 19512–19520.

Chapter  |  11.5  Proteomic Analysis of MEF2 Post-Translational Regulation in the Heart

Molkentin, J.D., Black, B.L., Martin, J.F., Olson, E.N., 1995. Cooperative activation of muscle gene expression by MEF2 and myogenic bHLH proteins. Cell 83, 1125–1136. Molkentin, J.D., Black, B.L., Martin, J.F., Olson, E.N., 1996a. Mutational analysis of the DNA binding, dimerization, and transcriptional activation domains of MEF2C. Mol. Cell Biol. 16, 2627–2636. Molkentin, J.D., Li, L., Olson, E.N., 1996b. Phosphorylation of the MADS-Box transcription factor MEF2C enhances its DNA binding activity. J. Biol. Chem. 271, 17199–17204. Morin, S., Charron, F., Robitaille, L., Nemer, M., 2000. GATA-dependent recruitment of MEF2 proteins to target promoters. Embo. J. 19, 2046–2055. Muzyczka, N., 1992. Use of adeno-associated virus as a general transduction vector for mammalian cells. Curr. Top. Microbiol. Immunol. 158, 97–129. Naaby-Hansen, S., Waterfield, M.D., Cramer, R., 2001. Proteomics – post-genomic cartography to understand gene function. Trends Pharmacol. Sci. 22, 376–384. Naldini, L., 1998. Lentiviruses as gene transfer agents for delivery to non-dividing cells. Curr. Opin. Biotechnol. 9, 457–463. Naya, F.J., Olson, E.N., 1999. MEF2: a transcriptional target for signaling pathways controlling skeletal muscle growth and differentiation. Curr. Opin. Cell Biol. 11, 683–688. Naya, F.J., Black, B.L., Wu, H., Bassel-Duby, R., Richardson, J.A., Hill, J.A., Olson, E.N., 2002. Mitochondrial deficiency and cardiac sudden death in mice lacking the MEF2A transcription factor. Nat. Med. 8, 1303–1309. Nesvizhskii, A.I., 2006. Protein identification by tandem mass spectrometry and sequence database searching. Methods Mol. Biol. 367, 87–120. O’Farrell, P.H., 1975. High resolution two-dimensional electrophoresis of proteins. J. Biol. Chem. 250, 4007–4021. Olson, E.N., Srivastava, D., 1996. Molecular pathways controlling heart development. Science 272, 671–676. Ornatsky, O.I., Andreucci, J.J., McDermott, J.C., 1997. A dominant-negative form of transcription factor MEF2 inhibits myogenesis. J. Biol. Chem. 272, 33271–33278. Ornatsky, O.I., Cox, D.M., Tangirala, P., Andreucci, J.J., Quinn, Z.A., Wrana, J.L., Prywes, R., Yu, Y.T., McDermott, J.C., 1999. Posttranslational control of the MEF2A transcriptional regulatory protein. Nucleic. Acids Res. 27, 2646–2654. Park, S.Y., Shin, H.M., Han, T.H., 2002. Synergistic interaction of MEF2D and Sp1 in activation of the CD14 promoter. Mol. Immunol. 39, 25–30. Pawson, T., 1995. Protein modules and signalling networks. Nature 373, 573–580. Pawson, T., Nash, P., 2000. Protein–protein interactions define specificity in signal transduction. Genes Dev. 14, 1027–1047. Pawson, T., Scott, J.D., 2005. Protein phosphorylation in signaling – 50 years and counting. Trends Biochem. Sci. 30, 286–290. Pinkse, M.W., Uitto, P.M., Hilhorst, M.J., Ooms, B., Heck, A.J., 2004. Selective isolation at the femtomole level of phosphopeptides from proteolytic digests using 2D-NanoLC-ESI-MS/MS and titanium oxide precolumns. Anal. Chem. 76, 3935–3943. Prince, H.M., 1998. Gene transfer: a review of methods and applications. Pathology 30, 335–347. Pruvost, A., Becher, F., Bardouille, P., Guerrero, C., Creminon, C., Delfraissy, J.F., Goujard, C., Grassi, J., Benech, H., 2001. Direct determination of phosphorylated intracellular anabolites of stavudine (d4T) by liquid chromatography/tandem mass spectrometry. Rapid. Commun. Mass. Spectrom. 15, 1401–1408.

823

Quadroni, M., James, P., 2000. Phosphopeptide analysis. Exs 88, 199–213. Quinn, Z.A., Yang, C.C., Wrana, J.L., McDermott, J.C., 2001. Smad proteins function as co-modulators for MEF2 transcriptional regulatory proteins. Nucleic. Acids Res. 29, 732–742. Regnier, F.E., Gooding, K.M., 1980. High-performance liquid chromatography of proteins. Anal. Biochem. 103, 1–25. Rejtar, T., Hu, P., Juhasz, P., Campbell, J.M., Vestal, M.L., Preisler, J., Karger, B.L., 2002. Off-line coupling of high-resolution capillary electrophoresis to MALDI-TOF and TOF/TOF MS. J. Proteome. Res. 1, 171–179. Rigaut, G., Shevchenko, A., Rutz, B., Wilm, M., Mann, M., Seraphin, B., 1999. A generic protein purification method for protein complex characterization and proteome exploration. Nat. Biotechnol. 17, 1030–1032. Rosenfeld, J., Capdevielle, J., Guillemot, J.C., Ferrara, P., 1992. In-gel digestion of proteins for internal sequence analysis after one- or two-dimensional gel electrophoresis. Anal. Biochem. 203, 173–179. Sartorelli, V., Huang, J., Hamamori, Y., Kedes, L., 1997. Molecular mechanisms of myogenic coactivation by p300: direct interaction with the activation domain of MyoD and with the MADS box of MEF2C. Mol. Cell Biol. 17, 1010–1026. Schlosser, A., Pipkorn, R., Bossemeyer, D., Lehmann, W.D., 2001. Analysis of protein phosphorylation by a combination of elastase digestion and neutral loss tandem mass spectrometry. Anal. Chem. 73, 170–176. Scott, J.D., Pawson, T., 2000. Cell communication: the inside story. Sci. Am. 282, 72–79. Seet, B.T., Dikic, I., Zhou, M.M., Pawson, T., 2006. Reading protein modifications with interaction domains. Nat. Rev. Mol. Cell Biol. 7, 473–483. Shevchenko, A., Chernushevich, I., Wilm, M., Mann, M., 2000. De novo peptide sequencing by nanoelectrospray tandem mass spectrometry using triple quadrupole and quadrupole/time-of-flight instruments. Methods Mol. Biol. 146, 1–16. Shore, P., Sharrocks, A.D., 1995. The MADS-box family of transcription factors. Eur. J. Biochem. 229, 1–13. Shou, W., Verma, R., Annan, R.S., Huddleston, M.J., Chen, S.L., Carr, S.A., Deshaies, R.J., 2002. Mapping phosphorylation sites in proteins by mass spectrometry. Methods Enzymol. 351, 279–296. Sickmann, A., Meyer, H.E., 2001. Phosphoamino acid analysis. Proteomics 1, 200–206. Sparrow, D.B., Miska, E.A., Langley, E., Reynaud-Deonauth, S., Kotecha, S., Towers, N., Spohr, G., Kouzarides, T., Mohun, T.J., 1999. MEF-2 function is modified by a novel co-repressor, MITR. Embo. J. 18, 5085–5098. Srivastava, D., 2006. Making or breaking the heart: from lineage determination to morphogenesis. Cell 126, 1037–1048. Steen, H., Kuster, B., Fernandez, M., Pandey, A., Mann, M., 2001. Detection of tyrosine phosphorylated peptides by precursor ion scanning quadrupole TOF mass spectrometry in positive ion mode. Anal. Chem. 73, 1440–1448. Takahashi, Y., Morita, T., Ueda, Y., 2002. Accurate mass measurement using multiple sprayer nano-electrospray mass spectrometry combined with nano-scale high-performance liquid chromatography on a magnetic sector instrument. J. Chromatogr. B. Analyt. Technol. Biomed. Life Sci. 776, 31–38. Tang, X.J., Boyd, R.K., 1992. An investigation of fragmentation mechanisms of doubly protonated tryptic peptides. Rapid. Commun. Mass. Spectrom. 6, 651–657.

824

Tang, X.J., Thibault, P., Boyd, R.K., 1993. Fragmentation reactions of multiply-protonated peptides and implications for sequencing by tandem mass spectrometry with low-energy collision-induced dissociation. Anal. Chem. 65, 2824–2834. Thomson, J., 1913. Rays of Positive Electricity and Their Application to Chemical Analysis. Longmans, Green and Co., London, UK. Todd, J., 1991. Ion-trap MS. Mass. Spectrom. Rev. 10, 3–52. Tretyakova, N., Chiang, S.Y., Walker, V.E., Swenberg, J.A., 1998. Quantitative analysis of 1,3-butadiene-induced DNA adducts in vivo and in vitro using liquid chromatography electrospray ionization tandem mass spectrometry. J. Mass. Spectrom. 33, 363–376. Tuthill, C.W., Rudolph, A., Li, Y., Tan, B., Fitzgerald, T.J., Beck, S.R., Li, Y.X., 2000. Quantitative analysis of thymosin alpha1 in human serum by LC-MS/MS. AAPS PharmSciTech. 1, E11. Wagner, Y., Sickmann, A., Meyer, H.E., Daum, G., 2003. Multidimensional nano-HPLC for analysis of protein complexes. J. Am. Soc. Mass. Spectrom. 14, 1003–1011. Westermarck, J., Weiss, C., Saffrich, R., Kast, J., Musti, A.M., Wessely, M., Ansorge, W., Seraphin, B., Wilm, M., Valdez, B.C., Bohmann, D., 2002. The DEXD/H-box RNA helicase RHII/Gu is a co-factor for c-Jun-activated transcription. Embo. J. 21, 451–460. Wilm, M., Mann, M., 1996. Analytical properties of the nanoelectrospray ion source. Anal. Chem. 68, 1–8. Wilm, M., Neubauer, G., Mann, M., 1996a. Parent ion scans of unseparated peptide mixtures. Anal. Chem. 68, 527–533. Wilm, M., Shevchenko, A., Houthaeve, T., Breit, S., Schweigerer, L., Fotsis, T., Mann, M., 1996b. Femtomole sequencing of proteins from polyacrylamide gels by nano-electrospray mass spectrometry. Nature 379, 466–469. Wollnik, H., 1993. TOF-MS. Mass. Spectrom. Rev. 12, 89–114. Wolschin, F., Wienkoop, S., Weckwerth, W., 2005. Enrichment of phosphorylated proteins and peptides from complex mixtures using metal oxide/hydroxide affinity chromatography (MOAC). Proteomics 5, 4389–4397. Wu, J.T., Qian, M.G., Li, M.X., Zheng, K., Huang, P., Lubman, D.M., 1998. On-line analysis by capillary separations interfaced to

PART  |  11  Cardiomics

an ion trap storage/reflectron time-of-flight mass spectrometer. J. Chromatogr. A. 794, 377–389. Yan, J.X., Packer, N.H., Gooley, A.A., Williams, K.L., 1998. Protein phosphorylation: technologies for the identification of phosphoamino acids. J. Chromatogr. A. 808, 23–41. Yang, C.C., Ornatsky, O.I., McDermott, J.C., Cruz, T.F., Prody, C.A., 1998. Interaction of myocyte enhancer factor 2 (MEF2) with a mitogenactivated protein kinase, ERK5/BMK1. Nucleic. Acids Res. 26, 4771–4777. Yang, L., Amad, M., Winnik, W.M., Schoen, A.E., Schweingruber, H., Mylchreest, I., Rudewicz, P.J., 2002. Investigation of an enhanced resolution triple quadrupole mass spectrometer for high-throughput liquid chromatography/tandem mass spectrometry assays. Rapid. Commun. Mass. Spectrom. 16, 2060–2066. Yates 3rd, J.R., 1998. Mass spectrometry and the age of the proteome. J. Mass. Spectrom. 33, 1–19. Yates 3rd, J.R., Speicher, S., Griffin, P.R., Hunkapiller, T., 1993. Peptide mass maps: a highly informative approach to protein identification. Anal. Biochem. 214, 397–408. Youn, H.D., Sun, L., Prywes, R., Liu, J.O., 1999. Apoptosis of T cells mediated by Ca2-induced release of the transcription factor MEF2. Science 286, 790–793. Yun, K., Wold, B., 1996. Skeletal muscle determination and differentiation: story of a core regulatory network and its context. Curr. Opin. Cell Biol. 8, 877–889. Zhang, F., Bartels, M.J., Stott, W.T., 2004. Quantitation of human glutathione S-transferases in complex matrices by liquid chromatography/ tandem mass spectrometry with signature peptides. Rapid. Commun. Mass. Spectrom. 18, 491–498. Zhang, Y., Wolf-Yadlin, A., Ross, P.L., Pappin, D.J., Rush, J., Lauffenburger, D.A., White, F.M., 2005. Time-resolved mass spectrometry of tyrosine phosphorylation sites in the epidermal growth factor receptor signaling network reveals dynamic modules. Mol. Cell Proteomics 4, 1240–1250. Zieske, L.R., 2006. A perspective on the use of iTRAQ reagent technology for protein complex and profiling studies. J. Exp. Bot. 57, 1501–1508.

Chapter 12.1

Evolution of Regeneration Jonathan M.W. Slack Stem Cell Institute, University of Minnesota, McGuire Translational Research Facility, Minneapolis, MN, USA

I.  Introduction The regeneration of missing parts is one of the most remarkable of biological phenomena. It remains one that embodies a number of genuinely unsolved questions for each of which there is no real understanding of the visible events in terms of genes and molecules. The first of these is the nature of the permissive conditions for regeneration. Why do some structures regenerate, while other apparently similar structures in related species do not? The second is the problem of cell lineage: to what extent does regeneration occur from reserve cells or stem cells; and to what extent from dedifferentiation and reprograming of existing tissue cells? The third is the issue of how a complex pattern of structures can be regenerated. For example, both the stump and the regenerate of a salamander limb contains muscles and cartilages, but the pattern of muscles and cartilages that regenerates in the new hand is quite different from that which was present at the cut surface. There are some partial answers to these problems, and in due course they will doubtless be solved at a molecular level by experimental biology. In the meantime, a useful perspective can be obtained by considering the evolution of regeneration, or at least by comparing what is known about regenerative abilities across the animal kingdom. There is also a basic philosophical issue at stake when we consider regeneration, and that is to decide whether it is a “pristine” or an “adaptive” quality (Goss, 1992). In other words, can any living tissue or structure regenerate so long as we are able to unlock some specific inhibitory mechanisms (pristine)? Or, can nothing regenerate unless it has a specific evolutionary adaptation to do so (adaptive)? It is quite important to be able to take a stance on this issue in order to decide how to approach a practical issue such as “what can we do to provoke regeneration in the human heart?” With regard to this, it is possible to obtain useful Heart Development and Regeneration Copyright © 2010 Elsevier Inc. All rights of reproduction in any form reserved.

insights from comparative zoology. These studies appear to suggest that many aspects of regeneration are pristine, but also that this ability is easily lost. In turn, this suggests that the retention of regenerative ability may incur an evolutionary cost of some sort.

II.  Phylogeny of animals As a background to this discussion, it is useful to consider the elements and categories which constitute the animal kingdom. There are generally considered to be about 34 phyla, representing groups of animals with a distinct body plan (Ruppert et al., 2004). However, the precise number of phyla varies, depending on the views of individual taxonomists and authors. Figure 1 shows a modern consensus phylogenetic tree for several of the most important phyla. Due to the fact that different phyla have different body plans, there are very few morphological characters in common between the adult animals of different phyla. For this reason, information from embryos and larvae has traditionally been used to attempt to draw up the phylogenetic tree for the animal kingdom as a whole. The first of these characters is the number of germ layers established during embryonic development. Most animals have formed three tissue layers by the end of gastrulation: the ectoderm; mesoderm; and endoderm. They are called triploblasts. Only three phyla are not triploblastic; the Porifera (sponges), which have no well-defined tissue layers at all, the Cnidaria (Hydra, jellyfish, sea anemones) and the Ctenophora (comb jellies), which both have two layers. The Cnidaria and Ctenophora are called diplo­ blasts. Animals of these phyla also, in the main, show radial symmetry, whereas most of the triploblastic animals show bilateral symmetry and are therefore also known as the Bilateria. However, both the diploblasts and the radial 827

828

PART  |  12  The Regenerative Heart

symmetry of the Cnidarians have recently been called into question (Finnerty et al., 2004). Bilateral animals always have some kind of head and tail, and in conformity with zoological usage the head end is here referred to as the “anterior” and the tail end as the “posterior”. Note that in human anatomy the convention is different, with “anterior” corresponding to the zoological “ventral” and “posterior” to the zoological “dorsal”. The next important feature is the presence or otherwise of a coelom. This is the cavity formed within the mesoderm, lined with a mesoderm-derived epithelium, the peritoneum, which often constitutes the principal body cavity. Platyhelminthes (flatworms) do not have a coelom, and are described as acoelomate. Many other invertebrate phyla, including the Nematoda, have a body cavity that is only partly surrounded by mesoderm, and are described as pseudocoelomate. Annelida (segmented worms), Mollusca, Nemertea (ribbon worms), Arthropoda, Echinodermata (sea urchins, starfish, etc.) and Chordata (including vertebrates) all have a coelom. The coelomate phyla were traditionally divided into two “super-phyla” called the Protostomia and the Deuterostomia. The latter group contains the Echinodermata and Chordata. The deuterostomes were defined by radial cleavage, formation of the anus from the blastopore, and enterocoely, which is formation of the coelom by budding of the mesodermal rudiment from the gut. In fact, true enterocoely is not found in vertebrates, but it is found in some protochordates and echinoderms and thus serves to link these two phyla together. The defining features of protostomes were spiral cleavage, early-acting cytoplasmic determinants and schizocoely

Cnidaria

Platyhelminthes

Bryozoa

Nemertea

(formation of the coelom by splitting of the mesodermal layer). Older textbooks may also include formation of the mouth from the blastopore, but this is incorrect. In protostomes the blastopore typically narrows to a ventral slit, from which both mouth and anus derive. Recently, the use of molecular taxonomy has changed this picture. The deuterostomes remain as a taxon, but the significance of the coelom has disappeared and there are now two new super-phyla making up the protostomes (Aguinaldo and Lake, 1998). The annelids and molluscs, which share a type of larva called the trochophore, are grouped with some other phyla in the Lophotrochozoa, while the arthropods and nematodes, together with all other molting animals, comprise the Ecdysozoa. From the perspective of this book, it should be noted that many, but not all, types of animals have a circulatory system or heart. There is no heart or blood vascular system in the Cnidarians, Platyhelminthes or Bryozoa (moss animals). Nemertea have a circulatory system with contractile vessels. Annelids have a segmentally-reiterated circulatory system with some of the dorsal vessels tending to be contractile and playing the role of a heart. Echinoderms have a centrally-located heart and radially-symmetrical haemal system, as well as a separate water vascular system. The tadpoles and zooids of ascidians, which are a type of invertebrate chordate, have hearts and circulatory systems. Amphioxus, a protochordate resembling the putative ancestor of vertebrates, has a circulatory system similar to vertebrates, but with no actual heart, the ventral aorta being the contractile organ. Although many of these groups have not been closely studied from a molecular genetic standpoint,

Annelida

Nematoda

Lophotrochozoa

Arthropoda

Echinodermata

Chordata

Ecdysozoa

Protostomia

Deuterostomia

Bilateria

Ancestral Metazoan Figure 1  Phylogenetic tree. This diagram indicates a generally-accepted view of the evolutionary relationship of the animal phyla mentioned in the text. Hydra belongs to the Cnidaria; polychaetes, oligochaetes and leeches belong to the Annelida; molluscs belong to the Lophotrochozoa; ascidians belong to the Chordata. Ctenophores (comb jellies) are a sister phylum of the Cnidaria, and Poriferans (sponges) are an outgroup connected to the ancestral metazoan.

829

Chapter  |  12.1  Evolution of Regeneration

the insects (mostly Drosophila) and the vertebrates (mouse, Xenopus, zebrafish) have been studied in great detail. In view of the large evolutionary gap between vertebrates and insects, it is likely that any elements they have in common will be found in most or all other animals. Most importantly, these common features include the use of the same cardiovascular transcription factors to drive heart development, most notably Nkx2.5/tinman (Chapter 9.1). This is despite the great anatomical difference between insect and vertebrate hearts; insects have an open body cavity filled with hemolymph and a dorsal contractile structure, while vertebrates have a two-, threeor four-chambered ventrally-located heart and circulatory system composed of arteries and veins (Tonissen et al., 1994; see Chapters 1.1, 1.2, 1.3, 1.4 and 1.5). Transcription factors simply act as switches regulating the expression of other genes, and so when an evolutionary novelty arises there is no particular reason to use one transcription factor rather than another. The use of the same transcription factor by different phyla to make a similar structure therefore suggests a relationship of descent or, in other words, homology, in the program for making that structure. The

complex issue of the relationship between homologous genetic pathways and the homology or otherwise of final anatomical structures is discussed in Bolker and Raff (1996).

II.A.  Distribution of Regenerative Ability It is often believed that regenerative ability is extensive in “lower” animals, and becomes lost in “higher” ones, but this is only partly true. More important than the grade of the organism is the nature of its cell renewal program. For example, nematodes generally do not regenerate at all, while some of the similarly complex annelids can regenerate all parts of the body from a small fragment. Nematodes do their growing during the embryonic and larval periods, and in the adult the cell number is fixed and determinate. By contrast, those annelids that can regenerate extensively also practice asexual reproduction by fission, so no new mechanism needs to be created to enable regeneration. The main types of regeneration are bidirectional, monodirectional and hyperplastic (Slack, 1980) (Fig. 2). Bidirectional

(A)

(B)

(C)

(D)

Figure 2  Types of regeneration: (A) bidirectional; (B,C) monodirectional, the lower part shows the celebrated experiment of (Butler, 1955); (D) hyperplastic. Regenerated parts are shown in orange.

830

means that a severed organism can regenerate in both directions, usually a head from an anterior-facing cut surface and a tail from a posterior-facing cut surface. Monodirectional means that an identical or at least similar distal structure will regenerate from either side of a cut surface. Monodirectional regeneration relates mostly to appendages such as arthropod and vertebrate limbs, which always regenerate distally, even if a proximal facing cut surface is engineered by experimenters (Butler, 1955). Hyperplastic regeneration refers to tissues such as the mammalian liver (Fausto, 2000), where the microstructure and overall size is regenerated, but the gross anatomical form is not. A distinction is also made between so-called epimorphic and morphallactic regeneration. Epimorphic regeneration refers to the type of regeneration in which a bud of dividing cells is formed and grows to produce the regenerated structure. This is the case for most instances of bidirectional and monodirectional regeneration. Morphallactic regeneration refers to the type of regeneration which is achieved by rearrangement of existing cells and is effectively confined to hydroids (Holstein et al., 2003). Bidirectional regeneration is truly dramatic, as it involves the ability to regrow the main body axis following a severing of the whole body (Berrill, 1951; Holstein et al., 2003). Recently, this process has been extensively studied in the planarian worms (Platyhelminthes) (Reddien and Alvarado, 2004), but it is actually shown most profoundly by many nemertean worms, which can grow a whole new body from a small fragment (Coe, 1934). Of more interest in the present context are the annelid worms, because of their possession of a circulatory system with contractile vessels, and they are discussed further below. Also showing bidirectional regeneration are most of the colonial type of animals, where the organism as a whole consists of a ramifying or branching colony spread out over a substratum, such as a submerged rock. Typically the colonies consist of a number of bodies called zooids, which are sea anemone-like in appearance, joined together by a stolon, which is a tube running across the substratum. The individual zooids have a defined and reproducible anatomy, characteristic of the species in question, but the colony as a whole may vary considerably in shape and size. Colonial animals include many hydrozoa, bryozoa and some ascidians. These are of phylogenetically very different grades. The hydrozoa are a class of the Cnidarians, conventionally regarded as diploblastic. The bryozoa are coelomates. The colonial ascidians are chordates, and so share some characteristics with the vertebrates. Most of these colonial animals will regenerate zooids from stolons and stolons from zooids, thus meeting the definition of bidirectionality. Among them, only the ascidians have hearts. In bidirectional regeneration the same cut surface can regenerate either a head or a tail, depending on its orientation. This means that there is an essential “polarity control” that makes a decision about whether the regenerate is

PART  |  12  The Regenerative Heart

to be a head or a tail. This must operate before the process of new part formation itself begins. Whole body bidirectional regeneration is usually associated with the type of asexual reproduction in which the body can spontaneously fragment into two or more parts, each of which develops to reconstitute a new individual. The requirements of asexual reproduction by fission and of bidirectional regeneration are obviously very similar. Bidirectional regeneration is also associated with a continuous turnover of cells in which there is a continuous flux from undifferentiated “stem cells” to the principal body parts. This is found in both planarians and hydroids, and probably also in some annelids. In planaria there is a considerable literature on the small undifferentiated cells called neoblasts that are responsible for continuous renewal of the whole body (Salo and Baguna, 2002; Reddien et al., 2005b). Neoblasts are the only cells that normally divide, and they are considered to comprise, or include, the stem cells that produce the 12–15 histologically-distinguishable cell types making up the tissues of the worm. In the steady state, a worm contains about 20% neoblasts and 80% differentiated cells. It is not known whether the neoblast population contains subpopulations committed to particular fates, or whether there is a small stem cell pool feeding a large transit amplifying pool, as is generally the case in higher animals. Following transection of a planarian there is an accumulation of undifferentiated cells under the wound epithelium which makes up the regeneration blastema. Blastema is a general term for a regeneration bud containing undifferentiated proliferating cells, although in planarians the term confusingly refers to an unpigmented distal region in which there is no cell division. The cells populating the planarian blastema come from the dividing neoblasts in the proximal region. The evidence that neoblasts give rise to the regenerate comes from two sources. First, BrdU-label­ ing of neoblasts before amputation leads to a blastema containing many BrdU-labeled cells. Second, the ability of worms to regenerate is destroyed by X-irradiation, which is followed by a rapid disappearance of neoblasts. After such a radiation dose, there is no further production of new cells and the worms will die a few weeks later. Monodirectional regeneration usually involves the replacement of amputated appendages such as legs, antennae or tails. It is displayed by many species of starfish and brittle star (Echinodermata) that can regenerate missing rays (Thorndyke et al., 2001) which, because of the radial symmetry of these animals, might be considered to be the main body. A few types of echinoderm, such as the starfish Linckia, can actually regenerate in a bidirectional manner, forming a whole body from an isolated ray (Kellogg, 1904). Regeneration of external appendages in insects is confined to the larval stages of Hemimetabola, those families which show a gradual progression to adulthood through a series of larval forms. If a leg or antenna is removed from a locust or cockroach, it will grow again

831

Chapter  |  12.1  Evolution of Regeneration

and will become visible at the next moult when the old cuticle is shed (French, 1982; Campbell and Tomlinson, 1995). It is not found in the Holometabolous insects which are those, like Drosophila, showing an abrupt metamorphosis during a pupal phase. Accordingly, Drosophila does not show any regeneration of adult structures, although the imaginal discs of the larva can regenerate if they are damaged prior to metamorphosis (Bryant, 1975). Regeneration in vertebrates is displayed mainly by the amphibians. Urodeles (newts and salamanders) are often able to regenerate limbs, tails and jaws, and many other structures, both before and after metamorphosis (Tsonis, 2000). Anurans (frogs and toads) can usually do so in the tadpole stage, but lose the ability at metamorphosis (Slack et al., 2004). Lizards are well-known for their ability to regenerate tails, but the new tail does not include all of the tissues and structures of the original (Simpson, 1970).

II.B.  Technical Issues This quick survey of regeneration shows that one of the problems for research is that extensive regenerative abilities are found mainly in animals that are not the familiar laboratory model species for developmental biology. This means that there is often a paucity of pre-existing collections of cloned genes, methods for introducing or removing genes, or methods for cell labeling or transplantation. As a result, new experimental approaches have often needed to be introduced for individual experiments, greatly reducing the rate of progress. However, this problem is now being overcome. The three major hurdles to applying molecular techniques to a new organism are: first, obtaining a reasonable inventory of genes to work with; second, being able to perform overexpression experiments by introducing specific genes into the organism; and third, being able to ablate gene activity selectively and specifically. It has now become fairly standard for projects with nonmodel organisms to commence with an expressed sequence tag (EST) project, whereby a large number of cDNAs are sequenced, cataloged and gridded. The labor required to do this has been vastly reduced by modern technology, and complete genome sequences are beginning to emerge for the more popular organisms, such as the planarian Schmidtea mediterranea. As far as gene introduction is concerned, there are two common solutions. One is electroporation, which can achieve transient introduction of DNA into a wide variety of specimens. The other is the use of pseudotyped retroviruses, which have a very broad host range and can also achieve integration of DNA into the host genome (Kumar et al., 2000). Specific ablation of gene activity has been achieved through the use of antisense oligonucleo­tides (usually the nuclease-resistant morpholino analogs) to destroy specific mRNA or prevent its translation; or by the overexpression of dominant-negative inhibitors of protein

function. In order to use such methods for regeneration research, a method of introduction into the cells of the regenerating structure, such as electroporation, is required. In recent years, RNA interference, in which specific mRNAs are destroyed by introduction of complementary sequences of short double-stranded RNA, has also become a serious contender as a generic method, particularly for invertebrates. An example of the assembly of a toolkit of methods is provided by the recent comprehensive studies of planarians (Alvarado et al., 2002; Reddien et al., 2005a).

II.C.  Annelid Regeneration Annelid regeneration deserves to be better understood, because annelids are the most complex animals capable of extensive bidirectional regeneration. Furthermore, annelids are among the simplest organisms possessing a circulatory system with contractile elements. From the point of view of the present book, the regeneration of the anterior end, or “head”, involves regeneration of the contractile vessels that constitute the heart. Annelids include three quite familiar groups of animals. The polychaetes comprise mostly marine worms which are sometimes free living and sometimes confined to burrows. The oligochaetes are mostly terrestrial and include the familiar earthworms. The leeches are well-known in their historic capacity as a method for removing blood from patients, and have also functioned on a modest scale as model organisms for developmental biology research (Weisblat and Shankland, 1985). The leeches were actually selected as model organisms because of their relatively stereotyped cell lineage and invariant segment number, but this is unfortunately associated with a complete inability to regenerate. On the other hand, some types of polychaetes and oligochaetes regenerate to a significant degree (Fig. 3). Many species of polychaete and oligochaete will regenerate posterior body regions (“tails”). This is associated with the fact that segment formation often continues for long periods at the posterior end, with a segment-forming growth-zone located between the last formed segment and a terminal structure called the pygidium (Kumé and Dan, 1957). More spectacular is the regeneration of anterior ends (“heads”). This form of regeneration is the exception rather than the rule, and tends to be associated with a normal mode of asexual reproduction of the worm by transverse fission. A recent study of an oligochaete called Enchytraeus japonensis exemplifies the main features that have been found in many other species (Myohara, 2004). This worm reproduces asexually by spontaneous fission into a number of pieces. Regeneration behavior is similar following spontaneous fragmentation, or cutting the worm into pieces. Posterior regeneration of each piece involves formation of a terminal pygidium with a growth zone, and the progressive formation of posterior segments from the

832

PART  |  12  The Regenerative Heart

Head

Pygidium

Figure 3  Annelid regeneration. In fully-bidirectional annelids the posterior regenerate forms a growth zone between the pygidium and stump, the anterior regenerate forms a growth zone between the regenerate and the stump, so in both cases new segments are formed from anterior to posterior.

growth zone. At the anterior end, each fragment regenerates just seven segments, corresponding to the most anter­ ior seven segments of the worm. The adjoining segments then undergo a metamorphic process to adjust the composition of the internal organs to the appropriate anteroposter­ ior level. Normally anterior-facing cut surfaces regenerate the head end, and posterior-facing cut surfaces regenerate the tail end, although it is also possible to obtain bipolar forms with a head at both ends when short (less than seven segment) pieces are cut from the anterior region of the body, or when longer fragments are placed in water instead of the normal agar plates. Recently, use has been made of long-range evolutionary homology in order to identify a population of pluripotent cells in Enchytraeus japonensis. A transcription factor called piwi is found across both animals and plants in primordial germ cells. In Enchytraeus, piwi is found in the germ cells of the gonads which are located in segments 7 and 8. Following anterior regeneration, a new gonad arises in the new segments 7 and 8, and becomes populated by piwi-positive germ cells. These are derived from a population of piwi-positive cells that are scattered along the length of the worm (Tadokoro et al., 2006). It will be most interesting to ascertain whether this population of cells corresponds to the neoblasts that were formerly identified by morphology in other species of annelid (Zhinkin, 1936; Stéfan-Dubois, 1954), and whether these cells form somatic tissues in addition to new germ cells. Serious unsolved questions exist regarding annelid regeneration. The most striking is that of polarity: how can the same cut surface form either a head or a tail? Tissue

that forms a head in one fragment would form a tail if the cut was made one segment more anteriorly. This problem is quite unsolved at the present time. The next problem is that of the cellular origin of the regenerate. In the planari­ ans, it is well-accepted that the source of the regenerate is the neoblasts. Neoblast-like cells have been described many times in annelids. They are thought to be associated with the main ventral nerve cord, and to migrate along it to the cut surface (Zhinkin, 1936). However, in a similar way to the amphibian limb, it is also considered that there is dedifferentiation of tissues to form a blastema.

II.D.  Urodele Limb Regeneration This subject has been extensively reviewed (Tsonis, 2000; Nye et al., 2003; Tanaka, 2003; Brockes and Kumar, 2005) and will just briefly be considered here for comparative purposes. After amputation of a limb, the stump rapidly forms a wound epithelium by migration of epidermal cells over the cut surface. Then the internal tissues dedifferentiate to a depth of about 1 mm from the cut. The cells from this region form a blastema consisting of loose-packed mesenchymal-type cells surrounded by a thick epidermal jacket. The blastema proliferates for a while, and then the structures of the limb redifferentiate in a proximal– distal sequence. The regenerate is completed in miniature, and then undergoes a long period of growth to retain its original size. The time required to regenerate the pattern in miniature varies with the age and size of the animal, but is typically 2–3 weeks for small larvae and a few months for large adults.

833

Chapter  |  12.1  Evolution of Regeneration

The source of cells for regeneration has been studied in some detail (Brockes and Kumar, 2002). The cells are known to originate near the amputation surface. This may be shown by X-irradiation, which can completely suppress regeneration. If the level of the future amputation site is irradiated, there is no regeneration, while if the region of the future amputation is protected, but neighboring tissues are irradiated, regeneration proceeds normally. This indicates that the blastema must arise from cells located no further than a few millimeters from the amputation surface, rather than from stem cells in the bone marrow or other distant sites. In terms of whether the new cells arise by dedifferentiation, this has been clearly demonstrated at least for mature myofibers. If newt myofibers are labeled by introduction of a marker gene and grafted into a limb which is then amputated, they will break up into mononuclear cells. These can still be identified by the expression of the marker gene, and they can be shown by BrdU-labeling to undergo cell division (Kumar et al., 2004). Metaplasia has been studied in regeneration using triploidy as a cell marker. If a graft is made of a pure tissue from a triploid to a diploid, allowed to heal, and the limb is amputated through the graft, then it can be asked which cell types in the regenerate carry the triploid marker (Namenwirth, 1974). Such experiments show that a graft of labeled epidermis can only generate more epidermis, and not internal tissues. By contrast, a graft of labeled cartilage can populate not only cartilage, but also all the connective tissues; dermis, tendons, ligaments and fibrous capsules. Furthermore, dermis can readily convert to cartilage. A graft of labeled myofibers will label predominantly muscle in the regenerate, with a few percent of other cell types, including cartilage (Kumar et al., 2000). Thus, these experiments suggest that there are probably three tissue lineages; epidermal, muscle and connective tissue, although extensive metaplasia can occur between the different types of connective tissue. Limb regeneration is absolutely dependent on a nerve supply to the limb. In animals in which the nerve supply has been transected, dedifferentiation occurs as normal, resulting in the formation of a blastema, but this blastema fails to grow and regeneration is aborted. The function of the nerves is to release mitogenic factors, often referred to as the neurotrophic factor(s). These probably include FGFs, neuregulins and transferrin (Mescher et al., 1997). The pattern of the regenerate conforms to the stump, so that all parts distal to the cut surface are replaced. The results of many experiments on regeneration suggest that the tissues of the limb carry cryptic codes for regional identity, often called positional values, which specify which structures shall be formed on differentiation. Respecification of positional value can be achieved by treating cone stage blastemas with retinoic acid (Maden, 1982). When an amputated, but otherwise anatomically normal limb is treated, the result is proximalization, such that the further

development of the blastema leads to serial duplication of structures. So, for example, a wrist-level blastema treated with retinoic acid can regenerate a complete arm. Since in these experiments the whole animal is immersed in a solution of retinoic acid, the blastema is probably exposed to a uniform dose of the reagent during treatment. Other effects can be demonstrated on the transverse axes of anatomically abnormal limbs, leading to posteriorization and ventralization (Bryant and Gardiner, 1992). One possible molecular component of positional value is the newt homolog of CD59, an inhibitor of complement activation (da Silva et al., 2002). This is more abundant in the proximal than the distal blastema, and is induced by retinoic acid. It is a cell-surface protein with a GPI anchor. If CD59 gene is introduced by electroporation into cells of an axolotl blastema, the cells containing them will move to more proximal positions than control cells, indicating a more proximal identity (Echeverri and Tanaka, 2005). As the limb is asymmetrical in three dimensions, it should in principle require three sets of codes to specify the pattern of differentiation in the three anatomical axes; proximodistal, anteroposterior and dorsoventral. Sonic hedgehog, known to be important in specifying anteroposterior pattern in the developing limb, is expressed in the posterior part of the blastema, and introduction of Shh into the anter­ ior side of a blastema using vaccinia virus will produce a double posterior duplication (Roy et al., 2000). But nothing is yet known of the targets of Shh, or of the control of the dorsoventral pattern.

II.E.  Mammalian Hyperplastic Regeneration With the exception of antler regeneration in deer (Price et al., 2005), there is no process in mammals that is truly comparable to the epimorphic regeneration found in lower vertebrates and invertebrates. However, there is a certain capacity to regenerate internal organs, and the mechanisms of this are of obvious interest when considering the prospects for repair and regeneration of the heart. Many tissues are arranged as structural-proliferative units (SPUs), in which the structural elements of the tissue, such as intestinal crypts, kidney nephrons or thyroid follicles, are also units of cellular renewal (Potten, 1978; Slack, 2000). When a part of a tissue is removed, the issue arises about whether there is any compensatory growth and, if so, whether it takes the form of an increase in cell size, an increase in cell number, or an increase in the number of structural-proliferative units (Fig. 4).

II.E.i.  Liver It is well-known that the mammalian liver will regenerate to its original size after a portion is removed. Liver has a well-organized histoarchitecture, with plates of hepatocytes extending from a periportal location where a small bile

834

PART  |  12  The Regenerative Heart

Regeneration of SPUs

Cellular hyperplasia

Cellular hypertrophy

Figure 4  Regenerative responses of mammalian tissues.

duct runs alongside a branch of the hepatic artery and the hepatic portal vein (the portal triad), to the perivenous location, adjacent to a branch of the hepatic vein. Various metabolic activities are localized or graded along the periportal–perivenous axis, although this does not seem to correspond to a pathway of cell maturation, as was once thought. Following the removal of significant hepatic tissue or a liver lobe, remaining hepatocytes start to proliferate within 24 hours. Other cell types (biliary, Kupffer and endothelial) commence proliferation somewhat later, so that all the liver cell populations eventually increase in number in a coordinated manner. Locally, the entry of hepatocytes into the cell-cycle is controlled by various cytokines, including hepatocyte growth factor (HGF), interleukin-6 (IL-6) and serotonin (5-HT) (Michalopoulos, 1997; Taub, 2004; Lesurtel et al., 2006). It has long been known that there are also circulating factors controlling the size of the liver, as removal of part of the liver from one of a pair of parabiosed animals will lead to additional growth in the intact liver (Moolten, 1967). It now seems likely that this includes bile acids that are released into the blood from the residual damaged liver (Huang et al., 2006). This regulatory feedback system is probably also operative in the situation of liver transplantation, when a small liver is grafted into a large host. In this scenario, the

graft grows rapidly until it achieves the size of the liver that was removed (Kam et al., 1987). There has been some debate about whether normal hepatic cell turnover involves stem cells, but the verdict from cell-labeling experiments is that it does not (Bralet et al., 1994; Kennedy et al., 1995). These experiments indicate that normal growth, cell turnover and regeneration arise from hepatocytes that may be at any position along the perivenous–periportal axis. However, if regeneration from hepatocytes is blocked, for example by dosing with the carcinogen acetyl aminofluorene, then the liver can regenerate from a stem cell population called the oval cells (Alison et al., 1997). Oval cells arise from the small bile ducts, and resemble the embryonic hepatoblasts that produce both hepatocytes and biliary epithelium (Shiojiri et al., 1991). Although overall hepatic size is unchanged, it appears that the lobules of the regenerated liver are larger than those of the original, in other words, the average distance from portal triad to central vein increases. This may not be true in the vicinity of the wound where remodeling may occur. An interesting question, as yet unanswered, is whether lobule size would continue to increase without limit in the event of repeated resection, and if so what effect this would have on the physiological function of the liver.

835

Chapter  |  12.1  Evolution of Regeneration

II.E.ii.  Other Organs According to Sidorova (1978), the regenerative behavior of mammalian organs differs, depending on the developmental stage at which the formation of structural-proliferative units (SPUs) normally concludes. For example, in rats the formation of new nephrons normally ceases 7–10 days after birth and of new pancreatic acini 3–4 weeks after birth. In general, the regenerative response to removal of part of an organ, or the removal of one of a pair of organs, involves similar processes to those already taking place. Therefore, if new SPUs are still being formed then this may occur in regeneration; if new cells but not SPUs are being formed then there may be hyperplasia leading to an increase in the size of the SPUs. If cell division has ceased, then regenerative behavior may be confined to an increase in cell size. In any event, the rate of growth is usually slightly increased, so that there is at least a partial catchup of overall size by the regenerating organ or part-organ compared to controls. It is well-known that if one kidney is removed the remaining kidney will increase in size. This normally involves an increase in both cell number and cell size, such that the nephrons become larger. In very young animals, in which nephrogenesis is still continuing, there is also an increase in the number of nephrons (Canter and Goss, 1975). There is a difference between the normal and hyperplastic state, in that the hyperplastic kidney can shrink when the metabolic demand decreases, and this may be demonstrated by transplantation. If one kidney is removed from a rat so that the other kidney is caused to enlarge, and then a second hyperplastic kidney is grafted into the site of the removed one, both hyperplastic kidneys will shrink towards normal size. On the other hand, grafting of a third normal kidney into an animal with two normal kidneys does not result in any shrinkage of donor or host organs (Silber, 1974). Therefore, the hyperplastic condition seems to be reversible, presumably in some way relating to the functional demand placed by the body on the organ, whereas the normal organ size is fixed. A situation of functional recovery with little or no increase in cell or SPU number is provided by the thyroid. Surgical ablation of half or three quarters of the thyroid is followed by an increase in follicle cell size and a recovery of thyroid hormone output within 25 days, however, the size of the organ as a whole increases only slightly (Logothetopoulos and Doniach, 1955).

II.F.  Cardiac Regeneration in Vertebrates Lower vertebrates have long been known to be able to regenerate parts of the heart. In amphibians and fish the ventricle is undivided, and the animal can survive removal of the ventricular apex so long as the cavity is not punctured (see Chapter 12.2). It was shown by Oberpriller (1974)

that 1/8 of the ventricle could be removed and would grow back. DNA synthesis was detected in the cardiomyocytes in the vicinity of the wound. Recently, the zebrafish has been established as a cardiac regeneration model (Poss et al., 2002, 2003; Raya et al., 2003; Chapter 12.2). Here a 20% ventricular resection will grow back in two months, with replacement of both cardiomyocytes and coronary vasculature. Initially the wound is closed by a clot, which is replaced by fibrin over 2–3 days. Cardiomyocytes in the vicinity enter S-phase. The notch pathway is activated in neighboring cells, along with transcription factors associated with heart development including Nkx2.5, Hand2 and Tbx20. More recent studies (Lepilina et al., 2006) indicate that the regeneration is not simply fed by the re-entry of cardiomyocytes into the cell-cycle. Use of two reporters for activity of the cardiac myosin light chain promoter, differing in time of expression and stability, indicates that some new cardiomyocytes are derived from noncardiomyocytes, presumably undifferentiated progenitor cells. The same work indicated an increase of FGF17 in the wounded area of myocardium, with FGF receptors in the epicardium. Induced expression of a transgene-encoding dominant-negative FGF receptor will suppress regeneration due to inhibition of neovascularization, leading to formation of a collagenous scar.

III.  Conclusions Overall, the main lesson from an evolutionary survey is that regeneration behavior generally corresponds to what the organism is currently doing in the normal course of its lifecycle. Animals that normally reproduce by fission have mechanisms enabling whole body regeneration, including heads. Hydra regenerate by morphallaxis, because they are constantly remodeling tissue into heads, tentacles and feet using a supply of cells from the midregion. The planarians regenerate by means of neoblasts, because the whole body is rapidly turning over, fed from this population of stem cells. Even a detailed study of the Xenopus tadpole tail indicates that regeneration embraces current ongoing processes, such as the addition of cells to the notochord or spinal cord. However, Xenopus tadpole tail regeneration does not recapitulate embryonic events that have terminated, such as the somite oscillator or the induction of the neural crest (Slack et al., 2004; Chen et al., 2006; Lin et al., 2007). The chief exception to this rule is perhaps the celebrated case of the urodele limb. In urodeles with significant regenerative capacity such as the axolotl, there probably is a continuous turnover of cells in each of the constituent tissue types in the limb. However, there is no ongoing production of distal pattern, or an ongoing situation of dedifferentiation. Therefore, the formation of the regeneration blastema, and the formation of a pattern of

836

structures by this growing blastema, appears to be a genuinely new behavior, although it probably employs the same molecular machinery as was used for embryonic development (e.g., Torok et al., 1999). It is perhaps ironic that this most highly-studied example of regeneration should prove to be, to some extent, an exception to the general rule. In terms of the prospects for heart regeneration the lessons are clear; if, as was once thought, the heart is entirely postmitotic and the cells of the young animal persist for life, then the prospects for regeneration will be poor. However if, as now seems likely, there is a population of cardiac progenitor cells that slowly replace damaged cardiomyocytes, then it will probably be possible to augment heart regeneration via their stimulation, culture, expansion and/or transplantation. All the signs indicate that the next few years will see decisive progress towards this objective. Since completion of this chapter two important relevant works have appeared. Firstly, the “highest” organism that is known to regenerate its head, including the heart, is now the hemichordate worm Ptychodera flava (Rychel and Swalla, 2008). Secondly, studies of 14C content have shown a small but measurable degree of cell turnover in cardiomyocytes of the human heart (Bergmann et al., 2009).

References Aguinaldo, A.M., Lake, J.A., 1998. Evolution of the multicellular animals. Am. Zool. 38, 878–887. Alison, M., Golding, M., El Nasir, L., Nagy, P., Thorgeirsson, S., Sarraf, C., 1997. Wholesale hepatocytic differentiation in the rat from ductular oval cells, the progeny of biliary stem cells. J. Hepatol. 26, 343–352. Sánchez Alvarado, A., Newmark, P.A., Robb, S.M., Juste, R., 2002. The Schmidtea mediterranea database as a molecular resource for studying platyhelminthes, stem cells and regeneration. Development 129, 5659–5665. Bergmann, O., et al., 2009. Evidence for Cardiomyocyte Renewal in Humans. Science 324, 98–102. Berrill, N.J., 1951. Regeneration and budding in worms. Biol. Rev. 27, 401–438. Bolker, J.A., Raff, R.A., 1996. Developmental genetics and traditional homology. Bioessays 18, 489–494. Bralet, M.P., Branchereau, S., Brechot, C., Ferry, N., 1994. Cell lineage study in the liver using retroviral-mediated gene-transfer – evidence against the streaming of hepatocytes in normal liver. Am. J. Pathol. 144, 896–905. Brockes, J.P., Kumar, A., 2002. Plasticity and reprogramming of differentiated cells in amphibian regeneration. Nat. Rev. Mol. Cell Biol. 3, 566–574. Brockes, J.P., Kumar, A., 2005. Appendage regeneration in adult vertebrates and implications for regenerative medicine. Science 310, 1919–1923. Bryant, P.J., 1975. Pattern formation in imaginal wing disk of Drosophila Melanogaster – fate map, regeneration and duplication. J. Exp. Zool. 193, 49–77. Bryant, S.V., Gardiner, D.M., 1992. Retinoic acid, local cell–cell interactions, and pattern formation in vertebrate limbs. Dev. Biol. 152, 1–25.

PART  |  12  The Regenerative Heart

Butler, E.G., 1955. Regeneration of the urodele forelimb after reversal of its proximodistal axis. J. Morphol. 96, 265–281. Campbell, G., Tomlinson, A., 1995. Initiation of proximodistal axis in insect legs. Development 121, 619–628. Canter, C.E., Goss, R.J., 1975. Induction of extra nephrons in unilaterally nephrectomized immature rats. Proc. Soc. Exp. Biol. Med. 148, 294–296. Chen, Y., Lin, G.F., Slack, J.M., 2006. Control of muscle regeneration in the Xenopus tadpole tail by Pax7. Development 133, 2303–2313. Coe, W.R., 1934. Analysis of regenerative processes in nemerteans. Biol. Bull. 66, 304–315. da Silva, S.M., Gates, P.B., Brockes, J.P., 2002. The newt ortholog of CD59 is implicated in proximodistal identity during amphibian limb regeneration. Dev. Cell 3, 547–555. Echeverri, K., Tanaka, E.M., 2005. Proximodistal patterning during limb regeneration. Dev. Biol. 279, 391–401. Fausto, N., 2000. Liver regeneration. J. Hepatol. 32, 19–31. Finnerty, J.R., Pang, K., Burton, P., Paulson, D., Martindale, M.Q., 2004. Origins of bilateral symmetry: Hox and Dpp expression in a sea anemone. Science 304, 1335–1337. French, V., 1982. Leg regeneration in insects – cell-interactions and lineage. Am. Zool. 22, 79–90. Goss, R.J., 1992. The evolution of regeneration – adaptive or inherent. J. Theor. Biol. 159, 241–260. Holstein, T.W., Hobmayer, E., Technau, U., 2003. Cnidarians: An evolutionarily conserved model system for regeneration?. Dev. Dyn. 226, 257–267. Huang, W.D., Ma, K., Zhang, J., Qatanani, M., Cuvillier, J., Liu, J., Dong, B.N., Huang, X.F., Moore, D.D., 2006. Nuclear receptor-dependent bile acid signaling is required for normal liver regeneration. Science 312, 233–236. Kam, I., Lynch, S., Svanas, G., Todo, S., Polim