354 95 179MB
English Pages 908 [900] Year 2016;2017
Any screen. Any time. Anywhere. Activate the eBook version of this title at no additional charge.
Student Consult eBooks give you the power to browse and find content, view enhanced images, share notes and highlights—both online and offline.
Unlock your eBook today. 1 Visit studentconsult.inkling.com/redeem 2
Scan this QR code to redeem your eBook through your mobile device:
Scratch off your code
3 Type code into “Enter Code” box 4
Click “Redeem”
5
Log in or Sign up
6
Go to “My Library” Place Peel Off Sticker Here
It’s that easy! For technical assistance: email [email protected] call 1-800-401-9962 (inside the US) call +1-314-447-8200 (outside the US)
Use of the current edition of the electronic version of this book (eBook) is subject to the terms of the nontransferable, limited license granted on studentconsult.inkling.com. Access to the eBook is limited to the first individual who redeems the PIN, located on the inside cover of this book, at studentconsult.inkling.com and may not be transferred to another party by resale, lending, or other means. 2015v1.0
CELL BIOLOGY
NOTE TO INSTRUCTORS: Contact your Elsevier Sales Representative for image banks for Cell Biology, 3e, or request these supporting materials at: http://evolve.elsevier.com
This page intentionally left blank
THIRD EDITION
CELL BIOLOGY THOMAS D. POLLARD, MD
Sterling Professor Department of Molecular, Cellular, and Developmental Biology Yale University New Haven, Connecticut
WILLIAM C. EARNSHAW, PhD, FRSE
Professor and Wellcome Trust Principal Research Fellow Wellcome Trust Centre for Cell Biology, ICB University of Edinburgh Scotland, United Kingdom
JENNIFER LIPPINCOTT-SCHWARTZ, PhD
Group Leader Howard Hughes Medical Institute, Janelia Research Campus Ashburn, Virginia
GRAHAM T. JOHNSON, MA, PhD, CMI Director, Animated Cell Allen Institute for Cell Biology Seattle, Washington; QB3 Faculty Fellow University of California, San Francisco San Francisco, California
1600 John F. Kennedy Blvd. Ste 1800 Philadelphia, PA 19103-2899
CELL BIOLOGY, THIRD EDITION IE
ISBN: 978-0-323-34126-4 ISBN: 978-0-323-41740-2
Copyright © 2017 by Elsevier, Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).
Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. With respect to any drug or pharmaceutical products identified, readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of practitioners, relying on their own experience and knowledge of their patients, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take all appropriate safety precautions. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability 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. Previous editions copyrighted © 2008, 2004 by Thomas D. Pollard, William C. Earnshaw, Jennifer Lippincott-Schwartz. Library of Congress Cataloging-in-Publication Data Names: Pollard, Thomas D. (Thomas Dean), 1942- , author. | Earnshaw, William C., author. | Lippincott-Schwartz, Jennifer, author. | Johnson, Graham T., author. Title: Cell biology / Thomas D. Pollard, William C. Earnshaw, Jennifer Lippincott-Schwartz, Graham T. Johnson. Description: Third edition. | Philadelphia, PA : Elsevier, [2017] | Includes bibliographical references and index. Identifiers: LCCN 2016008034| ISBN 9780323341264 (hardcover : alk. paper) | ISBN 9780323417402 (international edition) Subjects: | MESH: Cell Physiological Phenomena | Cells Classification: LCC QH581.2 | NLM QU 375 | DDC 571.6—dc23 LC record available at http://lccn.loc.gov/2016008034 Executive Content Strategist: Elyse O’Grady Senior Content Development Specialist: Margaret Nelson Publishing Services Manager: Patricia Tannian Senior Project Manager: Carrie Stetz Design Direction: Margaret Reid Printed in the United States of America Last digit is the print number: 9 8 7 6 5 4 3 2 1
The authors thank their families, who supported this work, and also express gratitude to their mentors, who helped to shape their views of how science should be conducted. Bill is proud to have both his longtime partner and confidante Margarete and his son Charles as advisors on the science for this edition. He would not be surprised if his daughter Irina were added to that panel for our next edition. His contributions are firstly dedicated to them. Bill also would like to thank Jonathan King, Stephen Harrison, Aaron Klug, Tony Crowther, Ron Laskey, and Uli Laemmli, who provided a diverse range of rich environments in which to learn that science at the highest level is an adventure that lasts a lifetime. Graham dedicates the book to his family, Margaret, Paul, and Lara Johnson; the Benhorins; friends Mari, Steve, and Andrew; and his partners Flower and Anna Kuo. He also thanks his mentors at the Scripps Research Institute, Arthur Olson, David Goodsell, Ron Milligan, and Ian Wilson, for developing his career. Jennifer thanks her husband Jonathan for his strong backing and her lab members for their enthusiasm for the project. Tom dedicates the book to his wife Patty, a constant source of support and inspiration for more than five decades, and his children Katie and Dan, who also provided advice on the book. He also thanks Ed Korn and the late Sus Ito for the opportunity to learn biochemistry and microscopy under their guidance, and Ed Taylor and the late Hugh Huxley, who served as role models.
Contributors Jeffrey L. Corden, PhD Professor Department of Molecular Biology and Genetics Johns Hopkins Medical School Baltimore, Maryland
vi
David Tollervey, PhD Professor Wellcome Trust Centre for Cell Biology University of Edinburgh Scotland, United Kingdom
Preface Our goal is to explain the molecular basis of life at the cellular level. We use evolution and molecular structures to provide the context for understanding the dynamic mechanisms that support life. As research in cell biology advances quickly, the field may appear to grow more complex, but we aim to show that understanding cells actually becomes simpler as new general principles emerge and more precise molecular mechanisms replace vague concepts about biological processes. For this edition, we revised the entire book, taking the reader to the frontiers of knowledge with exciting new information on every topic. We start with new insights about the evolution of eukaryotes, followed by macromolecules and research methods, including recent breakthroughs in light and electron microscopy. We begin the main part of the book with a section on basic molecular biology before sections on membranes, organelles, membrane traffic, signaling, adhesion and extracellular matrix, and cytoskeleton and cellular motility. As in the first two editions, we conclude with a comprehensive section on the cell cycle, which integrates all of the other topics. Our coverage of most topics begins with an introduction to the molecular hardware and finishes with an account of how the various molecules function together in physiological systems. This organization allows for a clearer exposition of the general principles of each class of molecules, since they are treated as a group rather than isolated examples for each biological system. This approach allows us to present the operation of complex processes, such as signaling pathways, as an integrated whole, without diversions to introduce the various components as they appear along the pathway. For example, the section on signaling mechanisms begins with chapters on receptors, cytoplasmic signal transduction proteins, and second messengers, so the reader is prepared to appreciate the dynamics of 10 critical signaling systems in the chapter that concludes the section. Teachers of shorter courses may concentrate on a subset of the examples in these systems chapters, or they may use parts of the “hardware” chapters as reference material. We use molecular structures as one starting point for explaining how each cellular system operates. This edition includes more than 50 of the most important and revealing new molecular structures derived from electron cryomicroscopy and x-ray crystallography. We explain the evolutionary history and molecular diversity
of each class of molecules, so the reader learns where the many varieties of each type of molecule came from. Our goal is for readers to understand the big picture rather than just a mass of details. For example, Chapter 16 opens with an original figure showing the evolution of all types of ion channels to provide context for each family of channels in the following text. Given that these molecular systems operate on time scales ranging from milliseconds to hours, we note (where it is relevant) the concentrations of the molecules and the rates of their reactions to help readers appreciate the dynamics of life processes. We present a wealth of experimental evidence in figures showing micrographs, molecular structures, and graphs that emphasize the results rather than the experimental details. Many of the methods will be new to readers. The chapter on experimental methods introduces how and why scientists use particularly important approaches (such as microscopy, classical genetics, genomics and reverse genetics, and biochemical methods) to identify new molecules, map molecular pathways, or verify physiological functions. The book emphasizes molecular mechanisms because they reveal the general principles of cellular function. As a further demonstration of this generality, we use a wide range of experimental organisms and specialized cells and tissues of vertebrate animals to illustrate these general principles. We also use medical “experiments of nature” to illustrate physiological functions throughout the book, since connections have now been made between most cellular systems and disease. The chapters on cellular functions integrate material on specialized cells and tissues. Epithelia, for example, are covered under membrane physiology and junctions; excitable membranes of neurons and muscle under membrane physiology; connective tissues under the extracellular matrix; the immune system under connective tissue cells, apoptosis, and signal transduction; muscle under the cytoskeleton and cell motility; and stem cells and cancer under the cell cycle and signal transduction. The Guide to Figures Featuring Specific Organisms and Specialized Cells that follows the Contents lists figures by organism and cell. The relevant text accompanies these figures. Readers who wish to assemble a unit on cellular and molecular mechanisms in the immune system, for example, will find the relevant material associated with the figures that cover lymphocytes/ immune system. vii
viii
PREFACE
Our Student Consult site provides links to the Protein Data Bank (PDB), so readers can use the PDB accession numbers in the figure legends to review original data, display an animated molecule, or search links to the original literature simply by clicking on the PDB number in the online version of the text.
Thomas D. Pollard
Throughout, we have attempted to create a view of Cell Biology that is more than just a list of parts and reactions. Our book will be a success if readers finish each section with the feeling that they understand better how some aspect of cellular behavior actually works at a mechanistic level and in our bodies.
William C. Earnshaw
Graham T. Johnson
Jennifer Lippincott-Schwartz
Contents SECTION I Introduction to Cell Biology 1 Introduction to Cells, 3 2 Evolution of Life on Earth, 15
SECTION II Chemical and Physical Background 3 Molecules: Structures and Dynamics, 31 4 Biophysical Principles, 53 5 Macromolecular Assembly, 63
SECTION VI Cellular Organelles and Membrane Trafficking 18 Posttranslational Targeting of Proteins, 303 19 Mitochondria, Chloroplasts, Peroxisomes, 317 20 Endoplasmic Reticulum, 331 21 Secretory Membrane System and Golgi Apparatus, 351 22 Endocytosis and the Endosomal Membrane System, 377
6 Research Strategies, 75
23 Processing and Degradation of Cellular Components, 393
SECTION III Chromatin, Chromosomes, and the Cell Nucleus
SECTION VII Signaling Mechanisms
7 Chromosome Organization, 107 8 DNA Packaging in Chromatin and Chromosomes, 123
24 Plasma Membrane Receptors, 411 25 Protein Hardware for Signaling, 425 26 Second Messengers, 443
9 Nuclear Structure and Dynamics, 143
27 Integration of Signals, 463
SECTION IV Central Dogma: From Gene to Protein
SECTION VIII Cellular Adhesion and the Extracellular Matrix
10 Gene Expression, 165 11 Eukaryotic RNA Processing, 189
28 Cells of the Extracellular Matrix and Immune System, 491
12 Protein Synthesis and Folding, 209
29 Extracellular Matrix Molecules, 505 30 Cellular Adhesion, 525
SECTION V Membrane Structure and Function
31 Intercellular Junctions, 543 32 Connective Tissues, 555
13 Membrane Structure and Dynamics, 227 15 Membrane Carriers, 253
SECTION IX Cytoskeleton and Cellular Motility
16 Membrane Channels, 261
33 Actin and Actin-Binding Proteins, 575
17 Membrane Physiology, 285
34 Microtubules and Centrosomes, 593
14 Membrane Pumps, 241
ix
x
CONTENTS
35 Intermediate Filaments, 613
42 S Phase and DNA Replication, 727
36 Motor Proteins, 623
43 G2 Phase, Responses to DNA Damage, and Control of Entry Into Mitosis, 743
37 Intracellular Motility, 639 38 Cellular Motility, 651 39 Muscles, 671
SECTION X Cell Cycle 40 Introduction to the Cell Cycle, 697 41 G1 Phase and Regulation of Cell Proliferation, 713
44 Mitosis and Cytokinesis, 755 45 Meiosis, 779 46 Programmed Cell Death, 797 Cell SnapShots, 817 Glossary, 823 Index, 851
Acknowledgments The authors thank their families and colleagues for sharing so much time with “the book.” Bill thanks Margarete, Charles, and Irina for sharing their weekends and summer holidays with this all-consuming project. He also thanks the Wellcome Trust for their incomparable support of the research in his laboratory and Melpomeni Platani and the Dundee Imaging Facility for access to the OMX microscope. Graham thanks Thao Do and Andrew Swift for contributions to the illustrations, and colleagues Megan Riel-Mehan, Tom Goddard, Arthur Olson, David Goodsell, Warren DeLeno, Andrej Sali, Tom Ferrin, Sandra Schmid, Rick Horwitz, UCSF, and the Allen Institute for Cell Science for facilitating work on this edition. He has special thanks for Ludovic Autin for programming the embedded Python Molecular Viewer (ePMV), which enabled substantial upgrades of many figures with complex structures. Jennifer thanks her family for sharing time with her part in the book. Tom appreciates four decades of support for his laboratory from the National Institutes of General Medical Sciences. Many generous individuals generously devoted their time to bring the science up to date by providing suggestions for revising chapters in their areas of expertise. We acknowledge these individuals at the end of each chapter and here as a group: Ueli Aebi, Anna Akhmanova, Julie Ahringer, Hiro Araki, Jiri Bartek, Tobias Baumgart, Wendy Bickmore, Craig Blackstone, Julian Blow, Jonathan Bogan, Juan Bonifacino, Ronald Breaker, Klaudia Brix, Anthony Brown, David Burgess, Cristina Cardoso, Andrew Carter, Bill Catterall, Pietro De Camilli, Iain Cheeseman, Per Paolo D’Avino, Abby Dernburg, Arshad Desai, Julie Donaldson, Charles Earnshaw, Donald Engelman, Job Dekker, Martin Embley, Barbara Ehrlich,
Roland Foisner, Nicholas Frankel, Tatsuo Fukagawa, Anton Gartner, Maurizio Gatti, David Gilbert, Gary Gorbsky, Holly Goodson, Jim Haber, Lea Harrington, Scott Hawley, Ron Hay, Margarete Heck, Ramanujan Hegde, Ludger Hengst, Harald Herrmann, Erika Holzbaur, Tim Hunt, Catherine Jackson, Emmanuelle Javaux, Scott Kaufmann, David Julius, Keisuke Kaji, Alexey Khodjakov, Vladimir Larionov, Dan Leahy, Richard Lewis, Kaspar Locker, Kazuhiro Maeshima, Marcos Malumbres, Luis Miguel Martins, Amy MacQueen, Ciaran Morrison, Adele Marston, Satyajit Mayor, Andrew Miranker, Tom Misteli, David Morgan, Peter Moore, Rachel O’Neill, Karen Oegema, Tom Owen-Hughes, Laurence Pelletier, Alberto Pendas, Jonathon Pines, Jordan Raff, Samara ReckPeterson, Elizabeth Rhoades, Matthew Rodeheffer, Michael Rout, Benoit Roux, John Rubinstein, Julian Sale, Eric Schirmer, John Solaro, Chris Scott, Beth Sullivan, Lee Sweeney, Margaret Titus, Andrew Thorburn, Ashok Venkitaraman, Rebecca Voorhees, Tom Williams, and Yongli Zhang. We thank David Sabatini, Susan Wente, and Yingming Zhao for permission to use their Cell SnapShots and Jason M. McAlexander for help with the final figures. Special thanks go to our colleagues at Elsevier. Our visionary editor Elyse O’Grady encouraged us to write this third edition and was a champion for the project from beginning to end as it evolved from a simple update of the second edition to an ambitious new book. Margaret Nelson, Content Development Specialist supreme, kept the whole project organized while dealing deftly with thousands of documents. Project Manager Carrie Stetz managed the assembly of the book with skill, patience, and good cheer in the face of many complicated requests for alterations.
xi
Guide to Figures Featuring Specific Organisms and Specialized Cells Organism/Specialized Cell Type
Figures
PROKARYOTES
Archaea Bacteria
Viruses
1.1, 1.2, 2.1, 2.4, 2.5 1.1, 1.2, 2.1, 2.4, 2.5, 2.7, 5.8, 5.12, 6.11, 7.4, 10.2, 10.5, 10.10, 10.11, 11.16, 12.6, 12.11, 13.9, 14.3, 14.9, 14.10, 15.4, 16.2, 16.3, 16.6, 16.13, 16.14, 18.2, 18.9, 18.10, 19.2, 19.7, 19.9, 20.5, 22.3, 22.10, 22.15, 27.11, 27.12, 27.13, 35.1, 37.12, 38.1, 38.24, 38.25, 42.3, 43.13, 44.27 5.10, 5.11, 5.12, 5.13, 22.15, 37.12
PROTOZOA
Amoeba Ciliates Other protozoa
2.1, 2.4, 2.8, 22.2, 22.5, 38.1, 38.4, 38.10, 41.7 2.4, 38.1, 38.13 2.4, 2.7, 36.7, 38.4, 37.10, 38.6, 38.21, 38.23
ALGAE AND PLANTS
Chloroplasts Green algae Plant cell wall Plant (general)
18.1, 18.2, 18.6, 19.7, 19.8, 19.9 2.8, 37.1, 37.9, 38.13, 38.14, 38.16, 38.18 31.4, 32.12, 32.13 1.2, 2.1, 2.4, 2.7, 2.8, 3.25, 6.6, 31.4, 33.1, 34.2, 36.7, 37.9, 38.1, 40.3, 44.26, 45.8
FUNGI
Budding yeast Fission yeast Other fungi
1.2, 2.4, 2.8, 6.15, 6.16, 7.3, 7.4, 7.7, 7.8, 8.22, 34.2, 34.20, 37.11, 42.4, 42.5, 42.15, 43.8 2.4, 2.8, 6.3, 7.8, 33.1, 40.6, 43.2, 44.23 2.8, 45.6
INVERTEBRATE ANIMALS
Echinoderms Nematodes Insects
2.8, 36.13, 40.11, 44.21, 44.22, 44.23 2.8, 36.7, 36.13, 38.11, 45.10, 46.9, 46.10 2.8, 7.4, 7.8, 7.15, 8.12, 8.13, 9.19, 14.19, 38.5, 38.11, 44.14, 44.12, 44.21, 44.25, 45.2, 45.8, 45.10
VERTEBRATE ANIMALS
Blood Granulocytes Lymphocytes/immune system Monocytes/macrophages Platelets Red blood cells Cancer Connective tissue Cartilage cells Extracellular matrix Fibroblasts Mast cells Bone cells Fat cells Epithelia Epidermal, stratified Glands, liver Intestine Kidney Respiratory system Vascular Muscle Cardiac muscle Skeletal muscle Smooth muscle Nervous system Central nervous system neurons Glial cells Peripheral nervous system neurons Synapses Reproductive system Oocytes, eggs Sperm Other human cells and disease Various organs
28.1, 28.4, 28.7, 30.13, 38.1 27.8, 28.1, 28.4, 28.9, 28.10, 46.7, 46.9, 46.18 28.1, 28.4, 28.7, 32.6, 32.11, 38.3, 46.2, 46.13 28.4, 28.5, 30.14, 32.11 13.8, 13.9, 13.11, 28.4, 32.11 34.19, 38.9, 41.2, 41.11, 41.12, 41.15, 42.10 28.1, 8.20 28.1, 28.1, 28.1, 27.7,
32.2, 32.3, 32.8, 32.9 28.2, 29.3, 29.4, 29.15, 32.1, 32.11, 35.1, 35.5, 37.1, 38.1 28.8 32.4, 32.5, 32.6, 32.7, 32.8, 32.9, 32.10 28.1, 28.3
29.7, 35.6, 40.1, 41.2, 41.5, 42.10, 46.8 21.26, 23.6, 34.20, 41.2, 44.2 17.2, 31.1, 32.1, 33.1, 33.2, 34.2, 46.19 17.3, 29.17, 35.1, 46.6, 46.7 17.4, 32.2, 34.3, 37.6, 38.17 22.6, 29.8, 29.17, 30.13, 30.14, 31.2, 32.11, 46.20 39.1, 39.13, 39.14, 39.18, 39.19, 39.20, 39.21, 39.22 17.9, 29.17, 33.3, 36.3, 36.4, 36.5, 39.1, 39.2, 39.3, 39.4, 39.5, 39.6, 39.7, 39.8, 39.9, 39.10, 39.11, 39.12, 39.13, 39.14, 39.15, 39.16, 39.17 29.8, 33.1, 35.8, 39.1, 39.23, 39.24 17.9, 17.10, 17.11, 30.8, 34.11, 34.12, 35.9, 37.7, 38.11, 39.12, 23.4 17.7, 17.9, 17.10, 29.17, 37.7 17.7, 17.9, 26.3, 26.16, 27.1, 27.2, 29.17, 30.15, 33.18, 35.9, 37.1, 37.3, 37.4, 37.5, 38.1, 38.6, 39.12 17.9, 17.10, 17.11, 29.17, 39.12 26.15, 34.14, 40.7, 40.8, 40.10, 40.11, 40.12, 45.14 38.1, 38.2, 38.14, 38.15, 38.20, 38.22, 45.1, 45.2, 45.4, 45.5, 45.8, 45.11 7.4, 7.6, 7.9, 7.11, 8.20, 9.10, 23.4, 41.2, 42.10
SECTION
Introduction to Cell Biology
I
This page intentionally left blank
1
CHAPTER
Introduction to Cells Biology is based on the fundamental laws of nature
embodied in chemistry and physics, but the origin and evolution of life on earth were historical events. This makes biology more like astronomy than like chemistry and physics. Neither the organization of the universe nor life as we know it had to evolve as they did. Chance played a central role. Throughout history and continuing today, the genes of all organisms have sustained chemical changes, some of which are inherited by their progeny. Many changes have no obvious effect on the fitness of the organism, but some reduce it and others improve fitness. Over the long term, competition between individuals with random differences in their genes determines which organisms survive in various environments. Surviving variants have a selective advantage over the alternatives, but the process does not necessarily optimize each chemical life process. Thus, students could probably design simpler or more elegant mechanisms for many cellular processes. Despite obvious differences, all forms of life share many molecular mechanisms, because they all descended from a common ancestor that lived 3 to 4 billion years ago (Fig. 1.1). This founding organism no longer exists, but it must have used many biochemical processes similar to those that sustain contemporary cells. Over several billion years, living organisms diverged from the common ancestor into three great divisions: Bacteria, Archaea, and Eucarya (Fig. 1.1). Archaea and Bacteria were considered to be one kingdom until the 1970s when the sequences of genes for ribosomal RNAs revealed that their ancestors branched from each other early in evolution. The origin of eukaryotes, cells with a nucleus, is still uncertain, but they inherited genes from both Archaea and Bacteria. One possibility is that eukaryotes originated when an Archaea engulfed a Bacterium that subsequently evolved into the mitochondrion. Multicellular eukaryotes (green, blue, and red in Fig. 1.1) evolved relatively recently, hundreds of millions of years
after single-celled eukaryotes appeared. Note that algae and plants branched before fungi, our nearest relatives on the tree of life. Living things differ in size and complexity and are adapted to environments as extreme as deep-sea hydrothermal vents at temperatures of 113°C or pockets of water at 0°C in frozen Antarctic lakes. Organisms also employ different strategies to extract energy from their environments. Plants, algae, and some Bacteria use photosynthesis to derive energy from sunlight. Some Bacteria and Archaea obtain energy by oxidizing inorganic compounds, such as hydrogen, hydrogen sulfide, or iron. Many organisms in all parts of the tree, including animals, extract energy from organic compounds. As the molecular mechanisms of life have become clearer, the underlying similarities among organisms are more impressive than their external differences. For example, all living organisms store genetic information in nucleic acids (usually DNA) using a common genetic Eucarya
Animals Plants Fungi
roplast Chlo
Amoeba ~1 billion years ago
drion on ch ito M
Bacteria
1–2 billion years ago, first eukaryote with a mitochondrion Ar chae on
~3.5 billion years ago, common ancestor emerged
Archaea
FIGURE 1.1 SIMPLIFIED PHYLOGENETIC TREE. This tree shows the common ancestor of all living things and the three main branches of life Archaea and Bacteria diverged from the common ancestor and both contributed to the origin of Eukaryotes. Note that eukaryotic mitochondria and chloroplasts originated as symbiotic Bacteria.
3
4
SECTION I n Introduction to Cell Biology
code, transfer genetic information from DNA to RNA to protein, employ proteins (and some RNAs) to catalyze chemical reactions, synthesize proteins on ribosomes, derive energy by breaking down simple sugars and lipids, use adenosine triphosphate (ATP) as their energy currency, and separate their cytoplasm from the external environment by means of phospholipid membranes containing pumps, carriers, and channels. Retention of these common molecular mechanisms in all parts of the phylogenetic tree is remarkable, given that the major groups of organisms have been separated for vast amounts of time and subjected to different selective pressures. These ancient biochemical mechanisms could have diverged radically from each other in the branches of the phylogenetic tree, but they worked well enough to be retained during natural selection of all surviving species. The cell is the only place on earth where the entire range of life-sustaining biochemical reactions can function, so an unbroken lineage stretches from the earliest cells to each living organism. Many interesting creatures were lost to extinction during evolution. The fact that extinction is irreversible, energizes discussions of biodiversity today. This book focuses on the molecular mechanisms underlying biological functions at the cellular level (Fig. 1.2). The rest of Chapter 1 summarizes the main points of the whole text including the general principles that
Nuclear envelope Nuclear lamina Nuclear pore Chromatin
apply equally to eukaryotes and prokaryotes and special features of eukaryotic cells. Chapter 2 explains what is known of the origins of life and its historic diversification through evolution. Chapter 3 covers the macromolecules that form cells, while Chapters 4 and 5 introduce the chemical and physical principles required to understand how these molecules assemble and function. Chapter 6 introduces laboratory methods for research in cell biology.
Universal Principles of Living Cells Biologists believe that a limited number of general principles based on common molecular mechanisms can explain even the most complex life processes in terms of chemistry and physics. This section summarizes the numerous features shared by all forms of life. 1. Genetic information stored in the chemical sequence of DNA is duplicated and passed on to daughter cells (Fig. 1.3). Long DNA molecules called chromosomes store the information required for cellular growth, multiplication, and function. Each DNA molecule is composed of two strands of four different nucleotides (adenine [A], cytosine [C], guanine [G], and thymine [T]) covalently linked in linear polymers. The two strands pair, forming a double helix held together by interactions between complementary pairs of nucleotide bases with one on each strand: A pairs
Rough endoplasmic reticulum Free ribosomes Protist
Centrioles
Nucleolus
Microtubule
Nucleus
Centrosome
Animal
Plant Cortex Microvillus
Lysosome
Coated pit
Peroxisome
Microtubule
A
Mold
Mitochondrion
Actin filaments
Golgi apparatus
Plasma membrane
Early endosome
Bacteria
B
Yeast
Archaea
FIGURE 1.2 BASIC CELLULAR ARCHITECTURE. A, Section of a eukaryotic cell showing the internal components. B, Comparison of cells from the major branches of the phylogenetic tree.
CHAPTER 1 n Introduction to Cells
Gene
Parent DNA strand
Replication intermediate
5
Two partially replicated DNA strands
DNA
Transcription
mRNA
Translation by ribosomes C
N
Polypeptide chain of amino acids
Folding
Folded protein Two identical DNA strands
FIGURE 1.3 DNA STRUCTURE AND REPLICATION. Genes stored as the sequence of bases in DNA are replicated enzymatically, forming two identical copies from one double-stranded original.
with T and C pairs with G. The two strands separate during enzymatic replication of DNA, each serving as a template for the synthesis of a new complementary strand, thereby producing two identical copies of the DNA. Precise segregation of one newly duplicated double helix to each daughter cell then guarantees the transmission of intact genetic information to the next generation. 2. Linear chemical sequences stored in DNA code for both the linear sequences and three-dimensional structures of RNAs and proteins (Fig. 1.4). Enzymes called RNA polymerases copy (transcribe) the information stored in genes into linear sequences of nucleotides of RNA molecules. Many RNAs have structural roles, regulatory functions, or enzymatic activity; for example, ribosomal RNA is by far the most abundant class of RNA in cells. Other genes produce messenger RNA (mRNA) molecules that act as templates for protein synthesis, specifying the sequence of amino acids during the synthesis of polypeptides by ribosomes. The amino acid sequence of most proteins contains sufficient information to specify how the polypeptide folds into a unique three-dimensional structure with biological activity. Two broad mechanisms control the production and processing of RNA and protein from tens of thousands of genes. Genetically encoded control circuits consisting of proteins and RNAs respond to environmental stimuli through signaling pathways. Epigenetic controls involve modifications of DNA or associated proteins that affect gene expression. Some epigenetic modifications can be transmitted during cell division and from a parent to an offspring. The basic plan for the cell contained in the genome, together with ongoing regulatory
=
FIGURE 1.4 Genetic information contained in the base sequence of DNA determines the amino acid sequence of a protein and its threedimensional structure. Enzymes copy (transcribe) the sequence of bases in a gene to make a messenger RNA (mRNA). Ribosomes use the sequence of bases in the mRNA as a template to synthesize (translate) a corresponding linear polymer of amino acids. This polypeptide folds spontaneously to form a three-dimensional protein molecule, in this example the actin-binding protein profilin. (For reference, see Protein Data Bank [www.rcsb.org] file 1ACF.) Scale drawings of DNA, mRNA, polypeptide, and folded protein: The folded protein is enlarged at the bottom and rendered in two styles—space-filling surface model (left) and a ribbon diagram showing the polypeptide folded into blue α-helices and yellow β-strands (right).
mechanisms (see points 7 and 8 below), works so well that each human develops with few defects from a single fertilized egg into a complicated ensemble of trillions of specialized cells that function harmoniously for decades in an ever-changing environment. 3. Macromolecular structures assemble from subunits (Fig. 1.5). Many cellular components form by selfassembly of their constituent molecules without the aid of templates or enzymes. The protein, nucleic acid, and lipid molecules themselves contain the information required to assemble complex structures. Diffusion usually brings the molecules together during these assembly processes. Exclusion of water from complementary surfaces (“lock-and-key” packing), as well as electrostatic and hydrogen bonds, provides the energy to hold the subunits together. In some cases, protein chaperones assist with assembly by preventing the aggregation of incorrectly folded intermediates. Important cellular structures assembled in this
6
SECTION I n Introduction to Cell Biology
A. Atomic scale 1,500,000× 10 nm
B. Molecular scale
C. Macromolecular scale
D. Organelle scale
DNA
DNA and proteins
Chromatin fiber
Chromosome
Protein backbone
Globular proteins
Actin filament
Filopodium with plasma membrane around actin filaments
E. Cellular scale 3000× 5,000 nm
Microtubule Fatty acids
Lipid bilayer with proteins
Membrane
FIGURE 1.5 MACROMOLECULAR ASSEMBLY. Many macromolecular components of cells assemble spontaneously from constituent molecules without the guidance of templates. This figure shows chromosomes assembled from DNA and proteins, a bundle of actin filaments in a filopodium assembled from protein subunits, and the plasma membrane formed from lipids and proteins.
way include chromatin, consisting of nuclear DNA packaged by associated proteins; ribosomes, assembled from RNA and proteins; cytoskeletal polymers, assembled from protein subunits; and membranes formed from lipids and proteins. 4. Membranes grow by expansion of preexisting membranes (Fig. 1.6). Cellular membranes composed of lipids and proteins grow only by expansion of preexisting lipid bilayers rather than forming de novo. Thus membrane-bounded organelles, such as mitochondria and endoplasmic reticulum, multiply by growth and division of preexisting organelles and are inherited maternally from stockpiles stored in the egg. The endoplasmic reticulum (ER) plays a central role in membrane biogenesis as the site of phospholipid synthesis. Through a series of vesicle budding and fusion events, membrane made in the ER provides material for the Golgi apparatus, which, in turn, provides lipids and proteins for lysosomes and the plasma membrane. 5. Signal-receptor interactions target cellular constituents to their correct locations (Fig. 1.6). Specific recognition signals incorporated into the structures of proteins and nucleic acids route these molecules to their proper cellular compartments. Receptors recognize these signals and guide each molecule to its appropriate compartment. For example, proteins destined for the nucleus contain short amino acid sequences that bind receptors to facilitate their passage through nuclear pores into the nucleus.
Similarly, a peptide signal sequence first targets lysosomal proteins into the lumen of the ER. Subsequently, the Golgi apparatus adds a sugar-phosphate group recognized by receptors that secondarily target these proteins to lysosomes. 6. Cellular constituents move by diffusion, pumps, and motors (Fig. 1.7). Most small molecules move through the cytoplasm or membrane channels by diffusion. However, energy provided by ATP hydrolysis or electrochemical gradients is required for molecular pumps to drive molecules across membranes against concentration gradients. Similarly, motor proteins use energy from ATP hydrolysis to move organelles and other cargo along microtubules or actin filaments. In a more complicated example, protein molecules destined for mitochondria diffuse from their site of synthesis in the cytoplasm to a mitochondrion (Fig. 1.6), where they bind to a receptor. Energyrequiring reactions then transport the protein into the mitochondrion. 7. Receptors and signaling mechanisms allow cells to adapt to environmental conditions (Fig. 1.8). Environmental stimuli modify cellular behavior. Faced with an unpredictable environment, cells must decide which genes to express, which way to move, and whether to proliferate, differentiate into a specialized cell, or die. Some of these choices are programmed genetically or epigenetically, but minute-to-minute decisions generally involve the reception of chemical or physical stimuli from outside the cell and
7
CHAPTER 1 n Introduction to Cells
A. Protein targeting from free ribosomes Protein synthesized on free ribosomes
Completed proteins released into cytoplasm
Transport into nucleus
Soluble enzymes Cytoskeleton
Incorporation into membranes and lumens of peroxisomes and mitochondria
B. Protein targeting from ER-associated ribosomes Complete proteins incorporated into ER membrane or transported into ER lumen
mRNA
processing of these stimuli to change the behavior of the cell. Cells have an elaborate repertoire of receptors for a multitude of stimuli, including nutrients, growth factors, hormones, neurotransmitters, and toxins. Stimulation of receptors activates diverse signal-transducing mechanisms that amplify the message and generate a wide range of cellular responses. These include changes in the electrical potential of the plasma membrane, gene expression, and enzyme activity. Basic signal transduction mechanisms are ancient, but receptors and output systems have diversified by gene duplication and divergence during evolution. 8. Molecular feedback mechanisms control molecular composition, growth, and differentiation (Fig. 1.9). Living cells are dynamic, constantly fine-tuning their composition in response to external stimuli, nutrient
Vesicles move from ER to Golgi apparatus and return
Diffusion down a concentration gradient
Ca2+
Pump
Microtubule track
Lumen proteins secreted FIGURE 1.6 PROTEIN TARGETING. Signals built into the amino acid sequences of proteins target them to all compartments of the eukaryotic cell. A, Proteins synthesized on free ribosomes can be used locally in the cytoplasm or guided by different signals to the nucleus, mitochondria, or peroxisomes. B, Other signals target proteins for insertion into the membrane or lumen of the endoplasmic reticulum (ER). From there, a series of vesicular budding and fusion reactions carry the membrane proteins and lumen proteins to the Golgi apparatus, lysosomes, or plasma membrane. mRNA, messenger RNA.
Transport up a concentration gradient ATP ADP
Channel
Membrane proteins delivered to target membrane
Vesicles move from the Golgi to lysosomes and to plasma membrane
Ca2+
Motor pulls membrane compartment ATP ADP
FIGURE 1.7 MOLECULAR MOVEMENTS BY DIFFUSION, PUMPS, AND MOTORS. Diffusion: Molecules up to the size of globular proteins diffuse in the cytoplasm. Concentration gradients can provide a direction to diffusion, such as the diffusion of Ca2+ from a region of high concentration inside the endoplasmic reticulum through a membrane channel to a region of low concentration in the cytoplasm. Pumps: Adenosine triphosphate (ATP)-driven protein pumps transport ions up concentration gradients. Motors: ATP-driven motors move organelles and other large cargo along microtubules and actin filaments. ADP, adenosine diphosphate.
A. Ligand binds receptor turning it on R
R* G
E
G*
B. Receptor activates GTP-binding proteins
E*
K ATP
K*
cAMP
C. Activated enzymes make second messenger cAMP
D. cAMP activates protein kinases
E. Kinases phosphorylate and activate enzymes
FIGURE 1.8 RECEPTORS AND SIGNALS. Activation of cellular metabolism by an extracellular ligand, such as a hormone. In this example, binding of the hormone (A) triggers a series of linked biochemical reactions (B–E), leading through a second messenger molecule (cyclic adenosine monophosphate [cAMP]) and a cascade of three activated proteins to regulate a metabolic enzyme. The response to a single ligand is multiplied at steps B, C, and E, leading to thousands of activated enzymes. GTP, guanosine triphosphate.
8
SECTION I n Introduction to Cell Biology
Tryptophan Precursor 1 + Precursor 2 Enz 1
Enz 2 Intermediate Enz 3 Tyrosine
A Mitosis M
Check for damaged or unduplicated DNA
Check for chromosome attachment to mitotic spindle
G2
Cytokinesis DNA
Check for DNA nicks
G1 Growth in mass
S Chromosome duplication
B
Centrosome duplication starts
Check for favorable environmental conditions
FIGURE 1.9 MOLECULAR FEEDBACK LOOPS. A, Control of the synthesis of aromatic amino acids. An intermediate and the final products of this biochemical pathway inhibit three of nine enzymes (Enz) in a concentration-dependent fashion, automatically turning down the reactions that produced them. This maintains constant levels of the final products, two amino acids essential for protein synthesis. B, Control of the cell cycle. The cycle consists of four stages. During the G1 phase, the cell grows in size. During the S phase, the cell duplicates the DNA of its chromosomes. During the G2 phase, the cell checks for completion of DNA replication. In the M phase, chromosomes condense and attach to the mitotic spindle, which separates the duplicated pairs in preparation for the division of the cell by cytokinesis. Biochemical feedback loops called checkpoints halt the cycle (blunt bars) at several points until the successful completion of key preceding events.
availability, and internal signals. The most dramatic example is the regulation of each step in the cell cycle. Feedback loops assure that the conditions are suitable for each transition such as the onset of DNA synthesis and the decision to begin mitosis. Similarly, cells carefully balance the production and degrada tion of their constituent molecules. Cells produce “housekeeping” molecules for basic functions, such as intermediary metabolism, and subsets of other proteins and RNAs for specialized functions. A hierarchy of mechanisms controls the supply of each protein and RNA: epigenetic mechanisms designate whether a particular region of a chromosome is active or not; regulatory proteins turn specific genes on and off and modulate the rates of translation of mRNAs into protein; synthesis balanced by the rates of degradation determines the abundance of specific RNAs and proteins; phosphorylation (covalent modification of
certain amino acids with a charged phosphate group) regulates protein interactions and activities; and other mechanisms regulate of the distribution of each molecule within the cell. Feedback loops also regulate enzymes that synthesize and degrade proteins, nucleic acids, sugars, and lipids to ensure the proper levels of each cellular constituent. A practical consequence of these common biochemical mechanisms is that general principles may be discovered by studying any cell that is favorable for experimentation. This text cites many examples of research on bacteria, insects, protozoa, or fungi that revealed fundamental mechanisms shared by human cells. For example, humans and baker’s yeast use similar mechanisms to control the cell cycle, guide protein secretion, and segregate chromosomes at mitosis. Indeed, particular proteins are often functionally interchangeable between human and yeast cells.
Features That Distinguish Eukaryotic and Prokaryotic Cells Although sharing a common origin and basic biochemistry, cells vary considerably in their structure and organization (Fig. 1.2). Bacteria and Archaea have much in common, including chromosomes in the cytoplasm, cell membranes with similar families of pumps, carriers and channels, basic metabolic pathways, gene expression, motility powered by rotary flagella, and lack of membranebound organelles. On the other hand, these prokaryotes are wonderfully diverse in terms of morphology and their use of a wide range of energy sources. Eukaryotes comprise a multitude of unicellular organisms, algae, plants, amoebas, fungi, and animals that differ from prokaryotes in having a compartmentalized cytoplasm with membrane-bounded organelles including a nucleus. The basic features of eukaryotic cells were refined more than 1.5 billion years ago, before the major groups of eukaryotes diverged. The nuclear envelope separates the two major compartments: nucleoplasm and cytoplasm. Chromosomes carrying the cell’s genes and the machinery to express those genes reside inside the nucleus. Most eukaryotic cells have ER (the site of protein and phospholipid synthesis), a Golgi apparatus (adds sugars to membrane proteins, lysosomal proteins, and secretory proteins), lysosomes (compartments containing digestive enzymes), and peroxisomes (containers for enzymes involved in oxidative reactions). Most also have mitochondria that convert energy stored in the chemical bonds of nutrients into ATP. Cilia (and flagella) are ancient eukaryotic specializations used for motility or sensing the environment. Membrane-bounded compartments give eukaryotic cells a number of advantages. Membranes provide a barrier that allows each type of organelle to maintain novel ionic and enzymatic interior environments. Each
CHAPTER 1 n Introduction to Cells
of these special environments favors a subset of the biochemical reactions required for life as illustrated by the following examples. The nuclear envelope separates the synthesis and processing of RNA in the nucleus from the translation of mature mRNAs into proteins in the cytoplasm. Segregation of digestive enzymes in lysosomes prevents them from destroying other cellular components. ATP synthesis depends on the impermeable membrane around mitochondria; energy-releasing reactions produce a proton gradient across the membrane that drives enzymes in the membrane to synthesize ATP.
9
transmembrane channels, carriers, and pumps (Fig. 1.10). These transmembrane proteins provide the cell with nutrients, control internal ion concentrations, and establish a transmembrane electrical potential. A single amino acid change in one plasma membrane pump and Cl− channel causes the human disease cystic fibrosis. Other plasma membrane proteins mediate interactions of cells with their immediate environment. Transmembrane receptors convert the binding of extracellular signaling molecules, such as hormones and growth factors into chemical or electrical signals that influence the activity of the cell. Genetic defects in signaling proteins, which mistakenly turn on signals for growth in the absence of appropriate extracellular stimuli, contribute to human cancers. Plasma membrane adhesion proteins allow cells to bind specifically to each other or to the extracellular matrix (Fig. 1.10). These selective interactions allow cells to form multicellular associations, such as epithelia (sheets of cells that separate the interior of the body from the outside world). Similar interactions allow white blood cells to bind bacteria so that they can be ingested and killed. In cells that are subjected to mechanical forces, such as muscle and epithelia, cytoskeletal filaments inside the cell reinforce the plasma membrane adhesion proteins. In skin, defects in these attachments cause blistering diseases.
Overview of Eukaryotic Cellular Organization and Functions This section previews the major constituents and processes of eukaryotic cells. With this background the reader will be able to appreciate cross-references to chapters later in the book.
Plasma Membrane The plasma membrane is the interface of the cell with its environment (Fig. 1.2). Owing to the hydrophobic interior of its lipid bilayer, the plasma membrane is impermeable to ions and most water-soluble molecules. Consequently, they cross the membrane only through
ANOTHER CELL
CYTOPLASM
C
Actin
B
Na+ K+
C ADP
– – –
D
E
F
G ++ +
Na+ K+
Na+ Glucose Na+
H
K+
Na+ Glucose Na+
ATP
–
– – –
G
+ +
+ + +
+
K+
A OUTSIDE FIGURE 1.10 STRUCTURE AND FUNCTIONS OF AN ANIMAL CELL PLASMA MEMBRANE. The lipid bilayer is a permeability barrier between the cytoplasm and the extracellular environment. Transmembrane adhesion proteins anchor the membrane to the extracellular matrix (A) or to like receptors on other cells (B) and transmit forces to the cytoskeleton (C). Adenosine triphosphate (ATP)-driven enzymes (D) pump Na+ out of and K+ into the cell (E) to establish concentration gradients across the lipid bilayer. Transmembrane carrier proteins (F) use these ion concentration gradients to transport of nutrients into the cell. Selective ion channels (G) regulate the electrical potential across the membrane. A large variety of receptors (H) bind specific extracellular ligands and send signals across the membrane to the cytoplasm.
10
SECTION I n Introduction to Cell Biology
Nuclear envelope Nuclear pore
Nuclear pore
Nucleolus Chromatin
FIGURE 1.11 ELECTRON MICROGRAPH OF A THIN SECTION OF A NUCLEUS. (Courtesy Don Fawcett, Harvard Medical School, Boston, MA.)
Nucleus The nuclear envelope is a double membrane that separates the nucleus from the cytoplasm (Fig. 1.11). All traffic into and out of the nucleus passes through nuclear pores that bridge the double membranes. Inbound traffic includes all nuclear proteins and ribosomal proteins destined for the nucleolus. Outbound traffic includes mRNAs and ribosomal subunits. The nucleus stores genetic information in extraordinarily long DNA molecules called chromosomes. Remarkably, portions of genes encoding proteins and structural RNAs make up only a small fraction (140,000 bp) segmental duplications of DNA distributed across a region of 2 × 106 bp flank a unique sequence region of approximately 1 × 106 bp. If recombination occurs between the segmental duplications, approximately 1.6 × 106 bp, including the unique sequence DNA, are lost. Because of the highly complex organization of this region and the large size of the duplications, this turned out to be the most difficult region of chromosome 7 to sequence.
The Human Genome: Variations on a Theme The human “reference genome” sequence does not come from a single person, but is instead an idealized assembly derived from the DNA of a number of people. Constructing an artificial reference genome is necessary, because although we might imagine that there is only
113
one “human genome,” data from sequencing many thousands of genomes have shown that there are dramatic variations in DNA content and sequence among individuals. Famously, analysis of some particularly variable regions of repetitive sequences forms the basis for DNA testing in criminology and paternity testing. Given the large number of genomes sequenced to date, it makes sense to talk of a “typical” genome and how this differs from the reference. Prepare to be amazed. A typical genome has 4 to 5 × 106 differences from the reference! The largest number of affected base pairs are in 2100 to 2500 “structural variants” (changes involving >50 bp). These include deletions, more than 120 LINE and more than 900 SINE insertions, and other changes not found in the reference genome. Overall, they encompass 20 × 106 bp and often occur in regions of repeated DNA sequence. Other variations occur in genes, with a typical genome having approximately 165 mutations that truncate proteins, approximately 11,000 mutations that change protein sequences, and a staggering 520,000 mutations in regions thought to be involved in regulating gene expression. Occasionally, these variations are linked to inherited human disease, and genome-wide association studies (GWAS) correlating sequence changes with human disease are a major ongoing focus of these sequencing efforts. At centromere regions of chromosomes, the content of repeated DNA sequences commonly varies by over 106 bp between different individuals. Overall, this rather staggering variability leads to the question, “What is a ‘normal’ human genome?”
The Centromere: Overview The centromere is at the heart of all chromosomal movements in mitosis and meiosis, as it nucleates on its surface the formation of the button-like kinetochore (see Fig. 8.21), the structure that attaches chromosomes to the mitotic spindle (the microtubule-based apparatus upon which chromosomes move; see Chapter 44). In mitotic chromosomes of most higher eukaryotes, the centromere forms a waist-like stricture or primary constriction where the two sister chromatids are most intimately paired. The centromere is a chromatin structure, and both DNA and proteins are essential to its function.
Variations in Centromere Organization Among Species In budding yeast, autonomous CEN (centromere) sequences specify protein-binding sites required for assembly of the kinetochore; if inserted into circular DNA molecules (plasmids), they render them capable of interacting with the mitotic spindle and segregating during mitosis (Fig. 7.7). In other organisms, including the fission yeast Schizosaccharomyces pombe, centromere sequences require an activation event to nucleate kinetochore formation. This event appears to
114
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus Uppercase = present in at least 14 of the 16 centromeres Lowercase = present in at least 9 centromeres – = any base can be present here CDE I
5'
CDE II
78-86 bp A A TCAC TG ~ 90% A + T G G
A. S. cerevisiae 125 bp
CDE III t G t t Tt t G– t TTCCGAAa – – – a a a a a 3'
CDE I
A
CDE II
CDE III
B. S. pombe chromosome III ~110 kb
Replication
Central core
Replication
Inner repeats
Outer repeats
C. D. melanogaster X chromosome
Mitosis
B
420 kb
Mitosis
+
+
or
All cells
+
= Plasmid
Random segregation
C
Functional CEN sequence CDE I, CDE II, CDE III
FIGURE 7.7 THE BUDDING YEAST CENTROMERE (CEN) IS SPECIFIED BY A 125-BP SEQUENCE. A, Three conserved DNA elements (CDE I to CDE III). CDE I and CDE III bind proteins in a sequence-specific manner. CDE III has mirror symmetry: a central C (dot) is flanked by two regions of complementary DNA sequence (arrows). All that seems to be important about CDE II is its abundance of A and T nucleotides and its overall length. B–C, The assay for mitotic stability of a plasmid used to clone CEN DNA from most budding yeast chromosomes. The plasmid carries a gene encoding an enzyme involved in adenine metabolism. When the plasmid is present, colonies are white. If the plasmid is lost, the colonies become red as a result of the accumulation of a metabolic by-product. If the plasmid is capable of replication but lacks a centromere, the colonies will be mostly red, reflecting the inefficient segregation of the plasmid at mitosis (B). If the plasmid carries a functional centromere, the colonies will be white, as the plasmid will be successfully transmitted at nearly every division (C).
involve epigenetic modification of the DNA and/or chromatin (discussed later). CEN sequences from all 16 budding yeast chromosomes have a common organization based around three conserved sequence elements (Fig. 7.7). These are designated (in the 5′ to 3′ direction) CDE I (centromere DNA element I, 8 bp), CDE II (78 to 86 bp), and CDE III (25 bp). A 125-bp region spanning CDE I to CDE III is sufficient to direct the efficient segregation of a yeast chromosome, which can reach a size of more than 1 million bp. This type of centromere, in which the kinetochore is assembled as a result of protein recognition of
Transposons
AAGAG satellite
AATAT satellite
Nonrepetitive DNA
FIGURE 7.8 ORGANIZATION OF THE CENTROMERIC DNAS OF BUDDING YEAST, FISSION YEAST, AND FRUIT FLY. A, The budding yeast (Saccharomyces cerevisiae) point centromere is specified by a 125-bp sequence. B, The fission yeast (Schizosaccharomyces pombe) regional centromeres all contain central core DNA flanked by complex arrays of repeated sequences. Embedded within these repeated sequences are a number of genes encoding transfer RNAs, not shown here. The minimum region required to construct a functional centromere in fission yeast artificial chromosomes is approximately 10 kb in length and includes the central core DNA plus a portion of the flanking repeated DNA. C, The fruit fly (Drosophila melanogaster) also has a regional centromere encompassing 420 kb. This is rich in satellite DNA and contains a number of transposable elements. The same satellite DNAs and transposons are also found at other, noncentromeric, regions of the chromosomes.
specific DNA sequences, is known as a point centromere and to date has been found only in budding yeasts. Kinetochores assembled on point centromeres bind a single microtubule. Even though the average size of S. pombe chromosomes is only fivefold larger than their counterparts in S. cerevisiae (4.6 Mb vs. 0.87 Mb), fission yeast centromeres are 300- to 600-fold larger (Fig. 7.8). The smallest S. pombe centromere consists of 35,000 bp, whereas the largest spans 110,000 bp. Fission yeast centromeric DNA is much more complex than its budding yeast counterpart, containing a central core of 4 to 7 kb of uniquesequence DNA flanked by complex arrays of repeated sequences. This type of centromere is known as a regional centromere. Kinetochores assembled on regional centromeres bind multiple microtubules (two to four in the case of S. pombe). Studies of S. pombe centromeres revealed in addition to the primary DNA sequence, an epigenetic activation step is required for CEN DNA to function as a centromere. Epigenetic events are inheritable properties of
chromosomes that are not directly encoded in the nucleotide sequence. They are typically explained either by enzymatic modification of the DNA (eg, methylation of cytosine) or by modification of proteins that are stably associated with the DNA. Epigenetic mechanisms also play an essential role in the assembly of centromeres in higher eukaryotes, including humans. In both S. pombe and metazoans, these epigenetic changes involve the construction of a special chromatin environment at centromeres. What this means in practice is that (except budding yeast), no single DNA sequence can be put into cells and function directly as a centromere. If a piece of S. pombe centromeric DNA is introduced into cells, it must undergo a series of packaging events and modifications that turn it into a functional centromere. These events are so rare that when candidate DNA molecules with CEN sequences are introduced into S. pombe cells, only about 1 in 105 assembles into a functional centromere. Regional centromeres are typically organized around a core region that nucleates kinetochore formation during mitosis. This core consists of a specialized form of chromatin called centrochromatin containing CENP-A, a specialized form of histone H3 that can replace H3 in nucleosomes (see Fig. 8.21). How the centrochromatin is organized varies dramatically between species (Fig. 7.8). Centrochromatin is typically flanked by constitutive heterochromatin, a form of chromatin that generally suppresses gene transcription and remains condensed throughout the cell cycle (see Fig. 8.7). Constitutive heterochromatin is characterized by the presence of special modifications of the histone proteins and other proteins that “read” (bind to) those modifications. (Chapter 8 discusses heterochromatin.) Both the core of the centromere and flanking heterochromatin are usually (but not invariably) comprised of repeated DNA sequences. The first fully sequenced centromere of a metazoan was that of rice chromosome 8. Sequencing was possible, because the rice centromere contains limited amounts of a centromeric satellite DNA (CentO) dispersed in blocks separated by transposons, retrotransposons, and fragments. All in all, 72% of this centromere is composed of repetitive sequences. The kinetochore, as defined by sequences associated with CENP-A, spans 750 kb and is interspersed with regions of chromatin containing normal histone H3 that is apparently packaged into heterochromatin. Surprisingly, this centromere region contains at least four genes that are actively transcribed. More recently it was discovered that chickens have three and the horse one chromosome with sequences composed of nonrepetitive DNA and lacking flanking heterochromatin. These centromeres are thought to be evolutionarily new, and may have originated from neocentromeres (see later). It is thought that such
CHAPTER 7 n Chromosome Organization
115
evolutionarily new centromeres gradually acquire repetitive DNA sequences, possibly because they provide as-yet unknown advantages over evolutionary time. The rice centromere is not evolutionarily new, having had its present organization for at least the last 10,000 years (since the indica and japonica cultivars of rice were separated) and appears to be intermediate between a canonical metazoan centromere and a neocentromere. The centromere organization of the fruit fly D. melanogaster shows important similarities and differences to the rice centromere. The centromere of the fly’s X chromosome occupies a stretch of roughly 420,000 bp (Fig. 7.8) that is composed mostly of simple-sequence satellite DNAs interspersed with transposable DNA elements. This resembles the situation in plants; however, in Drosophila, no sequences were found in this region that are unique to the fly centromeres; all sequences found at centromeres could also be found on the chromosome arms. Thus, it appears that something other than the DNA sequence alone must be responsible for conferring centromere activity on this region of the chromosome. In addition to point centromeres in budding yeast and regional centromeres found in most metazoans, many plants and insects as well as in the nematode C. elegans have a third variant, in which centromere activity is distributed along the whole length of the mitotic chromosomes. These holocentric chromosomes have binding sites for about 20 microtubules distributed along the whole poleward-facing surface of the chromosome during mitosis rather than a disk-like kinetochore at a centromeric constriction, as in humans. If a holocentric chromosome is fragmented, every piece can bind microtubules and segregate in mitosis. Perhaps surprisingly, the proteins of the holocentric kinetochore are the same as those found at disk-like regional kinetochores (see Chapter 8). Accordingly CENP-A is found in domains scattered across half of the worm genome that are characterized by low levels of transcription in the germline. One possibility is that in these chromosomes, any chromatin with the right transcriptional profile can serve to nucleate kinetochore assembly—perhaps the requirement for special epigenetic marks has been relaxed.
Vertebrate Centromere DNA Vertebrate centromeres initially proved extremely difficult to characterize in molecular detail, largely due to their large size and complex, highly repetitious organization. For example, the centromere of chromosome 21 (the smallest human chromosome at ~48 million bp) has been estimated to encompass more than 5 million bp. This entire region is composed of many thousands of copies of short DNA repeat sequences clustered together in head-to-tail arrays known as satellite DNA. Many lines of research have now converged to reveal that this centromere-associated satellite DNA is a preferred site of
116
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus Centromere Chromatid
1
2
3
4
Centromeric satellite DNA 300,000–5,000,000 bp
A1 B1 C1 D1 E1 F1 Higher-order arrays A4
B4
C4 D4 E4
F4
Monomer (171 bp) CTTCGTTGGAAPuCGGGA
CENP-B box (not in all monomers)
A TTCGTTGGAAACGGGA
Mouse CENP-B box
FIGURE 7.9 HIERARCHICAL ORGANIZATION OF αSATELLITE DNA AT HUMAN CENTROMERES. The numbers (1 to 4) indicate higher-order repeats of α-satellite DNA. These may contain from 2 to 32 monomers (indicated by A1, B1, and so on). DNA sequences of adjacent monomers within a repeat (eg, A1, B1, C1) may differ by as much as 40% from one another. DNA sequences of monomers occupying identical positions within the higher-order repeats (A1, A4, etc.) are nearly 99% identical to one another. The red sequence shown at the bottom represents the binding site for centromeric protein CENP-B.
centromere formation, but that centromeres can (rarely) form elsewhere. The trigger that seems to define any particular region of the chromosome as a centromere involves epigenetic modifications of the DNA and chromatin, at least one of which is the binding of the specialized histone H3 variant CENP-A. The major human centromeric satellite DNA, αsatellite, is a complex family of repeated sequences that constitutes approximately 5% of the genome. Monomers averaging approximately 171 bp long are organized into higher-order repeats (Fig. 7.9). Some of the monomers have a conserved 17-bp sequence (the CENP-B box) that forms the binding site for the centromeric protein CENP-B (mentioned earlier as having its origin in an ancient DNA transposon). The organization of higherorder repeats varies greatly from chromosome to chromosome, and numerous repeat patterns, comprising 2 to 32 monomers, have been described. Each chromosome has one or a few types of higher-order repeats of α-satellite DNA. The entire centromeric region of certain chromosomes may be composed of α-satellite monomers, apparently with little or no interspersed DNA of other types. The amount of α-satellite DNA at different centromeres varies widely: from as little as 300,000 bp on the Y chromosome to up to 5 million bp on chromosome 7. In
addition, the α-satellite DNA content of a given chromosome can vary by more than a million bp between different individuals. Clearly, a wide variation in the local organization of α-satellite DNA is tolerated. Human chromosomes also contain several other families of satellite DNA. Classical satellites I to IV, which together constitute 2% to 5% of the genome, are composed of divergent repeats of the sequence GGAAT. These satellites occur in blocks more than 20,000 bp long that are immediately adjacent to the centromeres of a number of chromosomes and may be found at lower levels near most centromeres. The so-called pericentromeric region adjacent to the centromere of chromosome 9 apparently contains 7 to 10 million bp of satellite III sequence. The long arm of the Y chromosome also contains huge amounts of satellite III DNA (up to 40% of its total DNA). If α-satellite DNA arrays longer than about 50,000 bp are introduced into cultured human cells, they occasionally form tiny minichromosomes with functional centromeres. For this to work, the α-satellite DNA arrays must have a highly regular organization, and some of the monomers must contain binding sites for CENP-B. Formation of these mammalian artificial chromosomes is very inefficient, so it is clear that α-satellite DNA arrays cannot automatically function as CEN DNA—some type of epigenetic activation is required. There is an interesting corollary of this role of epigenetic modifications in assembly of a functional centromere. Suppose a bit of noncentromeric DNA somehow acquired the right set of modifications. Could that now function as a centromere? The answer is yes. The formation of neocentromeres on noncentromeric DNA has been seen in S. pombe, fruit flies, chickens, and humans and was first described in plants. Rare individuals have a chromosome fragment that segregates in mitosis, despite loss of the normal centromere. Such chromosomes have acquired a new centromere or neocentromere in a new location on one of the chromosome arms. Remarkably, neocentromeres are composed of the normal DNA that exists at that location on the chromosome arm and yet somehow has acquired centromere function. Neocentromeres are bona fide centromeres; for example, they bind all known centromeric proteins except for CENP-B, which requires specific sequences on α-satellite DNA for binding. Different neocentromeres need have no sequences in common. These observations strongly support the hypothesis that epigenetic markers rather than the exact DNA sequence specify the centromere. The natural occurrence of α-satellite DNA at centromeres may reflect a propensity of α-satellite chromatin to acquire the epigenetic mark, rather than a sequence-specific mechanism as occurs in S. cerevisiae. In one study of a chicken cell line, more than 100 independent new neocentromeres formed after the
CHAPTER 7 n Chromosome Organization
normal centromere was deleted experimentally (Fig. 7.10). Amazingly, every neocentromere formed on a different DNA sequence with no common underlying sequence features. Regions containing or lacking genes could be incorporated into a neocentromere. The only common feature was a domain of chromatin roughly 40,000 bp long containing CENP-A nucleosomes. This
A. Entire Z chromosome Reads × 103
B. Entire Z chromosome with neocentromere
1000
1000
CENP-A
0
0
CENP-A
80
0
Position on Z chromosome
Reads × 103
15
Z centromere
0
80
Position on Z chromosome
15
BM23 neocentromere
CENP-A
0
42.62
42.65
Position on Z chromosome
CENP-A
0
3.78
3.81
Position on Z chromosome
FIGURE 7.10 DISTRIBUTION OF CENTROMERE HISTONE CENP-A AT THE NATURAL CENTROMERE AND AT A NEOCENTROMERE ON THE CHICKEN Z CHROMOSOME. Cells were treated with formaldehyde to crosslink proteins to DNA. Isolated DNA was fragmented into short pieces of a few hundred base pairs and an antibody used to pull down the DNA fragments crosslinked to CENP-A. Thousands of DNA fragments associated with CENP-A were then sequenced. These sequences were mapped along the Z chromosome (the female sex chromosome of the chicken). (Data from Hori T, Shang W-H, Toyoda A, Misu S, et al. Histone H4 Lys 20 mono-methylation of the CENP-A nucleosome is essential for kinetochore assembly. Dev Cell. 2014;29:740–749.)
A
117
corresponds to the size of the centromere on the starting chromosome. The epigenetic mark that defines an active centromere can be lost as well as gained. Thus, it is possible for a centromere to retain its normal DNA composition and yet lose the ability to assemble a kinetochore. This has been seen most clearly in naturally occurring human dicentric chromosomes. The chromosome shown in Fig. 7.11 arose through a breakage and fusion near the long arm of chromosome 13 and has two centromeres. As shown in the figure, one of these lost the ability to assemble a kinetochore even though it retained its α-satellite. What is the elusive epigenetic mark and how does it “magically” mark a region of the chromosome as a centromere? At present, all evidence suggests that the epigenetic mark has something to do with low level transcription of the CENP-A-containing DNA during mitosis. This is remarkable, because transcription is supposed to be entirely shut off during mitosis, and indeed, it seems that centromeres are the only region of the genome that is transcribed at that time. We do not yet know whether it is the process of transcription that is important or whether the RNA transcripts themselves serve an important role in specifying centromere chromatin. Once a DNA sequence has acquired the proper epigenetic mark, it can assemble a functional kinetochore that can regulate chromosome behavior in mitosis. This involves the binding and function of 100 or more proteins as discussed in Chapter 8 (see Fig. 8.21).
Ends of the Chromosomes: Why Specialized Telomeres Are Needed The ends of chromosomal DNA molecules pose at least two problems that cells solve by packaging the chromosome ends into specialized structures called telomeres.
B. CENP-B (a-satellite DNA)
C. CENP-C (kinetochore)
Active
Inactive
FIGURE 7.11 EPIGENETIC REGULATION OF HUMAN CENTROMERE FUNCTION. An unusual chromosome was discovered during prenatal screening of a fetus that sonography had indicated to be abnormal. This chromosome consisted of two copies of the maternal chromosome 13 linked end to end. It thus contained two centromeres and so was termed dicentric. Such dicentric chromosomes are normally unstable during mitosis, as the two centromeres on one chromatid often become attached to opposite spindle poles. This causes the chromosome to be stretched between opposite spindle poles and ultimately break. In the case of this particular dicentric chromosome, one of the centromeres was inactivated (presumably, it lost its epigenetic mark). This chromosome thus behaves perfectly normally in mitosis. When the distribution of centromere proteins at the active and inactive centromeres was compared, it was found that CENP-B was present at both but that CENP-C, a marker for kinetochores, was present only at the active centromere. A, Organization of the dicentric chromosome. B, Phase-contrast view of chromosomes from the amniocytes (left). Phase-contrast view taken with superimposed antibody staining for CENP-B (right). C, DNA stain of a different chromosome spread (left). Staining with antibody specific for CENP-C (right). (B–C, Courtesy William C. Earnshaw.)
118
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
First, it is essential that cells distinguish the ends of a chromosome from breaks in DNA. When cells detect DNA breaks, they stop their progression through the cell cycle and repair the breaks by joining the ends together (see Box 43.1). Telomeres keep normal chromosome ends from inducing cell cycle arrest and from being joined to other DNA ends by the repair machinery. Second, telomeres permit the chromosomal DNA to be replicated out to the very end.
Structure of Telomeric DNA Telomeres in all eukaryotes tested to date (with the exception of several insect species including Drosophila) are composed of many repeats of short DNA sequences. The sequence 5′ TTAGGG 3′ is found at the ends of chromosomes in organisms ranging from human to rattlesnake to the fungus Neurospora crassa. In the human, roughly 650 to 2500 copies of this sequence are found at the ends of each chromosome, a total length of approximately 4000 to 15,000 bp (this varies in different tissues). Higher plant telomeres have the sequence TTTAGGG, and other variations of this repeat sequence have been noted in protozoans and yeasts. The telomeric repeat is organized in a unique orientation with respect to the chromosome end. Thus, the end of every chromosome has one G-rich strand and one complementary C-rich strand. The G-rich strand always makes up the 3′ end of the chromosomal DNA molecule. Thus, the very 3′ end of the chromosome always has the following structure: …(TTAGGG)-OH. Furthermore, the end of the chromosome is not a blunt structure; the G-rich strand ends in a single-stranded overhang 30 to 400 bp long. This single strand of DNA is critical for telomere structure and function. It regulates telomerase activity and also “invades” the double helix of telomeric repeats, base-pairing and causing the ends of chromosomes to form large loops, called T loops that protect chromosome ends (see later discussion). A surprisingly complex balance of enzymatic activities maintains this single strand of DNA. These activities change throughout the cell cycle in dividing cells. How Telomeres Replicate the Ends of the Chromosomal DNA Telomeres prevent the erosion of the end of the chromosomal DNA molecule during each round of replication (for a more extensive discussion of DNA replication, see Chapter 42). All DNA replication proceeds with a polarity of 3′ to 5′ on the template DNA (5′ to 3′ in the newly synthesized DNA). Furthermore, all DNA polymerases (but not RNA polymerases) work by elongating a pre-existing stretch of double-stranded nucleic acid. During cellular DNA replication, this is achieved by making a short RNA primer and then elongating the RNA: DNA duplex with DNA polymerase. The primer is subsequently removed, and DNA polymerase fills the
A
Parent strands 3' 5'
DNA unwound
DNA replication
B 3' 5'
Lagging strand
Primer Daughter strands
Leading strand
3' 5'
RNA primer removal Okazaki fragment ligation
C 3' 5'
SS overhang on other end
SS overhang (unreplicated DNA) 3' 5'
FIGURE 7.12 DNA REPLICATION PROBLEM AT CHROMOSOME ENDS. DNA polymerases cannot initiate the formation of DNA on a template de novo; they can only extend preexisting nucleotide strands (see Chapter 42). In contrast, RNA polymerases can initiate synthesis without a primer. All replicating DNA chains start from a short region of RNA, which is used to “prime” DNA polymerase. A, DNA strand separation. B, RNA primer synthesis. Replication of DNA starts with the synthesis of an RNA primer (magenta) complementary to a short sequence of DNA, which is extended by DNA polymerase. C, The RNA primer is degraded and the gap is filled in by DNA polymerase. This being true, how can the DNA underneath the very last RNA primer replicated? SS, single stranded.
gap by elongation from the next upstream DNA end (Fig. 7.12). If the terminus of the chromosomal DNA is replicated from an RNA primer that sits on the very end of the DNA molecule, it follows that when this primer is removed, there is no upstream DNA on which to put a primer. How, then, is the DNA underneath the last RNA primer replicated? Years of searching for a DNA polymerase that could operate in the opposite direction proved fruitless. The answer that ultimately emerged turned out to be both elegant and unexpected. Most organisms have an enzyme called telomerase that specifically lengthens the 3′ end of the chromosomal DNA. Telomerases contain both protein and RNA subunits. The sequences of the RNA component provided an essential clue to how this enzyme works. The RNA component of human telomerase contains the sequence AUCCCAAUC, which can base-pair with the TTAGGG telomere repeat at the ends of the
CHAPTER 7 n Chromosome Organization
A (TTAGGG)n
Parent strand Lagging strand Primer
3' 5'
(AATCCC)n
RNA primer removal
Okazaki fragment ligation
B
3' 5'
Telomerase polymerization
C
Telomerase RNA template
3'
Telomerase translocation and reannealing cycle GGGTTAGGGTTAGGGTTAGGGTTAG3' CCCAATCCCAATCCC5' A AUCCCAAUCCCA CC AU C 3'
AU
C
5'
Elongation
GGGTTAGGGTTAGGGTTAGGGTTAGGGTTAG CCCAATCCCAATCCC
AU C
A AUCCCAAUCCCA CC AU
C
chromosome. The enzyme uses its own RNA as a template for the synthesis of DNA, which it “grows” from the end of the chromosome (Fig. 7.13). This hypothesis was confirmed by showing that changing the sequence of the telomerase RNA alters the telomere sequence at the end of the chromosome. According to this model, the telomerase actually synthesizes DNA using an RNA template. Thus, telomerase is a reverse transcriptase similar to that involved in the movement of the LINE retrotransposons (Fig. 7.5). When L1 family LINE retrotransposons insert themselves into the chromosome, a DNA end created at a nick in the chromosome is used to prime synthesis of a DNA strand using the LINE RNA as template, the newly synthesized DNA being a direct extension of the chromosomal DNA molecule. Human telomerase consists of hTERT (the telomerase reverse transcriptase) complexed to hTERC, the telomerase RNA, which is 450 nucleotides long. Telomerase RNA varies in size and sequence between species. Active human telomerase can be reconstituted in vitro from purified hTERC and hTERT in the presence of a cell-free lysate from reticulocytes (which appears to provide essential protein-folding factors). In cells, telomerase is associated with auxiliary protein subunits that are involved in telomerase RNA processing and maturation. Telomerase is subject to tight biological regulation. Active enzyme is detected in only a few normal tissues of adult humans. These include the stem cells of various tissues and male germ cells. In addition, approximately 90% of cancer cells express active telomerase and abnormal expression of telomerase has been linked to cancer. This telomerase is thought to enable the cancer cells to grow indefinitely without undergoing erosion of the ends of the chromosomes.
119
Translocation GGGTTAGGGTTAGGGTTAGGGTTAGGGTTAG CCCAATCCCAATCCC
A AUCCCAAUCCCA CC AU
AU C
C
Elongation GGGTTAGGGTTAGGGTTAGGGTTAGGGTTAGGGTTAG CCCAATCCCAATCCC
AU C
A AUCCCAAUCCCA CC
AU
C
FIGURE 7.13 TELOMERASE PROVIDES A ECHANISM FOR LENGTHENING CHROMOSOMAL ENDS. A–B, Normal mechanisms of DNA replication are unable to replicate the very 3′ end of the chromosomal DNA. C, Telomerase solves this problem by providing its own template in the form of an intrinsic RNA subunit. This RNA subunit contains a sequence complementary to that found at the chromosome terminus on the 3′ strand. This sequence is able to basepair with the DNA at the chromosome terminus and act as a template for DNA synthesis. In this case, the primer is the 3′ end of the chromosomal DNA, and the template is the RNA of the telomerase enzyme. Thus, the process of telomere elongation is a specialized form of reverse transcription (copying RNA into DNA), a process similar to that occurring during transposition of LINES (long interspersed nuclear elements) (Fig. 7.5), and during the life cycle of certain RNA-containing tumor viruses. The telomerase enzyme releases and rebinds its template after each 6 to 7 bp of new DNA has been synthesized. Up to several hundred base pairs may be added to the telomere in this way. D, In most cells, the 3′ end of the chromosomal DNA terminates in a single-stranded G-rich strand 30 to 400 nucleotides long that is essential for telomere structure and function.
D
(TTAGGG)n (AATCCC)n
3' OH 5'
~200 Base overhang
120
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
Paradoxically, hTERC is not tightly regulated. The hTERC RNA is detected in many tissues, most of which lack telomerase activity. By contrast, the expression of hTERT correlates tightly with telomerase activity. Indeed, introduction of a DNA-encoding hTERT into telomerasenegative cells produces telomerase activity. This can have extremely important consequences for the proliferation of the cells (Fig. 7.16). In cells that lack telomerase, a second pathway can help maintain the telomeric repeats at chromosome ends. This ALT (alternative lengthening of telomeres) process involves DNA recombination between telomeres. Cancer cells that lack telomerase expression have an activated ALT pathway. A third solution to this problem was taken by dipterans such as D. melanogaster, in which the ends of the chromosomes are composed of transposable elements. In the fly, a few bp are lost from the end of the chromosome at every round of replication. This erosion of the chromosome ends is remedied by the occasional transposition of specialized transposable elements to the chromosome end. Thus, this appears to be an example of an originally “selfish DNA” that has become recruited for an essential cellular function.
Structural Proteins of the Telomere Telomeres provide special protected ends for the chromosomal DNA molecule, in part by coating the end of the DNA molecules with protective proteins and by adopting a specialized DNA loop structure. In organisms with relatively short telomeric DNA sequences, those sequences are packaged into a specialized chromatin structure. In mammals, in which the telomeric sequences are much longer, the bulk of the telomeric DNA is packaged into conventional chromatin (see Chapter 8). A complex of six proteins called shelterin associates with telomeres in most organisms that have a telomerase (Fig. 7.14). Two subunits directly bind the TTAGGG duplex while one binds to the single stranded overhang. The other two subunits bridge the DNA binding subunits. S. cerevisiae has homologous subunits that bind to the telomeric repeats and the G-strand overhang. They protect the end of the recessed C-rich strand at telomeres, and this strand is rapidly degraded if these proteins are missing, with lethal consequences for the cell. Shelterin appears to both regulate telomerase activity and play an essential role in protecting chromosome ends.
A. Telomerase reverse transcriptase lengthens the chomosome end
B. The shelterin complex protects chromosome ends ATM signaling ATR signaling
AGGGTTAGGGTTAGGGTTAG3' OH TCCCAATCCCAATCC
TIN2
5'
3’ OH
Homology-directed repair Nonhomologous end joining 5’ end resection
TRF1
AU
TCCCAATCCCAATCC AAUCCCAAUCC C A CC AUC C strand 3' 5'
TRF2
RAP1
C. Loop model of telomere structure
100 nm
T loop
C
AA
AGGGTTAGGGTTAGGGTTAGGGTTAGGGTTAGGGTTAG3' OH TCCCAATCCCAATCC5' A A UC C C A UC C C AA U C C C Polymerase alpha 3'
U
extends C-strand
POT1
C
3'
AGGGTTAGGGTTAGGGTTAGGGTT
Telomerase extends G-strand
TPP1
AU
Shelterin AU recruits telomerase
A AUCCCAAUCCCA CC C
C
5' AGGGTTAGGGTTAGGGTTAGGGTTAGGGTTAGGGTTAG3' OH TCCCAATCCCAATCC AATCCCAATCCCA5'
D loop
FIGURE 7.14 TELOMERE STRUCTURES. A, Structure of telomerase. B, Organization and functions of shelterin, a complex of six subunits. TRF1 and TRF2 dimers bind to the double-stranded (TTAGGG)n repeats at telomeres. Together they bind TIN2, which in turn binds TPP1, which helps recruit POT1 to the single-stranded DNA at the chromosome end. If shelterin is lost, chromosomes fuse with one another, and many abnormalities are seen. C, T-loop model for vertebrate telomeres. Chromosomal ends may form a T-loop structure when a single-stranded G-rich 3′ end of the chromosome “invades” a double-stranded portion of the telomere, base-pairing with one strand and displacing the other strand (D loop). Inset, A T loop excised together with its chromatin proteins from a chicken erythrocyte chromosome. (Inset, From Nikitina T, Woodcock CL. Closed chromatin loops at the ends of chromosomes. J Cell Biol. 2004;166:161–165.)
CHAPTER 7 n Chromosome Organization
The Ku70/80 and MRN complexes are additional components of telomeres that are conserved from yeast to human. If mutations inactivate these complexes, telomeres frequently fuse together. This poses a conundrum, because elsewhere on chromosomes, these same proteins recognize DNA ends and participate in the repair of DNA breaks by joining bits of broken DNA together, a pathway known as nonhomologous end joining (NHEJ) (see Chapter 43). This is exactly the opposite of their role at telomeres. It thus appears that the breakage repair machinery recognizes chromosome ends, but the shelterin complex somehow changes its function from an end-joining role to an end-blocking protective role. Loss of shelterin results in a loss of the G-strand overhangs and a dramatic increase in the tendency of chromosomes to fuse end to end. This is because the chromosome ends are now recognized as DNA breaks, and the cell attempts to repair them using several of the DNA repair pathways discussed in Chapter 43. Fig. 7.15 shows fused chromosomes in a Drosophila mutant lacking a protein essential for the assembly of the fly equivalent of the shelterin complex at telomeres. In organisms with shelterin, the end protection may occur in part because subunit TRF2 can promote the formation of a special looped configuration of DNA in which the single-stranded G-strand overhang is base-paired with “upstream” TTAGGG DNA (Fig. 7.14C). Telomeres may also direct chromosome ends to their proper location within the cell. In budding yeast (and many other species), telomeres prefer to cluster together
A. Wild-type
B. Caravaggio
2 X
4
4
2
4
X
3
3
4
3
OH
HO 4 DNA repair/ end fusion
Telomeres protect ends 3
4
2
3
X
3
C. HOAP protein
3
4
FIGURE 7.15 DISRUPTION OF THE PROTECTIVE COMPLEX AT TELOMERES RESULTS IN CHROMOSOME FUSIONS. A, The chromosomes of a wild-type female Drosophila melanogaster seen at mitotic metaphase (see Chapter 44). B, The Caravaggio mutant is characterized by a “train” of chromosomes generated by telomeretelomere fusions. (Caravaggio is the name of an Italian train.) C, The cav gene encodes HP1/Orc2-associated protein (HOAP), which specifically localizes at all Drosophila telomeres. (A, Courtesy Gianni Cenci and Maurizio Gatti, University of Rome, Italy. B, From Cenci G, Siriaco G, Raffa GD, et al. Drosophila HOAP protein is required for telomere capping. Nat Cell Biol. 2003;5:82–84. C, Courtesy Nature Cell Biology.)
121
at the nuclear periphery. Mutants in telomere-binding proteins, or in regions of the histones with which they interact, disrupt this clustering in yeast. This results in activation of genes that are normally silenced when located in close proximity to telomeres. Thus, positioning of the telomere within the nucleus may be used to sequester genes into compartments where their transcriptional activity is repressed.
Telomeres, Aging, and Cancer Although the average length of telomeric repeats in humans is approximately 4000 bp, this length varies. Chromosomes of older individuals have shorter telomeres, and gametes have longer telomeres. This suggested the interesting possibility that chromosomes might lose telomeric sequences during the life of an individual. The relationship between telomere length and aging can be studied in cultured cells. Normal cells in culture grow for only a limited number of generations (often called the Hayflick limit) before undergoing senescence (this involves permanent cessation of growth, enlargement in size, and expression of marker enzymes, such as β-galactosidase). Because normal somatic cells lack telomerase activity, their telomeres shorten and eventually reach a critically short threshold before the cells senesce. In some cases, it is possible to force senescent cells to resume proliferation (eg, by expressing certain viral oncogenes). These “driven” cells continue to divide and their telomeres continue to shorten until a crisis point is reached. In crisis, cells suffer chromosomal instability (chromosomal fusions and breaks can occur) and cell death. In populations of human cells in crisis, very rarely (in approximately 1 in 106 cases), cells appear that once again grow normally. These cells now express telomerase. These observations with cultured cells led to the suggestion that senescence might occur in cells when the telomeric repeats of one or more chromosomes are reduced to a critical level. If correct, this model suggests very interesting (and controversial) implications for the regulation of cell life. Suppose that telomerase is active in the germline, so that all gametes have long telomeres. Now, if the enzyme were inactivated in somatic cells, this would effectively provide every cell lineage with a limitation on how many times it could divide before loss of telomeric sequences caused it to become senescent. Provided that the starting telomeres were sufficiently long and that telomerase was expressed in stem cells of tissues like testis and intestine, in which rapid division occurs throughout the life of the individual, this lack of telomerase in most cells would have no deleterious effect on the life span of the organism. In fact, such a mechanism might provide an important advantage by minimizing the chances that a clone of cells would escape from the normal regulation of growth control and become cancerous.
122
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
This model has been tested in two ways. First, mice were prepared in which the gene coding for the RNA component of telomerase or the telomerase reverse transcriptase was disrupted. These mice were healthy and fertile for six generations in the complete absence of telomerase but then subsequent generations became sterile as a result of cell death in the male germline. The cell death occurred when the telomeres shortened below a critical threshold. Having telomeres approximately seven times longer than humans might have contributed to their initial survival through several generations. Other studies show that mice age prematurely, when their telomeres shorten below a certain length. Remarkably, this ageing phenotype can be cured over the course of several weeks by activating hTERT in those mice. However, this “cure” can be a two-edged sword, as depending on the genetic makeup of the mice, the activation of hTERT can result in the formation of aggressive tumors! These experiments show that telomerase is not essential for the day-to-day life of the mouse, but clearly it is needed for the long-term survival of the species. In humans, a number of diseases (collectively termed “telomeropathies”) are associated with inheritance of mutant alleles of telomere components. These diseases include dyskeratosis congenita (a complex condition affecting the skin and nails that is associated with a complex array of other life-threatening conditions), aplastic anemia (loss of blood cell formation), bone marrow failure and others. These diseases are all associated with failures in cell proliferation. In a second experiment, the hTERT reverse transcriptase subunit of telomerase was introduced into normal
Dividing cells
Normal cells expressing TERT reverse transcriptase
Normal cells
0 0
20
40
60
80
Population doublings FIGURE 7.16 INTRODUCTION OF hTERT INTO NORMAL CELLS IS SUFFICIENT TO OVERCOME THE SENESCENCE LIMIT AND IMMORTALIZE THE CELLS. Following expression of hTERT, the human reverse transcriptase subunit of telomerase, cells act like normal cells (they are not transformed into cancer cells), but they can grow indefinitely. TERT, telomerase reverse transcriptase.
cells growing in culture. This caused an increase in the level of active telomerase with dramatic results. Instead of undergoing senescence, these cells kept dividing in culture, apparently indefinitely (Fig. 7.16). However, unlike cancer cells, which are also immortal, these cells did not acquire the ability to cause tumors. Thus, this experiment showed convincingly that telomeres are part of a mechanism that regulates the proliferative capacity of somatic cells. ACKNOWLEDGMENTS We thank Beth Sullivan, Rachel O’Neill, Vladimir Larionov, Maurizio Gatti, and Lea Harrington for their advice during revision of this chapter. SELECTED READINGS Aitman TJ, Boone C, Churchill GA, et al. The future of model organisms in human disease research. Nat Rev Genet. 2011;12:575-582. Beck CR, Garcia-Perez JL, Badge RM, Moran JV. LINE-1 elements in structural variation and disease. Annu Rev Genomics Hum Genet. 2011;12:187-215. Birchler JA, Gao Z, Sharma A, Presting GG, Han F. Epigenetic aspects of centromere function in plants. Curr Opin Plant Biol. 2011;14: 217-222. Bloom KS. Centromeric heterochromatin: the primordial segregation machine. Annu Rev Genet. 2014;48:457-484. Doolittle RF. Microbial genomes opened up. Nature. 1998;392: 339-342. Fukagawa T, Earnshaw WC. The centromere: chromatin foundation for the kinetochore machinery. Dev Cell. 2014;30:496-508. Gent JI, Dawe RK. RNA as a structural and regulatory component of the centromere. Annu Rev Genet. 2012;46:443-453. Heidenreich B, Rachakonda PS, Hemminki K, Kumar R. TERT promoter mutations in cancer development. Curr Opin Genet Dev. 2014;24: 30-37. Huang CR, Burns KH, Boeke JD. Active transposition in genomes. Annu Rev Genet. 2012;46:651-675. Martínez P, Blasco MA. Replicating through telomeres: a means to an end. Trends Biochem Sci. 2015;40:504-515. Palm W, de Lange T. How shelterin protects mammalian telomeres. Annu Rev Genet. 2008;42:301-334. Schueler MG, Sullivan BA. Structural and functional dynamics of human centromeric chromatin. Annu Rev Genomics Hum Genet. 2006;7: 301-313. Simonti CN, Capra JA. The evolution of the human genome. Curr Opin Genet Dev. 2015;35:9-15. Smit AF. Interspersed repeats and other mementos of transposable elements in mammalian genomes. Curr Opin Genet Dev. 1999;9: 657-663. Stanley SE, Armanios M. The short and long telomere syndromes: paired paradigms for molecular medicine. Curr Opin Genet Dev. 2015;33:1-9. Yan H, Jiang J. Rice as a model for centromere and heterochromatin research. Chromosome Res. 2007;15:77-84. The 1000 Genomes Project Consortium. A global reference for human genetic variation. Nature. 2015;526:68-74.
CHAPTER
8
DNA Packaging in Chromatin and Chromosomes E
ukaryotic chromosomal DNA molecules are thousands of times longer than the diameter of the nucleus and therefore must be highly compacted throughout the cell cycle. This compaction is accomplished by combining the DNA with structural proteins to make chromatin. Chromatin folding must compact the DNA but still permit access of the transcriptional machinery to regions of the chromosome required for gene expression. The first level of folding involves coiling DNA around a protein core to yield a nucleosome. The string of nucleosomes, known as a 10-nm fiber, shortens DNA approximately sevenfold relative to naked DNA. In some specialized cell types this is further condensed into a 30-nm fiber that shortens the DNA six- to sevenfold more. However, it appears that in most cells the further folding of the 10-nm fiber involves coils and looping, and is remarkably irregular and dynamic.
First Level of Chromosomal DNA Packaging: The Nucleosome The continuous DNA fiber of each chromosome is packaged into many hundreds of thousands of nucleosomes linked in series. Individual nucleosomes can be isolated following cleavage of DNA between neighboring parti cles by DNA-cutting enzymes called nucleases. Random digestion of chromatin initially yields a mixture of particles consisting of short chains of nucleosomes containing multiples of approximately 200 base pairs of DNA (Fig. 8.1). Continued nuclease cleavage yields a stable particle with 146 base pairs of DNA (1.75 turns of the DNA wrapped around the protein core). This is called a nucleosome core particle. The nucleosome core particle is disk-shaped, with DNA coiled in a left-handed superhelix around an octamer of core histones. This octamer consists of a central tetramer composed of two closely linked H3:H4 heterodimers, flanked on either side by two H2A:H2B
heterodimers. High-resolution crystal structures of nucleosome core particles revealed that each core histone has a compact domain of 70 to 100 amino acid residues that adopts a characteristic Z-shaped “histone fold” consisting of a long α-helix flanked by two shorter α-helices (Fig. 8.2). The amino-terminal approximately 30 amino acid residues of the core histones (referred to as N-terminal tails) are important for interactions both inside and outside the nucleosome. They project outward from the cylindrical faces of the nucleosomal core as well as between the adjacent winds of the DNA on the nucleosome surface. Although these N-terminal tails are not ordered either in crystals of nucleosome core particles or in solution, they are among the most highly conserved regions of these very highly conserved proteins. This is because they serve as signaling platforms and mediate packing interactions between nucleosomes. Modifications of the N-terminal tails regulate DNA accessibility within the chromatin fiber to the transcription, replication, and repair machinery.
Chromatin Modifications and Regulation of Chromatin Function The discovery that the sequence of bases in DNA provides a code to specify the primary structure of proteins triggered a revolution that culminated 50 years later with the near-complete sequencing of the human genome. To fully exploit this coding information, cells must control when to use it. Initial studies of the processes controlling gene expression focused on regulation of transcription by proteins that recognize specific DNA sequences at the 5′ end of genes (see Chapter 10), as this is how bacteria regulate gene expression. Eukaryotic gene regulation is much more elaborate. Human nuclei contain roughly 3.3 × 107 nucleosomes distributed along the DNA. Although more than 70% of the molecular surface of nucleosomal DNA is accessible 123
124
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
A
C
B
12 nm
200 bp
166 bp 146 bp 5.7 nm 166 bp = 2 full turns 146 bp = nucleosome core DNA
Increasing digestion with nucleases
FIGURE 8.1 NUCLEOSOMES. A, Electron micrograph showing chromosomal loops covered in nucleosomes, which look like beads on a string. B, Nuclease digestion of chromosomes releases fragments containing varying numbers of nucleosomes (left) in which the DNA fragments vary by multiples of 200 base pairs (center). More extensive nuclease digestion results in production of the nucleosome core particle, with 146 base pairs of DNA (right). C, Crystal structure of a nucleosome core particle. The DNA wraps around a compact core of histones. (A, Courtesy William C. Earnshaw. B, Left panel, modified from Woodcock CL, Sweetman HE, Frado LL. Structural repeating units in chromatin. II: Their isolation and partial characterization. Exp Cell Res. 1976;97:111–119. B, Center and right panels, modified from Allan J, Cowling GJ, Harborne N, et al. Regulation of the higher-order structure of chromatin by histones H1 and H5. J Cell Biol. 1981;90:279–288. C, For reference, see Protein Data Bank [PDB; www.rcsb.org] file 1KX5.)
A
H3 H4
H2A H2B
B
2 nm FIGURE 8.2 SECONDARY STRUCTURE OF THE HISTONES WITHIN THE CORE PARTICLE. A, A ribbon diagram shows that each histone protein in the octameric core of the nucleosome has a characteristic Z-shaped α-helical structure (the histone-fold). The flexible N-terminal portions of the histones, which have a critical role in regulating chromatin structure, did not occupy a unique location in the crystal and do not appear in this structure. B, The histone octamer surrounded by one of the two turns of DNA. (Modified from PDB file 1KX5 and Luger K, Mäder AW, Richmond RK, et al. Crystal structure of the nucleosome core particle at 2.8 Å resolution. Nature. 1997;389:251–260.)
CHAPTER 8 n DNA Packaging in Chromatin and Chromosomes
A. Histone modifications create chromatin states E M R EDITOR
MARK
READER
C
diverse activities
chromatin state
Polycomb chromatin
Transcription Heterochromatin
Transcription Polycomb: inhibits RNA Pol II elongation M M M M H3 M M M M KAARKSA QLAT PATG K H2N A R T K Q T A R K S T G U R H2A P GVKKPHR GKA A A A OOH A TESHH A KAKGK C KK A Mitotic Mitotic Mitotic eraser eraser M eraser M SGRGK H4 Ac KRHRKVLR ε-N-Acetyl lysine GG A D KGLGKGG A A O-Phospho serine A A
B
A
M (Mono, di, tri) methyl lysine M (Mono, di) methyl arginine U Mono-ubiquitin
C
125
+
+
H2N P E P SKS A
Lysine
A M
H2B APAPKKGSKKAITKAQKKDGKKR KRSRK A A A A A GKARAKAK H2N S G RGKQG
U VKYTS
SK
COOH
Tri-methyl lysine
FIGURE 8.3 HISTONE MODIFICATIONS. A, E → M → R → C pathways use posttranslational modifications to create different chromatin states. B, Modification of the amino- and carboxyterminal domains of the histones regulates nucleosome assembly, transcription, and mitotic chromosome condensation. Highlighted here are methylations of three lysines, which are associated with transcription, heterochromatin, and facultative heterochromatin respectively. Note that each residue is immediately adjacent to a residue phosphorylated in mitosis, which knocks the READER off the methylation mark. The modifications are described in the figure key. C, Structure of tri-methyl lysine. For other structures of modified amino acids see Fig. 3.3. Arginine, R; lysine, K; serine, S. (Modified from PDB file 1KX5 and Khorasanizadeh S. The nucleosome: from genomic organization to genomic regulation. Cell. 2004;116:259–272.)
to solvent, most nonhistone proteins involved in gene regulation bind nucleosomal DNA 10- to 104-fold less well than naked DNA. Thus, nucleosomes establish a general environment in which DNA replication and gene transcription are repressed unless signals are given to the contrary. The access of proteins to DNA in chromatin is regulated both by the density and specific localization of nucleosomes, and by specific modifications of the histones. The histones are acted on by enzymes we will call EDITORS (Fig. 8.3). EDITORS either place a MARK (a posttranslational modification) often, but not exclusively, on the histone N-terminal tail or remove an existing MARK. READERS then bind specifically to the MARK and recruit a variety of other activities. In some cases MARKs can act directly by influencing the charge properties of chromatin. The net result is the creation of a specific CHROMATIN STATE, of which two examples are: “open for transcription” and “inaccessible to transcription factors.” It has been proposed that the combination of MARKS on histones makes a kind of “code” that specifies the activity of various chromatin regions. This is disputed, however, as depending on context, individual MARKS can recruit different READERS with very different outcomes. Thus, if there is a code, it is far from simple and the significance of many of the histone MARKS remains to be deciphered. It has also been widely proposed that histone MARKS, together with methylation of the DNA itself are the basis of epigenetic regulation (see Fig.
7.11): the stable, heritable regulation of chromosomal functions by information that is not encoded in the DNA sequence. DNA methylation can be propagated through many cell divisions, but it is less clear that histone modifications are normally propagated in this way. Thus the role of histone modifications in epigenetic memory should be regarded as a popular hypothesis rather than an accepted fact.
Regulation of Chromatin Structure by the Histone N-Terminal Tails The N-terminal histone tails provide a molecular “handle” to manipulate DNA accessibility in chromatin. This complex area can only be outlined here. A wide range of MARKS has been identified at many sites in the histone N-terminal tails and elsewhere (Fig. 8.3; see Cell SnapShot 1). These modifications include acetylation, methylation and ubiquitination of lysine residues, phosphorylation of serine and threonine, and poly(ADP) ribosylation. Histones with acetylated lysines are generally associated with “open” chromatin that is permissive for RNA transcription, while histones with methylated lysines can be associated with either “open” or “closed” chromatin states. Because the histone modifications are read as combinations, individual modifications do not necessarily always have the same consequences. One example of this is the phosphorylation of histone H3 on serine 10 (H3-S10ph). In mitotic cells, this correlates with
126
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
chromatin compaction, but when combined with acetylation of surrounding amino acid residues, it can also be associated with the activation of gene transcription as nonproliferating cells reenter the cell cycle (see Chapter 41). During mitosis, phosphorylation of threonine 3, serine 10, and serine 28 disrupts the binding of READERS to methylation MARKS on lysines 4, 9, and 28, respectively (Fig. 8.3B). Thus, one MARK can regulate the activity of an adjacent MARK. Acetylation involves the transfer of acetate groups from acetyl coenzyme A to the ε-amino groups of lysine. This reduces the net positive charge of the N-terminal domain, causing chromatin to adopt an “open” conformation that is more favorable to transcription. The acetylation MARK acts as a binding site for protein READERs, one example of which is an approximately 100-aminoacid sequence motif called a bromodomain. Various bromodomain-containing READERS recruited to chromatin by acetylated histones often further modify histones in other ways that either promote or limit the accessibility of the DNA for transcription into RNA. Proteins called transcription factors regulate gene expression by binding specific DNA sequences and recruiting the transcriptional machinery (RNA polymerases and associated proteins) to activate gene expression (see Fig. 10.12). Many transcription factors recruit a protein complex, called a coactivator, that facilitates
Binding site
loading of the transcriptional apparatus onto the gene. Often, coactivators possess domains that recognize histone MARKS and have EDITOR activities to lay down new MARKS on N-terminal histone tails. For example, the yeast SAGA complex contains over 10 proteins, including READERS that recognize histone methylation and acetylation. It also has an EDITOR activity that removes the protein MARK ubiquitin (see Fig. 23.2) from target proteins plus a histone acetyltransferase EDITOR activity that acetylates lysine-14 and lysine-8 in the N-terminal tails of histone H3 (Fig. 8.4). Histone acetylation is dynamic. Just as transcriptional coactivators contain histone acetyltransferases that add acetyl groups to nucleosomes and promote gene activation, so corepressors, which are recruited in an analogous manner, can contain histone deacetylases that remove acetyl groups from selected lysine residues. Deacetylation tends to repress gene expression and is one strategy used to regulate cell-cycle progression during the G1 phase of the cell cycle (see Fig. 41.9). Histone acetylation is crucial for life. Yeast cells die if certain key lysines are mutated to arginines, thus preserving their positive charge but preventing them from being acetylated. In addition to marking nucleosomes by modification of their N-terminal tails, cells also use the energy provided by adenosine triphosphate (ATP) hydrolysis to actively remodel nucleosomes. This involves complex
Gene off TF
TATA
Nucleosome Histone N-terminal tails RNA polymerase
Transcription factor
Gene transcribed TF
TATA
AC
AC Histone acetyltransferase (HAT)
AC
AC
AC
AC
Other subunits? Spt8 Ada2 Ada3 GCN5
A simple HAT complex from yeast
Spt7
Ada2 Spt3 GCN5
Ada3 Spt20
A complicated HAT complex from yeast
FIGURE 8.4 Transcription factors (purple) bind specific DNA sequences and recruit coactivators to the 5′ ends of genes. Many of these coactivators work by acetylating the N-terminal tails or body of the core histones, thereby loosening the chromatin structure and promoting the binding and activation of the RNA polymerase holoenzyme (see Chapter 10). The coactivators vary in composition and complexity from relatively simple histone acetyltransferase complexes (bottom left) to the huge and elaborate SAGA complex (bottom right). In this side view, only one of the two turns of DNA around the nucleosome is seen. GCN5, Ada2, Ada3, Spt3, Spt7, Spt8, and Spt20 are the names of budding yeast genes whose products are found in these complexes. AC, acetylation; TATA, DNA sequence in the gene promoter [see Chapter 10]).
CHAPTER 8 n DNA Packaging in Chromatin and Chromosomes
protein “machines” that include a catalytic subunit that couples ATP hydrolysis to DNA translocation. All eukaryotes possess approximately 20 different classes of these chromatin remodeling enzymes. These different subclasses are capable of directing a range of different changes to nucleosome organization. For example, some enzymes reposition nucleosomes so that they are evenly spaced along DNA. Others remove histones from DNA. Still others direct replacement of core histone proteins with specialized variants.
Histone Deposition During Nucleosome Assembly During DNA replication, existing nucleosomes are partitioned randomly between daughter DNA strands. Newly assembled nucleosomes then fill the gaps. When not associated with DNA, histones are always bound to protein chaperones. Newly translated H3 and H4, which are acetylated on lysine-9 of H3 and lysine-5 and lysine12 of H4, associate with a chromatin assembly factor, called CAF1. One of the three subunits of CAF1 is a chaperone called retinoblastoma-associated protein of 48 kD (RbAp48). CAF1 is targeted to sites of DNA replication by interaction with proliferating cell nuclear antigen (PCNA), a doughnut-shaped protein that encircles the DNA and helps DNA polymerase slide along it during replication (see Fig. 42.12). Thus, CAF1 delivers newly synthesized histones to sites on the chromosome where new nucleosomes are required as DNA is synthesized during the S phase of the cell cycle (see Chapter 42). H3 and H4 are deposited first on the new DNA, followed by two H2A:H2B heterodimers to complete the assembly of the nascent nucleosome. Histone Variants Approximately 75% of histone H3 in chromatin is deposited during DNA replication by CAF1. The remaining 25% is a special isoform of H3, called H3.3, that is encoded by a different gene and deposited on chromatin by a different mechanism. Histone H3.3 is transcribed throughout the cell cycle and is not coordinated with DNA synthesis. Newly synthesized H3.3 binds to the RbAp48 chaperone, but they then associate with a protein called histone regulator A (HIRA) instead of the two CAF1 subunits. Some H3.3 assembles into nucleosomes at the time of DNA replication, just like the canonical H3. However, H3.3 can also be inserted into chromatin at other times of the cell cycle. For example, the HIRA–RbAp48 complex swaps H3.3/H4 dimers for H3/H4 dimers in chromatin during transcription, when the nucleosomes on the underlying gene are transiently perturbed. Although demethylases can remove the methyl groups from histone H3, replacement of histone H3 methylated on lysine 9 (H3-K9me) with unmethylated H3.3 is an efficient way to convert “closed” chromatin, where transcription is disfavored, into “open” chromatin that is favorable for transcription.
127
Other specialized histone variants also contribute to the microdiversity of chromatin. For example, the H3 isoform CENP-A is a key component of the kinetochore, the structure that assembles at centromeres to promote chromosome segregation during mitosis (see Fig. 8.21 below). The largest number of variant forms has been described for H2A. Interaction between the N-terminus of H4 and an acidic patch on the surface of H2A on the adjacent nucleosome has an important role in promoting chromatin fiber compaction. Therefore, altering the local H2A composition, which influences the strength of this interaction, provides another way to vary the accessibility of the DNA for gene expression. One variant, H2AX, which constitutes approximately 15% of the cellular H2A, helps maintain genome integrity. At sites of DNA damage, H2AX is rapidly phosphorylated by protein kinases. This serves as a MARK for the assembly of multiprotein complexes that signal and repair the damage (see Box 43.1).
Linker DNA and the Linker Histone H1 When examined by electron microscopy at low ionic strength, nucleosomal chromatin resembles a string of 10 nm diameter beads with linker DNA extended between adjacent nucleosomes (Fig. 8.1). Each nucleosome in chromosomes is typically associated with approximately 200 base pairs of DNA. Subtracting 166 base pairs for two turns around the histone octamer leaves 34 base pairs of linker DNA between adjacent nucleosomes. Linker DNA can vary widely in length in different tissues and cell types. A fifth histone, H1 or linker histone, binds to linker DNA where the DNA molecule enters and exits the nucleosome (Fig. 8.5). H1 histones have a “winged helix” central domain flanked by unstructured basic domains at both the N- and C-termini (Fig. 8.5). Mammals have at least eight variant forms (called subtypes) of H1 histones (H1a–e, H10, H1t, and H1oo). The amino acid sequences of these variants differ by 40% or more. H10 is found in cells entering the nondividing G0 state (see Chapter 41), whereas H1t and H1oo are found exclusively in developing sperm and oocytes, respectively. The role of H1 linker histone in chromatin remains enigmatic. The protein was originally assumed to regulate chromatin compaction, yet it is mobile in the nucleus, spending no more than a few minutes at any given location. Deletion of the sole linker histone genes from yeast and Tetrahymena (a ciliated protozoan) causes no obvious ill effects, but H1 is essential in mice. Although genes that encode individual H1 isoforms can be deleted in mice, simultaneous deletion of the genes for three isoforms causes embryonic death, apparently the consequence of alterations in chromatin structure that perturb normal patterns of gene expression.
128
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
Nucleosome model
C
A
Linker H1 histone
B. Bright field micrograph Heterochromatin
Euchromatin
C
Inactive X
N
H1 C. Fluorescence micrograph
D. Computer 3D reconstruction
Inactive X N
FIGURE 8.5 THE BINDING SITE OF HISTONE H1 ON THE NUCLEOSOME, NEAR THE SITE WHERE DNA STRANDS ENTER AND EXIT THE CORE PARTICLE. Orange, DNA; blue, H3; purple, H4; red, H2A; yellow, H2B. (For reference, see PDB files 1KX5 and 1HST.)
Functional Compartmentation of Chromatin: Heterochromatin and Euchromatin Chromatin has traditionally been categorized into two main classes based on structural and functional criteria. Euchromatin contains almost all genes, both actively transcribed and quiescent. Heterochromatin is transcriptionally repressed and is generally more condensed than euchromatin (Fig. 8.6). Heterochromatin was initially recognized because it stains more darkly with DNA-binding dyes than the remainder of the interphase nucleus. More recent analyses based on mapping patterns of the modifications of the histone N-terminal tails now suggest that there are at least five classes of chromatin environments in nuclei. These classes, defined some what arbitrarily and given colors for names, are green (classic heterochromatin with HP1; described later), yellow (active chromatin rich in H3-K4me3), red (active chromatin rich in histone remodelers), blue (facultative heterochromatin repressed by polycomb proteins), and black (repressed, but not via HP1). Another
FIGURE 8.6 EUCHROMATIN AND HETEROCHROMATIN. A, Electron micrograph of a thin section of a plasma cell nucleus. Euchromatin is decondensed. Heterochromatin (mostly clumped near the nuclear envelope and central nucleolus) remains condensed. B, Light micrograph of a female nucleus with four Barr bodies (arrows) (facultative heterochromatin composed of the inactive X chromosome). This woman has a highly abnormal genetic makeup, with five X chromosomes. The X chromosome inactivation system has a built-in counting mechanism that ensures that only one X chromosome remains active. C–D, The Barr body is structurally distinct from the active X chromosome. This figure is from a three-dimensional study in which the X chromosome was identified by in situ hybridization “painting” with probes that covered the entire chromosome. C, One slice through the three-dimensional data set. D, Two different views of the X chromosomes reconstructed in three dimensions. The inactive X chromosome is shown in red. Because of X chromosome inactivation, females are mosaic for functions encoded on the X chromosome. Each female embryo has two X chromosomes: Xpat and Xmat (for paternal and maternal). Following X chromosome inactivation, some cells will express genes from Xpat and others will express genes from Xmat. The inactivation is permanent; eg, all progeny of a cell with Xpat inactivated will also have Xpat inactivated. This inactivation occurs randomly in different cells of the embryo. In cats, genes responsible for coat color are encoded on the X chromosome. The patchy color pattern of calico cats reflects the underlying pattern of X chromosome inactivation. All classic calico cats are females. (A, From Fawcett DW. The Cell. Philadelphia: WB Saunders; 1981. B, Courtesy Barbara Hamkalo, University of California, Irvine. C–D, From Eils R, Dietzel S, Bertin E, et al. Three-dimensional reconstruction of painted human interphase chromosomes: active and inactive X chromosome territories have similar volumes but differ in shape and surface structure. J Cell Biol. 1996;135:1427–1440.)
approach has identified chromatin regions associated with the inner surface of the nuclear envelope. These lamina-associated domains tend to average approximately 106 base pairs in size and are mostly transcriptionally inactive.
129
CHAPTER 8 n DNA Packaging in Chromatin and Chromosomes
A typical nucleus has both euchromatin and heterochromatin, the latter often being concentrated near the nuclear envelope and around nucleoli. Much of the nuclear interior is occupied by pale-staining euchromatin rich in actively transcribing genes. Nuclei with low transcriptional activity have relatively more hetero chromatin. Classically two types of heterochromatin, constitutive and facultative, have been recognized. Constitutive heterochromatin is typically associated with repetitive DNA sequences, such as satellite DNAs (see Fig. 7.9), that are packaged into “closed” (green) chromatin in every cell type. In at least some cases, establishment of constitutive heterochromatin involves transcription of those repeated DNA elements to produce double-stranded RNAs that are cleaved into short fragments by the RNA interference (RNAi) machinery (see Fig. 11.13). The resulting short RNAs are thought to target activities that promote heterochromatin formation to their sites of transcription in the chromosome (see Fig. 11.14). The MARK that best defines constitutive heterochromatin is H3-K9me3 (Fig. 8.7), which is recognized by the READER heterochromatin protein 1 (HP1). The HP1 amino-terminus contains a 50-amino-acid motif called a chromodomain (chromatin modification organizer) that binds to H3-K9me3. Surprisingly, most HP1 is highly mobile in nuclei, moving on a time frame of seconds. It also binds and recruits enzymes that tri-methylate histone H3 lysine 9. HP1 can promote the lateral spreading of heterochromatin along the chromosome by recruiting other proteins that further modify the histone aminoterminal tails (Fig. 8.7). For example, the enzyme that trimethylates H3-K9 itself binds to HP1 and can modify adjacent nucleosomes. As a result, heterochromatin is not a static “closed” chromatin compartment but can “invade” nearby genes along the chromosome. If a chromosomal rearrangement moves an actively transcribed gene close to constitutive heterochromatin, heterochromatin may spread across it and repress transcription (Fig. 8.7). This is called position effect. HP1 and other repressive proteins can also recruit DNA methyltransferases that modify the underlying DNA by adding a methyl group to the 5′ position on cytosine in the dinucleotide CpG (cytosine phosphate guanine). Methylation can recruit READERS that inactivate gene transcription if it occurs near the 5′ promoters of genes (see Chapter 10) in regions with an above average concentration of CpG called CpG islands. Among the several binding proteins that recognize DNA containing 5-methyl-cytosine, methyl-cytosine binding protein (MeCP2) can repress expression of nearby genes by recruiting a histone deacetylase complex that removes acetyl groups from the core histone N-terminal tails (Fig. 8.7). MeCP2 is highly abundant in neurons, and mutations in the protein cause Rett syndrome, an X-linked
A. Gene translocation displayed on mitotic chromosome Chromosome breakage and rejoining Constitutive heterochromatin
B
Gene
Heterochromatin with bound HP1
M3
M3
Open, transcribed chromatin
A
M3
HP1 recruits histone deacetylase (HDAC)
M3
M3
M3
M3
A
Open, transcribed chromatin
A
M3
HDAC deacetylates H3 lysine 9
A
A
A
A
A A
M3
HP1 recruits Suv 39
M3
M3
M3
Suv 39 trimethylates H3 lysine 9
M3
M3
M3
M3
M3
M3
Heterochromatin spreads over entire region Gene off
M3
M3
M3
M3
M3
M3
M3
M3
FIGURE 8.7 POSITION EFFECT AND THE SPREADING OF HETEROCHROMATIN. A, If a transcriptionally active gene is moved next to a region of heterochromatin, it may be repressed as the heterochromatin spreads. The relative position of the gene is shown on mitotic chromosomes as it would be determined by in situ hybridization (Fig. 8.10). B, Diagrammatic representation of stages in the spreading of heterochromatin and silencing of the active gene: removal of acetyl groups from the histones by a histone deacetylase; addition of two or three methyl groups to lysine 9 of H3; and binding of HP1, which recruits a DNA methyltransferase plus other heterochromatin proteins to create heterochromatin. A, acetylation; Me, methylation.
neurodevelopmental disorder in which female infants develop apparently normally for 6 to 18 months, but then regress, losing language and adopting stereotypical postures and movements. How the MeCP2 defects lead to Rett syndrome is not known. Facultative heterochromatin consists of sequences that are in heterochromatin in some cell types and in euchromatin in others. X chromosome inactivation is
130
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
the classic example of facultative heterochromatin in mammals. In females, one X chromosome in each cell (selected at random) is inactivated early in development prior to implantation of the embryo. The inactivated X chromosome forms a discrete patch of heterochromatin at the nuclear periphery known as the Barr body (Fig. 8.6). Because most genes carried on the inactivated X chromosome become transcriptionally silent, females with two X chromosomes have the same levels of X chromosome-linked gene expression as males with a single X chromosome. Polycomb group proteins form facultative heterochromatin by modifying histones. They were identified in Drosophila as mutants in which particular body segments “forgot” their identity during development due to reactivation of the expression of several homeodomain transcription factors (see Fig. 10.14). Drosophila polycomb chromatin apparently locks genes that have been switched off in a silent epigenetic state that is stable through many generations of cell division. Two PRCs (polycomb repressive complexes) regulate transcription. The PRC2 complex initiates silencing by tri-methylating histone H3 on lysine 27 (H3-K27me3). Then chromodomain-containing members of the PRC1 complex bind specifically to H3-K27me3 (note the difference from the HP1 chromodomain, which READS H3-K9me3). PRC1 contains an E3 ubiquitin ligase (see Fig. 23.3) that transfers a single ubiquitin molecule to lysine 119 of histone H2A (H2A-K119ub). PRC1 binding also causes nucleosomes to form dense clumps that are resistant to remodeling and “opening” by ATP-dependent remodeling “machines.” Polycomb group proteins also function in X chromosome inactivation, in stem cell maintenance, and possibly in cancer stem cells. In mammals, the inactive X chromosome expresses a large (15 kb) noncoding RNA called XIST that associates with and “coats” the inactive X chromosome. Next, the PRC2 complex associates with the inactive X, transiently recruiting PRC1, which produces H2A-K119ub. This, together with low levels of histone acetylation, enrichment for the H2A variant macroH2A, and high levels of CpG methylation in many CpG islands combines to inhibit transcription of most genes. Several polycomb group proteins are required for the self-renewal of blood and neural stem cells, and may also perform a similar function in cancer stem cells. In stem cells, polycomb group proteins regulate transcription of factors that control the cyclin-dependent kinases that drive cell-cycle progression (see Chapter 40). They may also participate in the DNA damage response (see Chapter 43).
Imprinting: A Specialized Type of Gene Silencing The factors that produce heterochromatin are also involved in a very specific type of gene silencing known
as imprinting. An imprinted gene is stably turned off during formation of the egg or sperm. For example, if the maternal copy of a gene is imprinted, then expression can come only from the corresponding homologous chromosome contributed by the father. Currently, approximately 80 imprinted genes are known. One well-studied imprinting system involves the genes for insulin-like growth factor-2 (IGF2) and a noncoding RNA H19 in the mouse (Fig. 8.8). The DNA between these genes has an insulator element with binding sites for the CCTC-binding factor CTCF. Binding of CTCF to the insulator differs, depending on whether the chromosome is derived from the egg or sperm. On the maternal chromosome, it allows the H19 gene to be expressed but turns off the IGF2 gene by preventing access to a transcriptional enhancer. On chromosomes derived from the sperm, methylation of CpG sequences in the control region stops CTCF from binding. As a result, the paternal copy of IGF2 has access to its enhancer and is expressed, but the H19 gene is not expressed. This simple switch ensures that the offspring expresses only the paternal copy of the IGF2 gene and the maternal copy of H19.
Higher-Order Structure of Chromosomes Higher Levels of Chromosomal DNA Packaging in Interphase Nuclei Levels of chromatin structure beyond the nucleosome are poorly understood. This lack of clarity arises in part because dense packing of macromolecules in the nucleus makes it difficult to observe the details of higher-level folding of chromatin fibers directly. For more than 35 years the accepted dogma was that the next level of chromatin compaction beyond the 10-nm fiber was a solenoidal 30-nm fiber. Recent results, primarily using cryoelectron microscopy, now strongly question the existence of the 30-nm fiber in vivo in most cells. Visualization of specific DNA loci within fixed interphase nuclei by in situ hybridization (introduced in Fig. 8.10) can be used to estimate the degree of chromatin compaction by comparing the physical distance between two DNA sequences with a known number of base pairs between them. For regions of DNA up to approximately 250,000 base pairs apart, the chromatin fiber is shortened approximately 80- to 100-fold. When sequences are separated by tens of millions of base pairs, the shortening increases by another 20- to 30-fold. This suggests at least two levels of chromatin folding beyond the 10-nm fiber. The organization of chromatin fibers can be observed by superresolution fluorescence microscopy of living cells after labeling with a fluorescent marker, such as the jellyfish green fluorescent protein (GFP [see Fig. 6.3]) (Fig. 8.9). Individual nucleosomes are locally dynamic, changing their packing and locations as cells traverse the
CHAPTER 8 n DNA Packaging in Chromatin and Chromosomes
A. Maternal allele during oogenesis
B. Paternal allele during spermatogenesis
Enhancer DNMT
Enhancer DNMT
M
ICR
H19
CTCF binding creates boundary blocking the enhancer from accessing IGF2
M MM
CTCF IGF2
131
IGF2
M ICR H19
CTCF can't bind methylated IRC and now the enhancer can reach IGF2 Enhancer MM
IGF2
ICR
H19
Enhancer instead activates the much closer H19 gene
IGF2
M ICR H19
Methylation of the H19 promotor leaves it in a heterochromatinlike state
CTCF MM ICR
IGF2
H19
IGF2
M
M
H19 ICR
FIGURE 8.8 IMPRINTING OF THE INSULIN-LIKE GROWTH FACTOR-2 AND H19 LOCI. A, During oogenesis, CTCF binding to the imprinting control region (ICR) prevents methylation of the DNA. In the zygote, this methylated chromosome from the mother is bound by CTCF, which acts as an insulator, blocking the IGF2 (insulin-like growth factor-2) gene from gaining access to its enhancer. As a result, the maternal chromosome expresses H19 and not IGF2. B, During spermatogenesis, the ICR is methylated. In the zygote, the ICR on the chromosome derived from the father cannot bind CTCF. As a result, the IGF2 gene gains access to its enhancer and is expressed. The H19 gene is off.
A
B
5 µm Single nucleosomes (PALM)
Reconstruction from ~100,000 nucleosomes
FIGURE 8.9 SUPERRESOLUTION VISUALIZATION OF NUCLEO SOMES IN A LIVING HUMAN CELL. This experiment uses a clone of HeLa (Henrietta Lacks) cells expressing photoactivatable red fluorescent protein (mCherry) linked to histone H2B. This form of red fluorescent protein usually has almost no fluorescence, but some fractions become highly fluorescent by spontaneous activation. A, When imaged by photoactivated localization microscopy (PALM) ultraviolet laser microbeam, individual nucleosomes can be observed. B, When an image corresponding to approximately 100,000 nucleosomes is reconstructed, the nucleus is seen to be organized into chromatin domains. (Courtesy K. Maeshima, National Institute of Genetics, Japan.)
cell cycle. Some electron microscopy studies observed a fiber, 100 to 300 nm in diameter, which was called a chromonema fiber. In most studies, however, the chromatin appears to be relatively disordered. Together all these analyses are leading to a view of interphase chromatin as composed of irregular 10-nm chromatin fibers that are organized in dynamic loops. The 30-nm fibers and chromonema filaments may occur only under specialized conditions.
Large-Scale Structural Compartmentation of the Nucleus Although interphase nuclei lack a high degree of order, a number of general organizational principles are recognized. First, individual chromosomes tend to concentrate within discrete territories and intermingle with one another only to a limited extent. This is seen most clearly in human somatic cell nuclei when individual chromosomes are visualized by a special type of in situ hybridization called chromosome painting (Fig. 8.10). The volume of territories occupied by individual chromosomes correlates with the proportion of actively transcribing genes. In some cases, active genes are
132
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
A
B
C
Chromosomal DNA
Probe
D 5,000 bp
Probe
E 13
11
2
16
18
1 3
6 15
5
1
6
FIGURE 8.10 FLUORESCENCE IN SITU HYBRIDIZATION REVEALS THAT CHROMOSOMES OCCUPY DISCRETE TERRITORIES IN INTERPHASE NUCLEI. A, Chromosomes are spread on a slide as in Fig. 8.15. Following chemical fixation steps to preserve the chromosomal structure, the chromosomal proteins are removed by digestion with proteases and the genomic DNA strands are melted (separated) by heating. Next, a “probe DNA” (yellow) is added. This probe DNA is single-stranded so that it can base-pair (hybridize) to its complementary sequences in the chromosome. The probe DNA is chemically labeled with biotin. Next, the sites of hybridization on the chromosomes are detected with fluorescently labeled avidin, a protein from egg white that binds to biotin with extremely high affinity. The sites of avidin-binding appear yellow, whereas the remainder of the chromosomal DNA is counterstained with a red dye. B, The micrograph shows fluorescence in situ hybridization (FISH) analysis using a probe from near the von Hippel–Lindau locus on chromosome 3. C, Metaphase chromosome labeled by FISH using chromosome paint probes (probes distributed all along the chromosome, excluding repetitive DNA). In this 24-color FISH image, every chromosome is marked with two or three fluorochromes (true color image). D, The same combinatorial probe was used in 24-color FISH on a fibroblast nucleus under conditions preserving the 3D architecture. Every chromosome forms distinct chromosome territory. E, Every chromosome territory of the same nuclear optical section as on B was identified and false-colored after classification. (B, Courtesy Jeanne Lawrence, University of Massachusetts, Worcester. C–E, Courtesy I. Solovei, A. Bolzer, and T. Cremer, University of Munich, LMU, Germany.)
located well outside of the territories, as though their activation involved looping out a much larger domain from the remainder of the chromosome. These movements during gene activation may involve relocation from compartments where transcription is relatively infrequent to compartments where transcription is favored (Fig. 8.11C–D). Silent chromatin tends to be concentrated near the nuclear periphery in a wide range of cell types. This supports the hypothesis that particular chromosomal regions (eg lamina-associated domains near the nuclear lamina; see Chapter 9) might have preferred locations within the nucleus. As a result, chromosomes that are rich in actively transcribed genes tend to be localized toward the nuclear interior, while chromosomes with a lower gene content tend to be found near the nuclear periphery (Fig. 8.11A–B). These positions of chromosomes are mostly established by where chromosomes are located during the exit from the previous mitosis. Most movements of the chromatin during interphase are of 0.5 µm or less. These movements likely occur within topologically associating
domains (TADs) (see the next section), while the chromosomes overall remain relatively stationary.
Special Interphase Chromosomes With Clearly Resolved Loop Structures Studies of specialized chromosomes from organisms ranging from flies to mammals originally revealed a link between chromatin loops and regulated gene expression. Loops are clearly seen in lampbrush chromosomes during meiotic prophase in oocytes of many species (Fig. 8.12A). These loops are sites of intense transcriptional activity as oocytes stockpile huge stores of the components needed for rapid cell divisions during early development of the fertilized egg. The loops are easily seen because the DNA is coated with many RNA transcripts, together with proteins that package and process them. Similar loops are present in the giant polytene chromosomes found in some tissues of Drosophila larvae. Each polytene chromosome consists of more than 1000 identical DNA molecules packed side-by-side in precise linear register. Polytene chromosomes have a complex
CHAPTER 8 n DNA Packaging in Chromatin and Chromosomes
A
133
A
C
C
D
47 42 43
44
B
46
45
D
Chromosome 1 Chromosome 20 Puff
B
Puff
E
Chromosome territory 1 Chromosome territory 20
10 µm
FIGURE 8.11 CHROMOSOME POSITION IN THE NUCLEUS CORRELATES WITH TRANSCRIPTIONAL ACTIVITY. A, Metaphase chromosome spread from a healthy donor with painted chromosomes 1 (red) and 20 (green). B, The same paint probes were used in fluorescence in situ hybridization (FISH) experiments on threedimensionally preserved fibroblast nuclei (3D-FISH): they revealed two pairs of chromosome territories. Note the more central positioning of chromosome 20 territories and the more peripheral positioning of chromosome 1 territories. C–D, The CD4 gene (green) is located in the nucleoplasm in cells where it is expressed (C) but is associated with centromeric heterochromatin in cells where it is silent (D). (A–B, Courtesy I. Solovei, A. Bolzer, and T. Cremer, University of Munich, LMU, Germany. C–D, From Lamond AI, Earnshaw WC. Structure and function in the nucleus. Science. 1998;280:547–553; and Brown KE, Guest SS, Smale ST, et al. Association of transcriptionally silent genes with Ikaros complexes at centromeric heterochromatin. Cell. 1997;91[6]:845–854.)
pattern of thousands of bands (Fig. 8.12B–D). Stress or stimulation of gene expression by hormones causes certain bands to lose their compact shape and puff out laterally. Each puff is composed of hundreds of identical, actively transcribed chromatin loop domains (Fig. 8.12E).
Chromatin Conformation Capture and Topologically Associating Domains Powerful insights into the organization of chromatin fibers in somatic cell nuclei have followed from the development of a technique called 3C (chromosome
FIGURE 8.12 CHROMATIN LOOPS IN SPECIAL INTERPHASE CHROMOSOMES. A, Phase contrast view of the left end of meiotic lamp brush chromosome 6 from the newt Notophthalmus viridescens. B–D, Domain organization of polytene chromosomes. Once Drosophila larvae reach a certain developmental stage, most cells stop dividing, and larval growth proceeds via an increase in the size of individual cells. To keep the protein synthesis machinery of these huge cells supplied with messenger RNA, DNA replication is uncoupled from cell division so that ultimately, the cells contain many times the normal complement of cellular DNA (ie, they are polyploid). In certain tissues, the numerous copies of the chromosomes are maintained in strict alignment with respect to one another, making giant polytene chromosomes, the best known of which occur in the salivary gland. B, Giant polytene chromosomes are visible within isolated salivary gland nuclei. C, A portion of a high-resolution map of the Drosophila polytene chromosomes. D, Polytene chromosome showing puffs. The inset box shows an area analogous to that used in panel E. E, Electron micrograph of puff showing transcribing DNA loops. These loops are covered with a “fuzz” corresponding to growing RNA chains coated with proteins. (A, From Roth MB, Gall JG. Monoclonal antibodies that recognize transcription unit proteins on newt lampbrush chromosomes. J Cell Biol. 1987;105:1047–1054. B, From Robert M. Isolation and manipulation of salivary gland nuclei and chromosomes. Methods Cell Biol. 1975;9:377–390. C, Courtesy Margarete Heck, University of Edinburgh, United Kingdom. D, From Andersson K, Mahr R, Bjorkroth B, et al. Rapid reformation of the thick chromosome fiber upon completion of RNA synthesis at the Balbiani ring genes in Chironomus tentans. Chromosoma. 1982;87:33–48. E, From Lamb MM, Daneholt B. Characterization of active transcription units in Balbiani rings of Chironomus tentans. Cell. 1979;17:835–848.)
134
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
A. Hi-C
B. Interpreting the Hi-C map
DNA “handcuffed” by protein crosslinks Biotin Fill in ends with biotin label
0
Regions along chromosome
Crosslink nucleus digest with restriction endonuclease
0
C A compartments
Regions along chromosome 10
B compartments
10 0
0
10
Loops
Ligate ends TAD
7
Purify and shear to break the DNA
Pull down biotin to isolate and sequence millions of biotinylated junction fragments
Topologically associated domains (TADs)
10
TAD
8
Interpret the result
5
6
9
7 10 11 12
FIGURE 8.13 HI-C REVEALS CHROMATIN FOLDING PATTERNS IN NUCLEI. A, Diagram of important steps during the Hi-C procedure. B, Example of a Hi-C map with a diagram of how it is interpreted. The numbers along the axes are arbitrary and are supplied for demonstration purposes only. Each time two sequences are linked together, a red dot is inserted in the matrix. This map contains many millions of those dots. TADS are the square groupings of dark dots which indicate regions that are often linked together. C, Hi-C reveals the presence of both very long-range compartments and topologically associating domains (TADs) in chromosomes. (C, Modified from Dekker J, Marti-Renom MA, Mirny LA. Exploring the three-dimensional organization of genomes: interpreting chromatin interaction data. Nat Rev Genet. 2013;14:390–403. Micrograph from Thoma F, Koller T. Influence of histone H1 on chromatin structure. Cell. 1977;12:101–107.)
conformation capture) and its many derivatives, including Hi-C (Fig. 8.13A). In Hi-C, cells are treated with the fixative formaldehyde, which nonspecifically crosslinks proteins to the DNA. The idea is to “handcuff” together adjacent stretches of DNA. After cleaving the DNA with a restriction endonuclease, the ends are labeled with a biotinylated nucleotide, and then ligated together. This links pieces of DNA that were captured together by the crosslinking procedure and were therefore physically close to one another in the nucleus. Importantly, these sequences may be very far apart on the chromosomal DNA or even on different chromosomes! The biotin-containing DNAs are then sequenced by highspeed parallel methods (Fig. 3.16), yielding several hundred million sequence “reads” that generate a map (Fig. 8.13B) with an approximately 100-kb resolution of all regions of chromosomes that are close to other regions of chromosomes. This analysis reveals two levels of chromatin organization: TADs and compartments (Fig. 8.13C). A TAD is a region of the chromosome—usually spanning 100,000 to 1,000,000 base pairs—whose DNA sequences are preferentially captured together, presumably, because they form a cluster of loops. The bands seen in polytene
chromosomes correspond to TADs. A defining feature is that sequences in two adjacent TADs rarely come in contact with one another even though they may be closer to one another along the DNA strand than two more distant sequences that are found within the same TAD. One current model is that TADs form because the DNA is looped locally by CTCF and the cohesin complex (see later). Hi-C maps also show longer-range interactions known as compartments. They are thought to involve interactions between many TADs, and may correspond to larger domains of euchromatin and heterochromatin. The protein CTCF (CCCTC binding factor) marks approximately 75% of TAD boundaries at binding sites that define functional elements termed insulators. These were originally identified as short DNA sequence elements that frequently separate regions with active and inactive genes. For example, an insulator region containing CTCF and rich in H3 acetylated on K9 often separates an active gene cluster from an adjacent region of heterochromatin, Acetylation of H3 blocks methylation of H3-K9, thereby providing a barrier to the spreading of heterochromatin marked with H3-K9me3. CCTF binding in the insulator can physically block the DNA from being
CHAPTER 8 n DNA Packaging in Chromatin and Chromosomes
methylated, providing another defense against the spread of heterochromatin. Other TAD boundaries correspond to housekeeping genes undergoing active transcription, or to the presence of other insulators associated with transfer RNA genes. CTCF can recruit a ring shaped complex called cohesin, which is a key architectural factor in chromosomes (Fig. 8.18). Cohesin was originally identified, because it regulates the pairing between replicated DNA molecules (sister chromatids) when cells divide. However, defects in the cohesin loading machinery cause Cornelia de Lange syndrome, a group of developmental disorders characterized by abnormalities in regulation of gene expression but (surprisingly) no dramatic effects on sister chromatid segregation during mitosis. It later emerged that cohesin is associated with up to half of all actively transcribed genes. In mammals, 50,000 to 70,000 CTCF binding sites have been mapped, most of which are actually within TADs. One prominent CTCF binding site is found within Alu SINES, the short mobile genetic elements that comprise up to 15% of the human genome (see Chapter 7). Because CTCF and cohesin are thought to function together to bring regulatory elements together with genes, this association with a mobile DNA element has been suggested to be one factor that contributed to humans developing complex patterns of gene regulation. In terms of function, it seems likely that clustering of loops in TADs may provide a mechanism to coordinate the regulation of gene expression and, possibly, DNA replication. Clusters of loops have been suggested to form active chromatin hubs associated with locus control regions, which are responsible for coordinating the expression of groups of genes. Locus control regions (LCRs) were identified, because they could influence the transcriptional activity of cloned DNA sequences in transgenic animals. When genes are introduced into cultured cells, they normally insert at random into the chromosomes. Expression of such foreign transgenes depends on the site of insertion into the host chromosome. Transgenes are usually expressed when they insert into an active chromosomal domain but repressed when they insert into an inactive region. LCRs are DNA sequences that permit transgenes to be expressed no matter where they insert into the chromosomes, suggesting that they create active chromatin hubs independent of the surrounding chromosome. LCRs typically consist of clusters of multiple short 150 to 300 base pair regions that are rich in binding sites for transcriptional regulators (see Chapter 9). Experiments in which a single LCR drives the expression of a cluster of several genes reveal that the LCR stimulates the expression of only one gene at a time. Thus, LCRs appear to work by physically associating with a gene, forming a loop in the chromatin and establishing an active chromatin hub that turns on its expression. Because cohesin
135
can encircle pairs of DNA strands, it is now thought that this complex may anchor these DNA loops.
Organization of Mitotic Chromosomes When cells divide, the chromatin is dramatically reorganized, forming mitotic chromosomes that can be segregated efficiently to daughter cells. The formation of mitotic chromosomes involves two steps; compaction of the chromatin roughly threefold and organization of each sister chromatid (the replicated DNA molecule and proteins that package it) into a robust structure that can move as a unit when cells divide. It is still not known how the chromatin fiber is organized in mitotic chromosomes. Classic hierarchical coiling models suggested that the 30-nm chromatin fiber coils on itself, reaching larger and larger diameters and higher degrees of compaction. The 30-nm fiber is now largely disbelieved, but high-resolution Hi-C data reveal that chromatin fiber coiling is an important feature of mitotic chromosome formation. Hi-C technology also reveals that TADs disappear from chromosomes as cells enter mitosis and are replaced by a more-or-less uniform distribution of approximately 80,000 to 120,000 base pair loops. A variety of microscopy experiments had previously suggested that chromatin loops containing 15,000 to 100,000 base pairs provide the structural basis for large scale chromatin compaction in mitotic chromosomes. These loops radiate outward from the central chromatid axis and can be seen when metaphase chromosomes are swelled in hypotonic solutions (Fig. 8.14C). We favor a model proposing that mitotic chromosome formation involves both hierarchical coiling and looping of the chromatin fiber. During this process, key proteins become concentrated along the axial regions of the condensing chromosome arms and stabilize the overall structure (Fig. 8.14A). The mechanism of chromatin folding in mitotic chromosomes remains an area of active investigation and controversy. Although much less ordered than polytene chromosomes, the arms of typical diploid mitotic chromosomes nonetheless have a more-or-less reproducible substructure. If mammalian chromosomes from the early (prometaphase) stage of mitosis are subjected to a staining procedure called G-banding, up to 2000 discrete bands are observed (Fig. 8.15). Although the structural basis for the bands is not known, the pattern is highly reproducible. Dark G-bands tend to be gene-poor regions relatively enriched for DNA with a low A : T content and rich in long interspersed nuclear elements (see Fig. 7.5). They tend to replicate later in S phase than light G-bands (also called R, or reverse, bands). Cytogeneticists used these highly reproducible banding patterns for many years to identify individual human chromosomes. The quasi-reproducible higher-order structure of mitotic chromosomes is also seen when specific DNA
136
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
A. Chromosome packaging
B. DNA loops (chromatin proteins removed)
Extracted metaphase chromatid
Chromosome scaffold components
Coiled chromonema fiber (chromatid) at metaphase
C. Chromatin loops Chromosome scaffold components 100-nm chromonema fiber at prophase
Loop chromatin
DNA
Nucleosomes
FIGURE 8.14 CURRENT MODEL OF MITOTIC CHROMOSOME STRUCTURE. A, Filament of nucleosomes, chromatin looping, clustering of chromatin loops into coiled fiber. Nonhistone proteins complexes (blue dots) bind and end up concentrated along the central axis of the chromatid arm. Crosslinks between these complexes create the chromosome scaffold. When chromosomes are swollen or extracted, the scaffold remains compact, and loops of chromatin or DNA radiate out from it. B, DNA loops seen in a human mitotic chromosome from which the histones had been removed. C, Human chromosome showing loop domains. (B, From Paulson JR, Laemmli UK. The structure of histone-depleted chromosomes. Cell. 1977;12:817–828. C, Courtesy William C. Earnshaw.)
sequences marked by in situ hybridization appear as pairs of spots on the sister chromatids (Fig. 8.10). The two spots are distributed approximately symmetrically, indicating that the chromatin fiber is folded similarly, though not identically, in both chromatids.
Role of Nonhistone Proteins in Chromosome Architecture Mitotic chromosomes are composed of roughly equal masses of DNA, histones, and nonhistone proteins. Early evidence suggesting that nonhistone proteins might contribute to mitotic chromosome structure came from experiments in which chromosomes were treated with nucleases to digest the DNA and extracted to
remove most chromosomal proteins, including essentially all the histones. The surviving remnant of the chromosome contained approximately 5% of the proteins and less than 0.1% of the DNA, but still looked like a chromosome (Fig. 8.16). If the DNA was not digested, loops of DNA protruded from the protein (Fig. 8.14B). The protein remnant was called the chromosome scaffold because it looked like a structural backbone for the chromosome. Indeed, chromosome scaffold preparations contain several proteins with essential roles in the structure and maintenance of mitotic chromosomes. If isolated nuclei are subjected to the procedures used to isolate mitotic chromosome scaffolds, a residual structure is also obtained. This has been termed the nuclear
CHAPTER 8 n DNA Packaging in Chromatin and Chromosomes
A
Mitotic cell with chromosomes
B
C
3 2
p 1 1 Late prophase
2 q 3 4 Mid-metaphase
Early metaphase
Mid-metaphase
Late prophase
FIGURE 8.15 CHROMOSOME BANDING REVEALS A COMPLEX AND REPRODUCIBLE MULTIDOMAIN SUBSTRUCTURE OF MITOTIC CHROMOSOME ARMS. A, Mitotic cells in a hypotonic medium are dropped onto a slide to spread the chromosomes. In G-banding, chromosomes are given harsh treatments, such as exposure to concentrated sodium hydroxide, proteases, or high temperatures, and then stained with Giemsa dye. The chromosome arms then exhibit a characteristic pattern of light and dark bands. B, Photographs of G-banded human chromosome 2 from cells in late prophase, early metaphase, and mid-metaphase. Several examples are shown for each stage, illustrating the reproducibility of the banding patterns. C, Diagram summarizing the metaphase and prophase patterns. Because G-banding patterns are reproducible, this technique provides a way to identify individual chromosomes unambiguously. This was a major factor in the development of the field of cytogenetics, which is the study of the correlation between the structure of the chromosomes and genetics. (B–C, Modified from Yunis JJ, Sawyer JR, Ball DW. The characterization of high-resolution G-banded chromosomes of man. Chromosoma. 1978;67:293–307.)
matrix or nucleoskeleton. Although the existence and function of a nuclear matrix in vivo remains controversial, some components of the mitotic chromosome scaffold (eg, cohesin and condensin; discussed here) have roles in organizing chromosome territories and
137
chromatin loops. For example, cohesin and CTCF are thought to have important roles in organizing the architecture of interphase chromatin into TADs. Members of the SMC protein family have several important roles in chromosome dynamics. The name derives from their roles in the structural maintenance of chromosomes. SMC proteins are components of multiprotein complexes, such as condensin and cohesin, that are essential for mitotic chromosome architecture, the regulation of sister chromatid pairing, DNA repair and replication, and the regulation of gene expression. The two condensin complexes are composed of two SMC proteins (SMC2 and SMC4), plus two sets of three auxiliary subunits. Each SMC polypeptide folds back on itself at a hinge region to form a long antiparallel coiledcoil. This brings together two globular domains from either end of the molecule, each with half of an ATPbinding site (Figs. 8.17 and 8.18). ATP binding causes the two globular domains to associate with one another. This association is then reinforced by binding of a straplike kleisin (from the Greek for closure) subunit. The other auxiliary subunits bind to the kleisin and appear to regulate association of the complex with DNA. Condensin I and condensin II are thought to regulate distinct aspects of mitotic chromosome architecture. Condensin has a complex role in establishing the architecture of mitotic chromosomes. Condensin I regulates the timing of chromosome condensation and has an essential role in changing the genome organization from TADs to a brush-like array of loops as chromosomes form during entry of cells into mitosis. Condensin II apparently drives the compaction of the chromosome loops along the sister chromatid axes. The cell-cycle kinase Cdk1:cyclin B (see Chapter 40) regulates condensin binding to chromosomes by phosphorylation of an auxiliary subunit. During mitosis condensin is concentrated along the central axis of chromosome arms. When condensin binds to naked DNA in a test tube, it can use the energy of ATP hydrolysis to supercoil the DNA. The cellular role of this activity is unknown, but it may contribute to changing the conformation of chromatin loops. When condensin is depleted, mitotic chromatin condenses (apparently driven by changes in histone modifications), but the resulting chromosomes are fragile and appear disorganized if condensin depletion is rapid and complete. Cohesin is the second major SMC-containing pro tein complex of interphase and mitotic chromosomes. Cohesin is a tetramer containing SMC1 and SMC3 plus two auxiliary subunits. Cleavage of the kleisin Scc1, by a protease called separase initiates sister chromatid separation in mitotic anaphase (see Fig. 44.16). Cohesin, like condensin, is a ring-like molecule (Fig. 8.18). How cohesin holds the two sister chromatids together is still debated, although it is generally thought to physically encircle two sister DNA molecules. Cohesin
138
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
Chromosome scaffolds
Soluble proteins
B
Chromosomes
A
C. Scaffold
Topoisomerase II Condensin
Micrococcal nuclease 2M NaCl extraction
Mitotic cells
Isolated chromosomes
Insoluble proteins = chromosome scaffolds
FIGURE 8.16 ISOLATION OF MITOTIC CHROMOSOME SCAFFOLDS REVEALS IMPORTANT STRUCTURAL PROTEINS. A, Diagram of the procedure used to isolate mitotic chromosomes. B, Sodium dodecylsulfate polyacrylamide gel showing proteins of isolated chromosomes, proteins extracted during scaffold isolation, and proteins of isolated scaffolds. C, Chromosome scaffold centrifuged onto a thin carbon film and rotary-shadowed with Pt : Pd (platinum : palladium). The structure, which is approximately 95% protein, retains the overall shape of the mitotic chromosome. (B–C, Courtesy William C. Earnshaw.)
B
E. Distribution of condensin SMC2 subunit in mitotic chromosomes
Complete depletion
Add purified nuclei
Mock depletion
A
Smc4 CAP-D2 Smc2 CAP-G
Deplete condensin complex with antibody Xenopus mitotic egg extract
Add back purified condensin
Add purified nuclei
15 µm
C
Chromosome condensation
D
No chromosome condensation
Restored chromosome condensation
5 µm
FIGURE 8.17 IDENTIFICATION OF THE CONDENSIN COMPLEX. A, Experimental protocol showing that condensin is required for mitotic chromosome condensation in vitro. B, Sodium dodecylsulfate (SDS) polyacrylamide gel reveals the members of the condensin complex and demonstrates that they can be depleted from egg extract using a specific antibody. C–D, Chromatin lacking condensin does not form mitotic chromosomes in vitro, and this is restored by adding back condensin. E, Immunofluorescence micrograph showing the distribution of condensin subunit SMC2 on mitotic chromosomes of the chicken. The tiny chromosomes, called microchromosomes, are normal bird microchromosomes. (A–D, From Hirano T, Kobayashi R, Hirano M. Condensins, chromosome condensation protein complexes containing XCAP-C, XCAP-E and a Xenopus homolog of the Drosophila Barren protein. Cell. 1997;89:511–521. D, Micrograph courtesy William C. Earnshaw.)
assembles on chromosomes during DNA replication and is recruited to regions of heterochromatin by HP1. Recent evidence also suggests that cohesin also has an important role in regulating gene expression during interphase, possibly by stabilizing chromatin loops that assemble active chromatin hubs. CTCF can bind cohesin, and the two proteins are found at most boundaries
between TADs (though there are many more binding sites for the two proteins within TADs). DNA topoisomerase IIα, an enzyme that alters DNA topology by passing one double-helix strand through another, is a very abundant component of mitotic chromosomes. In mitosis, topoisomerase IIα is concentrated at centromeres and in axial regions along the
CHAPTER 8 n DNA Packaging in Chromatin and Chromosomes
C
D. Cohesin
Separase cleavage 268
Smc3
Separase cleavage 180
Hinge
Scc1
Smc2 N C
Smc2
CAP-H
Smc4
Smc2
Smc1
Scc1
C-term
B
E. Condensin I
N-term
Smc3
Hinge
Hinge
Smc1
A
139
SA1 CAP-G
Smc4 SA2
Smc4
CAP-D2
Pds5
FIGURE 8.18 CONDENSIN AND COHESIN COMPLEXES. A–B, Model of the isolated dimer of SMC2 and SMC4 from condensin. Chemical crosslinks between SMC2 and SMC4 used to constrain the modeling are shown in red. Colored spheres represent lysines involved in crosslinks. B, Structure of a portion of the cohesin complex showing the paired heads with a bound fragment of the Scc1 kleisin. C–D, Subunit composition and structural organization of the cohesin and condensin complexes. (B, From Gligoris TG, Scheinost JC, Bürmann F, et al. Closing the cohesin ring: structure and function of its Smc3-kleisin interface. Science. 2014;346:963–967.)
chromosome arms. Topoisomerase IIα is very dynamic in vivo, moving on and off chromosomes in a time frame of seconds. Mitotic chromosomes from cells lacking topoisomerase II are long and thin, and the protein is thought to have a role in untangling the DNA as the loops condense along the chromosome axis during chromosome formation. Topoisomerase II is also required for replicated sister chromatids to separate from one another during mitotic anaphase. Presumably, the enzyme separates tangles and intertwinings of DNA created during DNA replication. Remarkably, of the more than 4000 proteins found in mitotic chromosomes, only the histones, and fewer than 20 nonhistone proteins are known to have a role in mitotic chromosome formation. This does not count the more than 100 proteins that are required to form the kinetochores, which direct chromosomal movements in mitosis.
The Chromosome’s Control Center: The Kinetochore The centromere is the genetic locus that specifies the site where a kinetochore assembles on the chromosomal DNA molecule. The kinetochore is a button-like structure embedded in the surface of the centromeric chromatin of most eukaryotic mitotic chromosomes (Fig. 8.19). When thin sections of centromeres are examined by electron microscopy, the kinetochore often appears to have several layers. The inner kinetochore is embedded in the surface of the centromere and is composed of a specialized form of chromatin. The outer kinetochore consists of an outer plate with a fibrous corona on its outer surface. It is constructed from protein
A
Inner plate (kinetochore assembly and stability?)
B
Microtubules
Kinetochore
Outer plate (microtubule binding)
C
Corona (motors) FIGURE 8.19 KINETOCHORE STRUCTURE. A, Diagram of the major layers of the kinetochore. B, Thin-section electron micrograph of a kinetochore with attached microtubules. C, Thin-section micrograph of an unattached kinetochore. (B, Courtesy J.B. Rattner, University of Calgary, Alberta, Canada. C, Courtesy Rebecca L. Bernat and William C. Earnshaw.)
complexes that link the chromatin to microtubules of the mitotic spindle. During interphase, the kinetochore persists as a condensed ball of heterochromatin that resembles other areas of condensed chromatin within the nucleus. The distinct multilayered kinetochore structure forms on the surface of the centromere during an early stage of mitosis called prophase (see Chapter 44), reaching its mature state following nuclear envelope breakdown when the chromosome comes into contact with microtubules at the onset of mitotic prometaphase.
140
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
Chapter 7 describes the three types of centromeres known in eukaryotes (see Fig. 7.9). Point centromeres found in budding yeasts assemble kinetochores on defined DNA sequences and do not require epigenetic activation to function. They bind one microtubule. Regional centromeres, found in organisms ranging from fission yeast to humans, are based on preferred DNA sequences but require epigenetic activation to function. They bind two to 20 or more microtubules. In holocentromeres, as found in Caenorhabditis elegans and many plants and insects, the microtubules (roughly 20 in C. elegans) bind all along the poleward-facing surface of the mitotic chromosome. Given this diversity of centromeres, it is remarkable that the proteins responsible for kinetochore assembly and function are well conserved across evolution.
Mammalian Kinetochore Proteins The first three specific kinetochore proteins identified in any species were discovered in humans using autoantibodies present in the sera of patients with
A. Scleroderma patient
C
rheumatic disease (Fig. 8.20). These proteins, designated CENP-A (centromere protein), CENP-B, and CENP-C, are conserved from humans to yeasts. They are part of the 16-protein constitutive centromere-associated network, which is composed of proteins that remain bound to the inner kinetochore throughout the cell cycle. The inner kinetochore chromatin is based on specialized nucleosomes with the histone H3 variant CENP-A (Fig. 8.21). How CENP-A targets the DNA to assemble kinetochorespecific nucleosomes is unknown, but some factors include a specialized chaperone, histone modifications and specialized RNA transcription, which, remarkably, occurs during mitosis. CENP-B binds specifically to a 17–base pair sequence (the CENP-B box) in α-satellite centromere DNA (see Fig. 7.9) and is required for efficient kinetochore assembly, but the mechanistic details are unknown. CENP-B probably originated as the enzyme responsible for movement of an ancient transposon. CENP-C and CENP-T (identified much later) are essential DNA-binding proteins that bridge between the inner chromatin and outer microtubule-binding components
D. Immunoblot with autoimmune serum CENP-C • Links the outer and inner kinetochore • Binds CENP-A and Mis12 complex CENP-B • Binds centromere DNA sequence • Promotes efficient kinetochore assembly
B
Serum from patient
CENP-A • Centromeric histone H3 variant • Marks site of kinetochore assembly FIGURE 8.20 SOME PATIENTS WITH SCLERODERMA HAVE AUTOANTIBODIES THAT RECOGNIZE CENTROMERIC PROTEINS. Scleroderma (“hard skin”) is a serious connective tissue disease associated with excessive deposition of collagen in the skin and walls of blood vessels. Note the “purse string” appearance of the skin surrounding the mouth of this patient (A). When serum from a patient with anticentromere antibodies is added to chromosomes on a slide (B) and bound antibodies are detected with a fluorescent probe, the centromeric regions of the chromosomes “light up” (C). Anticentromere antibodies are useful to identify patients who are at risk for serious autoimmune disease. Up to 20% of the population has a mild condition—Raynaud phenomenon (hypersensitivity of the skin to cold)—that is very rarely a precursor to scleroderma. Sensitive assays for anticentromere antibodies revealed that patients with Raynaud phenomenon who also have these autoantibodies have an increased risk of progression to scleroderma. D, Centromere proteins (CENPs) detected with anticentromere antibodies from a scleroderma patient on an immunoblot following sodium dodecylsulfate gel electrophoresis of chromosomal proteins. (A, From Dana R. Scleroderma. In: Albert DM, ed. Albert & Jakobiec’s Principles & Practice of Ophthalmology, 3rd ed. Philadelphia: Elsevier; 2008. C–D, Courtesy William C. Earnshaw.)
CHAPTER 8 n DNA Packaging in Chromatin and Chromosomes
of the kinetochore (Fig. 8.21). It has been suggested that the CENP-C link is primarily involved in enabling chromosome movements, whereas the CENP-T linkage may monitor the status of kinetochore attachment. The best-characterized component of the outer kinetochore complex is the NDC80 complex—an elongated rod with globular ends linked by central coiledcoils. One end of this complex binds to microtubules. Some copies of the NDC80 complex form a network with six other components that is thought to be the main mechanical link between chromosomes and microtubules. Both CENP-C and CENP-T independently link the inner chromatin to this outer NDC80-associated network. One protein from the NDC80 network also recruits to kinetochores the signaling components of the mitotic checkpoint pathway that regulates progression of the cell through mitosis without errors (see Fig. 44.11).
CENP-T/W KNL1
Ndc80 complex
Microtubule
CENP-C CCAN Mis12 complex CENP-A nucleosome
Outer kinetochore assembles only in mitosis Links kinetochore to microtubules Inner kinetochore Associates with centromere chromatin across whole cell cycle
141
Centromere Proteins of the Budding Yeast The best-characterized kinetochores come from bud ding yeast. Yeast kinetochores have been isolated and subjected to both biochemical and biophysical characterization (Fig. 8.22). More than 65 kinetochore-associated proteins assemble a structure at least the size and complexity of a ribosome. Specific centromere DNA-binding factors (CBF) recognize the DNA sequences (CDE I and CDE III) that specify the point centromere (see Fig. 7.7) and wrap around a specialized nucleosome containing a centromerespecific histone H3 variant related to CENP-A (Fig. 8.22). A stretch of A : T-rich DNA called CDE II completes one turn around this nucleosome, juxtaposing the flanking CDE I and CDE III elements and their associated proteins. Several large complexes bind to this nucleosome/CBF platform (Fig. 8.22). Although the CBF proteins are unique to yeast, the other complexes are all conserved from yeast to humans. These include the NDC80 complex (which was first identified by yeast genetics). The NDC80 complex binds to a ring made by the 10-subunit Dam1 complex that encircles the microtubule. This may help the kinetochore hold onto microtubules that are shrinking as the chromosome moves poleward during anaphase. The Dam1 complex is poorly conserved during evolution, although a possible vertebrate counterpart has been identified.
Role of RNA Interference at Fission Yeast Centromeres FIGURE 8.21 HYPOTHETICAL MODEL FOR THE ORGANIZATION OF THE VERTEBRATE KINETOCHORE. Protein complexes discussed in the text are indicated.
The fission yeast Schizosaccharomyces pombe has the simplest well-characterized regional centromere. It assembles a kinetochore that binds two to four
Ndc80 complex
A Cse4 nucleosome Mif2
CBF3
B
Spc105 Dam1 complex
COMA
MIND Microtubule
KMN 100 nm
100 nm
FIGURE 8.22 MODEL FOR THE ORGANIZATION OF THE BUDDING YEAST KINETOCHORE AND MICROGRAPH OF AN ISOLATED KINETOCHORE. A, Hypothetical diagram of budding yeast kinetochore (for discussion of the yeast centromere DNA, see Chapter 7). B, Electron micrograph of a budding yeast kinetochore attached to a microtubule. (B, From Gonen S, Akiyoshi B, Iadanza MG, et al. The structure of purified kinetochores reveals multiple microtubule-attachment sites. Nat Struct Mol Biol. 2012;19:925–929.)
142
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
microtubules. Fission yeast have orthologs of all the proteins and protein complexes described here. The fission yeast centromere provides insights into the formation of centromeric heterochromatin. The “silent” repeated DNA in the S. pombe centromere is transcribed from both DNA strands, yielding short double-stranded RNAs that are processed by the RNAi machinery. This RNAi response is part of the pathway for assembly of centromeric heterochromatin (see Fig. 11.14). A wide range of S. pombe mutants affecting the RNAi machinery all compromise centromere function and mitotic chromosome segregation. Whether RNAi is also essential for centromere function in metazoans has been more difficult to determine, as genetic analysis is complicated by multiple redundancies in the genes encoding the RNAi machinery. However, careful analysis reveals that centromeric satellite DNAs are indeed transcribed. Remarkably, this transcription occurs during mitosis, and is the only transcription known to occur during that cell-cycle phase. The specialized regulation that enables mitotic centromere transcription is unknown, as is whether the transcripts participate in a functional RNAi pathway like that observed in yeast.
Conclusions Ironically, just as the sequence of the euchromatic portion of the human genome was completed, a shift in our understanding revealed that essential aspects of the control of gene activity and chromosome structure cannot be revealed by analysis of the DNA sequence alone, as these regulatory processes are “encoded” in transient epigenetic modifications of DNA and histones. Understanding the extraordinarily elaborate epigenetic code has only just begun, so watch this space for further exciting developments. ACKNOWLEDGMENTS We thank Julie Ahringer, Wendy Bickmore, Job Dekker, Margarete Heck, Kazuhiro Maeshima, and Tom OwenHughes for their suggestions on revisions to this chapter.
SELECTED READINGS Allshire RC, Ekwall K. Epigenetic regulation of chromatin states in Schizosaccharomyces pombe. Cold Spring Harb Perspect Biol. 2015;7:a018770. Bannister AJ, Kouzarides T. Regulation of chromatin by histone modifications. Cell Res. 2011;21:381-395. Belmont AS. Large-scale chromatin organization: the good, the surprising, and the still perplexing. Curr Opin Cell Biol. 2014;26:69-78. Bickmore WA. The spatial organization of the human genome. Annu Rev Genomics Hum Genet. 2013;14:67-84. Biggins S. The composition, functions, and regulation of the budding yeast kinetochore. Genetics. 2013;194:817-846. Chaligné R, Heard E. X-chromosome inactivation in development and cancer. FEBS Lett. 2014;588:2514-2522. de Graaf CA, van Steensel B. Chromatin organization: form to function. Curr Opin Genet Dev. 2013;23:185-190. Fukagawa T, Earnshaw WC. The centromere: chromatin foundation for the kinetochore machinery. Dev Cell. 2014;30:496-508. Gibcus JH, Dekker J. The hierarchy of the 3D genome. Mol Cell. 2013;49:773-782. Huang H, Sabari BR, Garcia BA, et al. SnapShot: histone modifications. Cell. 2014;159:458-458.e1. Hudson DF, Marshall KM, Earnshaw WC. Condensin: architect of mitotic chromosomes. Chromosome Res. 2009;17:131-144. Jeppsson K, Kanno T, Shirahige K, et al. The maintenance of chromosome structure: positioning and functioning of SMC complexes. Nat Rev Mol Cell Biol. 2014;15:601-614. Maze I, Noh KM, Soshnev AA, et al. Every amino acid matters: essential contributions of histone variants to mammalian development and disease. Nat Rev Genet. 2014;15:259-271. Merkenschlager M, Odom DT. CTCF and cohesin: linking gene regulatory elements with their targets. Cell. 2013;152:1285-1297. Narlikar GJ, Sundaramoorthy R, Owen-Hughes T. Mechanisms and functions of ATP-dependent chromatin-remodeling enzymes. Cell. 2013;154:490-503. Pombo A, Dillon N. Three-dimensional genome architecture: players and mechanisms. Nat Rev Mol Cell Biol. 2015;16:245-257. Simon JA, Kingston RE. Occupying chromatin: polycomb mechanisms for getting to genomic targets, stopping transcriptional traffic, and staying put. Mol Cell. 2013;49:808-824. Takeuchi K, Fukagawa T. Molecular architecture of vertebrate kinetochores. Exp Cell Res. 2012;318:1367-1374. Thadani R, Uhlmann F, Heeger S. Condensin, chromatin crossbarring and chromosome condensation. Curr Biol. 2012;22:R1012-R1021. Westhorpe FG, Straight AF. Functions of the centromere and kinetochore in chromosome segregation. Curr Opin Cell Biol. 2013;25: 334-340. Zhang T, Cooper S, Brockdorff N. The interplay of histone modificationswriters that read. EMBO Rep. 2015;16:1467-1481.
CHAPTER
9
Nuclear Structure and Dynamics T
he nucleus houses the chromosomes together with the machinery for DNA replication and RNA transcription and processing (Fig. 9.1). Immature RNAs must be kept apart from the translational apparatus because eukaryotic genes are transcribed into RNAs containing noncoding intervening sequences that are removed by splicing to assemble mature RNA molecules with a continuous open reading frame. Sequestration of immature RNAs is one function of the nuclear envelope, two concentric membrane bilayers that separate the nucleus and cytoplasm. The nuclear envelope also regulates the bidirectional transport of macromolecules in and out of the nucleus, participates in chemical, protein and mechanical signaling pathways, contributes to genome organization, and provides mechanical stability to the nucleus.
Nuclear envelope Nuclear pores
Nucleolus
Heterochromatin FIGURE 9.1 ELECTRON MICROGRAPH OF A THIN SECTION OF A CANCER CELL NUCLEUS WITH MAJOR FEATURES LABELED. (Courtesy Scott Kaufmann, Mayo Clinic, Rochester, MN.)
This chapter describes what is known about the structure of the nucleus, the nuclear envelope, and the transport of macromolecules into and out of the nucleus, and discusses their links to human diseases. Aspects of nuclear structure and function that are discussed elsewhere include genome and chromosome organization (Chapter 7), chromatin structure (Chapter 8), DNA replication (Chapter 42) and RNA transcription and processing (Chapters 10 and 11).
Overall Organization of the Nucleus Studies in which entire individual chromosomes are labeled by in situ hybridization (chromosome painting; see Fig. 8.10) reveal that chromosomes tend to occupy discrete regions within the nucleus called chromosome territories. The boundaries of adjacent territories, where more actively transcribed regions are generally located, overlap with one another such that approximately 40% of each territory intermingles with adjacent territories. The chromatin of these overlapping regions tends to be less compact than in the rest of the territory, and is referred to as the interchromosomal domain. Most RNA transcription and processing are thought to occur within this domain. Although the nucleoplasm is very crowded with chromosomes and ribonucleoproteins (RNPs), proteins can nonetheless diffuse surprisingly rapidly through the nucleus, possibly by moving in the interchromosomal domain. Evidence is accumulating that actin is present in nuclei, presumably in the interchromosomal domain. Although the role of this actin is unknown, an attractive hypothesis is that it forms a scaffold for other processes. Nuclear actin can influence the positioning of nuclear subdomains.
Specialized Subdomains of the Nucleus Cell nuclei contain numerous discrete subdomains or bodies with distinctive structural organizations and/ or biochemical composition (Fig. 9.2 and Table 9.1). 143
144
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
A
B
C Nucleoli PML bodies
Cajal bodies Nuclear envelope
Nucleoli Speckles Chromatin Cajal bodies
5 µm
D Speckles
PIKA
Nucleoli
FIGURE 9.2 EXAMPLES OF MAJOR SUBNUCLEAR STRUCTURES. A, Components involved in RNA processing are scattered throughout the nucleus but concentrated in domains called speckles that are rich in interchromatin granules. Inhibition of RNA processing causes splicing components to accumulate in enormous concentrations of interchromatin granule clusters. Several cells were injected with a short oligonucleotide that disrupts the function of the U1 small nuclear ribonucleoprotein (snRNP) in RNA processing (see Fig. 11.15), and were then stained with an antibody recognizing the Sm splicing components (green). The injected cells were marked by introducing an inert fluorescent dextran marker into the cytoplasm (red). B, Nucleus with simultaneous staining of nucleoli (blue), PML (promyelocytic leukemia) nuclear bodies (red), Cajal bodies (green), and the nuclear envelope (purple). C, Nucleus with simultaneous staining of chromatin (blue), nucleoli (red), speckles (green), and Cajal bodies (white). D, Nucleus with simultaneous staining of DNA (blue) and the polymorphic interphase karyosomal association (PIKA)/53BP1 nuclear body/OPT (Oct1/PTF/transcription) domain (red). Nucleoli appear as unstained areas. A number of proteins involved in the sensing and repair of DNA damage concentrate in the PIKA. (A, Courtesy David Spector, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. B–C, Courtesy Angus Lamond, University of Dundee, United Kingdom. D, Courtesy William S. Saunders and William C. Earnshaw.)
The most prominent of these is the nucleolus, discussed in the next section. Although often referred to as organelles, nuclear subdomains, unlike cytoplasmic organelles, are not membrane-bounded. In fact, many proteins that have been examined by the fluorescence recovery after photobleaching technique (see Fig. 6.3) exchange relatively rapidly between a nuclear body and a nucleoplasmic pool. Therefore, these bodies represent highly dynamic associations of macromolecular complexes. By concentrating particular RNAs and proteins with enzymes involved in their maturation, they accelerate macromolecular assembly and maturation processes. They may also concentrate components involved in gene regulation or repair at particular chromosomal loci. Structures associated with RNA transcription and processing are found at up to 10,000 sites spread throughout the typical mammalian nucleus as well as in a few more prominent domains. The dispersed sites likely correspond to structures called perichromatin fibrils, originally observed by electron microscopy on the surface of regions of condensed chromatin.
Perichromatin fibrils contain various splicing factors and RNA-packaging proteins. When factors involved in RNA processing are detected by fluorescence microscopy, 20 to 50 bright speckles are seen against a diffuse background of nucleoplasmic staining (Fig. 9.2). The diffuse staining probably corresponds to splicing factors associated with perichromatin fibrils at dispersed sites. Most speckles correspond to clusters of interchromatin granules, particles 20 to 25 nm in diameter distributed throughout the interchromosomal domain. Proteomic analysis reveals that isolated interchromatin granules contain more than 200 stably associated proteins, most involved with various aspects of RNA processing. When tagged with green fluorescent protein, components involved in pre–mRNA (messenger RNA) processing cycle between speckles and sites of transcription in less than 1 minute in live cells. Metabolic labeling experiments indicate that speckles are not major sites of active transcription, although most transcription sites are associated with the periphery of speckles. Speckles are less prominent in cells that transcribe RNA at high levels, and become strikingly
CHAPTER 9 n Nuclear Structure and Dynamics
145
TABLE 9.1 Major Nuclear Subdomains Structure
Comments
Nucleolus
The nucleolus (typically 1 to 5 structures of 0.5 to 5 µm diameter in mammalian cell nuclei) is the site of ribosomal RNA (rRNA) transcription and processing, as well as of preribosomal assembly. It is also the site of processing of several other noncoding RNAs, including the RNA component of the signal recognition particle (SRP; see Fig. 20.5). It plays an important role in helping organize the genome during interphase, as well as in regulating the stability of p53, a critical transcription factor that is involved in regulating the cell cycle, particularly when DNA damage occurs.
Speckles
Speckles are concentrations of components involved in RNA processing. They often correspond to clusters of interchromatin granules seen by electron microscopy. They may serve as storage depots of splicing factors, or they may play a more active role in splicing factor modification and/or assembly.
Cajal bodies
Formerly known as coiled bodies. Approximately 0.2 to 1.0 µm in diameter, Cajal bodies have a coiled fibrous substructure. First identified by electron microscopy, up to 10 of these structures are seen in transformed cells. They are usually absent from nontransformed normal cells. They contain the human autoantigen p80-coilin and survival of motor neurons (SMN) protein, which is encoded by the gene mutated in spinal muscular atrophy, a severe, inherited, human, muscular wasting disease. They are involved in small nuclear ribonucleoprotein (snRNP) and small nucleolar ribonucleoprotein (snoRNP) assembly and in maturation of telomerase (which also contains an RNA component).
PML bodies
Also known as PODs and ND10, 10 to 30 of these structures are scattered throughout the nucleus. They are thought to enhance gene repression by serving as assembly sites for certain transcriptional corepresser complexes. They also appear to be targeted during viral infections. Fusion of the marker protein PML to the α-retinoic acid receptor is often found in acute promyelocytic leukemia (hence the name PML), in which the PML bodies appear highly fragmented. Treatments that are effective against PML restore the normal morphology of PML bodies (see text).
Polycomb group bodies
Concentrations of the PRC1 and PRC2 complexes (see Chapter 8) involved in the silencing of facultative heterochromatin. One mechanism for gene inactivation may be translocation into these inactive domains.
53BP1 nuclear bodies
Defined as concentrations of the DNA repair-associated protein 53BP1. Also known as PIKA (polymorphic interphase karyosomal association) and OPT (Oct1/PTF/transcription) domain. These domains may be up to 5 µm in diameter during G1 phase, but their morphology and number vary across the cell cycle. They appear to correspond to sites of DNA damage during mitosis that arise as a result of incomplete DNA replication during S-phase.
prominent when RNA processing is inhibited (Fig. 9.2). Together these observations suggest that speckles may be dynamic depots where RNA processing factors accumulate then they are not active. They may also have a role in rendering mRNAs competent for export to the cytoplasm. Cajal bodies (formerly known as coiled bodies) are compact structures approximately 0.3 to 1.0 µm in diameter (Fig. 9.2B) that resemble balls of tangled threads in the electron microscope. Nuclei of rapidly growing transformed cells typically have one to 10 prominent Cajal bodies. These structures are absent from most nontransformed (normal) cells. They contain an 80-kD human autoantigen called p80-coilin and the survival of motor neurons (SMN) protein, which is encoded by a gene mutated in spinal muscular atrophy, a severe inherited human muscular wasting disease. The SMN protein participates in importing immature small nuclear ribonucleoproteins (snRNPs) into the nucleus after their assembly in the cytoplasm. p80-coilin recruits the SMN complex to Cajal bodies, where the snRNPs are further processed to render them functional in RNA splicing reactions (see Chapter 11). Cajal bodies also have a role in the maturation of the RNP enzyme telomerase as well as other functions.
Mammalian nuclei also contain approximately 10 to 30 bodies, varying in size from 0.3 to 1.0 µm, known as promyelocytic leukemia (PML) bodies (Fig. 9.2B). PML bodies were initially defined by the presence of a protein called PML, an important regulator of cell growth and genome stability. PML has a RING-finger amino acid sequence motif and is therefore probably an E3 ligase for ubiquitin or ubiquitin-like proteins (see Fig. 23.2). Its targets are unknown. In normal cells, PML bodies apparently have a role in assembling corepressor complexes that modify chromatin to repress transcription (see Chapter 8). Their other functions are not known. The PML gene was identified by analysis of a chromosome translocation between chromosomes 15 and 17 found in patients with acute promyelocytic leukemia (APL). In many patients, this translocation produces a gene fusion between PML and the retinoic acid receptor alpha (RARα). The fusion protein, PML-RARα, blocks differentiation of hematopoietic precursors and causes APL. In APL cells, PML-RARα is distributed in tiny punctate foci scattered throughout the nucleus. When APL cells are treated with arsenic trioxide, which is clinically effective in treating APL, PML-RARα aggregation causes prominent PML bodies to reform. PML-RARα is ubiquitylated within those bodies and subsequently degraded.
146
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
A
B
Dense fibrillar component Fibrillar center
Nucleus
Nucleoli
Nucleolus organizing regions
Granular component
Mitotic chromosomes
FIGURE 9.3 NUCLEOLUS AND NUCLEOLAR ORGANIZER REGION. A, Electron micrograph of a thin section of a typical nucleolus. The fibrillar centers, dense fibrillar component, and granular component are indicated. B, Use of silver staining to visualize the nucleolus in interphase nuclei and the nucleolar organizer regions on mitotic chromosomes of the rat kangaroo. (A, From Fawcett DW. The Cell. Philadelphia: WB Saunders; 1981. B, From Robert-Fortel I, Junéra HR, Géraud G, et al. Three-dimensional organization of the ribosomal genes and Ag-NOR proteins during interphase and mitosis in PtK1 cells studied by confocal microscopy. Chromosoma. 1993;102:146–157.)
This allows the hematopoietic precursors to differentiate and cures the cancer.
The Nucleolus: The Most Prominent Nuclear Subdomain The nucleolus, first described only 5 years after the nucleus, in 1835, is the most conspicuous and bestcharacterized nuclear subdomain (Figs. 9.1 and 9.3). Most mammalian cells have one to five nucleoli, which are specialized regions 0.5 to 5.0 µm in diameter surrounding transcriptionally active ribosomal RNA (rRNA) gene clusters. Nucleoli are the sites of most steps in ribosome biogenesis, from the transcription and processing of rRNA to the initial assembly of ribosomal subunits. The ribosome is a complex macromolecular machine with four different structural rRNA molecules and approximately 85 proteins that are assembled into two subunits (see Figs. 12.6 and 12.7). Transcription of rRNA by RNA polymerase I comprises nearly half of total cellular RNA synthesis in some cell types. This high level of synthesis is necessary to produce approximately 5 million ribosomes in each cell cycle, more than 30 every second in budding yeast. Nearly 700 proteins associate stably with human nucleoli. Many more may associate transiently, and this composition changes to reflect different metabolic states of the cell (Fig. 9.4). Many nucleolar proteins are involved with either rRNA synthesis and modification or with ribosome subunit assembly. The functions of many other nucleolar proteins remain unknown and may reflect the involvement of nucleoli in other biological processes.
A Pre-bleach
0s
10 s
60 s
5 min
30 min
0s
10 s
20 s
40 s
60 s
B Pre-bleach
FIGURE 9.4 ANALYSIS OF DYNAMICS OF CHROMATIN AND A MAJOR NUCLEOLAR COMPONENT. A, Fluorescence recovery after photobleaching (FRAP) of H2B-GFP (green fluorescent protein) shows that chromatin is immobile within the cell nucleus. B, FRAP of fibrillarin-GFP shows that this major component of nucleoli is highly dynamic. Scale bar: 5 µm. (A–B, Courtesy Tom Misteli. B, From Phair RD, Misteli T. High mobility of proteins in the mammalian cell nucleus. Nature. 2000;404:604–609.)
Other stable RNAs, including the RNA component of the signal recognition particle (see Fig. 20.5), are also processed in the nucleolus. Intriguingly, the nucleolus is involved in controlling the stability of the critical cell-cycle regulator protein p53 (see Fig. 41.5). Healthy cells keep p53 levels low by using ubiquitylation in the nucleolus to destabilize it. Under certain types of stress, cells defend themselves by
activating p53. They do this by having nucleolar proteins bind and inactivate Mdm2, the key factor that ubiquitylates p53.
Ribosomal Biogenesis in Functionally Distinct Regions of the Nucleolus Transmission electron micrographs of thin sections show three morphologically distinct regions in the nucleolus (Fig. 9.3). Fibrillar centers contain concentrations of rRNA genes, together with RNA polymerase I and its associated transcription factors. Actively transcribed ribosomal genes are found near the border between the fibrillar centers and a dense fibrillar component that surrounds them. The granular component is the site for many steps in ribosome subunit assembly and is made up of densely packed clusters of ribosomal precursors called preribosomal particles 15 to 20 nm in diameter. rRNA loci have a modular organization. Genes alternate with spacer regions in large tandemly arranged clusters (see Fig. 11.10). The repeat unit in this array (gene plus spacer) is approximately 40,000 base pairs in humans. Humans have approximately 300 to 400 copies of the ribosomal DNA (rDNA) repeat unit located in clusters on chromosomes 13, 14, 15, 21, and 22. Usually, only a fraction of these genes is actively transcribed. An additional rRNA, 5S RNA, is encoded by distinct genes and transcribed by RNA polymerase III (see Fig. 10.8). A simple yet efficient mechanism guarantees a balance between the RNA components of the two ribosomal subunits. The major rRNA components are encoded by a single precursor RNA molecule. In humans, this 13,000base precursor is commonly described by its sedimentation coefficient in sucrose gradients as 45S. Following its transcription, the RNA precursor is processed in a series of cleavages to yield the 18S, 5.8S, and 28S rRNA molecules (see Fig. 11.10). In addition to the cleavages, rRNA processing also involves extensive base and sugar modifications, including approximately 100 2′-O-methyl ribose and approximately 90 pseudouridine residues per molecule. The earliest stages of rRNA processing probably occur in the dense fibrillar component of the nucleolus. Later stages take place in the granular component. Ribosomal protein synthesis occurs in the cytoplasm on free ribosomes. The newly synthesized proteins are transported into the nucleus for assembly into ribosomes, predominantly in the granular component. Disassembly of the Nucleolus During Mitosis The nucleolus disassembles during each mitotic cycle, starting with the dispersal of the dense fibrillar and granular components during prophase. This disassembly is driven by specific phosphorylation of nucleolar proteins. Ultimately, the fibrillar centers alone remain associated with the mitotic chromosomes, forming
CHAPTER 9 n Nuclear Structure and Dynamics
147
what are termed nucleolus-organizing regions (NORs [Fig. 9.3B]), which often form a prominent secondary constriction of the chromosome. (The primary constriction is the centromere.) Several nucleolar proteins and RNA polymerase I remain bound at NORs as cells enter and exit mitosis but most nucleolar proteins coat the surface of the mitotic chromosomes forming a perichromosomal layer or “skin.” Nucleolar reformation begins in mitotic telophase as processing factors and unprocessed pre-RNA remaining from the previous cell cycle associate with NORs (10 in human), which then cluster into one to five foci. Next, a wide variety of nucleolar components assemble into particles termed prenucleolar bodies that associate with the NORs in a process requiring transcription of the rRNA genes. Normally, nascent transcripts, rather than ribosomal genes, nucleate assembly of the nucleolus in each cell cycle. If antibodies to RNA polymerase I are microinjected into mitotic cells, rRNA transcription is blocked, and nucleoli do not reform in the next G1 phase.
Structure of the Nuclear Envelope The nuclear envelope provides a selective permeability barrier between the nuclear compartment and the cytoplasm and acts as a platform that helps organize the chromosomes in discrete functional domains (Fig. 9.5). The barrier keeps pre-mRNAs in the nucleus until they are fully processed and licensed for export so that only mature mRNAs are delivered to ribosomes in the cytoplasm for translation into protein. It also provides an
Rough endoplasmic reticulum
CYTOPLASM
Ribosome
Nuclear pore complex
Outer membrane
PERINUCLEAR SPACE
Inner membrane Nuclear lamina NUCLEAR INTERIOR FIGURE 9.5 OVERVIEW ORGANIZATION.
Chromatin OF
NUCLEAR
ENVELOPE
148
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
Structure and Assembly of the Nuclear Lamina The nuclear lamina is a thin protein meshwork composed of type V intermediate filament proteins called nuclear lamins (Figs. 9.6 and 9.7). Lamins can be divided into two families. Lamin A is encoded by a gene that gives rise to four major polypeptides (including lamin C) by alternative splicing (see Fig. 11.6). Members of the lamin B family are the products of two distinct genes. The various families of lamin proteins assemble into distinct fibrous networks (Fig. 9.6), exhibit different patterns of gene expression, and appear to have distinct roles in nuclear structure. The pattern of lamin gene expression depends on the cell type and stage of development. The lamina of embryonic stem cells and early embryos is comprised of B-type lamins. Lamins A and C typically appear later in development as cells begin to differentiate, and their expression varies in different cell types. This variation in lamina composition may contribute to different patterns of gene expression and mechanical stability of the nucleus. Lamins A/C promote nuclear stiffness, whereas nuclei containing only B-type lamins are more elastic. Like other intermediate filament proteins (see Fig. 35.2), nuclear lamins have a central, rod-like domain that
additional level of genetic protection and control since various chromosomal events, including DNA replication and expression of certain genes, are regulated, at least in part, by changes in the ability of factors to move between the cytoplasm and nucleus. The nuclear envelope is composed of two concentric lipid bilayers termed the inner and outer nuclear membranes. The outer nuclear membrane is continuous with the rough endoplasmic reticulum and shares some of its functions, including the presence of ribosomes. It also has unique proteins and functions. For example, it contains proteins that help link the nuclear interior with the cytoskeleton. A fibrous nuclear lamina of intermediate filaments supports the inner nuclear membrane in many eukaryotes. These and other inner nuclear membrane proteins mediate interactions of the envelope with chromatin. The inner and outer nuclear membranes are separated by an approximately 50-nm luminal space that is continuous with the lumen of the endoplasmic reticulum. Nuclear pore complexes and the associated pore membrane bridge both nuclear membranes and provide the primary route for communication between the nucleus and cytoplasm during interphase.
OL
A
Nuclear pore complexes
Nuclear lamina
OL
C
D
LA
LB1
E
F
B
FIGURE 9.6 NUCLEAR LAMINA. A, Thin-section electron micrograph of a nuclear envelope with a prominent nuclear lamina and nuclear pores. B, Field emission scanning electron micrograph of the inner surface of an amphibian oocyte nuclear envelope. The nuclear pores are prominent, protruding above the underlying nuclear lamina. C–F, Visualization of lamins A and B1 in a HeLa (Henrietta Lacks) cell nucleus by structured illumination superresolution microscopy. Both lamins form short filaments that mostly do not colocalize. OL, overlay. (A, For reference, see Fawcett DW. The Cell. Philadelphia: WB Saunders; 1981, Fig. 156 [top]. B, From Zhang C, Jenkins H, Goldberg MW, et al. Nuclear lamina and nuclear matrix organization in sperm pronuclei assembled in Xenopus egg extract. J Cell Sci. 1996;109:2275–2286. C–F, From Shimi T, Kittisopikul M, Tran J, et al. Structural organization of nuclear lamins A, C, B1, and B2 revealed by superresolution microscopy. Mol Biol Cell. 2015;26:4075–4086.)
CHAPTER 9 n Nuclear Structure and Dynamics
A
B. Human Lamin A fiber
C. Human Lamin A Ig domain
N
Head
5 nm
α-helical coiledcoil dimerization Lamin A: ZMPSTE24 cleavage site
NLS
149
further, leaving the farnesylated cysteine at the carboxyl terminus. In contrast, once it is at the nuclear membrane, pre– lamin A is processed by a protease called Zmpste24 (zinc metalloprotease similar to yeast Sterile 24) that clips off 15 additional amino acids from the C-terminus including the farnesylated cysteine, thereby loosening its association with the membrane. Possibly as a result of this, some A-type lamins also distribute throughout the nucleoplasm. These intranuclear lamins have been suggested to have roles in cell-cycle regulation (see next section). The assembled lamina is tethered to the inner nuclear membrane both by the farnesyl group and by interactions with integral membrane proteins (see next section). The surface of the lamina facing the nuclear interior also interacts with the chromosomes. Thus, the lamina and its associated proteins both serve as a structural support for the nuclear envelope and influence chromosome distribution and function within the nucleus (see later).
CaaX Tail
FIGURE 9.7 LAMIN ORGANIZATION AND ASSEMBLY. A, Several stages in the assembly of isolated lamin B dimers into filaments in vitro. The dimers at left have two globular heads at the C-terminal end of a rod that is 52 nm long. B–C, Diagram of the structural organization of the nuclear lamins. The sequence CaaX (see text) is a signal for the attachment of a farnesyl group. NLS, nuclear localization sequence. (A, From Heitlinger E, Peter M, Haner M, et al. Expression of chicken lamin B2 in Escherichia coli: characterization of its structure, assembly, and molecular interactions. J Cell Biol. 1991;113: 485–495.)
is largely α-helical (Fig. 9.7). The basic building block of lamin assembly is a dimeric α-helical coiled-coil (see Fig. 3.10) of two identical parallel polypeptides. Lamin dimers self-associate end to end to form protofilaments that associate laterally in a process that is under active investigation. The coiled-coil is followed by a large C-terminal domain with a central globular fold and containing a nuclear localization sequence (see later section) that promotes the rapid import of newly synthesized lamin precursors into the nucleus. In most lamins, the C-terminus acquires a lipid posttranslational modification that targets them to the nuclear membrane. This involves enzymatic addition of the C15-isoprenoid hydrocarbon tail farnesyl (see Figs. 13.10 and 20.15). The farnesyl group is added to a cysteine side chain in an amino acid motif called the CaaX box (Ca1a2X, where C is a cysteine located four amino acids from the carboxyl terminus; a1 is any aliphatic amino acid; a2 is valine, isoleucine or leucine; and X is usually methionine or serine) at the carboxyl terminus of the protein. This motif was first recognized in the Ras proteins (see Fig. 25.7). The aaX residues are removed after addition of the farnesyl group. B-type lamins are not processed
Proteins of the Inner Nuclear Membrane Several hundred integral membrane proteins are associated with the inner nuclear membrane, often in in a tissue specific manner. Of these, the lamin B receptor, LAP2 (lamina-associated protein 2), emerin, MAN1, SUN1, and SUN2 have been characterized in detail. Some inner nuclear membrane proteins bind lamins to help anchor the lamina polymer to the membrane and many can interact with chromatin. The lamin B receptor binds heterochromatin protein HP1 (Fig. 9.8) and links the envelope to condensed chromatin. Codisruption of the lamin B receptor and lamin A releases most heterochromatin from the nuclear periphery. The LEM domain, a 40-amino-acid motif common to several nuclear proteins, including LAP2, emerin, and MAN1, binds to an abundant small protein called barrierto-autointegration factor (BAF), so named for a separate role facilitating viral genome integration for HIV. BAF binds directly to DNA and to histones and functions in organizing chromatin across the cell cycle. LAP2 can affect chromatin organization in multiple ways. Some of its several splice variants lack the transmembrane region for inner nuclear membrane association and are soluble. Both soluble and transmembrane forms have the LEM domain and bind BAF, but the transmembrane forms can also bind a histone deacetylase. Interactions of intranuclear lamin A and a soluble splice variant of LAP2 are important for cell-cycle regulation by forming a complex with the tumor suppressor retinoblastoma protein (pRb; see Chapter 41). This, in turn, regulates the transcription factor E2F (see Fig. 41.9), which is important for activating the G2-to-S transition. Several other inner nuclear membrane proteins also bind transcriptional activators, in some cases sequestering them at the nuclear periphery away from their gene targets.
150
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
CYTOPLASM Outer Nuclear Membrane
LBR
LAP2 Man-1 (β–γ)
Lamin B
LEM
KASH binding domain of SUN
LAP1
Emerin
Lamin A
HP1
KASH TM domain
BAF
SUN1/2 ~45nm SUN3 ~20nm SUN4/5 ~12nm
Heter ochromatin
NUCLEOPLASM
FIGURE 9.8 INTEGRAL PROTEINS OF THE INNER NUCLEAR MEMBRANE. Lamin B receptor (LBR), lamina-associated protein 2 (LAP2), Man-1, and emerin all bind lamin B. LBR associates with chromatin via HP1. The other three associate with chromatin via the barrier-to- autointegration factor (BAF). Emerin and lamina-associated protein 1 (LAP1) also bind to lamin A. The α form of LAP2 is not membrane associated and is not shown here. Three isoforms of SUN proteins link the inner nuclear membrane to KASH domain proteins of the outer nuclear membrane. KASH proteins asssociate with the cytoskeleton. (SUN proteins from Sosa BA, Kutay U, Schwartz TU. Structural insights into LINC complexes. Curr Opin Struct Biol. 2013;23[2]:285–291. For reference, see Protein Data Bank [www.rcsb.org] file 4DXT [ribbon diagram of SUN/KASH].)
The SUN proteins bind lamin A, and then connect across the nuclear envelope lumen to huge KASH domain proteins in the outer nuclear membrane that in turn bind to all three major cytoplasmic filament systems, actin filaments, intermediate filaments, and microtubules (see Fig. 9.8 and Chapters 33 to 35). Thus, the lamina is linked to the rest of the cytoskeleton. Deletions or mutations in several lamina proteins, including the SUN proteins, reduce the mechanical stability of the cell and interfere with cell migration. SUN proteins are also important for maintaining the uniform spacing of the nuclear envelope lumen. Their disruption results in uneven separation of the inner and outer nuclear membranes. These diverse functions in genome organization and regulation, cell cycle regulation, signaling cascades and cell and nuclear mechanical stability could explain the link between mutations in nuclear envelope proteins and human disease (see later).
Role of the Nuclear Envelope in Genome Organization A high-throughput method revealed that the nuclear lamina has an important role in chromosome organization within nuclei. The method uses a DNA-modifying enzyme fused to a lamin protein, so nearby DNA is modified and can be mapped along the genome. In human cells, approximately 40% of the genome is found in 1000 to 1500 LADs (lamina-associated domains; Fig. 9.9) ranging in size from 10 kb to approximately 10 Mb. Analysis of single cells revealed that approximately 15% of LADs are associated with the lamina in most cells, with the remainder varying from cell to cell. The constitutive LADs have low transcriptional activity. Indeed, heterochromatin-associated histone marks such as H3K9me3 (see Fig. 8.7) promote association of particular chromosomal regions with the lamina. The LADs
at a distance from the nuclear periphery are either associated with a similar repressive compartment surrounding nucleoli or are in the nuclear interior. Interestingly, association of the chromosomes with the nuclear envelope is perturbed in some nuclear envelope-associated diseases (see the next section). Chromatin interactions with nuclear pores can have both positive and negative effects on gene expression. In mammalian cells, the chromatin near pores appears less condensed (less heterochromatic) than most chromatin adjacent to the lamina. The significance of these interactions is still under study. Disassembly of the nuclear envelope during mitosis in metazoa releases the chromosomes so that they can be segregated to the daughter cells by the cytoplasmic mitotic spindle (see Fig. 44.1). Mitotic segregation of chromosomes to daughter cells takes place within the nucleus in many other eukaryotes, including yeasts.
Nuclear Envelope Defects Lead to Human Diseases In 1994, a gene mutated in patients with human X-linked Emery-Dreifuss muscular dystrophy was found to encode a protein of the inner nuclear envelope. The gene was named emerin. This link between the nuclear envelope and human disease was the tip of a huge iceberg. Genetic defects in nuclear envelope proteins cause at least 20 disorders, including muscular dystrophies, lipodystrophies, and neuropathies (diseases of striated muscle, fatty tissue, and the nervous system, respectively). The most dramatic of these is Hutchinson-Gilford progeria syndrome (Fig. 9.10). Affected individuals are essentially normal at birth, but they appear to age rapidly and die in their early teens of symptoms (including atherosclerosis and heart failure) that are typically associated with extreme age. More than 500 mutations scattered across the gene encoding lamin A/C cause at least 15 different diseases,
CHAPTER 9 n Nuclear Structure and Dynamics
CYTOPLASM
Nuclear envelope M M M
M
M M M
Methyl groups
M Dam methylase
Chromatin fiber
M
M
M
LAD
M
M
Isolated DNA M
A
M
M M
M M MM
M M M M
X µm
Lamina
LADs
Observed/expected (OE)
B
C
5 0
Cell 1
5 0
Cell 2
5 0
Cell 3
5 0
Cell 4
5 0
Cell n–1
5 0
Cell n
5 0
Average 0
LADs 40
Position on chromosome 17 (Mb)
80
FIGURE 9.9 NUCLEAR LAMINA HELPS ORGANIZE THE CHROMATIN INTO FUNCTIONAL DOMAINS. A, Dam methylase fused to a lamin protein methylates the DNA in chromatin that is closely associated with the nuclear lamina. Isolation of the DNA allows the methylation sites to be mapped. B, Constitutive LADs associate with the lamina even after the cell has gone through mitosis and the lamina has been disassembled and reassembled. C, Sites of increased methylation on chromosome 17 from six single human cells, plus an average of the entire population. Lamina-associated domains (LADs) are indicated. (B–C, Modified from Kind J, Pagie L, de Vries SS, et al. Genome-wide maps of nuclear lamina interactions in single human cells. Cell. 2015;163:134–147.)
collectively termed laminopathies, some of which are variants of the diseases mentioned above (Fig. 9.10). At least two laminopathies are also linked to mutations in the Zmpste24 protease. Some of the symptoms of laminopathies can be modeled in mice, where loss of lamin A causes nuclear envelope defects and leads to a type of muscular dystrophy. The most surprising aspect of the laminopathies is the fact that except for premature aging, the defects linked to each mutation are limited to a few tissues such as striated muscle, even though lamins A/C are ubiquitous in differentiated cells throughout the body. Lamin mutations appear to compromise the stability of the nuclear envelope, so it has been suggested that muscle nuclei
151
might be particularly sensitive to these mutations, owing to mechanical stress during contraction. However, this mechanism cannot account for the link between lamin mutations and lipodystrophy—fat is not a force-generating tissue—neuropathy, or progeria. An alternative suggestion is that these mutations change interactions between the inner nuclear membrane and chromatin and this alters gene expression patterns. Cells from patients with Hutchinson-Gilford progeria syndrome show signs of aging in culture that are accompanied by dramatic alterations in heterochromatin (see Fig. 8.6), but changes in gene expression are relatively small and vary between patients.
Nuclear Pore Complexes In a typical growing cell, nearly all traffic between the nucleus and cytoplasm passes through approximately 3000 channels, called nuclear pore complexes, that bridge the inner and outer nuclear membranes (Fig. 9.11). Nuclear pore complexes have a scaffold consisting of three stacked rings each with eightfold symmetry. Cytoplasmic and nuclear rings flank a prominent spoke ring that is intimately associated with the pore membrane linking the inner and outer nuclear membranes. The nuclear ring is anchored to the nuclear lamina. A less-prominent fourth luminal ring surrounds the pore membrane in the NE lumen. The minimum diameter of the central channel through the pore is approximately 40 nm, and the channel is approximately 50- to 70-nm long. Eight filaments project outward from both the nuclear and cytoplasmic rings. These are involved in docking of macromolecules to be transported through the pore. The nuclear filaments are linked at their outer ends by a terminal ring, much like the wire that secures the cork on a champagne bottle. This structure is called the nuclear basket. Vertebrate nuclear pore complexes are large structures with a mass of approximately 90 to 120 million Da as assessed by electron cryomicroscopy. Core com ponents identified by mass spectrometry account for approximately 70 million Da of the mass. Yeast nuclear pores are similar in overall structure but about half the mass. The protein composition of nuclear pore complexes is remarkably conserved. Approximately 30 core proteins, called nucleoporins (Fig. 9.12), are present in multiples of eight copies. Mass differences between electron cryomicroscopy and mass spectrometry measurements may be accounted for by transport factors and other auxiliary subunits that do not have a key structural role. Two multiprotein complexes, the 10-member Y complex (named because of its shape) and the fivemember Nup93 complex make up the scaffold of the pore. The cytoplasmic and nuclear rings are assembled
152
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
3
4
5
6
7
89
10
Lamin A Globular domain
CaaX
CYTOPLASM
B
r Membrane Outer Nuclea
Emerin
LBR
Lamin B
DCM – dilated cardiomyopathy
11
Globular domain
Coiled-coil helical domain
EMD – autosomal-dominant Emery-Dreifuss muscular dystrophy LGMD1B – limb girdle muscular dystrophy type 1B
R582H R584H G608G
2
Exon 1
Q294P R298C R336Q E358K R377H E386K V442A N456I or N456K G465D I469T R482Q or R482W K486N R527H or R527P T528K L530P R571S (LaC)
N195K E203G or E203K R249Q
5'
R133P T150P
R25P R50S or R50P R60G L85R
A. Selected lamin mutations and the diseases they cause
FPLD – familial partial lipodystrophy 3' 12
CMT2 – Charcot-Marie-Tooth neuropathy type 2 B1 MAD – mandibuloacral dysplasia HGP – Hutchinson-Gilford progeria
C
ZMPSTE24
Lamin A
Mandibuloacral dysplasia X-linked Emery-Dreifuss muscular dystrophy Pelger Huët anomaly NUCLEOPLASM Greenberg skeletal dysplasia FIGURE 9.10 HUMAN DISEASES ASSOCIATED WITH NUCLEAR ENVELOPE ABNORMALITIES. A, Some of the mutations in the gene encoding lamin A that are associated with human disease. The G608G mutation makes no change in the protein sequence but creates a splice site leading to the loss of 50 amino acid residues from lamin A, leading to an impaired processing of prelamin A and generation of stably farnesylated lamin A. This mutation causes Hutchinson-Gilford progeria. Colored numbers at the top refer to the amino acid that has been changed by each mutation. Mutations are distributed in exons all across the gene as shown. B, Mutations in other nuclear envelope proteins cause similar diseases. Three examples are shown. C, Two young boys with the premature aging disorder Hutchinson-Gilford progeria. Sam Berns (left) with friend John Tacket, Progeria Research Foundation Youth Ambassador. (A, Modified from Mounkes L, Kozlov S, Burke B, et al. The laminopathies: nuclear structure meets disease. Curr Opin Genet Dev. 2003;13:223–230. C, Courtesy the Progeria Research Foundation, Peabody, MA; http:// www.progeriaresearch.org.)
from 16 copies each of the Y complex. The Nup93 complex forms the framework of the spoke ring and interacts with four nucleoporins having transmembrane domains that bend and fuse the pore membrane. Several of these proteins share structural features with clathrin-like proteins that coat membrane transport vesicles (see Fig. 21.8), so they may have a common evolutionary origin. Eleven of the 30 nucleoporins contain repeats of the dipeptide FG (phenylalanine-glycine). Two common versions include XFXFG and GLFG. In all, the pore contains approximately 5000 of these FG repeats in highly flexible intrinsically disordered regions of the proteins. FG nucleoporins are anchored to the pore scaffold with their FG repeat regions projecting into the central pore, where they form the transport barrier (see later). Three experiments show that nucleoporins are required to transport proteins into the nucleus. First, antibodies to nucleoporins inhibit transport when added to isolated nuclei or when injected into live cells. Second, lectins, such as wheat germ agglutinin (which binds specifically to sugars attached to many nucleoporins), inhibit transport in similar experiments. Third, nuclear
pore complexes assembled in Xenopus egg extracts (see Box 40.3) in the absence of the highly conserved nucleoporin p62, the defining member of the FG-rich p62 complex of three nucleoporins, appear structurally normal but are inactive in transport. In metazoans, nuclear pore complexes are remarkably stable, with the proteins apparently persisting for the lifetime of the cell. New pore complexes continue to assemble throughout interphase, but they disassemble into soluble subcomplexes during mitosis. During the telophase stage of mitosis, pore complex reassembly begins with binding of the Y complex to chromatin. The Y complex then interacts with transmembrane nucleoporins and the Nup93 complex, which recruits factors that bend and fuse the membranes, forming the pore. If the Nup93 complex is depleted from Xenopus egg extracts, nuclear membranes form around added nuclei but are devoid of pores.
Traffic Between Nucleus and Cytoplasm The nuclear pore complex is a highly efficient conduit that can allow the passage of up to approximately 100 MDa of cargo per second. Traffic leaving the nucleus
153
CHAPTER 9 n Nuclear Structure and Dynamics
A
Cytoplasmic filaments Cytoplasmic ring
Nuclear envelope
Spoke ring
C
CYTOPLASMIC VIEW
Nup192
CNC
Nup170
Outer membrane Disordered FG repeats in lumen
Nuclear ring
Lumen
Basket filament
Inner membrane
Terminal ring
B
Nuclear basket
CNT Nic96
D CNC
Nup192
Nup170 CNT Nic96 SIDE VIEW
FIGURE 9.11 NUCLEAR PORE COMPLEX. A–B, Three-dimensional and central section views of models of the human nuclear pore complex. C, Two views of the molecular organization of the nuclear pore based on three-dimensional reconstructions of cryoelectron micrographs. The pore has eight-fold symmetry in the plane of the nuclear envelope and two-fold symmetry perpendicular to the nuclear envelope. Colored protein subunits are identified with labels; the membrane is gray. Disordered FG repeats (green dotted lines) fill the central pore. D, Detail of the molecular model illustrating the two-fold symmetry of the protein subunits perpendicular to the nuclear envelope, with colored subunits above and gray subunits below. (C–D, For reference, see Protein Data Bank file 5A9Q and von Appen A, Kosinski J, Sparks L, et al. In situ structural analysis of the human nuclear pore complex. Nature. 2015;526:140–143.)
Sc NUP2
N
C
Sc NUP1 Hs-p62 Hs POM121 Sc NUP49 Sc NUP116 100 Amino acids
GLFG region
Repeat motif
FG mixed region
XFXFG region
Hydrophobic span
FIGURE 9.12 SEQUENCE ORGANIZATION OF SEVERAL NUCLEOPORINS, THE STRUCTURAL COMPONENTS OF THE NUCLEAR PORES. Nucleoporins contain combinations of repeated sequences as shown. Letters refer to the amino acids (see Fig. 3.2). The hydrophobic FG (phenylalanine-glycine) repeats facilitate nuclear trafficking through the pores by interacting specifically with transport factors carrying cargo.
includes messenger ribonucleoproteins (mRNPs), ribosomal subunits, and transfer RNAs (tRNAs), all of which must be transported to the cytoplasm to function in protein synthesis. Traffic entering the nucleus includes transcription factors, chromatin components, and ribosomal proteins. Other molecules follow more complex routes. Small nuclear RNAs (snRNAs) are exported to the cytoplasm to acquire essential protein components; they are then reimported into the nucleus, where they undergo further maturation steps before functioning in RNA processing. Individual pores can simultaneously transport components in both directions. Nuclear pores have constitutive peripheral channels through which solutes and small proteins of up to 30 to 40 kD (~5–10 nm) can diffuse passively. However, the pores can also actively transport much larger macromolecular complexes via the central channel. Almost all physiological traffic through the pores, even of small molecules, is a facilitated process that involves specific carrier proteins traversing the central channel. For example, the 28-kD NTF2 dimer (the Ran transporter; see later) traverses the pore approximately 120 times
154
A
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
CYTOPLASM
C
A
N
C
N
C
N
NUCLEUS
0.5 h
FIGURE 9.13 Electron micrographs (upper panels) and an artist’s rendition (lower panels) show deformation of a large RNP particle as it passes through the nuclear pore complex (cytoplasm [top]; nucleus [bottom]). This RNA encodes a secreted protein, with a molecular weight of about 1 million Da, from the salivary gland of the fly Chironomus tentans. Once in the cytoplasm, the 5′ end of the RNA docks with ribosomes and begins synthesis of its protein even before the passage of the remainder of the RNP through the pore has been completed. (From Mehlin H, Daneholt B, Skoglund U. Translocation of a specific premessenger ribonucleoprotein particle through the nuclear pore studied with electron microscope tomography. Cell. 1992;69:605–613.)
more rapidly than does the 27-kD green fluorescent protein. The pore gate opens to a maximum of approximately 40 nm, but larger particles can squeeze through, provided that they are deformable. This is well documented for export of a well-studied enormous RNA that associates with roughly 500 packaging proteins to make an RNP particle approximately 50 nm in diameter. The RNP is deformed into a rod-shaped structure as it squeezes through the pore (Fig. 9.13). Rigid particles cannot usually exceed the 30- to 40-nm limit. Integral proteins of the inner nuclear membrane enter the nucleus by diffusion in the plane of the membrane. The lamin B receptor is highly mobile in the endoplasmic reticulum (ER), its site of synthesis, and rapidly diffuses to the nuclear envelope. There, it transits to the inner membrane through the peripheral channels of the nuclear pore complex. Once in the inner nuclear membrane, it becomes fixed in place, presumably by binding to the lamina and/or chromatin. This mechanism involving lateral diffusion and retention is a common mode of membrane protein translocation into the nucleus, although conventional transport through the pore may also occur (see later). Proteins that are imported into the nucleus bear a nuclear localization sequence (NLS), also called a nuclear localization signal, that is recognized by specific carrier proteins called transport receptors (Figs. 9.14 and 9.15). The best-studied NLS is a patch of basic amino
7h
48 h
NUCLEUS (N)
B NLS
CYTOPLASM (C)
Nucleoplasmin pentamer (145,000 Da)
Partial digestion with protease Microinject into frog oocyte
FIGURE 9.14 IDENTIFICATION OF A NUCLEAR LOCALIZATION SEQUENCE ON THE PROTEIN NUCLEOPLASMIN. This 29,000-Da protein exists in vivo as a pentameric complex with a molecular weight of 145,000. The monomer is small enough to diffuse passively through the nuclear pores, but the pentamer is too large to do so. A, Gentle cleavage of the pentamer with a protease removes a relatively small peptide from one end of the protein (left two gel lanes). When the cleaved pentamers were labeled with radioactivity and injected into the cytoplasm of a Xenopus oocyte, it was found that four species were produced that could still migrate into the nucleus and one species was produced that could not (right three pairs of gel lanes). B, The interpretation of this experiment is that each nucleoplasmin polypeptide contains a “tail” that can be removed by proteolysis and that this tail contains a nuclear localization sequence. Each pentamer can migrate into the nucleus as long as it retains at least one polypeptide with a tail. Tailless pentamers remain stuck in the cytoplasm. (A, From Dingwall C, Sharnick SV, Laskey RA. A polypeptide domain that specifies migration of nucleoplasmin in the nucleus. Cell. 1982;30:449–458.)
acids similar to the sequence PKKKRKV (single-letter amino acid code; see Fig. 3.2), first identified on the simian virus 40 (SV40) large T antigen. A point mutation, yielding PKNKRKV, inactivates this sequence as an NLS. A related type of bipartite NLS features two smaller patches of basic residues separated by a variable spacer (KRPAATKKAGQAKKKK [critical residues are underscored]). These two types of sequences are referred to as basic NLSs. Basic NLSs function autonomously and can direct the migration of a wide range of molecules into the nucleus in vivo. In one example, colloidal gold particles up to 23 nm in diameter coated with nucleoplasmin (a protein with a bipartite basic NLS) are transported through nuclear pores (Fig. 9.16). NLSs vary
CHAPTER 9 n Nuclear Structure and Dynamics
155
A. -NLS
B. +NLS
FIGURE 9.15 ICAD (inhibitor of caspase-activated DNase) protein (see Fig. 46.13) was fused to the green fluorescent protein (GFP; green here) and expressed in cultured cells. The DNA is blue. A, A mutant form of ICAD : GFP fusion protein lacking the ICAD nuclear localization sequence (NLS) accumulates randomly throughout the cell. B, The intact ICAD : GFP fusion protein with NLS accumulates quantitatively in the nucleus. (Courtesy K. Samejima, University of Edinburgh, United Kingdom.)
CYTOPLASM
0.1 µm
NUCLEUS
FIGURE 9.16 NUCLEAR LOCALIZATION SEQUENCE OF NUCLEOPLASMIN CAN CAUSE COLLOIDAL GOLD PARTICLES TO BE TRANSPORTED INTO THE CELL NUCLEUS. A thin-section electron micrograph shows gold particles coated with nucleoplasmin crossing the nuclear envelope by passing through the nuclear pore complexes. Much smaller gold particles coated with bovine serum albumin (BSA) remain in the cytoplasm. Both sets of gold particles were microinjected into the cytoplasm of Xenopus oocytes, and the cells were processed 1 hour later for electron microscopy. Scale bar: 0.1 µm. (From Dworetzky SI, Lanford RE, Feldherr CM. The effects of variations in the number and sequence of targeting signals on nuclear uptake. J Cell Biol. 1988;107:1279–1287.)
in size and sequence, and are recognized by a number of different kinds of transport receptors. For example, an alternative type of NLS rich in glycine promotes nuclear import by a similar mechanism (see later) but using a different transport receptor. Many proteins exported from the nucleus bear a nuclear export sequence (NES) that is recognized by transport receptors related to those used for nuclear import (Fig. 9.17). Like import signals, these signals vary
in size and complexity. The HIV I Rev protein provides one example of a leucine-rich sequence (LQLPPLERLTL) that is recognized by the carrier CRM1. Certain RNA sequences or structures may also serve as NESs. The following is a brief thumbnail of protein import into the nucleus (Fig. 9.18). A protein with an NLS (known as cargo) binds to an import receptor either by itself or in combination with an adapter molecule, forming a complex that then passes through pores into the nucleus. There, the cargo and adapter (if used) are displaced from the import receptor. The adapter then releases its cargo and is transported back to the cytoplasm as the cargo of an export receptor. Import receptors also shuttle back through pores, where they can meet more cargo or cargo/adapter complexes. Molecules exported from the nucleus use a variation of this cycle, being picked up by the transport machinery in the nucleus and discharged in the cytoplasm. The key to this system is that it is vectorial: Nuclear components are transported into the nucleus while components that function in the cytoplasm are transported out. This means that each carrier picks up its cargo on one side of the nuclear envelope and deposits it on the other. This directionality is regulated by a simple yet elegant system involving Ran, a small guanine triphosphatase (GTPase [see Figs. 4.6 and 4.7 for background material on GTPases]), and associated factors.
Components of Nuclear Import and Export The nuclear import and export system involves many components, but the general principles of its operation are simple. To understand how it works, this section
156
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
A. Nucleoplasmin
D. Rhodamine BSA
B. Nucleoplasmin minus NLS
E. Ovalbumin: HIV Rev-NES
C. Nucleoplasmin minus NLS nuclear injection
F. Ovalbumin: HIV Rev-NES + Leptomycin B
FIGURE 9.17 DEMONSTRATION OF THE EXISTENCE OF SPECIFIC NUCLEAR IMPORT AND EXPORT SIGNALS ON PROTEINS. Left, Nuclear import. A, Nucleoplasmin microinjected into the cytoplasm rapidly migrates into the nucleus. B, Nucleoplasmin lacking its nuclear localization sequence (NLS), when microinjected into the cytoplasm, stays in the cytoplasm. C, Nucleoplasmin lacking its NLS microinjected into the nucleus stays in the nucleus. Right, Nuclear export. D, Fluorescently labeled bovine serum albumin (BSA) microinjected into the nucleus stays in the nucleus. E, When ovalbumin conjugated to the nuclear export sequence (NES) of the HIV (the virus that causes AIDS) Rev protein is microinjected into the nucleus, it rapidly migrates into the cytoplasm. F, In the presence of leptomycin B (a drug that inhibits the activity of the nuclear export receptor CRM1), ovalbumin conjugated to the NES of HIV Rev protein stays in the nucleus after microinjection. (A–C, From Dingwall C, Robbins J, Dilworth SM, et al. The nucleoplasmin nuclear location sequence is larger and more complex than that of SV-40 large T antigen. J Cell Biol. 1988;107:841–849, copyright the Rockefeller University Press. D–F, From Fukuda M, Asano S, Nakamura T, et al. CRM1 is responsible for intracellular transport mediated by the nuclear export signal. Nature. 1997;390:308–311.)
first introduces several of the components (see Cell SnapShot 2) and then describes one transport event in detail. Adapters Adapters bind to the NLS or NES sequences on some cargo molecules and also to particular regions on receptors. The best-characterized adapter is importin α, which is responsible for recognition of small basic NLS sequences and works together with the transport receptor importin β (see later) in nuclear transport. Importin α consists of a highly flexible N-terminal NLS-like
importin β-binding domain followed by 10 repeats of a helical motif (the Armadillo repeat [Fig. 9.18D]) that give the structured portion of the molecule a slug-like shape. The importin β-binding motif can bind either the NLSbinding region on importin β or the NLS-binding domain on importin α itself (the “belly” of the slug). The latter provides an autoinhibitory mechanism that is thought to be important in regulating the release of cargo in the nucleus at the end of an import cycle. Binding to importin β uncovers the NLS binding site on importin α so that it can bind cargo more efficiently. Other nuclear trafficking pathways use different adapters. For example, two adapters bridge between snRNA and the export receptor CRM1 during snRNA export from the nucleus. Nuclear Transport Receptors Except for mRNP export from the nucleus (which uses special transport factors), all nuclear trafficking receptors are related to importin β, the import receptor for proteins bearing a basic NLS. At least 20 nuclear transport receptors are known in vertebrates (14 in yeast). These proteins are also called karyopherins. Some function in nuclear import, but others function in export. Importin β consists entirely of 19 copies of a helical protein interaction motif called a HEAT repeat, giving the protein the shape of a snail-like superhelix with the potential to interact with a large number of protein ligands. All importin β family members have a binding site for the Ran GTPase (Fig. 9.18D). Importin β binds many NLSs directly but also interacts with other cargoes via the importin α adapter. Nucleoporin FG repeats sandwich between importin β HEAT repeat helices during passage through the pore channel. Directionality/Recycling Factors Ran-GTPase and its bound nucleotides inform nuclear trafficking receptors whether they are located in the nucleus or cytoplasm. Ran-GTP (Ran with bound guanosine triphosphate [GTP]) dissociates import complexes but is required to form export complexes. The system imparts directionality because Ran-GTP is converted to Ran-GDP (guanosine diphosphate) in the cytoplasm and Ran-GDP is converted to Ran-GTP in the nucleus. Like other small GTPases, Ran has low intrinsic GTPase activity, but interactions with binding proteins (Ran-BP1 or Ran-BP2) and a GTPase-activating protein called Ran-GAP1 stimulate GTP hydrolysis. Ran-BP1 is anchored in the cytoplasm. Ran-BP2 is a component of the fibers projecting from the nuclear pore into the cytoplasm. This huge (>350 kD) protein can bind up to four Ran molecules as well as Ran-GAP1 and provides a structural scaffold for the conversion of Ran-GTP into Ran-GDP at the surface of the pore. Because Ran-BP1 and Ran-BP2 are both anchored in the cytoplasm, Ran-GTP is efficiently converted to Ran-GDP only in the cytoplasm,
157
CHAPTER 9 n Nuclear Structure and Dynamics
A. Simple nuclear import
B. Nuclear import with adaptor
Cargo
Cargo 1
2
1
3
2
3
Importin α GTP
GTP
Ran-GDP
Ran-GTP
Importin β
GDP
4
Ran-GDP
Ran-GTP
Importin β
GDP
Ran-GEF (on chromatin)
4
Ran-GEF (on chromatin) 6
5
8 Ran-GAP CYTOPLASM Ran-BP
5 NUCLEUS
D. Nuclear import / export proteins
Ran-GAP Ran-BP
C. Nuclear export Ran-GAP
6
Cas
Ran-GDP Ran-GEF Ran-GDP / GTP overlap
Importin β / Ran-GTP
GTP
Ran-GDP GDP
Ran-GTP
RanGTP
Ran-GEF (on chromatin)
Ran-BP / Ran-GTP 8 7
Importin α with NLS
Exportin
Exportin complexed with importin and Ran-GTP
Ran-GAP Ran-BP
Cargo (in this case importin α)
FIGURE 9.18 NUCLEAR TRAFFICKING OF MACROMOLECULES. Nuclear import of a cargo by the import receptor importin β without (A) or with (B) the use of an adapter protein. C, Export of a cargo by the importin β-related export receptor Cas. In this case, the cargo is the import adapter importin α. Directionality is given by Ran. Ran-GTP (guanosine triphosphate) releases import cargoes in the nucleus and is required for formation of the export complex. Numbers refer to the steps described in the text. D, Crystal structures of several of the components involved in nuclear transport. (Ribbon models courtesy F. Wittinghofer, MPI Dortmund, Germany.)
yielding a nuclear/cytoplasmic ratio of Ran-GTP of approximately 200 : 1. Ran-GDP must reenter the nucleus to be recharged with GTP. Efficient Ran-GDP transport into the nucleus requires nuclear transport factor 2 (NTF2). Back in the nucleus, Ran must release its bound GDP to acquire GTP. GDP dissociation is intrinsically slow but is stimulated by a guanine nucleotide exchange factor (GEF). This protein, called regulator of chromosome condensation 1 (RCC1), is tightly associated with chromatin throughout the cell cycle. This allows nuclear import to resume immediately after the nuclear envelope reforms at the end of mitosis. Because Ran is involved in
essentially every nuclear trafficking event, the flux of this small protein across the nuclear envelope is enormous—several million molecules per minute in cultured cells.
Description of a Single Import Cycle in Detail Consider the import into the nucleus of a typical protein (Fig. 9.18): 1. In the cytoplasm, the import complex forms as importin β binds to cargo either directly or complexed with an importin α adapter (the latter is true for cargos containing the very widely studied basic NLS discussed previously).
158
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
2. The import complex binds (docks) to the cytoplasmic filaments of the nuclear pore. 3. The complex is transferred through the pore in a process that is still under investigation. A popular model proposes that the highly concentrated FG repeat-containing unstructured regions of nucleoporins associate to form a hydrogel within the pore channel that blocks most diffusion through the pore. Nuclear transport receptors (eg, importin β) bind FG repeats by trapping them between their packed helices. This locally “melts” the hydrogel, allowing the receptor and its bound cargo to drift rapidly through the gel, ultimately crossing the pore in less than 20 ms. This process does not require energy from nucleoside triphosphate hydrolysis. 4. In the nucleus, Ran-GTP binds to importin β, displacing the cargo from it. 5. Importin β/Ran-GTP shuttles back through the pore to the cytoplasm. 6. In the nucleus, if the cargo was bound directly to importin β, it is now free to function. If it was actually a cargo/importin α complex, this now encounters a nuclear export receptor called CAS. Ran-GTP and CAS bind tightly to importin α, displacing the cargo. 7. CAS carries importin α and Ran-GTP through the nuclear pores back to the cytoplasm. Thus, importin α functions as an adapter in one direction and cargo in the other. The cargo is now in the nucleus, but the system is stalled. The import receptor, importin β, is back in the cytoplasm, but in a complex with Ran-GTP that cannot bind new cargo. The import adapter, importin α, is also in the cytoplasm, but it is locked in a complex with the CAS export receptor and Ran-GTP. The solution to this problem is simple. 8. Ran-BP1, Ran-BP2, and Ran-GAP1 associated with cytoplasmic filaments of the nuclear pore catalyze the hydrolysis of GTP bound to Ran. Ran-GDP dissociates from importin α, readying it for further cycles of nuclear import. In addition, GTP hydrolysis causes the importin α/CAS/Ran-GDP complex to dissociate, allowing CAS to return to the nucleus for further cycles as an export receptor and making importin α available in the cytoplasm to bind more cargo and function as an import adapter. The hydrolysis of GTP on Ran is the only source of chemical energy required to drive the accumulation of proteins in the nucleus against a concentration gradient. Although there are several names to remember, the nuclear trafficking system is actually quite straightforward, being regulated by the state of the guanine nucleotide bound by Ran. The key point is that the GEF that charges Ran-GDP with GTP is in the nucleus and the Ran-GAPs that promote hydrolysis of GTP bound to Ran are cytoplasmic. Cargo that is meant to be imported into the nucleus is released from its carriers in the presence
of high levels of nuclear Ran-GTP. Conversely, cargo that is destined for export to the cytoplasm is picked up by its carriers only in the presence of high levels of nuclear Ran-GTP and is released when the Ran is converted to Ran-GDP in the cytoplasm. In this way, the directionality of transport is defined by the different concentrations of Ran-GDP and Ran-GTP in the cytoplasm and nucleus.
A Distinct Pathway for mRNA Export From Nuclei Small RNAs are exported by karyopherin transport receptors using Ran-GTP for directionality, but the export of mRNA depends on a different mechanism that includes numerous quality controls. mRNA is exported as very large mRNP complexes that begin to assemble during RNA processing with binding of the transcription export (TREX) complex to the mRNA. These mRNP complexes dock on the inner surface of the pore, where they are subjected to quality control by the exosome (see Fig. 11.8) and other surveillance activities. Incorrectly processed mRNAs are degraded. Correctly processed mRNAs are guided through the nuclear pore by a dimeric transport receptor, Nxf1-Nxt, which is not related to karyopherins, but also interacts with FG repeats. Adenosine triphosphate (ATP), rather than GTP hydrolysis gives directionality to the process. The ATP is used in the cytoplasm by enzymes that change the RNA structure and dissociate Nxf1-Nxt1, thus preventing the RNP from reentering the pore. Regulation of Transport Across the Nuclear Envelope Cells regulate nuclear trafficking in several ways. The first of these is to change the number of pores. In rat liver, there are 15 to 20 pores per square micrometer of nuclear envelope (~4000 per nucleus), whereas nuclei of transcriptionally quiescent avian erythrocytes have very few nuclear pore complexes. Nuclear trafficking is often regulated by phosphorylation near the NLS on the cargo. Phosphorylation adjacent to a basic NLS inhibits nuclear import. This provides a mechanism to regulate the ability of a particular cargo to enter the nucleus in response to cell cycle (see Fig. 43.6) or other signals that can be coupled to specific protein kinase activation. Traffic across the nuclear envelope is also regulated by masking or unmasking NLSs. A “nuclear” protein whose NLS is covered up is trapped in the cytoplasm. A good example is the regulation of transcription factor nuclear factor κB (NF-κB) by inhibitor of nuclear factor κB (IκB; Fig. 9.19). IκB binds to NF-κB and covers up its NLS. Because IκB also has a nuclear export signal, the NF-κB:IκB complex is entirely cytoplasmic. Following an appropriate signal (see Fig. 10.21C), IκB is degraded. This uncovers the NLS on NF-κB, allowing it to enter the nucleus. This mechanism regulates gene expression
CHAPTER 9 n Nuclear Structure and Dynamics
A. Regulation of NF-κB localization Extracellular signal IκB degraded NLS
p50
IκB
Import receptor NLSs exposed
NLS
p65
159
46XY karyotype (normal male) to develop as females. Mutations in nuclear pore proteins are also associated with developmental diseases and chromosomal translocations involving pore components and are implicated in a variety of cancers. Nuclear transport defects are also found in numerous human neurodegenerative diseases (eg, Alzheimer disease), but the mechanism is not known.
Other Uses of the Importin/Ran Switch To NUCLEUS
B. Localization of dorsal in fly embryos
Lateral view
Transverse section
FIGURE 9.19 REGULATION OF NUCLEAR FACTOR κB (NFκB) LOCALIZATION. A, The transcription factor NF-κB is kept in the cytoplasm as a result of interactions with its inhibitor IκB (inhibitor of nuclear factor κB). IκB holds NF-κB in the cytoplasm in two ways. When it binds NF-κB, it covers up the NF-κB nuclear localization sequence (NLS). Second, IκB contains a nuclear export signal, so that any NF-κB associated with it that happens to enter the nucleus is rapidly exported to the cytoplasm. B, Localization of the dorsal transcription factor (a relative of NF-κB) in Drosophila embryos. These images represent a longitudinal (left) and cross-sectional (right) view of wild-type embryos. The dorsal protein is stained with specific antibody, which appears as dark spots where it has become concentrated in the cell nuclei in the ventral portion of the embryo. (B, From Roth S, Stein D, Nusslein-Volhard C. A gradient of nuclear localization of the dorsal protein determines dorsoventral pattern in the Drosophila embryo. Cell. 1989;59:1189–1202.)
during development (Fig. 9.19B) and activation of immune cells (see Fig. 27.8), among other examples.
Disorders Associated With Defective Nuclear Trafficking In many instances, protein function appears to be regulated by adjusting its location in the cell, and nuclear transport is one mechanism controlling localization. Thus, a myriad of examples undoubtedly exist in which disruption of transport leads to disease. This area has yet to be explored systematically, but in one interesting example, human sex determination is disrupted by mutations of an NLS on the SRY (sex-determining region Y) transcription factor, a master regulator of sex determination. These NLS mutants apparently disrupt the accumulation of SRY in the nucleus at a critical stage during development, causing individuals with a
The ability of Ran-GTP to release substrates bound to importin β provides a highly efficient switch for regulating protein availability. Cells use this system to regulate several supramolecular assembly processes, including assembly of the nuclear envelope, nuclear pore, and mitotic spindle. In these processes, importin β (and occasionally importin α) acts as a negative regulator of assembly by binding to and sequestering key proteins. In the case of mitotic spindle assembly in large cells such as eggs that lack centrosomes, sequestration of key proteins blocks spindle assembly. In eggs, this block is overcome in the vicinity of chromosomes, which bind high concentrations of the GEF RCC1. Spindle assembly is triggered only after nuclear envelope breakdown, when the chromosomes come in contact with the cytoplasm (see Fig. 44.2). Conversion of Ran-GDP to Ran-GTP near the chromosomes results in Ran-GTP binding to importin β. This releases bound proteins and triggers mitotic spindle formation. Importin β and Ran also appear to regulate nuclear pore assembly in a similar way by sequestering key pore components, including the Nup107-160 complex, until they are released by Ran-GTP. ACKNOWLEDGMENTS We thank Roland Foisner, Harald Herrmann, Tom Misteli, Michael Rout, and Eric Schirmer for their advice on revisions to this chapter. SELECTED READINGS Amendola M, van Steensel B. Mechanisms and dynamics of nuclear lamina-genome interactions. Curr Opin Cell Biol. 2014;28:61-68. Azuma Y, Dasso M. The role of Ran in nuclear function. Curr Opin Cell Biol. 2000;12:302-307. Dundr M. Nuclear bodies: multifunctional companions of the genome. Curr Opin Cell Biol. 2012;24:415-422. Fernandez-Martinez J, Rout MP. A jumbo problem: mapping the structure and functions of the nuclear pore complex. Curr Opin Cell Biol. 2012;24:92-99. Forbes DJ, Travesa A, Nord MS, et al. Nuclear transport factors: global regulation of mitosis. Curr Opin Cell Biol. 2015;35:78-90. Gruenbaum Y, Foisner R. Lamins: nuclear intermediate filament proteins with fundamental functions in nuclear mechanics and genome regulation. Annu Rev Biochem. 2015;84:131-164. Kabachinski G, Schwartz TU. The nuclear pore complex—structure and function at a glance. J Cell Sci. 2015;128:423-429.
160
SECTION III n Chromatin, Chromosomes, and the Cell Nucleus
Lamond AI, Earnshaw WC. Structure and function in the nucleus. Science. 1998;280:547-553. Pombo A, Dillon N. Three-dimensional genome architecture: players and mechanisms. Nat Rev Mol Cell Biol. 2015;16:245-257. Schmidt HB, Görlich D. Transport selectivity of nuclear pores, phase separation, and membraneless organelles. Trends Biochem Sci. 2016;41:46-61.
Sosa BA, Kutay U, Schwartz TU. Structural insights into LINC complexes. Curr Opin Struct Biol. 2013;23:285-291. Wickramasinghe VO, Laskey RA. Control of mammalian gene expression by selective mRNA export. Nat Rev Mol Cell Biol. 2015;16: 431-442.
SECTION
Central Dogma: From Gene to Protein
IV
This page intentionally left blank
SECTION IV OVERVIEW T
he hugely important prediction of a structure for DNA not only led Crick and Watson to propose a general strategy for the replication of DNA (discussed in Chapter 42) but also led Francis Crick to propose the central dogma of molecular biology: that DNA is transcribed into RNA and that this RNA is then translated into protein. Chapters 10 to 12 present how this central dogma plays out at the cellular level, with one crucial addition that could not have been foreseen by Crick. This new element is the complex battery of processing events that RNAs undergo before they function as messengers, transfer vehicles, processing machines, or protein synthesizing machines in the ribosome. Chapter 10 discusses transcription of DNA sequences into RNA, the initial step in recovering the information encoded in the genome. Three eukaryotic cellular RNA polymerases have distinct specialized tasks: polymerase I transcribes ribosomal RNAs; polymerase II transcribes all messenger RNAs (mRNAs) plus a number of small RNA molecules that are involved in RNA processing; and polymerase III transcribes transfer RNAs (tRNAs) and the smallest ribosomal RNAs. These three polymerases evolved from a common ancestor and retain many shared features. However, they have acquired significant differences in the ways they act on their target genes. Eukaryotic genes contain both upstream (5′) and downstream (3′) regulatory regions that are not transcribed into RNA. Each gene has a promoter located just upstream from the site where transcription begins. Enhancers are DNA sequences that regulate transcription from a distance. Both promoter and enhancer sequences form binding sites for regulatory proteins that either stimulate or repress transcription. The chromatin organization of the DNA template and its organization within the nucleus also influence the efficiency of transcription.
Fundamental differences in the ways in which eukaryotes and prokaryotes store their genomes have had a profound influence on the structure of genes and the fate of cellular RNAs. In prokaryotes, the DNA occupies a distinct region of cytoplasm that is not bounded by a membrane. This means that transcription of DNA sequences into mRNAs and translation of mRNAs into proteins can be coupled directly, with ribosomes attaching to nascent mRNAs even before they are fully copied from the DNA template. In contrast, eukaryotes house their genomes and the machinery for RNA transcription and processing in a nucleus bounded by a nuclear envelope. Eukaryotic protein-coding RNAs must be transported across the nuclear membrane prior to their translation by ribosomes in cytoplasm. This geographic segregation, in which mRNAs are created in one subcellular compartment and used in another, has allowed the evolution of structurally complex genes whose RNA products are spliced before use. The initial RNA products of transcription of most eukaryotic genes require extensive modifications by RNA processing before they are ready to function. Chapter 11 explains that most protein-coding genes of higher eukaryotes contain protein-coding regions called exons separated by noncoding intron regions. Consequently, the initial RNA copy of these genes must be processed to remove the introns before the finished mRNA is exported from the nucleus. The nucleus is the site of many other essential RNAprocessing events. These include the addition of 5′ cap structures to mRNAs, polyadenylation of the 3′ end of mRNAs, cleavage of some RNAs into functional smaller pieces, modification of RNA bases, and a host of sometimes bizarre editing events. Both the RNA substrates for these events and many enzymes that carry out the reactions are packaged into ribonucleoprotein particles by
DNA Gene
Cell nucleus
Gene expression (Ch 10)
RNA processing (Ch 11)
Protein synthesis (Ch 12)
mRNA
(DNA replication [Ch 42]) NUCLEOPLASM
CYTOPLASM
163
specific proteins, but RNAs themselves carry out a number of enzymatic reactions, including catalysis of peptide bond formation by the ribosome. Cells also contain enzymes that fragment doublestranded RNAs into small pieces, used by other proteins to direct the silencing of the genes that encoded them. This process of RNA interference (RNAi) is critical for defense against RNA viruses and in chromatin regulation. Cell biologists also use RNAi as a technique to study gene function in the laboratory. Chapter 12 describes how ribosomes translate the sequence of nucleotide triplets in mRNAs into proteins. tRNAs act as adapters, matching specific amino acids with triplet codons in the mRNA. The RNA component of ribosomes catalyzes the transfer of each successive amino acid from its tRNA onto the C-terminus of the growing polypeptide. Every step in the process is carefully regulated to ensure quality control of the finished polypeptide. Initiation factors select the proper AUG codon in the mRNA to begin the polypeptide with a methionine residue (or formylmethionine in the case of bacteria). Elongation factors check that the proper tRNA
164
is matched with each codon before peptide bonds are formed. Although polypeptides grow at 20 residues per second, errors occur at a rate of less than one residue in a thousand. Termination factors bring protein synthesis to a close at the C-terminus of the polypeptide and recycle the ribosomal subunits for another round of translation. Although some proteins fold spontaneously into their mature form following release from a ribosome, many proteins require a helping hand to reach their properly folded state. Chapter 12 covers four types of chaperones that help proteins fold by different mechanisms. Trigger factor, which is associated with ribosomes, provides a hydrophobic groove for protein folding. Heat shock protein (Hsp) 70 and Hsp90 chaperones bind hydrophobic residues in nascent polypeptides, prevent the unfolded protein from aggregating, and thereby promote folding. Cycles of binding and release are accompanied by hydrolysis of adenosine triphosphate (ATP). Chaperonins related to GroEL provide chambers to protect proteins during folding. ATP hydrolysis releases the protein from this chamber.
CHAPTER
10
Gene Expression* E
ach organism, whether it has 600 genes (Mycoplasma), 6000 genes (budding yeast), or 25,000 genes (humans), depends on reliable mechanisms to regulate the expression of these genes (ie, turn them on and off). This is called regulation of gene expression. In simple organisms, such as bacteria and yeast, environmental signals, such as temperature or nutrient levels, control much of gene expression. In multicellular organisms, genetically programmed gene expression controls development starting from a fertilized egg. Within these organisms, cells send each other signals that control gene expression either through direct contact or via secreted molecules, such as growth factors and hormones. Given the vast numbers of genes, even in simple organisms, regulation of gene expression is complicated. Control is exerted at multiple steps, including production of messenger RNA (mRNA), translation, and protein turnover. This chapter focuses on the first of these regulatory steps: the transcription mechanisms that lead to the production of mRNA and other RNA transcripts. Proteins called transcription factors (TFs) turn genes on or off by binding to DNA regulatory sequences associated with sequences encoding the protein or RNA product of the gene. The paradigm of this level of regulation is the bacterial repressor that controls expression of genes required for lactose metabolism in Escherichia coli. In eukaryotes, TFs are numerous, representing approximately 6% of human genes. They are also quite diverse, binding to a wide range of DNA regulatory sites. Fortunately, they fall into a limited number of families with similar structures and binding mechanisms. Three types of eukaryotic DNA-dependent RNA polymerases respond to these regulatory proteins and transcribe DNA sequence into RNA. Regulation of TFs is achieved by variations in a limited number of mechanisms that
*This chapter was written by Jeffry L. Corden.
control their synthesis, transport from the cytoplasm into the nucleus, activity through posttranslational modifications or binding to small molecular ligands. One key level of regulation is transcription initiation, the first step in production of RNA transcripts. This chapter examines the basic features of both prokaryotic and eukaryotic transcription units and the transcription machinery. Regulatory TFs that control the expression of groups of genes are discussed in the context of how external signals can reprogram patterns of gene expression. Finally, the chapter addresses the mechanisms that couple transcription to the downstream processing of nascent transcripts.
Transcription Cycle Synthesis of RNA by RNA polymerases is a cyclic process that can be broken down into three sets of events: initiation, elongation, and termination (Fig. 10.1). Each of these events consists of multiple steps that can be regulated independently. In the first step of the initiation process, RNA polymerase binds to the chromosome near the beginning of the gene, forming a preinitiation complex at a sequence termed a promoter. This binding must be highly specific to
Initiation
RNA polymerase
Elongation
Termination DNA
RNA
FIGURE 10.1 THE TRANSCRIPTION CYCLE. The transcription reaction consists of three basic steps in which the RNA polymerase initiates transcription at the promoter, elongates the nascent RNA copy of one of the DNA strands, and terminates transcription recognition of the appropriate signals.
165
166
SECTION IV n Central Dogma: From Gene to Protein
distinguish promoter from nonpromoter DNA. Next, a conformational change in the polymerase–promoter complex separates the DNA strands. This open complex allows RNA polymerase access to singlestranded nucleotide bases that serve as the template to start the transcript. After formation of a phosphodiester bond between the first two complementary ribonucleotides, the polymerase translocates one base and repeats the process of phosphodiester bond formation, resulting in elongation of the nascent RNA. The elongation reaction cycle continues at an average rate of approximately 20 to 30 nucleotides per second until the complete gene has been transcribed. However, the rate of elongation is not uniform, as RNA polymerase pauses at certain sequences. The final step in the transcription cycle, termination, occurs when the polymerase reaches a signal on DNA that causes an extended pause in elongation. Given enough time, the appropriate sequence context and factors, the nascent transcript dissociates from the elongating RNA polymerase, and the DNA template returns to a base-paired duplex conformation. Ultimately, RNA polymerase dissociates from the template and is free to search for a new promoter. Regulatory molecules target each of the steps in the transcription cycle. The frequency of initiation from different promoters varies as dictated by the need for the gene product. The initiation reaction is most often regulated, presumably because this prevents synthesis of transcripts that are not needed. Elongation and termination can also be regulated, as can splicing and further processing of mRNAs and noncoding RNAs (ncRNAs) (see Chapter 11). In eukaryotes, the sum of these nuclear regulatory steps, together with cytoplasmic regulation of mRNA stability and translation efficiency, contributes to the wide variation in the abundance of various mRNAs and proteins in particular types of cells.
Transcription Unit Genetic information in DNA is transcribed in segments corresponding to one or a few genes. Gene-coding and regulatory (cis-acting) DNA sequences that direct the initiation of transcription, elongation, and termination are collectively called a transcription unit. Prokaryotic transcription units, called operons, contain more than one gene, often encoding proteins with related physiological functions (Fig. 10.2A). DNA sequences flanking the operon direct the initiation and termination of transcription. A simple eukaryotic transcription unit, such as that encoding the human hemoglobin β-chain, also has flanking regulatory sequences, but the region encoding the polypeptide is interrupted by exons (Fig. 10.2B). Mutations that reduce β-globin levels in patients with β-thalassemias can occur either in the coding region, resulting in an unstable or truncated polypeptide, or in the adjacent control regions, leading to low levels of
A. Procaryotic transcription unit DNA
I
Z
Y
A
Transcription mRNA 5'
i
z
y
a
3'
B. Eukaryotic transcription unit β-globin transcription unit on genome DNA Transcription Pre-mRNA 5'
Exon
Intron
Exon
Intron
Exon
3'
Splicing Mature humanglobin mRNA Promoter mutations result in lower level of mRNA Nonsense, frameshift, missense mutations yield unstable or inactive protein Splice-site mutations result in aberrantly spliced mRNA 3' processing site mutations result in failure to polyadenylate mRNA FIGURE 10.2 PROKARYOTIC AND EUKARYOTIC TRANSCRIPTION UNITS. A, The two transcription units required for regulation of lactose metabolism in Escherichia coli. The I gene encodes the lac repressor, while the Z, Y, and A genes encode β-galactosidase, lactose permease, and thiogalactoside transacetylase. All three genes are required for the cell to grow on media containing lactose and are coregulated as the lac operon. B, The nucleotide sequence of one of the two DNA strands is transcribed into a complementary premessenger RNA (mRNA) copy. The pre-mRNA is processed by removing introns and splicing together the protein-coding exons (orange). The DNA sequences required for expression of a functional β-globin protein are indicated in different colors (see key). Mutations in any of these sequences can lead to decreased β-globin expression.
transcription or aberrant processing of the newly synthesized RNA (see Chapter 11). Thus, the transcription unit can be thought of as a linked series of modules, all of which must be functional for the gene to be transcribed at the correct level.
Biogenesis of RNA Typical cells contain more RNA than genomic DNA. The population of RNA molecules range in size from several tens to several thousand nucleotides. In prokaryotes, translation is initiated on newly synthesized mRNA during transcription. In eukaryotes, RNA is transported from its site of synthesis in the nucleus to the cytoplasm, where most RNA is used to synthesize proteins. Eukaryotic cells have four different types of RNA: 1. Ribosomal RNA (rRNA [see Fig. 11.9]) makes up approximately 75% of the total. 2. Small, stable RNAs, such as transfer RNA (tRNA [see Fig. 12.4]), small nuclear RNAs (snRNA [see
CHAPTER 10 n Gene Expression
Chapter 11]) involved in splicing, and 5S rRNA, make up approximately 15% of the total. 3. mRNA and its precursor heterogeneous nuclear RNA (hnRNA) account for only 10% of the total. 4. ncRNAs, including micro RNAs (miRNAs), are not abundant but regulate a variety of RNA-based processes. Transcription of eukaryotic DNA in the nucleus is linked to subsequent steps that process the nascent transcript in preparation for its eventual function (see Chapter 11 for a complete discussion of these steps). Processing of mRNA precursors includes capping and methylation of the 5′ end of the nascent transcript, splicing to remove introns and modifying the 3′ end by cleavage and addition of a stretch of adenosine residues. The finished mRNA is then transported to the cytoplasm, where it serves as the template for protein synthesis. Eukaryotic ribosomal RNA is encoded in tandemly repeated genes and each gene is transcribed as a long precursor molecule, which is cleaved and modified to give the final 28S, 5.8S, and 18S RNAs (Fig. 10.3). These RNAs are assembled, together with 5S RNA and approximately 80 proteins, into ribosomes in the nucleolus.
A
Ribosomal DNA repeat Transcription unit
Nontranscribed spacer
Transcription
45S precursor RNA Cleavage Ribosomal RNAs
18S
5.8S
28S 5S RNA and ribosomal proteins
Ribosome
B Nucleolar DNA
Transcription unit
Nascent pre-rRNA molecules
Direction of transcription
Transcription unit Nontranscribed spacer
FIGURE 10.3 RIBOSOMAL RNA TRANSCRIPTION UNIT. Ribosomal RNA (rRNA) is transcribed from a set of transcription units arrayed as tandem copies of the same transcription unit. A, Map showing the arrangement of sequences in a typical ribosomal DNA repeat. B, Electron micrograph showing two active rRNA transcription units. Note that each transcription unit is transcribed by multiple RNA polymerases. As the polymerases traverse the gene, the attached nascent RNA is extended, giving a tree-like appearance. (B, Courtesy of Yvonne Osheim, University of Virginia, Charlottesville.)
167
Transfer RNA is synthesized in the nucleus and transported to the cytoplasm, where it is charged with amino acids prior to participating in protein synthesis (see Fig. 12.5). snRNAs are synthesized and processed in the nucleus. From there, they migrate to the cytoplasm, where they acquire essential proteins, and then return to the nucleus to catalyze RNA splicing reactions (see Fig. 11.11). The postsynthetic processing pathway that a particular transcript follows is dictated, in part, by the transcription machinery that is used to initiate and elongate the transcript and by certain features of the nascent RNA.
RNA Polymerases Cellular RNA polymerases synthesize a strand of nucleic acid in the 5′ to 3′ direction that is complementary to one of the chromosomal DNA strands. Even though the enzymatic reaction is similar to DNA replication (see Fig. 42.1), there are several important differences. First, RNA polymerases synthesize a strand of ribonucleotides. Second, unlike DNA polymerase, RNA polymerases can initiate transcription without a primer. Finally, unlike replication, the newly transcribed sequences do not remain base-paired with the template but are displaced from the template approximately 10 base pairs (bp) from the growing end of the nascent RNA. All known RNA polymerases share these properties and have similar structures, since they arose from a common ancestor during evolution. Bacteria have a single RNA polymerase composed of six polypeptides. Two copies of the α subunit and one each of the β, β′, and ω subunits form a five-subunit core enzyme that synthesizes RNA. The sixth subunit, σ, binds to the core enzyme to form a holoenzyme that is able to recognize promoter sequences and initiate transcription. Most eukaryotes have three different RNA polymerases (some species of plants contain four) with the largest subunits closely related to bacterial β and β′ subunits. RNA polymerases I, II, and III each have 10 core subunits, most of which are unique to each enzyme (Fig. 10.4A). RNA polymerases I and III have additional subunits similar to RNA polymerase II general TFs discussed in a following section. RNA polymerase I concentrates in the nucleolus, where it synthesizes rRNA. Throughout the nucleoplasm RNA polymerase II synthesizes mRNA and several classes of ncRNAs including some snRNAs involved in RNA splicing, and long noncoding RNAs (lncRNAs) and miRNAs implicated in gene regulation. RNA polymerase III synthesizes tRNA, 5S rRNA, and the 7S RNA of the signal recognition particle (see Fig. 21.5). RNA polymerase IV is present only in plants, where it is involved in heterochromatin formation and gene silencing. The multiple eukaryotic RNA polymerases apparently originated through duplication of primordial genes,
168
A
SECTION IV n Central Dogma: From Gene to Protein
E. coli β' α
Pol I 1
β
3
2
Pol III
1
1
2
3
4
4
5 6 7 8 9 10
5 6 7 8 9 10
α
Pol II
3
B. Ribbon
2 4
5 6 7 8 9 10
Tandem repeats of the consensus aa sequence Tyr–Ser–Pro–Thr–Ser–Pro–Ser 90°
CTD
C. Conserved sequences Pol I
N
C 90°
Pol II
CTD
Pol III
Book icon
D. Conserved residues (red)
E. coli
Yeast pol I Yeast pol II Yeast pol III Human pol II H. halobium E. coli
Book icon
K G G G G G
KEG L KEGR KQGR KEGR KEGR KQGR
FR KHMMGKRVN I RGN LMGKRVD FRGN LS GKRVD VRGN LMGKRVD FRGS L SGKRVN FRQN L LGKRVD
FIGURE 10.4 MULTIPLE RNA POLYMERASES. A, Eukaryotic cells have three different polymerases (Pol) that share three common subunits (numbers 5, 6, and 8) and have a number of other related, but distinct, subunits (indicated by related colors and distinct shading). B, A ribbon diagram of the structure of RNA polymerase II showing the arrangement of different subunits (colored as in part A). Metal ions are indicated as red balls. A prominent cleft, large enough to accommodate a DNA template, is formed between the two largest subunits. The model DNA fragment is shown for size comparison only. C, Conserved amino acid sequences are dispersed throughout the largest subunits. Red indicates sequences that are conserved among both prokaryotes and eukaryotes. Yellow represents sequences that are conserved among the three different eukaryotic RNA polymerases. D, Conserved residues are located on the inner surface of the RNA polymerase cleft. E. coli, Escherichia coli; H. halobium, Halobacterium halobium. (B, For reference, see Protein Data Bank [PDB; www.rcsb.org] file 1I50 and Cramer P, Bushnell DA, Kornberg RD. Structural basis of transcription: RNA polymerase II at 2.8 angstrom resolution. Science. 2001;292:1863–1876. D, From Zhang G, Campbell EA, Minakhin L, et al. Crystal structure of Thermus aquaticus core RNA polymerase at 3.3 Å resolution. Cell. 1999;98:811–824.)
followed by evolution of specialized functions. RNA polymerase II is the most versatile, because it must transcribe approximately 25,000 different species of human mRNAs and perhaps an equal number of ncRNAs. The relative abundance of individual mRNAs can vary widely, often in response to external signals, from just a few copies to more than 10,000 copies per cell. Thus, RNA polymerase II must recognize thousands of different promoters and transcribe them with widely varying efficiencies. In contrast, RNA polymerase I is specialized to transcribe more than 100,000 copies of rRNA per cell and RNA polymerase III synthesizes several hundred species of highly abundant transcripts. Specialization has been balanced, however, by the need to retain the structural elements required for RNA
synthesis. The subunits of both prokaryotic and eukaryotic enzymes assemble into a roughly spherical structure with a diameter of approximately 150 Å and a cleft 25 Å wide, to accommodate the DNA template (Fig. 10.4B). The two largest subunits form the framework of the structure, with two lobes that clamp down on the template DNA and form the catalytic core (Fig. 10.4C). The most conserved residues are located on the inner surfaces of the enzymes with the site of nucleotide addition on the back wall of the cleft (Fig. 10.4D). Transcription does not necessarily require such large enzymes. Bacteriophages have evolved structurally distinct, DNA-dependent RNA polymerases that are one-fifth the size of the eukaryotic enzymes yet are able to carry out complete transcription cycles. The complexity of the
eukaryotic enzymes is likely attributable to the need for regulation, with additional subunits acting as sites for interaction with regulatory proteins. Domains that differ among the three types of eukaryotic RNA polymerases are likely to interact with factors that are unique to a particular class of polymerase. One example of a classspecific domain is the carboxyl-terminal domain (CTD) of the largest subunit of RNA polymerase II, which is composed of tandem repeats of the consensus heptapeptide TyrSerProThrSerProSer. The CTD is highly phosphorylated in vivo, and the timing of CTD phosphorylation suggests that this modification may be involved in the transition between the initiation and elongation steps of transcription. By serving as a scaffold binding numerous auxiliary factors, the CTD also couples transcription with the subsequent processing of the nascent mRNA as is discussed in a later section.
RNA Polymerase Promoters Initiation of transcription requires loading of RNA polymerase onto the chromosome at the promoter of a gene or operon. A promoter can be loosely defined as a DNA sequence where RNA polymerase binds, unwinds the template and initiates transcription. Strong promoters drive the expression of genes whose products are required in abundance. Weaker promoters regulate the expression of rare proteins or RNAs. In multicellular organisms, a promoter may direct expression at a high level in some cells, at an intermediate level in others, and be repressed in yet others. Promoters in bacteria are recognized by direct interactions of the RNA polymerase σ factor with specific DNA sequences. The most common σ factor in E. coli (σ 70) recognizes two conserved six-base sequences located 10 bases (minus 10) and 35 (minus 35) upstream of the transcription start site (Fig. 10.5A). Once initiation has occurred, σ is no longer required and can dissociate from the core enzyme. Bacterial cells have several distinct σ factors, each of which binds the core enzyme and directs RNA polymerase to a subset of promoters that contain different recognition sequences, thereby promoting independently regulated transcription of genes with diverse functions. Eukaryotic promoter sequences for RNA polymerases I and II are also situated upstream of the transcription start site. RNA polymerase I recognizes a single type of promoter located upstream of each copy of the long tandem array of pre-rRNA coding sequences (Figs. 10.3B and 10.5B). The core element of this promoter overlaps the transcription start site, while an upstream control element located approximately 100 bp from the start site stimulates transcription. Comparison of the first eukaryotic protein-coding gene sequences revealed a conserved consensus sequence—TATAAAA—called a TATA box, located approximately 30 bp upstream of the transcription start
169
CHAPTER 10 n Gene Expression
A. Prokaryotic promoter 10 bp 5'
-35 (6 bp)
(17–19 bp)
-10 (6 bp)
+1
3'
DNA
B. Eukaryotic Pol I promoter 5'
-200
-100
DNA
Upstream element
50 bp -50
+20
C. Eukaryotic Pol II promoter -37 to -32 -31 to -26 TATA 5' BRE
TFIIB TATA recognition box element
DNA
C C A C G C C TATA A A GGG
5'
+8 +20
A box
DNA
10 bp
+1 INR
+28 to +34 3' DPE
Initiator
Downstream promoter element
Py Py A NT APy Py
D. Eukaryotic Pol III promoter: tRNA genes +50 +61
AG A C G T G G T
E. Eukaryotic Pol III promoter: 5S rRNA gene 3'
5'
B box
3'
Core element
DNA
25 bp
+40
+80
C box
3'
FIGURE 10.5 PROKARYOTIC AND EUKARYOTIC PROMOTERS. The prokaryotic (A) and three eukaryotic (B–E) RNA polymerases recognize different promoter sequences. Positions of promoter elements are indicated with respect to the start of transcription (+1). For the RNA polymerase II (Pol II) promoter elements, the consensus sequences are shown. Not all polymerase II promoters contain all these elements. Pol, polymerase; rRNA, ribosomal RNA; TF, transcription factor; tRNA, transfer RNA.
site of many genes transcribed by RNA polymerase II (Fig. 10.5C). In addition to the TATA box, a less-conserved promoter element, the initiator, is found in the vicinity of the transcription start site of many genes. Some genes transcribed by polymerase II do not contain TATA boxes but may contain strong initiator elements. Together, these two elements account for the basal promoter activity of most protein-coding genes. Two types of RNA polymerase III promoters have key elements within the transcribed sequences (Fig. 10.5D– E). tRNA genes contain two 11-bp elements, the A box and B box, centered approximately 15 bp from the 5′ and 3′ ends of the coding sequence, respectively. The 5S-rRNA gene contains a single internal element, the C box, located in the center of the coding region. Given the differences in classes of eukaryotic promoters, it is not surprising that each type of polymerase uses different proteins to recognize the promoter sequences.
Transcription Initiation The loading of RNA polymerase onto double-stranded genomic DNA at a promoter sequence is best understood in prokaryotes and is discussed first before initiation by eukaryotes. Initiation takes place in a series of defined
170
SECTION IV n Central Dogma: From Gene to Protein
A. Closed complex (binding)
B. Open complex (melting)
C. Transcribing complex
Jaws of clamp
RNA exit channel
Nucleotide entry channel
FIGURE 10.6 THREE STEPS IN RNA POLYMERASE INITIATION. A, In the closed complex, the double-stranded promoter DNA is recognized by σ factor domains on the surface of the holoenzyme. Double-stranded DNA then transfers into the active site shown here. B, The open complex forms by unwinding DNA surrounding the transcription start site and positioning the single-stranded template in the active site of the polymerase. C, The initiation reaction in the context of the transcription cycle.
TABLE 10.1 Summary of Eukaryotic RNA Polymerase II General Transcription Factors Factor
Number of Subunits
Subunit M (kD)
Functions
TFIIA TFIIB
3
12, 19, 35
Stabilizes binding of TBP and TFIIB
1
25
TFIID
12
Binds TBP, selects start site, and recruits polymerase II
15–250
(TBP)
Interacts with regulatory factors
1
38
Subunit of TFIID; specifically recognizes the TATA box
TFIIE
2
34, 57
Recruits TFIIH
TFIIF
2
30, 74
Binds polymerase II and TFIIB
TFIIH
9
35–98
Unwinds promoter DNA; phosphorylates CTD (C-terminal domain of RNA polymerase II)
Polymerase II
12
10–220
Catalyzes RNA synthesis
TOTALS
42
~1000
TBP, TATA box–binding protein.
steps (Fig. 10.6). First, RNA polymerase holoenzyme binds to the double-stranded promoter, forming what is called the closed complex. Interactions between the σ factor and bases in the −10 and −35 elements of the promoter determine the specificity and strength of this interaction (Fig. 10.5). The second step in initiation is the formation of an open complex by separation of the two strands of DNA around the transcription start site producing a 14 base transcription bubble. This unpairing is accompanied by a conformational change in the polymerase that positions the single-stranded DNA template in the active site and narrows the DNA-binding cleft, effectively closing the polymerase clamp. In the next step, the DNA template in the active site base-pairs with the first two ribonucleotides, and the first phosphodiester bond is catalyzed. The process of single nucleotide addition is repeated until the nascent RNA is eight to nine bases long, at which point addition of bases to the growing RNA chain results in the unpairing of the 5′ RNA base of the RNA-DNA hybrid, and the nascent RNA begins to exit through a channel on the surface of the polymerase. The resulting conformational change in polymerase leads to the release of σ factor and formation
of a stable ternary (three-way) complex containing RNA polymerase, the DNA template, and the nascent RNA.
General Eukaryotic Transcription Factors Eukaryotic RNA polymerases require multiple initiation factors to start transcription. All the RNA polymerases use a TATA box–binding protein, but most of the other initiation factors are unique for each class. On the other hand, each RNA polymerase uses the same general transcription factors (GTFs) for most promoters. GTFs are remarkably conserved among eukaryotes. The next sections describe transcription initiation by the three forms of eukaryotic RNA polymerase. RNA Polymerase II Factors Initiation of transcription by RNA polymerase II in vitro depends on the ordered assembly of more than 20 GTFs at the promoter (Table 10.1). Assembly of this RNA polymerase II preinitiation complex begins with binding of TFIID, a large protein complex (~700 kD) consisting of TATA box–binding protein (TBP) and TBP-associated factors called TAFIIs (Fig. 10.7A). TBP alone is sufficient for basal transcription, while
CHAPTER 10 n Gene Expression
171
N
A
B TATA
II D
Gene C
TAFs TBP
N
C
TBP
II A
C
II B
CTD
Pol II II F
D.
TBP
II E
C
II H Preinitiation complex
TAFs B TBP A F
H
N
E
TF II B
Elongation factors
Direction of transcription +1
FIGURE 10.7 RNA POLYMERASE II PREINITIATION COMPLEX ON THE ADENOVIRUS-2 MAJOR LATE PROMOTER DNA. A, The sequential assembly of general transcription factors leads to a preinitiation complex with the promoter region in the closed complex. Helicase activities present in transcription factor IIH (TFIIH) use the energy of adenosine triphosphate (ATP) to unwind the promoter, leading to formation of an open complex. B, Binding of the TATA box–binding protein (TBP) leads to C, a pronounced bend in the DNA. D, TFIIB interacts both upstream and downstream of the TATA box and directs RNA polymerase to the transcription start site. (B–D, For reference, see PDB file 1VOL. TBP + DNA coordinates courtesy Stephen Burley, Rockefeller University, New York.)
TBP-associated factors (TAFs) apparently serve as targets for further activation of transcription (see subsequent sections). DNA binding by TBP is provided by a highly conserved C-terminal of 180 amino acids, which forms a saddle-shaped monomer with an axis of dyad symmetry (Fig. 10.7B). The underside of the TBP “saddle” binds to the minor groove of the TATA sequence, which is splayed open in the process. A
pronounced DNA bend is produced at each end of the TATAAA element by the intercalation of phenylalanine side chains (Fig. 10.7C). The TFIID-TATA box complex serves as a binding site for additional GTFs and positive and negative regulators. TFIIA binding stabilizes the TBP-DNA interaction and prevents the binding of repressors that would otherwise block further initiation complex formation.
172
SECTION IV n Central Dogma: From Gene to Protein
The next step in assembly of the initiation complex is adding TFIIB, which binds to one side of TBP, making contacts with DNA upstream and downstream of the TATA box (Fig. 10.7D). Mutations in the yeast gene encoding TFIIB alter mRNA start-site selection, indicating that TFIIB establishes the spacing between the TATA box and the transcription start site. TFIIB interacts directly with TBP and RNA polymerase II and is essential for the next steps in initiation complex assembly. RNA polymerase II joins the preinitiation complex (Fig. 10.7A) associated with TFIIF. This factor stabilizes the interaction of RNA polymerase II with TFIIB and TBP. TFIIF also binds to free polymerase and prevents interactions with nonpromoter DNA sites. TFIIH and its stimulatory factor TFIIE are the final general factors to enter the preinitiation complex. Their binding stabilizes contacts between proteins and DNA in the vicinity of the transcription start site. TFIIH contains eight polypeptides, several of which have functions outside of transcription initiation. Helicases associated with TFIIH use energy from adenosine triphosphate (ATP) hydrolysis to unwind a short stretch of promoter DNA at the transcription start site. This separation of DNA strands allows RNA polymerase II to recognize the template strand, bind the complementary nucleotides, and synthesize the first few phosphodiester bonds. TFIIH also contains a protein kinase that phosphorylates the CTD. This is Cdk-activating kinase (CAK), itself a Cdk-cyclin complex that phosphorylates and activates other cyclin-dependent kinases (see Fig. 40.14). In the initiation complex, phosphorylation of the CTD releases it from interactions with GTFs and mediator (see later section) allowing it to leave the promoter and enter the transcription elongation phase. Other TFIIH sub units have been identified as components of the DNA repair machinery. Several genes encoding TFIIH subunits are mutated in xeroderma pigmentosa, a human disease with defects in DNA excision repair. This suggests that TFIIH might link transcription to DNA repair (see Box 43.1).
Initiation by RNA Polymerases I and III Distinct initiation complexes initiate transcription at RNA polymerase I and III promoters (Fig. 10.8). RNA polymerases I (Pol I) and III (Pol III) contain subunits related to polymerase II (Pol II) GTFs TFIIF and TFIIE. Unique TFIIB-related factors provide additional GTF functions for Pol I and Pol III. The Pol I upstream binding factor binds to both the upstream control element and part of the core element of the promoter (Fig. 10.8A). A protein complex called SL1 stabilizes this initial complex. SL1 consists of TBP and TAFs specific to RNA Pol I, including one related to TFIIB. A unique factor Rrn3 binds Pol I and modulates rRNA transcription in response to nutrient availability.
+1
A. Pol I rRNA promotors UCE
Core element
Pre-rRNA gene
TBP TAFs Pol I UBF UBF
B. Pol III tRNA promotor
TFIIIC B''
TBP BRF
TFIIIB
Pol III
C. Pol III 5S-rRNA promotor
TFIIIC B''
TBP
TFIIIA
BRF
TFIIIB
Pol III
FIGURE 10.8 RNA POLYMERASE I AND III PREINITIATION COMPLEXES. A, Ribosomal RNA promoters assemble a preinitiation complex. (UCE, upstream control element.) This complex consists of an upstream binding factor (UBF) and a multisubunit factor that contains TATA box–binding protein (TBP). Together, these factors recruit RNA polymerase I. B–C, Initiation at RNA polymerase III promoters requires recognition of sequences within the transcribed sequences. These sequences differ for transfer RNA (tRNA) and 5S ribosomal genes. B, In the case of tRNA genes, only TFIIIC is required for specific binding. C, For 5S genes, the internal element is recognized by the specific DNA-binding factor TFIIIA. BRF, TFIIB-related factor.
The assembly of RNA Pol III initiation complexes differs at various promoters. Initiation at tRNA genes begins with TFIIIC binding to the A and B boxes (Fig. 10.8B); TFIIIB then binds upstream of the A box at a sequence determined both by an interaction with TFIIIC and through the DNA-binding capacity of TBP. Once the TFIIIC–TFIIIB complex has assembled, RNA Pol III initiates transcription. Multiple rounds of initiation can occur on the stable transfer DNA (tDNA)–TFIIIC–TFIIIB complex. Transcription of 5S rRNA genes requires an additional factor called TFIIIA that recognizes the C box located near the center of the 5S rRNA coding region. TFIIIC then binds with contacts on each side of TFIIIA, similar to the A and B boxes contacting tRNA genes. Finally, TFIIIB binds through interactions with TFIIIC and DNA, and the resulting preinitiation complex is recognized by RNA Pol III.
Summary of the Eukaryotic Basal Transcription Machinery Despite the evolutionary divergence of the multiple eukaryotic RNA polymerases and the specialization of each polymerase for a unique set of promoters, the fundamental mechanisms of transcription have been conserved. This conservation is reflected not only in similar sequences of the subunits of the polymerases themselves but also in the presence of TBP and TFIIB homologs among the GTFs used by each class of polymerase. Indeed, Archaea, which have only a single RNA polymerase, contain both TBP and TFIIB suggesting that initiation mechanisms employing GTFs evolved before the duplication of the RNA polymerases. Why are so many factors required to make a transcript? Part of the complexity might be necessary to generate multiple sites for interaction with regulatory factors that could either activate or repress the assembly or function of the preinitiation complex. A second role for the complex set of factors could be to target polymerases to specific sites in the nucleus. Finally, some factors could help load elongation, splicing, or termination factors onto the RNA polymerases.
Transcription Elongation and Termination The final stage of initiation leads to elongation and movement of the polymerase away from the promoter. This process of promoter clearance is associated with structural changes in the polymerase, which prepare it for efficient RNA synthesis and render it susceptible to the action of factors that regulate elongation. Such regulatory factors, together with structural features of the nascent transcript, influence elongation and can trigger the termination of transcription and the dissociation of the ternary elongation complex containing the DNA template, nascent RNA, and RNA polymerase. The termination reaction typically occurs at the 3′ end of the gene or operon and serves both to recycle RNA polymerase for additional initiation reactions as well as to ensure that adjacent genes are not inadvertently transcribed.
Transcription Elongation Complex Efficient synthesis of RNA requires balancing two competing demands. First, the elongation complex must be very stable, because premature dissociation from DNA produces defective partial transcripts and requires the polymerase to restart transcription from the promoter. However, the complex must also be bound loosely enough so that the polymerase can easily translocate along the DNA template. RNA polymerase evolved to meet these needs. The cleft formed at the interface between the two largest subunits is open when the polymerase is in the initiation
CHAPTER 10 n Gene Expression
173
complex. Once the first few RNA phosphodiester bonds form, the polymerase undergoes a conformational change. Subunits at the outer edge of the cleft close like jaws to encircle the DNA template. In this structure, the front end of the transcription “bubble” (an unpaired segment of the DNA template) is positioned at the back wall of the cleft, close to the catalytic center. The elongation complex is highly efficient and can function continuously for the 17 hours required to transcribe the more than 2 million-bp mammalian dystrophin gene (see Fig. 39.17).
Catalytic Cycle The DNA-dependent RNA polymerases catalyze synthesis of an RNA polymer from ribonucleoside 5′-triphosphates (ATP, guanosine triphosphate [GTP], cytidine triphosphate [CTP], and uridine triphosphate [UTP]) according to the following reaction: ( NMP )n + NTP → ( NMP )n+1 + PPi where (NMP)n is the RNA polymer; NTP is ATP, UTP, CTP, or GTP; and PPi is pyrophosphate. Polymerase extends the RNA chain in the 5′ to 3′ direction by adding ribonucleotide units to the chain’s 3′ OH end. Selection of the incoming nucleoside triphosphate (NTP) is directed by the DNA template and takes place at the transcription bubble (Fig. 10.9). The 3′ hydroxyl group acts as a nucleophile, attacking the α-phosphate of the incoming NTP in a reaction similar to that seen in DNA replication (see Fig. 42.1). The chain elongation reaction proceeds in vivo at a rate of 30 to 100 nucleotides per second and is facilitated by a set of flexible protein modules surrounding the polymerase active site.
Pausing, Arrest, and Termination Following the addition of each nucleotide, RNA polymerase may add an additional nucleotide, pause, move in reverse, or terminate (Fig. 10.9B). The relative probabilities of these alternative reactions depend on interactions between the transcription complex and the template, the nascent RNA transcript, and regulatory TFs. RNA polymerase does not elongate at a constant rate. Instead, it synthesizes RNA in short spurts between pauses. A pause of short duration can be caused by low NTP concentrations or alternatively by the transient unpairing of the 3′ end of the nascent transcript and template. Longer pauses are provoked by the presence, in the nascent RNA, of short (~20 bp) self-complementary sequences that can fold to form a stem-loop or hairpin, or the presence of a weak RNA–DNA hybrid. The presence of an unstable RNA–DNA hybrid can arise from the misincorporation of an NTP leading to an unpaired base in the hybrid. In this case, the RNA polymerase can backtrack or slide backward on the template (Fig. 10.9C).
174
SECTION IV n Central Dogma: From Gene to Protein
A. RNA polymerase
C. Elongating 5' 3'
3' 3'
5'
5' Backsliding
D. Paused
Nascent RNA
5' 3'
3' 5' 5'
B. Active site E. Arrested
Backsliding
5' Termination Editing Elongation 3' OH RNA transcript
3'
3'
5' 5' 3'
Template Position –1
1 +1
Next NTP
FIGURE 10.9 TRANSCRIPTION ELONGATION. A, Model of the transcription elongation complex consisting of RNA polymerase, template DNA, and nascent RNA transcript. RNA polymerases interact with the template upstream and downstream of the transcription bubble. B, The active site of RNA polymerase positions the growing end of the nascent transcript in the appropriate location for the addition of the next nucleoside triphosphate (NTP). After each single nucleotide addition, the polymerase may translocate forward and repeat the nucleotide addition (C), slide backward and pause for a variable time (D), or slide further backward, causing a transcription arrest that is reversed when the polymerase cleaves the nascent RNA (E).
Backward movement of the transcription bubble is accompanied by a zippering movement of the RNA–DNA hybrid in which the nascent RNA in the exit channel rehybridizes with upstream template sequences and the 3′ end of the transcript unpairs from the hybrid and is extruded through the same channel that NTPs use to enter the active site. If the polymerase backtracks more than a few nucleotides the complex becomes arrested and cannot resume elongation without assistance of additional factors. For example, transcription elongation factors can bind in the NTP channel of arrested complexes and activate the RNA polymerase to cleave the backtracked RNA. The new 3′ terminal residue is correctly positioned for incorporation of the next complementary NTP (Fig. 10.9C–E). This editing process increases the fidelity of transcription. Pausing also occurs following transcription of U-rich sequences, and in prokaryotes this is often associated with transcription termination.
Termination When elongating RNA polymerase reaches the end of a gene or operon, specific sequences in the RNA
called terminators trigger the release of the transcript and dissociation of the RNA polymerase. Bacteria have two types of terminators. The first are called intrinsic (or rho-independent) terminators, because they function in the absence of any protein factors (Fig. 10.10A). Intrinsic terminators consist of two sequence elements in the RNA: a stable GC-rich hairpin and a run of about eight consecutive U residues. As the first of these elements is synthesized, it forms a hairpin, causing polymerase to pause with less stable U:A bp (with only two H bonds [see Fig. 3.14]) in the hybrid. The nascent transcript is released from this complex, terminating transcription. The second type of prokaryotic termination requires a protein factor called rho (Fig. 10.10B). Rho is a hexameric helicase that binds cytosine-rich sequences and uses ATP hydrolysis to translocate along the nascent transcript in the 5′ to 3′ direction, essentially chasing the RNA polymerase. When polymerase pauses, rho can catch up and use the energy derived from ATP hydrolysis to pull the RNA out of the transcription elongation complex. Eukaryotic RNA polymerases evolved distinct mechanisms for termination. RNA Pol III requires no protein
CHAPTER 10 n Gene Expression
A. Rho-independent termination
175
B. Rho-dependent termination
CC CCC C
CG
C C
G C G G
Rho hexamer binds specific C-rich sequences of RNA
G- and C-rich selfcomplementary region forms hairpin
G- and C-rich
Rho migrates 5' to 3' to signal release of pol on contact
G C C G G C G C
Hairpin structure induces release of paused polymerase from polyU sequence
C G C G CG C G U
Rho's helicase activity unwinds RNA/DNA duplex releasing RNA
UU U U
U
FIGURE 10.10 PROKARYOTIC TRANSCRIPTION TERMINATION. A, Rho-independent termination is directed by sequences in the nascent transcript that operate in the absence of any additional factors. B, The bacterial termination factor rho translocates along the nascent RNA and on reaching the RNA polymerase (pol) causes the disassembly of the elongation complex.
factors but terminates efficiently after transcribing four to six consecutive U residues, presumably owing to instability of the RNA–DNA hybrid in the enzyme active site. RNA Pol I terminates in response to a protein factor that blocks further elongation by binding to a DNA sequence downstream of the termination site, leaving an inherently unstable U-rich RNA–DNA hybrid in the active site. The RNA Pol II termination mechanism is more complex, requiring a large multiprotein complex that recognizes the poly(A) addition signal sequence in the nascent transcript (see Fig. 11.3 for pre-mRNA processing). Deletion or mutation of the poly(A) signal results in a failure to terminate messages at the appropriate site. Thus, RNA Pol II termination is coupled to 3′-end processing (see Chapter 11).
Gene-Specific Transcription Regulation Transcription initiation is the critical first step in determining that each gene is expressed at the appropriate level in each cell. Depending mainly on the sequence of the promoter and other regulatory sequences, expression can be constitutive or influenced by regulatory proteins. This section discusses proteins that regulate transcription of specific genes either positively or negatively. The discussion starts with a prokaryotic example and then covers a variety of eukaryotic regulators. Although the details differ in prokaryotes
and eukaryotes, many of the basic principles are the same.
Regulation of Transcription Initiation in Prokaryotes Prokaryotes typically regulate gene expression in response to environmental cues such as the presence of nutrients in the growth medium (see Fig. 27.11). These signals are transmitted to the appropriate genes through regulatory proteins that bind to specific sequences near the genes they control to either activate or repress transcription. Both of these regulatory mechanisms come into play in regulation of the E. coli lactose (lac) operon (Figs. 10.2A and 10.11). The genes expressed from this operon are required for cells to metabolize lactose but are not expressed in the absence of lactose. Genetic studies in the 1960s showed that the gene upstream of the lac operon (I in Fig. 10.2A) encodes a repressor (lac repressor) that blocks expression of the lac operon in the absence of lactose (Fig. 10.11). The lac repressor binds to a site called an operator that overlaps the RNA polymerase binding site in the lac promoter. Lactose binding changes the conformation of the repressor, so it dissociates from DNA, allowing RNA polymerase to bind the promoter. Full expression of the lac operon requires the catabolite activator protein (CAP), another allosteric protein that binds DNA just upstream of the lac promoter. CAP is activated by a conformational change induced when
176
SECTION IV n Central Dogma: From Gene to Protein
Glucose
Lactose
A. Lac regulation physiology
High level of transcription
B. Lac regulation mechanics
Lac repressor (inactive)
CAP (inactive) Bacterial polymerase
Lactose inducer
cAMP CAP –35 –10 site
Lac Z
Lac repressor (active) Active CAP attracts polymerase
Low level of transcription
Repressor binding zone half-sites
Lac repressor CAP-binding zone
Transcription
5' CAACGCAATTAATGTGAGTTAGCTCACTCATTAGGCACCCCAGGCTTTACACTTTATGCTTCCGGCTCGTATGTTGTGTGGAATTGTGAGCGGATAACAATTTCACACAGGAAACAGCT 3' GTTGCGTTAATTACACTCAATCGAGTGAGTAATCCGTGGGGTCCGAAATGTGAAATACGAAGGCCGAGCATACAACACACCTTAACACTCGCCTATTGTTAAAGTGTGTCCTTTGTCGA -35 -10 +1
No transcription Lac operator
Lac operator
Polymerase-binding zone
FIGURE 10.11 REGULATION OF THE LACTOSE (LAC) OPERON. A, RNA polymerase (green) binding to the lac promoter is regulated by the binding of repressor or activator (catabolite activator protein [CAP]). B, Binding sites for CAP and the repressor at the lac operon. The main repressor-binding site overlaps the promoter and blocks access of RNA polymerase. Additional lac repressor-binding sites are located upstream and downstream of the promoter. Lac repressor can form a tetramer and thus bind two operators, forming a loop in the lac operon DNA. Inducer (eg, lactose) binding dramatically alters the conformation of the lac repressor diminishing its affinity for the operator. CAP binds just upstream of the promoter where it can stabilize the bound RNA polymerase.
it binds cyclic adenosine monophosphate (cAMP), which the cell produces when the intracellular glucose concentration is low. Active CAP bound to its site stabilizes the otherwise weak interaction of RNA polymerase with the promoter. The resulting activation allows maximum expression of the lac operon in the presence of lactose and the absence of glucose. Control of lac gene expression by opposing repressors and activators is an example of regulation at the first step in transcription initiation, binding of RNA polymerase to the promoter. Regulating access of RNA polymerase to promoters is a common form of transcription regulation in both prokaryotes and eukaryotes.
Overview of Eukaryotic Gene-Specific Transcription While recruitment of RNA polymerase to the promoter remains a key step in eukaryotic transcription regulation, there are additional layers of complexity. First, DNA is bound by histones and packaged in nucleosomes (see Fig. 8.1) that can block binding of TFs and RNA polymerase. Overcoming this generalized repressive effect requires activators that alter chromatin structure allowing the recruitment of RNA polymerase. Another major difference is that eukaryotic TFs bound tens to hundreds of kilobases away from the promoter
Gene-specific transcription factors
Enhancer
Coregulators: Mediator, ATP-dependent nucleosome remodelers, histone modifiers and negative cofactors
Promotor proximal elements
Mediator TAFs TBP TF IID
+1 Pol II
FIGURE 10.12 NETWORK OF INTERACTIONS THAT REGULATE RNA POLYMERASE II. Input comes from transcription factors bound to promoter proximal elements and enhancers and from coregulators that modify chromatin.
can activate transcription. In many cases these genespecific TFs do not act directly on polymerase but require coregulators that act as a bridge between gene-specific factors, the chromatin template and RNA polymerase with its associated GTFs (Fig. 10.12). The following sections explain how detailed mechanistic studies of a small set of model genes provided
CHAPTER 10 n Gene Expression
Transcription factor binding sites
D
Nucleosome locations Nucleosomefree region
C
Number of sequence reads
B
Nucleosomefree region
A
177
Histone modifications
FIGURE 10.13 CHROMATIN IMMUNOPRECIPITATION COUPLED WITH HIGH-THROUGHPUT SEQUENCING (CHIP-SEQ) MAPS PROTEIN BINDING SITES AND HISTONE MODIFICATIONS. A, Experimental protocol. B, Frequency of DNA reads of DNA associated with three transcription factors associated with the UCHL5 gene. C, Frequency of DNA reads of DNA associated nucleosomes along two budding yeast genes. Nucleosomes are spaced regularly along the DEP1 gene but not along the CYS3 gene. D, Histone modifications along an active and an inactive gene. The thickness of the bar represents the frequency of each modification. (B, From Farnham P. Insights from genomic profiling of transcription factors. Nat Rev Gen. 2009;10:605–616. C, From Barth TK, Imhof A. Fast signals and slow marks: the dynamics of histone modifications. Trends Biochem Sci. 2010;35:618–626. D, Based on data from Jiang C, Pugh F. Nucleosome positioning and gene regulation: advances through genomics. Nat Rev Gen. 2010;10:161–172.)
the concepts for our current understanding of how thousands of different proteins combine to regulate tens of thousands of different promoters. Genomewide studies have refined our understanding of how these regulatory mechanisms function in more global gene regulatory networks. Before addressing specific mechanisms, we consider techniques for mapping regulatory proteins to specific sites in the eukaryotic genome.
Mapping Transcription Components on the Genome One of the key advances in transcription research has been to map transcription regulators and transcripts on a genome-wide basis. Fig. 10.13 describes one of these approaches: chromatin immunoprecipitation coupled with high-throughput sequencing (ChIP-seq). This approach yields a genome-wide snapshot of the positions of RNA polymerase, TFs, and histones
on DNA and the modifications of these components. This comprehensive view of the distributions of transcription components has yielded novel insights about the locations of regulatory sequences and the presence of different combinations of histone modifications. This information will undoubtedly guide future experiments where the regulatory mechanisms are not yet clear.
Chromatin and Transcription DNA in eukaryotic cells associates with an equal mass of protein to form chromatin (see Chapter 8). Packaging DNA in arrays of nucleosomes compacts the DNA and restricts access of transcription proteins to the DNA template. Understanding how the transcription machinery interacts with nucleosomes is a key to understanding eukaryotic transcription regulation.
178
SECTION IV n Central Dogma: From Gene to Protein
Gene activation often involves disruption or displacement of nucleosomes located on specific regulatory regions. Before the discussion of specific mechanisms, it is useful to consider some aspects of nucleosome structure. The nucleosome consists of DNA wrapped in a left-handed helix around an octamer of histone subunits (see Fig. 8.1). The histone core makes numerous contacts with the DNA minor groove and phosphate backbone, leading to tight but relatively nonspecific binding. This aspect of the nucleosome allows for a dynamic association with DNA, because binding of the histone core to DNA is nearly as energetically favorable for all sequences. However, nucleosomes are not positioned uniformly along the DNA. First, some AT-rich sequences do not bend in a manner that can form a stable nucleosome. Such sequences are often found in promoter regions. Second, nucleosomes are less stable if the histones are modified, for example by acetylation or the inclusion of variant histone proteins. The presence of unstable nucleosomes enables the transcription machinery to access key regulatory sequences. Nucleosome remodeling complexes can either expose or shield regulatory elements by altering the location of nucleosomes on the DNA template. These multiprotein remodeling complexes use energy from ATP hydrolysis to destabilize interactions between histones and DNA thus altering the position of the nucleosome and “remodeling” the chromatin. One example is the SWI/SNF (yeast mating type switching defective/ sucrose nonfermenting) complex that is recruited to a specific subset of genes through interactions with transcription activators. The resulting remodeling of nucleosomes in the vicinity of promoters may be required to form a stable preinitiation complex. Genomic mapping of histones (Fig. 10.13) shows that most Pol II promoters are free of nucleosomes. Histone Modifications and Gene Expression Specific enzymes modify the histone tails with diverse chemical groups, often on lysine residues. Gene regulatory proteins recruit the modifying enzymes to chromatin generally as part of larger complexes (Table 10.2). Activator proteins generally recruit histone acetyltransfer-
ases, while histone deacetylases are part of corepressor complexes. The hundreds of chromatin regulatory complexes in cells give rise to different chromatin states defined both by their pattern of histone modification and by their transcription (Fig. 10.13). Silent chromatin is not transcribed and has nucleosomes with H3K9me3 or H3K27me3 modifications spanning multiple genes in heterochromatin (see Chapter 8). Active chromatin often contains nucleosomes with H3K4ac or H3K4me, H4K8ac modifications, often in promoter–proximal nucleosomes (Fig. 10.13). In stem cells (see Box 41.2), many promoter–proximal regions contain both activating and repressing marks, so the genes are thought to be poised to be either activated or repressed as downstream signals dictate. Most chromatin regulators are parts of larger complexes containing protein modules that recognize histone modifications such as bromodomains that interact with acetylated tails or chromodomains that bind methylated tails. For example, the SAGA histone acetyltransferase complex contains a bromodomain that anchors the complex to chromatin, facilitating further modification of regions that are already acetylated. A subunit of TFIID also contains a bromodomain that can facilitate the binding of TFIID to acetylated nucleosomes associated with active chromatin. Similarly, a number of histone methyltransferases contain chromodomains and are therefore targeted to their substrates by preexisting histone methylation. The following sections describe examples of how TFs, chromatin regulators, and the general transcription machinery interact to regulate eukaryotic genes.
Gene-Specific Eukaryotic Transcription Factors Eukaryotic TFs bind specific DNA sequences associated with the genes they regulate. This binding leads to activation or repression of transcription in a spatially and temporally controlled manner. In the simplest cases, the TF interacts directly with RNA Pol II and the GTFs but in more complex cases, the interaction may involve a coactivator or corepressor (see the following section). Current estimates indicate that approximately 6% of
TABLE 10.2 Nucleosome-Modifying Complexes Name
Subunits
Catalytic Activity
Histone-Interacting Domain
Target Histone(s)
SAGA
15
Histone acetylase
Bromodomain
H3, H2B
NuA4
6
Histone acetylase
Chromodomain
H4
P300
1
Histone acetylase
Bromodomain
H2A, H2B, H3, H4
NuRD
9
Histone deacetylase
Chromodomain
?
SIR2
3
Histone deacetylase
Neither
H4
MLL
7
Histone methylase
Neither
H3 (lysine 4)
CHAPTER 10 n Gene Expression
the coding capacity of the human genome (more than 1000 genes) is devoted to TFs that recognize specific DNA sequences. The following sections discuss the functional organization of these proteins, how they recognize DNA and how they interact with chromatin and the GTFs. DNA-Binding Domains Binding proteins to specific DNA sequences requires recognition of a pattern of bases along the double helix. The richest source of DNA sequence specificity comes from the chemical groups exposed in the major groove. Most specific DNA-binding proteins probe the major groove of the double helix with a small structural element (usually, an α-helix) with a shape complementary to the surface topography of a particular DNA sequence. The correct DNA sequence is recognized through multiple interactions between amino acid side chains in the recognition helix and the chemical groups on the DNA bases in the major groove. Single amino acid changes in the recognition helix can change the DNA sequence that is recognized. Protein-DNA complexes are stabilized by additional contacts between amino acid side chains and deoxyribose rings and phosphate groups or by bending the DNA.
A. Homeodomain
179
DNA recognition domains of specific TFs typically interact with only 3 to 6 bp of DNA. Given the size and complexity of the typical mammalian genome, a sequence must be approximately 16 bp long to occur by chance only once. How then can TFs recognize specific genes among the vast number of close but nonidentical sequences? Two strategies increase the length of the specific sequence to be recognized. The protein can either use several recognition elements or dimerize with itself or other DNA-binding proteins. Protein dimers can recognize sequences with twofold rotational symmetry. DNA-binding proteins can be grouped into families based on the structure of the domains used for DNA sequence recognition (Fig. 10.14). These include the helix-turn-helix (HTH) proteins, homeodomains, zinc finger proteins, steroid receptors, leucine zipper proteins, and helix-loop-helix proteins. Although these families include most known TFs, there remain other, less-common recognition domains. Within a given family, the recognition domain of each TF has an amino acid sequence that targets the protein to a particular DNA sequence. Conversely, different families of TFs can recognize the same regulatory sequence. The following sections discuss several of the more common eukaryotic DNA-binding domains.
B. Zinc fingers
NNTAATGGNN NNATTACCNN
C. Glucocorticoid receptor
NAGAACANNNTGTTCTN NTCTTGTNNNACAAGAN NNGCGTGGGCGNN NNCGCACCCGCNN
D. Basic region zipper
E. Factor 1 homodimer
F. Factor 2 homodimer
G. Factor 1/2 heterodimer
NNTGAGTCANN N NA C T CAG T N N
FIGURE 10.14 MOLECULAR STRUCTURES OF TRANSCRIPTION FACTOR DNA-BINDING DOMAINS. Recognition of specific DNA sequences requires interactions between amino acid side chains in the protein and chemical groups on the DNA bases. In each of the examples shown here, an α-helix interacts with specific bases through contacts in the major groove. A, The homeodomain α-helix recognizes a specific 6-bp sequence. B, A protein with three zinc fingers recognizes three consecutive 3-bp sequences. C, The glucocorticoid receptor forms a dimer that recognizes the same 6-bp sequence (a hormone response element) in opposite orientations spaced 3 bp apart. D, A leucine zipper factor dimerizes to recognize a pair of 4-bp sites with opposite orientation spaced 1 bp apart.
180
SECTION IV n Central Dogma: From Gene to Protein
Homeodomain This motif of 60 amino acids was discovered in Drosophila proteins that regulate development and is found in many eukaryotic TFs, including more than 150 human genes. Recognition is provided by an HTH motif composed of two helices, one of which sits in the major groove of the DNA-binding site contacting a recognition sequence of 6 bp (Fig. 10.14A). The HTH structure is not a stable domain on its own, but functions as part of a larger DNA-binding domain, such as the homeodomain. A flexible arm interacting with the minor groove provides the homeodomain with additional binding affinity. Zinc Finger Proteins The zinc finger protein sequence motif (Fig. 10.14B) is found in more than 600 human TFs. Each “finger” consists of 30 residues with conserved pairs of cysteines and histidines that bind a single zinc ion. The tip of the finger sticks into the DNA major groove, where it contacts three bases. Most zinc finger proteins contain multiple fingers, allowing longer sequences to be recognized to increase specificity. A related structure is present in the steroid hormone receptor family, although in this case, four cysteine residues coordinate the zinc ion and the finger is composed of two helices rather than one. Steroid hormone receptors also contain a dimerization domain, allowing recognition of sequences with dyad symmetry (Fig. 10.14C). Artificial zinc fingers can now be designed enabling synthetic proteins to recognize any desired DNA sequence for experimental manipulations.
Activation
A
DNA binding Transcription activation DNA binding
+ +
– –
B
– +
Transcription machinery
Factor 1 Gene 1
Factor 2 Gene 2
Swapped domains Gene 2 FIGURE 10.15 TRANSCRIPTION FACTORS CONSIST OF DISCRETE, FUNCTIONAL MODULES. A, Domain characterization. Although the entire factor is required for activation, the bottom domain is sufficient for DNA binding. B, Domain swapping. The activation domain of one factor (activating gene 1) can be fused to the DNAbinding domain of a heterologous factor (activating gene 2). The resulting chimeric factor will activate only genes containing the recognition site for the DNA-binding domain (gene 2).
Leucine Zipper Proteins Leucine zipper domains are made up of two motifs: a basic region that recognizes a specific DNA sequence and a series of leucines spaced 7 residues apart along an α-helix (leucine zipper) that mediate dimerization. These motifs form a continuous α-helix that can dimerize through formation of a coiled-coil structure involving paired contacts between hydrophobic leucine zipper domains (Fig. 10.14D; also see Fig. 3.10). Dimers of leucine zipper proteins recognize short, inverted, repeat sequences. The zipper family comprises many members, some of which can cross-dimerize and recognize asymmetrical sequences. Another family of factors comprises the helix-loop-helix proteins, which have the same type of basic region but differ in that they have two helical dimerization domains separated by a loop region.
Acidic activation domains are generally unstructured segments of polypeptide consisting of multiple acidic residues dispersed among a few key hydrophobic residues. Such domains activate transcription when experimentally grafted to a wide variety of different DNA-binding domains in a number of different cell types. Other types of activator domains have been characterized as being rich in proline or glutamine. The diverse activation domains use several mechanisms to activate transcription, the most direct being recruitment of the basal transcription machinery. For example, the glutamine-rich activation domain of the SP1 factor (see the next section) interacts with TFIID to recruit GTFs to the promoter. In many cases transcription activators and repressors do not contact the GTFs or Pol II directly but rather act via interactions with coregulator complexes as discussed in a following section.
Transcription Factors as Modular Proteins Binding of a TF to DNA per se does not activate transcription. A separate domain provides this function by interacting directly or indirectly with the basal transcription machinery to elevate the rate of transcription (Fig. 10.15). The best-characterized activation domain is an acidic region derived from the herpesvirus VP16 protein.
Transcription Factor Binding to Eukaryotic Promoter Proximal and Enhancer Elements Experiments analyzing eukaryotic promoter function in living cells revealed numerous DNA regulatory sequence elements in addition to the basal promoter elements recognized by the GTFs. The regulatory sequence elements fall into two broad categories based on their
CHAPTER 10 n Gene Expression
A A
Expression
CAT
Up to –10 kb E
181
Up to +10 kb TATA
5'
Exon 1
E
E
Exon 2
3'
Pre-mRNA Cohesin
B. Enhanceosome Enhancer DNA CAT reporter gene Coactivator
+1
+1
Promoter
?????
?????
+1
????
XXXXX XXXXX XXXXX XXXXX XXXXX XXXXX
Expression of CAT? + + – + – + – +
XXXXX CCAAT
GCGCG
TATA
Identified sequence elements
CAT gene XXX =
mutations
B. Promoter proximal elements of the human metallothionein gene GRE -300
-250
AP2 -200
AP2 MRE MRE AP2 AP1 MRE SP1 -150
-100
-50
TATA 0
FIGURE 10.16 RNA POLYMERASE II PROMOTER REGULATORY ELEMENTS. A, In vivo assays are used to identify key regulatory sequences. In the example shown, a promoter is placed in front of a gene encoding chloramphenicol acetyltransferase (CAT), and the resulting plasmid is transfected into cultured cells. This bacterial enzyme is easily assayed in eukaryotic cells because there is no corresponding endogenous activity. Targeted clusters of mutations, strategically placed throughout the promoter region, are tested for their effect on expression of the reporter gene. Mutations that reduce expression define important regulatory elements. B, The region immediately upstream of the metallothionein gene contains binding sites for several transcription factors. Each element is named for the factor that binds there: GRE (glucocorticoid response element), MRE (metal response element), and AP1, AP2, and SP1 (which bind protein factors with the same names as the DNA elements).
distances from the promoter. Promoter proximal elements are located within a few hundred base pairs upstream of the transcription start site. Enhancer sequences can be located from tens to hundreds of kilobases from the start of transcription. One example of a promoter proximal element is the CCAAT box in the promoter of the herpes simplex virus thymidine kinase gene. This site was identified by a technique called linker-scanning, in which clustered mutations are introduced at regular intervals in the promoter (Fig. 10.16A). In the case of the thymidine kinase promoter, the CCAAT and TATAAA sequences are
Enhanceosome complex
TA TA
Exon 1
FIGURE 10.17 ENHANCER ELEMENTS. A, These clusters of factor-binding sites can influence expression when located far from the promoter in either the upstream or downstream position. In addition, they work in either orientation with respect to transcription. B, Model enhancer showing the tight packing of several different DNA-binding proteins. These complexes fold into structures that have been called enhanceosomes.
required for full transcription. Thymidine kinase expression also requires the sequence GGCGCC, which was subsequently shown to serve as the binding site for SP1, a TF involved in expression of a number of so-called housekeeping genes, whose products are involved in constitutive cellular functions. Other promoter proximal elements are involved in regulated expression, for example, in response to cellular stress or exposure to heavy metals. Most promoters are paired with several different promoter proximal elements. This allows for regulation of transcription levels by varying the relative abundance or activity of the various factors. A good example is the human metallothionein gene, whose product protects cells from the toxic effects of metals (Fig. 10.16B). The location of numerous regulatory elements directly upstream of the TATA box suggests that a variety of different mechanisms regulate this gene. Enhancers are clusters of regulatory DNA sequences that resemble promoter proximal elements, but are considerably more complicated and have several distinguishing features. First, an enhancer can increase the rate of initiation from a basal promoter even if it is located up to 100 kb away along the chromosome. Second, the enhancer element will work in either orientation relative to the promoter (Fig. 10.17A). Third, enhancers can function with a heterologous promoter. Figure 10.17B shows an example of an enhancer sequence with a number of TFs bound, forming a complex called an enhanceosome. Many genes are associated with multiple enhancers. Each enhancer usually works in a cell type–specific fashion. An example is a sequence in an intron of the immunoglobulin heavy-chain gene that enhances transcription
182
SECTION IV n Central Dogma: From Gene to Protein
in lymphocytes but not in other cells. This regulation of enhancer function is accomplished by varying the levels of various enhancer-binding TFs in different tissues. In addition, enhancer chromatin structure is characterized by a nucleosome free region that allows TFs to bind, flanked by nucleosomes bearing histone H3K27ac and H4K3me1 modifications. The following sections discuss how enhancers interact with TFs and coregulators to increase transcription from promoters. Coactivators Coactivators are complexes of regulatory proteins that do not bind DNA themselves but are recruited by genespecific TFs. These complexes contain proteins that recruit the GTFs, alter chromatin structure and assist in the early stages of transcription. The most common coactivator is the Mediator, a complex of 26 proteins in human cells that bridges DNA-bound TFs to Pol II (Fig. 10.18A). Different Mediator subunits bind particular TFs and communicate regulatory signals to the initiation complex. One example: an interaction of Mediator with the Pol II CTD helps stabilize the preinitiation complex. Mediator also stimulates the CTD kinase activity of TFIIH thus releasing the CTD from the Mediator (Fig. 10.18A) Another class of coactivators has histone acetyltransferase activity that modifies histones and other proteins (Fig. 10.18B). One example is p300/CBP. This was initially identified as a protein interacting with a TF that
binds cAMP response elements. Many different TFs recruit p300 to chromatin to locations generally assumed to be enhancers. Histone H3K27 is one of the main targets of p300. This same H3 residue is the target of the polycomb repressive complexes (see Chapter 8), suggesting that p300 plays a role in switching between active and repressed chromatin states. A third class of chromatin coactivators regulates access to DNA by moving, ejecting, or altering the composition of nucleosomes. The SWI/SNF complex is an example of a nucleosome remodeler. When recruited to chromatin by TFs, this complex moves nucleosomes that block regulatory or promoter sequences. Coactivators often work together to activate genes. Initial binding of a TF may recruit a chromatin remodeler that exposes a second TF binding site. This second TF may then recruit mediator, thereby recruiting Pol II and the GTFs. Corepressors act in opposition to coactivators by repressing transcription. The most common form of repression involves chromatin modifications that block TF access (Fig. 10.18C). Histone deacetylase complexes like Sir2 and NuRD (Table 10.2) remove acetyl modifications leading to chromatin compaction and repression of transcription. Polycomb is another pair of corepressor complexes that methylates H3K27 leading to inhibition of RNA polymerase elongation. Polycomb repressive complexes play critical roles in early embryonic development.
A
B. Histone acetylation activates transcription
General transcription factors bind mediator
Mediator
Transcription factor
TAFs
Template DNA TPIID
TBP
Initiation
Coactivator
+1
CTD
Pol II
AC
AC
Pol II binds
AC
AC
AC
AC
AC
AC
Unphosphorylated CTD binds mediator
AC AC
Template DNA +1
TPIID
Pol II
C. Histone deacetylation represses transcription
Preinitiation complex
Transcription factor Corepressor
TFIIH phosphorylates the CTD allowing Pol II to escape promoter
Template DNA TPIID
TAFs
P
P P
P P
P
TBP
AC RNA
AC
AC
AC
AC
AC
AC AC
FIGURE 10.18 TRANSCRIPTION ACTIVATION MECHANISMS. A, General transcription factors and mediator form a scaffold for binding RNA polymerase II with unphosphorylated C-terminal domain (CTD) to form a preinitiation complex. Phosphorylation of CTD by TFIIH releases polymerase and starts transcription. B, Histone acetylases in a coactivator loosen chromatin in the vicinity of the promoter, allowing assembly of preinitiation complexes. C, Recruitment of histone deacetylases in a corepressor represses transcription by compacting the chromatin in the vicinity of the promoter.
CHAPTER 10 n Gene Expression
S Y SP T
SP S P SY
C-terminal domain
P S Y SP T
S Y SP T
S P SY
SP S P SY
183
SP
Preinitiation
Termination mRNA PP
P
S Y SP T
P
P
SP S P SY
PolyA tail
SY
P SY SP T S
Mediator SP
P S Y SP T
SP S P SY
Initiation
Capping enzyme Cap
Mediator disssociation
Elongation FIGURE 10.19 PHOSPHORYLATION OF THE C-TERMINAL DOMAIN OF RNA POLYMERASE II REGULATES TRANSCRIPTION. This cycle illustrates how phosphorylation influences each step in mRNA transcription. See the text for details.
Long-Range Regulatory Interactions Most genes are regulated by enhancers located many thousands of bases away from their promoter, so some means of communication between enhancer and promoter is required for gene activation. This is most commonly achieved through direct interaction when the chromatin fiber forms a loop bringing the enhancer and promoter into close contact. Such interactions between enhancers and promoters involve the cohesin complex discovered because it regulates separation of sister chromatids during cell division (see Fig. 8.18). The cohesin complex forms a ring around two DNA strands thus stabilizing the loop (Fig. 10.17). ChIP-seq screening for cohesin and high levels of Mediator has identified several hundred intergenic regions containing multiple enhancers clustered together. These “super enhancers” direct transcription of genes that specify cell fate and when associated with oncogenes lead to tumor pathogenesis. Post Initiation Regulation of Polymerase II Transcription After formation of a transcription preinitiation complex several steps lead to promoter clearance, elongation and termination. The CTD of Pol II not only orchestrates promoter proximal events but is also important for coupling transcription to splicing of the nascent transcript and to 3′-end formation (see Chapter 11). The CTD (Fig. 10.4C) is comprised of tandem repeats of the repeat of seven amino acids (Y1S2P3T4S5P6S7). The three serines
are phosphorylated at different stages of the transcription cycle (Fig. 10.19). TFIIH kinase CDK7 phosphorylates serine 5 (Ser5) in the preinitiation complex. This releases the Mediator from Pol II and creates a binding site for the capping enzyme that modifies the 5′ end of the message (see Fig. 11.2). At most promoters, inhibitory factors pause the early Pol II elongation complex after synthesizing approximately 30 nucleotides. Phosphorylation of CTD Ser2 by Cdk9 releases the paused polymerase and allows elongation to proceed. Signaling pathways regulate this process at many genes. Ser2 phosphorylation also helps recruit the RNA splicing machinery to the nascent transcript. At the 3′ end of the gene, Ser2 phosphorylation recruits the cleavage and polyadenylation machinery leading to formation of the mature mRNA and termination of elongation. Combinatorial Control The complexity of eukaryotic regulatory systems allows for the integration of multiple regulatory signals at individual genes. Such combinatorial control is seen in a limited way in prokaryotes. For example, the E. coli lac genes are regulated by both lactose and glucose. Only when glucose is low and lactose is present do the activator (CAP) and repressor (lac repressor) function to maximize lac expression. Regulation of transcription initiation in eukaryotes is based on similar principles with DNA-binding activators and repressors controlling individual genes. Each
184
SECTION IV n Central Dogma: From Gene to Protein
eukaryotic gene typically has binding sites for multiple factors. Integration of the individual binding events can take place in several ways. First, there is a degree of synergism to the binding of multiple factors. The enhanceosome is an example where binding of proteins that bend the DNA promotes binding of additional proteins. The key characteristic of the enhanceosome complex is that it stimulates transcription more strongly than the sum of the individual TFs. Synergy can also result from multiple interactions between activators bound to DNA at different upstream sites or different enhancers and targets in coactivators such as the mediator or nucleosome remodeling complexes. Many of the same mechanisms also can occur with repressors. Combinatorial control also can result from the interplay between factors that alter chromatin structure. For example, modification of histone tails by a histone acetyltransferase tethered to a DNA-bound TF can loosen chromatin at a particular site and create binding sites for additional factors. Subsequent binding of a nucleosomeremodeling complex can render sequences more accessible to the transcriptional machinery.
Modulation of Transcription Factor Activity Regulation of transcription initiation is fundamentally important in controlling gene expression. In many cases, the availability of factors that bind to specific sites in promoters is the switch that turns a gene on. Various strategies control the binding of specific factors to DNA regulatory elements (Fig. 10.20). One of the most straightforward is de novo synthesis of the specific factor (Fig. 10.20A). This requires an additional level of regulation of transcription and translation of the mRNA that encodes the specific factor. These steps take time, so this regulatory strategy is used more commonly to regulate developmental pathways than situations where rapid responses are required. Several mechanisms are used for rapid regulation of the activity of existing TFs. One mechanism involves the formation of an active factor from two inactive subunits (Fig. 10.20D). This association can be regulated through synthesis or by modification of preexisting subunits, leading to their association. Binding of small-molecule ligands is another means of controlling TF activity (Fig. 10.20B). In this case, the binding of the ligand induces a conformational change that leads to DNA binding and transcription activation. Interaction of TFs with inhibitory subunits is also used to regulate factor activity (Fig. 10.20E). The DNA binding or activation potential is held in check until the appropriate signal leads to dissociation or destruction of the inhibitory factor. Covalent modification—for example, by phosphorylation—is also used to convert inactive TFs to a functional form (Fig. 10.20C). Finally, the ability of TFs to bind DNA may be regulated by restricting their localization to the cytoplasm (Fig. 10.20F). These regulatory schemes are not mutually exclusive, and many regulatory pathways
A. De novo synthesis
D. Heterodimer formation Activation subunit
DNA-binding subunit
B. Ligand binding
E. Dimer dissociation
Ligand
C. Phosphorylation
Inhibitor
F. Subcellular localization Inhibitor
NUCLEUS FIGURE 10.20 REGULATION OF TRANSCRIPTION FACTOR ACTIVITY. Many strategies have evolved to regulate transcription factors in response to specific signals. A, The availability of a factor may be controlled by expressing it, de novo, only when it is needed. B, Factors may be synthesized in an inactive state and depend on a small molecule (ligand) for activity. C, Transcription factors that are synthesized in an inactive state can be activated by postsynthetic modification, such as phosphorylation. D, Some factors require an appropriate partner for activity. E, Constitutively active factors can be held in check by associating with inhibitory subunits. F, Active factors can be sequestered in the cytoplasm by blocking their transport to the nucleus.
(see the examples that follow) employ several different levels of regulation.
Transcription Factors and Signal Transduction One hallmark of eukaryotic gene regulation is the ability of cells to respond to a wide range of external signals. Cells detect the presence of hormones, growth factors, cytokines, cell surface contacts, and many other signals. They transmit this information to the nucleus, where changes in expression of specific genes are executed (see Fig. 27.4 for the three types of signaling pathways to the nucleus). TFs often execute the final step in these signal transduction pathways; the following sections discuss several examples not covered in Chapter 27.
Steroid Hormone Receptors Regulation of gene expression by steroid hormone receptors involves both ligand-binding and inhibitory subunits. This family of nuclear receptors includes TFs with a common sequence organization consisting of a specific DNA-binding domain, a ligand-binding domain
CHAPTER 10 n Gene Expression
A
Steroid
B
C
7-helix receptor
Receptor
185
Adenylcyclase Nuclear hormone receptor Steroid
Activated G-protein
Adaptor proteins cAMP
Hsp90
Kinases
R
C
CYTOPLASM
R
R
C
Active protein kinase A
C R C
Inactive protein kinase A
IκB
NF-κB
IκB complex
IκB degraded by proteasome
NUCLEUS CBP
CREB Polymerase FIGURE 10.21 TRANSCRIPTION FACTORS AS TARGETS OF SIGNAL TRANSDUCTION PATHWAYS. External signals are transmitted by a variety of pathways that eventually impinge on transcription factors. A, Steroid hormones diffuse through the cell membrane and bind to the hormone receptor in the cytoplasm (estrogen) or, more commonly, the nucleus. Hormone binding induces a conformational change that renders the receptor competent to activate transcription. B, Ligands bound to the extracellular surface of seven-helix receptors initiate a pathway that leads to the activation of protein kinase A, which then moves to the nucleus, where it phosphorylates transcription factor CREB (cyclic adenosine monophosphate [cAMP] response element–binding). (C, catalytic subunit of protein kinase A [PKA]; R, regulatory subunit of PKA that is dissociated from C by binding cAMP [R is shown smaller than actual size].) C, In a third strategy, constitutively active transcription factors are kept sequestered in the cytoplasm until a signaling pathway is activated. In this example, the transcription factor nuclear factor κB (NF-κB) is bound to an inhibitor called IκB (inhibitor of nuclear factor κB). Activation of the pathway leads to phosphorylation of IκB, which targets the inhibitory subunit for destruction by the proteasome. The free NF-κB is transported to the nucleus, where it activates the transcription of target genes.
that regulates DNA binding, and one or more transcription activation domains. The ligands that regulate these factors are small, lipid-soluble hormone molecules that diffuse through cell membranes and bind directly to the TF in the cytoplasm (Fig. 10.21A). Steroid hormones, retinoids, thyroid hormone, and vitamin D bind to distinct nuclear receptors, enabling them to recognize sequences in the promoters of a range of target genes. The specific sites of action in promoter DNA, termed hormone response elements, are related to either AGAACA or AGGTCA (Fig. 10.14C). Specificity of the response is generated by the spacing and relative orientation of the binding sites. Nuclear receptors can bind as homodimers, although some form heterodimers. In addition to heterodimerizing with other members of the nuclear receptor family, interactions with other types of TFs can link the steroid response to other pathways that signal through cell surface receptors. Heat shock protein 90 (Hsp90) blocks inactive steroid hormone receptors from interacting with DNA (Fig. 10.21A and see Fig. 17.13). This chaperone keeps the
receptor ligand-binding domain in a conformation ready to bind the ligand but unable to enter the nucleus. Hormone binding to the receptor dissociates Hsp90 and frees the receptor’s DNA-binding domain. The free ligand–bound receptor moves from the cytoplasm to the nucleus, where it binds its DNA target and activates transcription.
Cyclic Adenosine Monophosphate Signaling Changes in gene expression often develop in response to the binding of signal molecules to cell surface receptors. Binding of ligand induces a structural change in the receptor that sets off a chain of events leading to changes in transcription. Protein phosphorylation plays an important role in this process. The adenyl cyclase system controls not only metabolism (see Fig. 27.3) but also gene expression (Fig. 10.21B). Ligand binding to some seven-helix receptors leads to cAMP synthesis, which, in turn, activates protein kinase A (see Fig. 27.3). The promoters of cAMP-regulated genes contain a conserved DNA sequence element, called a cAMP response element,
186
SECTION IV n Central Dogma: From Gene to Protein
that mediates the transcriptional response to cAMP. A TF, termed cAMP response element–binding (CREB) protein, binds this sequence specifically. CREB protein is a leucine zipper TF that binds DNA as a dimer. The DNA-binding domain of CREB protein can be exchanged with other DNA-binding domains without loss of cAMP responsiveness. This indicates that cAMP does not work by altering CREB binding to DNA. Rather cAMP alters the transcription activation function by stimulating protein kinase A to phosphorylate a specific residue (serine 133) in CREB. Phosphorylation changes the conformation of CREB and allows interaction with a protein adaptor that recruits the transcription machinery leading to transcription of target genes.
Nuclear Factor κB Signaling The family of NF-κB TFs controls immune and inflammatory responses, development, cell growth, and apoptosis. The activity of NF-κB is normally tightly controlled and persistently active NF-κB is associated with cancer, arthritis, asthma, and heart disease. In most cells, NF-κB is held in an inactive form in the cytoplasm through interaction with an inhibitor called inhibitor of nuclear factor κB (IκB) (see Figs. 9.19 and 10.21C). When B lymphocytes (see Fig. 28.9) are stimulated to produce antibody, NF-κB binds to an enhancer in the immunoglobulin κ-chain gene and activates transcription. The stimulatory signal leading to NF-κB activity is transmitted through a protein kinase cascade that ultimately phosphorylates I-κB, signaling its destruction by proteolysis. I-κB destruction unmasks the NF-κB nuclear localization signal, leading to its transport to the nucleus, where it activates transcription of immunoglobulin genes. Transcription Factors in Development The previous discussion focused on how external signals can lead to changes in gene expression in the nucleus, which, in turn, changes cellular functions. A critical step in this genetic program is the regulation of one TF by another. Such cascades of TF activity are fundamental to gene regulation in development. Early cell divisions in multicellular organisms create different types of daughter cells that express distinct sets of genes. In this case, two types of information govern the expression of a gene. First, the interaction of the cell with its environment sends signals that are transduced to the nucleus and change the pattern of gene expression. How the nucleus interprets the transduced signals depends on the set of TFs that preexist within it. Thus, in addition to external signals, the history of the cell dictates which genes will respond to incoming signals. The programs of TF interaction during development are complicated, but the underlying principles of these pathways are well conserved. Many developmentally regulated TFs are autoregulated (Fig. 10.22), allowing
A. Cascade External signal DNA mRNAs Proteins
B. Autoregulatory inhibition External signal
C. Combinatorial activation External signal
Second external signal
FIGURE 10.22 GENE REGULATORY CIRCUITS. The complex patterns of gene expression observed in multicellular organisms arise from interactions among thousands of transcription activators and repressors, as illustrated by three examples. A, Transcription factors activate the expression of other factors leading to a cascade of changes in gene expression following the initial external signal. B, Some transcription factors can act as both activators or repressors. In this example, the external signal leads to expression of a transcription factor that goes on to activate the expression of other genes and repress its own expression. C, Multiple transcription factors regulate most genes, so activation requires more than one external signal—two in this example.
TFs to activate their own expression. This positive feedback creates a switch that leads to continued expression after the initial stimulus is gone. Other developmentally regulated TFs are, in turn, regulated by several different factors. This allows combinatorial signals to dictate expression (Fig. 10.22C). For example, some TFs activate certain promoters while repressing others. The basis of this contradictory property is thought to be the ability of TFs to cooperate with each other when bound at the same promoter. This cooperation can be either positive or negative. This allows the expression of a target gene to be regulated both by external signals (eg, proximity of an adjacent cell that expresses a signaling molecule) and by the preexistence of a given factor in the cell. In this way, only cells of a given lineage that are located in a certain area of an embryonic segment express the gene. As new TFs involved in development are discovered, the challenge will be to decipher the complicated combinatorial interactions among them.
SELECTED READINGS Corden JL. RNA polymerase II C-terminal domain: Tethering transcription to transcript and template. Chem Rev. 2013;113:8423-8455. de Laat W, Duboule D. Topology of mammalian developmental enhancers and their regulatory landscapes. Nature. 2013;502:499-506. Dekker J, Mirny L. The 3D genome as moderator of chromosomal communication. Cell. 2016;164:1110-1121. Delest A, Sexton T, Cavalli G. Polycomb: a paradigm for genome organization from one to three dimensions. Curr Opin Cell Biol. 2012;24:405-414. Hnisz D, Abraham BJ, Lee TI, et al. Super-enhancers in the control of cell identity and disease. Cell. 2013;155:934-947. Landick R. The regulatory roles and mechanism of transcriptional pausing. Biochem Soc Trans. 2006;34:1062-1066. Levine M. Transcriptional enhancers in animal development and evolution. Curr Biol. 2010;20:R754-R763. Murakami K, Calero G, Brown CR, et al. Formation and fate of a complete 31-protein RNA polymerase II transcription preinitiation complex. J Biol Chem. 2013a;288:6325-6332. Murakami K, Elmlund H, Kalisman N, et al. Architecture of an RNA polymerase II transcription pre-initiation complex. Science. 2013b; 342:1238724.
CHAPTER 10 n Gene Expression
187
Ong CT, Corces VG. CTCF: an architectural protein bridging genome topology and function. Nat Rev Genet. 2014;15:234-246. Ruthenburg AJ, Li H, Patel DJ, et al. Multivalent engagement of chromatin modifications by linked binding modules. Nat Rev Mol Cell Biol. 2007;8:983-994. Sainsbury S, Bernecky C, Cramer P. Structural basis of transcription initiation by RNA polymerase II. Nat Rev Mol Cell Biol. 2015;16: 129-143. Spitz F, Furlong EE. Transcription factors: from enhancer binding to developmental control. Nat Rev Genet. 2012;13:613-626. Teves SS, Weber CM, Henikoff S. Transcribing through the nucleosome. Trends Biochem Sci. 2014;39:577-586. Vannini A, Cramer P. Conservation between the RNA polymerase I, II, and III transcription initiation machineries. Mol Cell. 2012;45: 439-446. Yan J, Enge M, Whitington T, et al. Transcription factor binding in human cells occurs in dense clusters formed around cohesin anchor sites. Cell. 2013;154:801-813. Zaret KS, Carroll JS. Pioneer transcription factors: establishing competence for gene expression. Genes Dev. 2011;25:22272241. Zhou Q, Li T, Price DH. RNA polymerase II elongation control. Annu Rev Biochem. 2012;81:119-143.
This page intentionally left blank
CHAPTER
11
Eukaryotic RNA Processing* I
n all organisms, the genetic information is encoded in the sequence of the DNA. However, to be used, this information must be copied or transcribed into the related polymer, RNA. Eukaryotes synthesize many different types of RNA, but no RNA is simply transcribed as a finished product. The mature, functional forms of all eukaryotic RNA species are generated by posttranscriptional processing. These processing reactions are the major topic of this chapter. The major RNAs can be assigned to three major classes: (1) The cytoplasmic messenger RNAs (mRNAs) and their nuclear precursors (pre-mRNAs) carry the information that is used to specify the sequence, and therefore ultimately the structure, of all proteins in the cell. (2) Other RNAs do not encode protein but function directly, playing major roles in various metabolic pathways, including protein synthesis. These include the ribosomal RNAs (rRNAs) and transfer RNAs (tRNAs), which are the key components of the protein synthesis machinery; the small nuclear RNAs (snRNAs), which form the core of the pre-mRNA splicing system; and the small nucleolar RNAs (snoRNAs), which are important factors in ribosome biogenesis. These RNAs are generally much longer-lived than mRNAs and therefore often are referred to as stable or noncoding RNAs (ncRNAs). (3) The third and most recently identified class of RNA comprises several structurally related groups of very small (21 to 25 nucleotides) RNA species that play important roles in regulating gene expression. Base pairing between endogenous micro-RNAs (miRNAs) and target mRNAs in the cytoplasm represses their translation into protein. The packaging of DNA into a nontranscribed form termed heterochromatin (see Fig. 8.7) is promoted by a class of nuclear, small centromeric RNAs. The introduction of small double-stranded *This chapter was written by David Tollervey and includes some text and figures from a chapter in the first edition written by Barbara Sollner-Webb, with contributions from Christine Smith.
RNAs into many cell types and organisms results in cleavage of the target mRNA and consequent silencing of gene expression. This phenomenon is described as RNA interference (RNAi), and the RNAs are referred to as small interfering RNAs (siRNAs). In addition, a heterogeneous set of longer ncRNAs (lncRNAs) have been implicated in a variety of nuclear events.
Synthesis of Messenger RNAs Fig. 11.1 shows an overview of mRNA synthesis and degradation.
Messenger RNA Capping and Polyadenylation Two distinguishing features set mRNA apart from other RNAs: a 5′ cap structure and a 3′ poly(A) tail. These elements help protect the mRNA against degradation and act synergistically to promote translation in the cytoplasm. The mRNA cap is an unusual structure. It consists of an inverted 7-methylguanosine residue, which is joined onto the body of the mRNA by a 5′-triphosphate–5′ linkage (Fig. 11.2). Cap addition involves three enzymatic activities: (a) a 5′ RNA triphosphatase cleaves the 5′ triphosphate on the nascent transcript to a diphosphate; (b) RNA guanylyltransferase forms a covalent enzyme–guanosine monophosphate (GMP) complex and then caps the RNA by transferring this to the diphosphate; and (c) RNA (guanine-7) methyltransferase covalently alters the guanosine base by methylation, generating m7G. In addition, the first encoded nucleotides are frequently modified by methylation of the 2′ hydroxyl position on the ribose group, but the functional significance of these internal modifications is currently unclear. During 3′ end processing, the nascent pre-mRNA is cleaved by an endonuclease, and a tail of adenosine residues is added by poly(A) polymerase. Approximately 189
190
SECTION IV n Central Dogma: From Gene to Protein
Cotranscriptional mRNA capping Exon 1 m7G Capping enzymes
Recognition and cleavage of poly(A) site (termination competence)
Exon 2 Poly(A) site
Pol II
Transcription termination
CTD
Spliceosome EJC
m7G
Pre-mRNA surveillance NUCLEUS
or EJC
m7G
7G
Dcp1/2
m
AAAAA
Lsm1-7
Rat1
CYTOPLASM
7
m G
AAAAA
m7G
AAA
Nuclear exosome
EJC
C. ARE-mediated decay
B. mRNA turnover
AAAAA
m7G
AAAAA
7
AAAAA
Cotranslational deadenylation
m G
Caf1/Ccr4
Upf 1/2/3 m7G
m7G
mRNA nuclear export
A. Nonsense-mediated decay m 7G
AAAAA
or
AAAAA
ARE AAAAA ARE-BP
m7G
m7G
ARE Cytoplasmic exosome
m7G
Rapid 3´ degradation
Rapid 5´ and 3´ degradation
m
7
Lsm1-7
m7G
Ski7 Cytoplasmic exosome
Dcp1/2 7G
D. Nonstop decay
m7G
m G Lsm1-7
Xm1
P body
Cytoplasmic exosome
3´ degradation
Displacement of stalled ribosome and rapid 3´ degradation
FIGURE 11.1 SYNTHESIS AND DEGRADATION OF EUKARYOTIC MESSENGER RNAS. Nascent messenger RNA (mRNA) transcripts are transcribed by RNA polymerase II. Formation of the 5′ cap structure and cleavage and polyadenylation of the 3′ end of the mRNA both occur cotranscriptionally and involve factors that are recruited by the C-terminal domain (CTD) of the transcribing polymerase (see Fig. 10.4). The termination of transcription requires both the recognition of the site of polyadenylation and the activity of the 5′-exonuclease Rat1, which degrades the nascent RNA transcripts. Rat1 binds to the polymerase CTD via a linker protein. Pre-mRNA splicing can either be cotranscriptional or occur shortly after transcript release, and recruitment of splicing factors is not strongly dependent on the CTD. In human cells, the spliceosome deposits the exon-junction complex (EJC) around 24 nucleotides upstream of the site of splicing. Several steps in nuclear mRNA maturation are subject to surveillance. In yeast, nuclear pre-mRNAs can be either 3′ degraded by the nuclear exosome complex or decapped and 5′ degraded by the exonuclease Rat1. Nuclear decapping requires the Lsm2–8 complex and is probably performed by the Dcp1/2 decapping complex. Once in the cytoplasm, the mRNA is translated into proteins and undergoes degradation. Several different mRNA degradation pathways have been identified. A, Nonsense-mediated decay (NMD). If the EJCs all lie within or very close to the open reading frame (ORF), they will be displaced by the translating ribosomes. However, if an EJC lies beyond the end of the ORF, it will remain on the translated mRNA. This is taken as evidence that translation has terminated prematurely and triggers the NMD pathway. Recognition of the EJC requires the Upf1/2/3 surveillance complex, which also interacts with the ribosomes as they terminate translation. In yeast, NMD triggers both rapid decapping and 5′ degradation, without prior deadenylation, and 3′ degradation by the exosome. B, General mRNA turnover. During translation, most mRNAs undergo progressive poly(A) tail shortening. Loss of the poly(A) tail leads to rapid degradation. As in the nucleus, cytoplasmic mRNAs can be degraded from either the 5′ or the 3′ end. The 5′ degradation occurs largely in a specialized cytoplasmic region termed the P body in yeast or cytoplasmic foci in human cells. Here, the mRNAs are decapped by the Dcp1/2 heterodimer and then degraded by the cytoplasmic 5′-exonuclease Xrn1. Both activities are strongly stimulated by the cytoplasmic Lsm1–7 complex. Alternatively, deadenylated mRNAs can be 3′ degraded by the cytoplasmic exosome. C, ARE-mediated decay. In this pathway, specific A+U rich elements (AREs) are recognized by ARE-binding proteins (ARE-BPs) in the nucleus. These are transported to the cytoplasm in association with the mRNA and recruit the cytoplasmic exosome to rapidly degrade the RNA. D, Nonstop decay. If the mRNA lacks a translation termination codon, the first translating ribosome will stall and be trapped at the 3′ end of the RNA. The Ski7 protein, which is associated with the cytoplasmic exosome complex, is believed to release the stalled ribosome and target the RNA for 3′ degradation by the exosome. Note that this legend provides more complex information than that given in the text for interested readers.
200 to 250 A residues are added to mRNAs in human cells, while 70 to 90 are added in yeast. Cleavage and polyadenylation are performed by a large complex containing approximately 20 proteins that recognizes sequences in the mRNA, of which the best defined is a highly conserved AAUAAA motif located upstream of the site of polyadenylation (Fig. 11.3).
Links Between Messenger RNA Processing and Transcription The processes of cap addition and 3′ cleavage and polyadenylation are both linked to transcription of the mRNA by RNA polymerase II and occur cotranscriptionally on the nascent RNA (Fig. 11.1). The C-terminal domain
CHAPTER 11 n Eukaryotic RNA Processing
A. Signals for polyadenylation
A. Chemical structure of 5' capped mRNA OH
HO 2'
O
6
8
9N
5
4 3N 1 2
HN
m7G
NH2
B. 5' capping pathway 5' end of primary transcript γ βα
pppNpNp Triphosphatase
Pi
βα
ppNpNp αβ γ
Guanylyl transferase αβα
Gp p p PPi
Gp p p N p N p Guanine7-methyl transferase
SAM SAH
m7G p p p N p N p 2'-O-methyl transferase
SAM SAH
3'
4'
5' CH2
2'-O-methyl transferase
SAH
*
O –O P O O –O P O 5'— 5' linkage O –O P O O H2C 5' O Nuc 1 4'
H H
*
2'
O OCH3 –O P O O H2C Nuc 2 O H H
H H
O OCH3 P O O H2C Nuc 3 O
–O
H H
H H
O OH P O O H2C O
–O
H H –O
O P O O
GU-rich
Cleavage occurs after transcribing this signal
AAUAAA
AAA200
B. Frequency (%) of residues in animals
H H
3'
AAUAAA
Poly(A) added to new 3' end
1'
m7G p p p Nm p N p SAM
RNA polymerase
H H
H O
H3C +N 7
1'
191
Nuc 4
H H OH
m7G p p p Nm p
Nm p Capped mRNA
FIGURE 11.2 MESSENGER RNAS HAVE A DISTINCTIVE 5′ CAP STRUCTURE. A, The 5′ ends of messenger RNAs (mRNAs) are blocked by an inverted guanosine residue that is attached to the body of the mRNA by a 5′–5′ triphosphate linkage. The N7 position of the guanosine is methylated (red). The first encoded nucleotide of the mRNA (Nuc 1) is also methylated on the 2′-hydroxyl of the ribose ring. The second nucleotide (Nuc 2) may also be methylated. B, Capping of mRNAs is a multistep process.
(CTD) of the largest subunit of RNA polymerase II consists of many copies of a seven-amino-acid repeat (YS2PTS5PS), which undergo reversible modification by phosphorylation (see Fig. 10.4). A pronounced change in the CTD phosphorylation pattern coincides with the release of the polymerase from initiation mode into processive elongation mode. Immediately following transcription initiation, the repeats are largely phosphorylated on the serine residue at position 5. This modification is lost, while serine 2 phosphorylation increases, as the polymerase moves along the transcript. Capping of the 5′ end of the mRNA occurs by the time
97 98 100 100 100 97
A A U A A A
FIGURE 11.3 SIGNALS FOR PRE–MESSENGER RNA POLYADENYLATION. A, Poly(A) tails are added to pre–messenger RNAs (mRNAs) following transcription. After pol II transcribes the proteincoding region of the mRNA, it encounters two sequence elements: AAUAAA and a GU-rich element. These act as signals for the assembly of a large 3′ processing complex that cleaves the nascent pre-mRNA, releasing it from the transcription complex, and adds a tail of up to 200 adenosine residues. B, The poly(A) signal is highly conserved in vertebrates.
the transcript is approximately 25 to 30 nucleotides long, and the capping enzyme is recruited by the serine 5 phosphorylated CTD. This and other interactions with the polymerase result in strong allosteric activation of capping activity. In contrast, the cleavage and polyadenylation factors involved in 3′ end processing are recruited by interaction with the CTD phosphorylated at serine 2. The major termination pathway for RNA polymerase II on mRNAs is dependent on 3′ processing. Termination requires recognition of the poly(A) site by the cleavage and polyadenylation factors. These are carried along with the transcribing polymerase, and their offloading might make the polymerase competent for termination. Cleavage of the nascent transcript also allows the entry of a 5′ exonuclease—an enzyme that can degrade RNA from the 5′ end in a 3′ direction. This enzyme, which is called Rat1 in yeast and Xrn2 in humans, then chases after the transcribing polymerase, degrading the newly transcribed RNA strand as it goes. When the exonuclease catches the polymerase, it stimulates termination of transcription. This is referred to as the Torpedo model for transcription termination. Regulated 3′ End Formation on Histone Messenger RNAs A different 3′ end processing system operates for mRNAs encoding the major, replication-dependent histone proteins. These are highly expressed only during DNA replication, when they must package the newly synthesized DNA. A sequence in the 3′ untranslated region (3′
192
SECTION IV n Central Dogma: From Gene to Protein
UTR) of these mRNAs is recognized by base pairing to a small RNA: the U7 snRNA. In addition, a specific stemloop structure is recognized by a stem-loop binding protein. Endonuclease cleavage generates the mature 3′ end of the mRNA, which is not polyadenylated but is protected from degradation by the stem-loop binding protein. The efficiency of histone mRNA synthesis is increased during DNA replication at least in part by increased abundance of stem-loop binding protein. Other minor histone variants that are synthesized throughout the cell cycle are polyadenylated like other mRNAs.
Pre–Messenger RNA Splicing Important experiments in the 1950s and 1960s established that genes are collinear with their protein products. It therefore came as a considerable surprise when, in the late 1970s, it emerged that genes in animals and plants frequently had numerous strikingly large inserts whose sequence was not included in the mature mRNA or the protein product. It turns out that most human pre-mRNAs undergo splicing reactions, in which specific regions are cut out and the flanking RNA is covalently rejoined. The regions that will form the mRNA are termed exons, and the bits that are cut out (and are normally degraded) are called introns. In unicellular eukaryotes, introns are generally a few hundred nucleotides in length or shorter. In metazoans, however, they are often several kilobases in length, and pre-mRNAs can contain many introns. It is therefore remarkable that all the sites can be precisely identified and spliced. Signals for Splicing The signals in the pre-mRNA that identify the introns and exons are recognized by a combination of proteins and a group of small RNAs called the snRNAs. The snRNAs function in complexes with proteins in small nuclear ribonucleoprotein (snRNP) particles. Splicing occurs in a large complex termed the spliceosome, within which the pre-mRNA assembles together with five snRNAs (U1, U2, U4, U5, and U6) and approximately 100 different proteins. Particularly important protein-splicing factors are members of a large group of SR-proteins—so named because they contain domains rich in serinearginine dipeptides. Three conserved sequences within introns play key roles in their accurate recognition by the splicing machinery (Fig. 11.4). These lie immediately adjacent to the 5′ splice site and the 3′ splice site and surround an internal region that will form the intron branch point during the splicing reaction. The U1 and U6 snRNAs have sequences that are complementary to the 5′ splice site, whereas U2 is complementary to the branch point region. Although the spliceosome will finally bring together the sequences at each end of the intron, it is thought
that the splicing machinery initially recognizes the exons in a reaction termed exon definition. This makes sense because mRNA exons are generally quite small—up to a few hundred nucleotides in length—whereas the introns can be many kilobases long. No sequences in the exons are strictly required for splicing, but there are important stimulatory elements termed exonic splicing enhancers (ESEs), which generally bind members of the SR protein family. The ESEs have two major functions: to stimulate the use of the flanking 5′ and 3′ splice sites, promoting exon definition, and to prevent the exon in which they are located from being included in an intron. This latter function is particularly important in ensuring that all introns are spliced out without the splicing machinery skipping from the 5′ end of one intron to the 3′ end of a downstream intron. Pre–Messenger RNA Splicing Reaction The splicing reaction proceeds in two steps (Fig. 11.4). In the first, the 5′–3′ phosphate linkage that joins the 5′ exon to the first nucleotide of the intron—at the 5′ splice site—is attacked and broken. This reaction leaves the 5′ end of the intron attached to a downstream adenosine residue via an unusual 5′–2′ phosphate linkage. Because this adenosine remains attached to the flanking nucleotides by conventional 5′ and 3′ phosphodiester bonds, this creates a circular molecule with a tail that includes the 3′ exon. This structure is termed the intron lariat, and the adenosine to which the 5′ end of the intron is attached is termed the branch point, because it has a branched structure. In the second step of splicing, the free 3′ hydroxyl on the 5′ exon is used to attack and break the linkage between the last nucleotide of the intron and the 3′ exon—at the 3′ splice site. This leaves the 5′ and 3′ exons joined by a conventional 5′–3′ linkage and releases the intron as a lariat. This is linearized by the debranching enzyme and is probably rapidly degraded from both ends by exonucleases. The initial steps in splicing are the recognition of the 5′ splice site by the U1 snRNA and the binding of U2 snRNA to the branch point region, assisted by SR proteins (Fig. 11.5). Base pairing between U2 and the premRNA leaves a single adenosine bulged out of a helix and available for interaction with the 5′ splice site. The U4 and U6 snRNAs then join the spliceosome as a basepaired duplex, within a large complex that also contains the U5 snRNA. The U4 and U6 base pairing is opened, and the liberated U6 sequences displace U1 at the 5′ splice site. They also bind to U2—bringing the 5′ splice site and branch point into close proximity. At this point, the first enzymatic step of splicing occurs. This reaction is believed to be directly catalyzed by the intricate structure of the snRNA/pre-mRNA interactions rather than by the protein components of the spliceosome. The 5′
CHAPTER 11 n Eukaryotic RNA Processing
A. Signals for splicing
Exon 1
GU
Intron
5' splice site Exon
65 75 100 100 60 70 85 65
A G G U A A G U
80
AG
Exon 2
80 90 75 100 95
Py N Py Py Pu A* Py U A C U A A* C
Exon
3' splice site
Branch point A
Intron
A G G U
A
A G G
Exon
65 100 100 50
(Py)≥10 N C A G G Mammals Yeast
Exon 5' G3'
A G G
193
A
Exon 1
Intron
A
AG
GU
3'
GU
GU
O
B. Splicing mechanism
Exon 2
A
Lariat
AG
A
Exon 2
AG
Lariat
+
3' 5' 4'
Exon 1
2' 5'
Exon 2
Debranch Exon 1 Degrade (exonuclease)
FIGURE 11.4 SIGNALS AND MECHANISM OF PRE–MESSENGER RNA SPLICING. The precursors to most messenger RNAs (mRNAs) in humans and other eukaryotes contain regions (introns) that will not form part of the mature mRNA and do not encode protein products. During pre-mRNA splicing, the introns are removed and flanking regions (exons) are ligated. A, Introns contain three conserved sequence elements that are recognized during splicing. These lie at the 5′ and 3′ splice sites and surrounding the branch point adenosine within the intron. Numbers indicate the degree of conservation at each position in mammalian pre-mRNAs. The branch point sequence is much more highly conserved between different pre-mRNAs in yeast. The region between the branch point and the 3′ splice site frequently contains a run of pyrimidine residues, which is referred to as the polypyrimidine tract. B, Pre-mRNA splicing involves two catalytic steps. An attack by the branch point adenosine on the 5′ splice site releases the 5′ exon and intron as a circularized molecule (referred to as the intron lariat) joined to the 3′ exon. In the second step, the 3′ end of the 5′ exon attacks the 3′ splice site releasing the joined exons and the free intron lariat. The lariat is subsequently linearized (debranched) and degraded.
splice site is attacked and broken by the ribose 2′ hydroxyl group of the adenosine residue that is bulged out of the U2–intron duplex. The U5 snRNA and its associated proteins are responsible for holding onto the now free 5′ exon and correctly aligning it with the 3′ exon for the second catalytic step of splicing. Both catalytic steps in splicing are technically termed transesterification reactions, because nucleotides are linked by phosphodiester bonds, and the new bond is made at the same time as the old bond is broken. For this reason, the splicing reactions do not, in principle, require any input of energy. However, the assembly and subsequent disassembly of the spliceosome require numerous adenosine triphosphatases (ATPases). Most of these belong to a family of proteins that are generally termed RNA helicases. These are believed to use the energy of adenosine triphosphate (ATP) hydrolysis to catalyze structural rearrangements within the assembling and disassembly spliceosome.
AT-AC Introns The large majority of human mRNA splice sites have a GU dinucleotide at the 5′ splice site and AG at the 3′ splice site (Fig. 11.4). However, a minor group of introns contain different consensus splicing signals and are termed AT-AC (pronounced “attack”) introns because of the identities of the nucleotides located at the 5′ and 3′ splice sites. The splicing of the AT-AC introns involves a distinct set of snRNAs—U11, U12, U4ATAC, and U6ATAC— which replace U1, U2, U4, and U6, respectively. Only U5 is common to both spliceosomes. However, the underlying splicing mechanism is believed to be the same for both classes of intron. Alternative Splicing A surprising finding from the human genomic sequencing project was the relatively low number of predicted protein-coding genes, currently estimated at fewer than
194
SECTION IV n Central Dogma: From Gene to Protein 5' splice site
Exon 1
Intron
Branch point A
3' splice site
Exon 2
Pre-mRNA
U1 base pairs with 5' splice site U2 base pairs with branch site U1
U2 A
U5 brings exon 1 and exon 2 into close proximity U4 releases U6 which then base pairs to U2 U6 U4 Intron U6 U2 Exon 2
U1 Exon 1
G
U1 released and U6 binds the 5' splice site Exons clipped and ligated Lariat released
Exon 1
Exon 2
A
Lariat
C. Smith and G. Johnson. after P. Sharp
U5
FIGURE 11.5 SMALL NUCLEAR RNAS PLAY KEY ROLES IN PRE–MESSENGER RNA SPLICING. Although shown here as RNAs, the small nuclear RNA (snRNA)s function in large RNA-protein complexes termed snRNPs. Despite this fact, the major steps in both intron recognition and catalysis are believed to be performed by the snRNAs. The 5′ splice site and intron branch point are recognized by base pairing to the U1 and U2 snRNAs, respectively. The U5 snRNA enters the spliceosome in a complex with U4 and U6, which are tightly base-paired. U5 contacts both the 5′ and 3′ exons. U4 releases U6, which base-pairs to U2 and then displaces U1 in binding to the 5′ splice site. Within this very complex RNA structure, the 2′ hydroxyl group on the branch point adenosine, which is bulged out of the duplex between U2 and the pre-mRNA, attacks the phosphate group at the junction between the 5′ exon and the intron. In a transesterification reaction, the phosphate backbone is broken at the 5′ splice site. The 5′ exon is released with a 3′ OH group, and the 5′ phosphate of the intron is transferred onto the 2′ position of the ribose on the branch point adenosine, creating the intron lariat structure. U5 retains the 5′ exon and aligns it for a second transesterification reaction, during which the 3′ hydroxyl on the 5′ exon attacks the 3′ splice site, joining the exons and releasing the intron lariat.
20,000. This result has caused increased interest in the phenomenon of alternative splicing, which allows the production of more than one mRNA, and therefore more than one protein product, from a single gene. Several general forms of alternative splicing are commonly found. Exons can be excluded from the mRNAs, or
introns can be included. Some genes have arrays of multiple alternative exons, only one of which is included in each mRNA. In addition, the use of alternative splice sites can generate longer or shorter forms of individual exons (Fig. 11.6). Current estimates for the proportion of human genes that are subject to alternative splicing range from 30% to 75%. In some cases, this could potentially give rise to a very large number of different protein isoforms. Alternatively spliced proteins can have antagonistic functions, such as transcription activation versus transcription repression. For the vast majority of human genes, no information is available on the relative activities of different spliced isoforms. Compounding the difficulty in understanding is the fact that many genes show tissue-specific splicing. Thus, a gene could be transcribed in, say, both the liver and brain but generate products with substantially different functions in each tissue. In addition to generating protein diversity, alternative splicing can generate mRNAs with premature translation termination codons—“nonsense” codons. These are subject to rapid degradation by the nonsensemediated decay (NMD) surveillance pathway (see later). Switching splicing into a pathway that generates an NMD target is therefore a means of downregulating gene expression. It is likely that alterations in the activities of many different factors can lead to the preferential use of alternative splice sites. In at least some cases, changes in the abundance of a general splicing factor generates tissuespecific patterns of splicing. Localization of Pre–Messenger RNA Splicing The location of the splicing reaction within the nucleus was long a contentious topic. The snRNAs can be detected dispersed in the nucleoplasm but concentrate in small structures referred to as nuclear speckles or interchromatin granules, as well as in discrete larger structures known as Cajal bodies (see Fig. 9.2). It is now widely accepted that most splicing is performed by the dispersed snRNA population and can occur either cotranscriptionally or immediately following transcript release. Consistent with this, there is evidence that the recruitment of some splicing factors is promoted by association with the CTD of the transcribing polymerase. The speckles are likely to represent sites at which splicing factors are stockpiled ready for use. The Cajal bodies, in contrast, represent sites of maturation in which the snRNAs undergo site-specific nucleotide modification and perhaps assembly with specific proteins.
Modification of Messenger RNAs Site-specific nucleotide modifications take place at many sites on mRNAs, notably including formation of pseudouracil, 5-methylcytosine, N1-methyladenosine,
CHAPTER 11 n Eukaryotic RNA Processing
Alternative splicing 5'
5'
5' 5'
Exon 1
a b
c
Exon 1 2 3
1
Exon 2 Proteins produced
4
Exon 3 and exon 3'
5'
5'
or
c
c
f
Intron 2
3'
4 3
3'
4 or
e
1 d
Intron 1
e
d
b
a
195
f
c
3'
1 f
3'
Intron 3
c
**
4 or
3'
1
4
Exon 4 Proteins produced
3'
3'
3'
FIGURE 11.6 ALTERNATIVE SPLICING CAN GENERATE MULTIPLE DIFFERENT PROTEINS FROM A SINGLE GENE. Here are some of the possible mRNA and protein products of a gene whose pre-mRNA is subject to alternative splicing. Left, Examples show the effects of skipping one or more internal exons, which produces a set of related proteins with different combinations of “modules.” Right, Examples show the effects of alternative splice sites. In the case shown, the use of alternative 3′ splice sites redefines the 5′ end of the downstream exon. This can lead to the inclusion of additional amino acids in the protein product. Use of an alternative splice site can also cause the exon to be read in a different reading frame (green asterisk), changing the amino acid sequence. If the alternative reading frame contains a translation stop codon (red asterisk), a truncated protein will be produced, and the mRNA will generally be targeted for rapid degradation by the nonsense-mediated decay (NMD) pathway (Fig. 11.1).
and N6-methyladenosine (m6A). Proteins that interact with m6A have been characterized as WRITERS (methyltransferases that create the modification), READERS (proteins that specifically bind mRNAs with m6A modification and alter the processing, stability, or translation) and ERASERS (proteins that remove the modification by oxidative demethylation). Nuclear binding of m6A READERS can alter pre-mRNA alternative splicing and export. However, most m6A modifications in human mRNAs are close to the 3′ end and cytoplasmic binding of READERS can promote rapid mRNA turnover and translation repression. Methylation and demethylation of m6A are important in human development, particularly during spermatogenesis.
Editing of Messenger RNAs The term RNA editing in humans refers to covalent modifications that are made to individual nucleotides, which alter the base-pairing potential. Because the process of translation involves base pairing between mRNA and tRNAs, editing of the mRNA can have the effect of changing the amino acid that is incorporated and therefore the function of the protein. Like alternative splicing, this increases the diversity of protein products that can be synthesized from the genome. Slightly confusingly, the term editing is also used for quite different mechanisms that insert and delete nucleotides from RNAs in some single-celled eukaryotes. The best-characterized example is in the mitochondria of trypanosomes, which are protozoans that cause major human diseases, including African trypanosomiasis, Chagas disease, and leishmaniasis. Uracil residues are added and, less frequently, deleted from the mitochondrial mRNAs at many sites. These changes are specified by a large number of small guide RNAs. This form of editing is not known to occur in higher eukaryotes.
NH2 N N O (R)C'1 Cytidine
NH2 N N
N C'1(R) N Adenosine
H2O NH3 Apobec -1
O NH N O (R)C'1 Uridine
H2O NH3
O
ADAR
HN
N N C'1(R) N Inosine
FIGURE 11.7 RNA EDITING CHANGES NUCLEOTIDE BASE PAIRING. The coding potential of an mRNA can be altered by deamination. In C-to-U editing, the amino group at position 4 of the cytosine base is replaced with a carbonyl group, creating uracil. In A-to-I editing, replacement of the amino group at position 2 of adenosine creates inosine, which base-pairs with C residues rather than with U. ADAR, adenosine deaminase acting on RNA.
C-to-U Editing Deamination of cytosine to uracil is performed by an editing complex, sometimes referred to as the editosome, which includes the deaminase Apobec-1 (Fig. 11.7). Only a small number of nuclear-encoded targets have been identified, and in these, editing generates translation termination codons, producing shorter forms of the encoded proteins. The best-characterized example of C-to-U RNA editing involves the mRNA encoding intestinal apolipoprotein B (ApoB), where CAA-to-UAA editing in the loop of a specific stem-loop structure generates a stop codon. The truncated protein, ApoB48, has an important role in lipoprotein metabolism. In other cases editing may generate mRNAs that are targets for NMD (see later), leading to downregulation of protein expression.
196
SECTION IV n Central Dogma: From Gene to Protein
developmental decisions in oocytes and embryos. In addition, regulated cytoplasmic polyadenylation at synapses controls local translation in neuronal cells. This involves a family of distinct cytoplasmic polymerases. Their association with substrates and activity are both regulated by specific RNA-binding proteins.
A-to-I Editing The enzyme ADAR (adenosine deaminase acting on RNA) can convert adenine residues to inosine by deamination of the base (Fig. 11.7). Inosine acts like guanosine and base-pairs with cytosine rather than uracil, potentially altering the protein encoded by the mRNA. Most of the transcripts edited by ADAR encode receptors of the central nervous system, and RNA editing is required to create the full receptor repertoire. The amino acid substitutions that result from editing of the mRNAs can greatly alter the properties of ion channels, and aberrant editing occurs in various disorders ranging from epilepsy to malignant brain gliomas. ADAR binds as a dimer to imperfect double-stranded RNA duplexes, which are formed between the target site and sequences in a flanking intron. Editing is generally not 100% efficient, so heterogeneous populations of proteins are generated. In addition to specific editing of individual nucleotides, ADARs can hyperedit long double-stranded RNAs (dsRNAs). In mammals, dsRNAs elicit a strong antiviral response from the innate immune system and hyperediting is important to avoid inappropriate recognition of endogenous dsRNAs.
Messenger RNA Degradation and Surveillance Exosome Complex The RNA exosome is a multiprotein complex with exonuclease and endonuclease activities. The complex has a barrel-like structure. Substrates are threaded through the lumen of the barrel to reach the active site of the major 3′ exonuclease (DIS3/Rrp44) (Fig. 11.8). In addition, DIS3/Rrp44 harbors an endonuclease activity that is not accessed through the central channel. The nuclear exosome complex is associated with an additional 3′ to 5′ exonuclease (Rrp6 in yeast, PM-Scl100 in humans). In both the nucleus and cytoplasm, the activity of the exosome is dependent on cofactors. Chief among these are two related RNA helicases (proteins that can open RNA and RNA-protein structures using energy derived from ATP hydrolysis); Mtr4 in the nucleus and Ski2 in the cytoplasm. In the nucleus, Mtr4 is a component of the TRAMP (Trf4/5-Air1/2-Mtr4 polyadenylation) and NEXT (nuclear exosome targeting) complexes together with RNA binding proteins, while Ski2 forms the SKI complex, which is required for all known functions of the exosome in mRNA degradation. The nuclear exosome and Mtr4 participate in RNA maturation, notably in the processing of the 5.8S rRNA. However, the major functions of the nuclear exosome are probably in the surveillance and degradation of many different types of defective nuclear RNAs and RNA–protein complexes, and the clearance of numerous classes of
Cytoplasmic Polyadenylation The early steps of embryogenesis in metazoans occur before transcription of the genome commences. All mRNAs that are present in early embryos were therefore inherited from the mother. These “maternal messages” are frequently translationally inactive, at least in part because they lack a poly(A) tail. They can be activated for translation by polyadenylation in the cytoplasm. Cytoplasmic polyadenylation events are critical for many
5'
A
B. Nuclear exosome
RNA
S1/KH cap Central channel 3'
90°
TOP VIEW
PH A
NUCLEUS
Ski3
SKI complex
Ski2 Ski8
Cap
Rrp6
RNA degradation Rrp44
TRAMP complex Mtr4
Air1/2
Ski7
5'
Trf4/5
Rrp6 6-member ring
C. Cytoplasmic exosome RNA
Rrp44
A
A
3'
Cap PH CYTOPLASM
Rrp44
FIGURE 11.8 THE EXOSOME COMPLEX AND COFACTORS. The exosome has a barrel structure with the major active site in the 3′ exonuclease Rrp44/Dis3 located at the base of the central lumen of the complex. An additional 3′ exonuclease Rrp6 is located close to the entrance of the central channel, while an endonuclease active site on Rrp44 is located on the exterior of the structure. The exosome barrel is composed of an RNA-binding cap structure and a core containing six proteins that show sequence similarities to Escherichia coli RNase PH, but which have, surprisingly, all lost their catalytic activities in eukaryotes. Substrate RNAs are inserted into the complex by related nuclear and cytoplasmic localized RNA helicases: Mtr4 and the TRAMP (Trf4/5-Air1/2-Mtr4 polyadenylation) complex in the nucleus and Ski2 in the SKI complex in the cytoplasm.
ncRNAs. The cytoplasmic exosome functions, together with the SKI complex, in several different mRNA turnover pathways. Degradation of Messenger RNA Most analyses of the regulation of gene expression have concentrated on changes in the levels of mRNA transcription. However, the rate at which mRNAs are degraded is also important, influencing both the total amount of protein synthesized and the timing of protein synthesis following a transcription event. mRNAs are frequently described as having half-lives, but this is generally quite misleading. Degradation is not stochastic, and it is probably better to think of mRNA lifetimes. There are enormous variations in the lifetimes of different human mRNAs—from a very few minutes to many days—that have a large impact on protein expression levels. Different pathways of mRNA degradation can be classified as (a) the default pathway (ie, when we do not yet know of any specific activator or repressor of degradation), (b) regulated degradation pathways that respond to developmental or other signals, and (c) surveillance pathways that identify and rapidly degrade aberrant mRNAs or pre-mRNAs. A theme emerging from studies of all mRNA decay pathways is that RNA-binding proteins, which associate with the newly transcribed precursor in the nucleus, can be retained when the mRNA is exported to the cytoplasm. These proteins maintain a record of the nuclear history of the RNA that can be “read” by the cytoplasmic degradation machinery, and this plays a key role in determining the cytoplasmic fate of the mRNA. A key step in the timing of degradation of most mRNA is the slow, stepwise removal of the poly(A) tail by enzymes called deadenylases. The intact poly(A) tail is bound by multiple copies of the poly(A)-binding protein (PABP), at a stoichiometry of around one molecule per 10 to 20 A residues. Surprisingly, PABP antagonizes 5′ cap removal, probably via interactions with the translation initiation factor eIF4G, which, in turn, stabilizes the cap-binding protein eIF4E. These interactions effectively circularize the mRNA and strongly stimulate translation initiation (see Fig. 12.8). When the tail becomes too short for the last PABP molecule to bind, these interactions are lost. The cap can then be removed by a decapping complex, which cleaves the triphosphate linkage to the body of the mRNA, releasing m7GDP. Cap removal allows rapid 5′ to 3′ degradation of the mRNA by the 5′ exonuclease Xrn1. In addition, loss of the PABP/poly(A) complex allows 3′-degradation of the mRNA by the cytoplasmic exosome. A+U Rich Element–Mediated Degradation The degradation of many mRNA species in human cells is triggered by the presence of sequence motifs referred
CHAPTER 11 n Eukaryotic RNA Processing
197
to as A+U rich elements (AREs) (Fig. 11.1C). These are generally located in the 3′ UTR of the mRNA, where bound proteins will not be displaced by the translating ribosomes. This pathway plays an important regulatory role in gene expression, as it targets for rapid turnover mRNAs that encode proteins such as cytokines, growth factors, oncogenes, and cell-cycle regulators, for which limited and transient expression is important. Computational analyses indicate that up to 8% of human mRNAs carry AREs, and there is evidence that alterations in the activity of this pathway are associated with both developmental decisions and cancer. ARE-binding proteins associate with the nuclear pre-mRNAs and are exported to the cytoplasm, where they can either activate or repress ARE-mediated decay. Some ARE-binding proteins that activate degradation function by directly recruiting the exosome complex to degrade the mRNA from the 3′ end.
Surveillance of Messenger RNAs Nonsense-Mediated Decay The surveillance of mRNA integrity is important because defective molecules can encode truncated proteins, which are frequently toxic to the cell. The presence of a premature translation termination signal (or nonsense codon) strongly destabilizes mRNA via the NMD pathway (Fig. 11.1A). In human cells, termination codons are identified as being located in a premature position by reference to the sites of pre-mRNA splicing. Normal termination codons are within, or very close to, the 3′ exon, so no former splice sites lie far downstream. If any former splice site is located more than approximately 50 nucleotides downstream of the site of translation termination, the mRNA is targeted for degradation. The sites of former splicing events can be identified in the spliced mRNA product, because the spliceosome deposits a specific protein complex on the mRNA during the splicing reaction (Fig. 11.1). This exon-junction complex (EJC) binds to the 5′ exon sequence approximately 24 nucleotides upstream of the splice site. Several of the EJC components remain associated with the mRNA following its export to the cytoplasm. In normal mRNAs, the EJCs will all be displaced by the first translating ribosome, so if one (or more) remains on the mRNA, then translation has terminated too soon and NMD is activated. The identification of premature termination codons in yeast and Drosophila does not rely on cues provided by splice sites but probably involves recognition of other nuclear RNA-binding proteins that are retained on the cytoplasmic mRNAs. In all organisms tested, NMD also requires a surveillance complex, which bridges interactions between the terminating ribosome and the “place markers” on the mRNAs. In yeast and probably in humans, recognition of an mRNA as prematurely terminated activates both 5′ and
198
SECTION IV n Central Dogma: From Gene to Protein
3′ degradation. The mRNA can be decapped and 5′-degraded by Xrn1 without prior deadenylation or can be rapidly deadenylated and 3′-degraded by the exosome. In contrast, the degradation of mRNAs targeted by the NMD pathway in Drosophila is initiated by an endonucleolytic cleavage. Nonstop Decay Some mRNAs lack any translation termination codon because they have been inappropriately polyadenylated, inaccurately spliced, or partially 3′-degraded. Translating ribosomes efficiently stall at the ends of such nonstop mRNAs. This inhibits the repeated synthesis of truncated proteins (Fig. 11.1D). The cytoplasmic form of the exosome complex is associated with Ski7p, which is homologous to the guanosine triphosphatases that function in translation. The interaction of Ski7p with the stalled ribosome is believed to both release the ribosome and target the mRNA for rapid degradation. Nuclear RNA Degradation Analyses of RNA degradation have focused largely on cytoplasmic mRNA turnover, but most RNA synthesized in a eukaryotic cell is actually degraded within the nucleus. Pre-mRNAs are predominantly composed of intronic sequences, and almost all stable RNAs are synthesized as larger precursors that undergo nuclear maturation. In contrast to the role of poly(A) tails in stabilizing mRNAs in the cytoplasm, there is evidence that short, oligo(A) tails can act as destabilizing features during RNA degradation in the nucleus. The TRAMP complex cofactors include nuclear poly(A) polymerases and activate the exosome complex, probably by providing a single-stranded “landing pad,” during surveillance and degradation of many defective nuclear RNAs, including pre-mRNAs, pre-tRNAs and pre-rRNAs. In bacteria such as Escherichia coli, poly(A) tails are added to RNAs to make them better substrates for degradation. This has led to the proposal that the original function of polyadenylation was in RNA degradation, and this role is maintained in the eukaryotic nucleus. Following the appearance of the nuclear envelope in early eukaryotes, poly(A) tails took on a distinctly different function in promoting mRNA stability and translation in the cytoplasm.
Synthesis of Stable RNAs Transfer RNA Synthesis All tRNAs are excised from the interior of larger precursors (pre-tRNAs) (Fig. 11.9). Some pre-tRNAs are polycistronic, with two or more tRNAs excised from the same precursor. In yeast, at least, the genes that encode tRNAs cluster around the surface of the nucleolus, and pretRNA processing appears to occur largely within the nucleolus.
RNase P recognition site
A
5'
3'
3'
5'
RNase P 5'
B
5'
5'
CCA 3'
3'
1. Endonuclease 2. CCA adding enzyme
3'
C
5'
3'
5'
3'
1. Endonuclease 2. Multiactivity ligase 3. Phosphotransferase 3' splice site Anti-codon 5' splice site
3' 5'
FIGURE 11.9 MATURE TRANSFER RNAS ARE GENERATED BY PROCESSING. A, Transcription by RNA polymerase III generates a pre–transfer RNA (tRNA) that is 5′ and 3′ extended and may also contain an intron. Cleavage by RNase P generates the mature 5′ end. B, The 3′ end is cleaved by an unidentified nuclease, and the sequence CCA is added by a specific RNA polymerase. This sequence forms a single stranded 3′ end on all tRNAs. C, If an intron is present, it is removed in a splicing reaction that is distinct from pre-mRNA splicing and does not involve small RNA cofactors. The anticodon (green) is generally located 1 nucleotide away from the splice site.
The 5′ end of the mature tRNA is generated by cleavage by the ribozyme endonuclease RNase P, which recognizes structural elements that are common to all tRNAs. The 3′ ends of all mature tRNAs have the sequence Cp-Cp-AOH, to which the aminoacyl group is covalently attached. However, this CCA sequence is not encoded by the tRNA gene in eukaryotes, although it is encoded by tRNA genes in many bacteria. Instead, the pre-tRNA is initially trimmed, and the CCA sequence is then added by a specific RNA polymerase that belongs to the same family as the poly(A) polymerases that add tails to mRNAs. Many pre-tRNAs contain a single, short intron, which is removed by splicing. The enzymology of tRNA splicing is quite different from that of pre-mRNA splicing. The pre-tRNA is cleaved at the 5′ and 3′ splice sites by a tetrameric protein complex containing two endonucleases and two targeting factors. The cleavages leave products with 5′ hydroxyl residues and 2′ to 3′ cyclic phosphate. A separate tRNA ligase then recognizes these termini and rejoins the exons. In addition, tRNAs are subject to a bewildering array of covalent nucleotide modifications. Almost 100 different modified nucleotides have been identified in tRNAs,
ranging from simple methylation to the addition of very elaborate molecules. All are added without breaking the phosphate backbone of the RNA. The structures of all mature tRNAs are very similar, since each must fit exactly into the A, P, and E sites of the translating ribosome (see Fig. 12.7). It is likely that the modifications help the tRNAs fold into precisely the correct shape. They also aid accurate recognition of different tRNAs by the aminoacyl-tRNA synthases, which are responsible for charging each species of tRNA with the correct amino acid.
Ribosome Synthesis The synthesis of ribosomes is a major activity of any actively growing cell. Three of the four rRNAs—the 18S, 5.8S, and 25S/28S rRNAs—are cotranscribed by RNA polymerase I as a polycistronic transcript. This pre-rRNA is the only RNA synthesized by RNA polymerase I and is transcribed from tandemly repeated arrays of the ribosomal DNA (rDNA). In humans, approximately 300 to 400 rDNA repeats are present in five clusters (on chromosomes 13, 14, 15, 21, and 22). These sites often are referred to as nucleolar organizer regions, reflecting the fact that nucleoli assemble at these locations in newly formed interphase nuclei. The pre-rRNAs are very actively transcribed and can be visualized as “Christmas trees” in electron micrographs taken following spreading of the chromatin using low-salt conditions and detergent (Fig. 11.10A). The 5S rRNA is independently transcribed by RNA polymerase III. In most eukaryotes, the 5S rRNA genes are present in separate repeat arrays. Nucleolus Most steps in ribosome synthesis take place within a specialized nuclear substructure, the nucleolus (see Fig. 9.3). In micrographs, the nucleolus appears to be a very large and stable structure, but kinetic experiments indicate that it is in fact highly dynamic, with most nucleolar proteins rapidly exchanging with nucleoplasmic pools. A current view of the nucleolus is that its assembly is the consequence of many relatively weak and transient interactions between the nucleolar proteins. The result is a self-assembly process that greatly increases the local concentration of ribosome synthesis factors. This is envisaged to promote efficient preribosome assembly and maturation while allowing the rapid and dynamic changes in preribosome composition involved in this pathway. Similar mechanisms may generate other subnuclear structures such as Cajal bodies. The key steps in ribosome synthesis are (a) transcription of the pre-rRNA, (b) covalent modification of the mature rRNA regions of the pre-rRNA, (c) processing of the pre-rRNA to the mature rRNAs, and (d) assembly of the rRNAs with the ribosomal proteins (Fig. 11.10D). During ribosome synthesis, the maturing preribosomes move from their site of transcription in
CHAPTER 11 n Eukaryotic RNA Processing
199
the dense fibrillar component of the nucleolus, through the granular component of the nucleolus. They are then released into the nucleoplasm prior to transport through the nuclear pores to the cytoplasm. Here, the final maturation into functional 40S and 60S ribosomal subunits takes place.
Pre–Ribosomal RNA Processing The posttranscriptional steps in ribosome synthesis are extraordinarily complex, involving approximately 200 proteins and approximately 100 snoRNA species, in addition to the four rRNAs and approximately 80 ribosomal proteins. Ribosome synthesis is best understood in budding yeast, but all available evidence indicates that it is highly conserved throughout eukaryotes. A combination of endonuclease cleavages and exonuclease digestion steps generates the mature rRNAs in a complex, multistep processing pathway. Many pre-rRNA processing enzymes have been identified, although others remain to be found (Fig. 11.10E). The remaining species, 5S rRNA, is independently transcribed and undergoes only 3′ trimming. Modification of the Pre–Ribosomal RNA The rRNAs are subject to covalent nucleotide modification at many sites. Modification takes place on the pre-rRNA, either on the nascent transcript or shortly following transcript release from the DNA template. The most common modifications are methylation of the 2′-hydroxyl group on the sugar ring (2′-O-methylation) and conversion of uracil to pseudouridine by base rotation. The sites of these modifications are selected by base pairing with two groups of snoRNAs. The box C/D snoRNAs direct sites of 2′-O-methylation and carry the methyltransferase (called fibrillarin in humans and Nop1 in yeast) (Fig. 11.10B). The box H/ACA snoRNAs select sites of pseudouridine formation and carry the pseudouridine synthase (called dyskerin in humans and Cbf5 in yeast [Fig. 11.10C]). A small number of snoRNAs do not direct RNA modification but are required for pre-rRNA processing. The best characterized is the U3 snoRNA, which binds cotranscriptionally to the 5′–external transcribed spacer (ETS) region of the pre-rRNA. Base pairing between U3 and the pre-rRNA is required for the early processing reactions on the pathway of 18S rRNA synthesis and directs the assembly of a large pre-rRNA processing complex called the small subunit processome. This complex can be visualized as a “terminal knob” in micrographs of spread pre-rRNA transcripts (Fig. 11.10A). A subset of ribosome synthesis factors interacts with both the rDNA and RNA polymerase I. These interactions might promote both efficient pre-rRNA transcription and recognition of the nascent pre-rRNA. This is reminiscent of the association of mRNA processing factors with RNA polymerase II and suggests that maturation of different
200
SECTION IV n Central Dogma: From Gene to Protein
B. Box C/D snoRNAs C. Box H/ACA snoRNAs guide 2'-O-methylation guide pseudouridylation
A DNA
Nascent pre-rRNA molecules
Direction of transcription 2'OMe
D'
ΝΨ
C'
ΝΨ 5'
3'
rRNA
snoRNA
3'
C
D
snoRNA
rDNA
Processing 2'-O-methylation Box C + D snoRNAs Modification ϕ-formation Box H + ACA snoRNAs
Ribosomal proteins
Structural reorganization and transport
Protein synthesis CYTOPLASM Ribosomes
Late maturation
Primary transcript
3'
Cotranscriptional cleavage of 3’ ETS
Rnt1p
35s
Cleavage A0 A1 A2 ? ? ? 20s
Recycling
Cleavage E ?
Diffusion
NUCLEOPLASM
3'
rRNA
5'
Pre-rRNA
5S rRNA
ACA
E. S. cerevisiae pre-rRNA processing
Pol I transcription
Assembly NUCLEOLUS
H
5'
5' 3'
D
5'
2'OMe
Transcription unit Nontranscribed spacer
Transcription unit
Preribosomes
Processing and assembly factors
Processing and assembly factors
Cleavage A3
RNase MRP
18s
27sA3
Exonuclease A3 B1S
Processing B2 27sBS
Xrn1p Rat1p
Nuclear pore complex Preribosomes
27sA2
Exonuclease E C2
7s
?
Rex1p
Cleavage C2 25s
Exonuclease C1 C1
Exosome 5.8s
Rex1p Rex2p
Xrn1p Rat1p
25s
FIGURE 11.10 RIBOSOME SYNTHESIS. A, “Christmas trees” of nascent pre-rRNA transcripts. This electron micrograph shows ribosomal DNA (rDNA) genes in the process of transcription. Note the numerous molecules of RNA polymerase I (Pol I) along the rDNA, each associated with a pre–ribosomal RNA (rRNA) transcript. In the enlarged inset, the terminal balls can be seen on the transcripts. These large pre–rRNAprocessing complexes (small subunit processomes) assemble around the binding site for the U3 small nucleolar RNA (snoRNA) and are required for the early pre-rRNA processing steps. B–C, Roles of the modification guide snoRNAs. The pre-rRNAs undergo extensive covalent modification. Most modification involves methylation of the sugar 2′ hydroxyl group (2′-O-methylation) or pseudouridine (Ψ) formation, at sites that are selected by base pairing with a host of small nucleolar ribonucleoprotein (snoRNP) particles. Human cells contain well over 100 different species of snoRNPs, and each pre-rRNA molecule must transiently associate with every snoRNP in order to mature properly. Sites of 2′-O-methylation are selected by base pairing with the box C/D class of snoRNAs, which carry the methyltransferase Nop1/fibrillarin. Sites of pseudouridine formation are selected by base pairing with the box H/ACA class of snoRNAs, which carry the pseudouridine synthase Cbf5/dyskerin. D, Key steps in eukaryotic ribosome synthesis. Following transcription of the pre-rRNAs, most steps in eukaryotic ribosome synthesis take place within the nucleolus. The preribosomes are then released from association with nucleolar structures and are believed to diffuse to the nuclear pore complex (NPC). Passage through the NPC is preceded by structural rearrangements and the release of processing and assembly factors. Further ribosome synthesis factors are released during late structural rearrangements in the cytoplasm that convert the preribosomal particles to the mature ribosomal subunits. During pre-rRNA transcription and processing, many of the approximately 80 ribosomal proteins assemble onto the mature rRNA regions of the pre-RNA. E, The pre-rRNA processing pathway. The pathway is presented for the budding yeast Saccharomyces cerevisiae, but extensive conservation is expected throughout eukaryotes. The mature rRNAs are generated by sequential endonuclease cleavage, with some of the mature rRNA termini generated by exonuclease digestion. Scissors with question marks indicate that the endonuclease responsible is unknown.
CHAPTER 11 n Eukaryotic RNA Processing
A. mRNA
201
B. snoRNA/mRNA
Pol II promoter
Poly(A) signal
Exon
Intron
Pol II terminator
Exon
snoRNA
Pol II promoter
Exon
AAA
mRNA
+
Intron
+
Intron
Exonuclease degradation
Exonuclease degradation 5'p + snoRNA
Transcription Pol II pre-snRNA
Functions in pre-mRNA splicing Modification by scaRNPs CAJAL Assembly with BODY snRNA-specific proteins
Functions in ribosome synthesis
D. snoRNA processing snoRNA gene Transcription Pol II and assembly with snoRNP proteins
NUCLEOLUS
m32,2,7G
m32,2,7G
CBC m7G
Nuclear import m7G
Cap trimethylation and 3´ trimming NUCLEUS
m32,2,7G
m7G
Modification by snoRNPs
Nucleolar import
pre-snoRNA
Nuclear export
Sm-protein binding
AAA
mRNA
Debranching
snRNA gene
CBC m7G
Splicing m7G p p p
Debranching
C. snRNA processing
Pol II terminator
Exon
Intron
Splicing m7G p p p
Poly(A) signal
Cap trimethylation and 3´ trimming
CYTOPLASM
FIGURE 11.11 DIFFERENT PATTERNS OF STABLE RNA SYNTHESIS BY RNA POLYMERASE II. A, Primary transcripts encoding messenger RNAs (mRNAs) generally contain one or more introns, which are removed and degraded to produce the mature mRNA. B, In human cells, the small nucleolar RNAs (snoRNAs) that are involved in ribosomal RNA (rRNA) modification are generally synthesized by excision from the introns of highly transcribed protein-coding genes. The small nucleolar ribonucleoproteins (snoRNPs) bind to the snoRNA sequence within the pre-mRNA and protect it from degradation. C, The spliceosomal U1, U2, U4, and U5 small nuclear RNAs (snRNAs) are transcribed by RNA polymerase II (Pol II) and, like mRNAs, are capped with 7-methylguanosine and bound by the nuclear cap-binding complex (CBC). The pre-snRNA is exported to the cytoplasm, where it associates with the Sm-protein complex and is 3′ trimmed. The cap is then hypermethylated to 2,2,7-trimethylguanosine, and the RNA-protein complex is reimported into the nucleus. The newly imported snRNPs localize to the Cajal bodies, where the snRNA is covalently modified at sites selected by base pairing to the small Cajal RNAs (scaRNAs), another class of modification guide RNAs. Assembly with specific proteins then generates the mature snRNPs. D, Some snoRNAs, including U3, are individually transcribed by RNA polymerase II. Like the snRNAs, they are initially capped by with 7-methylguanosine and bind CBC. Following association with a set of snoRNAspecific proteins, they undergo cap-trimethylation and 3′ trimming. The snoRNPs then localize to the nucleolus, where they themselves undergo snoRNP-dependent modification and then participate in rRNA processing.
classes of RNA and their assembly with specific proteins might be functionally coupled to transcription.
Small Nuclear RNA Maturation The U1, U2, U4, and U5 snRNAs are encoded by individual genes transcribed by RNA polymerase II (Fig. 11.11C). Like mRNAs, the snRNA precursors undergo cotranscriptional capping with 7-methylguanosine, but they are not polyadenylated. In human cells, the newly synthesized precursors to these snRNAs are then exported to the cytoplasm. Once in the cytoplasm, the snRNAs form complexes with the Sm-proteins. This set
of seven different, but closely related, proteins assembles into a heptameric ring structure. Sm-proteins are named after the human autoimmune serum that was initially used in their identification. On their own, the Sm-proteins show low substrate specificity in RNA binding. However, in human cells, the assembly of the snRNAs with the Sm-proteins is highly specific and is mediated by a large protein complex. This complex includes the SMN protein (survival of motor neurons), which is the target of mutations in the relatively common genetic disease spinal muscular atrophy. While in the cytoplasm the snRNAs are further processed; the 3′ end of the RNA
202
SECTION IV n Central Dogma: From Gene to Protein
is trimmed, and the cap structure undergoes additional methylation to generate 2,2,7-trimethylguanosine. This hypermethylated cap structure is also present on snoRNAs (see later) and might be important to allow resident nuclear RNAs to be distinguished from mRNA precursors. Once the cap is trimethylated and bound by the Sm-proteins, the snRNAs can be reimported into the nucleus, where they initially localize to discrete sub nuclear structures termed Cajal bodies (see Fig. 9.2). Within the Cajal bodies, specific nucleotides in the snRNAs are modified by 2′-O-methylation and pseudouridine formation. The sites of these modifications are selected by base pairing with a group of resident small Cajal body RNAs (scaRNAs), which carry the RNA-modifying enzymes. The scaRNAs closely resemble the snoRNAs except that single scaRNAs can frequently direct both 2′-O-methylation and pseudouridine formation. Maturation of U6 snRNA is quite different from that of the other snRNAs. U6 is transcribed by RNA polymerase III and is not exported to the cytoplasm. Mature U6 retains the 5′ triphosphate and 3′ poly(U) tract that are characteristic of primary transcripts made by RNA polymerase III (see Chapter 10). However, the 5′ triphosphate is methylated on the γ-phosphate (ie, the position furthest from the nucleotide), while the terminal U of the poly(U) tract carries a 2′ to 3′ cyclic phosphate. Both of these modifications may help protect the RNA against degradation. U6 does not bind the Sm-proteins but instead associates with a related heptameric ring structure that is comprised of seven Lsm proteins (“like Sm”). Two distinct but related heptameric Lsm complexes are present in the nucleus and cytoplasm. The nuclear Lsm2–8 complex binds to the U6 snRNA and participates in the decapping of mRNA precursors that are destined for degradation in the nucleus (Fig. 11.1). In contrast, the Lsm1–7 complex participates in mRNA decapping and 5′ degradation in the cytoplasm. Nucleotides within the U6 snRNA are also modified at positions that are selected by guide RNAs, but this modification occurs in the nucleolus rather than the Cajal body.
Small Nucleolar RNA Maturation The snoRNAs are generally transcribed by RNA polymerase II (except in some plants in which polymerase III–transcribed snoRNAs can be found). However, the genes encoding snoRNAs can have a surprising variety of different organizations. In human cells, most snoRNAs are excised from the introns of genes that also encode proteins in their exons (Fig. 11.11B). The introns that encode snoRNAs are released by splicing and then linearized by debranching. The mature snoRNA is then generated by controlled exonuclease digestion. In contrast, most characterized snoRNAs in higher plants and several yeast snoRNAs are processed from polycistronic
precursors that encode multiple snoRNA species. Individual pre-snoRNAs are liberated by cleavage of the precursor by the double-strand–specific endonuclease RNase III (Rnt1 in yeast) and then trimmed at both the 5′ and 3′ ends. SnoRNAs can also be processed from single transcripts, and these have many features in common with snRNA transcripts. Like snRNAs, these individually transcribed snoRNAs carry trimethylguanosine cap structures (Fig. 11.11D). However, unlike snRNAs, which have a cytoplasmic phase, the maturation of snoRNAs and assembly of snoRNPs take place entirely within the nucleus, most steps probably occurring in the nucleolus.
Synthesis and Function of Micro-RNAs The terms siRNAs and miRNAs are used to describe groups of RNAs that are physically similar but have distinct functions and a variety of different names. All are approximately 22 nucleotides in length and associate with a protein complex called the RNA-induced silencing complex (RISC). Under different circumstances, siRNAs can lead to cleavage of target RNAs, repress translation of mRNAs, or inhibit transcription of target genes via formation of heterochromatin. It seems likely that miRNAs play major roles in regulating global patterns of gene expression in human cells. miRNAs are encoded in the genomes of many eukaryotes, including humans (Fig. 11.12). These are frequently transcribed as polycistronic precursors called primiRNAs (primarily miRNAs). Within the pri-miRNA, the precursors to the individual miRNAs (pre-miRNAs) form stem-loop structures. The stems are first cleaved by a nuclear double-strand–specific endonuclease called Drosha, releasing the individual pre-miRNAs. These are then exported to the cytoplasm, where cleavage by a second double-strand–specific endonuclease, Dicer, releases the miRNA in the form of a duplex with characteristic two-nucleotide 3′ overhangs and 5′ phosphate groups. These duplexes are incorporated into the RISC complex, where one of the strands becomes the functional miRNA. If the target mRNA sequence is incompletely complementary to the miRNA, its translation is repressed (Fig. 11.12). This is likely to be the normal function of most endogenous miRNAs. It has recently been estimated that 30% or more of human mRNAs are targets of miRNA regulation. miRNAs show tissue-specific patterns of expression and dynamic changes in expression during differentiation. Individual miRNAs can modulate the expression of many different mRNAs. Changes in miRNA expression levels have been correlated with many human developmental transitions and numerous cancers. The effects of individual miRNAs on the expression levels of target RNAs are generally quite small (less than twofold), but they play important roles by reinforcing or
CHAPTER 11 n Eukaryotic RNA Processing
Polycistronic miRNA genes
Monocistronic miRNA genes etc.
Transcription
dsRNA
203
Dicer
Dicer cleavage generates ~22nt dsRNA fragments with 2nt 3´ overhang
pri-miRNA
TRBP siRNA
Nuclear cleavage Drosha of pri-miRNAs
Small dsRNAs incorporated into RISC complex
pre-miRNA (~70nt) Exportin 5 Nuclear export Ran-GTP
NUCLEUS CYTOPLASM
pre-miRNA (~70nt)
Ago2 RISC complex One strand becomes functional siRNA used to recognize target sequences
Cytoplasmic cleavage of pre-miRNAs Dicer Mature miRNA(~22nt) Degradation of passenger strand RISC complex Target mRNA binding Target mRNA m7G
RISC
Ago2 cleaves target RNA within region base-paired to siRNA
Target mRNA
AAAAAAAAAA
Translation repressed FIGURE 11.12 microRNA MATURATION. The polycistronic micro-RNA (miRNA) precursors (termed primary-miRNAs, or primiRNAs) are cleaved by the double-strand-specific endonuclease Drosha within the nucleus. The individual pre-miRNAs are then exported to the cytoplasm by the export factor Exportin 5 in complex with Ran-GTP (see Fig. 9.18). Once in the cytoplasm, the pre-miRNAs are cleaved by the double-strand–specific endonuclease Dicer. One strand of the resulting duplex is then incorporated into the RNAinduced silencing complex (RISC) and becomes the functional miRNA. Imperfect duplexes are formed between the miRNA and target messenger RNAs (mRNAs); this results in the inhibition of the mRNA translation.
suppressing changes in gene expression programs that underlie cell fate decisions. The synthesis and stability of miRNAs are subject to functionally important regulation. Proteins binding to the pre-miRNAs can enhance of inhibit maturation, imposing tissue-specific expression patterns. In some cases, miRNA or pre-miRNA degradation is strongly stimulated by the activity of terminal uracil transferases (TUTases) that add uracil nucleotides to the 3′ ends of substrate RNAs. This targets them for degradation by a cytoplasmic exonuclease Dis3L2, which shows high specificity for RNAs with 3′ terminal U tracts. This pathway is important, for example, in regulating the abundance of the Let-7 miRNA, which is highly conserved in evolution. Let-7 functions as an oncogene in humans and is frequently overexpressed in cancers. If a target RNA sequence is perfectly complementary to the miRNA, it is cleaved by a component of the RISC
Exonucleases digest target mRNA
FIGURE 11.13 SMALL INTERFERING RNAS FUNCTION IN MESSENGER RNA CLEAVAGE. In contrast to the endogenous micro-RNAs (miRNAs), exogenously added small interfering RNAs (siRNAs) are designed by experimenters to be perfectly complementary to the target RNA, which is then cleaved by the Ago-2 component of the RNA-induced silencing complex (RISC) complex. In many organisms (including the nematode worm Caenorhabditis elegans and insects such as Drosophila), long double-stranded RNAs can be used. These are processed to approximately 22-nucleotide duplexes. In human cells, siRNAs are generally introduced as preformed 22- nucleotide duplexes or as stem-loops with structures that resemble endogenous pre-miRNAs. In either case, the siRNAs associate with Dicer, the double-strand RNA-binding protein TRBP, and Argonaut 2 to form the RISC complex. One strand becomes the functional siRNA, while the “passenger” strand is lost from the complex.
complex, Ago2 (“Slicer”). Target RNA cleavage occurs within the miRNA: mRNA duplex at a fixed distance (between nucleotides 10 and 11) from the 5′ end of the miRNA, which is specifically bound and used to precisely position the duplex relative to the catalytic site. This pathway can be exploited in a technique for the specific inactivation of target mRNAs, termed RNAi (Fig. 11.13). RNAi uses exogenously provided RNAs that are generally fully complementary to the target, typically provided as 22-nucleotide RNAs termed siRNAs. In many organisms (eg, in Drosophila or the nematode Caenorhabditis elegans), RNAi can be performed by introducing long dsRNAs. These are cleaved in vivo by
204
SECTION IV n Central Dogma: From Gene to Protein
Dicer into 22-bp fragments, which are then incorporated into the RISC complex. In mammals, including human cells, long dsRNAs cannot be used for RNAi, as they trigger an antiviral response and cell death. RNAi can, however, be performed in human cells by the introduction of precleaved 22-bp RNA fragments. Alternatively, small hairpin structures can be expressed that resemble endogenous pre-miRNAs. These are processed into functional 22-nucleotide siRNAs in vivo. The small size, ease of use, and potent function of siRNAs have made RNAi a powerful method for many analyses of eukaryotic gene function. In the nucleus, a closely related system is used to establish transcriptional silencing of RNA synthesis (Fig. 11.14). Although important gaps remain in our understanding, it appears that transcription of a region of the chromosomal DNA on both strands, generating a dsRNA, may be sufficient to induce its silencing. The dsRNA is likely to be cleaved by Dicer and/or Drosha to generate 22 nucleotide fragments, in this case termed siRNAs. These associate with a nuclear complex called RITS (RNA-induced transcriptional silencing [see Fig. 11.14]), which is related to the cytoplasmic RISC complex. These siRNAs identify the corresponding gene, possibly by binding to nascent RNA transcripts and, together with the RITS complex components, recruit a protein methyltransferase. This methylates histone H3 on lysine 9, a hallmark of repressive heterochromatin, which in turn recruits other heterochromatin proteins such as HP1 (see Fig. 8.7). The RITS complex includes an RNA-dependent RNA polymerase, and this may be able to generate new siRNAs, allowing the spreading of the heterochromatin into flanking sequences. The tendency of heterochromatin to spread into the flanking euchromatin has long been recognized and gives rise to the phenomenon of position effect variegation (see Fig. 8.7). In some eukaryotes, the methylated histone H3 can also recruit DNA methyltransferases that modify cytosine residues to 5′-methylcytosine. This reinforces heterochromatin formation and makes it heritable by daughter cells. This system may be important for the establishment of heterochromatin domains, such as those surrounding the centromeres in higher eukaryotes. It might also function as a defense system against the amplification of transposable elements. The irony is that it now seems likely that the largescale organization and transcriptional activity of the genome in many eukaryotes will involve RNAs that long eluded detection because they are so small.
Synthesis and Function of Piwi-Interacting RNAs The Piwi-interacting RNAs (piRNAs) form a distinct class of small RNAs, approximately 26 to 31 nucleotides in length. They are named because of their interaction
Bidirectional transcription Cleavage by dicer
DNA repeats (centromeric regions, transposons, etc.)
sRNA
Identification of DNA RITS sites homologous to sRNAs Methylation of lysine 9 on histone H3
Heterochromatin maintenance and spreading
Histone methyltransferase
M
M
Heterochromatin proteins Gene off M
DNA methylation (stable inheritance of repressed state)
DNA methyltransferase M M M M
M
M
FIGURE 11.14 SMALL HETEROCHROMATIC RNAS FUNCTION IN HETEROCHROMATIN FORMATION. The targets of microRNAs (miRNAs) and small interfering RNAs (siRNAs) are cytoplasmic messenger RNAs (mRNAs). However, siRNAs can also function in the nucleus. Small double-stranded RNAs (dsRNAs) in the nucleus can associate with the RNA-induced transcriptional silencing (RITS) complex. The siRNA-RITS complex then identifies the genomic site of transcription, possibly by recognition of the nascent transcripts. This leads to the establishment of heterochromatin at this location, via the recruitment of protein methyltransferases that methylate lysine 9 on histone H3, a hallmark of repressive heterochromatin (see Fig. 8.7). In some organisms, this is followed by methylation of the DNA, which makes the repressed heterochromatic state more stable and heritable.
with PIWI proteins that were first identified in Drosophila and are related to Argonaut. In invertebrates, piRNAs function in a silencing system in germline cells that blocks expression of transposons, thus protecting the DNA against recombination and mutation. In humans, piRNAs are also strongly expressed in the germline and are necessary for spermatogenesis.
Synthesis and Function of Other Noncoding RNAs RNA sequencing in human cells revealed that the majority (70%–90%) of the genome is detectably transcribed,
even though only less than 2% encodes proteins. This phenomenon is termed pervasive transcription, and it generates a bewildering array of ncRNA species. The definition of “ncRNAs” is somewhat vague, but it has come to mean the collection of transcripts that do not encode proteins, and do not fall neatly into one of the other major RNA classes (tRNA, rRNA, snRNA, etc). Human cells synthesis many thousands of different lncRNAs, defined as being longer than 200 nucleotides in length. These generally resemble mRNAs in being transcribed by RNA polymerase II and carrying cap and poly(A) modifications. The number of different lncRNAs increases with organismal complexity, suggesting that they may play important roles in generating tissue diversity; however, relatively few lncRNAs have been functionally characterized in detail. The best understood human lncRNA is Xist, which plays a key role in silencing one copy of the X chromosome in most female mammals, including humans. This is required to balance gene expression compared to males, which have only a single X chromosome. Xist RNA synthesized from a gene present on the silenced X forms protein complexes that appear to coat the compacted chromosome (see Fig. 8.6). Transcription silencing is achieved, at least in part, by recruitment of the polycomb repressive complex, which modifies specific residues in the histones that package the DNA.
Ribozymes Some RNAs have catalytic activity in the absence of proteins. Such RNA enzymes are termed ribozymes, and they play a number of key roles. Group I and Group II Self-Splicing Introns Two classes of introns can catalyze their own excision from precursor RNAs. These ribozymes are referred to as group I and group II self-splicing introns. Both classes of RNA fold into complex structures that catalyze splicing via two-step transesterification pathways (Fig. 11.15). The first group I intron was identified in 1981 as a 413-nucleotide fragment that was able excise itself from the pre-rRNA synthesized in the ciliate Tetrahymena. This was a major surprise, because at that time all known enzymes were proteins. The demonstration that an RNA could function as an enzyme had a major impact on subsequent RNA research. Group I introns are found in the pre-rRNAs of other unicellular eukaryotes, in the mitochondria and chloroplasts of many lower eukaryotes, and in the mitochondria of higher plants. Group II introns have been found in mitochondria of plants and fungi and in chloroplasts. The splicing mechanism of group II introns strikingly resembles nuclear pre-mRNA splicing (Fig. 11.15C–D). This led to the proposal that the nuclear pre-mRNA splicing system derived from ancestral group II introns. During early
CHAPTER 11 n Eukaryotic RNA Processing
205
eukaryotic evolution, the catalytic center of the group II intron might have become fragmented and separated into the present spliceosomal snRNAs. This would have converted a system that could work only on its own transcript into a system that could process other RNAs, greatly increasing the potential range of spliced RNAs. RNase P and RNase MRP Shortly after the identification of the group I intron in Tetrahymena, the RNA component of RNase P was also shown to function as a ribozyme. RNase P is an RNAprotein complex that cleaves pre-tRNAs at the 5′ end of the mature tRNA sequence in all organisms. The bacterial enzyme has one RNA component and one protein, but the RNA can cleave pre-tRNAs in vitro in the absence of the protein. In eukaryotes, RNase P has become more complicated, with one RNA and nine protein components. The eukaryotic RNA has not been shown to be active in the absence of proteins, but it does show structural similarities to the bacterial RNA, and it is assumed to be the catalyst. Eukaryotes also contain a second RNA-protein enzyme, called RNase MRP, which is closely related to RNase P. The RNA components share common structural features, and the complexes share eight common proteins. RNase MRP cleaves the pre-rRNA between the small and large subunit rRNAs (Fig. 11.10E). Notably, in many bacteria, RNase P can cleave the pre-rRNA at a similar position because of the presence of a tRNA within the pre-rRNA transcript. This suggests that RNase MRP arose in an early eukaryote as a specialized form of RNase P, with a specific function in pre-rRNA processing. By analogy to RNase P, cleavage by RNase MRP is predicted to be RNA catalyzed. RNase MRP also functions in mRNA turnover, at least in yeast, initiating the cell-cycle-regulated degradation of a small number of mRNAs. Large Subunit Ribosomal RNA The most important ribozyme is the rRNA component of the large ribosomal subunit, which does not participate in RNA processing but catalyzes peptide bond formation (see Fig. 12.9). During translation elongation, the peptidyltransferase reaction (the reaction by which amino acid residues are attached to each other to form proteins) is catalyzed by the rRNA itself. The peptidyltransferase reaction is energetically favorable, and it is currently thought that the catalytic activity derives primarily from the precise spatial positioning of the A-site and P-site tRNAs by the rRNA. The ribosomal proteins act as chaperones in ribosome assembly and as cofactors to increase the efficiency and accuracy of translation.
RNA-Based Gene Editing Rapid progress has been made in experimental gene editing techniques, driven by the development of
206
SECTION IV n Central Dogma: From Gene to Protein
A. Group II splicing
B. Group II RNA
C. Group II intron IV
III 5'
Exon 1
A
Intron
3'
Exon 2
V A
Lariat
Exon 2
Intron II
Exon 1 A
Lariat
+
VI
A
Exon 1 Exon 2
3'
Exon 2
I
Pre-mRNA 5'
F. Group I splicing 5'
Exon 1
Intron
G
Exon 1
E. Spliceosome RNAs
D. Spliceosome U4
3'
U6
Exon 2
U5 G
Exon 2 Exon 1 G
Intron
+
Exon 1 Exon 2
Exon 2 U1 5'
Pre-mRNA Exon 1
U5
U2
A
3'
Intron
U2 Catalytic center
U6 U5
FIGURE 11.15 COMPARISON OF SELF-SPLICING WITH PRE–MESSENGER RNA SPLICING. Group I and group II introns are catalytic RNAs or ribozymes that can excise themselves from precursor RNAs in the absence of proteins. A, The removal of group I introns is mechanistically distinct from nuclear pre–messenger RNA (mRNA) splicing and commences with the binding of an exogenous guanosine nucleotide (red G) within a pocket created by the intronic RNA structure. This G is used to attack and break the phosphate backbone at the 5′ splice site. Subsequently, the free 3′ end of exon 1 attacks the phosphodiester bond at the 3′ splice site, leading to exon ligation and the release of the linear intron. B, In contrast, the mechanism of splicing group II introns is very similar to pre-mRNA splicing. An adenine residue (A) near the 3′ end of the intron attacks the 5′ splice site, leading to the formation of a lariat intermediate. The subsequent attack of the free 3′ end of exon 1 on the phosphodiester bond at the 3′ splice site leads to exon ligation and the release of the intron lariat (compare to Fig. 11.4). C–D, Parallels can be drawn between structure and mechanism of group II self-splicing introns and pre-mRNA splicing. This suggested the model that group II introns gave rise to the nuclear pre-mRNA splicing system. The small nuclear RNAs (snRNAs) may be derived from fragments of a group II intron that developed the ability to function in trans (ie, on other RNAs) rather than acting only in cis on its own sequence. Specifically, domain VI of the group II introns functions like the U2-branch point duplex in activating the branch-point adenosine by bulging it out of a helix. Domain V acts like the U2-U6 duplex in bringing this adenosine to the 5′ splice site. Domain III resembles the U5 snRNA in base pairing to both the 5′ and 3′ exons at the splice sites.
RNA-based, site-specific DNA cleavage systems. These are derived from immunity systems that are present in many bacteria and most Archaea. DNA sequencing identified “clustered regularly-interspaced short palindromic repeats” (CRISPR) in which the regions between the repeats are generally derived from the genomes of bacteriophages (viruses that infect bacteria) or plasmids. Conserved proteins encoded adjacent to the CRISPR loci assemble into the Cascade complex. Cascade recognizes and cleaves novel, incoming phage DNA and incorporates small fragments into the CRISPR genomic locus. On subsequent phage infection, Cascade proteins use the CRISPR RNA transcript to recognize and degrade
the phage DNA. Numerous variants of the CRISPR/ Cascade system exist, probably reflecting evolutionary pressure from the development of antagonistic phage systems. It was subsequently shown that a single Cascade protein, together with a suitably engineered guide RNA, can perform highly specific, double-stranded DNA cleavage in almost any genome, including humans (Fig. 11.16). This has greatly facilitated genetic manipulation in many systems. At the time of writing, the Cas9 protein from Streptococcus pyogenes is predominately used as the RNA-directed endonuclease, but the field is developing rapidly. Future advances in genetic engineering of cells
CHAPTER 11 n Eukaryotic RNA Processing
A
Cas9
B
TOP
VIEW
UCGGUGCUUCG 3’ Stem G AGCCACGGUGAAA A loops A C AACUUG U A U sgRNA A GUCCGU GAA UC GGAAUAAAAUU CGAUACGACAAAA A A RNA-DNA hybrid 5’ GGCGCAUAAAGAUGAGACGCGUUUUAGAGCUAUGCUGUUUU GA 3’ CCGCGTATTTCTACTCTGCG ACCGCTAATC 5' NTS dsDNA Main cleavage sites TGGCGATTAG 3' NTS 5’ GGCGCATAAAGATGAGACGC Target DNA PAM Non-target DNA strand (protospacer adjacent motif)
207
FRONT
A
RN
VIEW
sg
60° 5' NTS
Target DNA
3' NTS
Cleavage sites
Stem loops
FIGURE 11.16 RNA-GUIDED DNA CLEAVAGE. The Cas9 protein has double-strand DNA (dsDNA) endonuclease activity at sites selected by base pairing with specific guide RNAs. Eukaryotic genome engineering commonly makes use of a complex between the Streptococcus pyogenes Cas9 double-strand nuclease and a single-guide RNA (sgRNA) guide. Cas9 can open the DNA duplex allowing potential base-pairing to the sgRNA. If the 20 nt complementary sequence is identified in the DNA together with a conserved element, called the protospacer adjacent motif (PAM), Cas9 can then cleave both DNA strands.
and organisms will revolutionize cell biology and medicine over coming decades.
discovered, so there is every reason to think that additional classes of RNA remain to be identified.
Conclusions
SELECTED READINGS
Eukaryotic cells have a bewildering array of RNA species that perform many different, key functions in gene expression. The mature forms of all these RNAs are generated by RNA processing reactions, so the RNA processing machinery is of considerable importance. Probably for this reason, RNA-processing enzymes and cofactors are generally highly conserved during eukaryotic evolution. For many RNA species, transcription and maturation are closely coupled and can be thought of as an integrated system. Finally, it is notable that many RNA species and functionally important modifications have only recently been
Cech TR, Steitz JA. The noncoding rna revolution—trashing old rules to forge new ones. Cell. 2014;157:77-94. Ebert MS, Sharp PA. Roles for microRNAs in conferring robustness to biological processes. Cell. 2012;149:515-524. Henras AK, Plisson-Chastang C, O’Donohue M-F, et al. An overview of pre-ribosomal RNA processing in eukaryotes. Wiley Interdiscip Rev RNA. 2015;6:225-242. Kilchert C, Wittmann S, Vasiljeva L. The regulation and functions of the nuclear RNA exosome complex. Nat Rev Mol Cell Biol. 2016;17:227-239. Lee M, Kim B, Kim VN. Emerging roles of RNA modification: m6A and U-tail. Cell. 2014;158:980-987. Papasaikas P, Valcárcel J. The spliceosome: the ultimate RNA chaperone and sculptor. Trends Biochem Sci. 2016;41:33-45.
This page intentionally left blank
12
CHAPTER
Protein Synthesis and Folding* W
hatever their final destination—cytoplasm, membranes, or extracellular space—proteins are synthesized in the cytoplasm of both prokaryotic and eukaryotic cells. The only exceptions are proteins encoded by genes in mitochondria and chloroplasts, which are synthesized in those organelles. The biochemical synthesis of proteins is called translation, as the process translates sequences of nucleotides in a messenger RNA (mRNA) into the sequence of amino acids in a polypeptide chain (Fig. 12.1). Translation of mRNA requires the concerted actions of small transfer RNAs (tRNAs) linked to amino acids, ribosomes (complexes of RNA and protein), and many soluble proteins. Guanosine triphosphate (GTP) binding and hydrolysis regulate several proteins that orchestrate the interactions of these components. Ultimately, RNA molecules in the ribosome catalyze the formation of peptide bonds. Some newly synthesized polypeptides fold spontaneously into their native structure in the cellular environment, but many require assistance from proteins called chaperones. All contemporary organisms share a common translation apparatus, so the mechanism of peptide bond formation must predate the common ancestor approximately 3.5 billion years ago. By the time of the common ancestor, many relatively complicated regulatory features were in place and were inherited across the phylogenetic tree.
Protein Synthetic Machinery Messenger RNA mRNAs have three parts: Nucleotides at the 5′ end provide binding sites for proteins that initiate polypeptide synthesis; nucleotides in the middle specify the sequence of amino acids in the polypeptide; and *This chapter was revised using material from the first edition written by William E. Balch, Ann L. Hubbard, J. David Castle, and Pat Shipman.
Large subunit Exit hole
4. Subunit recycling C 3'
AUG Small subunit 1. Initiation GTPase
N
Stop codon
tRNAs
3. Termination GTPase Elongating polypeptide
N
5'
end of mRNA
2. Elongation Two GTPases
Peptidyl tranferase site N
tRNAs
N
FIGURE 12.1 OVERVIEW OF THE TRANSLATION CYCLE SHOWING SIX RIBOSOMES ON A SINGLE mRNA. 1, Initiation. Initiator tRNAMet, mRNA, and accessory soluble factors assemble on the small subunit, which then joins with a large subunit. Met is the three-letter code for methionine. 2, Elongation. The polypeptide chain is synthesized, in the order specified by the mRNA, in sequential steps by recruitment of new aa-tRNAs that match the coding sequence of the mRNA, formation of peptide bonds, and dissociation of free tRNA. 3, Termination. Release factors recognize the stop codon (yellow) and terminate translation. The ribosome releases the polypeptide for folding in the cytoplasm. 4, Subunit recycling. The ribosomal subunits dissociate and are available for another round of translation. aa-tRNA, Aminoacyl-tRNA; AUG, initiation codon; GTPase, guanosine triphosphatase; tRNA, transfer RNA.
nucleotides at the 3′ end regulate the stability of the mRNA (Fig. 1.1). Within the protein-coding region, successive triplets of three nucleotides, called codons, specify the sequence of amino acids. The genetic code relating nucleotide triplets to amino acids is, with a few minor exceptions, universal. One to six different triplet codons encode each amino acid (Fig. 12.2). An initiation codon (AUG) specifies methionine, which begins 209
210
SECTION IV n Central Dogma: From Gene to Protein
Second Position C A
U
UUU UUC UUA UUG
C
CUU CUC CUA CUG
A
AUU AUC AUA AUG
G
GUU GUC GUA GUG
Phe Leu
Leu
Ile Met
Val
Ser
UAU UAC UAA UAG
CCU CCC CCA CCG
Pro
CAU CAC CAA CAG
ACU ACC ACA ACG
Thr
GCU GCC GCA GCG
Ala
UCU UCC UCA UCG
AAU AAC AAA AAG GAU GAC GAA GAG
G UGU UGC UGA UGG
Tyr Stop
CGU CGC CGA CGG
His Gln
AGU AGC AGA AGG
Asn Lys
GGU GGC GGA GGG
Asp Glu
Cys Stop Trp
Arg
Ser Arg
Gly
U C A G U C A G U C A G
Third Position (3' end)
First Position (5' end)
U
U C A G
= Chain-terminating codon = Initiation codon FIGURE 12.2 THE GENETIC CODE. The locations of the nucleotide in first, second, and third positions define the amino acid specified by the code.
all polypeptide chains, but may subsequently be removed. In addition, any one of three termination codons (UAA, UGA, UAG) stops peptide synthesis. Eukaryotic and bacterial mRNAs differ in three ways. First, eukaryotic mRNAs encode one protein, whereas bacterial mRNAs generally encode more than one protein. Second, most eukaryotic (and eukaryotic viral) mRNAs are capped by an inverted 7-methylguanosine residue joined onto the 5′ end of the mRNA by a 5′-triphosphate5′ linkage (see Fig. 11.2 and Fig. 12.3). This 5′ cap is stable throughout the life of the mRNA. It provides a binding site for proteins and protects the 5′ end against attack by nucleases. Third, most metazoan mRNAs require processing to remove introns (see Fig. 11.4). Most eukaryotic mRNAs have a 3′ tail of 50 to 200 adenine residues added posttranscriptionally to the 3′ end (see Fig. 11.3). This poly(A) tail binds a protein that promotes export from the nucleus and protects the mRNA from degradation in the cytoplasm. The 3′ poly(A) tails are shorter or absent on bacterial mRNAs. Many single-stranded mRNAs have some double-stranded secondary structure (see Fig. 3.19) that must be disrupted during translation to allow reading of each codon.
Transfer RNA tRNAs are adapters that deliver amino acids to the translation machinery by matching mRNA codons with their corresponding amino acids as they are incorporated into a growing polypeptide (Fig. 12.4). One to four different tRNAs are specific for each amino acid, generally
P P P 5'
m7G P P Cap P 5'
Prokaryotic
Eukaryotic
m7G mRNA Eukaryotic mRNA cap with associated proteins
FIGURE 12.3 mRNA CAP STRUCTURES. Prokaryotic mRNAs (messenger RNAs) end with a 5′ triphosphate. The 5′ cap of eukaryotic mRNAs consists of a 7-methylguanosine residue (m7G) linked to the mRNA by three phosphates. The protein eIF-4E binds the cap and protects against degradation by nucleases. (See Protein Data Bank [PDB; www.rcsb.org] file 1EJ1.)
reflecting their abundance in proteins. Specialized tRNAs carrying methionine (formylmethionine in bacteria) initiate protein synthesis. Transfer RNAs consist of ∼76 nucleotides that basepair to form four stems and three intervening loops. These elements of secondary structure fold to form an L-shaped molecule. A “decoding” triplet (the anticodon) is at one end of the L (the anticodon arm), and the amino acid acceptor site is at the other end of the L (the acceptor arm). Enzymes called aminoacyl-tRNA (aa-tRNA) synthetases catalyze a two-step reaction that couples a specific amino acid covalently to its cognate tRNA (Fig. 12.5).
CHAPTER 12 n Protein Synthesis and Folding
T stem 54
64
T loop 56
60 50 15
Variable loop 20
3'
7
69
44 26 Anticodon stem 38 32
A
3' Amino acid acceptor
1
72
4
69
72
12 D stem
15 R
D stem
12 A D loop A G G 20
B
T stem 60 C Y A T loop R G T ΨC 56 50 54 Y
U
R 26
Acceptor stem
64
7 Y
G
44
Anticodon stem
Variable loop
38
32 Y U
Anticodon
Anticodon loop
5'
PO4
5'
4
A C C
211
Anticodon
C
FIGURE 12.4 tRNA STRUCTURE. tRNAs (transfer RNAs) match an amino acid attached at the 3′ end with the mRNA (messenger RNA) triplet coding for that amino acid. A, Ribbon model, space-filling model, and textbook icon showing base pairing of the anticodon to an mRNA codon. B, Backbone model. C, Planar model showing stem loops of a generic tRNA. Single-letter code for the bases: adenine (A), any purine (R), any pyrimidine (Y), cytosine (C), guanine (G), pseudouridine (ψ), thymine (T), and uracil (U). (See PDB file 6TNA.)
aa-AMP +H N 3
R O C C O H O P O– O H2C O HO
Class I aa linkage
H3 N R C C O H O
+ Class I
Class II
HO
HO
Ad
O O Cyt O P O CH2 O–
OH
PPi
+
Class I synthetase
Two conjugations sites
tRNA
aa
HO Ad O O Cyt O P O CH2 O–
Ad
Synthetase + AMP
ATP
tRNA synthetase
Synthetase • aa-AMP
tRNA • synthetase • aa-AMP
aa-tRNA
FIGURE 12.5 CHARGING A tRNA WITH ITS CORRECT AMINO ACID. tRNA synthetases (shown schematically and as a space-filling atomic model in purple) provide a docking platform for a specific amino acid and its cognate tRNA (shown in orange as a schematic model and as a ribbon model bound to a synthetase). The amino acid is first activated by reaction with adenosine triphosphate (ATP). The carboxyl group of the amino acid is coupled to the α-phosphate of adenosine monophosphate (AMP) with the release of pyrophosphate. The synthetase then transfers the amino acid from the aminoacyl-AMP (aa-AMP) to a high-energy ester bond with either the 2′ (illustrated here) or 3′ hydroxyl of the adenine at the 3′ end of the tRNA. aa, Aminoacyl; PPi, inorganic phosphate. (See PDB file 1QTQ.)
212
SECTION IV n Central Dogma: From Gene to Protein A. Prokaryotic
16s
30s
RNA
21 proteins
30s 70s 50s
32 proteins
50s 5s
23s
RNA
RNA
B. Mammalian 18s
40s
RNA
33 proteins
40s 80s 60s
5.8s
60s
RNA
5s
RNA
49 proteins 28s RNA
FIGURE 12.6 MOLECULAR COMPONENTS OF RIBOSOMES. RNA is light gray in the small subunits and dark gray in the large subunits. Proteins are colored. A, Crystal structure of the 70S ribosome from the thermophilic bacterium Thermus thermophilus. B, Cryoelectron microscopic structure of the 80S ribosome from pig pancreas actively synthesizing protein. The three columns show space-filling models (left), inventories of ribosomal RNAs (rRNAs) and proteins (middle), and maps of the secondary structures of prokaryotic 16S rRNA and 18S eukaryotic rRNAs to illustrate their similarities despite divergent sequences. (A, See PDB file 4W2F. B, Cryo-EM density maps are EMData Bank [EMDB; www .emdatabank.org] files 2644, 2646, 2649, and 2650. Also see PDB file 3J7O.)
In the first step, adenosine triphosphate (ATP) and the amino acid react to form a high-energy aminoacyl (aa) adenosine monophosphate (AMP) intermediate with the release of pyrophosphate. The second step transfers the amino acid to the 3′ adenine of tRNA, forming an aa-tRNA. This reaction is called charging, as the high-energy ester bond between the amino acid and the tRNA activates the amino acid, preparing it to form a peptide bond with an amino group in the growing polypeptide chain. Each of the 20 aa-tRNA synthetases couples a particular amino acid to its several corresponding tRNAs. The two classes of aa-tRNA synthases have different evolutionary origins and attach their amino acids to two different hydroxyls of the adenine at the 3′ end of the tRNA (Fig. 12.5). The fidelity of protein synthesis depends on nearperfect coupling of amino acids to the appropriate tRNAs. Synthetases make this selection by interacting with as many as three areas of their cognate tRNAs: anticodon, 3′ acceptor stem, and the surface between these sites (Fig. 12.5). Some synthetases use proofreading steps to remove incorrectly paired amino acids from tRNAs.
Ribosomes Ribosomes are giant macromolecular machines that bring together an mRNA and aa-tRNAs to synthesize a polypeptide. Base pairing between mRNA codons and tRNA anticodons ensures that the sequences of the polypeptides synthesized are those prescribed by the
sequences of codons in the corresponding mRNAs. After many years of effort, crystal and cryoelectron microscopic (cryo-EM) structures are now available for many kinds of ribosomes (Fig. 12.6). Some surprises emerged from these structures. Ribosomes consist of a small subunit and a large subunit that bind together during translation of an mRNA (Fig. 12.7). Each subunit includes one or more ribosomal RNA (rRNA) molecules and many distinct proteins (Fig. 12.6). The sizes of these subunits and rRNAs are traditionally given in units of S (Svedburg), the sedimentation coefficient measured in an ultracentrifuge. Although all ribosomes derive from a common ancestor and have similar mechanisms of action, their structures have diverged. Mammalian ribosomes have larger RNAs and more proteins than prokaryotic and mitochondrial ribosomes. Ribosomal RNAs constitute the structural core of each ribosomal subunit (Fig. 12.7). The 18S rRNA of the small subunit of mammalian ribosomes contains approximately 1900 nucleotides, most of which are folded into base-paired helices. The large subunit of mammalian ribosomes includes three RNAs: a 28S rRNA consisting of approximately 5000 nucleotides, a 5.8S rRNA of 156 nucleotides, and a 5S rRNA of 121 nucleotides. The rRNAs fold into many based-paired helices, as first predicted by comparing the sequences of rRNAs from many different species (Fig. 12.6). These helices and their intervening loops pack to form a compact structure.
CHAPTER 12 n Protein Synthesis and Folding
A
B 40s
213
C
60s
A/P tRNA
60s
Pore
Nascent chain
D
E
CP
F
CP
L7/L12 stalk
E
P
A
* Tunnel
FIGURE 12.7 STRUCTURE OF THE MAMMALIAN RIBOSOME. The structure the pig Thermus thermophilus ribosome. RNA is shown in light gray for the small subunit and dark gray for the large subunit, and proteins are shown in a range of colors. A, Side view of a space-filling model. B, Cutaway side view of the large subunit showing the elongating polypeptide (as a black zigzag) in the exit channel through the core of the large subunit. C–E, Space filling models from different points of view. C, Bottom view showing the pore of the exit channel. D, Crown view showing the active site. E, Crown view showing three transfer RNAs (orange) bound in the active site. F, Crown view with ribbon diagrams of the proteins minus RNA. (Based on PDB file 4W2F.)
Although prokaryotic rRNAs differ in size and sequence from eukaryotic rRNAs, they fold similarly. Many features of rRNAs have been conserved during evolution, including the surfaces where subunits interact, the sites for binding tRNA, mRNA, and protein cofactors, and the nucleotides involved with peptide bond formation. Most ribosomal proteins associate with the surface of the rRNA core, although several extend peptide strands into the core (Fig. 12.7). Ribosomal proteins are generally small (10 to 30 kD) and basic. With one exception, ribosomes have just one copy of each protein. Decoding of mRNAs and polypeptide synthesis take place in the cavity between the subunits. The surfaces of this cavity are generally free of proteins, so (amazingly) rRNAs—not proteins—are largely responsible for mRNA binding, tRNA binding, and catalysis of peptide bond formation. tRNAs move sequentially through three sites shared by the two subunits: the A site (aa-tRNA), the P site (for peptidyl-tRNA), and the E site (for exit). The growing polypeptide chain exits in a tunnel that passes through the RNA core of the large subunit. The synthesis and assembly of a yeast ribosome requires the participation of all three RNA polymerases,
75 small nucleolar RNAs (snoRNAs), and more than 200 protein factors, in addition to the 80 ribosomal proteins and 4 rRNAs present in the mature ribosomes. Precursor RNAs are cleaved and modified to form the rRNAs (see Fig. 11.10). Assembly factors consisting of snoRNAs and numerous proteins then orchestrate the stepwise assembly of rRNAs and ribosomal proteins into the small and large subunits and guide their export from the nucleus into the cytoplasm. Although the genes for many ribosomal proteins are essential for viability, mutations in some can cause remarkably specific defects. For example, humans with just one functional gene for ribosomal protein RPSA are missing their spleen, but are otherwise normal. Mutations in genes for other subunits cause anemia and mutations in genes for certain assembly proteins cause liver disease.
Soluble Protein Factors Many soluble proteins cycle on and off ribosomes during protein synthesis, enhancing the rate and/or the fidelity of the reactions that occur there. The following sections highlight the role(s) of these soluble factors.
214
SECTION IV n Central Dogma: From Gene to Protein
Mechanism of Protein Synthesis Organisms in all three domains of life use homologous components and similar reactions to synthesize proteins, although the details differ as expected after 3 billion years of evolutionary divergence. In all three domains, protein synthesis takes place in four steps: initiation, elongation, termination, and subunit recycling (Fig. 12.1). Conformational changes move ribosomes along the mRNA as the gene sequence is read out. Few errors are made thanks to precise pairing of tRNAs with their amino acids and codons in the mRNA that occur on the ribosome. Guanosine triphosphatase (GTPase) proteins regulate the progress and fidelity each step (see Fig. 4.6 for details on GTPase cycles).
Initiation Phase The goal of initiation is to bring together the initiator tRNA carrying methionine (or N-formylmethionine, fMet, in Bacteria) and the AUG initiator codon of the mRNA in the appropriate site on the ribosome (Fig. 12.8). First, the two RNAs form a ternary complex on a small ribosomal subunit, which then associates with a large subunit to form a 70S ribosome in Bacteria and an 80S ribosome
in eukaryotes. Eukaryotes use more than 10 soluble protein factors (eukaryotic initiation factors [eIFs]) to coordinate the RNA interactions. Fewer protein factors (designated IF) participate in prokaryotes. In eukaryotes, several steps occur in succession: Step 1. Initiator Met-tRNA and the GTPase eIF-2A (with bound GTP) form a preinitiation complex on a small ribosomal subunit. Step 2. Several protein initiation factors assemble on the 5′ cap of the mRNA. The RNA helicase eIF-4A in this complex uses ATP hydrolysis to remove any secondary structure or bound proteins from the 5′ end of the mRNA. These cap recognition factors also interact with poly(A)-binding proteins on the far end of the mRNA, forming a circular complex that can either favor or inhibit initiation of translation. Step 3. The cap recognition complex targets the mRNA to a preinitiation complex. The order of these first three steps is still being investigated. For example, mRNA may bind to the small subunit before the initiation factors and Met-tRNA. Step 4. The small subunit and two initiation factors form a tunnel on the small subunit through which the mRNA is allowed to slide as the initiator tRNA in Stop
Small subunit
mRNA 5'
1. Preinitiation complex forms
40S initiation factors
eIF-2A • GTP • tRNAMet
Start
m7G
2. Initiation factors bind 5’ cap 3. mRNA binds preinitiation complex
AAA(A)n 3'
mRNA initiation factors Poly(A) binding protein
5'
AAA(A)n 3'
Some mRNAs circularize before binding ribosomes
tRNAMet
4. Ribosome scans mRNA for AUG
GTP
AA(A)n 3'
eIF-2A GDP
Pi
6. eIF-2A and initiation factors dissociate
5. Initiator tRNA binds AUG and GTP is hydrolyzed 7. Subunit joining
P A Large subunit
8. Elongation (see Fig. 12.9)
FIGURE 12.8 STEPS IN INITIATION IN EUKARYOTES. 1, Initiation factors (green) assemble with mRNA (messenger RNA), eIF-2A (purple, activated with GTP), and tRNAMet on a small ribosomal subunit to form the preinitiation complex. 2, Other initiation factors (blue) bind the 5′ cap of the mRNA. For some mRNAs, these 5′ cap-binding factors interact with poly(A)-binding proteins at the 3′ end of the mRNA. This circularization promotes initiation of some mRNAs and inhibits initiation of other mRNAs. 3, The preinitiation complex binds an mRNA. 4, The small subunit scans the mRNA for the AUG start codon (green). 5, When the initiator tRNA binds the start codon, eIF-2a hydrolyzes its bound GTP (guanosine triphosphate). 6, Phosphate, GDP (guanosine diphosphate), eIF-2a, and other initiation factors dissociate and recycle for further rounds of initiation. 7, The small subunit binds a large subunit. 8, Elongation begins. m7G, 7-Methylguanosine.
the P-site scans for the initiator AUG codon. This movement depends on ATP hydrolysis, but its role is not clear. Eukaryotic mRNAs tend to begin translation at the first AUG codon encountered, but the local sequence of the mRNA may also contribute to the specificity as it does in Bacteria. Step 5. When Met-tRNA base-pairs with the initiator AUG codon, eIF-2A hydrolyzes its bound GTP. Step 6. eIF-2A and the other initiation factors dissociate from the small subunit for recycling. Step 7. A large ribosomal subunit binds the small subunit complexed with both the mRNA and Met-tRNA. Another GTPase called eIF-5B hydrolyzes its bound GTP before elongation of the polypeptide begins. Initiation is the slowest and most highly regulated step in protein synthesis, frequently involving phosphorylation of initiation factors. For example, cells that are subjected to various stresses use phosphorylation of eIF-2A to inhibit translation. Phosphorylation increases the affinity of eIF-2A for its guanine nucleotide-exchange factor (eIF-2B), which competes with the initiator tRNA. In contrast, phosphorylation of eIF-4F favors translation by enhancing the interaction of this initiation factor with the 5′ cap of mRNAs. This mechanism can influence the selective translation of particular mRNAs, since the 5′ caps of mRNAs vary in affinity for eIF-4F.
Elongation Phase During elongation, the ribosome sequentially selects aa-tRNAs from the cellular pool in the order specified by the sequence of codons in the mRNA it is translating (Fig. 12.9). The ribosome catalyzes formation of a peptide bond between the amino group of the amino acid part of each new aa-tRNA and the carboxyl group at the C-terminus of the growing polypeptide chain and then moves on to the next codon. Codon-directed incorporation of amino acids into the polypeptide chain begins once the two ribosomal subunits are joined with an initiator tRNA and mRNA properly in place (Fig. 12.8). The elongation reactions occur in the cavity between the two ribosomal subunits. mRNA is threaded, codon by codon, along a bent path between the subunits. aatRNAs enter on one side of the cavity and bind successively to three sites between the two ribosomal subunits. Interactions with both subunits allows the tRNA to maintain contact with the ribosome as it moves, step by step, from the A site to the P site to the E site prior to dissociation. When the tRNAs are bound in the A and P sites, their anticodons base-pair with mRNA codons. Peptide bonds form at the other end of the tRNAs, which position the amino acid on the tRNA in the A site adjacent to peptidyl chain on the tRNA in the P site of the large subunit. The growing polypeptide exits through a 10-nm–long tunnel in the large subunit.
CHAPTER 12 n Protein Synthesis and Folding
215
Two GTPases called elongation factors (EF; eEF for eukaryotic elongation factors) bind near the A site and favor movements of the subunits relative to each other that facilitate the movements of the mRNA and tRNAs through the ribosome. Some of the energy from GTP hydrolysis also increases the accuracy, but makes elongation the most expensive phase of translation in terms of energy expenditure. The following paragraphs summarize the current understanding of the four elongation steps: (1) an aa-tRNA binds to the A site on the ribosome; (2) proofreading ensures that it is the correct aa-tRNA; (3) a peptide bond forms; and (4) translocation advances the mRNA by one codon and moves the peptidyl-tRNA from the A site to the P site on the ribosome. New structures and spectroscopic observations of single ribosomes will continue to reveal more details. Step 1. aa-tRNA binding. The first GTPase (called eEF1A in eukaryotes and EF-Tu in Bacteria; see Fig. 25.7) is charged with GTP by a nucleotide-exchange factor (called eEFX in eukaryotes and EF-Ts in Bacteria). This prepares eEF1A to bind an aa-tRNA, which it delivers to an empty A site of a ribosome. Cells contain enough eEF1A-GTP to bind all the aa-tRNAs and protect the labile ester bond anchoring the amino acid. Step 2. Proofreading. A proofreading mechanism retains aa-tRNAs in the A site if they are correctly base paired with the mRNA codon and allows other aa-tRNAs to dissociate. This “kinetic proofreading mechanism” uses two first-order reactions to discriminate between correct and incorrect aa-tRNAs: hydrolysis of GTP bound to eEF1A; and dissociation of guanosine diphosphate (GDP)-eEF1A from the aa-tRNA and the ribosome. If the aa-tRNA anticodon is base paired with the correct mRNA codon, then the ribosome stimulates GTP hydrolysis, phosphate release, a massive conformational change (see Fig. 25.7), and eEF1A dissociation in a few milliseconds. This allows the aminoacyl end of the aa-tRNA to move into the peptidyl transfer site on the large subunit and form a peptide bond. Those aa-rRNAs with weak, imperfect codonanticodon pairs dissociate from the A site before eEF1A can hydrolyze GTP and dissociate from the aminoacyl end of the tRNA. Step 3. Peptidyl transfer. The RNA of the large subunit forms the highly conserved active site that catalyzes the formation of peptide bonds (Fig. 12.9). This reaction eliminates water and transfers the carboxyl group esterified to the peptidyl-tRNA in the P site to the free amino group of the aa-tRNA in the A site. Catalysis of peptide bond formation depends on a combination of precise orientation of the substrates and stabilization of the transition state (just like protein enzymes). The chemistry is similar, but in reverse, to the hydrolysis of peptide bonds by
216
SECTION IV n Central Dogma: From Gene to Protein
tRNA charged (see Fig. 12-5)
mRNA Release factors
+
Termination
30s
Polypeptide
50S
RF
eEF2 release aa • tRNAaa • GTP • eEF1A complex
Elongation: repeat cycle
1
GDP eEF2
eEFX GTP
GTP hydrolysis Translocation
eEF1A • eEFX complex
Proofreading: incorrect tRNAs released due to low affinity
GTP hydrolysis
GDP
eEF2 binding
2
Pi
eEFX (GEF)
4
3
tRNA accepted eEF1A released deacyl-tRNA released
Peptidyl transfer
Peptidyl transfer P O O C H C R3 O NH C H C R2 O NH C H C R1 NH2
O O C H C R4 NH2
A
Peptidyl transferase catalyzes formation of new peptide bond
P OH
GTP
Hybrid states
Termination by puromycin A
O O C H C R4 O NH C H C R3 O NH C H C R2 O NH C H C R1 NH2
RIBOSOME INTERIOR
P
Puromycin A O O CH3 C O H C NH2
O O C H C R4 O NH Puromycin mimics C aa–tRNAtyr or aa–tRNAphe H C R3 O NH C H C R2 O NH C H C R1 NH2
P
A
OH
Polypeptide chain exits H3C N H N O H3C O CH N N 2 HO O NH C H C O NH2 C H C R4 O NH C H C R3
CH3 O
FIGURE 12.9 STEPS IN ELONGATION AND TERMINATION IN EUKARYOTES. Starting in the upper left, elongation factor eEF1A (EF-Tu in Bacteria) forms a ternary complex with GTP and each amino acyl-tRNAaa for (1) delivery to the matching the mRNA codon in the A site of the ribosome. This ternary complex dissociates rapidly if the anticodon–codon match is incorrect. (2) If the anticodon–codon match is correct, the ternary complex remains bound to the A site long enough for eEF1A to hydrolyze its bound GTP and dissociate from the tRNA still bound to the A site. (3) The ribosome catalyzes formation of a new peptide bond (inset). (4) After eEF2 (EF-G in Bacteria) binds the A site, GTP hydrolysis causes a conformational change that facilitates translocation of the tRNAs and mRNA through the ribosome. Release factors (RF, green) recognize the stop codon and terminate the polypeptide chain (blue), allowing the mRNA and ribosomal subunits to dissociate. The guanine nucleotideexchange factor eEFX promotes the exchange of GDP for GTP on eEF1A. The enlargements at the bottom show details of peptidyl transfer and the mechanism whereby the antibiotic puromycin terminates translation prematurely by mimicking the terminus of amino acyl-tRNATyr or tRNAPhe. It is incorporated on the C-terminus of the polypeptide, which then dissociates from the ribosome, because it lacks an activated carboxyl group.
proteolytic enzymes such as chymotrypsin. After for mation of the new peptide bond, the tRNA in the A site has the polypeptide on one end and its anticodon arm still base-paired to its mRNA codon on the small subunit. The antibacterial agent puromycin can disrupt elongation by mimicking a tRNAPhe or
tRNATyr (Fig. 12.10). Puromycin attacks the esterified carboxyl group of a peptidyl-tRNA in the P site, but lacking an appropriate acceptor site for further peptidyl transfer reactions, it terminates elongation. This results in premature release of the polypeptide chain from the ribosome.
CHAPTER 12 n Protein Synthesis and Folding
U -225
TS Slow
I
-245
A/t
Fast -265
0 0
10 4
Qs
20
8 12
Qc
30
N FIGURE 12.10 ENERGY LANDSCAPE IN PROTEIN FOLDING. As a protein matures from the unfolded state (U) through transition states (TS) to the native folded state (N), native-like contacts form, and the free energy of the system decreases. The two paths (folding trajectories) illustrate that fast protein folding (yellow line) is observed when more native-like contacts are made. When proteins become trapped in partially folded intermediate states, folding is slower (pink line) because energy barriers must be overcome. (Modified from Radford SE, Dobson CM. Computer simulations to human disease: emerging themes in protein folding. Cell. 1999;97:291–298.)
Step 4. Translocation. The second GTPase elongation factor (eEF2 in eukaryotes and EF-G in Bacteria) promotes three linked reactions that complete the elongation cycle. These GTPases have domains similar to domains 1 and 2 of EF-Tu (see Fig. 25.7) plus three domains that mimic the size and shape of a tRNA. Domain 1 binds and hydrolyzes GTP. Domains 3 to 5 target GTP-eEF2 to an empty A site on the ribosome. Binding of GTP-eEF2 to an empty A site favors rotation of the small subunit approximately 6 degrees relative to the large subunit. Hydrolysis of the GTP bound to eEF2 and phosphate dissociation promote the reverse rotation of the small subunit and movement of peptidyl-tRNA from the A site to the P site on the small subunit together with sliding of the mRNA three bases forward on the small subunit. This translocation step produces relatively large forces of approximately 13 pN (piconewtons) with the energy coming from peptide bond formation. At the same time, the deacylated tRNA in the P site is moved to the exit (E) site, where it dissociates later from the ribosome. Finally eEF2 with bound GDP dissociates from the A site, allowing another round of elongation. Addition of each new amino acid pushes the growing peptide through a 10-nm–long tunnel in the large subunit lined with RNA (Figs. 12.1, 12.6, and 12.7). The tunnel accommodates an extended polypeptide approximately 40 residues long with the N-terminus in the lead. The
217
distal parts of the tunnel are wide enough to pass an αhelix, but most folding of the polypeptide takes place outside the ribosome. Peptides longer than 40 residues protrude from the large subunit. Cells balance speed and accuracy during translation to achieve an error rate of about 1 in 104 incorrect amino acids. As a result of this compromise, ribosomes add about 20 amino acids per second to a polypeptide at 37°C, so synthesis of a protein of average size (300 amino acids) takes only 15 seconds. Greater precision could be achieved by slowing translation, but slower cellular growth might be an evolutionary disadvantage.
Termination Phase Termination occurs when the ribosome encounters a termination codon (UAA, UAG, or UGA) at the 3′ end of the coding sequence. Assembly of the polypeptide stops because a protein release factor, rather than an aa-tRNA, binds in the A site on the small subunit of the ribosome (Fig. 12.9). These release factors (called eRF1 in eukaryotes and RF1 or RF2 in bacteria) recognize stop codons and induce the ribosome active site to hydrolyze the peptidyl ester between the C-terminal amino acid of the polypeptide chain and the tRNA in the P site. The completed polypeptide chain threads through the ribosome and is released. Then a GTPase uses energy from GTP hydrolysis to promote dissociation of the mRNA and the ribosomal subunits, which are available for recycling to initiate translation of another mRNA. Further Features of Protein Synthesis Most mRNAs support protein synthesis by multiple ribosomes, forming polysomes (Fig. 12.1). Approximately 40 to 50 nucleotides of mRNA are associated with each ribosome. Consequently, once a ribosome has read approximately 60 nucleotides the initiation codon emerges and is available to assemble another ribosometRNA complex and start translation. Ribosomes can pack close together on one mRNA with all of newly synthesized polypeptides emerging around the periphery. This multiple occupancy of mRNAs explains why ribosomes are more abundant than mRNAs and how one mRNA molecule can guide the synthesis of several copies of its protein product simultaneously. This account of protein synthesis may give the impression of a homogeneous population of ribosomes moving steadily on mRNAs, but variation exists at every level. For example, three types of experiments show that ribosomes can pause during translation. Biochemical experiments and observations of single ribosomes showed that certain sequences, such as several consecutive prolines or mRNA secondary structures, can stall translation. Cells have a special elongation factor (called EF-P in bacteria and a/eIF-5A in eukaryotes) that binds stalled ribosomes and promotes peptide bond formation, so the ribosome can move on.
218
SECTION IV n Central Dogma: From Gene to Protein
New experiments using high-throughput DNA sequenc ing have documented pauses and revealed many other features of translation for all the mRNAs in a cell. This method, called “ribosomal profiling” or “ribosome footprints,” takes advantage of the fact that a ribosome protects approximately 30 bases of the associated mRNA from digestion by nucleases. Therefore, one can isolate polysomes, digest with a nuclease, and isolate the protected mRNA sequences. After copying into DNA, millions of fragments are sequenced in parallel (see Fig. 3.16) to show precisely to the nucleotide where ribosomes are located on mRNAs. The number of DNA reads is higher if ribosomes stall at certain positions. This broad view also revealed many surprising events that take place during translation, including unconventional start sites, pauses caused by environmental conditions and the association of many small RNAs with ribosomes. Ribosomes can vary in composition and posttranslational modifications. Single genes encode most mammalian ribosomal proteins, but plants have multiple genes for isoforms that are expressed in different cells. As they mature (see Fig. 11.10), rRNAs are methylated and some uridines are converted to pseudouridine (Fig. 12.4). Ribosomal proteins are modified by acetylation, methylation, phosphorylation, O-linked β-D-N-acetylglucosamine and ubiquitylation. These differences each have the potential to influence protein synthesis, although few examples have been characterized in detail. A large number of proteins associate with ribosomes and may also influence their activities.
Spontaneous Protein Folding Termination is the final step in translation, but just the beginning for a new protein. A polypeptide begins to experience its new environment while still being synthesized. When it is approximately 40 residues long, its N-terminus emerges from the protected tunnel of the large ribosomal subunit into cytoplasm, where it must fold into a three-dimensional structure (see Fig. 3.5) and find its correct cellular destination. The structure of folded proteins and the folding mechanism are both encoded in the amino acid sequence, making folding spontaneous under suitable conditions. For the soluble proteins, these conditions are aqueous solvent at physiological temperature, neutral pH, and moderate ionic strength. Folding of transmembrane proteins in a lipid bilayer is quite different (see Chapter 20). In test tube experiments, small soluble proteins can be denatured with high temperature, extremes of pH, or high concentrations of urea or guanidine. Denatured proteins exist as ensembles of unfolded polymers with little residual secondary structure. When denatured polypeptides of modest length are transferred to physiological conditions, many fold spontaneously into their native three-dimensional structures on a microsecond to millisecond time scale. (Proteins
such as collagen, which require isomerization of prolines, fold much more slowly; see Fig. 29.4.) Starting from many initial denatured states, a polypeptide converges toward a single low-energy native state (Fig. 12.10) driven by energy from numerous noncovalent interactions and the hydrophobic effect (see Fig. 4.5). The number of possible pathways to the native state is so numerous that if they were sampled individually, proteins would never fold. Thus, both theory and experiment indicate that folding involves a subset of the potential pathways, including the formation of an ensemble of loosely folded transition states with elements of secondary structure, certain turns, and hydrophobic contacts found in the core of the native protein. However, the free energy landscape for folding has hills and valleys, so proteins can be trapped in partially folded states. Many proteins fold spontaneously without assistance during biosynthesis in vivo. Folding begins when the N-terminus of the nascent polypeptide emerges from the ribosome. The vectorial nature of this very slow “cotranslational folding” has both advantages and liabilities. An advantage is that folding before the polypeptide is complete limits the routes to the folded state and might account for why many proteins fold more efficiently during biosynthesis than from the denatured state. On the other hand, vectorial folding precludes interactions of N-terminal sequences with C-terminal sequences until they have emerged from the ribosome. Such interactions are common in folded proteins. Folding of larger proteins is more complicated, especially in the crowded cytoplasm where partially folded proteins expose hydrophobic segments that are normally buried in the core of native proteins. These exposed core elements can aggregate irreversibly before folding is complete. Thus, many newly synthesized native proteins need assistance to avoid irreversible denaturation, aggregation, or destruction by proteolysis during folding. Misfolding of mutant proteins contributes to many human diseases. For example, the most common cause of cystic fibrosis is genetic deletion of the codon for a single amino acid in cystic fibrosis transmembrane regulator (CFTR), resulting in failure of the protein to fold properly (see Fig. 17.4). Beyond lacking function, misfolded proteins also poison the assembly of native proteins in blistering skin diseases (see Fig. 35.6), hypertrophic cardiomyopathies (see Table 39.1), and other “dominant negative” conditions. Folding of proteins into nonnative states causes prion and amyloid diseases (Box 12.1).
Chaperone-Assisted Protein Folding Several families of molecular chaperones (Fig. 12.11) facilitate folding of newly synthesized and denatured proteins. These chaperones do not fold polypeptides by directing the formation of secondary or tertiary
CHAPTER 12 n Protein Synthesis and Folding
219
BOX 12.1 Protein Misfolding in Amyloid Diseases Misfolding of diverse proteins and peptides results in spontaneous assembly of insoluble amyloid fibrils. Such pathological misfolding is associated with transmission of HIV, Alzheimer disease, Parkinson disease, transmissible spongiform encephalopathies (such as “mad cow disease”), and polyglutamine expansion diseases (such as Huntington disease, in which genetic mutations encode abnormal stretches of the amino acid glutamine). Accumulation of amyloid fibrils in these diseases is associated with slow degeneration of the brain. Pathological misfolding also results in amyloid deposition in other organs such as the endocrine pancreas in Type II diabetes. Some, but not all, amyloids are intrinsically toxic to cells. Some amyloid precursors are more toxic than the fibrils themselves. The precursor of a given amyloid fibril may be the wild-type protein or a protein modified through mutation, polyglutamine expansion, proteolytic cleavage, or posttranslational modification. In all cases, fibril initiation is unfavorable owing to very slow assembly of the first few molecules, but once formed, fibrils elongate quickly by adding protein subunits. Amyloid fibrils are extremely stable and resistant to proteolysis. Given that many unrelated proteins and peptides form amyloid, it is remarkable that these twisted fibrils all have similar structures: narrow sheets up to 10 µm long consisting of thousands of short β-strands that run across the width of the fibril. The β-strands can be either parallel or antiparallel, depending on the particular protein or peptide. Some amyloid fibrils consist of multiple layers of β-strands. The structures of the various parent proteins have nothing in common with each other or with amyloid cross–β-sheets, so these are examples of polypeptides with two stable folds. To form amyloid, the native protein must either be partially unfolded or cleaved into a fragment with a tendency to aggregate. In the common form of dementia called Alzheimer disease, proteolytic enzymes cleave a peptide (Aβ) from a transmembrane protein called β-amyloid precursor protein whose normal role is to participate in signal transduction. Aβ forms toxic oligomers and amyloid fibrils
structure. Rather, chaperones inhibit aggregation by binding exposed hydrophobic segments of nonnative polypeptides or providing sequestered environments. They release polypeptides in a folding-competent state for attempts at folding. If folding fails, the cycle of binding and release can be repeated. The following sections cover trigger factor (and other chaperones associated with ribosomes), Hsp70, Hsp90, and cylindrical chaperonins. In addition, specialized chaperones assist with the folding of particular proteins such as tubulin and actin. Mutations in several of these chaperones have been associated with human disease. See Fig. 20.6 for chaperones in the endoplasmic reticulum.
Trigger Factor Hydrophobic segments of the nascent polypeptide chain must be protected from aggregation until enough
that accumulate in the brain as neurons degenerate. Similarly, proteolytic fragments of an enzyme normally found in human semen form amyloid fibrils that enhance the transmission of HIV by many orders of magnitude. Therapeutic strategies include small molecules that stabilize native proteins or inhibit amyloid polymerization. “Infectious proteins” called prions cause transmissible spongiform encephalopathies, such as “mad cow” disease. Normally, these proteins do no harm, but once misfolded, the protein can act as a seed to induce other copies of the protein to form insoluble amyloid-like assemblies that are toxic to nerve cells. Such misfolding rarely occurs under normal circumstances, but the misfolded seeds can be acquired by ingesting infected tissues. Other proteins, including the peptide hormone insulin, the actin-binding protein gelsolin, the receptor protein β2microglobin and the blood-clotting protein fibrinogen, form amyloid in certain diseases. An inherited point mutation makes the secreted form of gelsolin susceptible to cleavage by a peptide processing protease in the trans-Golgi network. Fragments from the protein form extracellular amyloid fibrils in several organs. Exposure to copper during renal dialysis promotes β2-microglobin to form amyloid fibrils in joints. Given that amyloid fibrils form spontaneously and are exceptionally stable, it is not surprising that functional amyloids exist in organisms ranging from bacteria to humans. For example, formation of the pigment granules responsible for skin color depends on a proteolytic fragment of a lysosomal membrane protein that forms amyloid fibrils as a scaffold for melanin pigments. Budding yeast has approximately 10 proteins known to either assume their “native” fold or assemble into amyloid fibrils. The native fold of the protein Sup35p serves as a translation termination factor that stops protein synthesis at the stop codon (see Fig. 12.9). Rarely, Sup35p misfolds and assembles into an amyloid fibril. These fibrils sequester all the Sup35p in fibrils, where it is inactive. The faulty translation termination that occurs in its absence has diverse consequences that are inherited like prions from one generation of yeast to the next.
of the chain has emerged from the ribosome to participate in folding. Each growing polypeptide first encounters a chaperone bound next to the exit tunnel on the large ribosomal subunit. The chaperone associated with bacterial ribosomes is called trigger factor (Fig. 12.11). A structurally unrelated protein called nascent polypeptide-associated complex has a similar function in Archaea and eukaryotes. An extended array of hydrophobic patches on trigger factor binds hydrophobic features on the nascent polypeptide chain. These weak, rapidly reversible interactions prevent folding and pro tect the unfolded peptide from aggregation. The signal recognition particle binds on the other side of the exit tunnel, positioned so that its methionine-rich groove (see Fig. 20.5) also interacts with the growing polypeptide. Most bacterial polypeptides fold successfully after being released from trigger factor, while most
220
SECTION IV n Central Dogma: From Gene to Protein
A. Bacteria Trigger factor mRNA DnaK
B. Eukaryotes NAC DnaJ Hsp70
Native protein ~65–80% Native protein ~10–20% 7ATP + GroES
Hsp40
ATP + GrpE or other chaperones
GroEL
Hsp90 system
Prefoldin
ATP + cofactors? Native protein
Native protein ~15–20%
Native protein ~10–15%
ATP + cofactors?
TRiC
Native protein ~10%
FIGURE 12.11 COMPARISON OF CHAPERONE-ASSISTED FOLDING PATHWAYS. A, Bacteria. B, Eukaryotes. The percentages refer to estimates of the fraction of proteins using each pathway. Most proteins fold without the assistance of chaperones. Hsp, heat shock protein; NAC, nascent polypeptide-associated complex. (Modified from Hartl FU, Hayer-Hartl M. Molecular chaperones in the cytosol: From nascent chain to folded protein. Science. 2002;295:1852–1858. Copyright 2002 American Association for the Advancement of Science.)
eukaryotic polypeptides require assistance from additional chaperones.
Hsp70 Chaperones The most widespread chaperones are members of the heat shock protein 70 (Hsp70) family (Fig. 12.12). Their name came from the observation that cells subjected to stresses, such as elevated temperature, increase the synthesis of these proteins to protect against denatured proteins. Hsp70s are present in Archaea, Bacteria (called DnaK), and most compartments of eukaryotes. The family includes Hsp70 in mitochondria and BiP in endoplasmic reticulum (see Fig. 20.6). Budding yeasts have genes for 14 Hsp70s; vertebrates have more. Hsp70s enzymes consist of two domains: an Nterminal domain (folded like actin) binds and hydrolyzes ATP. It is connected by flexible hinge to a C-terminal domain that uses a clamp to bind and release a wide range of nascent segments of unfolded polypeptides with approximately eight hydrophobic residues. ATP hydrolysis and phosphate release close the clamp on the hydrophobic polypeptides, while ATP binding opens the clamp and releases the polypeptide. This cycle of peptide binding and release, protects hydrophobic peptides from aggregation during attempts at folding, delivery to mitochondria and chloroplasts, and import into these organelles (see Figs. 18.4 and 18.6). Hsp70 cooperates with other chaperones. Members of another family of heat shock proteins (Hsp40, called DnaJ in Bacteria) deliver unfolded proteins to bacterial Hsp70 (DnaK) and promote their binding by stimulating
DnaK to hydrolyze ATP. Another co-chaperone called GrpE promotes exchange of adenosine diphosphate (ADP) for ATP, which opens the clamp and releases the bound peptide. Animal Hsp70s have a mechanism of action similar to that of DnaK except that they have intrinsic nucleotide-exchange activity and do not require a nucleotide-exchange protein such as GrpE. Remarkably, Hsp70 can cooperate with an AAA adenosine triphosphatase found in bacteria, plants, and fungi to unfold aggregated proteins. Energy from ATP hydrolysis is used to pull a polypeptide from the aggregate through the central channel of the adenosine triphosphatase (ATPase). The polypeptide has a chance to fold once it emerges from the channel.
Hsp90 Chaperones Hsp90 cooperates with other chaperones to stabilize steroid–hormone receptors such as those for progesterone, glucocorticoids, estrogens, and androgens, before they bind their ligands (Fig. 12.13). The chaperones use cycles of ATP hydrolysis to maintain receptors in an “open” state, ready to bind hydrophobic steroids. Steroid binding completes the folding of the receptors and displaces the Hsp90 complex. Then the receptors move to the nucleus to regulate gene expression (see Fig. 10.21). Hsp90 also interacts with other signaling proteins including protein kinases. Chaperonins The chaperonin family of barrel-shaped particles promotes efficient protein folding (Fig. 12.14). They allow
221
CHAPTER 12 n Protein Synthesis and Folding
A. Hsp70 structure
Binding domain SHR
GrpE
Hsp90
Hsp70 Hsp40 HIP
Intermediate complex
Hsp90 HOP
ATPase domain B. DnaK cycle
IP
DnaJ delivers new polypeptide
Hsp90
Hsp70
HIP
Hsp90 P23 Open state Polypeptide
ATP
HOP Hsp40
P23 GA
Mature complex
IP
Hsp90 IP
Pi Hormone
GrpE
Hsp90 P23 SHR hormonebinding conformation
ATP ADP
GrpE
Closed state FIGURE 12.12 HEAT SHOCK PROTEIN 70 STRUCTURE AND FUNCTION. A, Ribbon diagrams of the atomic structures of DnaK (blue) and GrpE (green). B, The heat shock protein (Hsp) 70 folding cycle with bacterial DnaK as the example. DnaJ (Hsp40) delivers an unfolded peptide to the ATP-bound open state of DnaK and promotes ATP hydrolysis. The ADP-bound closed state of DnaK binds the peptide strongly. GrpE promotes dissociation of ADP. Rebinding of ATP dissociates GrpE and the peptide, which is free to attempt folding. Multiple Hsp70 cycles are usually required to complete protein folding. (For reference, see Zhu X, Zhao X, Burkholder WF, et al. Structural analysis of substrate binding by the molecular chaperone DnaK. Science. 1996;272:1606–1614; and Harrison CJ, Hayer-Hartl M, Hartl F, et al. Crystal structure of the nucleotide exchange factor GrpE bound to the ATPase domain of the molecular chaperone DnaK. Science. 1997;276:431–435.)
nascent and denatured polypeptides to fold or refold while sequestered in a cylindrical cavity protected from the complex environment of the cytoplasm. Although 85% of newly synthesized bacterial proteins fold spontaneously or with the assistance of Hsp70s, the remainder require the more isolated folding environment provided by chaperonins (Fig. 12.11). The mechanism of chaperonins is best understood for Escherichia coli GroEL and its co-chaperonin GroES. They assist with folding of nascent polypeptides, which in bacteria occurs largely after translation is complete. The GroEL/GroES complex consists of a cylinder with a central cavity composed of GroEL and a cap structure made of GroES. GroEL forms two rings of seven identi cal subunits. Mitochondrial (Hsp60/Hsp10), chloroplast
DNA binding FIGURE 12.13 STABILIZATION OF LIGAND-FREE STEROID HORMONE RECEPTORS BY HSP70, HSP90, AND VARIOUS ACCESSORY FACTORS (HOP, HIP, P23, GA, AND IP). Hormone binding releases the chaperones and allows the receptor-steroid complex to move to the nucleus. SHR, steroid hormone receptor. (For reference, see Buchner J. Hsp90 & Co.—a holding for folding. Trends Biochem Sci. 1999;24:136–142.)
(Cpn60/Cpn10), and eukaryotic chaperonins (TriC) are similar in design but more elaborate than GroEL/GroES, containing up to eight different gene products. This complexity represents evolutionary diversification for regulation of chaperonin function. ATP binding and hydrolysis set the tempo for folding cycles. Unfolded polypeptides interact with hydrophobic patches on the inner wall of the GroEL cylinder. Cooperative binding of ATP to each of the subunits in one of the two rings of seven changes their conformation (compare the upper and lower rings in Fig. 12.14B), expanding the internal volume by twofold and favoring binding of a heptameric ring of 10-kD GroES subunits. This closes the top of the cylinder and creates a folding cavity for proteins up to approximately 70 kD. After ATP hydrolysis on the ring surrounding the folding protein and ATP binding to the opposite ring of seven GroEL subunits, the GroES cap releases, and the cage opens. Folded polypeptides escape into the cytoplasm, whereas incompletely folded intermediates can rebind GroEL for another attempt at folding. ACKNOWLEDGMENT We thank Peter Moore for his suggestions on revisions to this chapter.
222
SECTION IV n Central Dogma: From Gene to Protein
Unfolded peptide
Folded peptide
GroES
7ATP + GroES
7ATP + GroES ATP
ATP
GroEL
ADP
ADP
ATP
ATP
7ADP + GroES
7pi
A Space-filling cross section
Ribbon (top view)
142 Å
B
140 Å
FIGURE 12.14 CHAPERONIN-MEDIATED FOLDING BY GroEL AND GroES. A, One folding cycle. B, Crystal structure of GroEL with a GroES cap bound to the upper, adenosine triphosphate (ATP)-bound ring of seven subunits. Unfolded polypeptides bind the rim of an uncapped ring. Cooperative binding of ATP to each of the seven GroEL subunits in one ring changes their conformation, favors GroES binding, and doubles the volume of the central cavity, where the protein folds. Following ATP hydrolysis, binding of ATP and GroES to the lower ring structure dissociates the upper GroES and discharges the folded protein. (B, Modified from Xu Z, Horwich AL, Sigler PB: The crystal structure of the asymmetric GroEL-GroES-(ADP)7 chaperonin complex. Nature. 1997;388:741–750. See PDB file 1AON.)
SELECTED READINGS Castellano LM, Shorter J. The surprising role of amylod fibrils in HIV infection. Biology (Basel). 2012;1:58-80. Chiti F, Dobson CM. Protein misfolding, functional amyloid, and human disease. Annu Rev Biochem. 2006;75:333-366. Daggett V, Fersht AR. Is there a unifying mechanism for protein folding? Trends Biochem Sci. 2003;28:18-25. Dobson CM. Protein folding and misfolding. Nature. 2003;426: 884-890. Hayer-Hartl M, Bracher A, Hartl FU. The GroEL-GroES chaperonin machine: A nano-cage for protein folding. Trends Biochem Sci. 2016;41:62-76. Hinnebusch AG. Molecular mechanism of scanning and start codon selection in eukaryotes. Microbiol Mol Biol Rev. 2011;75:434-467. Ibba M, Söll D. Aminoacyl-tRNAs: Setting the limits of the genetic code. Genes Dev. 2004;18:731-738. Ingolia NT. Ribosome profiling: new views of translation, from single codons to genome scale. Nat Rev Genet. 2014;15:205-213. Kim YE, Hipp MS, Bracher A, et al. Molecular chaperone functions in protein folding and proteostasis. Annu Rev Biochem. 2013;82: 323-355. Liu T, Kaplan A, Alexander L, et al. Direct measurement of the mechanical work during translocation by the ribosome. Elife. 2014;3:e03406. May BC, Govaerts C, Prusiner SB, Cohen FE. Prions: So many fibers, so little infectivity. Trends Biochem Sci. 2004;29:162-165. Mazumder B, Seshadri V, Fox PL. Translational control by the 3′-UTR: The ends specify the means. Trends Biochem Sci. 2003;28:91-98. Moore PB. How should we think about the ribosome? Annu Rev Biophys. 2012;41:1-19. Mumtaz MA, Couso JP. Ribosomal profiling adds new coding sequences to the proteome. Biochem Soc Trans. 2015;43:1271-1276.
Myers JK, Oas TG. Mechanisms of fast protein folding. Annu Rev Biochem. 2002;71:783-815. Ow SY, Dunstan DE. A brief overview of amyloids and Alzheimer’s disease. Protein Sci. 2014;23:1315-1331. Pearl LH, Prodromou C. Structure and mechanism of the Hsp90 molecular chaperone machinery. Annu Rev Biochem. 2006;75:271-294. Piper M, Holt C. RNA translation in axons. Annu Rev Cell Dev Biol. 2004;20:505-523. Ramakrishnan V. The ribosome emerges from a black box. Cell. 2014; 159:979-984. Ramakrishnan Lab. Ribosome Structure and Function. Movies and Overview Figures of the Ribosome. . Rodnina MV. The ribosome as a versatile catalyst: reactions at the peptidyl transferase center. Curr Opin Struct Biol. 2013;23: 595-602. Saibil HR. Biochemistry. Machinery to reverse irreversible aggregates. Science. 2013;339:1040-1041. Saio T, Guan X, Rossi P, Economou A, Kalodimos CG. Structural basis for protein antiaggregation activity of the trigger factor chaperone. Science. 2014;344:1250494. Selkoe DJ. Folding proteins in fatal ways. Nature. 2003;426:900-904. Sonenberg N, Dever TE. Eukaryotic translation initiation factors and regulators. Curr Opin Struct Biol. 2003;13:56-63. Voorhees RM, Ramakrishnan V. Structural basis of the translational elongation cycle. Annu Rev Biochem. 2013;82:203-236. Wilkie GS, Dickson KS, Gray NK. Regulation of mRNA translation by 5′- and 3′-UTR-binding factors. Trends Biochem Sci. 2003;28: 182-188. Xue S, Barna M. Specialized ribosomes: a new frontier in gene regulation and organismal biology. Nat Rev Mol Cell Biol. 2012;13: 355-369.
SECTION
Membrane Structure and Function
V
This page intentionally left blank
SECTION V OVERVIEW L
ife, as we know it, depends on a thin membrane that separates each cell from the surrounding world. These membranes, composed of two layers of lipids, are generally impermeable to ions and macromolecules. Proteins embedded in the lipid membrane facilitate the movement of ions, allowing cells to create an internal environment different from that outside. Membranes also subdivide the cytoplasm of eukaryotic cells into compartments called organelles. Chapter 13 introduces the features that are shared by all biological membranes: a bilayer of lipids, integral proteins that cross the bilayer, and peripheral proteins associated with the surfaces. Membranes are a planar sandwich of two layers of lipids that behave like two-dimensional fluids. Each lipid has a polar group coupled to hydrocarbon tails that are insoluble in water. The hydrocarbon tails are in the middle of the membrane bilayer with polar head groups exposed to water on both surfaces. Despite the rapid, lateral diffusion of lipids in the plane of the membrane, the hydrophobic interior of the bilayer is poorly permeable to ions and macromolecules. This impermeability makes it possible for cellular membranes to form barriers between the external environment, cytoplasm, and organelles. The selectively permeable membrane around each organelle allows the creation of a unique interior space for specialized biochemical reactions that contribute to the life of the cell. Chapters 18 to 23 consider in detail all the organelles, including mitochondria, chloroplasts, peroxisomes, endoplasmic reticulum, Golgi
Membrane organization Ch 13
apparatus, lysosomes, and the vesicles of the secretory and endocytic pathways. Peripheral membrane proteins found on the surfaces of the bilayer often participate in enzyme and signaling reactions. Others form a membrane skeleton on the cytoplasmic surface that reinforces the fragile lipid bilayer and attaches it to cytoskeletal filaments. Integral membrane proteins that cross lipid bilayers feature prominently in all aspects of cell biology. Some are enzymes that synthesize lipids for biological membranes (Chapter 20). Others serve as adhesion proteins that allow cells to interact with each other or extracellular substrates (see Chapter 30). Because cells need to sense hormones and many other molecules that cannot penetrate a lipid bilayer, they have evolved thousands of protein receptors that span the bilayer (Chapter 24). Hormones or other extracellular signaling molecules bind selectively to receptors exposed on the cell surface. The energy from binding is used to transmit a signal across the membrane and regulate biochemical reactions in the cytoplasm (Chapters 25 to 27). A large fraction of the energy that is consumed by organs such as our brains is used to create ion gradients across membranes. Several large families of integral membrane proteins control the movement of ions and other solutes across membranes. Chapter 14 introduces three families of pumps that use adenosine triphosphate (ATP) hydrolysis as the source of energy to transport ions or solutes up concentration gradients across membranes. For example, pumps in the plasma
Carriers Ch 15
Pumps Ch 14
Channels Ch 16
H+ K-channel
H+ Na+ ABC transporter
Na-Ca carrier Ca2+
Na/K ATPase pump K+
Membrane physiology Ch 17 Proton pump
225
membranes of animal cells use ATP hydrolysis to expel Na+ and concentrate K+ in the cytoplasm. Another type of pump creates the acidic environment inside lysosomes. A related pump in mitochondria runs in the opposite direction, taking advantage of a proton gradient across the membrane to synthesize ATP. A third family, called ABC transporters, use ATP hydrolysis to move a wide variety of solutes across plasma membranes. Carrier proteins (Chapter 15) facilitate the movement of ions and nutrients across membranes, allowing them to move down concentration gradients much faster than they can penetrate the lipid bilayer. Some carriers couple movement of an ion such as Na+ down its concentration gradient to the movement of a solute such as glucose up a concentration gradient into the cell. Carriers change their shape reversibly, opening and closing “gates” to transport their cargo across the membrane one molecule at a time. Channels are transmembrane proteins with selective pores that allow ions, water, glycerol, or ammonia to move very rapidly down concentration gradients across membranes (Chapter 16). Taking advantage of ion gradients created by pumps and carriers, cells selectively open ion channels to create electrical potentials across the plasma membrane and some organelle membranes. Many channels open and close their pores in response to local conditions. The electrical potential across the membrane regulates voltage-gated cation channels. Binding of a chemical ligand opens other channels. For instance, nerve cells secrete small organic ions (called neurotransmitters) to stimulate other nerve cells and muscles by binding to an extracellular domain of cation channels. The bound neurotransmitter opens the pore in the channel. In the cytoplasm, other organic ions and Ca2+ can also regulate channels. Cyclic nucleotides open plasma membrane channels in cells that respond to light and odors. Inositol triphosphate and Ca2+ control
226
channels that release Ca2+ from the endoplasmic reticulum. Through these diverse activities channels participate in all aspects of membrane physiology. All living organisms depend on combinations of pumps, carriers, and channels for many physiological functions (Chapter 17). Cells use ion concentration gradients produced by pumps as a source of potential energy to drive the uptake of nutrients through plasma membrane carriers. Epithelial cells lining our intestines combine different carriers and channels in their plasma membranes to transport sugars, amino acids, and other nutrients from the lumen of the gut into the blood. Many organelles use carriers driven by ion gradients for transport. Most cells use ion channels and transmembrane ion gradients to create an electrical potential across their plasma membranes. Nerve and muscle cells create fastmoving fluctuations in the plasma membrane potential for high-speed communication; operating on a millisecond time scale, voltage-gated ion channels produce waves of membrane depolarization and repolarization called action potentials. Each of our physiological systems depends on this cooperation among pumps, carriers and channels. Our abilities to perceive our environment, think, and move depend on transmission of electrical impulses between nerve cells and between nerves and muscles at specialized structures called synapses. When an action potential arrives at a synapse, voltage-gated Ca2+ channels trigger the secretion of neurotransmitters. In less than a millisecond, the neurotransmitter stimulates ligand-gated cation channels to depolarize the plasma membrane of the receiving cell. Muscle cells respond with an action potential that sets off contraction. Nerve cells in the central nervous system integrate inputs from many synapses before producing an action potential. Pumps and carriers cooperate to reset conditions after each round of synaptic transmission.
CHAPTER
13
Membrane Structure and Dynamics M
embranes composed of lipids and proteins form the barrier between each cell and its environment. Membranes also partition the cytoplasm of eukaryotes into compartments, including the nucleus and membranebounded organelles. Each type of membrane is specialized for its various functions, but all biological membranes have much in common: a planar fluid bilayer of lipid molecules, integral membrane proteins that cross the lipid bilayer, and peripheral membrane proteins on both surfaces. This chapter opens with a discussion of the lipid bilayer. It then considers examples of integral and peripheral membrane proteins before concluding with a discussion of the dynamics of both lipids and proteins. The following three chapters introduce three large families of membrane proteins: pumps, carriers, and channels. Chapter 17 explains how pumps, carriers, and channels cooperate in a variety of physiological processes. Chapters 24 and 30 cover plasma membrane receptor proteins. A. 1926
B. 1943
Development of Ideas About Membrane Structure Our current understanding of membrane structure began with E. Overton’s proposal in 1895 that cellular membranes consist of lipid bilayers (Fig. 13.1A). In the 1920s it was found that the lipids extracted from the plasma membrane of red blood cells spread out in a monolayer on the surface of a tray of water to cover an area sufficient to surround the cell twice. (Actually, offsetting errors—incomplete lipid extraction and an underestimation of the membrane area—led to the correct answer!) X-ray diffraction experiments in the early 1970s established definitively that membrane lipids are arranged in a bilayer. During the 1930s, cell physiologists realized that a simple lipid bilayer could not explain the mechanical properties of the plasma membrane, so they postulated a surface coating of proteins to reinforce the bilayer
C. 1972
D. 2001
EXTRACELLULAR
WATER
SPACE
Thy-1 Seven-helix receptor
Hydrocarbons
Fatty acid chains – +
Polar groups
Polar groups
+ –
Proteins anchored to phospholipid bilayer
Integral proteins
Protein on surface WATER
ic Dynamlipid pho s o h p bilayer
CYTOPLASM
Src
High-resolution protein structures
Receptor tyrosine kinase
FIGURE 13.1 DEVELOPMENT OF CONCEPTS IN MEMBRANE STRUCTURE. A, Gorder and Grendel model from 1926. B, Davson and Danielli model from 1943 reflecting beliefs of the time about the small sizes of proteins. C, Singer and Nicholson fluid mosaic model from 1972. D, Contemporary model with peripheral and integral membrane proteins. The lipid bilayer shown here and used throughout the book is based on a dynamic computational model (Fig. 13.5). The density of proteins in actual membranes is higher than shown here.
227
228
SECTION V n Membrane Structure and Function
(Fig. 13.1B). Early electron micrographs of thin sections of cells strengthened this view, since all membranes appeared as a pair of dark lines (interpreted as surface proteins and carbohydrates) separated by a lucent area (interpreted as the lipid bilayer). By the early 1970s, two complementary approaches showed that proteins cross the lipid bilayer. First, electron micrographs of membranes that are split in two while frozen (a technique called freeze-fracturing; see Fig. 6.6C) revealed protein particles embedded in the lipid bilayer. Later, chemical labeling showed that many membrane proteins are exposed on both sides of the bilayer. Light microscopy with fluorescent tags demonstrated that membrane lipids and some membrane proteins diffuse in the plane of the membrane. Quantitative spectroscopic studies showed that lateral diffusion of lipids is rapid but that flipping from one side of a bilayer to the other is slow. The fluid mosaic model of membranes (Fig. 13.1C) incorporated this information, showing transmembrane proteins floating in a fluid lipid bilayer. Subsequent work revealed structures of many proteins that span the lipid bilayer, the existence of lipid anchors on some membrane proteins, and a network of cytoplasmic proteins that restricts the motion of many integral membrane proteins
A. Alcohols +NH 3
CH2 CH2 OH Ethanolamine
B. Fatty acids O
O–
C CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 H C H H Palmitic acid
CH3 H3C +N CH3 CH2 CH2 OH Choline
+NH
3
O HC C – O CH2 OH Serine
O O– C C. Phospholipid synthesis CH2 CH2 CH2 CH2 CH2 CH2 Glycerol CH2 HC HC CH2 CH2 CH2 CH2 CH2 CH2 CH2 H C H H Oleic acid
(Fig. 13.1D). The density of proteins in actual membranes is higher than illustrated in the figure.
Lipids Lipids form the framework of biological membranes, anchor soluble proteins to the surfaces of membranes, store energy, and carry information as extracellular hormones and as intracellular second messengers. Lipids are organic molecules generally less than 1000 Da in size that are much more soluble in organic solvents than in water. They consist predominantly of aliphatic or aromatic hydrocarbons. This chapter explains the structures of the major lipids found in biological membranes and how the hydrophobic effect drives lipids to self-assemble stable bilayers. Membranes also contain hundreds of minor lipid species, some of which may also have important biological functions.
Phosphoglycerides Phosphoglycerides (also called glycerophospholipids) are the main constituents of membrane bilayers (Fig. 13.2). (These lipids are often called phospholipids, an
OH H C H H C OH H C H OH Glycerol
CDP CMP
OH H HO
H
OH
H HO
H OH
H H
H OH Inositol
H
H
OH HO
HO OH
OPO32–
H H
H OPO32– Inositol 4,5-biphosphate
D. Common phosphoglyceride Alcohol Phosphate O O P O– H H O Glycerol H C1 C2 C3 H2 O O O O C C CH2 CH2 CH2 CH2 CH2 CH2
C3 C2
C1
Fatty acids CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2
Phosphatidylcholine
FIGURE 13.2 STRUCTURE AND SYNTHESIS OF PHOSPHOGLYCERIDES. A, Stick figures and space-filling models of the alcohol head groups. B, Stick figures and space-filling models of two fatty acids. C, An alcohol, glycerol, and two fatty acids combine to make a phosphoglyceride. In some cases cytidine diphosphate (CDP) provides the phosphate linking glycerol to the alcohol. CMP, cytidine monophosphate. D, Diagram of a phosphoglyceride and a space-filling model of phosphatidylcholine.
CHAPTER 13 n Membrane Structure and Dynamics
TABLE 13.1 Common Fatty Acids of Membrane Lipids Name
Carbons
Double Bonds (Positions)
Myristate
14
0
Palmitate
16
0
Palmitoleate
16
1 (Δ9)
Stearate
18
0
Oleate
18
1 (Δ9)
Linoleate
18
2 (Δ9, Δ12)
Linolenate
18
3 (Δ9, Δ12, Δ15)
Arachidonate
20
4 (Δ5, Δ8, Δ11, Δ14)
imprecise term, as other lipids contain phosphate.) Phosphoglycerides have three parts: a three-carbon backbone of glycerol, two long-chain fatty acids esterified (or attached via an ether link in Archaea) to hydroxyl groups on carbons 1 and 2 (C1 and C2) of the glycerol, and phosphoric acid esterified to the C3 hydroxyl group of glycerol. Most also have an alcohol head group esterified to the phosphate. Fatty acids have a carboxyl group at one end of an aliphatic chain of 13 to 19 additional carbons (Table 13.1). More than half of the fatty acids in membranes have one or more double bonds. Fatty acids and phosphoglycerides are amphiphilic, as they have both hydrophobic (fears water) and hydrophilic (loves water) parts. The aliphatic chains of fatty acids are hydrophobic. The carboxyl groups of fatty acids and the head groups of phosphoglycerides are hydrophilic. The hydrophobic effect (see Fig. 4.5) drives amphiphilic phosphoglycerides to assemble bilayers (see later). Cells make more than 100 major phosphoglycerides using many different fatty acids and esterifying one of five different alcohols to the phosphate. In general, the fatty acids on C1 have no or one double bond, whereas the fatty acids on C2 have two or more double bonds. Each double bond creates a permanent bend in the hydrocarbon chain that contributes to the fluidity of the bilayer. The alcohol head groups give phosphoglycerides their names: phosphatidic acid [PA] (no head group) phosphatidylglycerol [PG] (glycerol head group) phosphatidylethanolamine [PE] (ethanolamine head group) phosphatidylcholine [PC] (choline head group) phosphatidylserine [PS] (serine head group) phosphatidylinositol [PI] (inositol head group) The various head groups confer distinctive properties to the various phosphoglycerides. All head groups have a negative charge on the phosphate esterified to glycerol. Neutral phosphoglycerides—PE and PC—have a positive charge on their nitrogens, giving them a net charge of zero. PS has extra positive and negative charges, giving it a net negative charge like the other acidic
229
phosphoglycerides (PA, PG, and PI). PI can be modified by esterifying one to five phosphates to the hexane ring hydroxyls. These polyphosphoinositides are highly negatively charged. The complicated metabolism of phosphoglycerides can be simplified as follows: Enzymes can interconvert all phosphoglyceride head groups and remodel fatty acid chains. For example, three successive enzymatic methylation reactions convert PE to PC, whereas another enzyme exchanges serine for ethanolamine, converting PS to PE. Other enzymes exchange fatty acid chains after the initial synthesis of a phosphoglyceride. These enzymes are located on the cytoplasmic surface of the smooth endoplasmic reticulum. Biochemistry texts provide more details of these pathways. Several minor membrane phospholipids are variations on this general theme. Plasmalogens have a fatty acid linked to carbon 1 of glycerol by an ether bond rather than an ester bond. They serve as sources of arachidonic acid for signaling reactions (see Fig. 26.9). Cardiolipin has two glycerols esterified to the phosphate of PA.
Sphingolipids Sphingolipids get their name from sphingosine, a nitrogen-containing base synthesized from serine and a fatty acid (Fig. 13.3). Sphingosine acts like the structural counterpart of glycerol plus one fatty acid of phosphoglycerides. Sphingosine carbons 1 to 3 have polar substituents. A double bond between C4 and C5 begins the hydrocarbon tail. Two variable features distinguish the various sphingolipids: the fatty acid (often lacking double bonds) attached by an amide bond to C2 and the nature of the polar head groups esterified to the hydroxyl on C1. Most sugar-containing lipids of biological membranes are sphingolipids. The head groups of glycosphingolipids consist of one or more sugars. Some are neutral; others are negatively charged. All of these head groups lack phosphate. Sugar head groups of some glycosphingolipids serve as receptors for viruses. Alternatively, a phosphate ester can link a base to C1. These so-called sphingomyelins have phosphorylcholine or phosphoethanolamine head groups just like PC and PE. Receptor-activated enzymes remove phosphorylcholine from sphingomyelin to produce the second messenger ceramide (see Fig. 26.11). Sphingolipids are longer than most phosphoglycerides and much more abundant in the thicker plasma membrane than in membranes inside cells (see Fig. 21.3). The hydrocarbon tails of sphingosine and the fatty acid contribute to the hydrophobic bilayer, and polar head groups are on the surface. Sterols Sterols are the third major class of membrane lipids. Cholesterol (Fig. 13.4) is the major sterol in animal
230
SECTION V n Membrane Structure and Function
Derived from serine H HO H OH C C3 C2 C1H2 CH H +NH3 CH2 CH2 CH2 CH2 CH2 Acyl chain CH (CH2)12
Sugar(s)
2
CH2 CH2 CH2 CH2 CH2 CH2 CH3
H HO H O C C C CH2 CH H HN O C CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 Fatty acid CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH3 CH3
C. Sphingomyelin Alcohol (choline or
Choline O ethanolamine) –O P O H HO H O C C C CH2 CH H HN O C CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 Fatty acid CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH2 CH3 CH3 Sphingomyelin
Phosphate Sphingosine
B. Glycosphingolipids
Sphingosine
Derived from fatty acid
A. Sphingosine
FIGURE 13.3 SPHINGOLIPIDS. A, Stick figure and space-filling model of sphingosine. B, Diagram of the parts of a glycosphingolipid. Ceramide has a fatty acid but no sugar. C, Stick figure and space-filling model of sphingomyelin.
A HO
B CH3 CH3 CH CH3
C
CH2 CH2 CH2 H3C C H CH3
FIGURE 13.4 CHOLESTEROL. A, Stick figure. B, Space-filling model. C, Disposition of cholesterol in a lipid bilayer with the hydroxyl oriented toward the surface. The rigid sterol nucleus tends to order fluid bilayers in the region between C1 and C10 of the fatty acids but promotes motion of the fatty acyl chains deeper in the bilayer owing to its wedge shape.
plasma membranes, with lower concentrations in internal membranes. Plants, lower eukaryotes, and bacteria have other sterols in their membranes. The rigid fourring structure of cholesterol is apolar, so it inserts into the core of bilayers with the hydroxyl on C3 oriented toward the surface. Cholesterol is vital to metabolism, being situated at the crossroads of several metabolic pathways, including those that synthesize steroid hormones (such as estrogen, testosterone, and cortisol), vitamin D, and bile salts secreted by the liver. Cholesterol itself is synthesized (see Fig. 20.15) from isopentyl (5-carbon) building blocks that form 10-carbon (geranyl), 15-carbon (farnesyl), and 20-carbon (geranylgeranyl) isoprenoids. As is described later, these isoprenoids are used as hydrocarbon anchors for many important
membrane-associated proteins. Isoprenoids are also precursors of natural rubber and of cofactors present in visual pigments.
Glycolipids Cells have three types of glycolipids: (a) sphingolipids (the predominant form), (b) glycerol glycolipids with a sugar chain attached to the hydroxyl on C3 of a diglyceride, and (c) glycosylphosphatidylinositols (GPIs). Some GPIs simply have a short carbohydrate chain on the hydroxyl of inositol C2. Others use a short sugar chain to link C6 of PI to the C-terminus of a protein (Fig. 13.10C). Triglycerides Triglycerides are simply glycerol with fatty acids esterified to all three carbons. Lacking a polar head group, they are not incorporated into membrane bilayers. Instead, triglycerides form oily droplets in the cytoplasm of cells to store fatty acids as reserves of metabolic energy (see Fig. 28.3). Mitochondria oxidize fatty acids and convert the energy in their covalent bonds into adenosine triphosphate (ATP) (see Fig. 19.4).
Physical Structure of the Fluid Membrane Bilayer Physical Properties of Bilayers of a Single Lipid In an aqueous environment, amphiphilic lipids spontaneously self-assemble into ordered structures in microseconds. The amphiphilic nature of phosphoglycerides and sphingolipids favors formation of lamellar bilayers, planar structures with fatty acid chains lined up more or less normal to the surface and polar head groups on the
CHAPTER 13 n Membrane Structure and Dynamics
A. Book icon
B. Computational model
C. H2O
D. Head groups
1.5 nm
231
E. Hydrocarbon tails
3.5 nm
FIGURE 13.5 COMPUTATIONAL MODEL OF A HYDRATED DIMYRISTOYLPHOSPHATIDYLCHOLINE BILAYER. A, Icon of the lipid bilayer used throughout this book, based on the model shown in B. B, Space-filling model of all the lipid atoms in the simulation. Stick figures of the water molecules are red. The polar regions of phosphatidylcholine (PC) from the carbonyl oxygen to the choline nitrogen are blue. Hydrocarbon tails are yellow. C, Stick figures of the water molecules only. D, Stick figures of the polar regions of PC from the carbonyl oxygen to the choline nitrogen only. E, Stick figures of the hydrocarbon tails only. This model was calculated from first principles starting with 100 PC molecules (based on an x-ray diffraction structure of PC crystals) in a regular bilayer with 1050 molecules of bulk phase water on each side. Taking into account surface tension and distribution of charge on lipid and water, the computer used simple Newtonian mechanics to simulate the molecular motion of all atoms on a picosecond time scale. After less than 100 picoseconds of simulated time, the liquid phase of the lipids appeared. The model shown here is after 300 picoseconds of simulated time. Such models account for most molecular parameters (electron density, surface roughness, distance between phosphates of the two halves, area per lipid [0.6 nm2], and depth of water penetration) of similar bilayers obtained by averaging techniques, including nuclear magnetic resonance (NMR), x-ray diffraction, and neutron diffraction. (Courtesy E. Jakobsson, University of Illinois, Urbana. Modified from Chiu S-W, Clark M, Balaji V, et al. Incorporation of surface tension into molecular dynamics simulation of an interface: a fluid phase lipid bilayer membrane. Biophys J. 1995;69:1230–1245.)
surfaces exposed to water (Fig. 13.5A). The two halves of the bilayer are called leaflets. Bilayer formation is favored energetically by the increase in entropy when the hydrophobic acyl chains interact with each other and exclude water from the core of the bilayer. The head groups of PC and PS have about the same cross-sectional areas as the aliphatic tails, so they are approximately cylindrical in shape, appropriate for flat bilayers. The hydrophobic effect is so strong that it drives lipid head groups into close packing, depleting water from the head group layer. The area per lipid molecule of a given type tends to be constant, so bilayers bend in response if molecules are added asymmetrically to one leaflet. The smaller head group makes PE adopt a slightly conical shape, favoring a curved bilayer. Bilayers of pure lipids are of two physical states depending on the temperature. The liquid disordered phase is a flexible, two-dimensional fluid with disordered acyl chains and the lipids diffusing rapidly (Fig. 13.5A). Tight packing of acyl chains in the gel state limits lateral diffusion. Low temperatures favor the gel state. Above a critical temperature the gel melts and transitions to the liquid disordered phase. The transition temperature depends on the saturation and lengths of the acyl chains. Short acyl chains favor the liquid state. Fatty acids with 18 or more carbons are solid at physiological temperatures unless they contain
double bonds that create a permanent bend and favor the liquid state by preventing tight packing of fatty acid tails in the middle of the bilayer. Phosphoglycerides in biological membranes are largely in the liquid phase owing to their compositions. The C14 and C16 fatty acids are saturated, but C18 fatty acids usually have one to three double bonds and C20 fatty acids have four double bonds (Table 13.1). The phosphoglycerides in particular biological membranes vary in both the lengths and saturation of the acyl chains. For example, abundant polyunsaturated acyl chains in synaptic vesicles (see Fig. 17.8) make the bilayer flexible and facilitate membrane traffic. Biophysical methods, including fluorescence recovery after photobleaching (Fig. 13.12), show that phosphoglycerides diffuse rapidly in the plane of a bilayer with a lateral diffusion coefficient (D) of about 1 µm2 s−1. Given that the rate of diffusion is 2(Dt)1/2 (t = time), a phosphoglyceride moves laterally about 1 µm/s in the plane of the membrane, fast enough to circumnavigate the membrane of a bacterium in a few seconds. Rarely (~10−5 s−1, corresponding to a half-time of 20 hours), a neutral phosphoglyceride, such as PC, flips unassisted from one side of a bilayer to the other. Flipping of charged phosphoglycerides is even slower. A computational method called molecular dynamics simulation is used to study the organization and
232
SECTION V n Membrane Structure and Function
dynamics of lipid bilayers (Fig. 13.5 explains the method). The model shown in Fig. 13.5 has the (short) 14-carbon acyl chains on the inside and polar head groups facing the surrounding water. The molecular density is lowest in the middle of the bilayer. The model emphasizes the tremendous disorder of the lipid molecules, as expected for a liquid. Fatty acid chains undergo internal motions on a picosecond time scale, making them highly irregular, with approximately 25% of the bonds in the bent configuration. Longer simulations show that the lipids wobble and rotate around their long axes in nanoseconds and diffuse laterally on longer time scales. Polar phosphorylcholine head groups vary widely in their orientations, some protruding far into water. This makes the bilayer surface very rough on the nanometer scale. Water penetrates the bilayer only to the level of the deepest carbonyl oxygens, leaving a dehydrated layer approximately 1.5 nm thick in the middle of the bilayer. Bilayers of phosphoglycerides have an electrical potential between the hydrocarbon (positive inside) and the aqueous phase, arising from the orientations of the carbonyl groups and the tendency of water molecules near the bilayer to orient with their positive dipole toward the hydrocarbon interior. These factors dominate over an oppositely oriented electrical dipole between the P and N atoms of the head groups. This inside positive potential may contribute to the barrier to the transfer of positively charged ions and polypeptides across membranes. Despite the disorder and lateral movement of the molecules, bilayers of phosphoglycerides are stable and impermeable to polar or charged compounds, even those as small as Na+ or Cl−. This poor electrical conductivity is essential for many biological processes (see Fig. 17.6). Small, uncharged molecules, such as water, ammonia, and glycerol, penetrate the hydrophobic core in small numbers passing only slowly across bilayers and much more rapidly through channels (see Figs. 16.14 and 16.15). Although bilayers neither stretch nor compress readily, they are very flexible, owing to rapid fluctuations in the arrangement of the lipids. Molecular dynamics simulations accurately reproduce these mechanical properties. Thus, one can draw out a narrow tube of membrane with little resistance by pulling gently on a vesicle composed of a simple bilayer (Fig. 13.6).
Physical Properties of Bilayers of Two or More Lipids All biological membranes consist of mixtures of lipids. Experiments on bilayers reconstituted from purified lipids revealed the physical properties of mixtures of two or more lipids. As expected from first principles, bilayers composed of mixtures of lipids can sort into domains with different compositions. For example, Fig. 13.6 shows a large vesicle formed from cholesterol and 2 forms of PC. The PC with saturated acyl chains segregated into a
Disordered phase Pipette
A
t=5s
Ordered phase
C Pipette t = 180 s
D t = 411 s
B
E
FIGURE 13.6 LIPID SORTING IN DOMAINS DRIVEN BY MEMBRANE CURVATURE. A, Schematic of the experiment. The giant lipid vesicle was composed of 37 mol% 1,2-dipalmitoyl-sn-glycero-3phosphocholine, 33 mol% cholesterol, 30 mol% 1,2-dioleoyl-snglycero-3-phosphocholine and 1 mol% of ganglioside GM1. This mixture of lipids spontaneously sorts into two domains: a disordered liquid domain marked with PE tagged with a red fluorescent dye; and an ordered liquid domain marked with a protein tagged with a green fluorescent dye that binds GM1. A suction micropipette on the left holds the vesicle. A second pipette pulled a narrow tube of membrane from the ordered domain. B, Immediately after the tube was pulled. C, D–E, successive time points showing partitioning of the disordered liquid domain into the tubule. Scale bars are 1 µm.
more ordered liquid phase with a high melting temperature distinct from PC with unsaturated acyl chains in a less-ordered liquid phase with a low melting temperature. Cholesterol has opposite effects on liquid and gel phases of phosphoglycerides, favoring the ordered liquid phase above the transition temperature but disrupting the order of the gel state. The presence of cholesterol in a bilayer makes the acyl chains pack more compactly. This allows lateral mobility of the lipids but restricts movement of small molecules across the bilayer. Sphingolipids are taller than most phosphoglycerides and tend to separate with cholesterol into thicker domains of the bilayer (Fig. 13.7B). These domains are much more abundant in the plasma membrane than in thinner membranes inside cells (see Fig. 21.3).
Structure and Physical Properties of Biological Membranes Biological membranes vary considerably in lipid composition. In addition to a variety of phosphoglycerides, plasma membranes of animal cells are approximately 35% cholesterol and more than 10% sphingolipids (Fig. 13.7), while internal membranes have lower amounts of these lipids. Prokaryotic membranes have different lipid compositions. Bacterial membranes consist of PE, PG, cardiolipin, and other lipids. Archaeal membranes have a mixture of glycolipids, neutral lipids, and ether-linked lipids, and some include single fatty acids. Most lipids are distributed asymmetrically between the halves of biological membranes. In animal cell plasma membranes, glycosphingolipids are outside, while PS,
CHAPTER 13 n Membrane Structure and Dynamics
A
Cholesterol
SM
GS
PC
PE
PS
233
A. Hypotonic
B. Isotonic
C. Hypertonic
D. Phase
E. Lipid
F. Membrane skeleton
∆P
Raft
B FIGURE 13.7 ASYMMETRICAL DISTRIBUTION OF LIPIDS IN THE PLASMA MEMBRANE OF AN ANIMAL CELL. A, Sphingomyelin (SM) and cholesterol form a small cluster in the external leaflet. GS, glycosphingolipid; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PS, phosphatidylserine. PS is enriched in the inner leaflet. B, Lipid raft in the outer leaflet of the plasma membrane enriched in cholesterol and sphingolipids. The lipids in the inner leaflet next to the raft are less well characterized.
PE, and PI are enriched in the cytoplasmic half of the bilayer (Fig. 13.7). PS asymmetry gives the cytoplasmic surface of the plasma membrane a net negative charge. Less is known about the lipid asymmetry of organelle membranes. Transmembrane proteins bind lipids with some specifically, so they also influence the lipid com position of membranes. Cholesterol is distributed more evenly between the two leaflets of membranes because it flips between the two sides of a bilayer on a second time scale. This happens because much less energy is required to bury its single hydroxyl than a polar head group. Lipid asymmetry is initially established during biosynthesis in the cytoplasmic leaflet of the endoplasmic reticulum (ER) (see Chapter 20). A protein (not yet identified) passively redistributes lipids synthesized on the cytoplasmic side of ER between the halves of the bilayer. Lipid asymmetry is reestablished along the secretory pathway and maintained in the plasma membrane by two families of enzymes that use energy from ATP hydrolysis to move lipid molecules from one side of a bilayer to the other. Flippases are P-type adenosine triphosphatase (ATPase) pumps for lipids (see Fig. 14.7). One isoform of the P4 ATPase pumps is found in the Golgi apparatus, while other isoforms are found in secretory vesicles, endosomes, or the plasma membrane to concentrate PS on the cytoplasmic sides of these membranes. A second family called floppases are ABC transporter pumps that move lipids from cytoplasmic leaflet to the extracellular leaflet. The same activity that mixes lipids in the ER also exposes PS on the outer surface of activated platelets
FIGURE 13.8 MEMBRANE DEFORMABILITY ILLUSTRATED BY THE PLASMA MEMBRANE OF HUMAN RED BLOOD CELLS. A–C, Differential interference contrast light micrographs. In an isotonic medium, the cell is a biconcave disk. In a hypotonic medium, water enters the cytoplasm, and the cell rounds up and bursts (arrows) if the area of the membrane cannot accommodate the volume. In a hypertonic medium, water leaves the cell and the membrane is thrown into spikes and folds. D, Phase-contrast micrograph showing that the plasma membrane is flexible enough to be drawn by suction into a capillary tube. E, Fluorescence micrograph showing that membrane lipids, marked with a fluorescent dye, evenly surround the membrane extension. F, The elastic membrane skeleton, marked with another fluorescent dye, stretches into the capillary but not to the tip of the extension. (D–F, Courtesy N. Mohandas, Lawrence Berkeley Laboratory, Berkeley, CA. For reference, see Discher D, Mohandas N, Evans E. Molecular maps of red cell deformation. Science. 1994;266:1032–1035.)
(see Fig. 30.14) and on cells marked for phagocytosis during programmed cell death (see Fig. 46.7). Because they interact favorably, cholesterol and sphingolipids form small domains in the outer leaflet of plasma membranes called rafts (Fig. 13.7B). Special invaginations of the plasma membrane called caveolae (see Fig. 22.7) are the best-characterized example of sphingolipid–cholesterol rafts. Some transmembrane proteins, GPI-anchored proteins, and fatty acid–anchored proteins (Fig. 13.10) associate with sphingolipids and cholesterol in artificial bilayers, so rafts are thought to participate in signaling. Like bilayers of pure phosphoglycerides cellular membranes have limited permeability to ions, high electrical resistance, and the ability to self-seal. Little force is required to deform bilayers into complex shapes. These features are illustrated by the response of a red blood cell plasma membrane to changes in volume (Fig. 13.8). The membrane area is constant, so a reduction in volume
234
SECTION V n Membrane Structure and Function
throws the membrane into folds, whereas swelling distends it to a spherical shape until it eventually bursts. If osmotic forces rupture a lipid bilayer, it will reseal. The lipid molecules comprising membranes are not soluble in water, but cytoplasmic lipid-binding proteins can take up specific lipids from a membrane and deliver them to another membrane. This process transfers lipids from their sites of synthesis in the ER to mitochondria as well as between other organelles (see Fig. 20.17).
Membrane Proteins Proteins are responsible for most membrane functions. The variety of membrane proteins is great, comprising more than one-third of proteins in sequenced genomes. Integral membrane proteins cross the lipid bilayer, and peripheral membrane proteins associate with the inside or outside surfaces of the bilayer. Transmembrane segments of integral membrane proteins interact with hydrocarbon chains of the lipid bilayer and have few hydrophilic residues on these surfaces. Like other soluble proteins, peripheral membrane proteins have hydrophilic residues exposed on their surfaces and a core of hydrophobic residues. Chemical extraction experiments distinguish these two classes of membrane proteins. Alkaline solvents (eg, 0.1 M carbonate at pH 11.3) solubilize most peripheral proteins, leaving behind the lipid bilayer and integral membrane proteins. Detergents, which interact with hydrophobic transmembrane segments, solubilize integral membrane proteins.
Integral Membrane Proteins Atomic structures of a growing number of integral membrane proteins and primary structures of thousands of others show how proteins associate with lipid bilayers (Fig. 13.9). Many integral membrane proteins have a single peptide segment that fulfills the energetic criteria (Box 13.1) for a membrane-spanning α-helix. Glycophorin from the red blood cell membrane was the first of these proteins to be characterized (Fig. 13.9A). Nuclear magnetic resonance experiments established that the single transmembrane segment of glycophorin is an α-helix. This helix interacts more favorably with lipid acyl chains than with water. By analogy with glycophorin, it is generally accepted that single, 25-residue hydrophobic segments of other transmembrane proteins fold into α-helices. In many cases, independent evidence has confirmed that the single segment crosses the bilayer. For example, proteolytic enzymes might cleave the peptide at the predicted membrane interface but cannot access the membrane interior. Potential glycosylation sites might be located outside the cell. Chemical or antibody labeling might identify parts of the protein inside or outside the cells. Transmembrane segments of integral membrane proteins that cross the bilayer more than once are folded
BOX 13.1 Amino Acid Sequences Identify
Candidate Transmembrane Segments
Amino acid sequences of integral membrane proteins provide important clues about segments of the polypeptide that cross the lipid bilayer. Each crossing segment must be long enough to span the bilayer with a minimum of charged or polar groups in contact with the lipid (Fig. 13.8). Polar backbone amide and carbonyl atoms are buried in α-helices or β-sheets to avoid contact with the lipid. Aromatic residues frequently project from transmembrane segments into the lipid near the level where acyl chains are bonded to the lipid head groups (red side chains in Fig. 13.8). A helix of 20 to 25 residues or a β-strand of 10 residues is long enough (3.0 to 3.8 nm) to span a lipid bilayer depending on the thickness of the bilayer. Quantitative analysis of the side chain and backbone hydropathy (aversion to water) of the sequence of an integral membrane protein usually identifies one or more hydrophobic sequences long enough to cross a bilayer (see the legend for Fig. 13.8 for details). The approach works best for helices that are inserted directly in the lipid, like the single transmembrane helix of glycophorin A that has mostly apolar side chains. If a protein has multiple transmembrane helices, some may escape detection by hydrophobicity analysis because they form a hydrophilic cavity lined with charged and polar side chains. For example, two of seven transmembrane helices of bacteriorhodopsin contain charged residues facing the interior of the protein, so they are less hydrophobic than the other transmembrane helices. Transmembrane β-strands are more challenging, as only half of the side chains face the membrane lipids. None of the transmembrane strands of porin qualify as transmembrane segments by hydrophobicity criteria. They are short, and many contain polar residues facing the central cavity.
into α-helices or β-strands. Hydrogen bonding of all backbone amides and carbonyls in the secondary structure minimizes the energy required to bury the backbone in the hydrophobic lipid bilayer. For the same reason, most amino acid side chains in contact with fatty acyl chains in the bilayer are hydrophobic. Membrane proteins can bind specific types of lipids that stabilize the protein. Chapter 20 considers how transmembrane proteins fold during their biosynthesis. Integral membrane proteins with all α-helical transmembrane segments are the most common. Examples are bacteriorhodopsin (Fig. 13.9B; see also Fig. 27.2), pumps (see Figs. 14.3, 14.4, 14.7, and 14.10), carriers (see Fig. 15.4), channels (see Fig. 16.3), cytochrome oxidase (see Fig. 19.5), and photosynthetic reaction centers (see Fig. 19.9). Where these proteins have polar and charged residues in the plane of the bilayer, they generally face away from the lipid toward the interior of the protein, in contrast to the opposite arrangement in water-soluble proteins.
CHAPTER 13 n Membrane Structure and Dynamics
A. Glycophorin
a
B. Bacteriorhodopsin
a
b
b
C. Porin
e
d
c
235
f
g a
e b
c
d
a b
Biological unit (dimer)
TOP VIEW a
Hydropathy index
3
Biological unit (trimer)
Biological unit (trimer) 3
c a
b
d
e f
g
3
0
0
0
-3
-3
-3
HYDROPHOBIC
HYDROPHILIC 20
100
Residue number
20
100
200
Residue number
20
100
200
Residue number
FIGURE 13.9 STRUCTURES OF REPRESENTATIVE INTEGRAL MEMBRANE PROTEINS. Top, Views across the lipid bilayer. Middle, Views in the plane of the lipid bilayer. Bottom, Hydrophobicity analysis. A, Glycophorin, a human red blood cell protein, has a single transmembrane α-helix. The extracellular and cytoplasmic domains are artistic conceptions. The transmembrane helices have a strong tendency to form homodimers in the plane of the membrane (see Protein Data Bank [PDB; www.rcsb.org] file 1MSR). B, Bacteriorhodopsin, a light-driven proton pump from the plasma membrane of a purple bacterium (see Fig. 14.3), has seven transmembrane helices. The green space-filling structure is retinal, the covalently bound, light-absorbing “chromophore.” This structure was first determined by electron microscopy of two-dimensional crystals and extended to higher resolution by x-ray diffraction (see PDB file 1AT9). C, Porin, a nonselective channel protein from the outer membrane of a bacterium, is composed largely of transmembrane β-strands. This structure was determined by x-ray crystallography of three-dimensional crystals (see PDB file 1PRN). Hydropathy plots are calculated from the energy required to transfer an amino acid from an organic solvent to water. One sums the transfer free energy for segments of 20 residues. Segments with large, positive (unfavorable) transfer free energies (around 1.5 on this scale) are more soluble in the hydrophobic interior of a membrane bilayer than in water and thus are candidates for membrane-spanning segments.
Many transmembrane proteins consist of multiple subunits that associate in the plane of the bilayer (Fig. 13.9). The transmembrane helix of glycophorin A has a strong tendency to form homodimers in the plane of the membrane. Dimers are favored because complementary surfaces on a pair of helices interact more precisely with each other than with lipids. The positive entropy change associated with dissociation of lipids from interacting protein surfaces (comparable to the hydrophobic effect in water) drives the reaction. Backbone carbonyl oxygens
also form unconventional hydrogen bonds with C-α hydrogens that stabilize dimers. Bacteriorhodopsin molecules self-associate in the plane of the membrane to form extended two-dimensional crystals. Many membrane channels form by association of four similar or identical subunits with a pore at their central interface (see Fig. 16.2). Bacterial cytochrome oxidase is an assembly of four different subunits with a total of 22 transmembrane helices (see Fig. 19.5). The purple bacterium photosynthetic reaction center consists of three unique
236
SECTION V n Membrane Structure and Function
helical subunits plus a peripheral cytochrome protein (see Fig. 19.9). A minority of integral membrane proteins use β-strands to cross the lipid bilayer. Porins form channels for many substances, up to the size of proteins, to cross the outer membranes of Gram-positive bacteria and their eukaryotic descendents, mitochondria and chloroplasts. Porins consist of an extended β-strand barrel with a hydrophobic exterior surrounding an aqueous pore (Fig. 13.9C). These subunits associate as trimers in the lipid bilayer. In addition to transmembrane helices or strands, many integral membrane proteins have structural elements that pass partway across the bilayer. Porins have extended polypeptide loops inside the β-barrel. Many channel proteins have short helices and loops that reverse in the middle of the membrane bilayer. These structural elements help form pores specific for potassium (see Fig. 16.3), chloride (see Fig. 16.13), and water (see Fig. 16.15).
Peripheral Membrane Proteins Six strategies bind peripheral proteins to the surfaces of membranes (Fig. 13.10). One of three different types of hydrophobic acyl chains can anchor a protein to a membrane by inserting into the lipid bilayer. Other proteins bind electrostatically to membrane lipids, and some insert partially into the lipid bilayer. Many peripheral proteins bind directly or indirectly to integral membrane proteins. Isoprenoid Tails A 15-carbon isoprenoid (farnesyl) tail (see Fig. 20.15) is added posttranslationally to the side chain of a cysteine
residue near the C-terminus of the guanosine triphosphatase (GTPase) Ras (see Fig. 4.6) and many other proteins. The enzyme making this modification recognizes the target cysteine followed by two aliphatic amino acids plus any other amino acid (a CAAX recognition site). Another enzyme cleaves off the AAX residues. This membrane attachment is required for Ras to participate in growth factor signaling (see Fig. 27.6).
Myristoyl Tails Myristate, a 14-carbon saturated fatty acid, anchors the tyrosine kinase Src (see Box 27.5) and other proteins involved in cellular signaling to the cytoplasmic face of the plasma membrane. Myristate is added to the amino group of an N-terminal glycine during the biosynthesis of these proteins. Insertion of this short, fatty acyl chain into a lipid bilayer is so weak (Kd: ~10−4 M) that additional electrostatic interactions between basic side chains of the protein and head groups of acidic phosphoglycerides are required to maintain attachment to the membrane. Glycosylphosphatidylinositol Tails A short oligosaccharide-phosphoglyceride tail links a variety of proteins to the outer surface of the plasma membrane. The C-terminus of the protein is attached covalently to the oligosaccharide, and the two fatty acyl chains of PI are in the lipid bilayer. In animal cells, this GPI anchors important plasma membrane proteins, including enzymes (acetylcholine esterase; see Fig. 17.9), adhesion proteins (T-cadherin; see Fig. 30.5), and cell surface antigens (Thy-1). The protozoan parasite
C
C. Thy-1 A. Ras
C
B. Src peptide
E. Prostaglandin synthase
F. Cadherin and catenin
N
D. Annexin
Tail of cadherin β-catenin
FIGURE 13.10 SIX MODES OF ASSOCIATION OF PERIPHERAL MEMBRANE PROTEINS WITH LIPID BILAYERS. A, A C-terminal isoprenoid tail attaches Ras to the bilayer (see PDB file 121P). B, An N-terminal myristoyl tail binds Src weakly to the bilayer. Electrostatic interactions between acidic lipids and basic amino acids stabilize the interaction. C, A C-terminal glycosylphosphatidylinositol (GPI) tail anchors Thy-1 (similar to an immunoglobulin variable domain) to the bilayer. D, Electrostatic interactions with phospholipids bind annexin to the bilayer (see PDB file 1A8A). E, Hydrophobic helices of prostaglandin H2 synthase partially penetrate the lipid bilayer (see PDB file 1CQE). F, The peripheral protein β-catenin (blue and purple; see PDB file 1CQE) associates with the cytoplasmic portion of the transmembrane adhesion protein cadherin (red and green; see PDB file 1FF5).
CHAPTER 13 n Membrane Structure and Dynamics
Trypanosoma brucei covers itself with a high concentration of a GPI-anchored protein. If challenged by an antibody response from the host, the parasite sheds the protein by hydrolysis of the lipid anchor and expresses a variant protein to evade the immune system.
Electrostatic Interaction With Phospholipids As postulated in the 1930s (Fig. 13.1), some soluble cytoplasmic proteins bind the head groups of membrane lipids. Annexins, a family of calcium-binding proteins implicated in membrane fusion reactions, bind tightly to PS (Fig. 13.10D). A second example is the “BAR” domain found in a variety of proteins. Positively charged residues on the concave surface of curved, dimeric BAR domains bind electrostatically to curved membranes or deform flat membranes into tubules (see amphiphysin in Fig. 22.11). Myosin-I motor proteins (see Fig. 36.7) also bind strongly to acidic phosphoglycerides of cellular membranes. Partial Penetration of the Lipid Bilayer Hydrophobic α-helices of prostaglandin H2 synthase (see Figs. 13.10E and 26.9) anchor the enzyme to membranes by partially penetrating the lipid bilayer. Another example is reticulons, proteins that insert into the cytoplasmic leaflet of the ER membrane and promote bending into narrow tubules and sharply curved edges of sheets (see Fig. 20.3). Association With Integral Proteins Many peripheral proteins bind cytoplasmic domains of integral membrane proteins. For example, catenins bind transmembrane cell adhesion proteins called cadherins (Fig. 13.10F). These protein–protein interactions provide more specificity and higher affinity than do the interactions of peripheral proteins with membrane lipids. Such protein–protein interactions anchor the cytoskeleton to transmembrane adhesion proteins (see Fig. 31.8) and guide the assembly of coated vesicles during endocytosis (see Fig. 22.9). Protein–protein interactions also provide a way to transmit information across a membrane. Ligand binding to the extracellular domain of a transmembrane receptor can change the conformation of its cytoplasmic domain, promoting interactions with cytoplasmic, signaltransducing proteins (see Chapter 24). The membrane skeleton on the cytoplasmic surface of the plasma membrane of human red blood cells (Fig. 13.11) provided the first insights regarding interaction of peripheral and integral membrane proteins. Two types of integral membrane proteins—an anion carrier called Band 3 and glycophorin—anchor a two-dimensional network of fibrous proteins to the membrane. The main component of this network is a long, flexible, tetrameric, actin-binding protein called spectrin (after its discovery in lysed red blood cells, “ghosts”; see Fig. 33.17). A linker protein called ankyrin binds tightly to both Band 3 and spectrin. Approximately 35,000 nodes consisting of a
A
237
D
B
C
Band 3 4.2 Ankyrin
Spectrin
Glycophorin C 4.1 β-actin Tropomyosin Tropomodulin
Dematin Adducin
FIGURE 13.11 THE MEMBRANE SKELETON ON THE CYTOPLASMIC SURFACE OF THE RED BLOOD CELL PLASMA MEMBRANE. A, Whole cell. B, Cutaway drawing. C, Detailed drawing. Nodes consisting of a short actin filament and associated proteins interact with multiple spectrin molecules, which, in turn, bind to two transmembrane proteins: glycophorin and (via ankyrin) Band 3. D, An electron micrograph of the actin-spectrin network. (D, Courtesy R. Josephs, University of Chicago, IL.)
short actin filament and associated proteins interconnect the elastic spectrin network. This membrane skeleton reinforces the bilayer, allowing a cell to recover its shape elastically after it is distorted by squeezing through the narrow lumen of blood capillaries.
Membrane Protein Dynamics Several complementary methods can monitor movements of plasma membrane proteins (Fig. 13.12A). The original approach was to label proteins with a fluorescent dye, either by covalent modification or by attachment of an antibody with a bound fluorescent dye. After a spot of intense light irreversibly bleaches the fluorescent dyes in a small area of the membrane, one observes the fluorescence over time with a microscope. If the test protein is mobile, unbleached proteins from surrounding areas move into the bleached area. The rate and extent of fluorescence recovery after photobleaching (FRAP) revealed that a fraction of the population of most membrane proteins diffuses freely in two dimensions in the plane of the membrane, but that a substantial fraction is immobilized because the recovery from photobleaching
238
SECTION V n Membrane Structure and Function
A. Fluorescence photobleaching
0 sec
A. Free diffusion
1 sec
B. Single bead
B. Partial confinement
C. Directed motion
10 sec
C. Laser trap
Bead Proteins 0 sec
1 sec
FIGURE 13.12 METHODS USED TO DOCUMENT THE MOVEMENTS OF MEMBRANE PROTEINS. A, Fluorescence recovery after photobleaching. Simulated experimental data with individual molecules are shown as green dots. B, Single-particle tracking. C, Optical trapping.
is incomplete. The same photobleaching method is used to study the mobility of fluorescent fusion proteins targeted to any cellular membrane (see Fig. 6.3). The second approach is to label individual membrane proteins with antibodies or lectins (carbohydrate-binding proteins) attached to small particles of gold or plastic beads (Fig. 13.12B). High-contrast light microscopy can follow the motion of a particle attached to a membrane protein. Despite their size, the particles have minimal effects on diffusion of membrane proteins. The third method is an extension of single-particle tracking. Instead of merely watching spontaneous movements, the investigator can grab a particle in an optical trap created by focusing an infrared laser beam through the microscope objective (Fig. 13.12C). Manipulation of particles with an optical trap reveals what happens when force is applied to a membrane protein. Membrane proteins exhibit a wide range of dynamic behaviors (Fig. 13.13). Some molecules diffuse freely. Others diffuse intermittently, alternating with periods of restricted movement. Substantial numbers of membrane proteins are immobilized, presumably by direct or indirect associations with the membrane skeleton or the cytoskeleton, or by forming large arrays through mutual interactions. The population of a given type of membrane protein (eg, a cell adhesion protein) may exhibit more than one class of dynamic behavior. For example, most proteins with GPI anchors diffuse freely, as is expected from their association with the lipid bilayer, but a fraction of any GPI-anchored protein has restricted mobility. Some transmembrane proteins also diffuse freely, but a fraction may become trapped or immobilized at any time. Diffusing proteins must be free of interactions with the membrane skeleton and with anchored membrane proteins. Cell adhesion proteins (cadherins; see Fig. 30.5) and
FIGURE 13.13 MOVEMENTS OF PROTEINS IN THE PLANE OF MEMBRANES. A, Free diffusion. B, Partial confinement by obstacle clusters, some associated with the membrane skeleton. C, Directed movement by a motor on an actin filament. (For reference, see Jacobson K, Sheets ED, Simson R. Revisiting the fluid mosaic model of membranes. Science. 1995;268:1441–1442.)
nutrient receptors (transferrin receptors; see Fig. 22.15) are examples of transmembrane proteins that diffuse intermittently. They alternate between free diffusion and temporary trapping for 3 to 30 seconds in local domains measuring less than 0.5 µm in diameter. In some cases, trapping depends on the cytoplasmic tails of transmembrane proteins, which are thought to interact reversibly with the cytoskeleton or with immobilized membrane proteins. Tugs with an optical trap show that the cages that confine these particles are elastic, as expected for cytoskeletal networks. Extracellular domains of these proteins may also interact with adjacent immobilized proteins. Immobilized proteins do not diffuse freely, and particles attached to them resist displacement by optical traps. The lipid bilayer can flow past immobilized transmembrane proteins without disrupting the membrane. If the plasma membrane of a red blood cell is sucked into a narrow pipette (Fig. 13.8D), lipids of the fluid membrane bilayer extend uniformly over the protrusion, leaving behind the immobilized membrane proteins and the membrane skeleton. Some membrane proteins undergo long-distance translational movements in relatively straight lines. Because disruption of cytoplasmic actin filaments by drugs impedes these movements, myosins (see Fig. 36.7) are the most likely motors for these movements. In some instances, members of the integrin family of adhesion proteins (see Fig. 30.9) use this transport system. Movements of membrane proteins in the plane of the membrane are essential for many cellular functions. Transmembrane receptors concentrate in coated pits before internalization during receptor-mediated endocytosis (see Fig. 22.12). Similarly, transduction of many signals from outside the cell depends on the formation of receptor dimers or trimers (see Figs. 24.5, 24.7, 24.8, 24.9, 24.10, and 46.18). Bound extracellular ligands stabilize collisions between freely diffusing receptor
proteins, juxtaposing their cytoplasmic domains and activating downstream signaling mechanisms. Similarly, movements in the plane of the plasma membrane allow clustering of adhesion receptors that enhances binding of cells to their neighbors or to the extracellular matrix (see Figs. 30.6 and 30.11). ACKNOWLEDGMENTS We thank Tobias Baumgart and Donald Engelman for their suggestions on this chapter. SELECTED READINGS Blaskovic S, Blanc M, van der Goot FG. What does S-palmitoylation do to membrane proteins? FEBS J. 2013;280:2766-2774. Curran AR, Engelman DM. Sequence motifs, polar interactions and conformational changes in helical membrane proteins. Curr Opin Struct Biol. 2003;13:412-417. Engelman DM. Lipid bilayer structure in the membrane of Mycoplasma laidlawii [bilayer structure established by x-ray diffraction]. J Mol Biol. 1971;58:153-165. Fleming KG. Energetics of membrane protein folding. Annu Rev Biophys. 2014;43:233-255. Forrest LR. Structural symmetry in membrane proteins. Annu Rev Biophys. 2015;44:311-337. Kiessling LL, Splain RA. Chemical approaches to glycobiology. Annu Rev Biochem. 2010;79:619-653.
CHAPTER 13 n Membrane Structure and Dynamics
239
McNeil PL, Steinhardt RA. Plasma membrane disruption: repair, prevention, adaptation. Annu Rev Cell Dev Biol. 2003;19:697-731. Nagle JF, Tristram-Nagle S. Structure of lipid bilayers. Biochim Biophys Acta. 2000;1469:159-195. Owen DM, Magenau A, Williamson D, Gaus K. The lipid raft hypothesis revisited—new insights on raft composition and function from super-resolution fluorescence microscopy. Bioessays. 2012;34: 739-747. Pandit SA, Scott HL. Multiscale simulations of heterogeneous model membranes. Biochim Biophys Acta. 2009;1788:136-148. Robertson JD. Membrane structure [historical perspective]. J Cell Biol. 1981;91:1895-2045. Sachs JN, Engelman DM. Introduction to the membrane protein reviews: the interplay of structure, dynamics, and environment in membrane protein function. Annu Rev Biochem. 2006;35: 707-712. Shevchenko A, Simons K. Lipidomics: coming to grips with lipid diversity. Nat Rev Mol Cell Biol. 2010;11:593-598. Simons K, Geri MJ. Revitalizing membrane rafts: new tools and insights. Nat Rev Mol Cell Biol. 2010;11:688-699. Stoeckenius W, Engelman DM. Current models for the structure of biological membranes [historical perspective]. J Cell Biol. 1969;42: 613-646. Wang L. Measurements and implications of the membrane dipole potential. Annu Rev Biochem. 2012;81:615-635. Wollam J, Antebi A. Sterol regulation of metabolism, homeostasis and development. Annu Rev Biochem. 2011;80:885-916. Zverina EA, Lamphear CL, Wright EN, Fierke CA. Recent advances in protein prenyltransferases: substrate identification, regulation, and disease interventions. Curr Opin Chem Biol. 2012;16:544-552.
This page intentionally left blank
CHAPTER
14
Membrane Pumps Introduction to Membrane Permeability Lipid bilayers provide a barrier to diffusion of ions and polar molecules larger than about 150 Da, so transmembrane proteins are required for selective passage for ions, and other larger molecules across membranes. The proteins controlling membrane permeability fall into three broad classes—pumps, carriers, and channels— each with distinct properties (Fig. 14.1). These proteins control traffic of small molecules across membranes, an essential of many physiological processes. • Pumps are enzymes that use energy from adenosine triphosphate (ATP), light, or (rarely) other sources to move ions (generally, cations) and other solutes across membranes at relatively modest rates. Pumps are also called primary active transporters, because they transduce electromagnetic or chemical energy directly into transmembrane concentration gradients between membrane-bound compartments.
• Carriers provide passive pathways for solutes to move across membranes down their concentration gradients from a region of higher concentration to one of lower concentration. Each conformational change in a carrier protein translocates a limited number of small molecules across the membrane. Carriers use ion gradients created by pumps as a source of energy to perform a remarkable variety of work. Translocation of an ion down its concentration gradient can drive another ion or solute up a con centration gradient, so these are called secondary transporters (see Chapter 15). • Channels are ion-specific pores that typically open and close transiently in a regulated manner. Open channels are highly specific but passive transporters, allowing a flood of an ion or other small solute to pass quickly across the membrane, driven by electrical and concentration gradients. The movement of ions through open channels controls the electrical
Pump
Carrier
Channel
Specificity
Absolute
Only 10–20×
Rate (ions/sec)
100
Intermediate 8 mM in disease states. This fourfold variation in Ko changes the membrane potential by 30 to 37 mV, enough to affect cellular processes that are sensitive to the membrane potential. Other channels open and close selectively in response to extracellular or intracellular ligands, membrane potential, physical forces, or other factors (see text). Selective activation of channels is responsible for action potentials and other behavior of excitable membranes (see Fig. 17.6).
APPENDIX 16.4
Charging and Discharging the Membrane Opening or closing ion channels influences the membrane potential and the flux of ions across the membrane. This discussion explains how movement of just a few ions allows cells to change their membrane potential without dissipating ion gradients across the membrane. Consequently, flux through a few ion channels rapidly changes the membrane potential during action potentials. The result of opening multiple channels with different ion selectivities and concentration gradients is also explained.
Membrane Capacitance The membrane potential (E) produced by a given net charge inside the cell (Q) depends on the physical properties of the membrane, summarized in a constant called capacitance (C):
Q C Capacitance depends on membrane area, thickness (physical separation between internal and external charges), and dielectric constant. If the capacitance is large, many ions must move to change the membrane potential. For cell membranes, the capacitance is approximately 1 mF/cm2. One farad is 6 × 1018 charges per volt. E=
Charge Movement for a Small Cell The following calculation shows why ion concentration gradients change little during most electrical events in cells. This is important to eliminate the requirement for excessive energy to restore ion gradients. A cell that is 18 μm in diameter might have a capacitance of 10−11 F, or 6 × 107 charges per volt of membrane potential. Thus,
CHAPTER 16 n Membrane Channels
movement of 6 million positive charges out of the cell produces a membrane potential of −0.1 V, or −100 mV. A cell of this size with an internal concentration of 150 mM K+ contains approximately 2.7 × 1011 K+, so movement of fewer than one in 40,000 K ions (0.0025%) from inside to outside creates a large membrane potential. This fraction of ions is far less for large cells, owing to their smaller ratio of surface area to volume. Thus, little energy is required for a large change in membrane potential, such as an action potential. When ion channels open, few ions cross the membrane before an opposing electrical field develops and retards further flux. In Chapters 14 and 15, pumps and carriers were also noted to produce opposing membrane potentials when moving ions across membranes. This can be avoided by opening ion channels that short-circuit the change in membrane potential by providing pathways for counterions to move in the same direction or similar ions to move in the opposite direction across the membrane.
Rate of Charge Movement Through Channels A current is the rate of movement of charge. The ionic current (I) across a membrane is taken as positive when charges move outward. According to this definition, the equation for conservation of charge in a cell is dQ = −I dt A positive current reduces the net charge inside the cell, and vice versa. Including the relationship for capacitance (E = Q/C), the equation relates the current to the rate of change of membrane potential: dE − I = dt C Because channels conduct about 6 × 106 charges per second, a single open channel changes E at a rate of −100 mV/sec on this 18-μm cell. Because most channels occur at densities of 50–200/μm2, an 18-μm cell will have 50,000 to 200,000 channels. If a few channels open together, the membrane potential rapidly approaches the Nernst potential for the selected ion. This explains why most electrical events in cells transpire in a millisecond time frame. Because the rate of current flow through ion channels is not limiting, the time course of electrical events depends on the kinetics of channel opening and closing. This focuses attention on factors that control whether channels are open or closed, also known as gating.
Net Current Through Ion-Selective Channels Another way to describe ionic current across a membrane is
283
complicated (Fig. 16.17B), so electrophysiologists approximate this current-voltage relationship of channels by a linear relationship, such as Ohm’s law (E = IR): I = g ( E − Eion ) where g is conductance (inverse of resistance) and Eion is the reversal potential of a particular ion channel (the potential at which current reverses from out to in). For perfectly selective pores, the reversal potential for each ion equals its Nernst potential, even in the face of other ionic gradients. The unit used for current is siemens (equivalent to 1 ampere per volt). Most channels have currents in the picosiemens range (10−12 S). For a simple pore, a plot of current versus membrane potential is linear, with no current at Eion; real channels are more complicated. Typical plots of current versus voltage deviate from a straight line. This is called rectification. Deviation may be attributable to voltagedependent conformational changes in the channel protein or to nonpermeant ions blocking the pore. Each channel contributes independently to the total current, so given n channels on a cell membrane, the total current is I = ng ( E − Eion ) Opening Na+ and K+ channels has opposite effects because the ion concentration gradients are reversed. The Nernst potential for Na+ is approximately +65 mV in a typical cell, given a 10-fold excess of Na+ outside the cell. Current through an Na+ channel is negative (ie, inward) at membrane potentials below ENa. Thus, if a Na+ channel opens on a cell in which E equals 0, the membrane potential rises toward ENa.
Consequence of Multiple Channel Types Opening Simultaneously More than one type of open channel creates a situation more complicated than the equilibrium described by the Nernst potential for a single-ion species (Figs. 16.18 and 16.19). Consider a cell with physiological ion gradients and two channels—one open K+ channel and one open Na+ channel—having conductances of gK and gNa. The total current through these two channels is the sum of the individual currents: I total = gK ( E − EK ) + gNa ( E − ENa ) Note from this relationship that current is zero at the midpoint between EK and ENa, and the line has twice the slope of a single channel (ie, twice the conductance). Which channel predominates? The equation for Itotal can also be written as I total = geff ( E − Eeff )
I = zeo ( J o − J i )
where the effective conductance geff and reversal potential Eeff are given by
where eo is the elementary charge. The dependence of current on membrane potential for real channels is
geff = gK + gNa
284
SECTION V n Membrane Structure and Function
conductances and a reversal potential that is the weighted average of their reversal potentials, that is, weighted by their relative conductances (Fig. 16.19A). Goldman, Hodgkin, and Katz formulated another equation for E. It uses permeability (P, in units of cm/s) to describe the membrane potential:
Current
A
r
+
K
Total
+
Eeff
EK –100
y onl +
K –50
0
d an
Na
e th ge o t
ENa
E 50 mV
100
ly + on Na
B
FIGURE 16.19 MEMBRANE POTENTIAL AND CURRENTS ACROSS A MEMBRANE WITH TWO TYPES OF CHANNELS. A, Dependence of currents on membrane potential resulting from opening either K+ channels or Na+ channels individually or together. In contrast to Fig. 16.17, which shows ion fluxes in each direction, this is a plot of net current. EK and ENa are the equilibrium potentials (zero current) when only potassium or sodium channels are open. When both types of channels are open, the equilibrium potential (Eeff) is midway between the equilibrium potentials of the two types of channels. B, Distribution of positive (red) and negative (blue) ions across the plasma membrane and around a cell having a negative membrane potential. Excess negative charge builds up near the inside of the membrane, with the excess positive charge near the outside.
and Eeff =
gK EK g E + Na Na gK + gNa gK + gNa
The two channels together act like a single channel with an effective conductance equal to the sum of their
E=
RT PNa [Na]o + PK [ K ]o + PCl [Cl]o + ln F PNa [Na]i + PK [ K ]i + PCl [Cl]i +
This equation summarizes the concepts presented here about membrane potentials. Just two factors determine the membrane potential: (a) the concentration gradients of different ions (eg, the Nernst potentials for each ion) and (b) the relative permeabilities of the membrane to these ions. When all Na+ and Cl− channels are closed (PNa, PCl = 0), the equation reduces to the Nernst relationship for K. When all K+ and Cl− channels are closed (PK, PCl = 0), the equation collapses to the Nernst relationship for Na+. In nerve cells, the resting membrane is most permeable to K+ but also slightly permeable to Na+, so the resting potential is near EK. Opening more K+ channels or lowering extracellular K+ makes the resting potential more negative. Opening more Na+ channels or raising extracellular Na+ makes the resting potential more positive.
Charge Redistribution by Electrical Conduction Most cellular ions have balancing counterions and are distributed randomly in solution, whereas unpaired ions contributing to membrane potentials are confined to boundary layers near the membrane (Fig. 16.19B). Likecharged ions repel one another, so unpaired ions tend to accumulate at boundaries where they can move no farther. During electrical events, unpaired ions redistribute over membrane surfaces by electrical conduction at rates much faster than diffusion. This works as follows: Ions are always in motion, exchanging places. Introduction of extra ions sets off a chain of movements as neighbors repel each other, resulting in rapid spread of unbalanced charge near the membrane. Diffusion of the entering ions over the plane of the membrane would take much longer than this electrical wave. Thus, electrical signaling is the fastest signaling process in cells.
17
CHAPTER
Membrane Physiology This chapter describes how pumps, carriers, and chan-
nels cooperate in living systems. These three components often work together in circuits or cycles. Pumps establish gradients of ions across membranes (see Chapter 9). Channels regulate membrane permeability to these ions to maintain the electrical potential (see Chapter 11) required for membrane excitability. Carriers use ion gradients as a source of energy to drive transport as well as to do other work (see Chapter 10). Coupling ion fluxes through pumps and carriers to do work is called a chemiosmotic cycle. Selective expression of a repertoire of pumps, carriers, and channels in specific membrane compartments enables cells to build sophisticated machines from a stockpile of standard components. If the pumps, carriers, and channels produced by a cell are known, it is relatively easy to explain complicated physiological processes. The examples in this chapter also show how defects in pumps and channels cause disease and how drugs can alleviate the symptoms.
membrane. This raises the concentration of a cation (C+) on one side and depletes it on the other side of a membrane-bounded compartment. An ion gradient is characterized by a chemical term, the concentration gradient, and an electrical gradient (the membrane potential explained in Fig. 11.17). The electrochemical potential across a membrane represents a reservoir of power and a capacity to do work, also known as an ion-motive force. A macroscopic analogy is using a pump to fill an elevated reservoir with fluid. Carriers and other membrane proteins use the potential energy of ion gradients to drive other processes. This is analogous to using fluid flow out of a reservoir to drive a turbine, which uses the energy for other types of work. Many carriers use energy derived from the downhill passage of one substrate to transport one or more other substances up their concentration gradients across the same membrane barrier. In Fig. 17.1, the carrier links
C+
Chemiosmotic Cycles A simple chemiosmotic cycle couples a cation transporting pump to solute transport by a carrier across the plasma membrane or an organelle membrane (Fig. 17.1). The driving reaction is called the primary transport step. This involves an input of energy and, in most cases, some chemical reaction. Other steps called secondary transport reactions depend on ion gradients. The transported substrate is the same chemically on both sides of the membrane. Although chemiosmotic cycles are simple in concept, their importance and power should not be underestimated. They operate in every membrane of every cell. Pumps use energy derived from adenosine triphosphate (ATP) hydrolysis, light absorption, or another chemical reaction (see Table 9.1) to move ions across a
C+ S
+ +++++ +++ +
– – – –– – – –
ATP hydrolysis C+
Pump
C+ S
Carrier
FIGURE 17.1 A MODEL CHEMIOSMOTIC CYCLE IN A MEMBRANE SURROUNDING A CLOSED SPACE. An adenosine triphosphate (ATP)-driven pump transports a cation C+ out of the compartment. The energy derived from ATP is stored as a concentration gradient of C+ (red triangle) and a membrane potential (yellow arrow) across the membrane. The carrier uses the electrochemical gradient of C+ to drive the transport of both C+ and a solute up a concentration gradient (green triangle) across the membrane.
285
286
SECTION V n Membrane Structure and Function
the transport of solute S to the movement of cation C+ down its gradient. Recirculation of cations allows a cell to accumulate solute against its concentration gradient. In addition to the osmotic work illustrated in Fig. 17.1, chemiosmotic cycles can do chemical work. During both oxidative and photosynthetic phosphorylation, proton cycles drive ATP synthesis by rotary ATP synthases (see Figs. 9.5, 19.5, and 19.8). Chemiosmotic cycles can also perform mechanical work. An electrochemical gradient of protons across the plasma membrane drives the rotation of bacterial flagella (see Fig. 38.25). Chemiosmotic cycles using protons dominate the biological world. Most bacterial cycles involve proton pumps, proton-linked carriers, or other proton-linked events. The same is true of early-branching eukaryotes, fungi, and plants. Plasma membranes of plant cells have a powerful proton pump and a collection of proton carriers. Proton chemiosmotic cycles are also characteristic of most eukaryotic organelles, including the Golgi apparatus, endosomes, lysosomes, mitochondria, and chloroplasts. Animal cell plasma membranes are a major exception, because they use predominantly sodium ions for their chemiosmotic cycles.
Epithelial Transport Net transport across an epithelium depends on tight junctions (see Fig. 31.2) that seal the extracellular space between the cells (Fig. 17.2). These junctions separate two extracellular compartments. The apical compartment is the free surface or lumen of the organ (eg, the intestine, respiratory tract, or kidney tubules—topologically continuous with the external world). The basolateral compartment lies between epithelial cells and is continuous with the underlying connective tissue and its blood vessels. Tight junctions restrict diffusion of solutes between the apical and basolateral compartments of the extracellular space. The extent of this seal varies from very tight to leaky. Tight junctions also separate the plasma membrane into apical and basolateral domains, restricting the movement of integral membrane proteins between these domains.
Glucose Transport in the Intestine, Kidney, Fat, and Muscle A chemiosmotic cycle transports glucose from food uphill from the lumen of the intestine to the blood (Fig. 17.2). Tight junctions restrict movement of glucose between the epithelial cells, so all the glucose must move through the epithelial cells using the following components: • Na+K+-adenosine triphosphatase (ATPase), located in the basolateral plasma membrane • SGLT1 (sodium glucose cotransporter 1) Na+/ glucose symporter, restricted to the apical plasma membrane
LUMEN/Apical
Na+
Glucose
Tight junction
Na+
Glucose
Basolateral membrane
Glucose
BASAL LAMINA
3 Na+
2 K+ Glucose
Glucose
3 Na+
Glucose
TISSUE SPACE exchangeable with blood FIGURE 17.2 GLUCOSE TRANSPORT BY THE INTESTINAL EPITHELIUM. Tight junctions seal the epithelium of polarized epithelial cells. Na+K+-adenosine triphosphatase (ATPase) pumps (spacefilling model) in the basolateral plasma membrane drive Na+/glucose symporters in the apical plasma membrane (upper inset) and glucose uniporters in the basolateral plasma membrane (left icon in lower inset) to move glucose from the lumen of the intestine to the blood. Basolateral K+ channels (middle icon) recycle K+ pumped into the cell.
• GLUT5 (glucose transporter 5) glucose uniporter, restricted to the basolateral plasma membrane Na+K+-ATPases use ATP hydrolysis to produce Na+ and + K gradients across the plasma membrane by continuously pumping Na+ out of and K+ into the cell. SGLT Na+/ glucose symporters use Na+ moving inward down its electrochemical gradient across the apical plasma membrane to accumulate high internal concentrations of glucose from the gut lumen. In this step, energy is expended (dissipation of the Na+ gradient) to move glucose uphill. GLUT (glucose transporter) uniporters in the basolateral membrane simply facilitate movement of cytoplasmic glucose down its concentration gradient out of the cell. The kidney uses a similar strategy to recapture glucose filtered from blood, transporting it across the renal proximal tubule cell and back into the blood. Fat and muscle cells use the GLUT4 D-glucose uniporter to take up glucose from the blood when it is plentiful following a meal. These cells store GLUT4 internally in membrane vesicles. After a meal, high blood glucose stimulates secretion of insulin into blood. Signal transduction mechanisms (see Fig. 27.7) lead to fusion of the GLUT4 vesicles with the plasma membrane. This increases the rate of glucose transport into fat and muscle by fivefold to 20-fold, lowering the blood glucose
287
CHAPTER 17 n Membrane Physiology
concentration and providing the cells with glucose to convert into glycogen and triglycerides for storage.
Salt and Water Transport in the Kidney In a section of the kidney tubule called the loop of Henle, the epithelium uses Na+K+-ATPase pumps and Na+/ K+/2Cl− symporters to reabsorb NaCl that is filtered from blood into the excretory pathway (Fig. 17.3). Without this mechanism, salt would be lost in urine. Tight junctions seal this epithelium, so that salt must pass through the cells to return to the blood. Na+/K+/2Cl− symporters in the apical plasma membrane allow NaCl from the urine to enter the cell down its concentration gradient. Abundant Na+K+-ATPases in the basolateral plasma membrane (5000/µm2) create a Na+ gradient to drive the symporter and to clear the cytoplasm of Na+ accumulated from the tubule lumen. KCl that enters with Na+ through the Na+/K+/2Cl− symporter leaves the cell through channels: K+ channels in apical and basolateral membranes and Cl− channels in basolateral membranes. Furosemide, a drug used to treat congestive heart failure, inhibits the Na+/K+/2Cl− symporter in the loop of Henle. A weak heart leads to accumulation of fluid in the lungs (causing shortness of breath) and other tissues (causing swelling of the ankles). Inhibiting the Na+/
K+/2Cl− symporter reduces NaCl reabsorption, so the kidney produces larger quantities of urine, clearing excess fluid from the body and relieving symptoms.
Cystic Fibrosis as a Transporter Disease Cystic fibrosis results from loss of function mutations of the cystic fibrosis transmembrane regulator (CFTR), an unorthodox ABC transporter that functions as a Cl− channel (see Chapter 9). Patients suffer from lung infections and impaired secretion of digestive enzymes by the pancreas. Understanding the pathology requires knowledge of the mechanisms that produce the fluid layer containing NaCl on the apical surface of the epithelial cells lining the airways of the lung (Fig. 17.4). Water and Cl− move through the cell, while Na+ moves between the cells. The complicated process depends on familiar pumps and carriers. Cl− moves into the base of the cell along with Na+ and K+ through Na+/K+/2Cl− symporters powered by the electrochemical gradient of Na+ created by Na+K+-ATPases in the basolateral membrane. This brings excess potassium chloride into the cell. K+ then
LUMEN
CFTR
Cl– ATP
Cl–
H2O
Apical
– – – – – – – – – – – –– – – – –
LUMEN
NaCl
Leaky tight junction
KCl
– – –– – – – – – –– – Na+ – – – – –
Cl– ATP Cl–
H2O
NaCl KCl Tight junction
Na+ Cl–
Cl–
2 Cl–
K+
3 Na+ ADP ATP
3 Na+
ADP ATP
BASAL
+ + LAMINA + + + + + +++ + + +
BASAL LAMINA
2 K+
Na+
2 K+ 3 Na+
2 Cl– K+
Cl–
Cl–
TISSUE SPACE exchangeable with blood FIGURE 17.3 SODIUM CHLORIDE TRANSPORT BY THE EPITHELIUM OF THE KIDNEY TUBULE. Tight junctions seal the space between these polarized epithelial cells of the thick ascending limb of the loop of Henle. Na+K+-ATPase pumps (space-filling model) in the basolateral plasma membrane drive Na+/K+/2Cl− symporters in the apical plasma membrane. K+ channels in the apical plasma membrane and K+ channels and Cl− channels in the basolateral plasma membrane provide paths for K+ to circulate and for Cl− to follow Na+ across the cell from the lumen of the tubule to the blood compartment.
H2O
+ +++ + ++ +
3 Na+
TISSUE SPACE exchangeable with blood FIGURE 17.4 SALT AND WATER TRANSPORT ACROSS THE EPITHELIUM LINING THE RESPIRATORY TRACT. Leaky tight junctions partially seal the space between these polarized epithelial cells. Na+K+-ATPase pumps in the basolateral plasma membrane drive Na+/K+/2Cl− symporters in the basolateral plasma membrane. CFTR (cystic fibrosis transmembrane regulator) Cl− channels in the apical plasma membrane allow Cl− to move into the lumen, creating a negative electrical potential that pulls Na+ between the cells into the lumen. CFTR also releases ATP, which activates additional Cl− channels. Water follows sodium chloride into the lumen through water channels and between the cells. Basolateral K+ channels allow K+ to circulate.
288
SECTION V n Membrane Structure and Function
recycles out of the cell by way of channels in the basolateral plasma membrane, leaving behind excess Cl− inside the cell. When activated by phosphorylation of its regulatory domain and ATP binding, CFTR in the apical plasma membrane opens a channel for Cl− and bicarbonate. Cl− moves down its electrochemical gradient out of the cell, carrying charge to the outside. The whole epithelium becomes polarized, with the lumen electrically negative relative to the basolateral fluid compartment. This electrical driving force draws Na+ between cells through leaky tight junctions from the extracellular fluid compartment to the surface of the epithelium. Sodium chloride on the apical surface creates an osmotic force that draws water down its concentration gradient across the cells to the lumen through water channels (see Fig. 11.15). CFTR also appears to inhibit transport mechanisms that reabsorb fluid from the lumen of the epithelium. A balance between this fluid secretion and fluid reabsorption normally keeps a layer of water on the surface of the epithelium allowing the cilia to clear secretions and bacteria from the lung. Loss of CFTR function leaves the lungs of cystic fibrosis patients too dry. This situation is life threatening because cilia in the respiratory tract cannot move sticky, dry, mucus containing bacteria and viruses out of the lungs, thereby predisposing to respiratory infections. Sticky secretions in the pancreatic ducts also interfere with the secretion of digestive enzymes by the pancreas. The severity of the disease depends on the particular mutation in the CFTR gene. Symptoms are relatively mild in patients with point mutations that reduce the open probability of the channel. A drug called ivacaftor increases the activity of these mutant channels and relieves many symptoms. The disease is more severe with the most common mutation (67% of cases), deletion of the codon for phenylalanine 508 (F508). The mutant Δ508 protein is temperature-sensitive, not folding
A. Hypertonic medium
–
HCO3
H+
Cl–
Water exits Cell shrinks
properly at 37°C, so it fails to negotiate the secretory pathway to the plasma membrane. Patients with two copies of the Δ508 mutation have classic cystic fibrosis. Heterozygotes with the Δ508 mutation and a normal gene (approximately 5% of humans) have no symptoms. Combining the ΔF508 mutation with one of more than a thousand different mutations in the other copy of the gene causes the typical lung disease and a range of severity in the pancreatic problems. Treating patients with the Δ508 mutation is challenging, but new therapies are being explored, including activation of other plasma membrane Cl− channels.
Transport Mechanisms to Improve Food Production Plants depend on transport mechanisms to take up nutrients, CO2, and water, and to export toxic materials, but their capacities vary widely. Thus some plants are more tolerant of salty conditions or concentrate higher concentrations of micronutrients such as iron. Carriers participate in most of these reactions, so providing food crops with optimal carriers is a promising strategy to improve production. For example, providing plants with carriers to export compounds that neutralize toxic ions such as Al3+ allows them to grow in acidic soils while supplying them with a channel-like Na+ transporter improves salt tolerance. Similarly, providing variants of carriers with a high affinity for iron results in higher concentrations of this micronutrient in rice.
Cellular Volume Regulation Cells employ both short- and long-term strategies involving pumps, carriers, and channels to maintain a constant volume (Fig. 17.5). These compensatory mechanisms are required because water moves across the plasma membrane through water channels and slowly through lipid bilayers if the osmotic strength of the environment
Normal volume
B. Hypotonic medium Water enters Cell swells
Na+ K+ 2Cl+ Na+ H2O
Cl– Taurine
K+ K+ Acute compensation: cell swells
Acute compensation: cell shrinks Normal volume
Cl– H2O
FIGURE 17.5 ACUTE CELLULAR VOLUME CONTROL. A, Cell is placed in hypertonic medium. B, Cell is placed in hypotonic medium. Cells compensate for volume changes by activating channels and carriers to move inorganic ions into or out of the cell. Swelling-sensitive LRRC8 channels also release taurine from the cell. Water follows passively through channels and across the lipid bilayer.
Regulation of membrane potential is particularly important in higher organisms, which use electrical signals generated by membrane channels for communication in their nervous and muscular systems. For example, reading and understanding this page depend on rapid creation and processing of electrical and chemical signals by cells in the visual system and brain. Ion channels produce the key event, a transient change in electrical potential of the plasma membrane, called an action potential. These energy-efficient electrical signals are the fastest means of communication in the body, spreading over the plasma membrane at tens of meters per second. Similarly, action potentials trigger skeletal muscle contraction, control the timing of the heartbeat, and coordinate the peristaltic motions of the gut and contractions of the uterus. Electrical excitability is not limited to nerves and muscles. Eggs use a form of action potential as an early
Description of an Action Potential If a microelectrode (see Fig. 11.16B) drives a small positive or negative current into a cell, a second microelectrode a short distance away detects a small voltage response. These electrotonic potentials decline rapidly with distance if the cell is not excitable. The plasma membrane of an excitable cell, such as neuron or muscle, reacts much differently to depolarization. Rather than responding with a small local current, an excitable cell generates a large, reproducible change in membrane potential, called an action potential when depolarized beyond a certain level (called a threshold) (Fig. 17.6). Voltage-gated ion channels (see Fig. 11.7) generate this powerful electrical signal that spreads rapidly (10 m/sec) over the entire plasma membrane. During an action potential, the membrane potential can reach a peak of +40 to 50 mV before repolarizing to the resting potential. Because action potentials are selftriggering, they travel without dissipation over long
ENa
50
80
Em 60 0
Voltage-gated Na+ channels
Threshold
40
Voltage-gated K+ channels
–50
20
0
EK
Open channels per µm2 of membrane
Excitable Membranes
step in blocking fertilization by more than one sperm. Chemotaxis by macrophages and secretion of insulin and other hormones both depend on electrical excitability. The reader should be familiar with the appendixes in Chapter 11 to appreciate the following material.
Membrane potential (mV)
differs even slightly from that inside the cell. Water moves to maintain an osmotic equilibrium, as is illustrated for red blood cells in Fig. 8.8. In a hypotonic medium, water moves into the cell to dilute the cytoplasm. In a hypertonic medium, water moves out to concentrate the cytoplasm. The mechanisms employed to compensate for these volume changes are well defined, but the mechanisms that sense volume changes and trigger these responses are still being investigated. Animal cells respond acutely to loss of water in a hypertonic environment by activating Na+/H+ antiporters, Cl−/HCO3− antiporters, and/or Na+/K+/2Cl− symporters that bring KCl and NaCl into the cell. Water follows, returning the cell to its original volume in minutes. Swelling in a hypotonic environment quickly activates plasma membrane K+ channels, Cl− channels, and/or a K+/Cl− symporter, releasing potassium chloride, osmolytes, and water from the cell. Osmolytes are small organic molecules, including amino acids and metabolically inactive polyalcohols (sorbitol and inositol) and methylamines. Bestrophins are the Cl− channels activated by swelling in flies (see Fig. 11.13), but vertebrates use LRRC8 channels related to gap junction connexins (see Fig. 31.7) that release the osmolyte taurine along with Cl−. Compensation by moving water along with ions and osmolytes works in the short run, but changes in the internal concentrations of K+, Na+, and Cl− affect the membrane potential and other physiological processes. In the long term, cells maintain their volume by adjusting the osmotic strength of cytoplasm with osmolytes. Adjusting osmolyte concentrations takes longer than ion concentrations, as it requires synthesis or degradation of osmolytes and their transport proteins, such as Na+/ osmolyte symporters.
289
CHAPTER 17 n Membrane Physiology
–100 0
1
Open
2
Time (msec) Closed
3
4
Open K+ leak channels
FIGURE 17.6 THE TIME COURSE OF AN ACTION POTENTIAL PASSING A MEASURING ELECTRODE INSERTED THROUGH THE PLASMA MEMBRANE OF A SQUID GIANT AXON. Spread of an action potential from an adjacent area of the membrane brings the membrane potential Em to threshold, triggering the action potential at this point on the membrane. The other curves show the conductance of the membrane at this point for Na+ and K+ expressed as the concentration of open channels. The lower trace shows the times during which K+ leak channels open and close at this point. ENa is the Na+ equilibrium potential, and EK is the K+ equilibrium potential.
290
SECTION V n Membrane Structure and Function
distances. This high-speed transmission is very efficient, requiring movement of very few ions across the membrane. The molecular events during an action potential were first characterized around 1950 in squid giant axons using microelectrodes coupled to an electronic feedback circuit. This clever “voltage clamp” holds the membrane potential constant by providing the cell with electrical current to compensate for changes in ion currents. Investigators discovered that changes in permeability to Na+ and K+ ions produced action potentials. Changing one variable at a time, they determined the time and voltage dependence of ion-specific conductance. They also determined the relationship between conductance and voltage. From these relationships, measured under controlled conditions, they could calculate the membrane response to virtually any experimental condition. To explain these changes in permeability, they postulated the existence of ion channels. The voltage clamp provided a direct measure of this channel activity.
Three Channels Generating Action Potentials Voltage-gated Na+ and K+ channels open and close in sequence to produce action potentials. Depending on the type of open channel, the membrane potential varies in time between the K+ equilibrium potential (EK) and the Na+ equilibrium potential (ENa) (see Fig. 11.18). Because membrane depolarization activates these ion channels, and because the response spreads this depolarization, triggering an action potential initiates a cascade of reactions that moves over the membrane, first to depolarize and then to repolarize the membrane. In nerves, just three types of voltage-gated channels are required to generate action potentials: • K+-selective leak channels of the Kir family and the TWIK family (see Fig. 11.2) are open at resting potentials. Cytoplasmic Mg2+ blocks Kir channels when the membrane depolarizes. • Voltage-gated Na+ channels are closed at the resting potential but open if the membrane depolarizes to approximately −40 mV. They open only transiently, because a first-order inactivation reaction closes the pore, even if the membrane potential is at or above zero (see Fig. 16.5). These channels return to the closed state without passing through the open state. • Delayed-rectifier voltage-gated K+ channels have a low probability of being open at the resting potential. They respond to membrane depolarization by opening, but more slowly than Na+ channels do. They stay open long enough to allow the repolarizing membrane potential to approach EK. The properties of these channels explain the time course of an action potential as follows: Stage 1: At rest, the membrane is slightly permeable to K+ but not to other ions, so the resting potential
is near EK, approximately −70 mV. K+-selective leak channels and a few open voltage-gated K+ channels contribute to this basal K+ permeability. Stage 2: If the membrane is depolarized by an oncoming action potential and reaches the threshold potential, K+-selective leak channels close and voltage-gated Na+ channels open. Because the membrane is permeable only to Na+ and because many Na+ channels open, Na+ moves into the cell and the membrane potential rapidly approaches ENa, approximately +45 mV. Stage 3: After 1 to 2 msec, Na+ channels spontaneously inactivate and slowly responding delayedrectifier K+ channels open. Now the membrane is strongly and selectively permeable to K+, so K+ moves out of the cell, and the membrane potential reverses all the way to EK, approximately −80 mV. K+ channels are less synchronized than Na+ channels, so the membrane potential falls slower than it rises. Stage 4: Delayed-rectifier K+ channels close progressively as the membrane repolarizes, and K+-selective leak channels open, returning the membrane potential to the resting voltage, just above EK. During an action potential, the membrane voltage changes by 100 to 150 mV in 1 to 2 msec. The membrane bilayer is approximately 7 nm thick, so this voltage corresponds to a field variation on the order of 150,000 volts/cm in 1 to 2 msec. Such strong forces elicit conformational changes in membrane proteins, such as voltagegated ion channels.
Membrane Depolarization: The Stimulus for Action Potentials The initial depolarization of the plasma membrane that triggers an action potential can arise from activation by a neurotransmitter (see “Synaptic Transmission” below) or spreading of an action potential from an adjacent region of the membrane or from an adjacent cell through a gap junction. Membrane depolarization must exceed a certain threshold to trigger an action potential. The threshold arises directly from the properties of the ion channels. Depolarization less than threshold activates a few Na+ channels, producing a small inward Na+ current, but it also activates some delayed-rectifier K+ channels, resulting in K+ efflux. If the Na+ conductance is small in relation to the K+ conductance, outward currents predominate and the membrane repolarizes. Depolarization greater than threshold activates additional Na+ channels, yielding inward Na+ currents greater than outward K+ currents, at least briefly. This positive feedback loop further depolarizes the membrane, amplifying activation of Na+ channels and producing the cascade of channel activation that makes action potentials an all-or-nothing event.
291
CHAPTER 17 n Membrane Physiology
Myelin Sheaths Speed Action Potentials Action potentials naturally spread rapidly over muscle cells and along extensions of neurons called axons, but some axons in the central and peripheral nervous system have insulation that speeds their propagation up to 10-fold. Supporting cells make the insulation by wrapping layers of plasma membrane around the axon to form a myelin sheath (Fig. 17.7). Gaps between these insulated sections expose the plasma membrane of the axon with voltage-gated ion channels. Action potentials jump at high speed from one gap to the next. Activity of the neuron can stimulate supporting cells to increase the thickness and extent of the sheath, which raises the speed of action potentials and contributes to learning certain motor skills.
cases, gap junctions (see Fig. 31.6) connect neurons at “electrical synapses,” where current moves directly between the two cells. All seven neurotransmitters shown in Fig. 17.8 are small organic molecules with an amino group. Secretory
Synaptic Transmission Most neurons use chemical messengers called neurotransmitters (Fig. 17.8) to communicate rapidly with each other and with effector cells, such as skeletal muscle and glands. This chemical communication occurs at sites called synapses (Figs. 17.9 and 17.10), where the sending cell is specialized to secrete a particular neurotransmitter and the receiving cell is specialized to respond to that neurotransmitter. The sending side of a synapse is referred to as presynaptic, whereas the receiving side is designated postsynaptic. Small vesicles containing neurotransmitter pack the presynaptic nerve terminal. Neurotransmitter receptors concentrate in the postsynaptic plasma membrane. Modest changes in either the presynaptic release of neurotransmitter or postsynaptic receptor activation can profoundly influence how a neuron processes this information. Analysis of synaptic transmission has revealed much about the mechanisms of secretion (see Chapter 22), signal transduction, and psychoactive drugs that affect behavior. Not all synapses use chemical transmitters. In special Acetylcholine
Structure
O CH3 C O CH2 CH2 H3C +N CH3 CH3
Receptors
Transmitter
Channels
Excitatory (nicotinic) Na+ / K+ channel
Seven-helix
Muscarinic receptor
Dopamine
HO
+H N 3
OH
CH2 CH2
Dopamine receptor
γ-Aminobutyric acid (GABA) O O– C CH2 CH2 +H N CH 3 2
FIGURE 17.7 MYELIN SHEATH ON COCHLEAR NERVE THAT TRANSMITS IMPULSES FROM THE EAR TO THE BRAIN. Electron micrograph of a thin section showing a Schwann cell and the myelin sheath that is wrapped around the axon. The inset shows a portion of the sheath at a higher magnification. (Courtesy Enrico Mugnaini. From Fawcett DW. The Cell, 2nd ed. Philadelphia: WB Saunders; 1981.)
Glutamate
Glycine
HO
COO– CH2 CH2 +H N 3
O C C – O H
Inhibitory Cl– channel
Excitatory Na+ / K+ channel or Na+ / K+ / Ca2+ channel
β-type GABA receptor
Metabotropic glutamate receptor
Norepinephrine
H +H N 3
O C C – O H
Serotonin
OH
HO CH CH2
+H N 3
NH HO +H N 3
CH2 CH2
Excitatory Na+ / K+ channel
Inhibitory Cl– channel Adrenergic receptor
Serotonin receptor
FIGURE 17.8 NEUROTRANSMITTERS AND THEIR LIGAND-GATED ION CHANNELS AND SEVEN-HELIX RECEPTORS.
292
SECTION V n Membrane Structure and Function
A
B Nerve
Glial cell Mitochondria Nerve ending Neurofilaments Microtubules Synaptic vesicles Basal lamina
Motor end plate
Muscle cell
H+ ACh synthesis
ATP ADP
ATP ADP
ACh
H+
rve Ne t ac
K+
io n
te po
Fusion
al nti
Na+
ATP ADP
Na+ Ca2+
Na+
Ca2+ ACh
K+ K+
Voltagegated
Ligandactivated K+
ACh
Endocytosis
K+ K+
Acetylcholine esterase
K+
Na+ Voltage-gated channels inactive
Na+
ADP ATP
Na+
ACh receptor closes upon ACh dissociation
Na+ Na+ l Muscle ac tion potentia
C. Excitatory transmission
K+
D. Recovery
FIGURE 17.9 NEUROMUSCULAR JUNCTION. A, A scanning electron micrograph of a motor nerve and the skeletal muscle cells that it innervates. B, An electron micrograph of a thin section of a frog neuromuscular junction. C, Excitatory synaptic transmission. The nerve action potential opens voltage-gated calcium channels. Entry of Ca2+ triggers fusion of a synaptic vesicle containing acetylcholine (ACh) with the plasma membrane. ACh binds and opens postsynaptic channels on the muscle cell, which trigger an action potential. D, Recovery includes ACh hydrolysis, recycling of synaptic vesicle membranes, and loading of synaptic vesicles with new ACh. (A, Courtesy Don Fawcett, Harvard Medical School, Boston, MA. B, Courtesy J.E. Heuser, Washington University, St. Louis, MO.)
293
CHAPTER 17 n Membrane Physiology
A
B
Axon hillock
Dendrite Myelin sheath Low Na+channel density
High Na+-channel density
C. Excitatory transmission
ATP
D. Recovery
ADP
PRESYNAPTIC
NEURON
Glut
H+
H+
Fusion
–
Ac tio np K+ otent ATP ADP – +++
+ +
Na+ +
+
–
–
Glut K+
ial
––
Ca2+
+ + +
++
–
– – –
– –
+
Glut
–
–
–
+
+ –
K+ POSTSYNAPTIC NEURON
+– – +
+
+
–––
–
––
– –
++
++
+++
++
+
Na+
K+ +
+ –
–
Na+ Glut stimulates Depolarization receptors to open
–
+
– ADP ATP
Glut Na+
ATP ADP – –
–
+
+
+
+
+ Na+ K
–
ATP
Ca2+
Na+
+
– ADP
H+
H+/Glut antiporter
–
– –
+
–
+
–
–
+
+
+
+
+
–
– +
+
Na+
+
Glut +
+ –
+ –
K+ Repolarization
+ –
+
+
–
–
+
+ –
+
+ –
–
Glut receptors close upon Glut dissociation
FIGURE 17.10 CENTRAL NERVOUS SYSTEM SYNAPSES. A, A neuron with its cell body and dendrites covered with a mixture of excitatory and inhibitory synapses. A high density of voltage-gated Na+ channels in the proximal part of the axon, called the axon hillock, favors the generation of an action potential when the sum of postsynaptic potentials brings the axon hillock to threshold. B, An electron micrograph of a thin section of brain showing synapses with vesicles (green) clustered in the presynaptic axon. The inset shows an anatomically correct molecular model of a synaptic vesicle. C, Synaptic transmission at a central nervous system (CNS) excitatory synapse. A presynaptic action potential opens voltage-gated Ca2+ channels. Entry of Ca2+ stimulates fusion of synaptic vesicles filled with glutamate (Glut) with the plasma membrane. Glutamate binds and opens postsynaptic α-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA) receptors that generate a local postsynaptic potential change. D, Recovery from excitatory stimulation includes retrieval of glutamate by a presynaptic Na+/glutamate symporter and concentration of glutamate in synaptic vesicles by a H+/glutamate antiporter. (B, Courtesy Don Fawcett, Harvard Medical School, Boston, MA. Inset, Modified from Takamori S, Holt M, Stenius K, et al. Molecular anatomy of a trafficking organelle. Cell. 2006;127:831–846.)
mechanisms are similar at all synapses, but each neurotransmitter requires its own biochemical machinery for synthesis, packaging in synaptic vesicles, and reception by postsynaptic cells. Such distinctive features of synapses using a particular transmitter make it possible
to modify synaptic transmission selectively in the clinic, such as in treatment with psychoactive drugs. In addition to activating ligand-gated ion channels, most neurotransmitters also stimulate particular sevenhelix receptors (Fig. 17.8; see also Fig. 24.3). For
294
SECTION V n Membrane Structure and Function
example, acetylcholine stimulates the seven-helix muscarinic acetylcholine receptor, which uses a trimeric G-protein intermediary to activate Kir3.1 K+ channels (Fig. 39.21). Glutamate stimulates seven-helix “metabotropic” receptors, which also act through trimeric G proteins. Disruption of the gene for metabotropic glutamate receptors leaves mice with defects in coordination and learning, and overstimulation of these receptors might contribute to some forms of intellectual disability in humans. The following sections compare synapses at the neuromuscular junction and in central nervous system. Both illustrate how pumps, carriers, and channels work together at synapses.
Neuromuscular Junction Motor neurons in the spinal cord and brainstem control contraction of skeletal muscle cells (see Fig. 39.14). Long axons from these neurons terminate in synapses on skeletal muscle cells, called neuromuscular junctions (Fig. 17.9A–B). Every neuronal action potential that reaches a neuromuscular junction evokes an action potential that spreads over the postsynaptic surface of the muscle cell and initiates contraction. This highly reliable, one-to-one communication depends on chemical transmission by acetylcholine between the nerve and muscle. Highly concentrated nicotinic acetylcholine receptors (see Fig. 11.12) in the postsynaptic membrane (~20,000/µm2) transduce the arrival of extracellular acetylcholine into membrane depolarization. Figure 17.9C illustrates the membrane proteins required for neuromuscular transmission. Both the nerve terminal and muscle depend on Na+K+-ATPase and Ca2+-ATPase pumps to maintain gradients of Na+, K+, and Ca2+ across their plasma membranes. Both presynaptic and postsynaptic cells need voltage-gated Na+ channels and K+ channels for action potentials. Additionally, the presynaptic membrane requires voltage-gated Ca2+ channels to trigger secretion of acetylcholine. A neuronal action potential initiates synaptic transmission by admitting Ca2+ into the presynaptic terminal through voltage-gated Ca2+ channels. In less than 1 msec, Ca2+ triggers fusion of synaptic vesicles containing acetylcholine with the plasma membrane. Within microseconds, acetylcholine released into the synaptic cleft between the cells reaches millimolar concentrations and binds postsynaptic acetylcholine receptors. Weak binding of acetylcholine to two subunits of the acetylcholine receptor (see Fig. 11.12) opens a nonselective cation channel. The open pore is about equally permeable to K+ and Na+ and less permeable to Ca2+, so the membrane potential collapses toward a reversal potential (see Fig. 16.19) of approximately 0 mV. This is above the threshold for triggering a self-propagating action potential in the muscle plasma membrane, which occurs with nearly 100% efficiency. The action potential
traveling over the muscle plasma membrane activates voltage-gated Ca2+ channels that trigger Ca2+ release from smooth endoplasmic reticulum, resulting in muscle contraction (see Fig. 39.15). Two different mechanisms terminate activation of acetylcholine receptors. An extracellular enzyme, acetylcholinesterase, degrades free acetylcholine in the synaptic cleft in a few milliseconds. In parallel, acetylcholine receptors automatically undergo a conformational change that increases the affinity for bound acetylcholine and closes the channel. Acetylcholine then dissociates slowly from these desensitized receptors, which return to the resting state. Nerve terminals retrieve synaptic vesicle membrane by pinching off some transiently opened vesicles and by endocytosis (see Chapter 22). Cytoplasmic enzymes synthesize new acetylcholine. A V-type ATPase proton pump (see Fig. 9.6) acidifies the lumen of synaptic vesicles, providing an electrochemical potential to drive an acetylcholine/H+ antiporter that concentrates acetylcholine in vesicles.
Central Nervous System Synapses Each of the approximately 100 billion (1011) neurons in the human brain receives synaptic inputs from many other neurons, forming a grand total of about 1015 synapses. Inputs at synapses covering the surfaces of the dendrites and the cell body (Fig. 17.10A) drive local changes in the membrane potential that are integrated at the base of the axon to start an action potential. Some synapses excite the postsynaptic cell, while others inhibit. Furthermore, most neurotransmitters (Fig. 17.8) activate both channels, producing point-to-point signals on a fast time scale, as well as seven-helix receptors that modulate the behavior of other neurons on longer time scales. In addition, many central nervous system (CNS) synapses secrete both a neurotransmitter and one of more than 100 neuropeptide hormones for communication through seven-helix receptors on neighboring cells. Finally, some neurons can secrete two neurotransmitters or switch their neurotransmitters during development. In all these ways synaptic transmission between neurons in the CNS (Fig. 17.10) differs fundamentally from the stable, efficient, one-to-one, excitatory coupling at neuromuscular junctions. Transmission at chemical synapses in the CNS depends on cooperation of pumps, carriers, and channels (Fig. 17.10C–D) similar to the neuromuscular junction. Incoming information takes the form of action potentials that arrive at synapses and open voltage-gated Ca2+ channels in the presynaptic membrane. The transient rise in cytoplasmic Ca2+ can trigger the fusion of synaptic vesicles with the presynaptic plasma membrane, releasing transmitter, but the probability of success ful fusion is lower than at neuromuscular junctions. Neurotransmitters secreted by the presynaptic cell
activate ligand-gated channels that control the postsynaptic membrane potential. Carriers in the presynaptic membrane and adjacent supporting cells terminate transmission by removing neurotransmitter from the synaptic cleft. Excitatory synapses using the neurotransmitters glutamate, acetylcholine, or serotonin activate receptor channels (see Figs. 16.11 and 16.12) permeable to cations that depolarize the membrane. Opening these channels causes a local, short-lived change in membrane potential, called a postsynaptic potential (PSP). Individual PSPs do not fire action potentials for two reasons. First, individual PSPs raise the membrane potential only a few millivolts, so they do not bring the postsynaptic membrane to threshold. Second, the plasma membrane of dendrites and the cell body contains few voltage-gated Na+ channels. Furthermore, inhibitory synapses on the same cell counteract excitatory synapses by secreting glycine or γ-aminobutyric acid (GABA) to activate Cl− channels that hyperpolarize the membrane, taking it farther from threshold (see Fig. 11.18). Neurons spatially average excitatory and inhibitory PSPs as they spread passively over the postsynaptic membrane and generate an action potential only when their sum at a particular time brings the membrane potential to threshold in the proximal part of the axon, called the axon hillock (Fig. 17.10A). This part of the plasma membrane is particularly sensitive to voltage, owing to a high concentration of voltage-gated Na+ channels. At threshold, they open, depolarizing the membrane (Fig. 17.6). Delayed-rectifier K+ channels then repolarize the membrane, resetting it in preparation for subsequent action potentials. The frequency of postsynaptic action potentials is proportional to the intensity of the presynaptic input above a threshold. Each action potential is identical and propagates down the axon. Both the pattern and frequency of action potentials carry information in the brain. Removal of neurotransmitters from the synaptic cleft terminates activation of postsynaptic receptors (Fig. 17.9D). The Na+ gradient across the plasma membrane drives symporters that return neurotransmitters to their presynaptic cells. Within the presynaptic cell, a V-type, proton-translocating ATPase acidifies the lumen of the synaptic vesicle and establishes a proton electrochemical gradient across the vesicle membrane to drive the antiporter that concentrates transmitter inside the vesicle.
Modification of Central Nervous System Synapses by Drugs and Disease The duration of synaptic stimulation depends on the rate of clearance of neurotransmitters from the synap tic cleft, so inhibiting these transport processes with drugs prolongs stimulation at particular classes of CNS synapses, with profound effects on brain function and
CHAPTER 17 n Membrane Physiology
295
behavior. Cocaine inhibits a plasma membrane dopamine transporter as well as transporters for serotonin and norepinephrine. Tricyclic antidepressants inhibit norepinephrine uptake, and other drugs inhibit serotonin uptake. These drugs have dramatic effects on the symptoms of depression as well as other psychiatric disorders. Excess stimulation of N-methyl-D-aspartate (NMDA) receptors rapidly kills postsynaptic neurons, most likely owing to the deleterious effects of excess cytoplasmic Ca2+. This occurs when glutamate is released from ischemic brain tissue during a stroke caused by compromising the blood supply to a region of the brain. Such damage might also contribute to neuron death in degenerative diseases of the nervous system, such as amyotrophic lateral sclerosis and Alzheimer disease. Nicotinic acetylcholine receptors in the CNS are found on both the postsynaptic and presynaptic membranes along with glial cells, so the effects of acetylcholine secreted by neurons and nicotine from tobacco are widespread. In some cases they carry out fast, excitatory synaptic transmission as at the neuromuscular junction. Nicotinic acetylcholine receptors in the presynaptic plasma membrane are highly permeable to Ca2+, so their stimulation admits Ca2+ into the presynaptic terminal. This enhances both the spontaneous release of neurotransmitter and release in response to action potentials. The isoform composition of CNS acetylcholine receptors differs from that of muscles (see Fig. 11.12). Some are homopentamers of α-subunits. Others are heteropentamers of α- and β-subunits. Activation of these ligand-gated channels in different regions of the brain may account for the enhancing effects of nicotine on learning and memory but also for tobacco addiction. Loss of CNS neurons that secrete acetylcholine might contribute to dementia in Alzheimer disease.
Modification of Central Nervous System Synapses by Use Memories are thought to depend on structural changes that modify the strength or numbers of synapses between neurons in the brain, processes called synaptic plasticity. Particular patterns of stimulation can produce longterm changes that enhance or reduce the efficiency of transmission at various glutamate-mediated synapses (Fig. 17.11). The hippocampus, a region of the vertebrate cerebral cortex known to participate in some forms of learning and memory, is often used for observing a simple form of cellular learning. Intense stimulation of excitatory glutamate synapses (20 pulses over a period of 200 msec) can increase synaptic strength for days or weeks. This is called long-term potentiation (LTP). Conversely, slow, prolonged stimulation of glutamate synapses reduces the response for hours. This is called long-term depression (LTD). The mechanisms of LTP and LTD have been investigated thoroughly, because
296
SECTION V n Membrane Structure and Function
A. Before LTP
B. LTP Induction
Probability of vesicle release low, 80%
NMDA receptors open
Postsynaptic responsiveness high
Ca2+ enters
Increase in AMPA active receptors
Ca2+ binds calmodulin Stimulates CAM kinase II
Increase in AMPA receptor conductance Increase in spines
AMPA receptor phosphorylation DENDRITE
Na+ Active AMPA-R Inactive AMPA-R NMDA-R Small postsynaptic potential change Output Dendritic spine
Action potential
Synaptic vesicles containing glutamate
Input Rapid-fire action potentials
Action potential
AXON Input
Lots of glutamate Na+ NO? Ca2+
Na+ Na+ Active AMPA-R Na+
New dendritic spine
Na+ Na+ Calmodulin
CAM Kinase II
Na+ Na+
Large postsynaptic potential change Output
FIGURE 17.11 LONG-TERM POTENTIATION OF SYNAPTIC TRANSMISSION AT EXCITATORY SYNAPSES IN THE HIPPOCAMPUS. A, Prior to long-term potential (LTP), postsynaptic responses to presynaptic action potentials are unreliable and small. B, Some acute responses to vigorous stimulation. C, After induction of LTP, postsynaptic responses are more reliable and larger. AMPA-R (α-amino-3-hydroxy-5methylisoxazole-4-propionate receptor) and NMDA-R (N-methyl-D-aspartate receptor) are two classes of glutamate receptors; NO is nitric oxide, a candidate for the retrograde signaling molecule; CAM kinase II is calcium-calmodulin kinase II.
they occur on appropriate time scales to modify synapses during development and high-order brain functions, such as learning and memory. Induction of LTP involves two types of glutamate receptor channels (see Fig. 11.11) located in postsynaptic specializations called dendritic spines (Fig. 17.11). AMPA (α-amino-3-hydroxy-5-methylisoxazole-4propionate) receptors open and close rapidly in response to glutamate and depolarize the membrane by admitting Na+. NMDA receptors respond slowly to glutamate and also depolarize the membrane by admitting Ca2+. The response of NMDA receptors to glutamate depends on the membrane potential, as partial depolarization is required to displace an extracellular Mg2+ ion blocking the channel. This dual dependence on glutamate and membrane potential makes NMDA receptors coincidence detectors, responsive to rapid stimulation or stimulation at nearby excitatory synapses. In principle, LTP and LTD might alter the efficiency of synaptic transmission by changing glutamate release from the presynaptic cell or the responsiveness of the
postsynaptic cell to glutamate. Investigators debate the relative importance of presynaptic and postsynaptic processes, but both likely contribute. Inconsistencies in observations likely reflect the fact that LTP involves multiple processes (Fig. 17.11), and the protocols used to study LTP can emphasize one feature over others. One presynaptic factor is the probability that an action potential will stimulate the fusion of a glutamatecontaining synaptic vesicle with the plasma membrane. In the resting state, glutamate release by these vesicles is unreliable. LTP increases the probability of exocytosis from less than 0.5 to greater than 0.8. The mechanism of enhanced transmitter release is still being investigated. Candidates include NMDA receptors in the presynaptic membrane, diffusive messengers such as nitric oxide (see Fig. 26.17) providing feedback from the postsynaptic side, and strengthened adhesion between the preand postsynaptic membranes. At least two factors on the postsynaptic side influence the response to glutamate: the number of active AMPA receptors at a synapse; and the number and sizes of
synapses with the stimulating axon. Inactive synapses with only NMDA receptors do not respond to weak stimulation with glutamate, but intense presynaptic release of glutamate during LTP arouses such silent synapses, especially if coordinated with membrane depolarization from neighboring synapses. Ca2+ enters the postsynaptic dendritic spine through active NMDA receptors, binds calmodulin (see Fig. 3.12C) and activates the processes that initiate and maintain LTP. Within seconds, calcium-calmodulin activates calciumcalmodulin (CAM)-kinase II (see Fig. 25.4A), which phosphorylates AMPA receptors. Phosphorylated AMPA receptors are more responsive to glutamate. In addition LTP shifts AMPA receptors from an intracellular pool of recycling endosomes to the postsynaptic membrane where they are anchored by proteins in the postsynaptic density on the inside of the plasma membrane. Within minutes, induction of LTP triggers signaling cascades that maintain the increased efficacy, leading to structural changes and increased protein synthesis. These changes may induce dendrites to stabilize existing spines or sprout new filopodia and spines that increase the number of synapses within an hour or so. Extension of these processes and remodeling of the shape of dendritic spines depend on actin filament assembly (see Fig. 38.7). Direct imaging in the brain of live mice has documented the formation of new spines associated with learning a motor skill, a change enhanced by sleep. Over the longer term, the postsynaptic cell initiates gene transcription and protein synthesis, bringing about further changes that stabilize enhanced synaptic transmission. LTD appears, in many ways, to be the reverse of LTP with multiple factors contributing. Small molecules released from the dendritic spine activate seven-helix cannabinoid receptors on the presynaptic membrane and reduce the reliability of exocytosis. In addition, postsynaptic AMPA receptors are less responsive owing to sequestration in internal compartments. Armed with the growing body of knowledge about synaptic plasticity at the cellular level, many
CHAPTER 17 n Membrane Physiology
297
neuroscientists are investigating the short- and long-term changes in the brain that account for learning and memory. ACKNOWLEDGMENTS We thank Pietro De Camilli for his suggestions on revisions to this chapter. SELECTED READINGS Birren SJ, Marder E. Plasticity in the neurotransmitter repertoire. Science. 2013;340:436-437. Dineley KT, Pandya AA, Yakel JL. Nicotinic ACh receptors as therapeutic targets in CNS disorders. Trends Pharmacol Sci. 2015;36: 96-108. Dorwart M, Thibodeau P, Thomas P. Cystic fibrosis: Recent structural insights. J Cyst Fibros. 2004;3:91-94. Feldman DE. Synaptic mechanisms for plasticity in the neocortex. Annu Rev Neurosci. 2009;32:33-55. Hoffmann EK, Lambert IH, Pedersen SF. Physiology of cell volume regulation in vertebrates. Physiol Rev. 2009;89:193-277. Hogg RC, Bertrand D. What genes tell us about nicotine addiction. Science. 2004;306:983-984. Keating MT, Sanguinetti MC. Molecular and cellular mechanisms of cardiac arrhythmias. Cell. 2001;104:569-580. Mall MA, Galietta LJ. Targeting ion channels in cystic fibrosis. J Cyst Fibros. 2015;14:561-570. Park P, Volianskis A, Sanderson TM, et al. NMDA receptor-dependent long-term potentiation comprises a family of temporally overlapping forms of synaptic plasticity that are induced by different patterns of stimulation. Philos Trans R Soc Lond B Biol Sci. 2013;369:20130131. Qiu Z, Dubin AE, Mathur J, et al. SWELL1, a plasma membrane protein, is an essential component of volume-regulated anion channel. Cell. 2014;157:447-458. Riordan JR. CFTR function and prospects for therapy. Annu Rev Biochem. 2008;77:701-726. Strange K. Cellular volume homeostasis. Adv Physiol Educ. 2004;28: 155-159. Vincent GM. The long QT syndrome: Bedside to bench to bedside. N Engl J Med. 2003;348:1837-1838. Voss FK, Ullrich F, Münch J, et al. Identification of LRRC8 heteromers as an essential component of the volume-regulated anion channel VRAC. Science. 2014;344:634-638. Yang G, Lai CS, Cichon J, et al. Sleep promotes branch-specific for mation of dendritic spines after learning. Science. 2014;344: 1173-1178.
This page intentionally left blank
SECTION
Cellular Organelles and Membrane Trafficking
VI
This page intentionally left blank
SECTION VI OVERVIEW E
ukaryotic cells evolved membrane-bounded compartments specialized to provide energy; to synthesize lipids, carbohydrates, proteins, and nucleic acids; and to degrade cellular constituents. These subcellular compartments, called organelles, have distinctive chemical compositions. Organelles vary in abundance and size in different cell types, including multicellular organisms, where each tissue and organ has specialized functions. An organelle often holds a monopoly on performing a given task; for example, endoplasmic reticulum (ER) synthesizes membrane proteins and certain membrane lipids, lysosomes contain enzymes to degrade many macromolecules, and mitochondria convert energy derived from the covalent bonds of nutrients into adenosine triphosphate (ATP) to provide energy for diverse cellular functions. A semipermeable membrane surrounds each organelle and establishes an internal microenvironment with concentrated enzymes, cofactors, and substrates to favor particular macromolecular interactions. Pumps (Chapter 14), carriers (Chapter 15), and channels
(Chapter 16) in each organelle membrane establish an internal chemical environment (pH, divalent cation concentration, reduction–oxidation [redox] potential) that is appropriate for particular biochemical functions. Mitochondria and chloroplasts use many enzymes embedded in their membranes to catalyze reactions that depend on the separation of reactants across the membrane or involve hydrophobic substrates and products soluble in the lipid bilayer (Chapter 19). Compartments also protect the rest of the cell from potentially dangerous activities, such as degradative enzymes in lysosomes and oxidative enzymes in peroxisomes. This division of labor among organelles has many advantages but also presents cells with challenges in terms of coordination of cellular activities, organelle biosynthesis, and cell division. Organelles are not autonomous, so their activities must be integrated to benefit the whole cell. Therefore, mechanisms are required to transport material between compartments and across the membranes that surround them. Many functional pathways require macromolecules and lipids to move
Translated polypeptide chains
Protein import Ch 18
Mitochondria and chloroplasts Ch 19
Endocytic pathway Ch 22
Secretory pathway import Ch 21
Endoplasmic reticulum Ch 20
Degradation Ch 23
301
from one organelle to another in a vectorial manner. This transport between organelles generally involves budding of vesicles from one membrane-bounded compartment followed by fusion with another, in a process collectively termed vesicular trafficking. In addition, contact sites between some membranes allow for exchange of lipids. This section focuses on two important processes as they pertain to the biogenesis and functions of the various organelles: the targeting of proteins, either during or after translation to their home organelle, and the bidirectional movement of vesicular traffic between organelles and the plasma membrane. The exocytic or secretory pathway from the endoplasmic reticulum to the plasma membrane and lysosomes coordinates organelle biosynthesis and secretion. The endocytic pathway takes in molecules and microscopic particles from outside the cell along with plasma membrane components. Operating together, the two pathways coordinate the distribution and turnover of membrane proteins and lipids. Proteins that are synthesized in the cytoplasm either remain there or move to their final destinations in the nucleus (Chapter 9), mitochondria, chloroplasts, and peroxisomes (Chapter 18). Hundreds of proteins destined for mitochondria and chloroplasts are synthesized in the cytoplasm and directed to these organelles by zip codes built into their polypeptide sequences. Usually, these guide sequences are removed by proteolytic processing once the polypeptide has moved through channels into one of the membranes or compartments inside these organelles. Different sorts of targeting sequences target dozens of proteins to peroxisomes. Chapter 19 explains how mitochondria and chloroplasts descended from bacteria that established symbiotic relationships with eukaryotes in two singular events about a billion years apart. Mitochondria brought along the capacity for ATP synthesis by oxidative phosphorylation, while chloroplasts contributed photosynthesis and oxygen production. Peroxisomes are derived from the ER by a process that is distinct from the secretory pathway. They carry out a number of oxidative reactions. The ER (Chapter 20) generates the secretory path way by synthesizing proteins for membranes and for secretion as well as many of the lipids that are used in membranes throughout the cell. Amino acid sequences called signal sequences direct ribosomes that synthesize integral membrane proteins and secreted proteins to receptors on the endoplasmic reticulum. Translation pushes these polypeptides through a protein pore into the lumen of the ER or into the lipid bilayer. After folding and modification by addition of oligosaccharides, these proteins exit from the ER in vesicles for transport to the Golgi apparatus and more distal parts of the secretory pathway, including lysosomes and plasma membrane. Retrograde vesicle traffic mediated by other proteins 302
retrieves membranes and proteins from the Golgi apparatus back to the ER. Despite this heavy bidirectional traffic between organelles, accurate sorting mechanisms allow each organelle to maintain its identity. Chapter 21 explains the mechanisms used for membrane trafficking. Cells use three different types of coat proteins with common evolutionary origins for budding membrane vesicles. Under the direction of membraneassociated guanosine triphosphatases (GTPases), these proteins form a coat on a donor membrane that distorts the lipid bilayer into a vesicle that buds from the surface. The vesicle carries membrane proteins and lipids and any material in the lumen. Sorting signals direct some proteins into these transport vesicles. After the vesicle moves by diffusion or by active transport along the cytoskeleton to a target membrane, different GTPases and peripheral proteins facilitate fusion of the vesicle with a target membrane. Cells employ at least five distinct mechanisms to internalize plasma membrane, along with a wide range of extracellular materials (Chapter 22). Ingestion of small particles, including bacteria, takes place by phagocytosis, in which a veil of plasma membrane surrounds the particle and takes it into a vacuole inside the cell. Fusion of vesicles containing lysosomal enzymes initiates the degradation of the contents. A second endocytic pathway takes receptors and their ligands into cells in small vesicles coated with clathrin. Other forms of endocytosis take up extracellular fluid and patches of plasma membrane enriched in cholesterol, sphingolipids, and certain signaling proteins. Inside the cell, the contents and membranes of these various endocytic vesicles are sorted in endosomes and then directed in vesicles back to the plasma membrane or onward to the Golgi apparatus or lysosomes for recycling. DNA is stable, but cells continuously replace most of their other constituents in a cycle of synthesis and degradation. Chapter 23 explains how cells degrade proteins and lipids, taken in from outside by endocytosis or from inside the cell. Each type of RNA, protein, and lipid has a natural lifetime, generally much shorter than that of the cell itself. Proteins are degraded and replaced, some every hour, others every day, and some every few weeks or months. Membrane lipids also turn over; some with lifetimes measured in minutes. Proteins and lipids taken in by endocytosis are degraded in lysosomes. In the process called autophagy, a double membrane surrounds a zone of cytoplasm, that can include entire organelles. Fusion of late endosomes and lysosomes with these autophagic vacuoles delivers enzymes that degrade the contents. A large protein complex called the proteasome degrades cytoplasmic and nuclear proteins, but only after they are marked for degradation by conjugation with the small protein, ubiquitin. A hierarchy of ubiquitin-conjugating enzymes controls the fate of proteins as they turn over during the cell cycle.
CHAPTER
18
Posttranslational Targeting of Proteins P
rotein synthesis is largely carried out by cytoplasmic ribosomes that provide all the proteins for the nucleus, cytoplasm, peroxisomes, and secretory pathway. Mitochondria and chloroplasts import most of their proteins from the cytoplasm, even though they originated as bacterial endosymbionts and have retained the capacity to synthesize a few of their proteins. Most of the original bacterial genes moved to the nucleus of the eukaryotic host. Given a common site of synthesis, accurate addressing is essential to direct proteins to their sites of action and to maintain the unique character of each cellular compartment. This is achieved by “zip codes” built into the structure of each protein (Fig. 18.1). Residues in the sequence of each protein—often, but not
Proteins synthesized on free ribosomes
necessarily, contiguous amino acids—form a signal for targeting. Targeting signals are both necessary and sufficient to guide proteins to their final destinations. Transplantation of a targeting signal, such as a presequence from a mitochondrial protein, to a cytoplasmic protein reroutes the hybrid protein into the organelle specified by the targeting sequence, mitochondria in this example. Some targeting signals are transient parts of the protein. For example, most mitochondrial proteins are synthesized with N-terminal extensions that guide them to mitochondria and then are removed. Alternatively, signals may be a permanent part of the mature protein, in some cases serving repeatedly to target a mobile protein between different destinations. Permanent nuclear targeting
Chloroplast Transit sequences are cleaved before folding in stroma or secondary targeting signal directs protein to final location STROMA
Polypeptide synthesized with chloroplast transit sequence Polypeptide synthesized with chloroplast transit sequence and secondary targeting sequence
Presequence is cleaved off and protein folds in target compartment MATRIX
Polypeptide synthesized with mitochondrial presequence Polypeptide synthesized with C-terminal PTS signal
Mitochondria
Peroxisome MATRIX
FIGURE 18.1 TARGETING SIGNALS DIRECT POLYPEPTIDES SYNTHESIZED ON CYTOPLASMIC RIBOSOMES TO CHLOROPLASTS, MITOCHONDRIA, AND PEROXISOMES. Signal peptidases remove some signals after the polypeptide enters the organelle.
303
304
SECTION VI n Cellular Organelles and Membrane Trafficking
signals can be located at the N-terminus, the C-terminus, or even the middle of a protein (see Chapter 9). Some proteins have more than one targeting signal: a primary code that directs the protein to the target organelle or pathway, and a second signal that steers the protein to its specific site of residence within the organelle or pathway. Targeting signals direct proteins to their destination by binding to organelle-specific receptors or using soluble escort factors as intermediaries. When necessary, proteins cross membranes via channels called translocons formed by integral membrane proteins (Fig. 18.2). Like ion channels (see Chapter 16), these proteintranslocating channels are gated to prevent indiscriminate transport of cellular constituents when not occupied by a polypeptide. Polypeptides fit so tightly in these channels during translocation that ions do not leak through. Ions traverse ion channels in a microsecond, whereas polypeptides take tens of seconds to move through translocons. Protein synthesis, adenosine triphosphate (ATP) hydrolysis, or the membrane potential provides the energy to power protein translocation across membranes. Three families of protein translocation channels are found in all three domains of life. Sec translocons direct proteins into the endoplasmic reticulum in eukaryotes and out of prokaryotes. The Tat family of pores translocates folded proteins into chloroplast thylakoids and out of prokaryotes. Membrane proteins related to Oxa1p help insert proteins synthesized in the mitochondrial matrix and prokaryotic cytoplasm into membranes. Mitochondria (Fig. 18.4), chloroplasts (Fig. 18.6), and
A. Protein export from bacteria
prokaryotes (Fig. 18.10) have additional families of protein translocation channels. Primary targeting can occur either cotranslationally, coincident with protein synthesis, or posttranslationally, after polypeptide synthesis. Chapter 20 covers protein targeting to the endoplasmic reticulum (ER) where, with a few exceptions, targeting is cotranslational. This chapter covers posttranslational targeting mechanisms that move proteins across membrane bilayers into mitochondria, chloroplasts, and peroxisomes, and out of Bacteria. Eukaryotes also secrete a few proteins directly across the plasma membrane. Chapter 9 covers posttranslational movements of proteins into and out of the nucleus through a large aqueous channel in the nuclear pore.
Transport of Proteins Into Mitochondria Mitochondrial outer and inner membranes define two spaces: one between the outer and inner membranes (termed the intermembrane space) and an interior space termed the matrix (Fig. 18.3). Each membrane and space has distinct functions and protein compositions, which are discussed in Chapter 19. Targeting signals and specific translocation machinery guide more than 1000 imported proteins selectively to these compartments. Genetic and biochemical experiments on fungi defined the molecular machinery for proteins to enter mitochondria. The TOM40 complex (translocase of the outer mitochondrial membrane) is the main portal into the mitochondrion. Thereafter proteins take one of four routes to different compartments: the outer membrane
B. Protein transport in eukaryotes
Omp85 Proteins synthesized on free ribosomes are released into the cytoplasm
Tat Sec
Chloroplast Toc
DNA
IM CYTOPLASM
Sec translocon translocates Sec proteins synthesized on the rough ER into the membrane or lumen
OM
PERIPLASMIC SPACE
Tic Toc and Tic translocons translocate proteins into the membranes or stroma Sec of chloroplasts PEX proteins insert proteins SAM into the membrane or lumen of peroxisimes Tom Mitochondria Tim PEX MATRIX MATRIX OM Peroxisome IM
Sec and Tat translocons translocate proteins into the thylakoid membrane or lumen STROMA
Tat Thylakoid membrane OM Tom and Tim translocons translocate proteins into the membranes or matrix of mitochondria
IM
FIGURE 18.2 TRANSLOCONS USED BY POLYPEPTIDES TO CROSS MEMBRANES. A, Bacterium with Sec and Tat translocons in the inner membrane and Omp85 in the outer membrane. B, Eukaryote translocons including Sec in the endoplasmic reticulum and thylakoid membrane of chloroplasts, Toc in the outer membrane of chloroplasts, Tic in the inner membrane of chloroplasts, Tat in the thylakoid membrane of chloroplasts, Tom and SAM in the outer membrane of mitochondria, Tim in the inner membrane of mitochondria, and PEX in peroxisomes.
305
CHAPTER 18 n Posttranslational Targeting of Proteins
using the SAM complex (sorting and assembly machinery of the outer membrane); the intermembrane space; the inner membrane via the TIM22 complex (translocase of the inner mitochondrial membrane); and the matrix using the TIM23 complex. Targeting signals direct proteins to the TOM40 complex and then on to other locations. Translocation requires energy and assistance from protein chaperones both outside and inside mitochondria.
a component of the electron transport chain in the intermembrane space (see Fig. 19.5), also has an internal signal for import into mitochondria. A succession of weak interactions with outer membrane receptors Tom20, Tom22, and Tom70 guides presequences and other target signals to the outer membrane translocon. The presequence initially contacts Tom20. Eight residues of the presequence fold into an amphipathic (hydrophobic on one side, hydrophilic on the other) α-helix that binds in a shallow hydrophobic groove on Tom20 with favorable electrostatic interactions (Fig. 18.4D). Other parts of the presequence are thought to interact with Tom40, the translocon itself. Although these associations are weak, collectively, they distinguish mitochondrial presequences from other proteins in the cytoplasm with high fidelity.
Delivery of Protein to Mitochondria After synthesis by cytoplasmic ribosomes, most proteins destined for mitochondria bind cytosolic chaperones of the Hsp70 (heat shock protein 70) family (see Fig. 12.12). This interaction maintains proteins in unfolded configurations competent for import (Fig. 18.3). Some imported proteins require additional factors for targeting to the translocation machinery. Targeting motifs for matrix proteins are called presequences, because they are usually removed by proteolytic cleavage in the mitochondrial matrix. Presequences are generally located at the N-termini of precursor polypeptides as contiguous sequences of 10 to 70 amino acids. They are rich in basic, hydroxylated, and hydrophobic amino acids, but share no defined sequences in common. The targeting sequences of many mitochondrial membrane proteins are in the middle of the polypeptide and are not cleaved after import. Cytochrome c,
A. Common pathway Hsp70
CYTOPLASM
Tom– – – 20– + –– +
OUTER
Translocation Across the Outer Membrane Tom40, a β-barrel protein similar to a porin (see Fig. 13.9C), forms the translocon of the outer membrane. Six proteins with single transmembrane helices associate around the periphery of Tom40: the three receptor subunits and three small subunits. These Tom40 complexes form dimers or trimers in the plane of the membrane (Fig. 18.3E) with pores approximately 2 nm in diameter. The outer membrane receptors transfer presequences and other targeting sequences to the translocon channel. Chemical crosslinking showed that both hydrophobic
B. Route to stroma
C. Route to intermembrane space C
C
Presequence bound to Tom20 Tom 70
Tom 40
MEMBRANE
Tom22
C
INTERMEMBRANE
SPACE
INNER MEMBRANE
+
+
+
+
+
+
∆ψ – –
–
– –
–
+
– –
+
+
+
–
Tim 23
Tim 17
+
+
+
+
+
+
∆ψ – –
–
+
+
+
+
+
∆ψ – –
–
Tim 44
MATRIX Signal peptidase
– –
MPP
+
+
+
+
+
+
∆ψ
–
– –
–
– –
–
+
+
+
∆ψ – –
–
Tim 44
New Hsp70
– –
–
MPP
Hsp 70
ATP ADP N
Cleaved presequence
ATP ADP Cleaved presequence
FIGURE 18.3 IMPORT OF MATRIX PROTEINS INTO MITOCHONDRIA. Models of the Tom and Tim complexes. A, Common pathway across the outer membrane. Hsp70 (heat shock protein 70) escorts polypeptides synthesized on cytoplasmic ribosomes to mitochondria where the presequence associates with Tom20/22. The basic presequence leads the polypeptide through Tom40, the translocase of the outer membrane, to the intermembrane space. B, Route to the stroma. The presequence enters the translocase of the inner membrane (Tim). The potential across the inner membrane (Δψ) pulls the presequence through Tim into the matrix, where it is cleaved by MPP (matrix processing protease). Cycles of Hsp70 binding to the peptide followed by adenosine triphosphate (ATP) hydrolysis and dissociation of Hsp70 from Tim44 ratchet the translocating peptide into the matrix, where it folds. C, Route to the intermembrane space. After cleavage of the presequence, the polypeptide backs up into the intermembrane space.
306
SECTION VI n Cellular Organelles and Membrane Trafficking
A. Mitochondrion CYTOPLASM Outer membrane
B. Tom20 Ser16'
C
Tom20
Gln75
Glu78
OUTER MEMBRANE
Tom 40
N Presequence
Tom22
INTERMEMBRANE
Intermembrane space MATRIX
Tom6TM
SPACE
Arg17'
C. Tom40 dimer
E. Model of Tom complex trimer
Tom 70
Arg14' Glu79
Inner membrane
CYTOPLASM
D. Model
Tim 23 Tim 44
Tim 17
INNER MEMBRANE
MATRIX
Tom5TM
FIGURE 18.4 MITOCHONDRIAL IMPORT COMPONENTS. A, Electron micrograph of a thin section of a mitochondrion. B, Structure determined by nuclear magnetic resonance spectroscopy of a presequence peptide bound to a hydrophobic patch on Tom20, a receptor from the mitochondrial outer membrane. Space-filling model of a cytoplasmic domain of Tom20. The presequence forms two turns of α-helix with two arginines exposed on the surface. N is the N-terminus and C is the C-terminus of the peptide. Yellow is a hydrophobic patch; orange is Gln-rich; red is Glu-rich. C, Three-dimensional reconstruction from electron micrographs of a dimer of Tom40, the translocase of the outer mitochondrial membrane. D, Schematic of the mitochondrial import apparatus, including Tom complex in the outer membrane and Tim complex in the inner membrane. E, Model of the active trimer of Tom complex. This face view shows cross sections of the transmembrane helices of four accessory subunits. (A, Courtesy Don W. Fawcett, Harvard Medical School, Boston, MA. B, Courtesy D. Kohda, Kyushu University. From Abe Y, Shodai T, Muto T, et al. Structural basis of presequence recognition by the mitochondrial protein import receptor Tom20. Cell. 2000;100:551–560 and Protein Data Bank [PDB; www.rcsb.org] file 1OM2. C, From Ahting U, Thun C, Hegerl R, et al. The Tom core complex: the general protein import pore of the outer membrane of mitochondria. J Cell Biol. 1999;147:959–968, copyright The Rockefeller University Press. E, From Shiota T, Imai K, Qiu J, et al. Molecular architecture of the active mitochondrial protein gate. Science. 2015;349:1544–1548.)
and charged polypeptides move through the central pore of the β-barrel. Proteins must be largely unfolded to fit through the 2-nm pore. Like other translocons Tom channels are likely to be gated, and they close when not occupied by a translocating polypeptide.
Assembly of Outer Membrane Proteins Some simple outer membrane proteins transfer laterally into the bilayer while they are in transit through Tom, while more complicated outer membrane β-barrel proteins, including Tom40 itself, require assistance. This is provided by another protein complex in the outer membrane called the SAM complex. The β-barrel protein SAM50 forms a channel and cooperates with a second subunit to mediate folding and insertion of β-barrel proteins into the membrane, similar to its bacterial counterpart. The TOM and SAM complexes interact, so the unfolded polypeptide can move from the TOM complex to the SAM complex with assistance from TIM chaperones associated with the inner membrane. Translocation Across the Inner Membrane to the Matrix Proteins use the Tim23 translocon to cross the inner membrane into the matrix. The integral membrane proteins Tim23 and Tim17 form the channel across the inner membrane and associate with three other subunits (Fig. 18.3). These proteins consist of four transmembrane helices rather than β-strands as in Tom40. Interactions of the N-terminal presequences of matrix proteins with Tim50 and Tim23 guide the presequence into the
translocation channel. Physical interactions of Tom and Tim complexes may facilitate the transfer of matrix proteins across both membranes. The MPP peptidase (matrix processing protease) cleaves off the presequences once they enter the matrix. Two energy sources—the electrical potential across the inner membrane and ATP hydrolysis by matrix chaperones—power polypeptide translocation across the inner membrane. The membrane potential (negative inside) pulls positively charged presequences across the membrane. Then the chaperone Hsp70 takes over and uses cycles of peptide binding and ATP hydrolysis to move the peptide into the matrix. One idea is that Hsp70 rectifies movements of the polypeptide in the pore, allowing movement forward into the matrix but not backward. Hsp70 binds when the polypeptide slides forward. After ATP hydrolysis, Hsp70 dissociates from the polypeptide and the exchange factor mGrp1 (see Fig. 12.12 for a related protein) rapidly recharges it with ATP, ready for another cycle of peptide binding, ATP hydrolysis, and release. This allows the polypeptide to slide forward into the matrix but not backward, so it eventually ends up as a folded protein in the matrix. Another model proposes that the energy from ATP hydrolysis is used to actively pull the polypeptide across the inner membrane. Proteins destined for the intermembrane space take the same route as those going to the matrix. However, after their presequences are cleaved by the MPP peptidase, they reverse into the space between the two membranes rather than continuing into the matrix.
CHAPTER 18 n Posttranslational Targeting of Proteins
Transport of Proteins Into Chloroplasts
A. Targeting sequence B. Chaperones C. Translocation binds Tom bind peptide from Tim22 into bilayer N
C
N
70
Tom 40 Tom22 +
+
Tim8/13
Tiny Tim chaperones +
+
+
∆ψ – – –
CYTOPLASM
C
+ – + – Tom + –
Tom20
+
Internal targeting sequence
Tim9/10 +
+
∆ψ – –
–
– – –
+
Tim 22
+
Tim 54
+
– –
+
+
+
+
+
C+
+
N
+
∆ψ –
–
–
–
–
–
307
– ++
–
MATRIX
FIGURE 18.5 IMPORT OF THE ADP/ATP ANTIPORTER ACC AND ITS INSERTION INTO THE INNER MEMBRANE BILAYER. A white bar across a translocon pore indicates that it is closed. A, An internal targeting sequence binds the ACC polypeptide to Tom70, which directs it into the Tom channel. B, In the intermembrane space, Tim9/10 and Tim8/13 capture the polypeptide and direct it to the Tim22/54 translocon that is used for import of matrix proteins. C, Tim22/54, in conjunction with the inner membrane potential (Δψ), promotes insertion of the six transmembrane helices into the inner membrane bilayer.
Translocation Into the Inner Membrane Bilayer The integral proteins of the inner membrane lack cleavable targeting signals, depending instead on targeting information contained in the intact protein to reach their destination. One example is the most abundant protein of the inner membrane, the adenosine diphosphate (ADP)/ATP antiporter that spans the inner membrane six times (see Fig. 15.4). Its signal sequence is located in the middle of the polypeptide. A family of small “tiny Tim” chaperone proteins (Tim8, Tim9, Tim10, Tim12, and Tim13) guide inner membrane proteins from the TOM complex across the intermembrane space to the Tim22 translocon in the inner membrane (Fig. 18.5). Complexes of Tim9/10 or Tim8/13 bind to hydrophobic segments of polypeptides during transit to the inner membrane. Many inner membrane proteins use the Tim22 translocon to access the lipid bilayer. Tim22 has three or four transmembrane helices and forms the channel, but little is known about its structure. It associates with Tim54, Tim12, and Tim18 in a 300-kD complex. Insertion of transmembrane segments into the bilayer depends on membrane potential. Export From the Matrix Insertion of proteins synthesized in the matrix into the inner membrane depends on an inner membrane protein called Oxa1p, which forms a translocon similar to bacterial YidC and chloroplast Alb3 (see later sections). Mitochondrial ribosomes are anchored to Oxa1p, allowing hydrophobic transmembrane segments to insert directly into the bilayer. At least one other protein complex participates in export of proteins from the matrix.
Eukaryotes acquired chloroplasts through symbiosis with a photosynthetic cyanobacterium (see Figs. 2.4 and 19.7). Over time, most of the bacterial genes moved to the nucleus, so more than 3000 chloroplast proteins are synthesized on cytoplasmic ribosomes and imported into one of three chloroplast membranes or the three compartments that they surround (Fig. 18.2). The innermost thylakoid membranes contain the photosynthetic apparatus inherited from cyanobacteria. The outer membrane likely came from the eukaryotic host, whereas the inner envelope membrane has both bacterial and eukaryotic features. Some organisms acquired their photosynthetic plastids by secondary or even tertiary rounds of endosymbiosis (see Fig. 2.7). These secondary or tertiary plastids are bounded by one or more additional membranes and have more complicated mechanisms to import the proteins expressed from nuclear genes. Although the protein import systems of chloroplasts and mitochondria both use zip codes on the imported proteins and translocons in the membranes (Fig. 18.6), the two systems share no common proteins. Another striking difference is that chloroplasts use specialized versions of some import components to acquire different proteins. This flexibility is important, as these organelles (collectively called plastids), have tissue-specific and age-specific functions, ranging from photosynthesis to starch storage (see Chapter 19). Mitochondria are more homogeneous, so a restricted set of import proteins suffices. In plants, N-terminal signal sequences called transit sequences target chloroplast proteins to the import machinery in the outer envelope. When added experimentally to the N-terminus of a test protein, transit sequences suffice to guide the test protein into the stroma of chloroplasts. These N-terminal targeting sequences vary in length from 13 to 146 residues, and their amino acid sequences have little in common beyond a net positive charge. The current understanding is that transit sequences are heterogeneous, because they contain many different types of zip codes that bind selectively to the specialized import receptors that plants express to allow different plastids to import the proteins appropriate for chloroplasts, starch storage organelles, and other specialized functions. All imported proteins use the same “general import pathway” to cross the outer and inner envelope membranes. The machinery consists of different protein complexes in each membrane called Toc (translocon at the outer envelope membrane of chloroplasts) and Tic (translocon at the inner envelope membrane of chloroplasts) (Fig. 18.6). These complexes were identified by chemical crosslinking of imported proteins to translocon proteins. Both Toc and a “super complex” of Toc with
308
SECTION VI n Cellular Organelles and Membrane Trafficking
B CYTOPLASMHsp70
A. Common pathway, six destinations
Toc 34
GTP
Toc 159
GTP Toc 36
Toc 75
Toc 34 OM
INTERMEMBRANE
GTP
Toc 75
OM
C
D
Transit sequence GTPGDP Toc 159
F
C
GTP Toc 36
Hsp70
INTERMEMBRANE
SPACE
E
SPACE
ATP
Tic 20/21 Tic 110
IM
STROMA
Tic 110
Signal peptidase
THYLAKOID MEMBRANE
Tic
Tic 20/21
IM
C
SP N
Signal peptidase
THYLAKOID MATRIX
ADP
STROMA
Cleaved transit signal Hsp70 Hsp60
FIGURE 18.6 CHLOROPLAST PROTEIN IMPORT VIA TOC AND TIC COMPLEXES. A, Common pathway across the outer membrane leading to six different destinations. B–F, Five stages in the movement of proteins into the stroma. B, Energy-independent binding of the transit sequence to outer membrane lipids and proteins, especially Toc159. C, Then guanosine triphosphate (GTP) hydrolysis by Toc34 and perhaps Toc159 promote insertion of the transit sequence through the outer membrane pore composed of Toc75. D, ATP hydrolysis by Hsp70 facilitates delivery to the inner membrane translocon Tic20/21. E, A stromal protease SP removes the N-terminal transit sequence. F, Hsp60 (heat shock protein 60) and Hsp70 promote folding of stromal proteins, while other proteins are rerouted to other compartments, including thylakoids. Tic, translocon at the inner envelope membrane of chloroplasts; Toc, translocon at the outer envelope membrane of chloroplasts. (Modified from Chen X, Schnell D. Protein import into chloroplasts. Trends Cell Biol. 1999;9:222–227; and May T, Soll J. Chloroplast precursor protein translocon. FEBS Lett. 2000;452:52–56.)
Tic can be isolated for analysis of their composition. Mutations that compromise chloroplast import have also contributed to understanding the process. The journey of a protein from its site of synthesis in cytoplasm into the stroma is understood in broad outline. The Toc159 and Toc34 proteins are receptors for transit sequences on the surface of chloroplasts. Plant cells express different isoforms of Toc159 and Toc34 to accommodate the import of different proteins appropriate to their differentiated state, although the details of these interactions are not yet clear. Both Toc159 and Toc34 have cytoplasmic guanosine triphosphatase (GTPase) domains similar to other small GTPases (see Fig. 25.7). Bound guanosine triphosphate (GTP) favors binding of transit sequences, and a bound transit sequence stimulates GTP hydrolysis and transfer of the transit sequence to the translocon. The β-barrel protein Toc75 forms the translocon across the outer membrane for all imported proteins. A homologous protein Omp85 translocates proteins in the opposite direction across the outer membrane of gramnegative bacteria. These proteins have a variable number of N-terminal POTRA (polypeptide transport associated) domains that may interact with many different transit sequences. Polypeptides are thought to be unfolded during transit through the narrow pore. Proteins destined for the outer membrane insert after passing through Toc75, similar to mitochondria.
The pore across the inner membrane consists of a complex of at least seven Tic proteins, but many details regarding their structures and functions are not yet settled. The abundant protein Tic110 not only forms some or all of a pore but also binds Hsp70 chaperones on the stromal side of the membrane. Smaller proteins called Tic20 and Tic21 appear to be distantly related to mitochondrial Tim23/17, so they may form a second type of channel composed of transmembrane helices in the inner membrane. As proteins emerge into the stroma, a signal peptidase cleaves off the transit peptide before the proteins fold or redistribute to their final locations. Hydrolysis of hundreds of ATP molecules in the stroma is required to complete the translocation and folding of imported proteins. The contributing enzymes include an AAA adenosine triphosphatase (ATPase), heat shock protein 93 (Hsp93), chaperone Hsp70, and a chaperonin heat shock protein 60 (Hsp60) (see Chapter 12). Some proteins remain in the stroma. Other proteins move on to thylakoid membranes or the thylakoid lumen using at least four different pathways. Some photosynthesis proteins insert directly into thylakoid membranes from the stroma. Others require help from proteins homologous to parts of the signal recognition particle (SRP) system used for export from bacteria (Fig. 18.10) and into the ER of eukaryotes (see Fig. 20.6). Although chloroplasts lack SRP RNA, GTPases
CHAPTER 18 n Posttranslational Targeting of Proteins
similar to an SRP protein and the SRP receptor cooperate with a protein that is homologous to translocon Oxa1p to mediate insertion into the thylakoid membrane. Hydrophilic proteins destined for the thylakoid lumen retain a secondary N-terminal signal sequence after the transit sequence is cleaved in the stroma. Some move across the thylakoid membrane into the thylakoid lumen through a translocon homologous to bacterial SecYE, powered by ATP hydrolysis by a homolog of SecA (Fig. 18.9). Other proteins with tightly bound redox factors cross the thylakoid membrane while compactly folded using translocon factors similar to the bacterial Tat system (Fig. 18.2). Secondary signal sequences with two arginine residues direct these proteins to a Tat translocon and the proton gradient drives the polypeptide across the membrane. After translocation, a peptidase in the thylakoid lumen removes both types of secondary signal sequences.
Transport of Proteins Into Peroxisomes Peroxisomes are simple organelles with a single membrane limiting a lumen containing many oxidative enzymes (see Fig. 19.10). Nuclear genes encode all proteins found in the membrane and lumen of peroxisomes. Their messenger RNAs are translated on cyto plasmic ribosomes, and the proteins are incorporated posttranslationally into peroxisomes (Fig. 18.1). Two types of targeting signals direct proteins from the cytoplasm to the peroxisome lumen (called matrix). The type 1 peroxisomal targeting signal (PTS1) is found at the extreme C-terminus of most peroxisomal enzymes (Fig. 18.7). PTS1 is just three amino acids
C
A
309
long with consensus sequence of serine-lysine-leucineCOOH, or a conservative variant. For example, alanine or cysteine can substitute at the −3 position, arginine or histidine can function at the penultimate position, and methionine can substitute for the C-terminal leucine. Amidation of the C-terminal carboxylate inactivates the signal. The type 2 peroxisomal targeting signal (PTS2) also targets a few proteins (only four are known in humans, one in yeast) to the peroxisome matrix. PTS2 sequences are located at or near the N-terminus and have a loose consensus sequence of RLXXXXXH/QL (where X is any amino acid). Proteins called peroxins deliver proteins from the cytoplasm to the peroxisomal membrane or lumen (Fig. 18.8 and Table 18.1). Loss-of-function mutations in humans and yeast revealed genes for more than 20 peroxins that participate in the biogenesis and proliferation of peroxisomes. Mutations of human PEX genes cause devastating diseases (Table 18.1). The PEX5 import receptor carries proteins with a PTS1 signal to the peroxisomal lumen by a highly unusual mechanism. The soluble protein PEX5 is proposed to bind the PTS1 motif and insert into the lipid bilayer surrounding the peroxisome, where it forms a transient pore to deliver the cargo protein into the lumen. Membrane proteins including a receptor called PEX14 participate in delivery of proteins into the lumen. Then PEX5 returns to the cytoplasm for further rounds of import. Recycling PEX5 depends on conjugation with ubiquitin and ATP hydrolysis by an AAA ATPase, a process with some similarities to ER-associated degradation (ERAD; see Fig. 20.12). A similar mode of action is proposed for PEX7, the import receptor for PTS2 proteins.
B Arg 520
Arg 378
H2O Leu (-1)
Ser (-3)
Gln (-4)
Tyr (-5)
H2O
N
TPRs 1–3
Lys (-2)
TPRs 5–7
Asn 524 Lys 490
Asn 497 Asn 489
Asn 531
FIGURE 18.7 BINDING OF A PEROXISOMAL TARGETING SIGNAL TYPE 1 (PTS1) TARGETING SIGNAL TO PEX5. A, PTS1 binds to the C-terminal tetratricopeptide repeat (TPR) domain of PEX5. The C-terminal, 40-kD TPR domain of PEX5, shown as a ribbon diagram, surrounds the PTS1 peptide, shown as a stick figure. Note TPRs 1 to 3 (yellow ribbons) and TPRs 5 to 7 (blue ribbons). An α-helical span (green ribbon) links the two triplet TPRs at the bottom of this structure; the C-terminal extension (white ribbons) also connects the two triplet TPRs. B, Detailed view of PEX5–PTS1 interactions between the PTS1 backbone (brown bonds) and PEX5 side chains (white bonds); the putative hydrogen bonds are shown as dashed green lines. This structure revealed the chemical basis of PEX5-PTS1 binding, as well as the sequence constraints of PTS1. (A, From PDB file 1FCH and Gatto GJ Jr, Geisbrecht BV, Gould SJ, Berg JM. Peroxisomal targeting signal-1 recognition by the TPR domains of human PEX5. Nat Struct Biol. 2000;7:1091–1095.)
310
SECTION VI n Cellular Organelles and Membrane Trafficking
A. De novo formation
Faresylated PEX16 from cytoplasm
B. Growth and division
Budding
PEX16- and PEX3mediated import of membrane proteins
PEX5- and PEX7mediated import of matrix proteins
Expansion by continued import of matrix and membrane proteins
Division
Nascent peroxisome
Preperoxisome PEX3 inserted into ER membrane
FIGURE 18.8 PEROXISOME BIOGENESIS. A, De novo formation by budding of a vesicle containing PEX3 and PEX16 from endoplasmic reticulum to form a preperoxisome. B, Growth and division of peroxisomes. PEX3 and PEX16 mediate the import of membrane proteins. The PEX5–PTS1 receptor, PEX7, and other peroxins mediate the import of proteins with PTS1 and PTS2 into peroxisomes.
TABLE 18.1 Peroxins: Proteins for the Assembly of Peroxisomes Groups
Peroxins
Functions
Diseases
Peroxisome Matrix Protein Import PEX5
PTS1 receptor and channel
NALD, ZS
PEX13, PEX14, PEX17, PEX33
Membrane docking complex for PEX5
NALD, ZS
PEX8 (PEX23)
Membrane partners of PEX5 & PEX14
PEX7
PTS2 receptor
PEX18, PEX20, PEX21
PTS2 coreceptors with PEX7
RCDP
Recycling Matrix Import Machinery PEX1, PEX6
AAA ATPases for recycling PEX5
PEX15, PEX26
Membrane anchors for PEX1 & PEX6
PEX2, PEX4, PEX10, PEX12
Ubiquitin-conjugating enzyme for cycling PEX5
PEX22
Membrane anchor for PEX4
IRD, NALD, ZS IRD, NALD, ZS
Peroxisome Membrane Protein Import PEX19
Cytoplasmic PMP receptor & chaperone
ZS
PEX3 (PEX16)
Membrane receptor for imported protein
IRD, ZS ZS
Peroxisome Biogenesis From Endoplasmic Reticulum PEX16
Recruits PMPs from ER membrane
PEX23, PEX30
Regulate de novo peroxisome formation
PEX25
Recruits Rho1
Peroxisome Fusion and Fission PEX11, PEX25, PEX27, PEX34
Regulate membrane elongation & fission
Mild ZS
PEX1, PEX6
AAA ATPases membrane fission & fusion
IRD, NALD, ZS
ATPase, adenosine triphosphatase; ER, endoplasmic reticulum; IRD, infantile Refsum disease; NALD, neonatal adrenoleukodystrophy; PBD, peroxisome biogenesis disorder; PMP, peroxisome membrane protein; PTS1/PTS2, peroxisomal targeting signal type 1/type 2; RCDP, rhizomelic chondrodysplasia punctate; ZS, Zellweger syndrome. For reference, see Smith JJ, Aitchison JD. Peroxisomes take shape. Nat Rev Mol Cell Biol. 2013;14:803–817.
Peroxisomal membranes form from lipids synthesized in the ER and proteins imported from the cytoplasm. The lipids are delivered in vesicles. Peroxisomal membrane proteins use one or more membrane peroxisomal targeting sequences (mPTSs) for delivery to peroxisomes. The sequences of mPTS motifs vary widely but include basic amino acids along with a transmembrane domain. PEX19 is the cytoplasmic receptor for mPTS motifs. It stabilizes these membrane proteins in the cytoplasm and delivers them to PEX3 and PEX16 on the peroxisome for insertion into the membrane (Fig. 18.8). Cells deficient in any of these three peroxins lack
peroxisomal membranes and the peroxisomal membrane proteins are degraded or mislocalized to other cellular membranes, particularly mitochondria. Peroxisomes may arise by either of two pathways (Fig. 18.8). Some peroxisomes form de novo by budding from the ER. PEX3 is inserted into the ER, where it recruits PEX16 and other peroxins. This specialized domain of ER then pinches off to form a nascent peroxisome de novo. By originating from the ER in this manner, peroxisomes can arise in cells that lack them without a preexisting peroxisome as template. Preexisting peroxisomes can grow by fusing with nascent peroxisomes and importing
proteins and lipids. They also divide by a fission process that depends the GTPase dynamin (see Fig. 22.9).
Translocation of Eukaryotic Proteins Across the Plasma Membrane by ABC Transporters Most proteins that are secreted by eukaryotic cells travel to the cell surface through the classical secretory pathway, including the ER and Golgi apparatus (see Chapters 20 and 21). But budding yeast use an ABC transporter (see Fig. 8.9) to transport their a-type mating factor directly from the cytoplasm across the plasma membrane. The a-factor is synthesized in the cytoplasm as part of a precursor, excised from the precursor by proteolytic cleavage, and then prenylated on its C-terminus before transport across the plasma membrane. This mechanism has been invoked to explain the secretion of a few mammalian proteins that lack the “signal sequences” that direct proteins to the classic ER secretory pathway. These include some cytokines, fibroblast growth factor, and some blood-clotting factors. This is a well-characterized route for secretion of some bacterial proteins.
Targeting to the Surfaces of the Plasma Membrane Many proteins synthesized in the cytoplasm are targeted to the cytoplasmic side (known as the cytoplasmic leaflet) of organelle and plasma membranes (see Fig. 13.9). These include peripheral membrane proteins that bind to cytoplasmic domains of integral membrane proteins or bind directly to the lipid bilayer. Other proteins are tethered to membrane bilayers by a covalently attached lipid added as a posttranslational modification following synthesis on cytoplasmic ribosomes. Lipid modifications on tethered proteins include long-chain, saturated fatty acids and isoprenoids. The saturated fatty acids are either myristate (14 carbons), which is added through amide linkage to aminoterminal glycine residues, or palmitate (16 carbons), which is usually added through a thioether linkage to cysteine residues found toward the C-terminus. The isoprenoids farnesyl (15 carbons) and geranylgeranyl (20 carbons) are added through thioether linkages to cysteine residues located at or near the C-terminus in specific structural motifs. Attachment of a lipid helps stabilize membrane association, but does not guarantee permanent anchoring to the membrane. Some proteins, such as the catalytic subunit of cyclic adenosine monophosphate–dependent protein kinase (PKA), are fatty acylated but remain mostly soluble in cytoplasm. Proteins attached to the external surface of plasma membranes by glycosylphosphatidylinositol anchors
CHAPTER 18 n Posttranslational Targeting of Proteins
311
arrive by a different route. These proteins are synthesized on ribosomes associated with the ER and then translocated into the ER lumen anchored by a C-terminal transmembrane segment. Inside the ER the protein is cleaved from its membrane anchor and transferred enzymatically to glycosylphosphatidylinositol before transport to the cell surface (see Fig. 20.8C).
Prokaryotic Protein Export Bacteria and Archaea employ at least 10 distinct strategies to transport proteins from the cytoplasm across the inner membrane and beyond. Seven of these pathways use a common pore across the inner membrane called the Sec translocon. These pathways are important because some contribute to human disease. In addition, they serve as important model systems, since eukaryotes use a homologous translocon to move proteins into the bilayer or lumen of the ER (see Fig. 20.7). This section begins with a discussion of six branches of the Sec secretory pathway and finishes with three distinct pathways.
Pathways Dependent on the SecYE Translocon Organisms in all three domains of life use Sec translocons to move proteins synthesized in the cytoplasm across membranes. Translocons in the plasma membranes of Bacteria and Archaea consist of two transmembrane proteins called SecY and SecE in Bacteria (Fig. 18.9). The translocons of the ER of eukaryotes consist of homologous protein subunits called Sec61α and γ (see Fig. 20.7). The narrow pore for translocating the secreted polypeptide is located in the middle of a bundle of α helices. Loss-of-function mutations of SecY or SecE compromise the secretion of most proteins by Bacteria or Archaea. Several accessory subunits assist in translocation, but they are not essential in Bacteria and are not present in eukaryotes. Posttranslational Protein Translocation Bacteria use Sec-signal sequences to direct many proteins to the SecYE translocon for transport across the plasma membrane or for insertion into the plasma membrane. Gram-positive bacteria such as Bacillus subtilis lack an outer membrane, so the proteins leave the cell after crossing the plasma membrane. In gram-negative bacteria, translocated proteins enter the periplasm, insert into the outer membrane, or leave the cell. Proteins targeted to the Sec translocon are synthesized in the cytoplasm with an N-terminal Sec-signal sequence. These targeting sequences consist of approximately 25 residues beginning with methionine, followed by a few basic residues, 10 to 15 hydrophobic residues, and a site for cleavage by a proteolytic enzyme called signal peptidase after translocation across the inner membrane. Chaperones such as SecB bind newly synthesized proteins to prevent folding and maintain a state that is
312
SECTION VI n Cellular Organelles and Membrane Trafficking
A. Bacterial protein export mechanism
N
Signal peptidase N
PERIPLASM
N
N
3
INNER
4
5
6
MEMBRANE
ATP
N
ADP
Pi
ATP SecA
7
SecYE
ATP
C
2 C
C
C
C
1 SecB
B. SecB
C. SecA
D. SecYE
C
N C
ADP
CYTOPLASM
FIGURE 18.9 SECRETION OF PROTEIN FROM BACTERIA THROUGH THE SecYE TRANSLOCON. A, Pathway of secretion. 1, After synthesis by a cytoplasmic ribosome, the polypeptide associates with the SecB chaperone. 2, SecA binds the presequence (blue) and docks on the SecYE translocon. 3, The presequence inserts into the translocon. 4, ATP-binding to SecA promotes insertion of the associated polypeptide into the translocon, followed by cleavage of the signal sequence. 5–7, The membrane potential and cycles of ATP hydrolysis by SecA drive the polypeptide across the inner membrane. B, Ribbon diagram of Haemophilus influenzae SecB. C, Ribbon diagram of Bacillus subtilis SecA. D, Ribbon diagram of Methanococcus jannaschii SecY complex translocon. (A, Modified from Danese PN, Silhavy TJ. Targeting and assembly of periplasmic and outer-membrane proteins in E. coli. Annu Rev Genet. 1999;32:59–94. For reference, see PDB file IOZB and Zhou J, Xu Z. Structural determinants of SecB recognition by SecA in bacterial protein translocation. Nat Struct Biol. 2003;10:942–948 [B]; PDB file 1TF2 and Osborne AR, Clemons WM, Rapoport TA. A large conformational change of the translocation ATPase SecA. Proc Natl Acad Sci U S A. 2004;101:10937–10942 [C]; and PDB file 1RHZ and van de Berg B, Clemons WM, Collinson I, et al. X-ray structure of a protein-conducting channel. Nature. 2004;427:36–44 [D].)
competent for translocation (Fig. 18.9). Unlike most other chaperones (see Figs. 12.13 and 12.14), SecB does not require ATP hydrolysis for cycles of interaction with substrates. Hsp70 homologs (DnaK) have a secondary role in chaperoning precursors for translocation. Translocation of many bacterial membrane and secreted proteins with cleavable signal sequences depends on the ATPase SecA. A system reconstituted from purified SecA, SecY, and SecE can translocate precursor proteins across lipid membranes in the presence of ATP. Remarkably, Archaea lack SecA, although they depend on translocon components that are homologous to SecYE. Eukaryotes use SecA only for translocation into chloroplast thylakoids (Fig. 18.2). SecA binds proteins associated with SecB in the cytoplasm and targets the signal sequence to the SecY translocon. SecY “proofreads” the signal sequence associated with SecA, releasing those with imperfect matches to the consensus prior to translocation. Binding to SecY
changes the conformation of SecA, bringing together two domains that form a clamp around the substrate peptide. Inside the clamp the peptide substrate binds weakly through hydrogen bonds to the edge of a small β-sheet. Cycles of ATP hydrolysis by SecA cause a rocking motion of a lever arm that pushes the unfolded peptide through the channel in small steps. The protein then folds after translocation. Signal peptidases located on the outer surface of the plasma membrane cleave signal peptides from translocated proteins soon after they cross the plasma membrane. Some bacterial signal peptidases are similar to eukaryotic homologs. Other bacterial signal peptidases are specialized to cleave lipoproteins just before an invariant cysteine. This cysteine is then conjugated to diacylglycerol, which anchors the lipoprotein to the outer surface of the plasma membrane or to the outer membrane of gram-negative bacteria. Signal peptidases also degrade cleaved signal peptides.
Translocation Dependent on the Signal Recognition Particle In eukaryotes, the signal recognition particle (SRP) is the adapter between signal sequences and the translocon of ER (see Fig. 20.5), but in bacteria, only a minority of integral membrane proteins and secreted proteins depend on SRP for targeting to the Sec translocon. Eukaryotic and Archaeal SRPs consist of a 7S RNA and several proteins, whereas Escherichia coli SRP consists of a smaller 4.5S RNA and a single protein called Ffh (for “fifty-four homolog,” after its eukaryotic counterpart) (see Fig. 20.5). SRP binds Sec-signal sequences and signalanchor sequences as they emerge from the ribosome. This interaction slows translation until SRP docks on the cytoplasmic surface of the inner membrane with its receptor FtsY and the Sec translocon. Resumption of translation drives the polypeptide through the translocon. See Chapter 20 for more details on SRP and eukaryotic cotranslational translocation. Insertion of Inner Membrane Proteins Proteins inserted into the inner membrane depend on YidC, a protein with six transmembrane helices related to Oxa1p and Alb3, that direct proteins into the inner membrane of mitochondria and thylakoid membranes of chloroplasts. Work is being done to determine if YidC inserts membrane proteins on its own or accepts them from the Sec translocon. Insertion of Proteins in the Outer Membrane of Gram-Negative Bacteria Outer membrane proteins are synthesized in the cytoplasm and directed to the Sec translocon by signal sequences. The signal sequence is cleaved from the unfolded protein after crossing the inner membrane into the periplasm. Several periplasmic chaperones and assembly factors participate in protein folding, including enzymes that catalyze the isomerization of proline peptide bonds and oxidation/reduction of cysteine thiol groups. Insertion into the outer membrane depends on β-barrel proteins, but little is known about the targeting signals. Outer Membrane Autotransporter Pathway Some proteins, including secreted proteolytic enzymes and toxins as well as membrane-anchored adhesins and invasins, hitch a ride to the cell surface on their own outer membrane transporters (Fig. 18.10A). These proteins have an N-terminal secreted domain and a C-terminal domain that forms a transmembrane β-barrel like a porin (see Fig. 13.9). The protein uses the Sec pathway to cross the inner membrane and the β-barrel inserts into the outer membrane. The N-terminal functional domain then translocates across the outer membrane through its β-domain pore. An outer membrane protease releases toxins and proteases, whereas adhesins that follow this route remain on the surface attached to the β-domain.
CHAPTER 18 n Posttranslational Targeting of Proteins
313
Outer Membrane Single Accessory Pathway Some hemolysins and hemagglutinins move to the periplasm through the Sec pathway and then use a single accessory protein to translocate across the outer membrane. The accessory protein forms a β-barrel in the outer membrane with a pore like autotransporters and the porins that transport peptides across the outer membranes of chloroplasts. Chaperone/Usher Pathway Gram-negative bacteria use a novel mechanism, downstream of the Sec pathway, to transport and assemble pili on their outer surface. Pili are appendages involved with bacterial pathogenesis, including urinary tract infections. A periplasmic chaperone binds the pillus peptide and promotes folding. The pilus subunit is folded similar to an immunoglobulin domain (see Fig. 3.13), but lacks the seventh β-strand. This exposes core hydrophobic residues. The chaperone consists of two immunoglobulin-like domains, one of which donates a strand to complete the immunoglobulin domain of the pilus subunit. The chaperone delivers a pilus subunit to an outer membrane translocon called usher (Fig. 18.10A). There it transfers its bound subunit to the end of a growing chain of pilus subunits, all bound together, head to tail, by strands that complete the seven-strand β-sheet of the adjacent subunit. On the outer surface, the pilus subunits rearrange into a helical pilus. The assembly reaction is thought to provide the energy for translocation. The chaperone prevents premature assembly of the pilus. Type II Secretion Bacteria use an alternate route downstream of the Sec pathway to secrete other toxins and enzymes with cleaved signal sequences (Fig. 18.10A). At least a dozen protein subunits participate in this complicated pathway. The pore in the outer membrane is composed of a secretin, a protein with relatives that also participate in type III secretion, phage biogenesis, and formation of one type of pilus. The secretin pore is a ring of 12 to 14 subunits around a large gated channel that is 5 to 10 nm in diameter. Type IV Secretion Bacteria secrete a few proteins using an apparatus similar to that used for DNA transfer between two bacteria during conjugation and for DNA injection into plant cells by Agrobacterium. DNA is transferred directly from the cytoplasm of one bacterium to the cytoplasm of another bacterium or plant cell. Proteins that are secreted by this pathway include pertussis toxin by Bordetella pertussis and another toxin by Helicobacter pylori. This pathway starts with synthesis in the cytoplasm and translocation across the plasma membrane by the Sec translocon. If present, the signal sequence is cleaved
314
SECTION VI n Cellular Organelles and Membrane Trafficking
A. SecYE-dependent pathways Type V autotransporter
Single accessory subunit
Pillus
N
Type II
Usher Secretin
OM Forms channel and self-inserts PERIPLASM
Folded protein N
IM
Type IV To eukaryotic cell
N
Chaperone
Folded protein
N
SecYE Cleaved signal sequence
DNA
CYTOPLASM
B. SecYE-independent pathways
To eukaryotic cell
Folded protein Type I
Flagella
Folded protein
Type III
Tol C Membrane fusion protein C
N
ABC transporter Flagellin subunit
ATP ADP
ATP ADP
N
C
FIGURE 18.10 SECRETION ACROSS THE OUTER MEMBRANE OF GRAM-NEGATIVE BACTERIA. A, Pathways dependent on SecYE. The cleaved signal sequence is shown in blue. The β-domain of autotransporters forms a pore for the translocation of part of its own chain, which may remain attached, as shown, or be cleaved for escape from the cell. Single accessory proteins form a pore for secretion of separate proteins. Usher forms a pore for the translocation and assembly of pili. Type II secretion uses a secretin pore for translocation. Type IV secretion employs a large translocon similar to that used by Agrobacterium for secretion of DNA. B, Pathways independent of SecYE. Type I secretion uses an ABC transporter to cross the inner membrane and additional subunits to cross the periplasm and outer membrane. Left panel, Ribbon model of TolC, one type of translocon that spans the periplasm and outer membrane. Right panel, Each TolC subunit contributes four β-strands to a porin-like structure that spans the outer membrane. α-Helical continuations of these β-strands form a tube having an internal diameter of 3.5 nm for transport of proteins across the periplasm. Bacterial flagella transport flagellin subunits across both membranes and then through the central channel of the flagellar filament for incorporation at the growing tip. Type III secretion uses components similar to the basal body of flagella. Gray illustration (far right) shows a three-dimensional reconstruction of the type III secretion apparatus from Salmonella typhimurium. IM, inner membrane; OM, outer membrane. (A–B, Modified from Thanassi DG, Hultgren SJ. Multiple pathways allow protein secretion across the bacterial outer membrane. Curr Opin Cell Biol. 2000;12:420–430. B, TolC ribbon diagram based on PDB file 1EK9. For reference, see Koronakis V, Sharff A, Koronakis E, et al. Crystal structure of the bacterial membrane protein TolC central to multidrug efflux and protein export. Nature. 2000;405:914– 919. Reconstruction of the type III secretion complex from S. typhimurium based on Marlovits TC, Kubori T, Sukhan A, et al. Structural insights into the assembly of the type II secretion needle complex. Science. 2004;306:1040–1042.)
before translocation across the outer membrane by the type IV secretion system (Fig. 18.10A).
Pathways Independent of the Sec Translocon Type I ABC Transporters Bacteria use ABC transporters (see Fig. 8.9) to secrete a small number of toxins (eg, E. coli hemolysin), proteases, and lipases. C-terminal signal sequences of 30 to 60 residues target these proteins to the ABC transporter, the only component required for secretion by gram-positive bacteria. Gram-negative bacteria require not only a transporter in the inner membrane but also two proteins that form a continuous channel across the periplasm and outer membrane (Fig. 18.10B). ATP hydrolysis by the ABC transporter provides energy for translocation. Protein conduits across the periplasm and outer membrane engage ABC transporters presenting substrates for export and then disengage when translocation is complete. Genes for secreted proteins are generally in the same operon as the export machinery. Flagellar and Type III Secretion Systems The basal bodies of bacterial flagella transport flagellin subunits through a central pore that crosses both membranes (Fig. 18.10B) and extends the length of the flagellar shaft to the tip, where subunits add to the distal end (see Fig. 5.8). This flagellar pathway transports a few other proteins, including a phospholipase that contributes to the virulence of Yersinia, the cause of the black plague. Pathogenic gram-negative bacteria, such as Yersinia, use the syringe-like type III apparatus, similar to a bacterial flagellum, to transport toxins from the cytoplasm into the medium or directly into target cells. In the target cell, these toxins disrupt cellular physiology, in part by forming pores in target cell membranes. The type III secretion complex consists of approximately 20 different protein subunits including some with homology to rotary ATPases (see Fig. 8.4). A complex base consisting of several protein rings spans the periplasm and both membranes. A polymer of a single type of protein forms a hollow needle up to 40 nm long for injection of toxins directly into target animal or plant cells. Several signals direct proteins to this pathway. One such signal is a noncleavable signal sequence in the secreted protein that binds a chaperone dedicated to targeting toxins to the type III pathway. A cytoplasmic ATPase separates secreted proteins from chaperones, and the transmembrane electrochemical gradient of protons provides most of energy for transport. Double Arginine Pathway Many but not all Bacteria and Archaea use proteins homologous to chloroplast Tat proteins to translocate
CHAPTER 18 n Posttranslational Targeting of Proteins
315
folded proteins across the plasma membrane. In both prokaryotes and chloroplasts, some of these cargo proteins participate in redox reactions and have bound cofactors such as flavins or FeS clusters. These cofactors are incorporated as the proteins fold in the cytoplasm or chloroplast stroma. In contrast to the Sec translocon, the Tat translocon accommodates folded proteins. The N-terminal signal sequences for this pathway have a pair of arginines (RR) in a conserved sequence (Ser/Thr-ArgArg-X-Phe-Leu-Lys, where X is any amino acid) adjacent to a stretch of at least 13 uncharged residues. Translocation of these proteins in E. coli requires three Tat proteins. One forms the transmembrane pore, and the others appear to participate in targeting. Virtually all Archaeal proteins that move through Tat remain anchored to the cell surface. SELECTED READINGS Berks BC. The twin-arginine protein translocation pathway. Annu Rev Biochem. 2015;84:843-864. Chacinska A, Koehler CM, Milenkovic D, et al. Importing mitochondrial proteins: machineries and mechanisms. Cell. 2009;138: 628-644. Chatzi KE, Sardis MF, Karamanou S, Economou A. Breaking on through to the other side: protein export through the bacterial Sec system. Biochem J. 2013;449:25-37. Costa TR, Felisberto-Rodrigues C, Meir A, et al. Secretion systems in gram-negative bacteria: structural and mechanistic insights. Nat Rev Microbiol. 2015;13:343-359. Dautin N, Bernstein HD. Protein secretion in Gram-negative bacteria via the autotransporter pathway. Annu Rev Microbiol. 2007;61: 89-112. Demarsy E, Lakshmanan AM, Kessler F. Border control: selectivity of chloroplast protein import and regulation at the TOC-complex. Front Plant Sci. 2013;5:483. Diepold A, Armitage JP. Type III secretion systems: the bacterial flagellum and the injectisome. Philos Trans R Soc Lond B Biol Sci. 2015; 370:20150020. Gutensohn M, Fan E, Frielingsdorf S, et al. Toc, Tic, Tat et al.: Structure and function of protein transport machines in chloroplasts. J Plant Physiol. 2006;163:333-347. Kedrov A, Kusters I, Driessen AJ. Single-molecule studies of bacterial protein translocation. Biochemistry. 2013;52:6740-6754. Li HM, Chiu CC. Protein transport into chloroplasts. Annu Rev Plant Biol. 2010;61:157-180. Li HM, Teng YS. Transit peptide design and plastid import regulation. Trends Plant Sci. 2013;18:360-366. Neupert W, Herrmann JM. Translocation of proteins into mitochondria. Annu Rev Biochem. 2007;76:723-749. Shiota T, Imai K, Qiu J, et al. Molecular architecture of the active mitochondrial protein gate. Science. 2015;349:1544-1548. Smith JJ, Aitchison JD. Peroxisomes take shape. Nat Rev Mol Cell Biol. 2013;14:803-817. Szabo Z, Pohlschroder M. Diversity and subcellular distribution of archaeal secreted proteins. Front Microbiol. 2012;3:207. Thanassi DG, Hultgren SJ. Multiple pathways allow protein secretion across the bacterial outer membrane. Curr Opin Cell Biol. 2000;12: 420-430.
This page intentionally left blank
CHAPTER
19
Mitochondria, Chloroplasts, Peroxisomes T
his chapter considers three organelles formed by posttranslational import of proteins synthesized in the cytoplasm. Mitochondria and chloroplasts both arose from endosymbiotic bacteria, two singular events that occurred about 1 billion years apart (see Fig. 2.4B). Both mitochondria and chloroplasts retain remnants of those prokaryotic genomes but depend largely on genes that were transferred to the nucleus of the host eukaryote. Both organelles brought biochemical mechanisms that allow their eukaryotic hosts to acquire and use energy more efficiently. In oxidative phosphorylation by mitochondria and photosynthesis by chloroplasts, energy from the breakdown of nutrients or from absorption of photons is used to energize electrons. As these electrons tunnel through transmembrane proteins, energy is extracted to create proton gradients. These proton gradients drive the rotary adenosine triphosphate (ATP) synthase (see Fig. 14.5) to make ATP, which is used as energy currency to power the cell. Peroxisomes contain no genes and depend entirely on nuclear genes
A
to encode their proteins. Their evolutionary origins are obscure. Peroxisomes contain enzymes that catalyze a wide range of oxidation reactions that are essential for cellular homestasis. Patients who lack peroxisomes have severe neural defects.
Mitochondria Evolution of Mitochondria Mitochondria (Fig. 19.1) arose about 2 billion years ago when a bacterium fused with an archaeal cell (see Fig. 2.4B and associated text). The bacterial origins of mitochondria are apparent in their many common features (Fig. 19.2). The closest extant relatives of the bacterium that gave rise to mitochondria are Rickettsia, aerobic α-proteobacteria with a genome of 1.1 megabase (Mb) pairs. These intracellular pathogens cause typhus and Rocky Mountain spotted fever. It now appears likely that the actual progenitor bacterium had the genes required for both aerobic and anaerobic metabolism.
B
FIGURE 19.1 CELLULAR DISTRIBUTION AND STRUCTURE OF MITOCHONDRIA. A, Fluorescence light micrograph of a Cos-7 tissue culture cell with mitochondria labeled with green fluorescent antibody to the β-subunit of the F1-ATPase (adenosine triphosphatase) and microtubules labeled red with an antibody. B, Electron micrograph of a thin section of a mitochondrion. (A, Courtesy Michael Yaffee, University of California–San Diego. B, Courtesy Don Fawcett, Harvard Medical School, Boston, MA.)
317
318
SECTION VI n Cellular Organelles and Membrane Trafficking
A. Mitochondria Outer membrane Inner membrane DNA Book icon
Cristae Matrix
B INTERMEMBRANE SPACE I
II
III
IV
C
PERIPLASM I
II
MATRIX
III
IV
CYTOPLASM
D. Bacterium
DNA CYTOPLASM Periplasmic space
Inner membrane Outer membrane
FIGURE 19.2 The compartments of a mitochondrion (A–B) compared with a bacterium (C–D). Respiratory chain complexes I to IV are labeled with roman numerals.
By the time of the last eukaryotic common ancestor (LECA), most of the bacterial genes were lost or moved to the nucleus, but all known mitochondria retain some bacterial genes. A few eukaryotes that branched from the last eukaryotic common ancestor, such as Entamoeba, subsequently lost the organelle, leaving behind a few mitochondrial genes in the nucleus. Chromosomes of contemporary mitochondria vary in size from 366,924 base pairs (bp) in the plant Arabidopsis to only 5966 bp in Plasmodium. These small, usually circular genomes encode RNAs and proteins that are essential for mitochondrial function, including some subunits of proteins responsible for ATP synthesis. The highly pared-down human mitochondrial genome with 16,569 bp encodes only 13 mitochondrial membrane proteins, two ribosomal RNAs, and just enough transfer RNAs (tRNAs) (22) to translate these genes. The number of proteins encoded by other mitochondrial genomes ranges from just three in Plasmodium to 97 in a protozoan. Nuclear genes encode more than 1000 other mitochondria proteins, including those required to assemble ribosomes and synthesize proteins in the matrix. All mitochondrial proteins encoded by nuclear genes are synthesized in the cytoplasm and imported into mitochondria (see Figs. 18.3 and 18.4).
Structure of Mitochondria Mitochondria consist of two membrane-bounded compartments, one inside the other (Fig. 19.2). The
outer membrane surrounds the intermembrane space. The inner membrane surrounds the matrix. Each membrane and compartment has a distinct protein composition and functions. Porins in the outer membrane provide nonspecific channels for passage of molecules of less than 5000 Da, including most metabolites required for ATP synthesis. Some proteins in the intermembrane space participate in ATP synthesis but, when released into the cytoplasm, trigger programmed cell death (see Fig. 46.16). The matrix is the site of fatty acid oxidation, the citric acid cycle, and mitochondrial protein synthesis. The highly impermeable inner membrane has two domains. The boundary domain next to the outer membrane is specialized for protein import at contacts with the outer membranes (see Fig. 18.4). The rest of the inner membrane forms folds called cristae that are specialized for converting energy provided by breakdown of nutrients in the matrix into ATP. Cristae may be tubular or flattened sacs and vary in number and shape, depending on the species, tissue, and metabolic state. Four complexes (I to IV) of integral membrane proteins use the transport of electrons to create a gradient of protons across the inner membrane (Fig. 19.2B). The F-type rotary ATP synthase (see Fig. 14.5) uses this proton gradient to synthesize ATP. A complex of five transmembrane proteins and two soluble proteins stabilizes the junction of cristae with the inner membrane. The complex is called MICOS for mitochondrial con tact site and cristae organizing system, because loss of these proteins results in disorganized cristae. MICOS links the inner and outer membranes and separates the transmembrane proteins of the inner membrane into two domains.
Biogenesis of Mitochondria Mitochondria grow by importing most of their proteins from the cytoplasm and by internal synthesis of some proteins and replication of their own genome (Fig. 19.3). Targeting and sorting signals built into the mitochondrial proteins that are synthesized in the cytoplasm direct them to their destinations (see Fig. 18.4). Similar to cells, mitochondria divide, but unlike most cells, they also fuse with other mitochondria. These fusion and division reactions were first observed nearly 100 years ago. Now it is appreciated that a balance between ongoing fusion and division determines the number of mitochondria within a cell. Both fusion and division depend on proteins with guanosine triphosphatase (GTPase) domains related to dynamin (see Fig. 22.8). In fact, eukaryotes might have acquired their dynamin genes from the bacterium that became the mitochondrion. Mitochondrial division starts when a tubule of endoplasmic reticulum (ER) membrane encircles a mitochondrion to mark the site for division. Then a dynamin-related
CHAPTER 19 n Mitochondria, Chloroplasts, Peroxisomes
Nuclear genes on nuclear DNA Over 1,000 mitochondrial proteins synthesized in cytoplasm Mitochondrial genes on mitochondrial DNA Import into mitochondria
• 13 mitochondrial membrane proteins • 22 tRNAs • 2 rRNAs
FIGURE 19.3 BIOGENESIS OF HUMAN MITOCHONDRIA. The relative contributions of nuclear and mitochondrial genes to the protein composition. rRNA, ribosomal RNA; tRNA, transfer RNA.
GTPase forms a spiral around the mitochondrion and cooperates with actin filaments to pinch mitochondria in two. During apoptosis (see Chapter 46), this GTPase also participates in the fragmentation of mitochondria. Mitochondrial fusion involves two GTPases, one anchored in the outer membrane and the other in the inner membrane, both linked by an adapter protein in the intermembrane space. Fusion of the outer membranes requires a proton gradient across the inner membrane, whereas fusion of the inner membranes depends on the electrical potential across the inner membrane. Loss-of-function mutations in fusion proteins lead to cells with numerous small mitochondria, some lacking a mitochondrial DNA (mtDNA) molecule. Human mutations in the genes for fusion proteins result in defects in the myelin sheath that insulates axons (one form of CharcotMarie-Tooth disease) and the atrophy of the optic nerve. Mitochondrial fusion proteins are also required for apoptosis.
Synthesis of ATP by Oxidative Phosphorylation Mitochondria use energy extracted from the chemical bonds of nutrients to generate a proton gradient across the inner membrane. This proton gradient drives the F-type rotary ATP synthase to produce ATP from adenosine diphosphate (ADP) and inorganic phosphate. Enzymes in the inner membrane and matrix cooperate with pumps, carriers, and electron transport proteins in the inner membrane to move electrons, protons, and other energetic intermediates across the impermeable inner membrane. This is a classic chemiosmotic process (see Fig. 17.1). Mitochondria receive energy-yielding chemical intermediates from two ancient metabolic pathways, glycolysis and fatty acid oxidation (Fig. 19.4), that both evolved in the common ancestor of living things. Both pathways feed into the equally ancient citric acid cycle of energy-yielding reactions in the mitochondrial matrix:
319
• The glycolytic pathway in the cytoplasm converts the six-carbon sugar glucose into pyruvate, a three-carbon substrate for pyruvate dehydrogenase, a large, soluble, enzyme complex in the mitochondrial matrix. The products of pyruvate dehydrogenase (carbon dioxide, the reduced form of nicotinamide adenine dinucleotide [NADH], and acetyl coenzyme A [CoA]) are released into the matrix. NADH is a high-energy electron carrier. Acetyl-CoA is a two-carbon metabolic intermediate that supplies the citric acid cycle with energy-rich bonds. • Breakdown of lipids yields fatty acids linked to acetylCoA by a thioester bond. These intermediates are transported across the inner membrane of mitochondria, using carnitine in a shuttle system. In the matrix, acyl-carnitine is reconverted to acyl-CoA. Enzymes in the matrix degrade fatty acids two carbons at a time in a series of oxidative reactions that yield NADH, the reduced form of flavin adenine dinucleotide (FADH2, another energy-rich electron carrier associated with an integral membrane enzyme complex), and acetylCoA for the citric acid cycle. Breakdown of acetyl-CoA during one turn of the citric acid cycle produces three molecules of NADH, one molecule of FADH2, and two molecules of carbon dioxide. Energetic electrons donated by NADH and FADH2 drive an electron transport pathway in the inner mitochondrial membrane that powers a chemiosmotic cycle to produce ATP (Fig. 19.5). Electrons use two routes to pass through three protein complexes in the inner mitochondrial membrane. Starting with NADH, electrons pass through complex I to complex III to complex IV. Association of these three complexes in a “super complex” may facilitate electron transfer. Electrons from FADH2 pass through complex II to complex III to complex IV. Along both routes, energy is used to transfer multiple protons (corresponding to at least 10 electrons per NADH oxidized) across the inner mitochondrial membrane from the matrix to the inner membrane space. The resulting electrochemical gradient of protons drives ATP synthesis (see Fig. 14.5). This process is called oxidative phosphorylation, because molecular oxygen is the sink for energy-bearing electrons at the end of the pathway, and because the reactions add phosphate to ADP. Eukaryotes that live in environments with little or no oxygen use other acceptors for these electrons and produce nitrite, nitric oxide, or other reduced products rather than water. Oxidative phosphorylation is understood in remarkable detail, thanks to atomic structures of F-type ATP synthase and each electron transfer complex. Nuclear genes encode most of the protein subunits of these complexes, but mitochondrial genes encode a few key subunits. Bacteria and mitochondria share homologous proteins for the key steps in oxidative phosphorylation (Fig. 19.2), although the machinery in mitochondria is usually
320
SECTION VI n Cellular Organelles and Membrane Trafficking
A. Glycolysis (in the cytoplasm) OPO32– HOH2C H2C O H ATP ADP H O H H H H HO OH H OH HO OH H OH OH H Glucose
Hexokinase
OH H Glucose 6-phosphate
OPO32– O H2C Phosphoglucose isomerase
OPO32– OH O ADP ATP H CH2 2C
H H HO OH OH H Fructose 6-phosphate
Phosphofructokinase
H H HO OH OH H Fructose 1,6-biphosphate
NAD+ COO – HO C H Malate CH2 – COO H2O
Phosphoglycerate kinase
2 ATP H H O 3-phosphoglycerate H C C C O– (2 molecules) O OH PO32–
H2O
Phosphoglyceromutase
H H O 2-phosphoglycerate – (2 molecules) H C C C O HO OPO32–
H2O
COO – CH Fumarate CH COO –
COO – Isocitrate CH2 H C COO – HO C H COO –
Enolase
2 H2O H H O Phosphoenolpyruvate – (2 molecules) H C C C O OPO32– 2 ADP Pyruvate
NAD+
FADH2
kinase
CO2 + NADH + H+
FAD COO – CH2 Succinate CH2 COO – GTP + HSCoA GDP + Pi
COO – CH2 α-Keto- CH2 glutarate C O COO –
Succinyl CoA O SCoA C CH2 CH2 COO –
2 NADH + 2 H+
H H O 1,3-bisphosphoglycerate H C C C OPO32– (2 molecules) O OH PO32– 2 ADP
COO – CH2 HO C COO – Citrate CH2 COO –
COO – CH2 cis-Aconitate C COO – CH COO –
NADH + H+
Triosephosphate isomerase
3-phosphate dehydrogenase
HSCoA
COO – C O Oxaloacetate CH2 COO –
Aldolase
HO O H H C C C OPO3H– O H
H H O Glyceraldehyde H C C CH 3-phosphate O OH (2 molecules) PO32– 2 NAD + 2 Pi Glyceraldehyde
B. Citric acid cycle (in the mitochondrial matrix) O H2O + CH3 C SCoA Acetyl-CoA
Dihydroxyacetone phosphate
OPO32– CH2
2 ATP
H O O Pyruvate – (2 molecules) H C C C O H
C. Integration of metabolic pathways in mitochondrium ADP
ATP
Lipid breakdown
Glycosis CO2 Pyruvate
NAD+ + HSCoA CO2 + NADH + H+
Fatty acids Pyruvate
H+ Pi O2
ADP O2 H2O + Pi ATP H+ H+ H+ e–
FADH2
H+ FADH2
Acetyl-CoA
Citric acid cycle
CO2 NADH
NADH FIGURE 19.4 METABOLIC PATHWAYS SUPPLYING ENERGY FOR OXIDATIVE PHOSPHORYLATION. A, Glycolysis. ADP, adenosine diphosphate; ATP, adenosine triphosphate; NAD, nicotinamide adenine dinucleotide; NADH, reduced form of nicotinamide adenine dinucleotide. B, Citric acid cycle. Production of acetyl-coenzyme A (CoA) by the glycolytic pathway in cytoplasm and fatty acid oxidation in the mitochondrial matrix drive the citric acid cycle in the mitochondrial matrix. This energy-yielding cycle is also called the Krebs cycle after the biochemist H. Krebs. NADH and FADH2 (reduced form of flavin adenine dinucleotide) produced by these pathways supply high-energy electrons to the electron transport chain. GTP, guanosine triphosphate; HSCoA, reduced coenzyme A. C, Overview of metabolic pathways. Note energy-rich metabolites (yellow).
321
CHAPTER 19 n Mitochondria, Chloroplasts, Peroxisomes
A
B
INTERMEMBRANE SPACE
Cytochrome c H+
H+
H+
ATP
H+ Pi
+
+ +
Q
e–
– –
+
+
+
QH2
e–
e-
+
e–
–
–
+
O2 H2O
–
H+
+
–
–
H+
ADP +
+
–
–
ADP + Pi FADH2 FAD Succinate Complex II
Complex IV ATP Complex III
Complex V
NADH NAD+ Complex I
C
D
Cytochrome c1 Cytochrome b Rieske protein
Carriers MATRIX
INTERMEMBRANE Subunit II
SPACE
F0
Subunit I Subunit III Complex IV Complex III MATRIX
F1 Complex V
FIGURE 19.5 CHEMIOSMOTIC CYCLE OF THE RESPIRATORY ELECTRON TRANSPORT CHAIN AND ADENOSINE TRIPHOSPHATE (ATP) SYNTHASE. A, A mitochondrion for orientation. B, The electron transport system of the inner mitochondrial membrane. Note the pathway of electrons through the four complexes (red and yellow arrows) and the sites of proton translocation between the matrix to the intermembrane space (black arrows). The stoichiometry is not specified, but at the last step, four electrons are required to reduce oxygen to water. The F-type ATP synthase uses the electrochemical proton gradient produced by the electron transport reactions to drive ATP synthesis. FAD, flavin adenine dinucleotide. C, Arrangement of F-type rotary ATP synthases and the electron transport chain in the inner membrane. D, Some atomic structures of the electron transport chain components. In the cytochrome bc1 complex III, the 3 of 11 mitochondrial subunits used by bacteria are shown as ribbon models. The supporting subunits found in mitochondria are shown as cylinders. The four subunits of complex IV encoded by the mitochondrial genome are shown as ribbon models. They form the functional core of the complex, which is supported by additional subunits shown as cylinders. See Fig. 14.4 for further details of ATP synthase (complex V). (D, Images of complex III and complex IV courtesy of M. Saraste, European Molecular Biology Laboratory, Heidelberg, Germany. For reference, see Zhang Z, Huang L, Schulmeister VM, et al. Electron transfer by domain movement in cytochrome bc1. Nature. 1998;392:677–684; and Yoshikawa S, Shinzawa-Itoh K, Nakashima R, et al. Redoxcoupled crystal structural changes in bovine heart cytochrome c oxidase. Science. 1998;280:1723–1729. Also see Protein Data Bank [PDB; www.rcsb.org] file 2OCC.)
more complex. Plasma membranes of bacteria and inner membranes of mitochondria have equivalent components, and the bacterial cytoplasm corresponds to the mitochondrial matrix (Fig. 19.2). Thus, bacteria are useful model systems with which to study the common mechanisms. Energy enters this pathway in the form of electrons that are produced when NADH is oxidized to the oxidized form of nicotinamide adenine dinucleotide (NAD+), releasing one H+ and two electrons (Fig. 19.5B). If the
proton and electrons were to combine immediately with oxygen, their energy would be lost as heat. Instead, these high-energy electrons are separated from the protons and then passed along the electron transport pathway before finally rejoining molecular oxygen to form water. Along the pathway, electrons associate transiently with a series of oxidation–reduction acceptors, generally metal ions associated with organic cofactors, such as hemes in cytochromes and iron-sulfur centers (2Fe2S) and copper centers in complex IV. Electrons
322
SECTION VI n Cellular Organelles and Membrane Trafficking
move along the transport pathway at rates of up to 1000 s−1. To travel at this rate through a transmembrane protein complex spanning a 35-nm lipid bilayer, at least three reduction–oxidation (redox) cofactors are required in each complex, because the efficiency of quantum mechanical tunneling of electrons between redox cofactors falls off rapidly with distance. Two cofactors, even with optimal orientation, would be too slow. Electrons give up energy as they move step by step along the transport pathway. In three complexes along the pathway, this energy is used to pump protons from the matrix to the inner membrane space. This establishes an electrochemical proton gradient across the inner mitochondrial membrane that is used by the F-type rotary ATP synthase to drive ATP production. Direction is provided to the movements of electrons by progressive increases in the electron affinity of the acceptors. The final acceptor, oxygen (at the end of the pathway), has the highest affinity. The first component of the electron transport path way, complex I (or NADH:ubiquinone oxidoreductase), handles electrons obtained from NADH. Vertebrate mitochondrial complex I with 46 different protein subunits is much more complex than bacterial complex I with 14 subunits. NADH donates two electrons to flavin mononucleotide associated with protein subunits located on the matrix side of the inner membrane. A crystal structure of the cytoplasmic domain of the bacterial complex shows the path for the electrons from flavin mononucleotide through seven iron sulfur clusters to quinone in the lipid bilayer. For each molecule of NADH oxidized, the transmembrane domains of complex I transfer four protons from the matrix into the inner membrane space. The second component of the electron transport pathway is complex II or succinate:ubiquinone reductase, a transmembrane enzyme that makes up part of the citric acid cycle. Complex II couples oxidation of succinate (a four-carbon intermediate in the citric acid cycle) to fumarate with reduction of flavin adenine dinucleotide (FAD) to FADH2. Complex II does not pump protons but transfers electrons from FADH2 to ubiquinone. Reduced ubiquinone carries these electrons to complex III. The third component of the electron transport pathway is complex III, also called cytochrome bc1. This well-characterized, transmembrane protein complex consists of 11 different subunits. The homologous bacterial complex has only three of these subunits, the ones that participate in energy transduction in mitochondria. Eight other subunits surround this core. Complex III couples the oxidation and reduction of ubiquinone to the transfer of protons from the matrix across the inner mitochondrial membrane. Energy is supplied by electrons from both complex I and complex II that move through the cytochrome b subunit to a subunit with a 2Fe2S redox center. This subunit then rotates into
position to transfer the electron to cytochrome c1, another subunit of the complex. Cytochrome c1 then transfers the electron to the water-soluble protein cytochrome c in the intermembrane space (or periplasm of bacteria). Cytochrome oxidase, complex IV, takes electrons from four cytochrome c molecules to reduce molecular oxygen to two waters and to pump four protons out of the matrix. Mitochondrial genes encode the three subunits that form the core of this enzyme, carry out electron transfer, and translocate protons. Nuclear genes encode the surrounding 10 subunits. The electrochemical proton gradient produced by the electron transport chain provides energy to synthesize ATP. Chapter 14 explained how the F-type rotary ATP synthase (complex V) can either use ATP hydrolysis to pump protons or use the transit of protons down an electrochemical gradient to synthesize ATP (see Figs. 14.4 and 14.5). The proton gradient across the inner mitochondrial membrane drives rotation of the γ-subunit. The rotating γ-subunit physically changes the conformations of the α- and β-subunits, bringing together ADP and inorganic phosphate to make ATP. An antiporter in the inner membrane exchanges cytoplasmic ADP for ATP synthesized in the matrix (see Fig. 15.4A). Cryoelectron tomography revealed that dimers of F-type rotary ATP synthases are located in rows along the crests of the cristae, where they are responsible for the sharp bend in the inner membrane (Fig. 19.5C). The other components of the electron transfer machinery occupy the flat sides of the cristae. This arrangement of proteins and the small volume inside cristae facilitates the movements of protons and cytochrome c between the components.
Mitochondria and Disease As expected from the central role of mitochondria in energy metabolism, mitochondrial dysfunction contributes to a remarkable diversity of human diseases (Fig. 19.6), including seizures, strokes, optic atrophy, neuropathy, myopathy, cardiomyopathy, hearing loss, and type 2 diabetes mellitus. These disorders arise from mutations in genes for mitochondrial proteins encoded by both mtDNA and nuclear DNA. Many of the known disease-causing mutations are in genes for mitochondrial tRNAs. The existence of approximately 1000 copies of mtDNA per vertebrate cell influences the impact of deleterious mutations. A mutation in one copy would be of no consequence, but segregation of mtDNAs may lead to cells in which mutant mtDNAs predominate, yielding defective proteins. For example, a recurring point mutation in a subunit of complex I causes some patients to develop sudden onset of blindness in middle age owing to the death of neurons in the optic nerve. Patients with the same mutation in a larger fraction of mtDNA
323
CHAPTER 19 n Mitochondria, Chloroplasts, Peroxisomes
A. Disorders due to mutations in nuclear DNA-encoded proteins Complex II Complex III Number of subunits Complex I nDNA-encoded ~35 4 10 Leigh syndrome Leigh syndrome Leukodystrophy Paraganglioma
+ +
Q
e-
– –
+
e–
H+
+
+
+
–
–
–
QH2
SPACE
Complex IV Complex V 10 ~14 Leigh syndrome Cardioencephalomyopathy Leukodystrophy/tubulopathy
Cytochrome c H+
H+
INTERMEMBRANE
ATP
H+
ee-
O2 H2O
Pi +
H+
–
+
–
–
H+
ADP +
+
–
–
ADP + Pi NADH
NAD+
FADH2 FAD Succinate
ATP
B. Disorders due to mutations in mitochondrial DNA-encoded proteins Complex II Number of subunits Complex I 7 0 mtDNA-encoded LHON LHON + dystonia Sporadic myopathy
Complex III 1 Sporadic myopathy
Complex IV 3 Sporadic anemia Sporadic myopathy Encephalomyopathy
Complex V 2 NARP MILS FBSN
Carriers
MATRIX
FIGURE 19.6 Mutations in both mitochondrial and nuclear genes for mitochondrial proteins cause a variety of diseases by compromising the function of particular mitochondrial subsystems. FBSN, familial bilateral striatal necrosis; LHON, Leber hereditary optic neuropathy; MILS, maternally inherited Leigh syndrome; mtDNA, mitochondrial DNA; NARP, neurogenic muscle weakness, ataxia, retinitis pigmentosa; nDNA, nuclear DNA; Pi, inorganic phosphate. (Modified from Schon EA. Mitochondrial genetics and disease. Trends Biochem Sci. 2000;25:555–560.)
molecules suffer from muscle weakness and intellectual disability as children. Mutations in the genes for subunits of ATP synthase cause muscle weakness and degeneration of the retina. Slow accumulation of mutations in mtDNA may contribute to some symptoms of aging. Mutations in mitochondrial DNA are passed from a mother to her children, as sperm mitochondria do not contribute to the embryo. Methods are being developed to combine genetically normal nuclei from affected mothers with enucleated cytoplasm from healthy donors to eliminate these mutations in infants conceived by in vitro fertilization. Mutations in nuclear genes for mitochondrial proteins cause similar diseases (Fig. 19.6A). A mutation in one subunit of the protein import machinery (see Fig. 18.5), Tim8, causes a type of deafness.
Chloroplasts Structure and Evolution of Photosynthesis Systems Photosynthetic bacteria and chloroplasts of algae and plants (Fig. 19.7) use chlorophyll to capture the remarkable amount of energy carried by single photons to boost electrons to an excited state. These high-energy electrons drive a chemiosmotic cycle to make nicotinamide adenine dinucleotide phosphate (NADPH) and ATP. Photosynthetic organisms use ATP and the reducing power of NADPH to synthesize three-carbon sugar phosphates from carbon dioxide. Glycolytic reactions (Fig. 19.4)
running backward use this three-carbon sugar phosphate to make six-carbon sugars and more complex carbohydrates for use as metabolic energy sources and structural components. Some bacteria and Archaea, such as Halobacterium halobium, use a completely different lightdriven pump lacking chlorophyll to generate a proton gradient to synthesize ATP (see Fig. 14.3). In that case, retinol associated with bacteriorhodopsin absorbs light to drive proton transport. Photosynthesis originated approximately 3.5 billion years ago in a bacterium, most likely a gram-negative purple bacterium (see Fig. 2.4). These bacteria evolved components to assemble a transmembrane complex of proteins, pigments, and oxidation/reduction cofactors called a reaction center (Fig. 19.8). Reaction centers absorb light and initiate an electron transport path way that pumps protons out of the cell. Such photosystems turn sunlight into electrical and chemical energy with 40% efficiency, better than any human-made photovoltaic cell. Given their alarming complexity and physical perfection, it is remarkable that photosystems emerged only a few hundred million years after the origin of life itself. Broadly speaking, photosynthetic reaction centers of contemporary organisms can be divided into two different groups (Fig. 19.8). The reaction centers of purple bacteria and green filamentous bacteria use the pigment pheophytin and a quinone as the electron acceptor, similar to photosystem II of cyanobacteria
324
SECTION VI n Cellular Organelles and Membrane Trafficking
A
B. Chloroplast
Thylakoid membrane
Grana
C
THYLAKOID SPACE
STROMA
Thylakoid membrane
Outer membrane
Cell wall Plasma membrane
Starch granule
Inner membrane
CYTOPLASM
E. Cyanobacterium
Thylakoid space
Stroma
PERIPLASM
Outer membrane Inner membrane Stroma
Thylakoid space
D
Thylakoid membrane DNA Ribosomes
Thylakoid space Cytoplasm DNA Ribosomes
FIGURE 19.7 MORPHOLOGY OF CHLOROPLASTS AND CYANOBACTERIA. A, Electron micrograph of a thin section of a spinach chloroplast. B, Chloroplast. C–D, Comparison of the machinery in the photosynthetic membranes of chloroplasts and cyanobacteria. E, Drawing of a cyanobacterium illustrating the internal folds of the plasma membrane to form photosynthetic thylakoids. (A, Courtesy K. Miller, Brown University, Providence, RI.)
and chloroplasts. The reaction centers of green sulfur bacteria and heliobacteria have iron-sulfur centers as electron acceptors, similar to photosystem I of cyanobacteria and chloroplasts. Cyanobacteria are unique among bacteria in that they have both types of photosystems as well as a manganese-containing enzyme that splits water, releasing from two water molecules four electrons, four protons, and oxygen (Fig. 19.7E). Coupling this enzyme to photosynthesis was a pivotal event in the history of the earth, as this reaction is the source of most of the oxygen in the earth’s atmosphere. Chloroplasts of eukaryotic cells arose from a symbiotic cyanobacterium (see Fig. 2.7). Much evidence indicates that this event occurred just once, giving all chloroplasts a common origin. Thereafter chloroplasts moved by lateral transfer to various organisms that diverged prior to the acquisition of chloroplasts, for example, from a green alga to Euglena. Chloroplasts have retained up to 250 original bacterial genes on circular genomes. As in the case of mitochondria, many bacterial genes were lost or moved to the nucleus of host eukaryotes. Chloroplast genomes encode subunits of many proteins responsible for photosynthesis and chloroplast division, ribosomal RNAs and proteins, and a complete set of tRNAs. More than 2000 chloroplast proteins encoded by nuclear genes are synthesized in the cytoplasm and transported posttranslationally into chloroplasts (see Fig. 18.6). The organization of cyanobacterial membranes explains the architecture of chloroplasts (Fig. 19.7C–E). In cyanobacteria, light-absorbing pigments, as well as
protein complexes involved with electron transport and ATP synthesis, are concentrated in invaginations of the plasma membrane. The F1 domain of the F-type rotary ATP synthase faces the cytoplasm, and the lumen of this membrane system is periplasmic. This internal membrane system remains in chloroplasts but is separated from the inner membrane (the former plasma membrane). These thylakoid membranes contain photosynthetic hardware and enclose the thylakoid membrane space. Like the bacterial plasma membrane, the chloroplast “inner membrane” is a permeability barrier, containing carriers for metabolites. The inner membrane surrounds the stroma, the cytoplasm of the original symbiotic bacterium, a protein-rich compartment devoted to synthesis of three-carbon sugar phosphates, chloroplast proteins, and all plant fatty acids. The stroma also houses the genomes and stores starch. The outer membrane, like the comparable bacterial and mitochondrial membranes, has large pore channels that allow free passage of metabolites.
Plastids Chloroplasts are just one manifestation of a class of organelles called plastids. Depending on the developmental stage and tissue type plastids have a range of compositions and functions made possible by selective synthesis and import of proteins. Chloroplasts are specialized for photosynthesis in green plant tissues but differ considerably in composition and physiology in developing and senescent plant tissues. Some plastids, such as the starch-storing amyloplasts in potatoes, lack the photosynthetic machinery.
325
CHAPTER 19 n Mitochondria, Chloroplasts, Peroxisomes
A. Purple bacteria, green filamentous bacteria Cytochrome c2
-1.0
H+
3 H+
+
QB
e-
QH2
ADP + Pi
H+
ATP
Light
PERIPLASM H+
Light
Fes Fes
H+ Light-harvesting complex
NAD reductase
D. Electron energy BChl2*
– –
ATP synthase
2 H+ H Light-harvesting complexes
Fes H+
Fes Ferridoxin
Photosystem II Cytochrome b6 f Photosystem I complex
eNADP NADPH + H+ NADP reductase
0
BChl2 0.5
Green sulfur bacteria
F. Electron energy Chl2*
H+
-1.0
+
QH2
FX FA/B NAD
1.0
LUMEN
Plastocyanin
H+
+ +
– –
–
Chl
-0.5
CYTOPLASM
e-
e-
BChl2
Purple bacteria
Light
QB
0.5
+ +
PERIPLASM/
Light Mn2+
QA
QB Cytochrome b c1 Cytochrome c2
ATP
E. Cyanobacteria, algae, plants 2 H2O
QA 0
ADP + Pi
eNAD NADH + H+
Cytochrome b c Type I complex photosystem
H+
–
H+
Ferridoxin
3 H+
BChl BPhe
-1.0 +
4 H+ + O2
-0.5
1.0
ATP synthase CYTOPLASM
C. Green sulfur bacteria Heliobacteria Cytochrome c2 H+
– –
–
2 H+ H Light-harvesting complex Cytochrome b c1 Type II photosystem complex
+ +
ADP + Pi ATP
ATP synthase STROMA/ CYTOPLASM
Energy (volts)
QA
H+
BChl2*
Energy (volts)
Light
B. Electron energy
Energy (volts)
Cytochrome
PERIPLASM
Chl2* -0.5
Phe QA QB
0
0.5
H2O
1.0
Chl2
Chl Q
FX FA/B NADP
Chl2
YZ
Chloroplasts and cyanobacteria Photosystem II Photosystem I
FIGURE 19.8 COMPARISON OF PHOTOSYNTHETIC COMPONENTS, ELECTRON TRANSPORT PATHWAYS, AND CHEMIOSMOTIC CYCLES TO MAKE ADENOSINE TRIPHOSPHATE. A–B, Type II photosystem only. C–D, Type I photosystem only. E–F, Both photosystem II and photosystem I. Right diagrams, The energy levels of electrons in the three types of photosynthetic organisms, showing excitation of an electron by an absorbed photon (vertical arrows), electron transfer pathways through each reaction center (arrows sloping right), and electron transfer steps outside the reaction centers (arrows sloping left). (A, C, and E, For reference, see Kramer DM, Schoepp B, Liebl U, et al. Cyclic electron transfer in Heliobacillus mobilis. Biochemistry. 1997;36:4203–4211. B, D, and F, For reference, see Allen JP, Williams JC. Photosynthetic reaction centers. FEBS Lett. 1998;438:5–9.)
Light and Dark Reactions Photosynthetic mechanisms capture energy from photons to drive two types of reactions: • Light reactions depend on continuous absorption of photons. These reactions occur in or on the surface
of thylakoid membranes. They include the generation of high-energy electrons, electron transport to make NADPH, creation of a proton gradient across the thylakoid membrane for the chemiosmotic synthesis of ATP, and generation of oxygen.
326
SECTION VI n Cellular Organelles and Membrane Trafficking
• Dark reactions convert carbon dioxide into threecarbon sugar phosphates. These reactions continue for some time in the dark. However, they depend on ATP and NADPH produced by light reactions, so they eventually stop when ATP and NADPH are exhausted in the dark. These reactions account for most of the carbon dioxide converted to carbohydrates on earth. (Alternatively specialized prokaryotes drive carbon fixation by oxidation of hydrogen sulfide and other inorganic compounds.) All photosynthetic systems use similar mechanisms to capture energy from photons (Fig. 19.8). Pigments associated with transmembrane proteins in photosynthetic reaction centers absorb photons and use the energy to boost electrons to a high-energy excited state. Subsequent electron transfer reactions partition this energy in several steps to generate a proton gradient across the membrane. Generation of this proton electrochemical gradient and chemiosmotic production of ATP are similar to oxidative phosphorylation (Fig. 19.5). Specific photosynthetic systems differ in the complexity of the hardware, the source of electrons, and the products (Fig. 19.8). Most photosynthetic bacteria use either a type I photosystem or a type II photosystem to create a proton gradient to synthesize ATP. Cyanobacteria and green plants use both types of reaction centers in series to raise electrons to an energy sufficient to make NADPH in addition to ATP. These advanced systems also use water as the electron donor and produce molecular oxygen as a by-product.
Energy Capture and Transduction by Photosystem II The reaction center from the purple bacterium Rhodopseudomonas viridis (Fig. 19.9A) serves as a model for the more complex photosystem II of cyanobacteria and chloroplasts. This bacterial reaction center consists of just four subunits. A cytochrome subunit on the periplasmic side of the membrane donates electrons. Two core subunits form a rigid transmembrane framework to bind 10 cofactors in orientations that favor transfer of high-energy electrons from two “special” bacteriochlorophylls through chlorophyll b and bacteriopheophytin b. Photosynthesis begins with absorption of a photon by the special pair bacteriochlorophylls. Photons in the visible part of the spectrum are quite energetic, 40 to 80 kcal mol−1, enough to make several ATPs. The purple bacterium reaction center absorbs relatively lowenergy, 870-nm red light. The energy elevates an electron in the special pair bacteriochlorophylls to an excited state (Fig. 19.8B). This excited state can decay rapidly (109 s−1), causing the energy to dissipate as heat or emission of a less-energetic photon by fluorescence or phosphorescence. However, reaction centers are optimized to transfer excited-state electrons rapidly and efficiently from the special pair bacteriochlorophylls to
bacteriopheophytin (3 × 10−12 s) and then to tightly bound quinone A (200 × 10−12 s). Transfer is by quantum mechanical tunneling right through the protein molecule. Because the tunneling rate falls off sharply with distance, four redox centers must be spaced close together to allow an energetic electron to transfer across the lipid bilayer faster than spontaneous decay of the excited state. On the cytoplasmic side of the membrane, two electrons transfer from quinone A to quinone B (100 × 10−9 s), where they combine with two protons to make a high-energy reduced quinone, QH2 (Fig. 19.8A). In purple bacteria, these cytoplasmic protons are taken up through water-filled channels in the reaction center, contributing to the proton gradient. QH2 has a low affinity for the reaction center and diffuses in the hydrophobic core of the bilayer to the next component in the pathway, the chloroplast equivalent of the mitochondrial cytochrome bc1 complex III (Fig. 19.8A). As in mitochondria, passage of energetic electrons through this complex releases protons from QH2 on the periplasmic side of the membrane, adding to the electrochemical gradient. The electron circuit is completed by transfer of low-energy electrons from complex bc1 to a soluble periplasmic protein, cytochrome c2. Electrons then move to the cytochrome subunit of the reaction center, which supplies special pair chlorophylls with electrons for the photosynthetic reaction cycle. The net result of this cycle is the conversion of the energy of two photons into transport of three protons to the periplasm. A diagram of the energy levels of the various intermediates in the cycle (Fig. 19.8B) shows how energy is partitioned after an electron is excited by a photon and then moves, step by step, through proteinassociated redox centers back to the ground state. The proton electrochemical gradient established by photosynthetic electron transfer reactions is used to drive an F-type rotary ATP synthase (see Fig. 14.5) similar to those of nonphotosynthetic prokaryotes and mitochondria.
Light Harvesting Reaction center chlorophylls absorb light, but both chloroplasts and bacteria increase the efficiency of light collection with proteins that absorb light and transfer the energy to a reaction center. Most of these lightharvesting complexes are small, transmembrane proteins that cluster around a reaction center, although some bacteria and algae also have soluble light-harvesting proteins. Transmembrane, light-harvesting proteins con sist of a few α-helices associated with multiple chlorophyll and carotenoid pigments (Figs. 19.8A and C and 19.9B). Using multiple pigments broadens the range of wavelengths absorbed and increases the efficiency of photon capture. Leaves are green because chlorophylls
327
CHAPTER 19 n Mitochondria, Chloroplasts, Peroxisomes
A. Purple bacterium type II photosystem reaction center
B. Cyanobacterium type I photosystem
Cytochrome Hemes
L
Electron pathway
PsaM
PsaA/B PsaK
PsaF/J
Clb
eC1 Car
eC2 eC3
Phb QB M
Fe
QA
H CYTOPLASM
FX
PsaE
PsaL/I PsaD PsaC
FIGURE 19.9 STRUCTURES OF PHOTOSYSTEM HARDWARE. A, Ribbon diagram of type II photosystem from the purple bacterium Rhodopseudomonas viridis, with ball-and-stick models of bacteriochlorophyll and other cofactors to the right in their natural orientations. Similar core subunits L and M each consists of five transmembrane helices. This pair of subunits binds four molecules of chlorophyll b (Clb), two molecules of bacteriopheophytin b (Phb), one nonheme iron (Fe), two quinones (QA, QB), and one carotenoid (Car) in a rigid framework. A cytochrome with four heme groups binds to the periplasmic side of the core subunits. Subunit H associates with the core subunits via one transmembrane helix and with their cytoplasmic surfaces. The atomic structure of this photosynthetic reaction center was the Nobel Prize work of J. Diesenhofer, R. Huber, and H. Michel. B, Ribbon diagram of photosystem I of Synechococcus elongatus, with ball-and-stick models of chlorophyll and other cofactors to the right in their natural orientations. This trimeric complex consists of three identical units, each composed of 11 polypeptide chains. Within each of these units, this 4-Å resolution structure includes 43 α-helices, 89 chlorophylls, 1 quinone, and 3 iron-sulfur centers, but other details (eg, amino acid side chains) are not resolved. The photosynthetic reaction center consists of the C-terminal halves of the two central subunits (PsaA/PsaB, red-brown) associated with six chlorophylls, one or two quinones, and a shared iron-sulfur cluster. Plastocyanin or cytochrome c6 on the lumen side donates electrons to reduce the P700 special pair chlorophylls (eC1) of the reaction center. Light energizes an electron, which passes successively through two other chlorophylls, a quinone, and the shared iron-sulfur cluster (red), Fx. The electron then transfers to the iron-sulfur clusters of the accessory subunit PsaC on the stromal side of the membrane. The surrounding eight subunits (red, gray), associated with approximately 80 chlorophylls, compose the core antenna system, forming a nearly continuous ring of α-helices around the reaction center. Absorption of light by additional light-harvesting complexes and these antenna subunits puts chloroplast electrons into an excited state. This energy passes from one pigment to the next until it eventually reaches the reaction center. (A, Copyright Diesenhofer & Michel, Nobel Foundation, 1988. For reference, see PDB file 1PRC and Diesenhofer J, Michel H. The photosynthetic reaction center from the purple bacterium Rhodopseudomonas viridis. Science. 1989;245:1463–1473. A 3.5-Å crystal structure of the photosystem II complex from the cyanobacterium Thermosynechococcus elongatus, including 19 subunits, is now available. Also see Ferreira KN, Iverson TM, Maghlaoui K, et al. Architecture of the photosynthetic oxygen-evolving center. Science. 2004;303:1831–1838. B, For reference, see PDB file 2PPS and Schubert W-D, Klukas O, Krauss N, et al. Photosystem I of Synechococcus elongatus at 4 Å resolution: comprehensive structure analysis. J Mol Biol. 1997;272:741–769.)
and carotenoids absorb purple and blue wavelengths (620 nm), reflecting only yellow-green wavelengths in between. Light absorbed by light-harvesting proteins boosts pigment electrons to an excited state. This energy (but not the electrons) moves without dissipation by fluorescence resonance energy transfer from one closely spaced pigment molecule to another and eventually to the special pair chlorophylls of a reaction center. This rapid (10−12 s), efficient process transfers energy captured over a wide area to a reaction center to initiate a cycle of electron transfer and energy transduction.
Energy Capture and Transduction by Photosystem I The reaction centers of green sulfur bacteria and heliobacteria are similar to photosystem I of cyanobacteria and chloroplasts. Generation of a proton gradient by photosystem I has many parallels with photosystem II. Direct absorption of light or resonance energy transfer from surrounding light-harvesting complexes excites special-pair chlorophylls in photosystem I (Fig. 19.8C– D). Excited-state electrons move rapidly within the reaction center from these chlorophylls through two accessory chlorophylls to an iron-sulfur center. The pathway includes a quinone in cyanobacteria and
328
SECTION VI n Cellular Organelles and Membrane Trafficking
chloroplasts. Electrons then move to the iron-sulfur center of a subunit on the cytoplasmic side of the membrane. The subsequent events in green sulfur bacteria and heliobacteria include electron transfer by the soluble protein ferredoxin to an NAD reductase, followed by transfer by a lipid intermediate to cytochrome bc complex, and then back to the reaction center via a cytochrome c.
Oxygen-Producing Synthesis of NADPH and ATP by Dual Photosystems Chloroplasts and cyanobacteria combine photosystem II and photosystem I in the same membrane to form a system capable of accepting low-energy electrons from the oxidation of water and producing both a proton gradient to drive ATP synthesis and reducing equivalents in the form of NADPH (Fig. 19.8E–F). Both photosystems are more elaborate in dual systems than in single systems. Although plant photosystem II, with more than 25 protein subunits, is much more complicated than is the homologous reaction center of purple bacteria, the arrangement of transmembrane helices and chlorophyll cofactors in the core of the plant reaction center is similar to the simple reaction center of purple bacteria. Photosynthesis involves a tortuous electron transfer pathway powered at two waystations by absorption of photons. This process begins when the special pair chlorophylls of photosystem II are excited by direct absorption of light or by resonance energy transfer from surrounding light-harvesting complexes (Fig. 19.8E– F). Electrons come from splitting two waters into molecular oxygen and four protons. Excited-state electrons tunnel through the redox cofactors and combine with protons from the stroma (or cytoplasm in bacteria) to reduce quinone QB to QH2, a high-energy electron donor. QH2 diffuses to complex b6-f, the chloroplast equivalent of the mitochondrial bc1 complex. Passage of electrons through complex b6-f releases protons from QH2 into the thylakoid lumen (or bacterial periplasm), contributing to the proton gradient across the membrane. Complex b6-f donates electrons from QH2 to photosystem I. Direct absorption of 680-nm light or resonance energy transfer from surrounding light-harvesting complexes boosts special pair chlorophyll electrons to a very high-energy, excited state (Fig. 19.8F). Excited-state electrons pass through chlorophyll and iron-sulfur centers of photosystem I to the iron-sulfur center of the redox protein, ferredoxin, on the cytoplasmic/stromal surface of the membrane. The enzyme nicotinamide adenine dinucleotide phosphate (NADP) reductase combines electrons from ferredoxin with a proton to form NADPH, the final product of this complex electron transfer pathway powered at two waystations by absorption of photons. Uptake of stromal protons during NADPH formation contributes to the transmembrane proton
gradient for the synthesis of ATP. Antiporters in the inner membrane exchange ATP for ADP, as in mitochondria.
Synthesis of Carbohydrates ATP and NADPH produced by light reactions drive the unfavorable conversion of carbon dioxide into sugars. This is the first step in the earth’s annual production of approximately 1010 tons of carbohydrates by photosynthetic organisms. This process is very expensive, consuming three ATPs and two NADPHs for each carbon dioxide added to the five-carbon sugar ribulose 1,5-bisphosphate. The responsible enzyme, ribulose phosphate carboxylase (called RUBISCO), is the most abundant protein in the stroma and might be the most abundant protein on the earth. The products of combining the five-carbon sugar with carbon dioxide are two molecules of the three-carbon sugar 3-phosphoglycerate. An antiporter in the inner chloroplast membrane exchanges 3-phosphoglycerate for inorganic phosphate, so 3-phosphoglycerate can join the glycolytic pathway in the cytoplasm (Fig. 19.4). Driven by this abundant supply of 3-phosphoglycerate, the glycolytic pathway runs backward to make six-carbon sugars, which are used to make disaccharides such as sucrose to nourish nonphotosynthetic parts of the plant, the glucose polymer starch to store carbohydrate, and cellulose for the extracellular matrix (see Figs. 3.25A and 32.13).
Peroxisomes Peroxisomes are organelles bounded by a single membrane (Fig. 19.10), named for their content of enzymes that produce and degrade hydrogen peroxide, H2O2. Oxidases produce H2O2 and peroxidases such as catalase break it down. Peroxisomes also contain diverse enzymes for the metabolism of lipids and other metabolites, including the β-oxidation of fatty acids and oxidation of bile acids and cholesterol. All peroxisomal proteins are encoded by nuclear genes, translated on cytoplasmic ribosomes, and then subsequently incorporated into peroxisomes (see Fig. 18.8). Peroxisomes form in two different ways: de novo synthesis by budding from the ER and growth and division of preexisting peroxisomes (see Fig. 18.8). Cells that lack preexisting peroxisomes can form peroxisomes without a template by differentiation and budding of ER membranes. Two key proteins known as peroxins, PEX3 and PEX16, are targeted to the ER, where they recruit other peroxins to form a specialized domain that pinches off to form a nascent peroxisome. Defects in peroxisomal biogenesis cause a spectrum of lethal human diseases known as the peroxisomal biogenesis disorders (see Table 18.1). These diseases include Zellweger syndrome, neonatal adrenoleukodystrophy, infantile Refsum disease, and rhizomelic
CHAPTER 19 n Mitochondria, Chloroplasts, Peroxisomes
A
329
B
FIGURE 19.10 PEROXISOMES. A, Fluorescence micrographs of a CV1 cell expressing green fluorescent protein fused to PTS1, which labels peroxisomes green. Microtubules are stained red with labeled antibodies, and nuclear DNA is stained blue with propidium iodide. B, Electron micrograph of a thin section of a tissue culture cell showing three peroxisomes. Peroxisomes have a single bilayer membrane and a dense matrix, including a crystal (in some species) of the enzyme urate oxidase. (A, Courtesy S. Subramani, University of California–San Diego. For reference, see Wiemer EAC, Wenzel T, Deernick TJ, et al. Visualization of the peroxisomal compartment in living mammalian cells. J Cell Biol. 1997;136:71– 80. B, Courtesy Don W. Fawcett, Harvard Medical School, Boston, MA.)
chondrodysplasia punctata. These diseases are moderately rare, occurring in approximately 1 in 50,000 live births. Most patients with peroxisomal biogenesis disorders display no defect in peroxisome membrane synthesis or import of peroxisomal membrane proteins, but they do have mild-to-severe defects in matrix protein import. However, in rare cases, patients lack peroxisome membranes altogether. Studies of both yeast pex mutants and cells from patients with peroxisomal biogenesis disorders have provided clues regarding peroxisome biogenesis (see Table 18.1). SELECTED READINGS Chacinska A, Koehler CM, Milenkovic D, et al. Importing mitochondrial proteins: machineries and mechanisms. Cell. 2009;138: 628-644. Demarsy E, Lakshmanan AM, Kessler F. Border control: selectivity of chloroplast protein import and regulation at the TOC-complex. Front Plant Sci. 2013;5:483. Emma F, Montini G, Parikh SM, et al. Mitochondrial dysfunction in inherited renal disease and acute kidney injury. Nat Rev Nephrol. 2016;doi: 10.1038/nrneph.2015.214; [Epub ahead of print]. Hohmann-Marriott MF, Blankenship RE. Evolution of photosynthesis. Annu Rev Plant Biol. 2011;62:515-548.
Hosler JP, Ferguson-Miller S, Mills DA. Energy transduction: Proton transfer through the respiratory complexes. Annu Rev Biochem. 2006;75:165-187. Jarvis P, López-Juez E. Biogenesis and homeostasis of chloroplasts and other plastids. Nat Rev Mol Cell Biol. 2013;14:787-802. Keeling PJ. The number, speed, and impact of plastid endosymbiosis in eukaryotic evolution. Annu Rev Plant Biol. 2013;64:583-607. Kühlbrandt W. Structure and function of mitochondrial membrane protein complexes. BMC Biol. 2015;13:89. Kühlbrandt W, Davies KM. Rotary ATPases: A new twist to an ancient machine. Trends Biochem Sci. 2016;41:106-116. Labbé K, Murley A, Nunnari J. Determinants and functions of mitochondrial behavior. Annu Rev Cell Dev Biol. 2014;30:357-391. Nelson N, Junge W. Structure and energy transfer in photosystems of oxygenic photosynthesis. Annu Rev Biochem. 2015;84:659-683. Pfanner N, van der Laan M, Amati P, et al. Uniform nomenclature for the mitochondrial contact site and cristae organizing system. J Cell Biol. 2014;204:1083-1086. Poole AM, Gribaldo S. Eukaryotic origins: How and when was the mitochondrion acquired? Cold Spring Harb Perspect Biol. 2014;6: a015990. Schon EA, DiMauro S, Hirano M. Human mitochondrial DNA: roles of inherited and somatic mutations. Nat Rev Genet. 2012;13:878-890. Smith JJ, Aitchison JD. Peroxisomes take shape. Nat Rev Mol Cell Biol. 2013;14:803-817.
This page intentionally left blank
CHAPTER
20
Endoplasmic Reticulum T
he endoplasmic reticulum (ER) is the largest membrane-delineated intracellular compartment within eukaryotic cells, having a surface area up to 30 times that of the plasma membrane (Fig. 20.1). The ER performs many essential cellular functions, including protein synthesis and processing, lipid synthesis, compartmentalization of the nucleus, calcium (Ca2+) storage and release, detoxification of compounds, and lipid transfer and signaling to other organelles (Table 20.1). It also has roles in the biogenesis of the Golgi apparatus, peroxisomes and lipid droplets, and helps mitochondria to divide. Approximately one-third of all cellular proteins are imported into the lumen of the ER or integrated into ER membranes. Import occurs at rates of 2 to 13 million new proteins synthesized per minute. The ER retains some of these imported proteins for its own functions, some are degraded, and others are exported into the secretory pathway (see Chapter 21) for targeting to other compartments within the cell. A
B
The ER is organized as an extensive array of tubules and flat saccules called cisternae (cisterna means “reservoir”) that form an interconnected and contiguous threedimensional network (a reticulum) stretching from the nuclear envelope to the cell surface. This system has several structural domains. The ER that flattens around the cell nucleus to form a double membrane bilayer barrier is called the nuclear envelope (see Fig. 9.5). The peripheral ER that extends from the nuclear envelope is comprised of both a polygonal network of tubules and flat, stacked membrane cisternae close to the nucleus. The stacked cisternae are covered with ribosomes for the synthesis, import, and folding of membrane, luminal, and secreted proteins. The tubule network extends throughout the cytoplasm (Fig. 20.1C); its functions include lipid synthesis, Ca2+ storage and release, and making contacts with the membranes of other organelles and the plasma membrane. This chapter describes (a) the overall functions and organization of the ER, (b) insertion of proteins into and C
Nuclear pore Stacked cisterane
Nuclear envelope Stacked cisterane
Ribosomes Peripheral sheet Tubules
FIGURE 20.1 OVERVIEW OF THE ENDOPLASMIC RETICULUM/NUCLEAR ENVELOPE. A, Diagram of the endoplasmic reticulum (ER) and nuclear envelope. B–C, Fluorescence micrographs of cells expressing an ER marker tagged with green fluorescent protein (white in this image). B, The expansive character of ER is emphasized. C, Peripheral ER tubules. (A, From Goyal U, Blackstone C. Untangling the web: mechanisms underlying ER network formation. Biochim Biophys Acta. 2013;1833:2492–2498. B–C, Courtesy Drs. Chris Obara and Aubrey Weigel, Janelia Research Campus, Ashburn, VA.)
331
332
SECTION VI n Cellular Organelles and Membrane Trafficking
TABLE 20.1 Subdomains of the Endoplasmic Reticulum
A
ER Domain
Function
Associated Proteins
Rough ER
Protein translocation Protein folding and oligomerization Carbohydrate addition ER degradation
Sec61 complex, TRAP, TRAM, BiP PDI, Calnexin, Calreticulin, BiP Oligosaccharide transferase EDEM, Derlin1
Smooth ER
Detoxification Lipid metabolism Heme metabolism Calcium release
Cytochrome P450 enzymes HMG-CoA reductase Cytochrome b5 IP3 receptors
Nuclear envelope
Nuclear pores Chromatin anchoring
POM121, GP210 Lamin B receptor
ER export sites
Export of proteins and lipids into secretory pathway
Sar1p, Sec12p, Sec16p
ER contact zones
Transport of lipids
LTPs
BiP, biding immunoglobulin protein; EDEM, ER degradation-enhancing α-mannosidase-like protein; ER, endoplasmic reticulum; HMG-CoA, β-hydroxy-β-methylglutaryl-coenzyme A; IP3, inositol triphosphate; LTP, lipid-transfer protein; PDI, protein disulfide isomerase; TRAM, translocating chain-associating membrane protein; TRAP, transloconassociated protein.
across the ER membrane, (c) the mechanisms of folding, assembly, and degradation of proteins in the ER, and (d) the synthesis and metabolism of lipids by the ER. Chapter 21 covers the secretory pathway, which begins at the ER. Chapter 26 explains how the ER stores and releases Ca2+.
Overview of Endoplasmic Reticulum Functions and Organization The foremost function of the ER is producing most proteins that are secreted from the cell as well as lipids that make up the membranes of the other organelles, including the Golgi apparatus, endosomes, lysosomes, and plasma membrane. The ER also supplies much of the lipid for the membranes of mitochondria and peroxisomes. The area of ER specialized for protein synthesis, folding, and degradation is called the rough ER, because its cytoplasmic surface is studded with ribosomes (Fig. 20.2B). The rough ER contains specialized receptors and channels that transfer proteins synthesized by ribosomes in the cytoplasm across ER membranes. Inside the ER lumen, newly synthesized proteins are exposed to a dense meshwork of chaperones and other modifying enzymes (estimated to be 200 mg/mL) that catalyze their folding and assembly. Proteins that are incorrectly folded or misfolded can be exported back into the cytoplasm, where they are degraded. Misfolded proteins, when
CYTOPLASM
rER
NUCLEUS
sER
ER export domain Golgi apparatus
Nuclear envelope
B
FIGURE 20.2 ENDOPLASMIC RETICULUM SUBDOMAINS. A, Drawing of a cell with specialized regions of the endoplasmic reticulum (ER). Rough ER (rER) with bound ribosomes extends from the nuclear envelope to the cell periphery. Smooth ER (sER) without ribosomes is specialized for drug metabolism and steroid synthesis and includes tubulovesicular elements composing ER exit sites. The nuclear envelope consists of ER membrane wrapped around the chromosomes and other nuclear elements. B, Electron micrograph of a thin section of rough ER and neighboring mitochondrion from the pancreas. The lumen is colored blue. (Micrograph by Keith R. Porter; courtesy Don W. Fawcett, Harvard Medical School, Boston, MA.)
accumulated in the ER at high levels, can trigger an unfolded protein response. This activates specific genes in the nucleus whose products help modify or destroy the misfolded proteins and compensate for the decreased capacity of ER folding. Both soluble and transmembrane proteins are exported from the ER at sites called ER export domains. These tubulovesicular membranes lack ribosomes and bud off vesicle intermediates for delivery to the Golgi apparatus (see Chapter 21). The surface of the ER forming the outer nuclear envelope, which faces the cytoplasm, is indistinguishable from the rest of the ER except for the presence of nuclear pores that allow passage of molecules between the nucleus and cytoplasm (see Fig. 9.18). By contrast, the ER surface forming the inner nuclear envelope, which faces the nucleoplasm, contains specialized proteins that interact with the nuclear lamina and chromatin (see Fig. 9.8). In mitosis, the ER maintains its morphology as an interconnected network, whereas the nuclear envelope either disassembles (in most animal cells; see Fig. 44.6) or remains intact (in yeasts and most other fungi). In cells where the nuclear envelope
CHAPTER 20 n Endoplasmic Reticulum
The smooth ER, composed of tubular elements lacking ribosomes, is dedicated to enzyme pathways involved in drug metabolism (hepatocytes), steroid synthesis (endocrine cells), or calcium uptake and release (see Fig. 26.12). The cytochrome P450 family of hemecontaining membrane proteins resides in the smooth ER. These enzymes use an electron transfer process to detoxify endogenous steroids, carcinogenic compounds, lipid-soluble drugs, and environmental xenobiotics. The lumen of the ER is an oxidizing environment that favors disulfide bond formation, which helps stabilize proteins that are exported from the ER to the outside of the cell. The abundance of a particular ER region varies in specialized cells. Cells dedicated to the production, storage, and regulated secretion of proteins (such as exocrine cells and activated B cells) are rich in rough ER. By contrast, smooth ER is abundant in endocrine cells that synthesize steroid hormones and in muscle cells owing to their requirement to store and release Ca2+ to control contraction.
disassembles during mitosis, nuclear pores disassemble and integral membrane proteins of the nuclear envelope diffuse into surrounding ER membranes. Peroxisomes, lipid droplets, and the Golgi apparatus all depend on the ER for their biogenesis and maintenance. During peroxisome biogenesis, the ER provides the initial scaffold for recruiting core components (ie, Pex16p and Pex3p) involved in peroxisomal protein import (see Fig. 18.8). Biogenesis of the Golgi apparatus depends on the ER to synthesize resident enzymes and on constitutive cycling of membrane between the ER and Golgi apparatus. Consequently, perturbing the export of proteins from the ER impacts the structure and function of the Golgi apparatus. Lipid droplets emerge directly from the ER by the accumulation of neutral lipids between the leaflets of the ER bilayer. This forms an oil droplet that eventually buds toward the cytoplasm. Subsequent growth of lipid droplets may take place by neutral lipid synthesis either on their surface or at the ER. The ER lumen is the major Ca2+ storage site in cells, owing to calcium pumps in the membrane (see Figs. 14.7 and 26.12) and many Ca2+-binding proteins in the lumen. The second messenger IP3 (inositol triphosphate) releases Ca2+ from the lumen by activating calcium release channels (see Fig. 26.12). Carefully regulated release and uptake of Ca2+ by the ER control muscle contraction (see Fig. 39.15) and many other cellular processes. The ER can also release Ca2+ next to neighboring organelles at ER–organelle contact sites. A
333
Endoplasmic Reticulum Shape Generation Several classes of proteins determine the unique morphology of the ER as it undergoes continual rearrangements. These proteins ensure that the ER maintains itself as a single polygonal network of tubules and stacked cisternae extending from the nuclear envelope to the cell periphery (Fig. 20.3). Proteins that mediate ER tubule formation include the reticulons and DP1/Yop1 proteins B
GTPase
Atlastin/Sey1p
Reticulons or DP1/Yop1p
CYTOSOL
ER tubules
LIPID ER LUMEN
C Reticulons
SMOOTH ER TUBULE
BILAYER
D GTP hydrolysis Membrane fusion
ER LUMEN
Atlastins FIGURE 20.3 ENDOPLASMIC RETICULUM SHAPING AND FUSION PROTEINS. A, Superresolution fluorescence micrograph of the peripheral network of tubular ER membranes stained with an antibody to reticulon. See Fig. 6.5 for details. B, Insertion of reticulons and atlastin into the ER membrane bilayer. C, Drawings of tubular ER membranes showing reticulon proteins in the cytoplasmic leaflet of the bilayer and atlastin proteins connecting two tubules. D, Proposal for the fusion of ER membranes by atlastin with energy coming from guanosine triphosphate (GTP) hydrolysis. (B, Modified from Chen S, Novick P, Ferro-Novick S. ER structure and function. Curr Opin Cell Biol. 2013;25:428–433. C–D, Modified from Goyal U, Blackstone C. Untangling the web: mechanisms underlying ER network formation. Biochim Biophys Acta. 2013;1833:2492–2498. D, Modified from Lin S, Sun S, Hu J. Molecular basis for sculpting the endoplasmic reticulum membrane. Int J Biochem Cell Biol. 2012;44:1436–1443.)
334
SECTION VI n Cellular Organelles and Membrane Trafficking
(Fig. 20.3B–C). Although not related by sequence, these proteins share hydrophobic segments predicted to form α-helical hairpins that partially span the lipid bilayer (Fig. 20.3B). Insertion of these hydrophobic segments into the cytoplasmic leaflet of the ER membrane bilayer is thought to help generate the highly curved, tubular ER morphology, along with directly scaffolding reticulons on the surface of the ER. Reticulons and DP1/YOP1 proteins also participate in nuclear pore formation, most likely by stabilizing curved membranes. Members of the atlastin/RHD3/Sey1p family of dynamin-related guanosine triphosphatases (GTPases) localize to highly curved ER membranes and mediate the formation of three-way junctions responsible for the polygonal structure of the tubular ER network (Fig. 20.3B–C). Atlastins consist of an N-terminal cytoplasmic GTPase domain and a three-helix bundle followed by two very closely spaced transmembrane segments and a C-terminal amphipathic helix. Guanosine triphosphate (GTP) binding stimulates interactions between atlastin oligomers in two adjacent membranes, forming a tethered complex. Fusion between ER tubules depends on a conformational change in the cytosolic domain linked to GTP hydrolysis that pulls the membranes together, leading to oligomerization of the transmembrane segments, and interaction of the C-terminal tail with the membranes undergoing fusion. After membrane fusion, atlastin releases GDP before another round of membrane fusion. Interactions with microtubules can remodel the ER in two ways: motor proteins can pull ER along the side of a microtubule; or the ER membrane can attach to +TIP attachment complexes that track the end of a growing microtubule (see Fig. 37.6). One example of TIP tracking uses the transmembrane ER protein stromal interaction molecule 1 (STIM1) (see Fig. 26.12), which binds directly to the +TIP protein EB1 (see Fig. 34.4C). This interaction allows ER tubules containing STIM1 to reach the plasma membrane and activate Orai (see Fig. 26.13A), the storeoperated Ca2+-channel in the plasma membrane that admits Ca2+ to refill calcium stores in the ER. Interactions with the actin cytoskeleton can also drive ER remodeling. This is particularly important in plants (see Fig. 37.9) and yeast cells (see Fig. 37.11), but also occurs in animal cells such as neurons, where myosin-Va transports ER along actin filaments (see Fig. 36.8) into the dendritic spines of neurons. Defects in proteins that control the shape of the ER often lead to disease. For example, autosomal dominant mutations in atlastin or reticulons result in lengthdependent degeneration of the distal parts of the axons of corticospinal upper motor neurons in hereditary spastic paraplegias, and contribute to the pathogenesis of amyotrophic lateral sclerosis, which involves degeneration of both upper and lower motor neurons. The large size and highly polarized geometry of neurons
makes shaping and distributing the ER network especially important in these cells.
Endoplasmic Reticulum– Organelle Contacts The ER makes many types of contacts with other membranes, which are mediated by specific proteins. Chapter 26 describes contacts between Ca2+-sensing proteins in the ER membrane with plasma membrane Ca2+ channels that cells use to resupply the ER with Ca2+ (see Fig. 26.12). Other membrane contact sites (Fig. 20.4 and Table 20.2) depend on a conserved receptor on the cytoplasmic surface of the ER called VAP (for VAMP [vesicle-associated membrane protein]-associated protein; see Fig. 20.17). The cytoplasmic domain of VAP binds a wide range of proteins with an FFAT motif containing two phenylalanines (FF) in an acidic tract. FFAT proteins link the ER to mitochondria, endosomes, Golgi, peroxisomes or the plasma membrane. Within the gap of 10 to 30 nm between the organelles, tethered proteins with lipid binding domains transfer lipid molecules between the lipid bilayers (see Fig. 20.17). Mutations of one of the two genes for VAP cause some cases of neurodegenerative diseases (amyotrophic lateral sclerosis, Parkinson disease). Membrane contact sites are also important for Ca2+ signaling. Because Ca2+ diffuses only about 100 nm from its release site (see Chapter 26), Ca2+ signals from the ER to other organelles are more efficient at tight interfaces. ER–organelle contacts also control organelle division and many aspects of cytoplasmic and plasma membrane organization.
Pla
em am sm
ne bra
Endosome Lysosome
Mitochondrion Lipid droplet Golgi
Peroxisome
ER
NUCLEUS FIGURE 20.4 ENDOPLASMIC RETICULUM–ORGANELLE CONTACT SITES. Drawing of a cell showing contacts of the endoplasmic reticulum (ER) with endosomes, Golgi apparatus, lipid droplets, mitochondria, peroxisomes, and plasma membrane.
CHAPTER 20 n Endoplasmic Reticulum
335
TABLE 20.2 Location and Proposed Functions of Proteins at Endoplasmic Reticulum–Organelle Contacts ER-Mitochondria Contacts MFN1-MFN2
Calcium transfer
VAPs-PTPIP51
Lipid transfer
ERMES complex
Tethers and possibly transfers lipids
ER-Endosome VAP-A-ORP1L
Senses sterol levels and regulates endosome positioning
VAP-A-STARD3
Transfers sterols
VAP-A-Protrudin
Transfers kinesin-1 from the ER to late endosomes to facilitate their to the cell periphery
ORP5-NPC1
ORD domain of ORP5 transfers cholesterol
ER-Golgi Contacts VAP-OSBP
Transfers PtdIns(4)P from the Golgi apparatus to ER and sterols from ER to Golgi apparatus
VAP-CERT
Transfers ceramide
VAP-NIR2
Maintains diacylglycerol levels in the Golgi
ER-Lipid Droplet Contacts FATP1-DGAT2
Coordinates lipid droplet expansion at lipid droplet-ER MCSs
CERT, ceramide transport protein; ER, endoplasmic reticulum; MCS, membrane contact site; OSBP, oxysterol-binding protein; VAP, VAMP (vesicleassociated membrane protein)-associated protein.
Overview of Protein Translocation Into the Endoplasmic Reticulum A major function of the ER is to import and process newly synthesized proteins from the cytoplasm. This is necessary for growth of other organelles, including the Golgi apparatus, nucleus, endosomes, lysosomes, and plasma membrane, as well as for production of nearly all proteins secreted from the cell. All proteins are synthesized in the cytoplasm and must be targeted specifically to the ER, where they are either fully translocated across the ER membrane and released into the ER lumen (soluble proteins) or only partly translocated across the ER membrane and transferred into the lipid bilayer of the ER membrane (transmembrane proteins). The orientation of a protein in the lipid bilayer or its localization to the lumen is established during initial protein translocation and maintained as vesicles transfer the protein between membranes of the secretory pathway (see Fig. 21.2). Thus, domains of transmembrane proteins to be exposed on the cell surface must be in the ER lumen when their transmembrane domains are inserted into the ER membrane. Similarly, secreted soluble proteins must be fully translocated into the lumen of the ER. Because the ER lumen is topologically equivalent to the extracellular space, transport of proteins into the ER is analogous to transport into or across a prokaryotic plasma membrane (see Fig. 18.9). In both cases transported substrates must be recognized, targeted from the cytoplasm to the membrane, and translocated across the membrane through a protein channel without other molecules leaking across the membrane. Overcoming these obstacles involves specialized and regulated factors in the cytosol and target membrane. Insertion of proteins in the ER can occur either as the protein is being made by membrane-bound ribosomes
(cotranslational translocation) or after synthesis is complete (posttranslational translocation), each by distinct mechanisms. In the posttranslational pathway, the protein is first fully synthesized in the cytoplasm and then translocated independently of the ribosome.
Cotranslational Translocation Signal Sequences All soluble and membrane proteins destined for ER translocation using the cotranslational pathway contain a hydrophobic “leader” sequence that serves as a recognition signal for direction to the ER membrane. N-terminal leader sequences (termed signal sequences) are typically 15 to 35 amino acids long and contain a hydrophobic core of at least 6 residues. For many membrane proteins, the first transmembrane segment (a hydrophobic stretch of 16 to 25 residues) serves as a signal sequence. Aside from hydrophobicity, these signal sequences have no other features in common. Nevertheless, when attached to proteins that are not normally targeted to the ER, these signal sequences direct the protein to the ER and not to other organelles such as to mitochondria or peroxisomes, which use different targeting signals (see Fig. 18.1). Signal Recognition Particle and Signal Recognition Particle–Receptor The cotranslational pathway begins once the first hydrophobic residues of either a signal sequence or a transmembrane segment emerges from the ribosome and is recognized by the signal recognition particle (SRP), a large ribonucleoprotein complex (Fig. 20.5). SRP binds these hydrophobic peptides, slows translation by that ribosome and delivers the ribosome to a protein-conducting channel in the ER membrane called
336
SECTION VI n Cellular Organelles and Membrane Trafficking
FIGURE 20.5 STRUCTURE AND MECHANISM OF THE SIGNAL RECOGNITION PARTICLE. A, Secondary structure of human and Escherichia coli signal recognition particle (SRP) RNAs with sites of protein interactions in blue. B–E, Cryoelectron microscopic structures of SRP bound to mammalian ribosomes. SRP (green) wraps around the ribosome from the exit tunnel of the large subunit to the guanosine triphosphatase (GTPase) center near transfer RNA (tRNA) in the P-site. B, Structure with the SRP54 subunit (green ribbon diagram) engaging the transmembrane domain (TMD) of a nascent polypeptide that has just emerged from the exit tunnel of the ribosome. C, Two views of the structure of the SRP– ribosome complex in the scanning mode. D, Steps in the interaction of SRP with the ribosome, nascent polypeptide, and eukaryotic elongation factor 2 (eEF2). E, In the engaged mode the transmembrane domain of the nascent chain displaces one of the SRP54 helices. (B–E, From Voorhees RM, Hegde RS. Structures of the scanning and engaged states of the mammalian SRP-ribosome complex. eLife. 2015;4:e07975.)
a translocon. Then translation resumes, driving the polypeptide through the translocon into the lumen of the ER or inserting it into the membrane bilayer. Human SRP is composed of six proteins (named by their apparent molecular weights) assembled on a 300-nucleotide RNA that spans the distance from the exit tunnel on the large ribosome subunit to the GTPase center where translation factor GTPases (eEF1, eEF2, eEF3) bind. The SRP54 protein subunit binds signal sequences in a deep, hydrophobic groove lined by the flexible side chains of several methionines. Like bristles of a brush, the methionines accommodate the various hydrophobic side chains of different signal sequences. Phosphates of the SRP RNA near one end of the hydrophobic groove interact with basic residues that are often (but not always) adjacent to the hydrophobic core of signal sequences and transmembrane signal segments. Once SRP54 has engaged a signal sequence, the opposite end of SRP associates with the other side of the large ribosomal subunit at the GTPase center, where it can compete with binding of elongation factors eEF1 and eEF2 thereby slowing translation (Fig. 20.5E). This modest slowing of translation provides time to target the ribosome to the translocation channel in the ER before excessive polypeptide synthesis precludes cotranslational transport.
The complex consisting of ribosome, nascent chain, and SRP targets selectively to the ER membrane by binding the SRP receptor, a heterodimer consisting of one subunit that binds SRP and another that spans the ER membrane (Fig. 20.6). Both the SRP54 subunit and the SRP receptor have GTPase domains (Fig. 20.6B) similar to Ras (see Fig. 4.6). Neither SRP nor SRP receptor hydrolyzes GTP on its own. Instead, association of the two GTPase domains that accompanies successful targeting completes both of their active sites, resulting in hydrolysis of both bound GTPs. Dissociation of the γ-phosphates reduces the affinity of SRP for its receptor. The net result is that they dissociate from each other as well as from the ribosome and nascent chain. Dissociation of SRP54 from its binding site near the ribosome exit tunnel now allows the protein-conducting Sec61 channel to bind at the same site. The ribosome is now successfully targeted to and docked at the translocon, providing an opportunity for subsequent translocation or membrane insertion of the nascent polypeptide. SRP is released into the cytoplasm for another round of targeting, while SRP receptor diffuses away in the membrane to capture the next SRP–ribosome complex. The targeting cycle thus delivers the ribosome– nascent chain complex to the translocon and recycles the SRP and SRP receptor. GTP binding and hydrolysis
337
CHAPTER 20 n Endoplasmic Reticulum
A
B
Small ribosomal subunit
mRNA
SRP-Receptor
GT
P GTP
Ribosome binds mRNA Large ribosomal subunit
SRP
SRP
SRP-R
Translation begins
Binding completes both active sites resulting in hydrolysis of both GTPs, dissociation & resumption of translation
Polypeptide chain
GDP GT
GTP
P
SRP
Signal sequence
N
SRP binds signal TP and ribosome, slowing polypeptide translation
TRAM Sec61 channel
GDP GTP
DP
TRAP
G
TP G TP
G
DP
SRP receptor
SRP released
Unit binds to SRP receptor on ER membrane
G
CYTOPLASM
ER LUMEN
Pause
G
Translation
N
BiP
ATP ADP
BiP ratchets as translation continues FIGURE 20.6 COTRANSLATIONAL PATHWAY FROM RIBOSOME TO THE ENDOPLASMIC RETICULUM LUMEN. A, Signal recognition particle (SRP) and SRP-receptor use a cycle of recruitment and GTP hydrolysis to control delivery of a ribosome with an messenger RNA (mRNA) and nascent chain with a signal sequence to the Sec61 translocon in the ER membrane. SRP binds a signal sequence emerging from a ribosome and slows polypeptide translation. SRP also directs the ribosome to the SRP-receptor on the ER membrane, where the ribosome docks on the translocon and continues translation. B, Ribbon diagrams of two views of the complex between SRP and SRP-receptor showing the close association between the two GTPase domains, which activates GTPase hydrolysis. ADP, adenosine diphosphate; ATP, adenosine triphosphate; BiP, binding immunoglobulin protein; GDP, guanosine diphosphate; TRAM, translocating chain-associating membrane protein; TRAP, transloconassociated protein.
by SRP and SRP receptor provide directionality and order to the sequence of reactions that bring the nascent chain to the translocation channel.
Sec61 Complex: The Protein-Conducting Channel The nascent chain emerging from the ribosome engages and opens the translocon for transport across the ER membrane. The translocon is an evolutionarily conserved complex of three transmembrane proteins associated with proteins inside the ER. The eukaryotic Sec61 complex consists of an α subunit with 10 transmembrane helices and smaller β and γ subunits with single helices crossing the membrane (Fig. 20.7). The subunits of the homologous bacterial and archaeal SecY complex are called SecY, SecE, and SecG (see Fig. 18.9D). The Sec61 complex provides a high-affinity docking site for the ribosome–nascent chain complex. Additional factors thought to facilitate binding of the signal sequence to the Sec61 complex include the protein
TRAM (translocating chain-associating membrane protein) and the protein complex TRAP (translocon-associated protein). Bacterial SecY interacts with fully translated peptides associated with the chaperone SecA (see Fig. 18.9A). Prokaryotic and eukaryotic translocons have nearly identical structures. The 10 transmembrane helices of Sec61α are arranged around a narrow, hourglass-shaped pore that is most constricted near the center of the lipid bilayer (Fig. 20.7B). A seam between two pairs of helices, termed the lateral gate, can potentially separate to open the pore toward the membrane like a clamshell. The small accessory subunits form supporting transmembrane helices. Interactions with the ribosome and signal sequence control opening of the Sec61 translocon across or toward the membrane. Without these ligands translocons are quiescent with closed pores and lateral gates. A ring of hydrophobic side chains fills the pore and a short helix
338
SECTION VI n Cellular Organelles and Membrane Trafficking
FIGURE 20.7 STRUCTURE OF THE SEC61 PROTEIN-CONDUCTING TRANSLOCON. Structures were determined by cryoelectron microscopy of single particles. A, Cross section through the large subunit of a ribosome with a nascent chain in the ribosome exit tunnel and Sec61 associated with the exit pore. B–E, Ribbon diagrams and interpretative drawings of the three functional states of Sec61: B–C, quiescent without a translocating polypeptide and with a plug in the closed channel; D, primed while bound to a ribosome; and E, engaged with a bound signal sequence helix (in blue), an open channel and open lateral gate. (A, From Voorhees RM, Fernandez IS, Scheres SH, Hedge RS. Structure of the mammalian ribosome-Sec61 complex to 3.4 Å resolution. Cell. 2014;157;1632–1643. B–E, From Voorhees RM, Hegde RS. Toward a structural understanding of co-translational protein translocation. Curr Opin Cell Biol. 2016;41:91–99.)
called a plug further blocks the channel. When the cytoplasmic loops of Sec61α interact with the ribosome, the translocon undergoes a subtle conformational change to a “primed” state that contains a partially opened lateral gate. A signal sequence emerging from the ribosome exploits this cracked lateral gate to fully open the channel. The signal sequence entering the pore of the translocon forms an α-helix that binds to a hydrophobic patch in the lateral gate. The bound signal sequence is oriented with its N-terminus toward the cytoplasm, so it forms a loop in the pore of the translocon extending back to the exit site on the ribosome. This interaction leads to parting of the lateral gate, which is held open by the bound signal peptide. A parted lateral gate widens the translocon pore, leading to displacement of the plug in preparation for protein translocation. The ability to recognize signal sequences allows Sec61 to discriminate substrates for translocation from other proteins. Traditionally, this was thought to be a constitutive process predetermined by the sequences on the substrate. However, various cell types differ in the efficiency with which they recognize particular signal sequences. This can be explained if additional proteins at the translocation site influence signal sequence recognition. For example, proteins Sec62, Sec63, p180, p34, TRAM, and TRAP complex may stimulate or inhibit the translocation of selected substrates by recognizing diversity within the signal sequence. Selective changes in expression or modifications of these accessory components in different cell types could then affect the outcome of translocation for different substrates.
Polypeptides can take one of two possible paths through the engaged translocon as translation proceeds. Secretory proteins and soluble parts of membrane proteins move through the translocon pore into the ER lumen. Transmembrane domains of integral membrane proteins leave the translocon pore through the lateral gate to enter the hydrophobic environment of the lipid bilayer. Elongation of the polypeptide chain by the ribosome provides the energy for the nascent chain to pass through the channel across the membrane or into the bilayer. Thus, the energy used for protein synthesis is harnessed to drive translocation of the polypeptide to its destination.
Translocation of Soluble Proteins Into the Lumen of the Endoplasmic Reticulum Soluble proteins destined for secretion or retention in the ER lumen are translocated through the central pore of the translocon (Fig. 20.8A). The small, flexible side chains lining the pore fit snugly around a translocating peptide, preventing passage of ions or other small molecules. Thus the nascent chain has a continuous path from the peptidyl transferase center in the ribosome, through the translocation channel into the ER. Inside the ER lumen binding and release of chaperones such as binding immunoglobulin protein (BiP) may help translocate the polypeptide across the membrane, although this has not been firmly established. Once the translocating polypeptide has grown to approximately 150 residues, a signal peptidase associated with the translocon in the ER lumen cleaves off the signal sequence. After signal peptide cleavage, the new
339
CHAPTER 20 n Endoplasmic Reticulum
A. Protein released in lumen
Start-transfer sequence Stop-transfer sequence
CYTOPLASM Sec61 channel ER LUMEN
Signal cleaved, translocation continues
Signal sequence
Signal degraded, peptide folded
N
Mature protein
B. Single transmembrane protein with C-terminus in cytoplasm (type 1)
Translocation
C
Signal peptide cleaved, protein released from channel
Stop-transfer sequence stops in channel
N-terminal start-transfer peptide in Sec61 channel
C. GPI-anchored protein
N
Mature transmembrane protein
C
Translocation
Stop-transfer sequence stops in channel
N-terminal start-transfer peptide in Sec61 channel
D. Single transmembrane protein with N-terminus in cytoplasm (type 2)
N
Start-transfer peptide cleaved, protein released from channel
N
Protein is cleaved and transferred to a GPI lipid anchor
E. Two transmembrane-containing protein with N-terminus in cytoplasm
C N
C
N
N
Translocation
Translocation
Completed protein released from channel
Internal start-transfer peptide in Sec61 channel
C
Completed protein released from channel
Internal start-transfer peptide in Sec61 channel
F. Multiple transmembrane protein N
N
N
N
Stop-transfer sequence stops translocation Internal start-transfer peptide in Sec61 channel
Lateral transfer from the channel to the bilayer
C
Process repeated until all start-transfer and stop-transfer peptides are inserted into bilayer
FIGURE 20.8 COMPARISON OF PATHWAYS USING SEC61 TO TARGET PROTEINS TO THE ENDOPLASMIC RETICULUM LUMEN OR MEMBRANE. The drawings show how signal sequences, start-transfer sequences and stop-transfer sequences target protein to Sec61 channel and then to the lumen and membrane of the endoplasmic reticulum (ER). GPI, glycosylphosphatidylinositol.
340
SECTION VI n Cellular Organelles and Membrane Trafficking
N-terminus of the growing polypeptide continues to pass through the translocon until it is released into the ER lumen. The remaining cleaved signal peptide either is degraded or may have other functions elsewhere in the cell.
Insertion of Membrane Proteins Into the Endoplasmic Reticulum Bilayer Most proteins destined for insertion into the ER membrane bilayer use the Sec61 protein-conducting channel, which recognizes and inserts their transmembrane domains into the membrane bilayer. The machinery for targeting and insertion is physically coupled to the ribosome near the polypeptide exit tunnel, so the transmembrane domains are shielded from the aqueous cytoplasm. Transmembrane proteins are categorized as type 1, type 2, or polytopic depending on the orientation of their transmembrane domains across the lipid bilayer (Fig. 20.8). This orientation is established during translation and maintained as the protein moves to its final destination in the cell by membrane budding and fusion events (see Chapters 21 and 22). The N-terminal signal sequence of type 1 transmembrane proteins initiates translocation, similar to soluble proteins by engaging two helices adjacent to the lateral gate (Fig. 20.8B). As translation proceeds, a stop-transfer signal (usually a transmembrane domain) stops the transfer process before the polypeptide chain is completely translocated. After the signal sequence (also called a start-transfer signal) is cleaved off by the ER signal peptidase, the transmembrane segment slides out of the translocon through the lateral exit site. This leaves the protein oriented with its N-terminus in the ER lumen, its transmembrane segment spanning the ER membrane and its C-terminus in the cytoplasm. Some type 1 proteins, called glycosylphosphatidylinositol (GPI)-anchored proteins, exchange their C-terminal transmembrane segment for an oligosaccharide anchored to the lipid phosphatidylinositol (Fig. 20.8C; also see Fig. 13.10). An enzyme in the ER lumen cleaves off the transmembrane segment and transfers the new C-terminus to a preassembled GPI membrane anchor. Many cell-surface proteins are attached to the plasma membrane in this manner. This allows them to be readily released from the cell when specific phospholipases in the plasma membrane are activated. For example, during sperm capacitation, many GPI-anchored proteins are cleaved and released from the sperm plasma membrane. This reorganization of the sperm’s cell surface is essential for a sperm to fertilize an egg. Type 2 transmembrane proteins use a transmembrane domain in the middle of the polypeptide as an internal signal sequence (Fig. 20.8D). Once such an internal signal sequence emerges from a ribosome, it is recognized by SRP and brought to the ER membrane,
where it serves as a start-transfer signal to initiate protein translocation. When the protein is fully synthesized, the start-transfer signal, which is not cleaved off, slides out of the translocation channel into the surrounding lipid bilayer, where it serves as a transmembrane anchor. Polytopic proteins that span the membrane multiple times (such as ion channels and carriers) use multiple stop-transfer signals, none of which are cleaved by signal peptidases (Fig. 20.8E–F). SRP is probably required to target the first signal sequence to the ER membrane. Thereafter, the dynamics of the channel must accommodate sequences that specify translocation of loops in the cytoplasm or lumen alternating with the transfer of transmembrane segments to the lipid bilayer.
Association of Lipid-Anchored Proteins With the Cytoplasmic Surface of the Endoplasmic Reticulum Many classes of lipid-anchored proteins, including N-Ras and H-Ras GTPases, are targeted from the cytoplasm to the cytoplasmic leaflet of the ER bilayer by posttranslational modification of a C-terminal CAAX motif (where C is cysteine, A is any aliphatic residue, and X is any residue). First, a soluble enzyme attaches a farnesyl or geranylgeranyl lipid to the cysteine residue in CAAX with a thioether bond, anchoring the protein to the cytoplasmic surface of the ER membrane. Then, a prenylCAAX protease in the ER bilayer cleaves off the AAX residues, leaving the prenylcysteine as the new C terminus. Finally, another enzyme in the ER bilayer methylates the carboxyl group of the modified cysteine. Lipid-anchored proteins reach the plasma membrane by following the secretory pathway out of the ER. One or two other cysteine residues upstream of the CAAX motif of N-Ras and H-Ras can be tagged with palmitic acid via a labile thioester bond. Posttranslational Translocation by Sec61 Translocons Proteins targeted for posttranslational translocation into the ER have signal sequences and use the Sec61 translocon for entering the ER, but use proteins other than SRP or SRP receptor to guide them to the translocons after being released from the ribosome into the cytoplasm (Fig. 20.9). Posttranslational translocation is most common in yeast and bacteria. Mammalian cells use it primarily for translocating small proteins (2000/µm2) is much higher than that in the cell body (6/µm2). Binding of at least two cAMP molecules increases the probability that the channel is open from near 0 to about 0.65. Because the probability is not 1.0, individual activated channels flicker open and closed on a millisecond time scale. The ensemble of many activated channels admits enough Na+ and Ca2+ to depolarize the membrane. Ca2+-activated chloride channels carry additional current. The lag of 200 to 500 milliseconds between binding of the odorant and peak membrane depolarization is attributable to the relatively slow binding of cAMP to the channel. Triggering an action potential differs from the role of cAMP in most other tissues, where the main target is protein kinase A (PKA [see Fig. 25.3]). Depolarization of the ciliary membrane initiates an action potential (see Fig. 17.6) by activating voltagegated sodium channels (see Fig. 16.2) in the cell body. The action potential propagates along the axon to a chemical synapse with the second neuron in the pathway located in the olfactory bulb of the brain. The two stages of amplification downstream of the receptor allow a few active receptors to produce an action potential. Adaptation Desensitization—the waning of perceived odorant intensity despite its continued presence—results from a combination of processes in both olfactory neurons and the brain. Even with constant exposure to odorant the system adapts at each step in the signaling pathway owing to the transient G-protein activation, the selflimited increase in cAMP, and the brief membrane depolarization (Fig. 27.1C). The sequential nature of many of these feedback circuits implies that they have intrinsic delays and therefore serve not only to alter the magnitude of the response but also to shape its time course.
466
SECTION VII n Signaling Mechanisms
G-protein–coupled receptors are desensitized by protein kinases that phosphorylate the receptor and by proteins called arrestins that bind phosphorylated receptors (see Fig. 24.3). These modifications inhibit the interaction of activated receptors with G-proteins and provide negative feedback at the first stage of signal amplification. Negative feedback is coupled to receptor stimulation, because the olfactory receptor kinase is brought to the plasma membrane by binding the Gβγ subunits released by receptor-induced G-protein dissociation. Ca2+ entering the cell through cyclic nucleotide–gated channels binds calmodulin, which provides two types of negative feedback. Calcium-calmodulin activates the cAMP phosphodiesterase, which rapidly converts cAMP to inactive 5′-adenosine monophosphate (AMP). It also binds to the cyclic nucleotide–activated channels, reducing their affinity for cAMP by 10-fold and also reducing the probability of opening at less than saturating cyclic nucleotide concentrations. These two effects of Ca2+ alter the responsiveness of the neuron to initial odorant exposure, limit the time course of the response, extend the dynamic range over which the cell can respond, and make a cell transiently refractory to additional stimulation.
Processing in the Brain Mammals discriminate vastly more odorants than the number of available receptors. This is achieved by combining information from multiple types of receptors in their central nervous systems. The molecular specificity established in the olfactory epithelium between an odorant and its receptor is preserved at the first step in the brain, because all the axons from the sensory neurons that express a particular odorant receptor converge on only two to three target areas in the olfactory bulb called glomeruli (Fig. 27.1D). There the axons from like sensory neurons form synapses with dendrites of about 50 secondary neurons in the pathway. Given approximately 1000 odorant receptors in the mouse, each mouse olfactory bulb has approximately 2000 glomeruli. Most of the secondary neurons send their axons to higher levels, where they terminate in a combinatorial manner on cortical neurons. Of special interest, the odorant receptor itself is an important determinant of axon targeting to the glomeruli. Expression of different odorant receptors results in the axons selecting new glomerular targets. The discrimination of a particular odorant is achieved in two stages: At the first stage, each odorant activates several different receptors, and each receptor can bind a group of related odorants. Therefore, each odorant activates a particular pattern of olfactory sensory neurons and their coupled glomeruli. At the next level, neurons in the cerebral cortex receive information from a combination of glomeruli, leading to eventual discrimination
BOX 27.1 Sex and the Second Olfactory System Animals use olfaction to find their mates, identify their offspring, and mark their territories. Some of the odorants used for social interactions are volatile chemicals that stimulate the main olfactory system. A second accessory olfactory system detects other social odorants. Some are volatile chemicals found in urine; others are not volatile, including major histocompatibility complex (MHC) class II peptide complexes that are shed from the surfaces of cells into the urine and other secretions. Accessory sensory neurons are located in a special part of the epithelium lining nasal cavity called the vomeronasal organ. Each of these neurons expresses one of approximately 300 sevenhelix receptors from a different family than the main odorant receptors. Odorant binding activates a signal transduction pathway distinct from main olfactory neurons, dependent on a Trp channel (see Fig. 16.9) rather than a cyclic nucleotide–gated channel. The axons project to the accessory olfactory bulb in the brain.
of many different smells at higher levels of the brain. Box 27.1 has information on our second olfactory system.
Photon Detection by the Vertebrate Retina Overview of Visual Signal Processing Photons are energetic but unconventional agonists. They are tiny, move very fast, and penetrate most biochemical materials. These properties create a formidable challenge for detecting photons and transducing their intensities and wavelengths into a signal that can be transmitted to the brain. Nevertheless, vertebrate photoreceptor cells capture single photons and convert this energy into a highly amplified electrical response (Fig. 27.2). Phototransduction is the best-understood eukaryotic sensory process, because the system is amenable to sophisticated biophysical, biochemical, and physiological analysis. Single-cell organisms use similar mechanisms to respond to light (see Fig. 38.16). Vertebrate photoreceptor cells are neurons located in a two-dimensional array in the retina, an epithelium inside the eye. The cornea and lens of the eye form an inverted real image of the outside world on the retina, so the intensity of the light across the field of view is encoded by the array of geographically separate photoreceptor cells. Photoreceptor cells lie at the base of a complex neural processing system. Having detected the rate of photon stimulation at a particular place in the visual field, photoreceptor neurons communicate this information to higher levels of the visual system. Initial processing of the information takes place in the retina, where secondary and tertiary neurons take input from multiple photoreceptors to derive local information regarding image contrast, as well as color and intensity. Neuroscience texts present more detailed information
467
CHAPTER 27 n Integration of Signals
A. Rod anatomy
B. Dark
C. Activation
D. Recovery
Photon
Outer segment
Plasma membrane Rhodopsin Disks
α
Dγ
β
GTP Inner segment
Connecting cilium ER Golgi Mitochondria
Nucleus Fiber Synaptic ending
Gα T D GTP GDP
GTrimeric cGMP Guanylyl cyclase
Phosphodiesterase GMP cGMP
Arrestin Gβγ ATP ADP
RGS
Kinase
RGS promotes hydrolysis cGMP D GTP
Na+
cGMPCa2+ gated channels open
Na+ cGMPgated channels close
GTP
+ + – + – +
– +– – + + +– – – + + +
Constitutive secretion of– – glutamate
cGMPCa2+ gated channels open K+
4 Na+ Carrier – + – – + –
GTP
– + + + + – +
Plasma membrane –– hyperpolarizes –– – + – +
Hyperpolarized membrane inhibits + + secretion – + –
–
+
4 Na+ Carrier Current restores –+ ++ resting membrane –– +– potential –+ +– – +– – + + +– – – + + + Glutamate – –
secretion resumes
Retina Glutamate
Glutamate
FIGURE 27.2 VERTEBRATE VISUAL TRANSDUCTION. A, Drawing of a rod cell. Disks in the outer segment are rich in rhodopsin. ER, endoplasmic reticulum. B–D, Drawings of small portions of an outer segment (upper panels) and the synaptic terminal of a rod cell (lower panels) in three physiological states. Active components are highlighted by bright colors. B, Resting cell in the dark. Constitutive production of cyclic guanosine monophosphate (cGMP) keeps a subset of the plasma membrane cGMP-gated channels open most of the time, allowing an influx of Na+ and Ca2+. At this membrane potential, the synaptic terminal constitutively secretes the neurotransmitter glutamate. Ca2+ leaves the outer segment via a sodium/calcium exchange carrier in the outer segment, whereas Na+ leaves the cell via a sodium pump in the plasma membrane of the inner segment. C, Absorption of a photon activates one rhodopsin, allowing it to catalyze the exchange of GTP for GDP bound on many molecules of transducin (GT). This dissociates GTα from Gβγ. Each GTα-GTP binds and activates one molecule of phosphodiesterase (attached to the disk membrane by N-terminal isoprenyl groups), which rapidly converts cGMP to guanosine monophosphate (GMP). As the concentration of free cGMP declines, the cGMP-gated channels close, leading to hyperpolarization of the plasma membrane and inhibition of glutamate secretion at the synaptic body. D, Recovery is initiated when rhodopsin kinase phosphorylates activated rhodopsin. Binding of arrestin to phosphorylated rhodopsin prevents further activation of GT. Phosphodiesterase and an RGS protein cooperate to stimulate hydrolysis of GTP bound to GT, returning GT to the inactive GTα-GDP state. Synthesis of cGMP by guanylyl cyclase returns the cytoplasmic concentration of cGMP to resting levels and opens the cGMP-gated channels. Constitutive secretion of glutamate resumes. ADP, adenosine diphosphate; ATP, adenosine triphosphate.
on higher levels of visual processing in the retina and brain. The response of photoreceptor cells depends on the intensity of the light, that is, the flux of photons. Vertebrate retinas detect light with intensities that range over 10 orders of magnitude. Rod photoreceptors (Fig. 27.2A) detect low levels of light from approximately 0.01 photon/µm2/s (dim stars) to 10 photons/µm2/s but do not discriminate light of different colors. Cone photoreceptors (cones) respond to more intense light, up to about 109 photons/µm2/s (full sunlight). Three classes of cones with chromophores sensitive to different wavelengths of light allow humans to encode wavelength. This is the basis for color vision (Box 27.2).
BOX 27.2 Color Vision Three types of cones allow us to discriminate colors. Cones are organized much like rods but express one of three different seven-helix photoreceptors, each with a distinct visual pigment. The absorption spectra of the photoreceptors overlap, so the central nervous system can perceive colors from deep purple to deep red by comparing the relative activation of the three types of cones at each point in the visual field. As in rods, signal transduction in cones depends on degradation of cyclic guanosine monophosphate (cGMP) and closure of cGMP-gated channels. Activation of cones requires much more intense light, but cones respond faster than rods.
468
SECTION VII n Signaling Mechanisms
Rods and cones have three specialized regions with different molecular components and functions. The nucleus and the organelles in the inner segment maintain the cell’s structure and metabolism. A vestigial cilium connects the inner segment to the outer segment, which consists of a stack of internal membrane disks containing the photoreceptor protein, rhodopsin in the case of rods, surrounded by the plasma membrane. Disks form by invagination and pinching off of flattened sacks of plasma membrane. Rhodopsin synthesized in the cell body is transported to the plasma membrane along the secretory pathway and segregated into disk membranes. The lumen of the disks corresponds topologically to the lumen of the endoplasmic reticulum or the extracellular space. The base of the photoreceptor cell forms a synapse with the next neuron in the circuit. Absorption of a photon activates rhodopsin and initiates a signaling cascade (Fig. 27.2) involving a trimeric G-protein that activates a cyclic guanosine monophosphate (cGMP) phosphodiesterase. The phosphodiesterase lowers the cytoplasmic concentration of cGMP and closes cGMP-gated channels in the plasma membrane. Closing these channels hyperpolarizes the plasma membrane and reduces the release of glutamate at the synapse with the next neuron in the visual circuit. Feedback loops operate at every level in this pathway, turning off the response to a flash of light. The following sections explain how these reactions achieve their spectacular sensitivity in rods.
Rhodopsin Rhodopsin is a seven-helix, G-protein-coupled receptor with a light-absorbing chromophore, 11-cis retinal, covalently attached to lysine 296 through a protonated Schiff base (Fig. 27.2B). Although 11-cis retinal is bound to a site in the bundle of transmembrane helices similar to sites where ligands bind other seven-helix receptors, this conformation of rhodopsin is inactive with respect to catalyzing nucleotide exchange on its trimeric G-protein. Thus, rhodopsin is a seven-helix receptor with a covalently attached, but inactive, ligand. The ability of rods to detect single photons depends on two favorable properties. First, the noise level is very low, owing to the stability of the 11-cis retinal. In vertebrate rods, fewer than one molecule in 4 × 1010 isomerizes spontaneously every second, so the background level of activated rhodopsin is very low, even with more than 108 molecules of rhodopsin per cell. Thus, one does not perceive spots of light in the dark. Second, rods absorb photons very efficiently by virtue of the high density of rhodopsin in disks (~25,000 rhodopsins/µm2) and stacking thousands of disks on top of each other in the direction of incoming photons. Rhodopsin constitutes 90% of the disk membrane protein and 45% of the disk membrane mass. About half of the photons that traverse the outer segment are absorbed, and about
two-thirds of absorbed photons produce an electrical change in the plasma membrane. Absorption of light initiates the signal transduction pathway. The energy from the absorbed photon isomerizes the 11-cis retinal chromophore to all-trans retinal in picoseconds, creating a cascade of intramolecular reactions that activate rhodopsin by changing its con formation. The active conformation called metarhodopsin II catalyzes nucleotide exchange on transducin, its trimeric G-protein partner (see Fig. 25.9). Following activation, rhodopsin is inactivated by hydrolysis of the Schiff base linking all-trans retinal to the protein and dissociation of the chromophore. Attachment of a fresh molecule of 11-cis retinal, derived from vitamin A or regenerated from all-trans retinal, regenerates rhodopsin.
Positive Arm of the Signal Cascade Two enzymatic reactions amplify the signal initiated by absorption of light. Metarhodopsin II catalyzes the exchange of GDP for GTP on the α subunit of transducin, which then dissociates its βγ subunits also attached to the cytoplasmic face of the disk membrane by covalently bound lipid groups. Each metarhodopsin II produces hundreds of activated transducins in a fraction of a second, nearly as fast as the molecules collide while they diffuse in the plane of the very crowded disk bilayer. Nevertheless, nucleotide exchange on transducin is ratelimiting in the whole transduction cascade, even with a 107 acceleration by metarhodopsin II. Transducin α-GTP activates phosphodiesterase associated with the disk membrane by binding the enzyme’s two inhibitory γ subunits. This frees the catalytic α and β subunits of phosphodiesterase to break down cGMP to guanosine monophosphate (GMP) at a high rate. The cytoplasmic concentration of cGMP depends largely on its rate of destruction by lightactivated phosphodiesterase, as it is made continuously by guanylyl cyclase. As the concentration of cGMP falls, cGMP-gated cation channels (see Fig. 16.10) in the plasma membrane close resulting in hyperpolarization of the membrane. This change in membrane potential inhibits glutamate release at the synapse. The light-induced decline in glutamate release has opposite effects on the two types of “bipolar neurons” connected to rods: It stimulates “on-type” bipolar neurons to fire action potentials and hyperpolarizes the “off-type” bipolar neurons. This combination of responses is the first step in the discrimination of contrast in our visual world. Amplification in this pathway is spectacular. Within 1 second after absorption of a single photon, rhodopsin activates 1000 transducins and a similar number of phosphodiesterases, which break down 50,000 cGMPs. This change in concentration closes hundreds of cGMPgated channels, each of which blocks the entry of more
CHAPTER 27 n Integration of Signals
BOX 27.3 Electrical Circuits in the Photoreceptor Absorption of light changes currents flowing through electrical circuits in photoreceptor cells. In the dark, the resting cyclic guanosine monophosphate (cGMP) concentration in the outer segment keeps open 1% of the cGMPgated channels. These open channels produce an inward “dark current” of Na+ and Ca2+, which is balanced by an outward current of K+ through channels in the inner segment. Sodium-potassium adenosine triphosphatase (ATPase) pumps in the inner segment compensate for the accumulation of Na+ and the depletion of K+. Ca2+ entering the outer segment is exported by a carrier in the plasma membrane of the outer segment that exchanges Ca2+ and K+ for Na+. Following absorption of a photon, both the cGMP concentration and the probability of the cGMP-gated channels being open declines on a millisecond time scale. In parallel, the cation current into the outer segment falls, hyperpolarizing the plasma membrane. The cytoplasmic Ca2+ concentration also declines from about 300 nM to 50 nM. The magnitudes of these responses depend on the number of photons absorbed and the size of the amplified signal. Two useful properties emerge from the fact that the extracellular concentration of Ca2+ largely blocks open photoreceptor channels, similar to the cyclic nucleotide– gated channel of olfactory neurons. First, this reduces the burden on the pumps that maintain the ionic gradients in the cell. Second, using multiple channels with low ionic conductance improves the signal-to-noise ratio. For example, if only two channels carried the dark current, the statistical opening or closing of one channel would create large fluctuations in the current. If 100 partially blocked channels carried the same current, then opening or closing single channels has a modest effect on total current.
than 10,000 cations. Box 27.3 provides more details about the electrical circuit in the rod cell.
Recovery and Adaptation After a dim flash of light, the dip in cytoplasmic cGMP concentration and hyperpolarization of the plasma membrane are short-lived, on the order of 2 seconds in rods, and even less in cones. Rods reset the signaling pathway by inhibiting metarhodopsin II, inactivating transducin α-GTP, and stimulating the synthesis of cGMP. Phosphorylation turns off active metarhodopsin II. Transducin βγ subunits activate rhodopsin kinase (called GRK1 for G-protein–coupled receptor kinase 1), which phosphorylates several residues near the Cterminus of the receptor. Phosphorylation not only reduces the ability of rhodopsin to activate transducin, but it also creates a binding site for arrestin, which prevents further production of transducin α-GTP (see Fig. 24.3).
469
BOX 27.4 Second Visual System to Set Circadian
Clocks
Many organisms, including humans, use the regular variation in light during the day and night to entrain a network of transcription factors that control a 24-hour circadian cycle of metabolic activities throughout the body. For example, mice that are kept in complete darkness continue to run (searching for food) during the hours corresponding to night and sleep during the hours corresponding to day. Without light input, they gradually drift from a precise 24-hour cycle. Neither rods or cones are required to receive the light that synchronizes the internal cycle with the 24-hour day. Instead, a subset of retinal ganglion cells absorbs the light and sends signals to the hypothalamic region of the brain. At least two different photoproteins absorb the light: melanopsin, an opsin family member, and cryptochromes, proteins with a flavin chromophore. So the eye is two photodetectors in one.
The second level of negative feedback comes from the hydrolysis of GTP bound to transducin α in less than 1 second. Both phosphodiesterase (just activated by transducin α) and an RGS protein (regulator of G-protein signaling; see Fig. 25.8) stimulate transducin α to hydrolyze the bound GTP. Humans with mutations that disable the retinal RGS protein cannot adapt to rapid changes in light, so they are blinded for several seconds when they step out of a dark room into full sunlight. Dissociation of transducin α-GDP from phosphodiesterase inhibitory subunits terminates cGMP breakdown. The reduction in cytoplasmic Ca2+ that accompanies closure of cGMP-gated cation channels stimulates the guanylyl cyclase that rapidly restores the cGMP concentration. This change opens the cation channels and returns the membrane potential to the resting level. Box 27.4 has information on a second visual system that mammals use to set circadian clocks.
Regulation of Metabolism Through the β-Adrenergic Receptor Epinephrine, a catecholamine that is also called adrenaline (Fig. 27.3), is secreted by the neuroendocrine cells of the adrenal gland and other tissues when an animal is startled, is stressed, or otherwise needs to respond vigorously. Norepinephrine, a closely related catecholamine, is secreted by sympathetic neurons, including those that regulate the contractility of the heart. These hormones flow through the blood and stimulate cells of many types throughout the body to heighten their metabolic activity. The particular physiological response in each tissue depends on selective expression of a family of nine adrenergic receptors and their associated signaling hardware in differentiated cells (Table 27.1). Epinephrine binding
470
SECTION VII n Signaling Mechanisms
OH
OH
OH
Epinephrine
OH
OH
Norepinephrine
HO CH H N CH H3C H
Tyrosine
HO CH H2N CH H
CH2 O H2N C C OH H
A. Ligand binds receptor, turning it on B. Receptor activates trimeric GTPase Gβγ
+
γ
α GDP
β
GTP
Gα
GTP
GDP
C. G,α-GTP activates adenylyl cyclase
cAMP D. cAMP activates PKA
ATP
BAR receptor kinase PKA Glucose-6 P
108
105 Log
ATP
P-phosphorylase kinase
106
PKA-RII
E. PKA phosphorylates PK
P-phosphorylase
107
RII
cAMP PKA
Calmodulinphosphorylase kinase
PP1
Calmodulin-phosphorylase kinaseCa2+
104 103
Ca-CM-phosphorylase kinase-
100
Gα-GTP
G. Phosphorylase activated by phosphorylation
10 Active receptor
1 0
F. Ca2+ completes activation of PK
0
Time (minutes)
Phosphorylase
H. Active phosphorylase makes glucose-6 P
ATP
PP1
Phosphorylase-
Glycogen
Glucose-6 ATP
FIGURE 27.3 β-ADRENERGIC SIGNALING MECHANISM. Active components are shown in bright colors. Upper right, Pathway of epinephrine synthesis. Lower left, Time course of the amplification of the signal by the catalytic cascade of signal-transducing enzymes. A, Epinephrine binds the seven-helix β-adrenergic receptor, shifting its equilibrium to the active conformation. B, Active receptor catalyzes the exchange of GDP for GTP on GSα, dissociating GSα-GTP from Gβγ. C, GSα-GTP activates adenylyl cyclase, which produces multiple cAMPs. D, cAMP activates protein kinase A (PKA) by dissociating the regulatory subunit, RII. E, PKA phosphorylates and partially activates multiple molecules of phosphorylase kinase (PK). F, Ca2+ binds to calmodulin (CM) associated with PK, completing activation. G, PK phosphorylates and activates phosphorylase. H, Phosphorylase catalyzes the conversion of glycogen to glucose-6 phosphate. Negative feedback loops (red) terminate stimulation. cAMP activates PKA and Gβγ activates β-adrenergic receptor kinase, both of which phosphorylate and inhibit the receptor from catalyzing nucleotide exchange. PP1, protein phosphatase 1.
to these β-adrenergic receptors is the classic example of a pathway using a seven-helix receptor (see Fig. 24.2), a trimeric G-protein (see Fig. 25.9), and adenylyl cyclase (see Fig. 26.2) to produce cAMP. This second messenger mediates a wide variety of cellular responses by activating PKA (see Fig. 25.3), which phosphorylates many different cellular proteins, thereby changing their activities. Differentiated cells vary in their responses to epinephrine and norepinephrine, because they express different
targets for PKA. In the heart, PKA phosphorylates voltage-gated Ca channels, increasing cytoplasmic Ca2+, and phospholamban, a small membrane protein that stimulates Ca2+ pumps in the smooth endoplasmic reticulum to clear Ca2+ from the cytoplasm (see Fig. 39.14). These changes stimulate the heart to contract more frequently and with greater force (see Fig. 39.21). On the other hand, PKA in smooth muscle cells of arteries phosphorylates and inhibits myosin light chain kinase, which inhibits contraction and increases blood flow (see
CHAPTER 27 n Integration of Signals
471
TABLE 27.1 Four Examples of Adrenergic Receptors and Physiological Responses Receptor
Tissue
Signaling Pathway
Responses
α1
Smooth muscle, blood vessels Smooth muscle, GI tract Liver
Gq, PLC-β, IP3, Ca , MLCK Gq Gq
Contraction Relaxation Glycogenolysis
α2
Smooth muscle, blood vessels Pancreatic islets
Gi, Ca2+ Gi, inhibit A-cyclase, K+ channel open
Contraction Secretion inhibition
β1
Heart
Gs, A-cyclase, cAMP, PKA, phospholamban
Increased contraction
β2
Liver Skeletal muscle
Gs, A-cyclase, cAMP, PKA, phosphorylase Gs, A-cyclase, cAMP, PKA, phosphorylase
Glycogenolysis Glycogenolysis
2+
cAMP, cyclic adenosine monophosphate; GI, gastrointestinal; IP3, inositol triphosphate; MLCK, myosin light-chain kinase; PKA, protein kinase A; PLC-β, phospholipase Cβ.
Fig. 39.24). Skeletal muscle and liver cells respond to epinephrine and PKA by breaking down glycogen to release glucose into the circulation (Fig. 27.3), while PKA in brown fat cells dissipates energy as heat (see Fig. 28.3). This section explains how β-adrenergic receptors regulate the production of glucose-6 phosphate from glycogen (Fig. 27.3), a process termed glycogenolysis. As with vision and olfaction, the response to epinephrine is sensitive, highly amplified (see the graph in Fig. 27.3), and subject to negative feedback control. Five stages of amplification along the seven-step pathway allow binding of a single molecule of epinephrine to a receptor to activate millions of enzyme molecules that produce many millions of molecules of glucose-6 phosphate: 1. Epinephrine binds to a β-adrenergic receptor, trapping it in its active conformation. The resting system is on the verge of activation, because the ligand-free receptor is in a rapid equilibrium between activated and unactivated states. Even without agonist, a small fraction of receptors is active at any given time. Thus, experimentally increasing the total concentration of receptors (and therefore the concentration of spontaneously active receptors) can maximally activate downstream pathways, even in the absence of agonists. 2. Each activated receptor catalyzes the exchange of GDP for GTP on many molecules of the trimeric protein Gs, causing the dissociation of GTP-Gsα from the Gβγ subunits and amplifying the signal up to 100fold. The G-protein subunits remain attached to the membrane by their lipid anchors but separate to activate different targets. This is the first branch point in the pathway. 3. GTP-Gsα binds and activates adenylyl cyclase, an integral membrane protein (see Fig. 26.2) that produces many molecules of cAMP. This is the second stage of amplification. 4. cAMP activates PKA by binding and dissociating its inhibitory RII subunit (see Fig. 25.3). Activation of
PKA mediates most effects of cAMP, but cAMP also activates cyclic nucleotide–gated ion channels in some cells. 5. Each activated PKA amplifies the signal by phosphorylating many substrate molecules, including phosphorylase kinase. This enzyme requires Ca2+ for activity, but phosphorylation by PKA reduces the Ca2+ requirement, so the kinase is active even at resting (0.1 µM) Ca2+ concentrations. On the other hand, high cytoplasmic Ca2+ concentrations alone, as during muscle contraction, can activate phosphorylase kinase without phosphorylation. 6. Each activated phosphorylase kinase further amplifies the signal by phosphorylating and activating many molecules of the enzyme phosphorylase b. PKA enhances the phosphorylation of both phosphorylase kinase and phosphorylase b by inhibiting the regulatory G subunit of protein phosphatase 1 (see Fig. 25.5), which dephosphorylates both enzymes. 7. Each activated phosphorylase molecule (called phosphorylase a) removes many glucose subunits from glycogen, one at a time. This fifth stage of amplification produces glucose-6-phosphate, which can enter the energy-releasing glycolytic pathway of the cell (see Fig. 19.4) or, in the case of liver, can be dephosphorylated and released into the bloodstream to provide an energy source for other cells in the body. If the concentration of epinephrine in the blood declines, epinephrine dissociates rapidly from β-receptors and their equilibrium shifts promptly toward the inactive state. Even in the continued presence of epinephrine the cell adapts owing to reactions that counterbalance the positive arm of the pathway as follows. First, each activating reaction is reversible, either spontaneously or catalyzed by specific enzymes. Activated GTP-Gsα hydrolyzes its bound nucleotide slowly, at a rate of approximately 0.05 s−1, then rebinds Gβγ, returning the complex to its inactive state. A Ca2+calmodulin-activated phosphodiesterase degrades cAMP to inactive 5′-adenosine monophosphate (AMP). This
472
SECTION VII n Signaling Mechanisms
convergence allows signaling pathways that release Ca2+ (see Fig. 26.12) to modulate the β-adrenergic pathway. Second, the system has several negative feedback loops that operate on a range of time scales. As in olfaction and vision, active β-adrenergic receptors are inhibited by phosphorylation. In seconds to minutes inhibitory sites in the C-terminal cytoplasmic tail are phosphorylated by PKA activated by cAMP and β-adrenergic receptor kinase (now called GRK2) recruited by Gβγ subunits released from Gsα-GTP. β-arrestin binding to phosphorylated receptors rapidly blocks interactions of the active receptor with G-proteins and attracts cAMP phosphodiesterase to the membrane. On a time scale of many minutes, interactions of β-arrestin with clathrin and adapter proteins (see Fig. 22.9) mediate removal of receptors from the cell surface by endocytosis. Prolonged stimulation by epinephrine results in receptor ubiquitination (see Fig. 23.2), endocytosis, and degradation. In addition to these effects on glucose metabolism, active β-adrenergic receptors produce at least two other signals. Gβγ subunits activate calcium channels in some cells. This Ca2+ can augment glycogen breakdown at the phosphorylase kinase step. In addition to its negative effects, β-arrestin activates the mitogen-activated protein (MAP) kinase pathway (Fig. 27.6) by binding to phosphorylated receptors and also serving as a membraneanchoring site for the cytoplasmic tyrosine kinase, c-Src.
Signaling Pathways Influencing Gene Expression Many extracellular ligands influence gene expression, via just three kinds of generic pathways (Fig. 27.4). The ligands for the first generic pathway are small and hydrophobic, such as steroids, vitamin A, and thyroid hormone. These ligands penetrate the plasma membrane and bind nuclear receptors in the cytoplasm. The ligands for the other two generic pathways include small charged molecules, peptides, and proteins that cannot penetrate the plasma membrane. Therefore, they must bind receptors on the cell surface to initiate pathways that activate transcription factors. In all cases, activated transcription factors cooperate with other nuclear proteins to regulate the expression of specific genes: 1. Nuclear receptor pathways: Ligands such as steroid hormones cross the plasma membrane into the cytoplasm, where they bind latent transcription factors called nuclear receptors. Ligand-bound receptors move from the cytoplasm into the nucleus and, in combination with other proteins, activate transcription of specific genes (see Fig. 10.21A). 2. Pathways activating mobile kinases in the cytoplasm: Some plasma membrane receptors turn on pathways to activate cytoplasmic protein kinases, which enter the nucleus, where they phosphorylate
A
Small hydrophobic ligands
B
Charged small molecules, peptides, proteins
Charged small molecules, peptides, proteins
Plasma membrane receptors
ECM
Activate cytoplasmic kinase
Bind nuclear receptor
C
CYTOPLASM
Activate latent cytoplasmic transcription factor (phosphorylation, proteolysis)
NUCLEUS Phosphorylate latent transcription factor Gene
Gene
1. Steroid hormones 1. PKA CREB 2. Thyroid hormone 2. Receptor tyrosine kinases MAP kinase 3. T-cell receptor cytoplasmic tyrosine kinase
Gene 1. Cytokines STAT 2. TGF-β Smads 3. TNF NF-κB 4. Delta Notch 5. Hedgehog 6. Wnt β-catenin 7. TCR NF-AT
FIGURE 27.4 THREE SIGNALING PATHWAYS BY WHICH EXTRACELLULAR LIGANDS INFLUENCE GENE EXPRESSION. A, Nuclear receptor pathway for small hydrophobic ligands that penetrate the plasma membrane (see Fig. 10.21A for an example). B, Pathways employing a plasma membrane receptor and a cytoplasmic protein kinase that enters the nucleus to activate a latent transcription factor. (See Fig. 10.21B for the PKA pathway, Fig. 27.6 for a receptor tyrosine kinase pathway, and Fig. 27.8 for a cytoplasmic tyrosine kinase pathway.) C, Pathways employing a plasma membrane receptor and activating a latent transcription factor in the cytoplasm. The list includes six known pathways of this type. (See Fig. 27.8 for nuclear factor–activated T cells [NF-AT]; Fig. 27.9 for a signal transducer and activator of transcription [STAT] pathway; Fig. 27.10 for a SMAD [Sma- and Mad-related protein] pathway; Fig. 10.21C for the nuclear factor κB [NF-κB] pathway; Chapter 24 for the Notch and Hedgehog pathways; and Fig. 30.7 for the β-catenin pathway.) CREB, cAMP response element–binding protein; ECM, extracellular matrix; TCR, T-cell receptor; TNF, tumor necrosis factor.
latent transcription factors. These mobile kinases include PKA (see Fig. 10.21B) and several MAP kinases (Figs. 27.5 through 27.8). 3. Pathways activating latent transcription factors in the cytoplasm: Other plasma membrane receptors activate latent transcription factors in the cytoplasm, generally by phosphorylation or proteolysis. These activated transcription factors then enter the nucleus. These mobile transcription factors include nuclear factor–activated T cells (NF-AT) (Fig. 27.8); signal transducer and activator of transcriptions (STATs) (Fig. 27.9), SMADs [Sma- and Mad-related proteins] (Fig. 27.10); nuclear factor κB (NF-κB) (see Fig.
CHAPTER 27 n Integration of Signals
10.21C); Notch (see Chapter 24); Hedgehog (see Chapter 24); and β-catenin (see Fig. 30.7).
Mitogen-Activated Protein Kinase Pathways to the Nucleus Cascades of three protein kinases terminating in a MAP kinase relay signals from diverse stimuli and receptors to the nucleus (Fig. 27.5). The first kinase activates the second kinase by phosphorylating serine residues. The second kinase activates MAP kinase by phosphorylating both a tyrosine and a serine residue in the activation loop (see Fig. 25.3F). Active MAP kinase enters the nucleus and phosphorylates transcription factors, which regulate gene expression. Key targets include genes that advance or restrain the cell cycle, depending on the system. MAP kinases also regulate the synthesis of nucleotides required for making RNA and DNA. A variety of cell surface receptors initiate pathways that activate MAP kinase cascades. Many of these pathways pass through the small guanosine triphosphatase (GTPase) Ras, allowing cells to integrate diverse growthpromoting signals to control the cell cycle (see Fig. 41.9). Receptor tyrosine kinases for growth factors (Fig. 27.6) and insulin (Fig. 27.7) send signals through Ras. Other receptors use nonreceptor tyrosine kinases coupled to Ras and MAP kinase, such as T-lymphocyte receptors via zeta-associated protein kinase (ZAP-kinase) (Fig. 27.8). Seven-helix receptors can also activate MAP kinase pathways. For example, β-arrestin not only inactivates β-adrenergic receptors, it also couples them to a MAP kinase pathway (see Fig. 24.3). Budding yeast activates MAP kinase pathways in two ways. Mating pheromones
Inputs Receptors
Growth factors
Oxidative stress
Receptor Receptor tyrosine kinase tyrosine kinase
bind seven-helix receptors that release Gβγ subunits of trimeric G-proteins, which activate the first kinase in the cascade. Alternatively, osmotic shock activates a twocomponent receptor (Fig. 27.11) upstream of another MAP kinase pathway that regulates the synthesis of glycerol, which is used to adjust cytoplasmic osmolarity. Animal cells have multiple MAP kinase cascades with particular isoforms of the three kinases linked in series and leading to different effectors (Fig. 27.5). The kinases that make up these pathways are expressed selectively in various cells and tissues. Deletion of single MAP kinases in mice is generally not lethal, so crosstalk between pathways is likely to be extensive. A cascade of kinases provides opportunities to integrate inputs from converging pathways and to amplify signals. Amplification can be so strong that a MAP kinase cascade can act like an all-or-nothing switch. For example, frog oocytes that are arrested in the G2 stage of the cell cycle react to the hormone progesterone by either remaining arrested or entering the cell cycle at full speed. Progesterone activates a MAP kinase cascade consisting of Mos, MEK1, and the p42 MAP kinase. In individual cells, the MAP kinase is either unphosphorylated and inactive or doubly phosphorylated and fully active. This bistable switch-like response depends in part on the fact that both MEK1 and MAP kinase require two independent phosphorylation events for activation. In addition, active MAP kinase provides two types of positive feedback (Fig. 27.5B). MAP kinase both activates Mos by phosphorylation and also drives Mos expression. Consequently, a marginal stimulus turns on some cells strongly and others not at all rather than producing a partial response in all the cells.
B
Inputs
Interleukin-1
Extracellular matrix
TNF family receptor
Integrins
Receptor
Ras
Src
TRAF6/TAB1/2
Rac1
Activator
MAP kinase kinase kinase (S/T kinases)
c-Raf
MEKK2
TAK1
MEKK1
MAP KKK
MAP kinase kinase (dual specificity kinases)
MKK1
MKK5
MKK3/6
MKK4/7
MAP KK
MAP kinase (S/T kinase)
Erk1/2
Erk5
p38(α, β, γ, δ)
JNK1/2/3
Substrates
p90RSK (kinase)
MEF2 (TF)
MNK1 (kinase)
c-Jun (TF)
Outputs
Growth, differentiation
Gene expression
Cytokine production
Growth, differentiation
MAP kinase Transcription factor
Negative feedback
Activators
Positive feedback
A
473
Gene expression: (MAP KKK) & (MAP kinase phosphatase)
FIGURE 27.5 A, Four mitogen-activated protein (MAP) kinase pathways. The pathways are arranged vertically. The levels of the pathways are arranged horizontally with the nuclear compartment at the bottom. Dual-specificity protein kinases phosphorylate S/T and Y residues. B, Positive and negative feedback loops along a generic MAP kinase pathway. K, kinase; TF, transcription factor.
474
SECTION VII n Signaling Mechanisms
Autoinhibited monomer 130º
EGF
B. Extracellular domains dimerize A. EGF binding opens extracellular domains
One kinase domain binds and activates other kinase domain
D. Other effectors with SH2 or PTB domains bind P-tyrosines
C. Receptor binds and activates PLCγ
E. Ras activation
PIP2
DAG+IP3
Ras GTP
Ras GDP
Minimally active kinase domain
PLCγ
PKC
SOS
Raf
Raf activation
Grb2
Ubiquitination & endocytosis MEK
MEK activation
ERK
Transcription factors
ERK activation
G. Cellular proliferation, differentiation, etc
F. MA
Pk
se in a
y wa th a p
FIGURE 27.6 EPIDERMAL GROWTH FACTOR (EGF) RECEPTOR TYROSINE KINASE SIGNALING PATHWAY THROUGH MITOGENACTIVATED PROTEIN (MAP) KINASE. A, Ligand binding changes the conformation of the extracellular domains of the receptor. B, Extracellular domains dimerize, bringing together the tyrosine kinase domains of two receptor subunits in the cytoplasm. Direct interactions and transphosphorylation activate the kinases and create specific docking sites for effector proteins with SH2 (Src homology 2) domains. C, Phospholipase Cγ (PLCγ) binds one phosphotyrosine and is activated by phosphorylation to break down PIP2 into diacylglycerol and inositol triphosphate (IP3). D, A complex of the adapter protein Grb2 and the nucleotide exchange factor SOS binds another phosphotyrosine. (The gene for SOS protein got its name—“son of sevenless”—as a downstream component of the sevenless growth factor receptor gene required for the development of photoreceptor cell number seven in the fly eye.) E, SOS catalyzes the exchange of GDP for GTP on the membrane-associated small GTPase Ras. Ras-GTP attracts the cytoplasmic serine/threonine kinase Raf to the plasma membrane. F, Raf phosphorylates and activates the dual-function kinase mitogen activated protein/extracellular signal-related kinase (MEK). MEK phosphorylates and activates MAP kinase. G, MAP kinase enters the nucleus and activates latent transcription factors. (Receptor drawings based on originals by Daniel J. Leahy, Johns Hopkins University, Baltimore, MD. For reference, see Protein Data Bank [PDB; www.rcsb.org] file 2AHX for the unliganded receptor and Bouyain S, Longo PA, Li S, et al. The extracellular region of ErbB4 adopts a tethered conformation in the absence of ligand. Proc Natl Acad Sci U S A. 2005;102:15024–15029.)
Both yeast and mammals anchor two or three of the kinases in certain MAP kinase pathways to a common scaffold protein. Physical association of the enzymes insulates these pathways from parallel pathways but precludes amplification.
Growth Factor Receptor Tyrosine Kinase Pathway Through Ras to Mitogen-Activated Protein Kinase Protein and polypeptide growth factors control the expression of genes required for growth and development. For example, the protein epidermal growth factor (EGF) controls proliferation and differentiation
of epithelial cells in vertebrates. Platelet-derived growth factor (PDGF) stimulates the proliferation of connective tissue cells required to heal wounds (see Fig. 32.11). Growth factor signaling pathways transfer information from the cell surface through at least seven different protein molecules to the nucleus (Fig. 27.6). Conservation of the main features of the mechanism in vertebrates, nematodes, and flies made it possible to combine information from different systems. Genetic tests identified the components and established the order of their interactions. Many components were identified independently as oncogenes and by biochemical isolation and reconstitution of individual steps.
Information flows from growth factors to the nucleus as follows: 1. Growth factors bind to the extracellular domain of their receptors, bringing two receptors together either by linking them (see Fig. 24.5A) or inducing a conformational change (see Fig. 24.5B). 2. Dimerization of receptors activates their cytoplasmic tyrosine kinase domains either by transphosphorylating tyrosine residues on the activation loop of their partner (see Fig. 24.5A) or by allosteric interactions (see Fig. 24.5B). The active kinases phosphorylate other tyrosines on the cytoplasmic domain of the receptor. 3. Each newly created phosphotyrosine is flanked by unique amino acids that form specific binding sites for SH2 domains of downstream effectors, including phospholipase Cγ1, phosphatidylinositol 3-kinase (PI3K) and a preformed complex of the adapter protein Grb2 with SOS. Grb2 consists of three Src homology domains: SH3/SH2/SH3 (see Fig. 25.10). The SH2 domain binds tyrosine-phosphorylated growth factor receptors. The SH3 domains anchor proline-rich sequences (PPPVPPRR) of SOS, a guanine nucleotide exchange factor for the small GTPase Ras (see Fig. 4.6). 4. Association of Grb2-SOS with the receptor raises its local concentration near Ras, which is anchored to the bilayer by farnesyl and palmitoyl groups. Proximity alone suffices for SOS to activate Ras, by exchanging GDP for GTP, as forced experimental targeting of SOS to the plasma membrane by other means also activates Ras. Ras-GTP is active for some time, as it hydrolyzes GTP very slowly (0.005 s−1). Two mechanisms inactivate Ras-GTP: a GTPase-activating protein (Ras-GAP) binds to the receptor and stimulates GTP hydrolysis; and removal of the palmitate releases Ras from the plasma membrane. 5. Ras-GTP binds and activates Raf-1, a serine/threonine kinase (a MAP kinase kinase kinase). 6. Active Raf-1 phosphorylates and activates the dualfunction protein kinase MEK (also called MAP kinase kinase or MKK1). 7. MEK activates MAP kinase by phosphorylating both serine and tyrosine residues on the activation loop. 8. Active MAP kinase has some cytoplasmic substrates, and also enters the nucleus to phosphorylate and activate transcription factors already bound to DNA (Fig. 27.5). These transcription factors control the expression of genes for proteins that drive the cell cycle (see Fig. 41.9), as well as phosphatases that generate negative feedback by inactivating the kinases along these pathways (Fig. 27.5B). The routes from the cell surface through Ras and MAP kinase to nuclear transcription factors are not simple linear pathways. The signal is amplified at some steps and influenced by both positive and negative feedback loops at multiple levels (Fig. 27.5). For example, negative
CHAPTER 27 n Integration of Signals
475
feedback comes from pathways through phospholipase Cγ1 and PI3K that produce Ca2+ and lipid second messengers that activate protein kinase C (PKC) isoforms (see Fig. 26.6). PKC inhibits growth factor receptors by phosphorylation. Active receptors are also modified by addition of a single ubiquitin, a signal for inactivation by endocytosis (see Figs. 22.13 and 23.2). Growth factor pathways are essential for normal growth and development, but malfunctions can cause disease by inappropriate cellular proliferation. One example is the release of PDGF at the sites of blood vessel injury. Normally, PDGF stimulates wound repair (see Fig. 32.11), but excess stimulation of the proliferation of smooth muscle cells in the walls of injured blood vessels is an early event in the development of arteriosclerosis. Many components of growth factor signaling pathways were discovered during the search for genes that cause cancer. As Jean Marx put it, “growth pathways are liberally paved with oncogene products.”* Several of the genes were identified in cancer-causing viruses as oncogenes capable of transforming cells in tissue culture. Oncogenes include sis, a retroviral homolog of PDGF; erbB, a homolog of the EGF receptor; and raf kinase. Subsequently, the normal homologs of these genes were found to have mutations in human cancers. Cancercausing mutations typically make the protein constitutively active, producing a positive signal for growth in the absence of external stimuli (see Fig. 41.12). Two types of mutations increase the concentration of active Ras-GTP and transmit positive signals for growth in the absence of external stimuli, predisposing individuals to malignant disease (see Fig. 41.12). Point mutations (such as substitution of valine for glycine-12) can constitutively activate Ras by reducing its GTPase activity. Alternatively, mutations that inactivate GTPase-activating proteins (GAPs), such as NF1 (the gene causing neurofibromatosis, the so-called elephant man disease; see Fig. 4.7), reduce the rate that Ras hydrolyzes GTP.
Insulin Pathways to GLUT4 and Mitogen-Activated Protein Kinase The insulin receptor tyrosine kinase triggers an acute response that allows muscle and adipose cells to take up blood glucose following a meal (Fig. 27.7). High blood glucose levels stimulate β cells in the islets of Langerhans of the pancreas to secrete insulin, a small protein hormone. Insulin receptors are found on many cells. Insulin is also a growth factor acting through the MAP kinase pathway. The insulin receptor is a stable, dimeric tyrosine kinase composed of two identical subunits, each consisting of two polypeptides covalently linked by a disulfide bond. In the absence of insulin the extracellular domains hold *Marx J: Forging a path to the nucleus. Science. 1993;260:1588–1590.
476
SECTION VII n Signaling Mechanisms
Insulin receptor Insulin
Flotillin
PTB domain PH domain PIP2 PIP3 F G PI3K IRS
A. Insulin binds Tyrosine kinase domain (inactive)
CAP B. Transphosphorylation Cbl activates tyrosine kinase domains GEF
D
C
E
SHC
Grb2-SOS
Ras-GDP Ras-GTP
Ras-GDP
TC10-GDP TC10-GTP
Inactive
Active (triphosphorylated)
Glucose
Akt
J. Indirectly activates Rabs on vesicle to promote fusion
I. Releases vesicle from cytoplasmic tethers MAP kinase
ThrMAP Tyrkinase
K
H
Glycogen synthase L. Promotes conversion of glucose into glycogen
Vesicle with GLUT4 glucose transporters
FIGURE 27.7 INSULIN SIGNALING PATHWAYS IN AN ADIPOSE CELL. A, Insulin binds the preformed dimeric receptor, allowing the association of the transmembrane domains and bringing together the tyrosine kinase domains in the cytoplasm. B, The tyrosine kinase domains activate each other by transphosphorylation of activation loops (Fig. 27.3F). Receptor kinases then phosphorylate a variety of downstream targets: (C) the adapter protein Cbl, which activates a guanine nucleotide exchange protein (GEF) that in turn activates the small GTPase TC10; (D) the adapter protein SHC, which binds Grb2-SOS and slowly initiates the MAP kinase pathway; (E) the adapter protein IRS, which binds Grb2-SOS and rapidly initiates the MAP kinase pathway; and (F) another IRS phosphotyrosine, which binds phosphatidylinositol 3-kinase (PI3K). G, PI3K phosphorylates PIP2 to make phosphatidylinositol 3,4,5-trisphosphate (PIP3). H, PIP3 binds and activates several protein kinases: Akt (protein kinase B [PKB]), protein kinase Cλ (PKCλ), and PKCζ. I, Activated TC10 initiates the release of glucose transporter 4 (GLUT4) storage vesicles from cytoplasmic tethers. J, Akt phosphorylates and inactivates Rab GTPase-activating proteins (GAPs), activating Rab GTPases on GLUT4 storage vesicles, stimulating their fusion with the plasma membrane. K, GLUT4 transports glucose into the cell. L, Akt indirectly activates glycogen synthase. CAP binds Cbl to the plasma membrane protein flotillin. The phosphotyrosine-binding (PTB) domain of IRS binds phosphotyrosine and the PH domain binds PIP3. (For reference, see Kavran JM, McCabe JM, Byrne PO, et al. How IGF-1 activates its receptor. Elife. 2014;25:3.)
the transmembrane helices and cytoplasmic tyrosine kinase domains apart. Insulin binding to the extracellular domains allows the transmembrane domains to come together. This brings the tyrosine kinase domains into proximity so they can transphosphorylate each other (see Fig. 25.3F), turning on their kinase activities (Fig. 27.7D). The kinases propagate the signal by phosphorylating adapter proteins including IRS (insulin receptor substrate, isoforms 1 to 4), SHC (for SH2 and collagen-like), and Cbl. Each plays a distinct role in the ensuing response. This strategy differs from growth factor receptors, which phosphorylate tyrosines on the receptor itself to create binding sites for effector enzymes with SH2-domains. The short-term effects of insulin are to stimulate glucose uptake from blood (particularly into skeletal muscle and white fat) and the synthesis of glycogen, protein, and lipid. Glucose uptake in these cells depends on the glucose carrier, GLUT4 (see Fig. 15.1 for closely related carriers). In the absence of insulin glucose uptake is limited, because these cells have few GLUT4 uniporters in their plasma membranes. Most GLUT4 molecules are stored in the membranes of vesicles trapped near the Golgi apparatus inside the cell. Insulin stimulates fusion of these GLUT4 storage vesicles with the plasma membrane, increasing the capacity to transport glucose into the cell.
Delivery of GLUT4 to the plasma membrane requires two separate signals from the insulin receptor, one to release the storage vesicles from their tethers and the other to target the vesicles for fusion with the plasma membrane. The first signal begins with phosphorylation of the adapter protein Cbl, which activates a guanine nucleotide exchange factor (GEF). The GEF activates the small GTPase TC10, leading to the release of GLUT4 vesicles trapped near the Golgi apparatus. The second signal begins with PI3K binding to a phosphotyrosine on IRS. PI3K synthesizes phosphatidylinositol 3,4,5trisphosphate (PIP3), which activates protein kinase PKB/ Akt and other kinases. Akt indirectly activates Rab GTPases associated with the GLUT4 vesicles by phosphorylating and inactivating a Rab GAP. Active Rab-GTP promotes fusion of GLUT4 storage vesicles with the plasma membrane. Akt also stimulates the conversion of glucose entering the cell through GLUT4 into its storage form, glycogen, by releasing glycogen synthase (the enzyme that makes glycogen) from inhibition by glycogen synthase kinase (GSK) 3. Inactivating mutations of Akt and the downstream Rab-GAP cause rare forms of diabetes, showing their importance in the response to insulin. Insulin is also a growth factor for some cells, acting through the Ras/MAP kinase pathway to nuclear
477
CHAPTER 27 n Integration of Signals
A. Resting
B. Contact with antigenpresenting cell
ANTIGENPRESENTING CELL
T-cell receptor complex CD4
γε
β α
C. ZAP-70 activated by binding ζ-chain phosphotyrosines
ANTIGENPRESENTING CELL
MHC complex Antigen
ζδε
EXTRACELLULAR SPACE
T CELL
T CELL Lck (inactive)
ITAM
Active Lck phosphorylates ITAM tyrosines
SH2 SH2 Kinase
Cytoplasmic tails of CD3
D. Ribbon diagram of MHC class II with bound peptide
ZAP-70 (active)
ZAP-70 (active) ZAP-70 (inactive)
E. Propagation of signals
ANTIGEN-PRESENTING CELL EXTRACELLULAR SPACE
PIP2
+DAG
IP3
NF-AT
RAFT
PLCγ
Ca2+ Calcineurin
LAT
PKC
Vav RacRac-GDP GTP
PAK
Cytoskeleton
thrMAP MAP thrkinase kinase
NUCLEUS
T CELL
H
F
Raf
SLP 76
NFκB
Gene transcription
Ras GTP
Raf
Grb2SOS
Gads SH2
LAT
Ras GDP
ANTIGEN-PRESENTING CELL ICAM1
Seconds
Minutes
G. Immunological synapse
Hours ANTIGEN-PRESENTING CELL
T CELL
LFA T CELL
FIGURE 27.8 T-LYMPHOCYTE ACTIVATION. A, Drawing of the T-cell receptor (TCR) complex on a resting T lymphocyte including inactive nonreceptor tyrosine kinase Lck and unphosphorylated phosphorylation sites (ITAMs [immunoreceptor tyrosine-based activation motifs]) on the cytoplasmic tails of CD3 subunits. B, An encounter with an antigen-presenting cell with a major histocompatibility complex (MHC)-antigenic peptide complex complementary to the particular TCR initiates signaling. Active Lck phosphorylates various ITAMs. C, The nonreceptor tyrosine kinase ZAP-70 (zeta-associated protein of 70 kD) is activated by binding via its two SH2 (Src homology 2) domains to phosphorylated ITAMs on the zeta chains. D, Ribbon diagrams of MHC II (green) with bound peptide from moth cytochrome c (orange). The main model is reduced in size and tilted 90 degrees forward in the view in the upper right corner, the same orientation as in the panels B, C, E, and H. E, Active ZAP-70 phosphorylates various targets, including the transmembrane protein LAT and the adapter protein SLP76, which then propagate the signal. Phospholipase Cγ binds a LAT phosphotyrosine and produces inositol triphosphate (IP3) and diacylglycerol (DAG). IP3 releases Ca2+ from vesicular stores. Ca2+ activates calcineurin (protein phosphatase 2B), which activates the latent transcription factor nuclear factor–activated T cells (NF-AT). DAG and Ca2+ activate protein kinase C (PKC), which activates latent transcription factor nuclear factor κB (NFκB). Vav, the nucleotide exchange factor of the small GTPase Rac, is activated by binding to SLP76. Grb2-SOS binds another phosphorylated ITAM and initiates the MAP kinase cascade. F, Micrographs of the time course of the interaction of a T cell with an artificial membrane mimicking a specific antigen-presenting cell. Each image comprises a superimposition interference reflection micrograph, showing the closeness of contact as shades of gray (with white being closest apposition), and a fluorescence micrograph, showing TCRs (green) and intercellular adhesion molecule 1 (ICAM1) (red). The stable arrangement of ICAM1 around concentrated TCRs is called an immunologic synapse. G–H, Immunologic synapse with a central zone of TCRs bound to MHC complexes and peripheral ICAM1 bound to the integrin LFA. Gads is an adapter protein; RAFT is a lipid raft. (D, For reference, see PDB file 1KT2 and Fremont DH, Dai S, Chiang H, et al. Structural basis of cytochrome c presentation by IEk. J Exp Med. 2002;195:1043–1052. F, Courtesy M. Dustin, New York University, New York.)
478
SECTION VII n Signaling Mechanisms
transcription factors. The signaling circuit to Ras has two arms that operate on different time scales. The fast pathway, acting within seconds, is through tyrosine phosphorylation of IRS, which binds Grb2-SOS and initiates the MAP kinase pathway. The slow arm, acting over a period of minutes, is through phosphorylation of SHC, which binds larger quantities of Grb2-SOS and slowly initiates a sustained response of the MAP kinase pathway. Normal growth and tissue differentiation of many animals depend on insulin-like growth factors, which act on receptors similar to insulin receptor and use IRS1 to channel growth-promoting signals to the nucleus.
T-Lymphocyte Pathways Through Nonreceptor Tyrosine Kinases Some signaling pathways that control cellular growth and differentiation operate through cytoplasmic protein tyrosine kinases separate from the plasma membrane receptors. The best-characterized pathways control the development and activation of lymphocytes in the immune system. T lymphocytes are the example used in this discussion. T lymphocytes defend against intracellular pathogens, such as viruses, and assist B lymphocytes in producing antibodies (see Fig. 28.9). Antigen-presenting cells activate T cells by presenting peptide antigens on their surface bound to histocompatibility proteins for interactions with T-cell receptors (TCRs) and accessory proteins (Fig. 27.8). Engagement of the TCR triggers a network of interactions among protein tyrosine kinases, adapter proteins, and effector proteins on the inner surface of the plasma membrane. Tyrosine phosphorylation of multiple membrane and cytoplasmic proteins activates three separate pathways to the nucleus. Two pathways activate cytoplasmic transcription factors; the third uses the Ras/MAP kinase pathway to activate transcription factors in the nucleus. The T-cell antigen receptor is a complex of eight transmembrane polypeptides (Fig. 27.8A). The α and β chains, each with two extracellular immunoglobulin-like domains, provide antigen-binding specificity. Similar to antibodies (see Fig. 28.10), one of these immunoglobulin domains is constant and one is variable in sequence. The genes for TCRs are assembled from separate parts, similar to the rearrangement of antibody genes (see Fig. 28.10). Genomic sequences for variable domains are spliced together randomly in developing lymphocytes from a panel of sequences, each encoding a small part of the protein. Assembly of TCRs in the endoplasmic reticulum requires six additional transmembrane polypeptides, each with one or more short sequence motifs, called immunoreceptor tyrosine activation motifs (ITAMs), in their cytoplasmic domains. This combinatorial strategy creates a diversity of T-cell antigen receptors, with one type expressed on any given
T cell. Variable sequences of α and β chains provide binding sites for a wide range of different peptide antigens bound to cell surface proteins on antigen-presenting cells. These are collectively termed the major histocompatibility complex (MHC) antigens (Fig. 27.8D). The peptide antigens are fragments of viral proteins or other foreign proteins that are degraded inside the cell, inserted into compatible MHC molecules during their assembly in the endoplasmic reticulum, and transported to the cell surface. The expression of single types of α and β chains provides each individual T cell with specificity for a particular peptide. Although T-cell antigen receptors bind specifically, their affinity for the peptide-MHC complex is low (Kd [dissociation equilibrium constant] in the range of 10 µM). Given the small number (hundreds) of unique MHC-peptide complexes found on the target cell surface, this low affinity would not be sufficient for a lymphocyte to form a stable complex with an antigen-presenting cell. Accessory proteins called coreceptors, such as CD4 (also the receptor for human immunodeficiency virus [HIV]) and CD8 (see Fig. 30.3), bind directly to any MHC protein and reinforce the interaction of the two cells. Activation by antigen stimulation also depends on parallel nonspecific stimuli from inflammatory mediators. Two classes of cytoplasmic protein tyrosine kinases transmit a signal from the engaged TCR to effector systems. The first class of kinases, including Lck and Fyn, are relatives of the Src tyrosine kinase (see Fig. 25.3C), the first oncogene to be characterized (Box 27.5). These tyrosine kinases are anchored to the plasma membrane by myristoylated N-terminal glycines and inhibited by a phosphotyrosine near the C-terminus (see Fig. 25.3C). This tyrosine is phosphorylated by a kinase, Csk, and dephosphorylated by the transmembrane protein tyrosine phosphatase, CD45 (see Fig. 25.6B). Apparently, CD45 keeps Lck partially dephosphorylated and therefore partially active in resting lymphocytes. Zeta-associated protein–70 kD (ZAP-70) is the most important of the second class of protein tyrosine kinases in this pathway. Two SH2 domains allow ZAP-70 to bind tyrosine-phosphorylated ITAMs on ζ chains. Physical contact of a T lymphocyte with an antigenpresenting cell carrying an MHC-peptide specific for its TCR generates multiple signals as follows: 1. Engagement of TCRs leads to activation of Lck by dephosphorylation of the inhibitory C-terminal tyrosine and phosphorylation of its activation loop. 2. Active Lck phosphorylates ITAMs on the various TCR accessory chains. 3. ZAP-70 is activated by binding phosphorylated ITAMs and phosphorylation of its activation loop by Lck. 4. Active ZAP-70 phosphorylates targets that include the key transmembrane protein LAT and the adapter protein SLP76, which then propagate the signal. 5. Signals reach the nucleus by three pathways. First, phospholipase Cγ1 is activated by binding a
CHAPTER 27 n Integration of Signals
BOX 27.5 Src Family of Protein Tyrosine Kinases The founding member of the Src family of protein tyrosine kinases has a prominent place in modern biology. During the 1920s, Peyton Rous discovered the first cancer-causing virus in a mesodermal cancer of chickens called a sarcoma. Later, the Rous sarcoma virus was found to be a virus with an RNA genome (a retrovirus). Comparisons with similar viruses that did not cause cancer revealed that one gene, named src, is responsible for transforming cells into cancer cells. Many years later, a gene very similar to src was found in normal chicken cells. The cellular protein product, c-Src, is a carefully regulated protein tyrosine kinase that controls of cellular proliferation and differentiation. Mutations in the gene for viral src, v-src, activate its protein product constitutively, driving cells to proliferate and contributing to the development of cancer. The family of Src-like proteins shares a common structure (see Fig. 25.3C). Five functionally distinct segments are recognized in the sequences. An N-terminal myristic acid anchors the protein to the plasma membrane. Without this modification, the protein is inactive. The next domains are the founding examples of Src homology domains SH3, which bind proline-rich peptides, and SH2, which bind peptides containing a phosphorylated tyrosine (see Fig. 25.10). The kinase domain is followed by a tyrosine near the C-terminus. Phosphorylation of this tyrosine and its intramolecular binding to the SH2 domain lock the kinase in an inactive conformation. Dephosphorylation of the C-terminal tyrosine and phosphorylation of the activation loop activate the kinase. Expression of c-Src is highest in brain and platelets, but a null mutation in mice produces relatively few defects, except in bones, where a failure of osteoclasts to remodel bone leads to overgrowth, a condition called osteopetrosis (see Fig. 32.6).
phosphotyrosine on LAT and by its own tyrosine phosphorylation. Active phospholipase Cγ1 produces inositol triphosphate (IP3) and diacylglycerol. Release of Ca2+ from vesicular stores by IP3 activates calcineurin (protein phosphatase 2B [see Fig. 25.6A]). This activates the latent transcription factor nuclear factor–activated T cells (NF-AT). Second, Grb2-SOS binds another phosphotyrosine on LAT and initiates the MAP kinase cascade by activating Ras. Third, Vav, a nucleotide exchange factor for small GTPases, is anchored indirectly to LAT. This initiates a pathway that degrades IκB, freeing nuclear factor κB (NF-κB) (see Fig. 10.21C) to enter the nucleus. 6. The signal ultimately reaches the cytoskeleton via Vav, which catalyzes the exchange of GTP for GDP on the Rho-family GTPase Rac, which leads to actin filament assembly (see Fig. 38.7). When a T cell recognizes an antigen-presenting cell with an appropriate peptide bound to MHC on
479
its surface, the TCRs and adhesion proteins in the interface between the cells rearrange to form an “immunologic synapse” (Fig. 27.8F). TCRs initially gather around a region of contact between integrins (lymphocyte function–associated antigen) on the T cell and immunoglobulin-cellular adhesion molecules (ICAMs) on the antigen-presenting cell. Over time, actin polymerization around the periphery of this zone drives TCR complexes into a central region where the plasma membranes are close together surrounded by a ring of relatively tall adhesion molecules (Fig. 27.8G–H). Active TCR clusters in this central region generate a signal to the nucleus and are then internalized and degraded. The best immunosuppressive drugs used in human organ transplantation block lymphocyte proliferation by inhibiting calcineurin, the phosphatase that activates NF-AT. These drugs completely block T-cell activation providing that they are given within 1 hour of the stimulus; that is, before gene expression has been initiated. Cyclosporine and FK506 bind to separate cytoplasmic proteins, cyclophilin, and FK-binding protein. Both of these drug-protein complexes bind calcineurin and inhibit its phosphatase activity. Considering that many cells express calcineurin, the effects of these drugs on lymphocytes is amazingly specific, with relatively few side effects. Specificity arises from the low concentration of calcineurin in lymphocytes: only 10,000 molecules in T cells compared with 300,000 in other cells. Hence, low concentrations of inhibitor can selectively block calcineurin in T lymphocytes. Cyclosporine made human heart and liver transplantation feasible. The response to TCR activation depends on the particular state of differentiation of the T cell that encounters its partner antigenic peptide. Stimulation causes some T cells to secrete toxic peptides that kill the antigenpresenting cell, others synthesize and secrete lymphokines (immune system hormones), others proliferate and differentiate, and yet others commit to apoptosis (see Fig. 46.9).
Cytokine Receptor, JAK/STAT Pathways Many polypeptide hormones and growth factors (collectively called cytokines) regulate gene expression through a three-protein relay without a second messenger—the most direct signal transduction pathway from extracellular ligands to the nucleus (Fig. 27.9). Growth hormone uses this mechanism to drive overall growth of the body, erythropoietin directs the proliferation and maturation of red blood cell precursors, and several interferons and interleukins mediate antiviral and immune responses. Slime molds and animals use these pathways, but these proteins are not present in fungi or plants. The three components in these pathways are a dimeric plasma membrane receptor that lacks intrinsic enzymatic activity, a tyrosine kinase (JAK) constitutively
480
SECTION VII n Signaling Mechanisms
A. Ligand binds Ligand
B. Transphosphorylation D. Tyrosine phosphorylation releases STATs
JAK
E. Secondary STAT phosphorylation via growth factor receptor tyrosine kinase pathway
SH2
(Negative feedback)
STAT
C. STAT binds phosphotyrosine
SH2
F. Reciprocal binding of SH2 to phosphotyrosine forms STAT dimer
STAT
Synthesis of SOCS1
CYTOPLASM
G. P-STAT dimer enters nucleus
DNA
SOC S1
H. STAT activates expression of various genes including SOCS1
NUCLEUS
FIGURE 27.9 CYTOKINE JAK KINASE/SIGNAL TRANSDUCER AND ACTIVATOR OF TRANSCRIPTION (STAT) SIGNALING PATHWAY. A, Cytokine binds a preformed receptor dimer (see Fig. 24.7), changing arrangement of the JAK tyrosine kinases bound to the cytoplasmic domains of the receptor. B, Active JAKs phosphorylate each other and tyrosines on the receptor. C, The SH2 (Src homology 2) domain of the latent transcription factor STAT binds a receptor phosphotyrosine. D, JAK phosphorylates the STATs, which then dissociate from the receptor. E, Growth factor receptor tyrosine kinases can also activate STATs. F, STATs form active dimers by reciprocal SH2-phosphotyrosine interactions. G, The STAT dimer enters the nucleus. H, The STAT dimer activates the expression of various genes. One of these genes encodes SOCS1, which creates negative feedback by inhibiting further STAT activation. JAK, just another kinase.
associated with each receptor subunit, and a latent, cytoplasmic transcription factor called a STAT (signal transducer and activator of transcription). JAKs, originally given the lighthearted name “just another kinase,” are often called Janus kinases (for the Greek god who opens doors). The N-terminal halves of JAKs mediate their association with receptors. In the absence of a ligand the kinases are inactive owing to interactions with the partner kinase (Fig. 24.6). Some cytokine receptors bind and activate a single type of JAK; others are promiscuous. The signal from ligand binding moves from the cytokine receptor to JAK to STAT to the nucleus (Fig. 27.9) as follows: 1. Ligand binding changes the conformation of the receptor and relieves the mutual inhibition of the associated JAKs (Fig. 24.6). 2. The JAKs activate each other by transphosphorylation and phosphorylate tyrosines on the cytoplasmic tails of the receptors, creating docking sites for STATs. 3. SH2 domains target preformed STAT dimers to phosphotyrosine binding sites on the receptor. 4. The JAK phosphorylates the STAT, changing the orientation of the subunits. This dissociates the STAT from the receptor. 5. Active STAT dimers enter the nucleus and activate expression of various genes. Three different mechanisms turn down the response to cytokine activation. Phosphatases inactivate the
receptor, kinase, and intranuclear STATs. Endocytosis also turns off active receptors. A slowly acting, negative feedback loop limits the duration of the response. One of the genes expressed in response to STAT encodes SOCS1. Once synthesized, the SOCS1 protein inhibits further STAT activation by interacting with the cytokine receptor. Selective expression of specific cytokine receptors, four JAKs, and six STATs prepares differentiated mammalian cells to respond specifically to various cytokines. Active STATs are either homodimers or heterodimers of two different STATs. A variety of STATs, with some unique and some common subunits, binds regulatory sites of genes required to activate the target cells. The products of genes controlled by STATs not only contribute to differentiated cellular functions, but some also drive proliferation. Accordingly, loss of JAK function causes certain immune deficiencies, while patients with mutations in STAT5b are resistant to growth hormone and fail to grow. The opposite effect follows from a mutation that renders JAK2 constitutively active: red blood cells proliferate out of control, independent of stimulation by erythropoietin. The three-protein pathway from a cytokine receptor to JAK to activated STAT is appealing in its simplicity, but in reality, these pathways do not operate in isolation. On one hand, converging signals from EGF- and PDGFreceptors can phosphorylate and activate STATs, a second input to STAT-responsive genes (Fig. 27.9). On
CHAPTER 27 n Integration of Signals
the other hand, some cytokine receptors can regulate gene expression by acting through Shc and Grb2-SOS to Ras and MAP kinases and other pathways.
481
such as SMAD2 and SMAD3 (Fig. 27.10). Phosphorylated R-SMADs dissociate from RI and associate with SMAD4, called a co-SMAD, because it is not subjected to phosphorylation itself. A trimer consisting of two R-SMADs and one co-SMAD enters the nucleus and associates with other DNA binding proteins to activate or inhibit transcription of specific genes as well as influencing chromatin structure. The number of targeted genes varies from a handful to hundreds depending on the other transcription factors produced by the cell and epigenetic modifications of the target gene chromatin. Other SMADs regulate these pathways by inhibiting phosphorylation of R-SMADs. The SMAD pathway activated by TGF-β regulates cellular proliferation and differentiation of many cell types, including epithelial and hematopoietic cells. Although its name implies that it should drive transformation, TGF-β actually stops the cell cycle of normal cells in G1 by promoting expression of negative regulators of cyclindependent kinases (see Fig. 41.3). On the other hand, TGF-β can stimulate the growth of some cancers. In accord with the ability of TGF-β pathways to inhibit cellular growth, many human tumors have loss-of-function mutations in the genes for TGF-β receptors or SMADs. Mutations in an accessory receptor for TGF-β cause malformed blood vessels in the human disease hereditary hemorrhagic telangiectasia. Mice with homozygous
Serine/Threonine Kinase Receptor Pathways Through SMAD All metazoans use a family of dimeric polypeptide growth factors related to transforming growth factor-β (TGFβ [see Fig. 24.7]) to specify developmental fates during embryogenesis and to control cellular differentiation in adults. In humans more than 40 genes in this family of ligands are divided into two classes: (a) those closely related to TGF-β and activins and (b) a large family of bone morphogenetic proteins. All activate a short pathway consisting of receptor serine/threonine kinases and a family of eight mobile transcription factors called SMADs (Sma- and Mad-related proteins). The receptors consist of two types of subunits called RI (seven isoforms in human) and RII (five isoforms). Various combinations of these receptors bind about 30 different ligands, some of which antagonize each other. Ligand binding brings together two RI and two RII receptors, allowing the RII receptors to activate the RI receptors by transphosphorylation (see Fig. 24.7). Active RI receptors phosphorylate “regulated SMADs” (R-SMADs),
A. TGFβ brings together two RI receptors with two RII receptors
RI • RII • TGF-β
B. RII kinases phosphorylate and activate RI kinases
SARA
C. R-SMAD binds RI kinase in complex with SARA MH1
Auto inhibited R-SMAD (2/3) MH2
Inactive RI kinase domain
Active RII kinase domain
MH2 MH1
D. RI kinase phosphorylates R-SMAD, promoting dissociation from the kinase and SARA
MH1 MH2 Co-SMAD
F. Active SMAD trimer enters nucleus DNA
or NUCLEUS
Co-SMAD E. Two R-SMADs bind a Co-SMAD
G. SMADs associate with other DNA-binding proteins to activate or inhibit transcription of specific genes
CYTOPLASM
FIGURE 27.10 TRANSFORMING GROWTH FACTOR (TGF)-β/SMAD SIGNALING PATHWAY. A, Binding of a TGF-β dimer assembles a complex consisting of two RII receptors and two RI receptors. B, RII phosphorylates and activates RI. C, An autoinhibited R-SMAD binds RI in a complex with the adapter SARA. D, RI kinase phosphorylates R-SMAD, which promotes its dissociation from RI. E, A co-SMAD binds an R-SMAD dimer to form an active trimer. F, The SMAD trimer enters the nucleus. G, The SMAD trimer associates with other DNA-binding proteins to activate or inhibit transcription of specific genes.
482
SECTION VII n Signaling Mechanisms
loss-of-function mutations in genes for the components of the TGF pathway die during embryonic development. Like other signaling pathways, these receptors and SMADs do not operate in isolation. MAP kinases and cyclin dependent kinases can phosphorylate SMADs, and active RI receptors can activate MAP kinase pathways and other signaling pathways.
Bacterial Chemotaxis by a TwoComponent Phosphotransfer System The two-component system (Box 27.6) regulating bacterial chemotaxis (Fig. 27.12) is the best-understood signaling pathway of any kind. E. coli cells use five types of plasma membrane receptors to sense a variety of
BOX 27.6 “Two-Component” Signaling Prokaryotes, fungi, and plants transduce stimuli ranging from nutrients to osmotic pressure using signaling systems consisting of as few as two proteins, a receptor-linked histidine kinase, and a “response regulator” activated by phosphorylation of an aspartic acid (Fig. 27.11). Extensive collections of mutants in these pathways and sensitive single-cell assays for responses, such as flagellar rotation, provide tools for rigorous tests of concepts and mathematical models derived from biochemical experiments on isolated components. Two-component systems are abundant in bacteria with 32 response regulators and 30 histidine kinases in Escherichia coli. Archaea have genes for up to 24 response regulators. Some eukaryotes have a few two-component systems, but these genes were lost in metazoans. The slime mold Dictyostelium has more than 10 histidine kinases, whereas fungi have just one or two of these systems. Plants use a twocomponent system to regulate fruit ripening in response to the gas ethylene. Two-component receptors either may include a cytoplasmic histidine kinase domain (Fig. 27.11C) or may bind a separate histidine kinase, such as the aspartate chemotactic receptor Tar (Fig. 27.11B). Tar consists of two identical subunits. Three Tar dimers are anchored together at their bases in the cytoplasm (Fig. 27.11B). Binding of aspartic acid between the extracellular domains of two subunits changes their orientation by a few degrees. Transmission of this conformational change across the membrane alters the activity of CheA, a histidine kinase associated with the cytoplasmic domains of the receptor. Histidine kinases have a conserved catalytic domain of approximately 350 residues that is structurally unrelated to eukaryotic serine/threonine/tyrosine kinases (shown in Fig. 25.3). Another domain allows them to form homodimers. Histidine kinases are incorporated into a wide variety of proteins, including transmembrane receptors (Fig. 27.11C) and cytoplasmic proteins such as CheA (Fig. 27.11B). The catalytic domain transfers the γ-phosphate from adenosine triphosphate (ATP) to just one substrate, a histidine residue of its homodimeric partner. This histidine is usually located in the dimerization domain. All response regulators have a domain of approximately 120 residues folded like CheY (Fig. 27.11B; also see Fig. 3.7). Transfer of phosphate from the phosphohistidine of a kinase to an invariant aspartic acid changes the conformation of the response regulator. Most response regulators such as CheB (Fig. 27.11B) and OmpR (Fig. 27.11C) are larger than CheY,
having C-terminal effector domains. Many effector domains, including OmpR, bind DNA and regulate transcription of specific genes when the response regulator is activated by aspartate phosphorylation. Other response regulators are included as a domain of the histidine kinase itself. Reversible phosphorylation transfers information through two-component systems. The mechanism differs fundamentally from eukaryotic kinase cascades, which transfer phosphate from ATP to serine, threonine, or tyrosine, forming phosphoesters at every step. By contrast, twocomponent systems first transfer a phosphate from ATP to a nitrogen of a histidine of the kinase, the first of the two protein components. The high-energy his~P phosphoramidite bond is unstable, so the phosphate is readily transferred to the side chain of an aspartic acid of the response regulator (RR): ATP + kinase-his ↔ ADP + kinase-his~P kinase-his~P + RR-asp ↔ kinase-his + RR*-asp~P RR*-asp~P + H2O ↔ RR-asp + phosphate Phosphorylation activates response regulators (RR*) by changing their conformation. Details differ depending on the response regulator. In the case of OmpR, phosphorylation relieves autoinhibition of the DNA-binding domain (Fig. 27.11C). Phosphorylation of CheY reveals a binding site for the flagellar rotor. The signal dissipates by dephosphorylation of the response regulator, either by autocatalysis or stimulated by accessory proteins. Lifetimes of the highenergy aspartic acylphosphate vary from seconds to hours. A minimal two-component system, such as a bacterial osmoregulatory pathway (Fig. 27.11C), consists of a dimeric plasma membrane receptor with a cytoplasmic histidine kinase domain and a cytoplasmic response regulator protein. Signal transduction is carried out in four steps. A change in osmolarity alters the conformation of the receptor, activating the kinase activity of its cytoplasmic domain. The kinase phosphorylates a histidine residue on the other subunit of the dimeric receptor. This phosphate is transferred from the receptor to an aspartic acid side chain of the response regulator protein OmpR. Phosphorylation changes the conformation of the response regulator domain of OmpR, allowing its DNA-binding domain to activate the expression of target genes.
CHAPTER 27 n Integration of Signals
A. Tar structure α3 α2 α4
B. Chemotaxis Receptor
Transmembrane N C
Linker
Methylation
α5
Che R Methyl
D
Che B
RR 5
3 5 4 ATP ATP 1 2
Che A
RR
H
N
1
2
1
2
3 3
N
4
C
RR Che Y Rotary motor
C. Osmoregulation
4
5
4
5
Kinase
H
3
1
Che Y
N
C C
Che A dimer
MethylD esterase
D
H Signaling
Che W + ADP ATP
Methyl
Che W α4
Ligand
Receptor
Ligand binding
α1
483
N
Response C Che B regulators
RR Receptor
Env Z dimer N N
Sensor CheA Tar helices
Env Z
HPt
H
CheW
H
ADP +
ATP ATP
Kinase
ATP
D
OmpR
RR
C C
D N
RR
DNA binding
C
OmpR
Gene expression
FIGURE 27.11 TWO-COMPONENT BACTERIAL SIGNALING SYSTEMS. A, Atomic model of the aspartate receptor Tar. The atomic structures of the extracellular and cytoplasmic domains were determined by X-ray crystallography. The transmembrane α-helices are models based on the primary structure. The two polypeptides are shown in red and blue. Each polypeptide starts in the cytoplasm and passes twice through the lipid bilayer. B, Bacterial chemotaxis signaling proteins. Scale models of the molecular components and pathway of information transfer. The ribbon diagrams of domains shown on the right are color coded in the molecular models on the left. An accessory protein, CheW, facilitates binding of the histidine kinase CheA to the aspartate receptor Tar. CheY and CheB are response regulators. CheR is a methyl transferase. C, Bacterial osmoregulation. The histidine kinase forms the cytoplasmic domain of the receptor. OmpR is the response regulator with a DNAbinding domain. Scale models of the molecules (left) and pathway of information transfer (right). (A, Modified from material courtesy S.H. Kim, University of California, Berkeley. For reference, see PDB file 1QU7 and Kim KK, Yokota H, Kim SH. Four helical-bundle structure of the cytoplasmic domain of a serine chemotaxis receptor. Nature. 1999;400:787–792. B–C, Based on material courtesy A.M. Stock, Robert Wood Johnson Medical School. For reference, see Stock AM, Robinson WL, Goudreau PN. Two-component signal transduction. Annu Rev Biochem. 2000;69:183–215.)
different chemicals, which range from nutrients to toxins. These receptors are also called methyl-accepting chemotaxis proteins, as they are regulated by methylation. The most abundant receptor, with thousands of copies per cell, is Tar (Fig. 27.11A–B), the receptor for the nutrients aspartic acid (Tar-D) and maltose, protons (as part of pH sensing), temperature, and the repellent nickel. The receptors form clusters of various sizes over the cell membrane, with the highest concentration at one end of the cell (Fig. 27.12A). This polarized distribution facilitates interactions between receptor molecules but has nothing to do with sensing the direction of chemical gradients.
The chemotactic signaling system guides swimming bacteria toward attractive chemicals and away from repellents in a biased random walk. Environmental chemicals influence the behavior of the cell by biasing the rotation direction of the flagellar motor (see Figs. 38.24 and 38.25 for details of the motor itself). In the absence of a response regulator, the motor turns counterclockwise, and the bacterium swims smoothly in a more or less linear path. When the flagella turn clockwise, the bacterium tumbles about in one place. A tumble allows a bacterium to reorient its direction randomly, so when it resumes smooth swimming, it usually heads in a new direction. In the absence of
484
SECTION VII n Signaling Mechanisms
A. Messenger paths
D. Chemotaxis
Nose spot with complex of chemoreceptors including Tar
Bacteria heading toward attractant tend to keep running
Flagella
Inner membrane Peptidoglycan layer Outer membrane
B. Steady swimming
Flagella form a bundle
Flagella and bundle rotate counterclockwise
Postulated path of Che Y Rotary motor
C. Tumble
Bundle of flagella flies apart
ATTRACTANT
Bacterium tumbles in place
Reoriented bacteria tend to keep running toward attractant Bacteria heading away from attractant tend to tumble Tumbling bacteria reorient randomly
Reoriented bacteria run toward attractant
Flagella rotate clockwise
Bacteria tumble until a random tumble orients them toward the attractant
FIGURE 27.12 MOTILITY OF BACTERIA IS DETERMINED BY THE DIRECTION OF ROTATION OF THEIR FLAGELLA. A, Drawing of Escherichia coli showing the flagella and arrangement of signal transduction components with the highest concentration of receptors at one pole of the cell. B, Flagella that rotate counterclockwise (viewed from the tip of the flagella) form a bundle that pushes the cell smoothly forward. C, When flagella rotate clockwise, the bundle flies apart, and the cell tumbles in place. D, An attractive chemical biases movement toward its source by modulating the frequency of runs and tumbles.
chemoattractants, bacteria randomly switch between periods of swimming that last about 0.9 second and brief tumbles lasting about 0.1 second, allowing for random reorientations every second. A gradient of chemical attractant promotes the length of runs up the gradient by suppressing tumbling if the concentration of attractant increases over time (Fig. 27.12D). A two-component signaling pathway senses the concentration of attractant and controls the frequency of tumbling through phosphorylation of the response regulator CheY, which acts on the flagellar motor. Most components of the system were discovered by mutagenesis and named “Che” for chemotaxis gene with a lowercase “p” to indicate phosphorylation. Ligand-free Tar receptors stimulate the phosphorylation of the associated histidine kinase CheA, which is bound to the receptor by a “scaffold” protein CheW. CheAp activates the response regulator, CheY, by transferring phosphate from histidine to aspartic acid 57 (D57) of CheY. CheYp has a higher affinity for the flagellar motor than CheY, so ligand-free receptors maintain a steady state with the rotors partially saturated
with CheYp. With several bound CheYps, the motor switches from its free-running, counterclockwise state to a brief clockwise tumble approximately once per second. Information about aspartate in the environment flows rapidly through the pathway as changes in the concentrations of the phosphorylated species CheAp and CheYp. A key point is that Tar with bound aspartate, Tar-D, ceases to activate histidine phosphorylation of CheA. Hence aspartate binding to Tar reduces the saturation of the flagellar motors with CheYp and the frequency of tumbles. For the cell to respond to aspartate on a subsecond time scale, an accessory protein, CheZ, is required to increase the rate of CheYp dephosphorylation more than 100-fold from its slow spontaneous rate of 0.03 s−1. Given this fast dissipation of CheYp, phosphate must flow constantly from adenosine triphosphate (ATP) to CheAp to CheYp to maintain a tumbling frequency of about 1 s−1 in the absence of an attractant (Fig. 27.13A). (Most other two-component pathways respond in minutes rather than milliseconds, because
CHAPTER 27 n Integration of Signals
A. No aspartate
B. Early events after sudden increase in aspartate concentration
5
Outer membrane
5
C. Late events after a sudden increase in aspartate concentration
Asp
1
Asp
1
485
5
Asp Asp
Asp
1
Che R
Me Me
Asp
PM Methyl
Methyl Che R
Che B (on)
Che R
Strong
ATP
3
ADP
4 2
Weak
Methyl
Methyl
Che Y Che A
Che Z
Che B (on)
Methyl
4 2 Che Z
Me
Che B (off) 4 2
ATP ADP
3
Che Z
Che Y FIGURE 27.13 SIGNALING DURING BACTERIAL CHEMOTAXIS. Drawing of the signaling system under three conditions. A, Absence of aspartate. Ligand-free Tar (1) allows CheA to phosphorylate CheY and CheB (3). Constant dephosphorylation of CheYp drives a cycle of phosphorylation (2). The steady-state concentration of CheYp keeps the motor partially saturated (4). The partially saturated motor turns counterclockwise 90% of the time (runs [thick arrow]) and clockwise 10% of the time (tumbles [thin arrow]) (5). B, Rapid response to the presence of aspartate. Aspartate binding turns off Tar (1). Constant dephosphorylation depletes CheYp on a time scale of tens of milliseconds (2). CheY dissociates from the motor (4). The motor, without CheYp, rotates counterclockwise, so the bacterium runs continuously (5). C, Slower, adaptive response to the presence of aspartate. Inactive CheA stops phosphorylating CheB, allowing dephosphorylation of CheBp on a time scale of seconds; this inactivation of CheB and the higher affinity of CheR for inactive Tar result in higher methylation of Tar (1). Even with bound aspartate, methylated Tar is partially active, allowing phosphorylation of CheA (2). CheY is phosphorylated (3). CheYp rebinds the motor (4). The flagella turn clockwise part of the time, returning to the steady-state frequency of runs and tumbles (5).
dephosphorylation of the response regulator is much slower. The following sections examine chemotaxis on the system level.) Temporal Sensing of Gradients Unlike eukaryotic cells, most bacteria are too small and move too fast to detect a spatial gradient along the length of the cell body. Instead, they sense the gradient as a perceived change in concentration of attractant or repellent as a function of time. When a bacterium swims up a gradient of chemoattractant, the concentration of attractant increases with time, and the signaling mechanism suppresses tumbling. Fewer tumbles result in a biasing of smooth swimming toward the attractant. When a cell swims down the gradient, tumbling is more frequent, allowing for reorientation. A sudden increase in the concentration of aspartate yields a smooth swimming response within 200 milliseconds as a result of rapid reequilibration of the concentrations of all the cytoplasmic signaling components (Fig. 27.13B). Aspartate binds Tar and inhibits autophosphorylation of CheA. CheYp has a half-life of less than 100 milliseconds, so the concentrations of CheAp and CheYp decrease rapidly. CheYp dissociates from the flagellar motor and the tendency of the motor to stay in the counterclockwise, smooth swimming direction increases. If the concentration change with time is due to a persistent gradient of aspartate, the bacterium tends to move on average up the gradient toward the source, although
it will make several runs up and down the gradient during this time. The opposite sequence of events takes place if a bacterium swims down a gradient of aspartate. The fraction of Tar with bound aspartate declines, CheAp and CheYp concentrations rise, and tumbling is more frequent, providing opportunities to reorient and swim back up the gradient. Such is the case with avoiding a repellant gradient, as well. Adaptation If the aspartate concentration suddenly increases everywhere, bacteria respond quickly with smooth swimming, but within tens of seconds to minutes, they return to their normal frequency of intermittent tumbling. Thus, the steady-state tumbling frequency depends on changes in the concentration of aspartate relative to background levels rather than the absolute concentration. This remarkable capacity to adapt is accomplished by a feedback control mechanism provided by reversible methylation of the receptor (Fig. 27.13C). Two relatively slow enzymes determine the level of Tar methylation (Fig. 27.11B). CheR adds methyl groups to four glutamic acid residues on each receptor polypeptide, whereas the response regulator CheB removes them. Methylated Tar has a somewhat lower affinity for aspartate than unmethylated Tar, but Me-Tar with bound aspartate is more effective at stimulating CheA phosphorylation than Tar with bound aspartate.
486
SECTION VII n Signaling Mechanisms
CheR is constitutively active but sensitive to the overall metabolic state of the cell, as it depends on the concentration of S-adenosyl methionine, a methyl donor that is used in many metabolic reactions. The CheB methylesterase is autoinhibited by its response regulator domain but activated by phosphorylation by CheAp. Thus, CheB methylesterase activity depends on the concentration of CheAp, and rate of demethylation determines the level of receptor methylation. Adaptation occurs because aspartate binding to Tar activates two different pathways on different time scales. On a millisecond time scale, the concentrations of both CheAp and CheYp decline, CheYp dissociates from the motor, and the cell swims smoothly. As Tar molecules convert to the aspartate-bound state, CheR has a greater affinity for demethylated glutamate residues on the inactive Tar-D and begins to methylate them, which occurs on a time scale of seconds. As Tar molecules accumulate methyl groups, the reduced activity associated with ligand binding is slowly reversed. In this way, Tar molecules with bound aspartate have their activity restored to the baseline level that they had before the aspartate concentration increased. This robust adaptation mechanism is an integral feedback system, just like a thermostat on a heater. The system works in part, because CheR preferentially methylates inactive receptors and CheBp preferentially demethylates active receptors. Extended Range of Response An amazing feature of this system is its ability to respond with fast changes in flagellar rotation and slow adaptation to changes of just a few percentage points in aspartate concentration over a range of five orders of magnitude. Clearly, a simple bimolecular reaction of aspartate with Tar cannot change the fractional saturation of Tar over such an extended range of concentrations. This extended range of sensitivity is valuable for the survival of the bacterium and depends on amplification at the level of the receptor whereby aspartate binding to one Tar activates many surrounding Tars in the receptor clusters at the end of the cell. This physical communication is achieved by a lattice of CheA and CheW between the cytoplasmic tips of the receptors. Bacterial chemotaxis illustrates some of the classical features of signaling pathways, including high sensitivity due to amplification at the level of CheA phosphorylation, feedback control through methylation of Tar, and branching networks that respond on different time scales to the same stimulus. The mechanism has been tested thoroughly by mutating all the signaling components and observing the consequences. Furthermore, random variations in the numbers of these proteins allow individuals in populations of bacterial cells to take
advantage of a wide range of environments and improve the fitness of the community. ACKNOWLEDGMENTS We thank Jonathan Bogan and Nicholas Frankel for suggestions on revisions of this chapter. SELECTED READINGS Babon JJ, Lucet IS, Murphy JM, Nicola NA, Varghese LN. The molecular regulation of Janus kinase (JAK) activation. Biochem J. 2014;462: 1-13. Bogan JS. Regulation of glucose transporter translocation in health and diabetes. Annu Rev Biochem. 2012;81:507-532. Boucher J, Kleinridders A, Kahn CR. Insulin receptor signaling in normal and insulin-resistant states. Cold Spring Harb Perspect Biol. 2014;6:a009191. Bray D. The propagation of allosteric states in large multiprotein complexes. J Mol Biol. 2013;425:1410-1414. Call ME, Wucherpfennig KW. The T cell receptor: Critical role of the membrane environment in receptor assembly and function. Annu Rev Immunol. 2005;23:101-125. Capra EJ, Laub MT. Evolution of two-component signal transduction systems. Annu Rev Microbiol. 2012;66:325-347. Dustin ML, Groves JT. Receptor signaling clusters in the immune synapse. Annu Rev Biophys. 2012;41:543-556. Ferrell JE Jr. Self-perpetuating states in signal transduction: Positive feedback, double-negative feedback and bistability. Curr Opin Cell Biol. 2002;14:140-148. Frankel NW, Pontius W, Dufour YS, et al. Adaptability of non-genetic diversity in bacterial chemotaxis. Elife. 2014;3:doi:10.7554/eLife. 03526. Goh LK, Sorkin A. Endocytosis of receptor tyrosine kinases. Cold Spring Harb Perspect Biol. 2013;5:a017459. Hubbard SR. The insulin receptor: both a prototypical and atypical receptor tyrosine kinase. Cold Spring Harb Perspect Biol. 2013;5: a008946. Johnson GL, Lapadat R. Mitogen-activated protein kinase pathways mediated by ERK, JNK and p38 protein kinases. Science. 2002;298: 1911-1912. Jones CW, Armitage JP. Positioning of bacterial chemoreceptors. Trends Microbiol. 2015;23:247-256. Krzysztof P. G protein–coupled receptor rhodopsin. Annu Rev Biochem. 2006;75:743-767. Massagué J. TGFβ signalling in context. Nat Rev Mol Cell Biol. 2012;13: 616-630. Mombaerts P. Genes and ligands for odorant, vomeronasal and taste receptors. Nat Rev Neurosci. 2004;5:263-278. Ridge KD, Abdulaev NG, Sousa M, et al. Phototransduction: Crystal clear. Trends Biochem Sci. 2003;28:479-487. Rieke F, Baylor DA. Single photon detection by rod cells of the retina. Rev Mod Phys. 1998;70:1027-1036. Ritter SL, Hall RA. Fine-tuning of GPCR activity by receptor-interacting proteins. Nat Rev Mol Cell Biol. 2009;10:819-830. Shi Y, Massague J. Mechanisms of TGF-β signaling from cell membranes to the nucleus. Cell. 2003;113:685-700. Tu Y. Quantitative modeling of bacterial chemotaxis: signal amplification and accurate adaptation. Annu Rev Biophys. 2013;42:337-359. van der Merwe PA, Dushek O. Mechanisms for T cell receptor triggering. Nat Rev Immunol. 2011;11:47-55. Venkatakrishnan AJ, Deupi X, Lebon G, et al. Molecular signatures of G-protein-coupled receptors. Nature. 2013;494:185-194.
SECTION
Cellular Adhesion and the Extracellular Matrix
VIII
This page intentionally left blank
SECTION VIII OVERVIEW T
his section covers the variety of extracellular materials that provide mechanical support for the tissues of multicellular organisms. After their divergence about 1 billion years ago, animals and plants evolved completely different macromolecules to construct their extracellular matrices. The main biopolymer in animals is the protein collagen, whereas plants use the polysaccharide cellulose. Both can make impressively strong structures, including cartilage and bone in animals and wood that supports giant trees. This section also explains the mechanisms that cells of all sorts use to adhere to each other and the objects in their environments, including the extracellular matrix. Cell surface adhesion proteins enable cells to establish intimate relationships with each other and macromolecules in the extracellular matrix. These interactions are essential for tissue integrity and intercellular communication in complex tissues, including the brain, heart, and other organs. Chapter 28 describes the cells that are found in the extracellular matrix of vertebrate animals. Fibroblasts synthesize and secrete the macromolecules that form the extracellular matrix. Fat cells store high-energy lipid molecules. Specialized phagocytic cells and immune system cells patrol the extracellular matrix of
connective tissues, seeking out and destroying foreign cells and molecules throughout the body. Chapter 29 describes the biosynthesis of the macromolecules that form the extracellular matrices of vertebrates. In connective tissue, fibroblasts secrete the protein subunits of collagen fibrils and elastic fibers as well as adhesion proteins and complex polysaccharides that reinforce the protein fibers in the extracellular matrix. The proportions of these macromolecules vary considerably. Tendons, ligaments, and some layers of the intestinal wall are composed largely of massive collagen fibrils with relatively few cells. The vitreous body of the eye is composed mostly of gelatinous polysaccharides with few fibers. The simplest type of extracellular matrix is the basal lamina, a thin layer of matrix that is secreted as rug beneath epithelial cells and a sheath around muscle cells and neurons. Remarkably, just four families of plasma membrane adhesion proteins, immunoglobulin cell adhesion molecules (IgCAMs), cadherins, integrins, and selectins, account for much of cellular adhesion. Chapter 30 introduces these adhesion proteins and explains some of their diverse structures and functions. IgCAMs and cadherins make specific interactions with
Cells Ch 28
Extracellular matrix Ch 29
Adhesion Ch 30
Primitive mesenchymal cell
Junctions Ch 31
Proliferation/differentiation
Various cell types
Connective tissues Ch 32
489
complementary proteins on the surface of partner cells. Most integrins bind to extracellular matrix molecules, but some engage adhesion proteins on other cells. Selectins interact with glycoproteins called mucins on the surfaces of other cells. Expression of a limited repertoire of adhesion protein isoforms allows cells of multicellular organisms to establish specific interactions with appropriate partner cells while avoiding inappropriate interactions. The specificity provided by these adhesion molecules is required to form epithelia during embryonic development, assemble specialized connective tissues (Chapter 32), heal wounds, and transmit the force of muscle contraction to the extracellular matrix. Chapter 30 features two examples of dynamic, selective adhesion: adhesion of platelets to each other during the repair of damage to small blood vessels and blood clotting, and adhesion of white blood cells to the endothelial cells lining blood vessels of inflamed tissues. Even unicellular organisms require molecular mechanisms to adhere to other cells and objects that they encounter in their environments. For example, unicellular algae and yeast adhere to each other during mating, and slime mold amoebas adhere to each other as they develop into fruiting bodies. Bacteria also form complex biofilms to help them survive in hostile environments. Intercellular junctions are specialized sites of adhesion between cells in some tissues (Chapter 31). Tight junctions allow a sheet of epithelial cells to create semipermeable barriers between tissue compartments such as the lumen and the wall of the intestine. Such barriers allow epithelia to concentrate materials on one side or the other. Gap junctions are composed of nonselective
490
channels that connect the cytoplasms of two cells. These channels enable action potentials to spread directly from one cell to the next, as in the heart. Gap junction channels also allow solutes that are less than 1000 Da in molecular weight to move between the coupled cells. Cadherins connect adjacent cells to the cytoskeleton at two types of adhesive junctions. The cytoplasmic domains of cadherins are anchored to actin filaments at adherens junctions and to intermediate filaments at desmosomes. The abundance, organization, and proportions of macromolecular components determine the mechanical properties of the extracellular matrix (Chapter 32). Plant cells secrete cellulose, a polymer of glucose units, plus a mixture of other polysaccharides, glycoproteins, and organic molecules to make a wall around each cell. These cell walls form materials ranging from soft cotton fibers to the wood that supports giant Sequoia trees. Connective tissues of animals also exhibit striking variety, owing to their particular mixtures of matrix molecules. Skin and blood vessels are resilient, because of numerous elastic fibers. Tendons have great tensile strength, owing to the high density of collagen fibrils. Bone is incompressible and rigid, because of its calcified collagen matrix. On the level of gross anatomy, cells and fibers form fascia, tendons, cartilage, and bones that support the organs of the body. Connective tissue also provides avenues for communication and supply within the body. Both the circulatory system and the peripheral nervous system run through connective tissue compartments of each organ. The vascular system transports phagocytic and immune system cells to sites where they are needed for defense.
CHAPTER
28
Cells of the Extracellular Matrix and Immune System A
derive from multipotential mesenchymal stem cells (see Box 41.2 and Fig. 28.1). In adults, small numbers of these inconspicuous precursor cells associate with small blood vessels but are difficult to identify by light microscopy. By electron microscopy (Fig. 28.2), mesenchymal cells resemble fibroblasts but with fewer organelles of the secretory pathway. The nature of the mesenchymal stem cells in adult tissues is still being investigated.
remarkable variety of specialized cells populate the connective tissues of animals. These cells manufacture extracellular matrix, defend against infection, and maintain energy stores in the form of lipid (Fig. 28.1). Some of these cells arise in connective tissue and remain there. These indigenous cells are specialized: Fibroblasts make the collagen, elastic fibers, and proteoglycans of the extracellular matrix; fat cells store lipids; chondrocytes secrete the matrix for cartilage; and osteoblasts manufacture the calcified matrix of bone. The remaining cells arise elsewhere, travel through blood and lymph, and enter connective tissue as needed, so they are known as immigrant cells. These visitors are part of the immune system, which defends against pathogens. This chapter introduces all these cells.
Fibroblasts Fibroblasts are the connective tissue workhorses, synthesizing and secreting most macromolecules of the extracellular matrix (Fig. 28.2). Chapter 29 considers the synthesis of these matrix molecules in detail. As appropriate for a secretory cell, mature fibroblasts have abundant rough endoplasmic reticulum and a large Golgi apparatus. They are generally spindle-shaped, with a flattened, oval nucleus, but can assume many other shapes depending on the mechanical forces in the surrounding matrix. The migratory patterns of the fibroblasts determine the patterns of collagen fibrils in tissues.
Indigenous Connective Tissue Cells Mesenchymal Stem Cells During embryogenesis the indigenous cells of connective tissues (fibroblasts, fat cells, chondrocytes, and osteoblasts) A
B Mesenchymal stem cell
Mesenchymal stem cell
White fat cell From Thomas Lentz
Differentiation and proliferation
Fibroblast
Monocyte Mast cell Neutrophil Eosinophil
White fat cell
Brown fat cell Fibroblast Chondrocyte Osteoblast
Lymphocyte
Plasma cell
Macrophage
FIGURE 28.1 CONNECTIVE TISSUE CELLS. A, Indigenous connective tissue cells all originate from a stem cell called a primitive mesenchymal cell. B, Connective tissue near a small blood vessel showing indigenous cells in pink and immigrant cells in green.
491
492
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
A
B
Golgi Collagen
t las
ob br Fi
lc ma hy nc se Me ell
FIGURE 28.2 FIBROBLASTS. A, Scanning electron micrograph of fibroblasts migrating through collagen fibrils. B, Transmission electron micrograph of a thin section of a fibroblast illustrating the abundant organelles of the secretory pathway (endoplasmic reticulum and Golgi apparatus) and extracellular collagen fibrils. A primitive mesenchymal cell is shown in the lower left. (A, Courtesy E.D. Hay, Harvard Medical School, Boston, MA. B, Courtesy D.W. Fawcett, Harvard Medical School, Boston, MA.)
In response to tissue damage, fibroblasts proliferate and migrate into the wound, where they synthesize new matrix to restore the integrity of the tissue (see Fig. 32.11). This process can get out of control if inflammatory cells secrete excessive transforming growth factor (TGF)-β (see Fig. 27.10) and other factors that stimulate fibroblasts to produce matrix molecules. Fibrosis, excess accumulation of extracellular matrix, can compromise the functions of the heart, liver, lung, or skin.
White Fat Cells White fat cells (adipocytes) of vertebrates store lipids as a readily accessible reserve of energy. They arise from mesenchymal progenitors and are distributed in connective tissues beneath the skin and in the abdominal mesentery. These round cells vary in diameter depending on the size of their single, large, lipid droplet (Fig. 28.3) containing triglycerides, neutral lipids with a fatty acid esterified to all three carbons of glycerol (see Fig. 13.2 for the structures of glycerol and fatty acids). Specialized proteins coat the cytoplasmic surface of lipid droplets. Intermediate filaments and endoplasmic reticulum separate the lipid droplet from the thin rim of cytoplasm. Fat cells respond to the metabolic needs of the body. After a meal, parasympathetic nerves stimulate fat cells to take up fatty acids and glycerol from blood and synthesize triglycerides for storage. During fasting or when the body requires energy, sympathetic nerves acting through β-adrenergic receptors (see Fig. 27.3), stimulate
adipocytes to hydrolyze fatty acids from triglycerides for release into the blood for use by other organs. If a mammal ingests excess calories, white fat cells enlarge their lipid stores and increase in number. White fat cells are long lived, with a half-life approximately 9 years, so the excess cells persist in obese individuals. White fat tissue is also an endocrine organ that secretes several cytokines and polypeptide hormones including leptin. Many factors including nutritional status, mass of fat cells, exercise and sleep influence the rate of leptin secretion. Leptin suppresses appetite and hunger by stimulating cytokine receptors on neurons in the brain. These neurons respond by secreting other polypeptide hormones that regulate appetite. Massive obesity results from congenital absence of leptin or defects in its receptor. A variety of mutations cause inherited lipodystrophies, conditions with loss of fat in one or more tissues. Examples include loss of function mutations of an enzyme required to synthesize triglycerides or a nuclear receptor that stimulates differentiation of fat cells. It is an ongoing mystery why mutations in the gene for nuclear lamins A and C (see Fig. 9.7) or a protease that processes lamin A should cause selective loss of fat from the trunk and limbs.
Brown and Beige Fat Cells Brown and beige fat cells of placental mammals derive their color from cytochromes in numerous mitochondria, which they use to generate heat in response to cold
CHAPTER 28 n Cells of the Extracellular Matrix and Immune System
A
C
493
E
IF
B
D
F
IF
Mitochondria Lipids
B and F from Thomas Lentz
Lipid
ER
FIGURE 28.3 FAT CELLS. A, Light micrograph of a section of white adipose (fat) cells stained with hematoxylin and eosin. B, Drawing of a white adipose cell. C, Transmission electron micrograph of a thin section of the edge of a lipid droplet showing the circumferential sheath of vimentin intermediate filaments (IF [see Chapter 35]). D, Interpretive drawing of a lipid droplet with its associated filaments and endoplasmic reticulum (ER). E, Light micrograph of a section of brown fat. F, Drawing of a brown fat cell. (A, C, and E, Courtesy D.W. Fawcett, Harvard Medical School, Boston, MA. B and F, Modified from T. Lentz, Yale Medical School, New Haven, CT. D, Modified from Werner Franke, University of Heidelberg, Germany.)
or (in lean rodents) excess food intake. Fat is stored in multiple, small droplets (Fig. 28.3F). Brown fat is less abundant than white fat, being concentrated in connective tissue between the scapulae in mammals. Newborn humans have more brown fat than do adults in order to generate heat during the adjustment to a new environment after birth. Hibernating animals use brown fat to raise their temperatures when emerging from hibernation. Beige fat cells are found among white fat cells and increase in numbers when a mouse is exposed to cold. A fourth type of fat cell is found in bone marrow. Both brown and beige fat cells generate heat by short-circuiting the proton gradient that is usually used to generate adenosine triphosphate (ATP) in mitochondria (see Fig. 19.5). Sympathetic nerves acting through β-adrenergic receptors (see Fig. 27.3) and protein kinase A stimulate brown fat cells to break down lipids to provide fatty acids for oxidation by mitochondria. β-Adrenergic receptors also drive expression of “uncoupling protein,” a carrier in the inner mitochondrial membrane similar in structure to the mitochondrial ATP/adenosine diphosphate (ADP) antiporter (see Fig. 15.4A). Uncoupling protein-1 dissipates the proton electrochemical gradient across the inner mitochondrial membrane (perhaps acting as a fatty acid/proton symporter), so energy is lost as heat rather than being used
to synthesize ATP. Beige fat cells also run a futile cycle of creatine phosphorylation and dephosphorylation to produce heat. Thermogenesis may be an “energy buffer” that, when defective in animals, can contribute to obesity. Therefore, stimulating thermogenesis is one goal in the treatment of obesity. Brown fat cells express thermogenic proteins constitutively, whereas β-adrenergic stimulation drives expression of these proteins in beige fat cells. The thermogenic fat cells have different precursors. Brown fat cells arise from the same mesenchymal stem cells as skeletal muscle, while beige fat cells are more closely related to white fat cells.
Origin and Development of Blood Cells The blood of vertebrates contains a variety of cells, each with a specialized function (Fig. 28.4 and Table 28.1). Red blood cells transport oxygen, platelets repair damage to blood vessels, and various types of white blood cells defend against infections. All blood cells derive ultimately from pluripotent stem cells (Fig. 28.4B; also see Box 41.2). These hematopoietic stem cells can restore the production of all blood cells in mice that have been irradiated to destroy their own blood cell precursors or after transplantation of human bone marrow. Destruction of stem cells (eg, by drugs such as
494
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
A
B
Neutrophil RBC
Common lymphoid progenitor
Platelets Monocyte
B lymphocyte Granulocyte
Lymphocyte T lymphocyte
Lymphocyte
Monocyte
Platelets
Pleuripotential stem cell
Neutrophil
Eosinophil
Basophil Platelets
Myeloid stem cell
RBC Committed stem cells
Platelets Megakaryocyte
FIGURE 28.4 BLOOD CELLS. A, Light micrograph of a dried blood smear prepared with Wright stain. B, Family tree of blood cells showing the developmental relationships of the various lineages. Looping-back arrows indicate renewal of the cell type. Forward-oriented arrows indicate differentiation and proliferation. RBC, red blood cell. (A, Courtesy J.-P. Revel, California Institute of Technology, Pasadena.)
TABLE 28.1 Blood Cells (as Seen on a Stained Smear of Blood) Type
Concentration
Features
Platelets
300,000/µL
Anucleate; 2–3 µm wide; purple granules
Erythrocytes
~5 × 106/µL
7 µm wide biconcave disks; no nucleus; pink cytoplasm
Neutrophils
~60% of total WBCs
10–12 µm wide; multilobed nucleus; many unstained granules; few azurophilic (blue) granules
Eosinophils
~2% of total WBCs
Bilobed nucleus; numerous, large, refractile, pink-stained granules; ~12 µm wide
Basophils
~0.5% of total WBCs
Lobed nucleus; large, blue-stained granules; ~10 µm wide
Lymphocytes
~30% of total WBCs
Small, round, intensely stained nucleus; some small azurophilic granules; variable amount of clear blue cytoplasm, so may be classified as either small (~7–8 µm wide), medium, or large
Monocytes
~5% of total WBCs
Up to 17 µm wide; large, indented nucleus and gray-blue cytoplasm with a few azurophilic granules
WBCs, white blood cells.
chloramphenicol) leads to aplastic anemia, a condition in which few blood cells are produced, owing to a lack of precursors. Hematopoietic stem cells do not grow well in tissue culture, because they require a special environment
provided in the bone marrow. In these niches located next to small blood vessels, endothelial cells, mesenchymal cells, and sympathetic nerves provide contacts and growth factors. Sympathetic nerves also drive the release of small numbers of hematopoietic stem cells into the blood every day following a circadian rhythm. Pluripotent hematopoietic stem cells give rise to much larger numbers of more mature stem cells that proliferate and produce blood cells. At several stages along the differentiation pathway, precursors undergo irreversible differentiation that commits them to a particular lineage. The first branch in the pathway of differentiation separates the precursors of lymphocytes from the precursors of the other blood cells called myeloid stem cells. Next, myeloid stem cells differentiate into committed stem cells that reside in different locations in bone marrow. One branch gives rise to red blood cells, megakaryocytes, and platelets. In the other branch a common committed stem cell differentiates into monocytes and the three types of granulocytes (neutrophils, eosinophils, and basophils). Mast cells also arise from myeloid stem cells, although the lineage is uncertain. Through differentiation, each mature cell type acquires unique functions. Platelets, red cells, granulocytes, and monocytes develop in bone marrow. Lymphocytes develop in both bone marrow and lymphoid tissues (thymus, spleen, and lymph nodes). Minute quantities of specific glycoprotein growth factors control the balance between self-renewal and proliferation at each stage of blood cell development, starting with pluripotential stem cells. Feedback mechanisms control production of these growth factors. For example, the oxygen level in the kidney controls the synthesis of erythropoietin, the growth factor for the
CHAPTER 28 n Cells of the Extracellular Matrix and Immune System
red blood cell series. (See Fig. 27.9 for the cytokine signaling pathway.) A dimeric transcription factor called hypoxia inducible factor (HIF)-1α/HIF-1β regulates the expression of erythropoietin. When oxygen is abundant, HIF-1α is hydroxylated on a proline residue, marking it for ubiquitination and destruction (see Fig. 23.3), turning down the expression of erythropoietin. When kidney cells lack oxygen (owing to low levels of red blood cells, poor blood circulation in the kidney, or high altitude) HIF-1α/HIF-1β accumulates, and erythropoietin is expressed and secreted to stimulate red blood cell production by bone marrow. The reciprocal relationship between oxygen and erythropoietin that is achieved by this feedback mechanism sets red blood cell production at a level required to deliver oxygen to the tissues. Many other cells use the HIF-1α/HIF-1β system to adjust gene expression to local oxygen levels. Mutations altering the growth control (see Fig. 41.12) of a stem cell can give rise to proliferative disorders, such as leukemia. In chronic myelogenous leukemia, a chromosomal rearrangement in a single white blood cell precursor creates a fusion between the genes for BCR (breakpoint cluster region) and ABL (a Src family cytoplasmic tyrosine kinase; see Fig. 25.3 and Box 27.5). The BCR-ABL protein is constitutively active and drives proliferation of a clone of immature white blood cells that crowd out and inhibit the production of other blood cells, leading to anemia and platelet deficiency. Affected individuals are prone to infection, because the immature white blood cells are ineffective phagocytes. Fortunately, a small-molecule inhibitor of the kinase activity of BCR-ABL ( imatinib mesylate [GLEEVEC]) suppresses this clone in many patients. Mutations rendering the JAK2 (just another kinase 2) kinase (see Fig. 24.6) constitutively active drive proliferation in other leukemias. Uncontrolled proliferation of a clone of red blood cell precursors causes a similar condition, characterized by excess red cells, called polycythemia vera.
Cells Confined to the Blood Erythrocytes (Red Blood Cells) Red blood cells (RBCs) (Fig. 28.4; also see Fig. 13.8) contain more than 300 mg/mL of hemoglobin to carry oxygen from the lungs to tissues and carbon dioxide from tissues to the lungs. As they proliferate and differentiate in bone marrow, RBC precursors accumulate hemoglobin and shed all their organelles. A resilient, spectrin–actin membrane cytoskeleton (see Fig. 13.11) maintains the biconcave shape even after the cell is heavily distorted each time it squeezes through a small capillary. The elasticity of the membrane skeleton allows it to regain its shape. Human bone marrow produces about 100 billion RBCs each day. After circulating in blood for 120 days, erythrocytes abruptly become senescent, and
495
phagocytes in the spleen, liver, and bone marrow remove them from the blood. The biochemical basis of this precise cellular aging and clearance process is still being investigated. Mutations in many different genes cause RBC diseases. In hereditary spherocytosis (and other hemolytic anemias), the membrane cytoskeleton loses its resiliency as a result of mutations, causing deficiencies or molecular defects of spectrin or other component proteins. These defective cells are easily damaged and eventually become smaller and rounder than normal. Many different mutations of the globin genes may compromise the synthesis of a particular globin gene or decrease the stability or oxygen-carrying capacity of hemoglobin. In sickle cell disease, hemoglobin S is prone to assemble into tubular polymers that distort the cell and clog up the circulation.
Platelets Platelets are small cellular fragments without a nucleus that contribute to blood clotting and repair of minor defects in the sheet of endothelial cells that lines blood vessels. A long, coiled microtubule presses out against the plasma membrane, like a spring, to maintain the platelet’s disk shape (Fig. 28.5). The most prominent organelles are two types of membrane-bound granules. Dense granules contain ADP and serotonin. Alpha granules contain stores of adhesive glycoproteins including fibrinogen, fibronectin (see Fig. 29.14), and thrombospondin as well as the potent protein hormone called platelet-derived growth factor. Platelet-derived growth factor has a role in wound healing (see Fig. 32.11), but can contribute to atherosclerosis by stimulating the abnormal proliferation of smooth muscle cells in the walls of damaged arteries. Another secreted cytokine kills malaria parasites inside RBCs. Platelets containing a full complement of organelles bud from the tips of protrusions on the surface precursor cells—giant polyploid megakaryocytes in the bone marrow. Thrombopoietin, a protein hormone related to erythropoietin, controls platelet production by stimulating the thrombopoietin cytokine receptor. Liver and kidney cells secrete thrombopoietin at a constant rate. Receptors on circulating platelets bind part of the thrombopoietin. Consequently, the blood concentration of thrombopoietin available to stimulate megakaryocyte maturation and platelet formation is inversely related to the total number of platelets. This feedback loop stimulates platelet production if the platelet supply is low. Like RBCs, platelets are confined to the blood where they circulate for about seven days. Two pools of platelets freely exchange with each other: About two-thirds of the total platelets circulate, whereas one-third of the platelets are stored in the blood vessels of the spleen. The stored pool may increase when the spleen is enlarged, decreasing the platelet count in the blood.
496
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
A
B
C
D
MT
E
Actin
Granules
MT
F
G
H
I
Platelet
Damage exposes basal lamina
Activated platelets secrete ADP
Activated platelets aggregate over defect
Endothelium Basal lamina
Platelet binds
ADP ADP
MT FIGURE 28.5 PLATELETS AND THEIR ROLE IN HEMOSTASIS. A–B, Transmission electron micrographs of thin sections of platelets. C, Interpretive drawing showing the circumferential band of microtubules (MT), actin filaments in the cortex, and granules in the cytoplasm. D, Role of platelets in blood clot retraction. Filopodia grasp strands of fibrin in a blood clot (upper panel) and pull them together (lower panel). E, Electron micrograph of a thin section showing platelets adhering to the basal lamina through a small defect in the endothelium and to a second platelet in the lumen. F–I, Stages in the repair of a defect in the endothelial lining of a blood vessel. F, Circulating platelets do not bind to normal endothelial cells. G, Damage to the endothelium exposes the basal lamina, and a platelet binds to the collagen. H, Collagen activates the platelet to secrete adenosine diphosphate (ADP), which activates passing platelets. I, Activated platelets bind together, covering the defect in the endothelium. (A–B, Courtesy O. Behnke, University of Copenhagen, Denmark.)
Platelets control bleeding in three ways. First, they adhere and change shape to cover damaged vascular surfaces. Second, platelets stimulate blood clotting. A surface protein on activated platelets stimulates a cascade of proteolytic reactions culminating in the cleavage of plasma fibrinogen to form fibrin, which polymerizes to clot blood. The fibrin gel stops the flow of blood from damaged blood vessels. Third, platelets help close holes in damaged blood vessels by pulling on fibrin strands and contracting the clot. When the mild trauma of daily existence produces defects in the endothelial cells lining blood vessels, platelets repair the damage. Platelets bind to von Willebrand factor and collagen in the basal lamina when it is exposed by damage to the endothelium. This triggers one of the best-understood examples of regulated cellular adhesion (for more detail, see Fig. 30.14). Activated platelets aggregate, extend actin-containing filopodia, and secrete the contents of their granules. The secreted ADP sets up a positive feedback loop, activating more platelets that form a cluster to fill the defect in the endothelium. Other platelets are consumed by phagocytic cells in the liver and spleen. Patients with defective platelets or reduced circulating platelets (a complication of bone marrow disease and cancer chemotherapy) bruise easily, owing to unrepaired damage in small blood vessels, and may even bleed spontaneously. Conversely, hyperactive platelets may initiate pathological clots in the blood vessels of the heart, causing heart attacks or thrombosis in the veins of the legs.
Cellular Basis of Innate Immunity All multicellular animals use two forms of innate immunity to defend themselves against infection by microorganisms. Phagocytic cells track down, ingest, and kill bacteria and fungi (see Fig. 22.3). A second line of defense is secretion of cytokines and small antimicrobial proteins by white blood cells and epithelial cells of the skin and intestine. The following sections describe the main mammalian phagocytes, macrophages and neutrophils (Fig. 28.4). They originate in bone marrow from a common committed stem cell (see Fig. 28.4B) and acquire unique functions as they differentiate. Innate immune cells detect the presence of microorganisms using Toll-like receptors (TLRs) (Fig. 28.6). The Toll gene was discovered in Drosophila encoding a receptor that was first linked to dorsal–ventral polarity in early development and later shown to be required for resistance to fungal infections. TLRs are called “pattern recognition receptors,” because they bind generic, repeating structures of polymeric macromolecules rather than recognizing fine molecular details like antibodies. These “pathogen-associated molecular patterns” or PAMPs are found on components essential for the functions of the microorganisms such as viral doublestranded RNA, bacterial flagellin, lipopolysaccharide from the outer membrane of gram-negative bacteria, and zymosan from the cell walls of fungi. Mammals have about a dozen TLRs in three classes. All TLRs have cytoplasmic TIR (Toll interleukin-1 receptor)
CHAPTER 28 n Cells of the Extracellular Matrix and Immune System
Ligand-binding domain with leucine-rich repeats
497
TLR1 and TLR2 receptors dimerized by triacetylated lipopeptide
Receptor TIR domain of TLR1
Adaptor proteins
Adaptor proteins bound to TIR domain
Kinases NF-κB activated then imported into nucleus
NF-κB
TNF gene expressed
TLR3 bound to double-stranded RNA in an endosome
FIGURE 28.6 TOLL-LIKE RECEPTORS SIGNALING FROM THE PLASMA MEMBRANE AND ENDOSOMES. Binding of a triacetylated bacterial lipopeptide stabilizes a heterodimer of Toll-like receptor (TLR)-1 and TLR2 in the plasma membrane shown as ribbon diagrams with space-filling surfaces. Dimers of TLR3 in the membranes of endosomes bind double-stranded RNAs released from viruses. Ligand binding to receptor dimers aligns their cytoplasmic Toll-interleukin receptor (TIR) domains and allows binding of adapter proteins that initiate a signal transmitted to kinases, which activate cytoplasmic transcription factors including nuclear factor κB (NF-κB). NF-κB moves to the nucleus and stimulates expression of tumor necrosis factor (TNF) and other inflammatory mediators. (Molecular structures are based on Protein Data Bank [www.rcsb .org] file 2Z2X. For reference, see Jin MS, Kim SE, Heo JY, et al. Crystal structure of the TLR1-TLR2 heterodimer induced by binding of a triacylated lipopeptide. Cell. 2007;130:1071–1082; and Liu L, Botos I, Wang Y, et al. Structural basis of toll-like receptor 3 signaling with doublestranded RNA. Science. 2008;320:379–381.)
domains, but differ in other respects. Classic TLRs have ligand binding domains consisting of leucine rich repeats (Fig. 28.6). Interleukin (IL)-1 receptors have extracellular immunoglobulin (Ig) domains. PAMP binding to a dimeric TLR on the plasma membrane of white blood cells or antigen-processing dendritic cells stimulates the secretion of inflammatory mediators such as tumor necrosis factor (TNF) and IL-1 and IL-6. TNF and ILs then alert distant cells to respond to the infection. The signaling pathway from TLRs to TNF (Fig. 28.6) involves cytoplasmic adapter proteins and kinases that activate transcription factors including nuclear factor κB (NF-κB) (see Fig. 10.22C) and expression of some long noncoding RNAs. Plasma membrane TLRs also activate mitogen-activated protein (MAP) kinase pathways (see Fig. 27.5). The response also includes secretion of approximately 40 small proteins called chemokines that attract motile phagocytic cells. Chemokines with many unrelated names (IL-8, RANTES [regulated upon activation, normal T-cell expressed and secreted], eotaxin, monocyte chemotactic protein [MCP]-1, etc.) have similar structures and bind to a family of 14 different chemokine receptors expressed selectively by lymphocytes, monocytes, and granulocytes. These seven-helix receptors are coupled to trimeric G-proteins (see Fig. 24.3) that mediate chemotaxis (see Fig. 38.10) toward the source
of the chemokine. Secretion of antimicrobial peptides such as defensins and cathelicidins not only kill pathogens by interacting with their membranes but also attract and activate cells of both the innate and the adaptive immune systems. Cytoplasmic receptors with TIR domains bind nucleic acids. TLRs 3, 7, 8, and 9 are located in endosomes, where they bind double-stranded RNA released from viruses and single stranded RNAs from viruses and bacteria. Signaling pathways using some of the same components as the plasma membrane TLRs stimulate the expression and secretion of TNF and interferon (IFN)-α. The cytoplasmic system uses a helicase called RIG-I and other proteins to recognize foreign RNAs and to respond via NF-κB to produce IFN-β. Other pattern recognition receptors reside in the cytoplasm of innate immune cells and other cells. For example, nucleotide-binding oligomerization domain (NOD)-like receptor proteins activate a kinase that stimulates both the transcription factor NF-κB and the MAP kinase pathways. Other NOD-like receptor proteins activate the proteolytic enzyme caspase-1 (see Fig. 46.11), which processes inflammatory cytokine precursors for secretion. In addition to phagocytes, mammals have special lymphocytes called natural killer cells that express a variety of receptors to detect and kill cells infected with
498
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
a virus. Like other innate immune cells, natural killer cells also secrete a variety of cytokines that influence immune responses and inflammation. These elements of the innate immune system are programmed genetically, so they respond without prior exposure to the pathogen. Despite their generic nature, these innate responses work remarkably well, defending all metazoans against infection.
Neutrophils Neutrophils, also known as polymorphonuclear leukocytes or “polys,” are the main phagocytes circulating in blood, ready to enter connective tissues at sites of infection. They are distinguished by a multilobed nucleus and two types of granules (Fig. 28.7). The more abundant specific granules contain lysozyme (an enzyme that digests bacterial cell walls) and alkaline phosphatase. These granules do not stain with either the basic or acidic dyes used for blood smears, so these cells are called neutrophils. Azurophilic granules are true lysosomes containing hydrolytic enzymes bound to acidic
A. Neutrophil
proteoglycans. Neutrophils have few mitochondria, so they produce ATP in poorly oxygenated wounds by breaking down stores of glycogen by glycolysis. They are among the most motile cells in the body. Human bone marrow produces about 100 billion (80 g) neutrophils each day. In response to infection or injury, a circulating factor releases neutrophils from the bone marrow into the blood. Neutrophils spend about 10 hours in blood, spending part of the time adherent to endothelial cells, chiefly in the lung. Exercise and epinephrine release adherent neutrophils into the circulating pool; smoking increases the adherent pool. Neutrophils leave the blood by receptor-mediated attachment to endothelial cells and then crawling between them into the connective tissue (see Fig. 30.13), where they perish after a day or two of phagocytosis. Neutrophils are humans’ first line of defense against bacterial infection, and they are highly specialized for finding and destroying bacteria. Provided that the concentration of neutrophils is high enough (approximately 10 million cells per milliliter), these motile phagocytes
D. Monocyte
B. Eosinophil Ilustrations A-E from Thomas Lentz
E. Macrophage
C. Basophil
FIGURE 28.7 WHITE BLOOD CELLS. Transmission electron micrographs of thin sections of each cell and interpretive drawings with lysosomes shown in brown. A, A neutrophil showing the multilobed nucleus (the connections between lobes are in other sections) and the two classes of granules. B, An eosinophil showing the bilobed nucleus and the large, specific granules containing a darkly stained crystalloid. C, Basophil with large specific granules colored blue. D, Blood monocyte. E, Macrophage grown in tissue culture. (Micrographs courtesy D.W. Fawcett, Harvard Medical School, Boston, MA. Drawings modified from T. Lentz, Yale Medical School, New Haven, CT.)
CHAPTER 28 n Cells of the Extracellular Matrix and Immune System
can find and destroy bacteria faster than the invaders can reproduce. Bacterial products, especially N-formylated peptides, attract neutrophils by binding plasma membrane receptors and stimulating locomotion, similar to chemotaxis by other cells (see Fig. 38.10). Neutrophils bind and ingest bacteria by phagocytosis (see Fig. 22.3). Both types of granules fuse with phagosomes, delivering antibacterial proteins and proteolytic enzymes that kill the ingested bacteria. Some granules fuse with the plasma membrane, releasing antibacterial proteins outside the cell. Phagosome membranes produce millimolar concentrations of superoxide (O2−) radicals and other reactive oxygen species that help disperse the granule enzymes and contribute to killing bacteria. These toxic oxygen species may also cause collateral damage to the neutrophil. Genetic defects in the enzymes that produce superoxide cause chronic granulomatous disease, a serious human disease, because neutrophils cannot kill ingested bacteria and fungi.
Eosinophils Eosinophils are members of the granulocyte lineage, present in low numbers in the blood. They are identified in blood smears as cells with a bilobed nucleus and large specific granules that stain brightly red with eosin (Fig. 28.7B). Specific granules contain a cationic protein crystalloid, a ribonuclease and peroxidase, in addition to a crystalloid of a basic protein. Like neutrophils, eosinophils transit the blood for hours on their way to connective tissues, especially in the gastrointestinal tract, where they survive for a few days. Chemotactic factors generated by the complement system, basophils, some tumors, parasites, and bacteria all attract eosinophils to tissues. Many of the same factors attract other leukocytes, but particular chemokines are specialized for eosinophils. Eosinophils accumulate in blood and in tissues infected with parasites, but experts do not agree on whether eosinophils kill bacteria or parasites. Activated eosinophils contribute to inflammation in some allergic disorders such as asthma but they also secrete factors that promote immune responses by lymphocytes. Macrophages Macrophages are a diverse group of professional phagocytes with many common features but two different origins. All have a receptor tyrosine kinase (colonystimulating factor receptor) that drives their differentiation into phagocytes. Macrophages in brain, liver, lung and some other tissues arise from cells in the embryonic yolk sac, while adult tissue macrophages develop from monocytes that develop in bone marrow and circulate in the blood. Monocytes are the largest blood cells with an indented nucleus and a small number of azurophilic granules (Figs. 28.4A and 28.7D). After about 3 days in the blood, monocytes enter tissues and differentiate into
499
macrophages under the influence of local growth factors, including lymphokines secreted by lymphocytes (Fig. 28.9). Macrophages enlarge and amplify their machinery for locomotion, phagocytosis, and killing microorganisms and tumor cells. Tissue macrophages may divide and survive for months. Local growth factors in bone stimulate monocytes to fuse and differentiate into multinucleated osteoclasts that degrade bone matrix during bone remodeling (see Fig. 32.6). Macrophages generally follow neutrophils to wounds or infections to clean up debris and foreign material. Plasma membrane receptors for antibodies allow macrophages to recognize foreign matter marked with antibodies and to facilitate its ingestion. Primary lysosomes fuse with phagosomes to degrade the contents. Eventually, the cytoplasm fills with residual bodies containing the remains of ingested material (see Fig. 23.4). When confronted with large foreign bodies, macrophages can fuse together to form giant cells. Giant multinucleated microphages will even try to ingest a Petri dish if it is coated with antibody. Macrophages and their cousins, called dendritic cells (see below), participate in the immune response by degrading ingested protein antigens and presenting antigen fragments on their surface bound to major histocompatibility complex (MHC) class II proteins (Fig. 28.9). This complex activates helper T lymphocytes carrying the appropriate T-cell receptors (see Fig. 27.8). Activated T cells proliferate and secrete growth factors that stimulate B lymphocytes to produce antibodies. Engagement of TLRs stimulates macrophages, which secrete dozens of factors involved with host defense, inflammation, and normal development. Among these, IL-1, TGF-α, TGF-β, and platelet-derived growth factor stimulate the proliferation and differentiation of the cells required to heal wounds (see Fig. 32.11). Chemokines attract cells of the immune system to sites of inflammation.
Mast Cells and Basophils Basophils (Fig. 28.6C) and mast cells (Fig. 28.8) both have histamine-containing granules that are secreted when antigens bind cell-surface IgE molecules. The large, abundant granules contain, by mass, 30% heparin– basic protein complex, 10% histamine, and 35% basic proteins, including proteases. Plasma membrane receptors bind a random selection of IgE antibodies made by the immune system in response to exposure to antigens. Binding of the corresponding antigen to IgE on the surface of a basophil or mast cell activates a signaling pathway with a Src family tyrosine kinase similar to those in T lymphocytes (see Fig. 27.8). A cytoplasmic Ca2+ pulse triggers fusion of granules with the plasma membrane (see Fig. 21.16) and other pathways stimulate production of cytokines and lipid second messengers. Mechanical trauma, radiant energy (heat, X-rays),
500
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
A
B
C
connective tissue blood vessels (Fig. 28.8) and beneath epithelial cells lining airways and the gastrointestinal tract. Basophils circulate in small numbers in the blood. They look much like neutrophils, but have a bilobed nucleus and large, basophilic, specific granules containing serotonin and all the histamine in the blood (Fig. 28.7C). Basophils are weak phagocytes. Humans have both circulating basophils and tissue mast cells, but this is not universal. Mice, for example, have mast cells but no basophils, and turtles have basophils but no mast cells.
Cellular Basis of Adaptive Immunity
Granule
C from Thomas Lentz
FIGURE 28.8 MAST CELLS. A, Light micrograph of loose connective tissue, stained with toluidine blue, illustrating mast cells scattered along a blood vessel (drawn in to enhance the contrast). Large mast cell granules stain intensely purple with basic thiazine dyes such as toluidine blue. B, Transmission electron micrograph of a thin section of a mast cell. C, Drawing of a mast cell. (A–B, Courtesy D.W. Fawcett, Harvard Medical School, Boston, MA. C, Modified from T. Lentz, Yale Medical School, New Haven, CT.)
bacterial toxins, and venoms are less-specific stimuli, but can also trigger secretion. Outside the cell, the carrier proteins release heparin and histamine. On the positive side, secretion of histamine and other granule contents rapidly attracts other cells to fight infections as part of “immediate hypersensitivity” reactions. On the negative side, histamine binds to cellular receptors, causing plasma to leak from blood vessels, contraction of smooth muscle, and itching sensations. This results in congestion and constriction of the respiratory tract in allergic reactions and swelling of the skin after an insect bite. Secreted fibrinolysin and heparin inhibit blood clotting. Stimulated mast cells also secrete TNF-α and eicosanoids, contributing to the activation of other inflammatory cells in chronic conditions including asthma and arthritis. Basophils and mast cells both arise from bone marrow myeloid stem cells (Fig. 28.4B), with mast cells differentiating on a different, and still uncertain, pathway from that of basophils and other granulocytes. After arising in the bone marrow immature mast cells move quickly to other tissues where they mature and distribute along
Starting with cartilaginous fish, vertebrates developed a sophisticated adaptive immune system. The response is slower than innate immunity, because it depends on the selection and multiplication of lymphocytes that produce soluble antibodies or cell surface receptors precisely targeted to foreign molecules. This response depends on rearrangement and mutation of genes to produce highly selective antibodies and receptor proteins. Although this adaptive response takes about a week to mobilize, it produces specialized lymphocytes that survive for years, providing the host with a faster adaptive response when exposed to the pathogen a second time. In response to infection, lymphocytes of the immune systems of vertebrates produce two kinds of adaptive responses: humoral (in the body fluids) and cellular. B lymphocytes produce the humoral response by secreting antibodies (immunoglobulins), soluble proteins that diffuse in the blood and tissue fluids. Many types of T lymphocytes mediate the cellular arm of the adaptive immune response. Of these, cytotoxic T lymphocytes (killer T cells) destroy cells infected with viruses, whereas helper T cells regulate other lymphocytes. These responses protect against infection but fail in AIDS when the HIV kills helper T cells. A blood smear reveals lymphocytes of various sizes and shapes (Fig. 28.4), but not their remarkable heterogeneity at the molecular level (Fig. 28.9). Antibodies produced by B cells provide a chemical defense against viruses, bacteria, fungi, and toxins. Antibodies, or immunoglobulins, are an incredibly diverse family of proteins, each with a binding site that accommodates one of millions of different ligands termed antigens. Antigens include proteins, polysaccharides, nucleic acids, lipids, and small organic molecules produced biologically or chemically. Antibody binding can mark an antigen for phagocytosis or neutralize its toxicity. The huge repertoire of antigen-binding sites present in the collection of antibodies that circulate in a single individual arises through rearrangement and somatic mutations of Ig genes (Fig. 28.10). This remarkable
CHAPTER 28 n Cells of the Extracellular Matrix and Immune System
A. Proliferation and genetic recombination
B-cell lineage
CD4
Ag
Soluble antigen
Ig • Ag
Ag binds a subset of cells
C Ab
MHC II
CD8
TCR
MHC II • Ag CD4 TCR
Clone of cells secretes antibody to specifec antigen
Proliferation
CD4 TCR
Clone of helper T cells secretes growth factors
TCR
MHC class I displays digested Ag on virus-infected somatic cells MHC I
MHC I • Ag
Subset of T cells binds macrophage
Growth factors stimulate B cells
Ig'
Killer T-cell lineage
MHC class II displays digested Ag on macrophages and other antigenpresenting cells
B
Proliferation
Lymphoid stem cell
Helper T-cell lineage
Ig
501
CD8 TCR
Subset of T-cells binds infected cell Proliferation
Clone of killer T cells targets virus-infected cell CD8 TCR
FIGURE 28.9 THE IMMUNE RESPONSE BY THREE CLASSES OF LYMPHOCYTES THROUGH THREE PARALLEL STEPS. A, Genetic recombination produces populations of cells with a wide variety of antigen specificities provided by cell-surface immunoglobulins (Ig) or T-cell receptors (TCR). B, The binding of specific antigens (Ag) to surface immunoglobulins or TCRs selects a subset of the cells for proliferation. MHC, major histocompatibility complex. C, Proliferation of clones of selected cells yields many cells specialized to produce antibody (Ab) to soluble antigens, secretion of growth factors by helper T cells in response to ingested and degraded antigens, or killing of virus-infected cells identifiable by the viral peptides on their surface. The helper and killer T cells use a common set of TCRs and are guided to the appropriate target cells by the CD4 and CD8 accessory molecules. See Figs. 27.8 and 46.9 for details on T-lymphocyte activation and selection, respectively.
process was exploited during evolution specifically for the use of the immune system. Immunoglobulins of most mammals are composed of four polypeptide chains—two identical heavy chains and two identical light chains—each encoded by different genes (Fig. 28.10). Light chains and heavy chains both contribute to the antigen-binding site. Camels and llamas are an exception; their antibodies consist of a single polypeptide. Immunoglobulin genes exist in segments aligned along a vertebrate chromosome. Several of these gene segments must be combined in the proper order to make a functional antibody gene. Some gene segments encode the framework of the antibody protein, which is essentially identical within each antibody class. Other gene segments, present in many variations, encode the part of the polypeptide chain that forms the antigenbinding site. During maturation of a particular B cell, recombination enzymes (RAG1 and RAG2) assemble immunoglobulin gene segments into one unique full-length gene
for a heavy chain and one for a light chain. As a result of random gene arrangements, each B cell assembles and expresses novel immunoglobulin genes. The process is precise in that the right number of segments is always chosen to make a heavy chain or a light chain, but it is also random in that any one of the variable segments may be chosen. The resulting antibody contains two identical but unique antigen-binding sites. The gene segments can be assembled in many different combinations, and most heavy chains can assemble with most light chains. The diversity arising from the combinatorial process is expanded further in two ways. First, the recombination process inserts a variable number of nucleotides between the gene segments. Second, pre–B cells use enzymes to mutate codons for amino acids in the antigen-binding site, creating unique variations in the antigen-binding specificity in different cells. In principle, approximately 3000 different light chains and 60,000 heavy chains can combine to produce approximately 100 million different antibodies—even without taking mutations into account.
502
SECTION VIII n Cellular Adhesion and the Extracellular Matrix A. Germline DNA V segments
5’
D segments
C segments
3’
Rearrangement of D and J
B. Pro–B-cell DNA V
5’
J segments
Variable antigen-binding sites
DJ
C. Pre–B-cell DNA V
5’
D. B-cell DNA 5’
C
Rearrangement of V C
DJ
3’
Somatic mutations of VDJ sequences and rearrangement of C V
DJ
Constant region IgG antibody
3’
E. Messenger RNA 5’ V DJ C
C
3’
Gene expression with alternative splicing of C domains 3’
FIGURE 28.10 ASSEMBLY OF IMMUNOGLOBULIN GENES BY REARRANGEMENT OF GENE SEGMENTS. A, Region of germline DNA with multiple, tandem V, D, J, and C segments for assembling an immunoglobulin (Ig) heavy chain. B, Same region after rearrangement of D and J segments in a Pro–B cell. C, After rearrangement of V in a Pre–B cell. D, B cell. E, Immunoglobulin messenger RNA produced by a B cell. (Modified from Chiorazzi N, Rai KR, Ferrarini M. Chronic lymphocytic leukemia. N Engl J Med. 2005;352:804–815.)
Accordingly, it is possible experimentally to find in a mouse an antibody that is specific for almost any naturally occurring or synthetic chemical. Infection by a pathogen results in the selection of antibodies that bind to the pathogen, but not to any of the individual’s own molecules. This response comes from activation and proliferation of preexisting B cells making antibodies that bind to molecules of the pathogen. Activation requires a chance encounter of particular B cells with T cells presenting antigenic molecules from the pathogen (see later) and stimulates the cell to mature into a factory for secreting antibodies. Alternate splicing of messenger RNA (mRNA [see Fig. 11.6]) selects domains required to direct the same antibody to either the plasma membrane or the secretory pathway. Antigen binding to the surface immunoglobin displayed on a particular B cell stimulates that cell to go through multiple rounds of cell division, giving rise to a clone of cells, all making the same antibody. In an animal, this process of clonal expansion amplifies multiple clones of B cells. All the clones produce antibodies that react with the antigen, but most clones make novel antibodies to that antigen. Therefore, an animal makes many different antibodies to the same antigen. By isolating single B cells and expanding their numbers in the laboratory, one may obtain enough identical cells to make useful amounts of the same exact antibody, called a monoclonal antibody, because it arose from the proliferation of a single cell into a clone of cells.
The B-cell response produces two types of mature cells. Plasma cells are each highly specialized to secrete large amounts of one specific antibody. Long-lived memory cells display a specific antibody on their surface, ready to mount an amplified response on subsequent exposure to the same antigen. This immunologic memory explains why exposure to a particular pathogen or vaccination against a pathogen results in protection, in the form of antibodies, for many years. Specialized B lymphocytes and plasma cells secrete different antibody isoforms or isotypes. Formation of immunoglobulins with the various isotypes requires further recombination events to join the variable region containing the antigen-binding site to the isotype constant domain. IgG isotypes, produced in lymph nodes and spleen, circulate in blood and tissue fluids. IgA isotypes, secreted by B cells in lymphoid nodules of the respiratory and gastrointestinal tracts and by mammary glands, are taken up by epithelial cells and resecreted into the lumens of these organs (transcytosis [see Fig. 22.6]). IgE isotypes bind to receptors on the surface of mast cells and basophils (see earlier discussion). T lymphocytes provide cellular responses to pathogens. Cytotoxic T cells kill tumor cells and virus-infected cells. Helper T cells stimulate antibody production by B cells. The specificity of these responses is provided by variable cell surface receptors called T-cell receptors (see Fig. 27.8). A set of segmented genes analogous to immunoglobulin genes encodes T-cell receptors. In
CHAPTER 28 n Cells of the Extracellular Matrix and Immune System
contrast to antibodies, T-cell receptors do not bind free antigens but rather recognize peptide antigens bound to proteins called MHC antigens on the surface of a target cell (see Fig. 27.8). These highly variable MHC proteins are responsible for the rejection of tissue grafts from nonidentical individuals. The two types of MHC proteins—class I and class II—acquire their antigenic peptides differently. All somatic cells produce class I MHC proteins. In cells that have been infected by a virus, cytoplasmic proteasomes degrade some viral proteins to peptides (see Chapter 23), which ABC transporters (TAP1, TAP2) move from the cytoplasm (see Fig. 14.9) into the endoplasmic reticulum (ER). In the lumen of the ER, peptides insert into the binding site of compatible class I molecules and the complex moves to the plasma membrane. In contrast, macrophages and other antigen-presenting cells, such as dendritic cells, ingest foreign matter and degrade it in endosomes and lysosomes. Such peptide fragments bind to class II proteins in endosomes and thence move to the cell surface of these antigen-presenting cells. T lymphocytes patrol the body, inspecting the surfaces of other cells. A chance encounter with a cell displaying a peptide-MHC complex complementary to its T-cell receptor stimulates the T cell (see Fig. 27.8). The response is proliferation and expansion of a clone of identical T cells. Accessory membrane proteins CD4 and CD8 on the T-cell surface cooperate with T-cell receptors to direct the two types of T cells to target cells with the appropriate MHC proteins. T-cell receptors provide antigen specificity. Immature T-cells express both CD4 and CD8 but lose one of them as they mature into cytotoxic (CD8+) or helper (CD4+) T cells. CD8-positive cytotoxic T cells are specialized to kill cells infected with viruses. The presence of virus inside is revealed by MHC class I proteins displaying vital peptides on the surface of the infected cell. CD8 binds to a constant region of MHC class I proteins carrying viral peptides. Cytotoxic T cells use three weapons to kill the target cell: First, T cells carry a ligand for the Fas receptor on the target, which stimulates apoptosis of the target cell (see Fig. 46.18). Second, activated T cells bind to the target cell, forming an immunological synapse (see Fig. 27.8) into which the T cell secretes perforin, a protein that inserts into the plasma membrane of the target cell, forming large (10 nm) pores that leak cytoplasmic contents and ultimately lyse the cell. Third, T cells secrete toxic enzymes into the synapse that enter target cells through the plasma membrane pores. CD4 on helper T cells binds a constant part of the MHC class II protein and targets helper T cells to cells presenting ingested antigens. The progeny of stimulated helper T cells secrete growth factors (lymphokines or interleukins) in the vicinity of B cells with the foreign antigen bound to immunoglobulins on their surface.
503
Helper T cells are required for B cells to make antibodies against most antigens. This explains how HIV causes AIDS. The virus uses CD4 as a receptor to infect and eventually kill helper T cells. Loss of helper T cells severely limits the capacity of B cells and cytotoxic T cells (which also require T-cell help) to mount antibody and cellular responses to microorganisms. Infections that the immune system normally dispatches with ease then become life-threatening. Genetic defects cause a wide variety of immunodeficiency diseases. For example, defects in Bruton tyrosine kinase result in failure to produce B cells. Remarkably, humans who lack function of the enzyme adenosine deaminase have no B cells or T cells, but are otherwise normal. Deficiencies of many specialized lymphocyte proteins (cytokine receptors, interleukin receptors, Lck tyrosine kinase, ZAP-70 [zeta-associated protein of 70 kD] tyrosine kinase, RAG1 or RAG2, TAP1 or TAP2) also lead to immunodeficiencies. ACKNOWLEDGMENTS We thank Matthew Rodeheffer for suggestions on revisions to this chapter. SELECTED READINGS Berry R, Rodeheffer MS, Rosen CJ, et al. Adipose tissue residing progenitors. Curr Mol Biol Rep. 2015;1:101-109. Beutler B. Inferences, questions and possibilities in Toll-like receptor signalling. Nature. 2004;430:257-263. Bianco P. “Mesenchymal” stem cells. Annu Rev Cell Dev Biol. 2014;30: 677-704. Boes M, Ploegh HL. Translating cell biology in vitro to immunity in vivo. Nature. 2004;430:264-271. Busiello RA, Savarese S, Lombardi A. Mitochondrial uncoupling proteins and energy metabolism. Front Physiol. 2015;6:36. Call ME, Wucherpfennig KW. The T cell receptor: Critical role of the membrane environment in receptor assembly and function. Annu Rev Immunol. 2005;23:101-125. Chakraborty AK, Weiss A. Insights into the initiation of TCR signaling. Nat Immunol. 2014;15:798-807. Eaves CJ. Hematopoietic stem cells: concepts, definitions, and the new reality. Blood. 2015;125:2605-2613. Gay NJ, Symmons MF, Gangloff M, Bryant CE. Assembly and localization of Toll-like receptor signalling complexes. Nat Rev Immunol. 2014;14:546-558. Grinnell F. Fibroblast biology in three-dimensional collagen matrices. Trends Cell Biol. 2003;13:264-269. Hargreaves DC, Medzhitov R. Innate sensors of microbial infection. J Clin Immunol. 2005;25:503-510. Hartwig J, Italiano J Jr. The birth of the platelet. J Thromb Haemost. 2003;1:1580-1586. Ho MS, Medcalf RL, Livesey SA, et al. The dynamics of adult haematopoiesis in the bone and bone marrow environment. Br J Haematol. 2015;170:472-486. Howell WM. HLA and disease: guilt by association. Int J Immunogenet. 2014;41:1-12. Hubbi ME, Semenza GL. Regulation of cell proliferation by hypoxiainducible factors. Am J Physiol Cell Physiol. 2015;309:C775-C782.
504
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
Lacy P, Rosenberg HF, Walsh GM. Eosinophil overview: structure, biological properties, and key functions. Methods Mol Biol. 2014; 1178:1-12. Mitchell WB, Bussel JB. Thrombopoietin receptor agonists: a critical review. Semin Hematol. 2015;52:46-52. Morrison SJ, Scadden DT. The bone marrow niche for haematopoietic stem cells. Nature. 2014;505:327-334. Murray PJ, Wynn TA. Protective and pathogenic functions of macrophage subsets. Nat Rev Immunol. 2011;11:723-737. Perez-Lopez A, Behnsen J, Nuccio SP, et al. Mucosal immunity to pathogenic intestinal bacteria. Nat Rev Immunol. 2016;16:135-148. Raje N, Dinakar C. Overview of Immunodeficiency Disorders. Immunol Allergy Clin North Am. 2015;35:599-623. Rockey DC, Bell PD, Hill JA. Fibrosis—a common pathway to organ injury and failure. N Engl J Med. 2015;372:1138-1149. Rosen ED, Spiegelman BM. What we talk about when we talk about fat. Cell. 2014;156:20-44. Rot A, von Adrian UH. Chemokines in innate and adaptive host defense: Basic chemokinese grammar for immune cells. Annu Rev Immunol. 2004;22:891-928.
Schmetzer O, Valentin P, Church MK, Maurer M, Siebenhaar F. Murine and human mast cell progenitors. Eur J Pharmacol. 2016;778: 2-10. Schroder K, Tschopp J. The inflammasomes. Cell. 2010;140:821-832. Segal AW. How neutrophils kill microbes. Annu Rev Immunol. 2005; 23:197-223. Trombetta ES, Mellman I. Cell biology of antigen processing in vitro and in vivo. Annu Rev Immunol. 2005;23:975-1028. Varol C, Mildner A, Jung S. Macrophages: development and tissue specialization. Annu Rev Immunol. 2015;33:643-675. Waggoner SN, Reighard SD, Gyurova IE, et al. Roles of natural killer cells in antiviral immunity. Curr Opin Virol. 2015;16:15-23. Wynn TA, Chawla A, Pollard JW. Macrophage biology in development, homeostasis and disease. Nature. 2013;496:445-455. Yang D, Biragyn A, Hoover DM, et al. Multiple roles of antimicrobial defensins, cathelicidins, and eosinophil-derived neurotoxin in host defense. Annu Rev Immunol. 2004;22:181-215.
29
CHAPTER
Extracellular Matrix Molecules T
his chapter introduces the macromolecules of the extracellular matrix. Although the extracellular matrix is composed of only five classes of macromolecules— collagens, elastin, proteoglycans, hyaluronan, and adhesive glycoproteins—it can take on a rich variety of different forms with vastly different mechanical properties. This is possible for two reasons. First, each of these classes of macromolecule comes in a number of variants (encoded by different genes or produced by alternative splicing), each with distinctive properties. Second, the cells that constitute the extracellular matrix secrete different proportions of these isoforms in various geometrical arrangements. As a result, the extracellular matrix in different tissues is adapted to particular functional requirements, which vary widely in tendons, blood vessel walls, cartilage, bone, the vitreous body of the eye, and subcutaneous fat. Beyond providing mechanical support, the extracellular matrix also strongly influences embryonic development, provides pathways for cellular migration, provides essential survival signals, and sequesters important growth factors.
Collagen The collagen family is the most abundant and versatile classes of proteins in the human body. Collagens form a wide range of different structures with remarkable mechanical properties. Weight for weight, fibrous collagens are as strong as steel. Their name, which comes from the Greek words for “glue” and “producing,” reflects the long-known adhesive properties of denatured collagen extracted from animal tissues. The defining feature of collagens is a rod-shaped domain composed of a triple helix of polypeptides (Fig. 29.1). Each polypeptide folds into a left-handed polyproline II helix that repeats every third residue with the side chains on the outside. Three of these helices associate to form a triple helix that may be up to 420 nm long.
The triple helical domains have a repeating amino acid sequence: glycine-X-Y, where X is most often proline and Y is most often hydroxyproline. The small glycine residues allow tight contact between the polypeptides in the core of the triple helix. Larger residues, even alanine, interfere with packing. Poly-L-proline has a strong tendency to form a left-handed helix like individual collagen chains but does not form a triple helix, owing to steric interference. The triple helix is most stable if all X residues are proline and all Y residues are hydroxyproline,
A
B
C
Chain B Y G G
Y X
G
G G
G
X
Chain A X
Chain C
FIGURE 29.1 COLLAGEN TRIPLE HELIX. A, End-on view of three left-handed polyproline type II helices with glycines (G) in the core. B, Longitudinal view of the strands of a triple helix. C, Spacefilling model of the structure of a short collagen triple helix. (A, Modified from van der Rest M, Garrone A. Collagen family of proteins. FASEB J. 1991;5:2814–2823. C, See Protein Data Bank [PDB; www.rcsb.org] file 1BKV.)
505
506
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
but other residues at some of these positions are essential for collagen to assemble higher-order structures. Humans have approximately 100 genes with collagen triple repeats, and more than 20 specialized collagen proteins have been characterized (Fig. 29.2 and Appendix 29.1). Most are components of the extracellular matrix, but a few are transmembrane proteins (see Fig. 31.8C). Collagen proteins were named with Roman numerals in the order of their discovery. The polypeptides are called α-chains and numbered separately. Appendix 29.1 groups collagens according to function. The size and shape of collagens vary according to function. Some collagens are homotrimers of three identical α-chains. Others are heterotrimers of two or three
MOLECULES
different α-chains. Some chains (eg, [α1(II)]) are used in more than one type of collagen. Other proteins, including the extracellular enzyme acetylcholine esterase (see Fig. 17.9) and some cell surface receptors, have similar triple helical domains but are not classified as collagens. To be a collagen, a protein must also form fibrils or other assemblies in the extracellular matrix. Nematodes, which lack connective tissue, seem to have lost the genes for fibrillar collagens but have elaborated a family of 160 genes for collagens that form their cuticle. Collagen biochemistry is challenging, because many tissue collagens are insoluble, owing to covalent crosslinking between proteins. Historic purification protocols began with proteolytic digestion to liberate
AGGREGATES
AGGREGATE MICROGRAPHS
A. Fibrillar collagens N
100 nm
C
Types I, II, III, V, and XI
Overlaps
B. Sheet-forming collagens
Gaps
7S NC1
7S N
NC1
NC1 C
Type IV
Tetramer ("spider") (NC1)2 7S
N
C
N
C
Type VIII
Type X(?)
C. Anchoring/linking collagens N
Type VI
Dimer
Tetramer
C
Beaded filament AF Dimer
NC1 N
Type VII
N
NC1
NC1 C
Anchoring fibril (Af)
NC4 GAG
Type IX
C
NC4
Type IX
BL Type II collagen fibril
N
Type XII and XIV
C
FIGURE 29.2 COMPARISON OF MAJOR COLLAGEN FAMILIES. Scale drawings and micrographs of collagen molecules and their assembly into higher-order structures. AF, anchoring fibrils; BL, basal lamina; NC1, noncollagenous domain 1; NC4, noncollagenous domain 4; 7S, a domain of type IV collagen. (Modified from van der Rest M, Garrone R. Collagen family of proteins. FASEB J. 1991;5:2814–2823.)
CHAPTER 29 n Extracellular Matrix Molecules
protease-resistant triple helical fragments. Now intact collagens can be isolated after secretion by cells in tissue culture.
Fibrillar Collagens Triple helical rod-shaped collagen molecules about 300 nm long self-associate to form strong but flexible banded fibrils (Fig. 29.2) that reinforce all the tissues of the body. Collagen fibrils form a variety of higher-order structures. Loose connective tissue (see Fig. 32.1A) has an open network of individual fibrils or small bundles of fibrils that support the cells. In many tissues, the fibrils of type I and associated collagens aggregate to form the so-called collagen fibers that are visible by light microscopy (Fig. 29.3A). In extreme cases, such as in tendons, the extracellular matrix consists almost exclusively of tightly packed, parallel bundles of collagen fibers (see Fig. 32.1B). In bone, type I collagen fibrils form regular layers reinforced by calcium phosphate crystals (see Fig. 32.4). Layers of orthogonal collagen fibers make the transparent cornea through which one sees (Fig. 29.3C). In cartilage and the vitreous body of the eye, type II collagen fibrils trap glycosaminoglycans and proteoglycans, which retain enough water for the matrix to resist compression (see Fig. 32.3) and, in the case of the eye, to provide an optically clear path for light. Fibrillar collagens are widespread in nature and have been highly conserved during evolution, so the homologs from sponges to vertebrates are similar. Each fibrillar collagen can form homopolymers in vitro; but in vivo, most form heteropolymers with at least one other type of fibrillar collagen (Appendix 29.1). This mix of the fibrillar collagen subunits is one factor that regulates the size of collagen fibers. Proteoglycans also participate in regulating collagen assembly (Appendix 29.2). Biosynthesis and Assembly of Fibrillar Collagens The biosynthesis of collagen is noteworthy for the extensive number of processing steps required to prepare the
A
B
507
protein for assembly in the extracellular matrix (Fig. 29.4). Fibroblasts synthesize type I collagen. Collagen follows the exocytic pathway used by other secreted proteins (see Chapter 21), but along the way it undergoes several rounds of precise proteolytic cleavage, glycosylation, catalyzed folding, and chemical crosslinking. The final product is a smooth fibril with staggered molecules crosslinked to their neighbors. Other fibrillar collagens are likely to be produced by similar mechanisms. Large genes with 42 exons encode the α-chains of type I collagen. All the exons for the triple helical domain were derived during evolution by duplication and divergence from a primordial exon of 54 base pairs (bp) coding for 18 amino acids or six turns of polyproline helix. Approximately half of the exons consist of 54 bp; a few with 45 bp have lost one Gly-X-Y; and the rest are 108 (2 × 54) or 162 (3 × 54) bp. Distinctive exons encode the N- and C-terminal globular domains. The initial transcript, referred to as preprocollagen, translocates into the lumen of the rough endoplasmic reticulum, where intracellular processing begins (Fig. 29.4). First, removal of the N-terminal signal sequence yields procollagen with unfolded α-chains with N- and C-terminal nonhelical propeptides. Second, enzymes hydroxylate most prolines and some lysines in the Y-position. Third, enzymes add sugars (gal-glu or gal) to the delta-carbon of some lysines, by a mechanism distinct from the typical glycosylation of asparagine or serine (Fig. 3.26). A novel mechanism initiates the folding of collagen in the endoplasmic reticulum: the C-terminal propeptides of three α-chains form a globular structure stabilized by cysteines linked with disulfide bonds. The enzyme protein disulfide isomerase catalyzes the formation of these disulfides. Formation of this globular domain has three important consequences. First, it ensures the correct selection of α-chains (two α1-chains and one α2-chain in the case of type I collagen). Second, it aligns the three polypeptides with their C-terminal
Fibroblast
C
Collagen longitudinal sections Collagen cross sections Elastic fiber FIGURE 29.3 MICROGRAPHS OF COLLAGEN FIBRILS IN CONNECTIVE TISSUES. A, Collagen fibrils (pink) in the dense connective tissue of the dermis. B, Electron micrograph of a thin section of a fibroblast, collagen fibrils, and elastic fibers. C, Orthogonal layers of collagen fibrils in the cornea of the eye. (A, Courtesy D.W. Fawcett, Harvard Medical School, Boston, MA. B, Courtesy J. Rosenbloom, University of Pennsylvania, Philadelphia. C, Courtesy E.D. Hay, Harvard Medical School, Boston, MA.)
508
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
A
Glc–Gal
Gal
B
Chain selection and registration N Glc–Gal
Proteolytic trimming C
Collagen molecule assembly
Gal
Folding Crosslinking Folded procollagen
FIGURE 29.4 BIOSYNTHESIS AND ASSEMBLY OF FIBRILLAR COLLAGEN ILLUSTRATING DETAILS COVERED IN THE TEXT. A, Translation of α-chains, chain registration, and folding. B, Secretion, assembly, and crosslinking. (Modified from Prokop DJ. Mutations in collagen genes as a cause of connective tissue diseases. N Engl J Med. 1992;326:540–546. Copyright 1992 Massachusetts Medical Society. All rights reserved.)
Gly-X-Y repeats in register, ensuring that the triple helix forms with all three chains in phase. Third, the globular propeptides prevent assembly of procollagen into insoluble fibrils during transit through the secretory pathway. Given their repeating Gly-X-Y structure, separated collagen chains without propeptides associate indiscriminately and out of register with other chains. For example, gelatin is simply a mixture of collagen chains without propeptides. Boiling dissociates the chains from each other. When cooled, the chains randomly associate out of register at random positions along their lengths, forming a branching network that solidifies into the gel that is used in food preparation. Following selection and registration of the three α-chains, the helical rod domains zip together, beginning at the C-terminus. Correct folding of the triple helix requires all-trans peptide bonds. Because proline forms cis and trans peptide bonds randomly, the slow isomerization of cis prolyl-peptide bonds to trans limits the rate of triple helix folding in vitro. The enzyme prolylpeptide isomerase catalyzes the interconversion of these prolyl-peptide bonds and speeds up folding of the triple helix in vivo. The resulting rod-shaped, triple-helix glycoprotein is called procollagen. Procollagen is too large (>300 nm long) to fit into conventional COPII coated vesicles that bud from the endoplasmic reticulum (ER) with cargo destined for the Golgi apparatus (see Fig. 21.3), so accessory proteins are required. These transmembrane proteins interact with both procollagen inside the ER and the forming COPII
coat on the cytoplasmic side of the membrane. They allow the COPII vesicle to grow large enough to accommodate protocollagen. The COPII vesicles deliver protocollagen to the Golgi apparatus. Humans with mutations in the gene for the COPII protein Sec23A have defects in collagen secretion and bone formation. Less is known about the path of procollagen through the Golgi apparatus and vesicles that transport protocollagen to the cell surface, where it is secreted. Some cells have specialized collagen assembly sites (Fig. 29.4). Like a spider trailing its silk web behind, fibroblasts help determine the arrangement of collagen fibrils as they move through tissues (Fig. 29.3C). Outside the cell, matrix metalloproteinases (Fig. 29.19) cleave the propeptides from the triple helical domain, forming the mature collagen molecule (formerly called tropocollagen). Relieved of its inhibitory propeptides, collagen selfassembles into fibrils by a classical entropy-driven process (Fig. 29.5). Weak, noncovalent bonds between collagen molecules specify the self-assembly of fibrils but provide little tensile strength. Adjacent collagen molecules are staggered by 67 nm, so a 35-nm gap is required between the ends of the collagen molecules (5 staggers at 67 nm = 335 nm = 1 molecular length of 300 nm + a 35-nm gap). The size of the fibrils is influenced by incorporation of minor fibrillar collagens (eg, collagen V) and interactions of FACIT (fibril-associated collagens with interrupted triple helices) collagens and other matrix molecules with their surfaces.
CHAPTER 29 n Extracellular Matrix Molecules
The great tensile strength of mature collagen fibrils comes from covalent crosslinking between the inextensible triple helices. For most fibrillar collagens, the enzyme lysyl oxidase catalyzes the formation of covalent bonds between the ends of collagen molecules (Figs. 29.4 and 29.6). The enzyme oxidizes the ε amino groups of selected lysines and hydroxylysines to aldehydes. These aldehydes react spontaneously with nearby lysine and hydroxylysine side chains to form a variety of covalent crosslinks between two or three polypeptides. Disulfide bonds, rather than modified lysine side chains, crosslink type III collagen fibrils. Point mutations or deletions in collagen genes or lack of function of one of the enzymes that processes collagen (lysyl hydroxylase, lysyl oxidase, or procollagen proteases) can each cause defective collagen fibrils (Appendix 29.1). These defects cause a number of deforming and even lethal human diseases: brittle bones (osteogenesis imperfecta), fragile cartilage (several forms of dwarfism), and weak connective tissue (EhlersDanlos syndrome). Chapter 32 discusses these diseases in more detail.
509
organs, epithelia, or even whole animals. Six different human genes for type IV collagen encode proteins that form net-like polymers that assemble into the basal lamina beneath epithelia (Fig. 29.7) and around muscle and nerve cells. The concluding section of this chapter provides details about basal lamina structure, function, and diseases. Hexagonal nets of type VIII collagen form a special basement membrane (Descemet membrane) under the endothelium of the cornea. Related collagens form the cuticle of earthworms and the organic skeleton of sponges.
Linking Collagens Specialized connecting and anchoring collagens (also called FACIT) link fibrillar and sheet-forming collagens
EPIDERMAL CELL
IFs
Hemidesmosomes
Sheet-Forming Collagens Collagens in this second group polymerize into sheets rather than fibrils (Fig. 29.2). These sheets surround
Basal lamina
Anchoring fibrils
Gold-labeled antibody to type VII collagen 64 nm
DERMIS 64 nm
64 nm
FIGURE 29.7 ANCHORING FIBRILS OF TYPE VII COLLAGEN. Electron micrograph of a thin section of human skin reacted with a gold-labeled antibody to the C-terminal domain of type VII collagen. Top to bottom, Basal epithelial cell with keratin intermediate filaments (IFs) attached to hemidesmosomes, which link to the basal lamina. Short fibrils of type VII collagen link the basal lamina to plaques in the dermis. Both ends of these bipolar fibrils (Fig. 29.2) are labeled with gold. Bar is 0.1 µm. (Courtesy D.R. Keene, Portland Shriners Hospital, OR.)
335 nm FIGURE 29.5 STRUCTURE OF COLLAGEN FIBRILS. Electron micrographs and drawing of molecular packing. (Micrographs courtesy Alan Hodges, Marine Biological Laboratory, Woods Hole, MA.)
Hydroxypyridinium crosslink
Collagen chain 1 Lysine Hydroxylysine OH NH2
NH2
NH2
Oxidation by lysyl oxidase
NH OH
O
O
C
C
H
OH C C
H Condensation
C N
C OH C OH
OH
Hydroxylysine Collagen chain 2 FIGURE 29.6 COVALENT CROSSLINKING OF COLLAGEN MOLECULES. After lysyl oxidase oxidizes hydroxylysine side chains, the aldehydes condense with each other and a lysine to form two- and three-membered (shown) crosslinks between adjacent collagen molecules.
510
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
to other structures (Fig. 29.2). For example, type VII collagen forms anchoring fibrils that link of the type IV collagen in the basal lamina of stratified epithelia to plaques in the underlying connective tissue (Fig. 29.7). The type VII collagen homotrimer has an exceptionally long triple-helix domain with nonhelical domains at the N-terminus of each chain. The tails of type VII molecules overlap to form antiparallel dimers that associate laterally to form anchoring fibrils. Mutations in type VII collagen cause both dominant and recessive forms of a severe blistering disease, dystrophic epidermolysis bullosa. In heterozygotes, mutated chains interfere with the assembly of anchoring fibrils by normal type VII collagen chains. Without anchoring fibrils, the basal lamina adheres weakly to the connective tissue matrix. Even mild physical trauma to the skin causes the epithelium to pull away from the connective tissue, forming a blister. Mutations in intermediate filaments cause similar defects (see Fig. 35.6). Type IX collagen heterotrimers do not polymerize. Instead, they associate laterally with type II collagen fibrils with an N-terminal helical segment and a glycosaminoglycan on a serine projecting from the surface (Fig. 29.2). In the vitreous body of the eye, these polysaccharides fill most of the extracellular space.
A
B
Endothelium Internal elastic lamina Smooth muscle
Elastic Fibers Rubber-like elastic fibers are found throughout the body and are prominent in the connective tissue of skin, the walls of arteries (Fig. 29.8), and the lung. They are entropic springs that recoil passively after tissues are stretched. For example, each time the heart beats, pressurized blood flows into and stretches the large arteries. Energy stored in elastic fibers pushes blood through the circulation between heartbeats. Elastic fibers are composite materials; a network of fibrillin microfibrils is embedded in an amorphous core of crosslinked elastin, which makes up 90% of the organic mass (Fig. 29.9). Fibroblasts produce both components. Loose bundles of microfibrils initiate assembly. A third protein, called fibulin, is required for elastin subunits to assemble between the microfibrils.
A
FIGURE 29.8 ELASTIC FIBERS IN THE WALL OF A SMALL ARTERY. A, Light micrograph of a cross section stained to bring out the internal elastic lamina (box) and wavy elastic fibers among the muscle cells. The boxed area includes the internal elastic lamina between the endothelial cells lining the lumen and the underlying smooth muscle cells. B, Electron micrograph of a longitudinal thin section illustrating the internal elastic lamina. In such standard preparations, elastic fibers stain poorly and appear amorphous except for occasional 10-nm microfibrils on the surface. (Courtesy Don W. Fawcett, Harvard Medical School, Boston, MA.)
B
FIGURE 29.9 ELECTRON MICROGRAPHS OF DEVELOPING ELASTIC FIBERS FROM A FETAL CALF. A, Longitudinal section. B, Cross section. Fibrillin microfibrils form a scaffolding for elastin, which stains darkly in this preparation. (Courtesy J. Rosenbloom, University of Pennsylvania, Philadelphia.)
CHAPTER 29 n Extracellular Matrix Molecules
511
thought to form α-helices with pairs of lysines adjacent on the surface. As tropoelastin assembles on the surface of elastic fibers, lysyl oxidase oxidizes paired lysines of tropoelastin to aldehydes. Oxidized lysines condense into a desmosine ring that covalently crosslinks tropoelastin molecules to each other (Fig. 29.11). The four-way crosslinks, involving pairs of lysines from two tropoelastin molecules, are unique to elastin. The same enzyme catalyzes the crosslinking of collagen, but it forms only twoand three-way crosslinks. Elastic fibers are similar to rubber except that elastic fibers require water as a lubricant. Hydrophobic segments between the crosslinks are thought to form extensible random coils that extend and become aligned when an elastic fiber is stretched (Fig. 29.11C). A difference in entropy of the polypeptide in the contracted and stretched states is thought to be the physical basis for the elasticity (see the Gibbs-Helmhotz equation in Chapter 4). Stretched fibers store energy, owing to ordering (low entropy) of the polypeptide chains. Fibers shorten when the resistance is reduced, because the
Fibrillin is the primordial component of elastic fibers, having arisen in Cnidarians (see Fig. 2.8). It is a long, floppy protein consisting of a tandem array of domains some of which are glycosylated (Fig. 29.10). Humans have three fibrillin genes. Fibrillin-1 is the main component of 10-nm microfibrils, along with several glycoproteins. Microfibrils are composed of parallel fibrillin molecules that interact head to tail, reinforced by disulfide bonds made by the first hybrid domains. Parts of neighboring subunits overlap in globular beads connected by flexible arrays of domains. Microfibrils are about 100 times stiffer than elastin, and they stretch by rearrangement of molecules and domains rather than unfolding. Fibrillins and related proteins called latent-TGFβ (transforming growth factor-β)-binding proteins, act as repositories for TGFβ family proteins in connective tissues. Elastin subunits are a family of closely related 60-kD proteins called tropoelastins, the products of alternative splicing from a single elastin gene. Long sequences rich in hydrophobic residues are interrupted by short sequences with pairs of lysines separated by two or three small amino acids (Fig. 29.11). Lysine-rich sequences are N
C
TB 8 cysteine
EGF-like
Calcium-binding EGF-like
Hybrid domains
Proline-rich
FIGURE 29.10 DOMAIN ORGANIZATION OF HUMAN FIBRILLIN-1. A tandem array of independently folded domains, including 47 epidermal growth factor–like (EGF-like) domains, forms a linear molecule. (Modified from Rosenbloom J, Abrams WR, Mecham R. Extracellular matrix 4: the elastic fiber. FASEB J. 1993;7:1208–1218.)
A
C K
K K K K
K K
K K
K K
K K
K
K K
K
K
Contracted, low energy, high entropy
K
K K
K
B. Crosslinking reactions
CH2 NH2 CHO OHC CHO CH2
C
CH2 NH2 CHO
NH CH2 CH2 OHC C
CH2 N CH
N C
Desmosine
C C
C C
CH2 HN LNL CH2
Stretched, high energy, low entropy
FIGURE 29.11 ELASTIN POLYPEPTIDES AND CROSSLINKING REACTIONS. A, Lysine-rich helical domains separate random chains rich in hydrophobic residues. B–C, Lysyl oxidase converts lysine amino groups to aldehydes, which react with other lysines to form simple linear crosslinks or six-membered rings linking two polypeptides. If the peptide bonds are hydrolyzed experimentally (not shown here), the linear crosslink is released as leucyl-norleucine (LNL) and the six-membered crosslink is released as the amino acid desmosine. C, Comparison of the contracted state with low-energy disordered chains having high entropy with the stretched state with high-energy ordered chains having low entropy. Elastin polypeptides form a continuous, covalently bonded network. Application of force stretches the chains between the crosslinks. This is a lowentropy, high-energy state. Reduced force allows the chains to contract into a more disordered, higher-entropy state with lower energy. (Modified from Rosenbloom J, Abrams WR, Mecham R. Extracellular matrix 4: the elastic fiber. FASEB J. 1993;7:1208–1218.)
512
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
polypeptide chains return to their disordered, lowerenergy, higher-entropy state. Only embryonic and juvenile fibroblasts synthesize elastic fibers, which turn over slowly, if at all, in adults. Consequently, adults must make do with the elastic fibers that are formed during adolescence. Fortunately, these fibers are amazingly resilient. Arterial elastic fibers withstand more than 2.5 billion cycles of stretching and recoil during a human life. Many tissues become less elastic with age, particularly the skin, which is subjected to damage from ultraviolet irradiation. Compare, for example, how readily the skin of a baby recoils from stretching compared with that of an aged person. The loss of elastic fibers in skin is responsible for wrinkles. Dominant mutations in the elastin gene cause a human disease called cutis laxa. The skin and other tissues of patients with this disease lack resilience. Collagens are found across the phylogenetic tree, but only vertebrates are known to produce elastin. Invertebrates evolved two completely different elastic proteins. Mollusks have elastic fibers composed of the protein abductin. Insects use another protein, called resilin, to make elastic fibers. Dominant mutations in the fibrillin-1 gene cause Marfan syndrome and illustrate the physiological functions of elastic fibers. Most of the thousand known fibrillin-1 mutations make the protein unstable and susceptible to proteolysis. Other point mutations interfere with folding. All patients are heterozygotes. Elastic fibers of patients with Marfan syndrome are poorly formed, accounting for most of the pathological changes. Most dangerously, weakness of elastic fibers in the aorta leads to an enlargement of the vessel, called an aneurysm, which is prone to rupture, with fatal consequences. Prophylactic replacement of the aorta with a synthetic graft and medical treatment with drugs that block β-adrenergic receptors (see Fig. 27.3) allow patients a nearly normal life span. In some patients, a floppy mitral valve in the heart causes reflux of blood from the left ventricle back into the left atrium. Weak elastic fibers that suspend the lens of the eye result in dislocation of the lens and impaired vision. Weak elastic fibers result in lax joints and curvature of the spine. Most affected patients are tall, with long limbs and fingers, but the connection of these features to fibrillin is not known. The manifestations of the disease are quite variable, even within one family, for reasons that are not understood. Mutations in the fibrillin-2 gene cause congenital contractural arachnodactyly, a disease characterized by joint stiffness.
Glycosaminoglycans and Proteoglycans Glycosaminoglycans (GAGs, formerly called mucopolysaccharides) are long polysaccharides made up of repeating disaccharide units, usually a hexuronic acid and a
hexosamine (Fig. 29.12). With one important exception— hyaluronan—GAGs are synthesized as covalent, posttranslational modifications of a large family of proteins called proteoglycans. These proteins vary in structure and function, but their associated GAGs confer some common features. All vertebrate cells synthesize proteoglycans. Most are secreted into the extracellular matrix, where they are major constituents of cartilage, loose connective tissue, and basement membranes. Mast cells package the proteoglycan serglycin, along with other molecules in secretory granules. A few proteoglycans, including syndecan and CD44, are transmembrane proteins with their GAGs exposed on the cell surface. Of the known GAGs, hyaluronan (formerly called hyaluronic acid) is exceptional in two regards. First, enzymes on the cell surface synthesize the alternating polymer of [D-glucuronic acid β (1 → 3) D-N-acetyl glucosamine β (1 → 4)]n (Fig. 29.12). Other GAGs are synthesized as posttranslational modifications of a core protein. Second, hyaluronan is not modified postsynthetically, as are all other GAGs. The linear polymer, often exceeding 20,000 disaccharide repeats (a length >20 µm) is released into the extracellular space. In contrast to proteins, nucleic acids, and even N-linked oligosaccharides, which are precisely determined macromolecular structures, the GAG chains of proteoglycans appear to vary both in length and the sequence of the sugar groups. The four-step synthesis of GAGs (Fig. 29.12) explains this variability: 1. Ribosomes associated with ER synthesize the core protein, which enters the secretory pathway. 2. In compartments between the ER and the trans-Golgi apparatus, glycosyltransferases initiate GAG synthesis by adding one of three different, short, link oligosaccharides to serine or asparagine residues of the core proteins (Fig. 29.12A–B). The structural clues identifying these sites are not understood, as they do not have a common amino acid sequence motif. A tetrasaccharide attached to serine anchors dermatan sulfate, chondroitin sulfate, and heparan sulfate. Branched oligosaccharides anchor keratan sulfate to serine or asparagine. 3. In the trans-Golgi network, other glycosyltransferases elongate the polysaccharide by adding, sequentially, two alternating sugars to the growing chain (Fig. 29.12D–F). The primary products are homogeneous, linear polymers, each with one pair of alternating sugars. 4. Enzymes modify some but not all the residues along these alternating sugar polymers by adding sulfate to hydroxyl or amino groups, or by isomerizing certain carbons to convert D-glucuronic acid to its epimer L-iduronic acid (Fig. 29.12D–F). The result is a heterogeneous polymer. The mechanisms that select particular sites for modification are not understood.
CHAPTER 29 n Extracellular Matrix Molecules
A CS
3
U
N
D. Chondroitin/Dermatan Sulfate n G
U
HS
4
U
2 X
G
O Ser
-1,4-glcUA-β
O
B
S
S
3
G
4
SO3–
-1,3-galNAc-β-1,4-idoUA-α
H n
KS
3
G
S
4
3
H
G
6 4 N O 3
Direction of synthesis
Ser (Thr)
3
G
3
S
G
4
4
H
H
2
2
M
M
CO2–
4
HO
3
HO 4
O
F 6 6 4 4 M H H 3
O N C Asn H
-1,3-gal-β-1,4-glcNAc-βHO CH2OH 4 O Direction of O 3 synthesis OH
Glucuronic acid Galactose Phosphate
S H X
Sialic acid glcNAc Xylose
F M N
–
O
1
O
4
HO
CO2 3
O HO
OH
1
4
O
3
NH Ac
3
n
Direction of synthesis
NH Ac
1
O n
-1,4-glcNAc-αCO2
4
HO
3
SO3–
O
OH
1
O
4
HO O
O n
-1,4-idoUA-α
Direction of synthesis 1
1
CH2OH O
4
-1,4-glcUA-β
O CH2OH O
NH Ac
F. Heparan Sulfate/Heparin
C. Hyaluronan -1,4-glcUA-β-1,3glcNAc-β-
CH2OH O 3
HO Fucose Mannose galNAc
SO3–
SO3–
N-linked
G
O
1
OH
E. Keratan Sulfate
O-linked
H
S
U
513
SO3–
CH2OH O 3
NH Ac
1
O n
FIGURE 29.12 SYNTHESIS OF GLYCOSAMINOGLYCANS (GAGS). A–B, Three short oligosaccharides link GAGs (left) to proteoglycan core proteins (right). A, A tetrasaccharide anchors chondroitin sulfate (CS), dermatan sulfate, and heparan sulfate (HS) to serine residues. B, Two different, branched oligosaccharides link keratan sulfate (KS) to either serine or asparagine. C–F, Four parent polymers and postsynthetic modifications. C, Hyaluronan [D-glucuronic acid β (1 → 3) D-N-acetylglucosamine β (1 → 4)]n (n ≥25,000) is not modified postsynthetically. D, Chondroitin sulfate and dermatan sulfate are synthesized as [D-glucuronic acid β (1 → 3) D-N-acetylgalactosamine β (1 → 4)]n (n usually 1 s−1) for dissociation of ligand. This actually makes good biological sense. Rapidly reversible interactions allow white blood cells to roll along the endothelium of blood vessels (Fig. 30.13). Transient adhesion also allows fibroblasts to migrate through connective tissue. In contrast, the interactions of cells in epithelia and muscle appear to be more stable, perhaps owing to multiple weak interactions between clustered adhesion proteins cooperating to stabilize adherens junctions and desmosomes (see Fig. 31.8). The combined strength of these bonds is said to increase the “avidity” of the interaction.
Fifth Principle of Adhesion Many adhesion receptors interact with the cytoskeleton inside the cell. Adapter proteins link cadherins and integrins to actin filaments or intermediate filaments. These interactions provide mechanical continuity from cell to cell in muscles and epithelia, allowing them to transmit forces and resist mechanical disruption. Sixth Principle of Adhesion Association of ligands with adhesion receptors activates intracellular signal transduction pathways, leading to changes in gene expression, cellular differentiation, secretion, motility, receptor activation, and cell division. Signaling through adhesion receptors allows cells to adjust their behavior based on physical interactions with the surrounding matrix or cells.
Identification and Characterization of Adhesion Receptors The ability of mixed populations of cells to sort into homogeneous aggregates revealed that mechanisms exist to bind like cells together. Similar experiments showed that cells also bind matrix macromolecules, such as fibronectin, laminin, collagen, and proteoglycans. Biochemical isolation of the responsible adhesion proteins was challenging, but progressed rapidly once monoclonal antibodies (see Fig. 28.10) that inhibit adhesion were available. These antibodies provided assays for purification of adhesion proteins and cloning of their complementary DNAs (cDNAs). With representatives from each family in hand, cloning cDNAs for related proteins was straightforward. Once the first crystal structures were determined, the structures of other family members could be modeled using the structures of shared functional domains.
CHAPTER 30 n Cellular Adhesion
527
Insights about the functions of adhesion receptors came in several steps. Localization of a protein on specific cells frequently provided the first clues. Typically, the expression of each protein is restricted to a subset of cells or to a specific time during embryonic development or both. Next, investigators used specific antibodies to test for the participation of the adhesion protein in cellular interactions in vitro or in tissues. Both human genetic diseases and mutations in mice and other organisms produce defects caused by the absence of adhesion proteins. Blistering skin diseases called pemphigus illustrate the serious consequences when pathological autoantibodies disrupt adhesion between skin cells expressing the antigen (see Fig. 31.8). In leukocyte adhesion deficiency, white blood cells lack the β2-integrin that is required to bind the endothelial cells that line blood vessels. These defective white blood cells fail to bind to blood vessel walls or to migrate into connective tissue at sites of infection. Patients with a bleeding disorder called Bernard-Soulier syndrome lack one of the adhesion receptors for von Willebrand factor, a protein that promotes platelet aggregation. Loss of cadherins contributes to the spread of some cancer cells.
Immunoglobulin Family of Cell Adhesion Molecules The IgCAM family includes hundreds of adhesion proteins that bind ligands on the surfaces of other cells. Some interactions are homophilic binding to the same IgCAM on another cell; others are heterophilic with different IgCAMs, integrins, other proteins or proteins with sialic acid. These interactions help specify interactions between different cell types in developing and mature animals. IgCAMs have one to seven extracellular immunoglobulin domains anchored to the plasma membrane by a single transmembrane helix (Fig. 30.3 and Table 30.1). The compact immunoglobulin (Ig) domains consist of 90 to 115 residues folded into seven to nine β-strands in two sheets, usually stabilized by an intramolecular disulfide bond. The N- and C-termini are at opposite ends of the Ig domains, so they can form linear arrays. Some nervous system IgCAMs have three or four fibronectin III (FN-III) domains (see Fig. 13.13) between the Ig domains and the membrane anchor. Most IgCAMs consist of a single polypeptide, but others are multimeric, with two (CD8) or four subunits (see Fig. 27.8 for the T-cell receptor). The C-terminal cytoplasmic tails of these receptors vary in sequence and binding sites for intracellular ligands. The cytoplasmic domains of the lymphocyte accessory receptors CD4 and CD8 bind protein tyrosine kinases required for cellular activation (see Fig. 27.8 for CD4). The cytoplasmic domains of neuronal IgCAMs bind PDZ domain proteins or the membrane skeleton (see Fig. 13.11). An adapter protein links IgCAMs in the nectin family to cytoplasmic actin filaments.
528
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
A
2
4
3 2
1 2
5
3
ICAM-1 CD54
5
1
Nectin
1 2
ICAM-2 CD102
4
3
1 3
2
5
6
6
7
7
VCAM-1 CD106
2
1
1 2
1
3
3
MAdCAM-1
B
CD4D1D2
CD4D3D4
CD8α/α
FIGURE 30.3 STRUCTURES OF REPRESENTATIVE IMMUNOGLOBULIN CELL ADHESION MOLECULES. A, Domain maps of examples with their common names and CD numbers. B, Ribbon diagrams of the lymphocyte coreceptors CD4 (domains 1 and 2 on the left and domains 3 and 4 in the middle) and CD8. ICAM, intercellular adhesion molecule; MAdCAM, mucosal addressin cell adhesion molecule; VCAM, vascular cell adhesion molecule. (For reference, see Protein Data Bank [PDB; www.rcsb.org] files 3CD4, 1CID, and 1CD8.)
Differentiated metazoan cells express IgCAMs selectively, especially during embryonic development, when they contribute to the specificity of cellular interactions required to form the organs. Neurons and glial cells express specific IgCAMs that guide the growth of neurites, mediate synapse formation, and promote the formation of myelin sheaths. In adults, interaction of endothelial cell intercellular adhesion molecule (ICAM)-1 with a white blood cell integrin is essential for adhesion and movement of the leukocytes into the connective tissue at sites of inflammation (Fig. 30.13). Like other cell adhesion proteins, IgCAMs participate in signaling processes. Best understood are interactions of lymphocytes with antigen-presenting cells during immune responses. IgCAMs reinforce the interaction of T-cell receptors with major histocompatibility complex molecules carrying appropriate antigens on other cells (see Fig. 27.8). Although individual interactions are weak, the combination of specific (T-cell receptor) and nonspecific (CD2 and CD4) interactions with the target cell suffices to initiate signaling.
Cadherin Family of Adhesion Receptors The complex architecture of organs in vertebrates relies on Ca2+-dependent associations between the cells mediated by more than 80 cadherins (Table 30.2). Their name derives from “calcium-dependent adhesion” protein.
TABLE 30.1 Immunoglobulin Family of Cell Adhesion Molecules* Examples
Structure
Extracellular Ligands
Intracellular Ligands
CD2†
2Ig-1TM
LFA-3 (CD58)
CD4†
4Ig-1TM
Class II MHC
CD8†
Dimer: 1Ig-1TM
Class I MHC
C-CAM
4Ig-1TM
Self
F11 (contactin)
6Ig-4FN-II-1TM
ICAM-1†
5Ig-1TM
ICAM-2
2Ig-1TM
L1 (Ng-CAM) [mouse]
6Ig-3FN-III-1TM
Self
Neurons, Schwann cells
Adhesion
LFA-3 (CD58)
2Ig-1TM or GPI anchor
CD2
WBCs, epithelia, fibroblasts
Adhesion
MAG
5Ig-1TM
Neurons
Glial cells
Myelin formation
NCAM
5Ig-3FN-III-1TM
Self
Neurons, other cells
Adhesion
Neurofascin [chick]
6Ig-4FN-III-1TM
? Self
Neurites
Bundling neurites
PECAM-1 (CD31)
6Ig-1TM
Self
Platelets, endothelium, myeloid cells
Adhesion
TAG-1
6Ig-4FN-III-GPI anchor
? Self
Neurons
Neuron migration
VCAM-1
7Ig-1TM
WBC α4 integrin
Endothelium (regulated)
WBC/endothelium adhesion
Expression
Functions
T cells
T-cell activation
Lck
T cells, macrophages
T-cell coreceptor
Lck
Cytotoxic; other T cells
T-cell coreceptor
Liver, intestine, WBCs
Cell adhesion
Neurons
Neurite fasciculation
Epithelia, WBCs
WBC adhesion
LFA-1, MAC-1
Endothelium, WBCs Ankyrin
Ankyrin
CAM, cell adhesion molecule; CD, cellular differentiation antigen; FN-III, fibronectin-III domain; GPI, glycosylphosphatidylinositol; ICAM, intercellular adhesion molecule; Ig, immunoglobulin domain; Lck, nonreceptor tyrosine kinase; LFA, lymphocyte function–associated antigen; MAG, myelin-associated glycoprotein; MHC, major histocompatibility complex; NCAM, neural cell adhesion molecule; PECAM, platelet endothelial cell adhesion molecule; TAG, transient axonal glycoprotein; TM, transmembrane domain; VCAM, vascular cell adhesion molecule; WBC, white blood cells. *Hundreds are known. † Partial atomic structure.
CHAPTER 30 n Cellular Adhesion
529
TABLE 30.2 Cadherin Family of Adhesion Molecules* Type (Examples)
Extracellular Ligands
Intracellular Ligands
Expression
Functions
E-cadherin
Self
Catenins (actin)
Epithelia, others
Adherens junctions
N-cadherin
Self
Catenins (actin)
Neurons, muscle, endothelium
Adhesion
R-cadherin
Self
Catenins (actin)
Retina, neurons
Adhesion
Desmocollins
Self, desmogleins
Plakoglobin (desmoplakin, plackophilin, IF)
Epithelia
Desmosomes
Desmogleins
Self, desmocollins
Plakoglobin (desmoplakin, plackophilin, IF)
Epithelia, heart
Desmosomes
Self
None (GPI anchor)
Early embryos, neurons
Intercellular adhesion
Self
Fyn tyrosine kinase (some)
Vertebrate neurons, other cells
Self-avoidance
Self
None
Endocrine glands, neurons
Intercellular adhesion
Classic Cadherins
Desmosomal Cadherins
Atypical Cadherins T-cadherin Protocadherins α-, β-, and γ- Protocadherins Signaling Cadherins RET protooncogene
GPI, glycosylphosphatidylinositol; IF, intermediate filament. *More than 80 are known. Plakoglobin is also known as γ-catenin.
Genes for cadherin domains appeared in unicellular precursors of sponges, representing an early step toward the evolution of metazoan organisms. Cadherins generally interact with like cadherins on the surfaces of other cells in a calcium-dependent fashion, but some cadherins form heterophilic interactions. Homophilic interactions of cadherins link epithelial and muscle cells to their neighbors, especially at specialized adhesive junctions called adherens junctions and desmosomes (Fig. 30.4; also see Fig. 31.8). The structural hallmark of the cadherin family is the CAD domain (Figs. 30.5 and 30.6), which consists of approximately 110 residues folded into a sandwich of seven β-strands. This fold resembles Ig and FN-III domains, but appears to be a case of convergent evolution. N- and C-termini are on opposite ends of CAD domains. Ca2+ bound to three sites between adjacent CAD domains links them together into rigid rods. Without Ca2+, the domains rotate freely around their linker peptides. Classic cadherins interact head to head through their N-terminal CAD1 domains (Fig. 30.6D) forming strong “trans-interactions” with a partner on another cell. Cadherins on the same cell can interact laterally in “cis-interactions.” Three-dimensional reconstructions of electron micrographs of desmosomes show both transand cis-interactions (Fig. 30.6B). Cadherins are synthesized with a small domain before the N-terminal interaction strand, which must be removed by proteolysis to allow binding to another cadherin. A single α-helix links classic cadherins and desmosomal cadherins to the plasma membrane, but T-cadherin
A. Desmosome
B. Adherens junction
Cadherins Cadherins
FIGURE 30.4 ELECTRON MICROGRAPHS OF ROD-LIKE CADHERINS CONNECTING THE PLASMA MEMBRANES OF ADJACENT CELLS. Intestinal epithelial cells were prepared by rapid freezing, freeze-fracture, deep etching, and rotary shadowing. A, Desmosome with associated intermediate filaments in the cytoplasm. B, Adherens junction with associated actin filaments. (Courtesy N. Hirokawa, University of Tokyo, Japan. Modified from Hirokawa N, Heuser J. Quick-freeze, deep-etch visualization of the cytoskeleton beneath surface differentiations of intestinal epithelial cells. J Cell Biol. 1981;91:399–409, copyright The Rockefeller University Press.)
has a glycosylphosphatidylinositol (GPI) anchor (see Fig. 13.10). Cytoplasmic domains vary in size, sequence, and binding sites for associated proteins. The protooncogene RET is a cadherin with a cytoplasmic tyrosine kinase domain. Adapter proteins link the cytoplasmic domains of cadherins to actin filaments or intermediate filaments
530
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
E-cadherin
ICS β-catenin binding
Desmocollin-1a
Desmoglein-1
ICS
ICS
T-cadherin Glycosylphosphatidylinositol anchor Protocadherins Src-family kinase binding RET Tyrosine kinase FIGURE 30.5 DOMAIN MAPS OF CADHERINS. All have extracellular CAD domains. A single transmembrane segment anchors five examples. A glycosylphosphatidylinositol (GPI) tail anchors T-cadherin. Intracellular cadherin segment (ICS) domains interact with catenin adapters to link E-cadherin to actin filaments and desmocollin and desmoglein to intermediate filaments.
to reinforce adhesion and maintain the physical integrity of tissues (see Fig. 31.8). The cytoplasmic tails of classic cadherins bind along the entire length of the adapter protein β-catenin (catenin is “link” in Greek), a long, twisted coil of 36 short α-helices (see Fig. 13.10F). Monomers of α-catenin link β-catenin to actin filaments, an interaction strengthened by tension. α-Catenin dimers also stabilize actin filaments. The more complicated cytoplasmic domains of desmosomal cadherins (desmocollins and desmogleins) interact with γ-catenin (a relative of β-catenin also called plakoglobin) and desmoplakin. Desmoplakin links these cadherins to keratin intermediate filaments (see Fig. 31.8B). The tails of some cadherins interact with formins, proteins that nucleate and elongate actin filaments (see Fig. 33.14).
Signaling by Cadherins and Catenins In addition to helping with the mechanical sorting of embryonic cells, cadherins produce signals that influence cellular proliferation, migration, and differentiation. For example, interactions between cadherins on epithelial cells result in “contact inhibition” of both growth and motility (see Figs. 41.3 and 41.11). Rho-family guanosine triphosphatases (GTPases) (see Fig. 33.19) mediate contact inhibition of movements. The GTPase Rho stimulates contraction at the contact site, while the GTPase Rac drives protrusion of the side of the cell facing away from the contact site. Both cadherins and
Eph receptor tyrosine kinases participate in this reaction to cell-cell contact. Engagement of cadherins at sites of contact also stops cellular proliferation through the Hippo signaling pathway of protein kinases. This pathway inhibits expression of genes required for cell cycle progression such as cyclin-dependent kinases (see Fig. 41.3). The mechanism involves phosphorylation of a transcription factor, excluding it from the nucleus. Contact inhibition of growth and motility suppresses the spread of cancer cells, so mutations disabling E-cadherin can contribute to the transition from benign to invasive malignant tumors. For example, genetic defects in E-cadherin predispose people to stomach cancer. The oncogenic tyrosine kinase Src (see Box 27.5) phosphorylates both E-cadherin and β-catenin. This phosphorylation is associated with loss of adhesion of epithelial cells, suggesting one way in which transformation might alter cellular adhesion. In addition to linking cadherins to actin filaments, β-catenin is an active component in the Wnt signal transduction pathway that regulates gene expression during differentiation of embryonic cells (Fig. 30.7). Extracellular signaling proteins called Wnts regulate the concentration of β-catenin available to regulate gene expression. The pathway was discovered in Drosophila, where it helps to determine the polarity of segments in early embryos. Wnts were named from the original Drosophila gene Wingless and the mouse protooncogene Int-1. The human genome encodes 29 Wnts. Most β-catenin in cells is bound to cadherins, but a second pool exchanges between the cytoplasm and the nucleus, where it recruits transcription factors to regulate the expression of genes for cellular proliferation and tissue differentiation. Cells synthesize β-catenin continuously, but it turns over rapidly in resting cells, so little accumulates in the nucleus. A cytoplasmic complex controls degradation of β-catenin. The complex consists of two kinases—glycogen synthase kinase (GSK) and casein kinase-1α—and the product of the APC gene (defective in patients with familial adenomatous polyposis coli, giving rise to multiple precancerous polyps in the large intestine). Phosphorylated β-catenin is ubiquitylated and degraded by proteasomes. Wnts suppress the degradation of β-catenin by binding to seven-helix receptors and another class of receptors in the plasma membrane. Several steps downstream in an incompletely characterized pathway, the Wnt signal inhibits GSK and casein kinase-1α. Inhibition of the kinases stops proteolysis of β-catenin, so ongoing synthesis of β-catenin raises the concentration that is free to accumulate in the nucleus. Stem cell proliferation is one of many developmental events influenced by Wnt signaling and adhesion by cadherins (see Box 41.2). Wnt binding its receptors also activates a parallel “noncanonical pathway” involving Rho-family GTPases that control the cytoskeleton during cell migration.
CHAPTER 30 n Cellular Adhesion
A. EM data
531
D. Crystal structures
B. 3D reconstruction from EM E-CAD1
E-CAD2
E-CAD3
E-CAD4
C. Atomic models fit to EM surfaces
E-CAD5
β-catenin
FIGURE 30.6 ADHESION BY CADHERINS. A, Electron micrograph of a thin section of a desmosome, colorized to emphasize the plasma membranes (red) and cadherin extracellular domains (blue). B, Three-dimensional reconstructions of the plasma membrane and cadherin extracellular domains. C, Crystal structure of the C-cadherin extracellular domains fit into electron microscopic reconstructions of intercellular links between the cells. D, Ribbon diagrams of the crystal structure of a dimer of C-cadherin extracellular domains compared with the book icon for cadherins. The inset highlights the antiparallel intermolecular interaction of the two CAD1 domains mediated by flexible N-terminal peptides. A conserved tryptophan fits into a hydrophobic pocket of the partner CAD1 domain, forming the reciprocal interactions. Calcium ions (blue) stabilize interactions between CAD domains. (A–C, From He W, Cowin P, Stokes DL. Untangling desmosomal knots with electron tomography. Science. 2003;302:109– 113, copyright the American Association for the Advancement of Science. D, For reference, see PDB file 1L3W and Boggen TJ, Murray J, Chappuis-Flament S, et al. C-Cadherin ectodomain structure and implications for cell adhesion mechanisms. Science. 2002;296:1308–1313.)
Wnt Cadherin Wnt receptor β-catenin bound to cadherin
Free β-catenin β-catenin bound to APC
CYTOPLASM
Transcription factors NUCLEUS
Tcf
DNA
Gene expression
APC
Casein kinase 1α GSK E3 ubiquitin ligase SCFβTRCP Degradation of β-catenin by proteasomes
FIGURE 30.7 PARTICIPATION OF β-CATENIN IN GENE EXPRESSION. Free β-catenin is in equilibrium with binding sites on cadherins and APC (adenomatous polyposis coli) and may also enter the nucleus, where it combines with Tcf/LEF-1 transcription factors. When β-catenin concentrations are low in the nucleus, Tcf/LEF-1 represses gene expression, but the complex of β-catenin with Tcf/LEF-1 activates the expression of genes for cellular growth and differentiation. Synthesis of β-catenin is constant, so degradation determines its concentration in cytoplasm: glycogen synthetase kinase (GSK) and casein kinase-1α phosphorylate β-catenin bound to APC, triggering its ubiquitinylation and degradation. Extracellular Wnt acts through a seven-helix receptor and another receptor to promote gene expression by inhibiting the phosphorylation and degradation of β-catenin.
532
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
Mutations can alter the balance of β-catenin synthesis and turnover. Loss of APCs or mutations of phosphorylation sites on β-catenin result in excess β-catenin that enters the nucleus and stimulates proliferation. The cadherin expressed from the RET protooncogene signals through its cytoplasmic tyrosine kinase domain (Fig. 30.5). Point mutations in the segment between the CAD domains and the plasma membrane or tyrosine kinase of RET cause constitutive dimerization of the receptor or activation of the tyrosine kinase or both. These mutations cause dominantly inherited cancers of endocrine glands. On the other hand, mutations that disable RET cause Hirschsprung disease. Autonomic nerves in the wall of the intestines fail to develop, causing severe dysfunction.
Roles of Cadherins in Organ Formation Differential expression and regulation of cadherins help guide organ formation during embryonic development (Fig. 30.8). Cells with matching cadherins bind together and exclude cells without those cadherins (or other appropriate adhesion receptors), although the mechanism is more complicated than differential affinities of cadherins for each other. For example, cadherins can be activated or inactivated from inside the cell by signaling pathways in response to growth factors or other adhesion proteins. In other situations, IgCAMs facilitate the assembly of cadherins in adhesive junctions. All cells of early embryos express several different cadherins, but as soon as the embryo forms three germ layers, the ectoderm on the outside surface expresses E-cadherin. In its absence, embryos die. Subsequently, when ectoderm folds inward to form the neural tube, those cells switch to expressing N-cadherin. Neurons use protocadherins to avoid making synapses with themselves (see below). Later in development, cells in specialized organs typically express characteristic
A
B Ectoderm
Neural tube L-CAM (E-cadherin) Cadherin 6B N-cadherin FIGURE 30.8 RESTRICTED EXPRESSION OF CADHERINS DURING FORMATION OF THE NEURAL TUBE. A, Distribution of three cadherins before and after the neural tube forms. B, Fluorescent antibody staining reveals the selective expression of cadherin 6B (green) and N-cadherin (red) in the neural tube of a developing chick embryo. (Courtesy M. Takeichi, Kyoto University, Japan.)
cadherins, such as those in osteoblasts (OB-cadherin), kidney (K-cadherin), and muscle (M-cadherin). A giantsized cadherin links sensory stereocilia on the hair cells in the inner ear. These tip links pull open ion channels when the stereocilia move in response to sound waves.
Integrin Family of Adhesion Receptors Integrins are the main cellular receptors for ECM molecules (Table 30.3), but some integrins bind adhesion molecules on other cells. Their genes arose early in metazoan evolution, so they are present in sponges and corals that branched early in the evolution of animals (see Fig. 2.8). Fibroblasts and white blood cells use integrins to adhere to fibronectin and collagen as they move through the ECM. Integrins bind epithelial and muscle cells to laminin in the basal lamina, providing the physical attachments necessary to transmit internal forces to the matrix and to resist external forces (see Fig. 31.8). These interactions generate signals that control cell growth and structure. When defects in small blood vessels need repair, integrins allow platelets to adhere to basement membrane collagen and to each other via plasma fibrinogen (see Fig. 30.14). Together, these interactions are essential for tissue development and integrity in multicellular organisms. Genetic losses of integrin function result in several human diseases.
Structure of Integrins Integrins are heterodimers of two transmembrane polypeptides called α- and β-chains, which both contribute to ligand-binding specificity (Fig. 30.9). Vertebrate cells use a combinatorial strategy to establish their integrin repertoire by selectively expressing a subset of 18 different α-chains and eight β-chains. These chains combine to form at least 24 different kinds of dimers that bind different ligands. Alternative messenger RNA (mRNA) splicing (see Fig. 11.6) adds to the diversity of integrin isoforms. The ligand-binding domains of the α- and β-chains form a globular head connected to the plasma membrane by 16-nm legs (Fig. 30.9). More than 25 disulfide bonds stabilize these domains. All integrin β-chains and a subset of integrin α-chains have an I domain (inserted domain) with a bound divalent cation that interacts with acidic residues of ligands. All α-chains have an N-terminal β-propeller domain similar to a trimeric G protein β-subunit (see Fig. 25.9). Interaction of the α-chain propeller domain with the β-chain I domain holds the integrin dimer together. Single transmembrane segments anchor both integrin chains to the cell membrane. Short (α ≤77 residues; β = 40 to 60 residues, except β4 = 1000 residues) C-terminal cytoplasmic tails contribute to efficient heterodimer assembly. The I domains and the β-propeller bind at least two sites on ligands. For instance, integrin α5β1 binds two
533
CHAPTER 30 n Cellular Adhesion
TABLE 30.3 Integrin Family of Cell Adhesion Molecules* Examples
Structure
Extracellular Ligands
Me2+
Intracellular Ligands
Expression
Fibronectin receptors
α5β1, others
Fibronectin
Ca
Talin, paxillin
Fibroblasts, other cells
Cell-matrix adhesion
GPIIb/GPIIIa
αIIbβ3
Fibrinogen, von Willebrand factor
Ca
Talin, paxillin
Platelets
Platelet aggregation
Laminin receptor
α6β1, α7β1
Laminin
Yes
Talin, paxillin
Epithelia, muscle
Cell-matrix adhesion
LFA-1 (CD11/CD18)
αLβ2
Ig-CAM-1, -2, -3
Mg
Talin, paxillin
All WBCs
WBC/endothelium adhesion
MAC-1
αMβ2
Ig-CAM-1, fibrinogen
Yes
Talin, paxillin
WBCs except lymphocytes
WBC/endothelium adhesion
Vitronectin receptor
αVβ3
Vitronectin, fibronectin
Ca
Talin, paxillin
Endothelium, smooth muscle, others
VLA-4
α2β1
Collagen, laminin
Mg
Talin, paxillin
WBCs, epithelium, endothelium
Function
WBC/matrix adhesion
CD, cellular differentiation antigen; GP, glycoprotein; ICAM, intercellular adhesion molecule; LFA, lymphocyte function–associated antigen; Me2+, divalent cation dependence; VLA, very late antigen; WBC, white blood cell. *Twenty-four are known.
sites on fibronectin: an arginine-glycine-aspartic acid (RGD) sequence on a surface loop of FN-III domain 10 and a secondary site on the adjacent FN-III domain 9 (see Fig. 29.14). Both sites are required for binding, so simple RGD peptides can compete fibronectin from the integrin. Integrin binding sites of some ligands are on separate polypeptide chains. The RDD binding site for integrin α1β1 is on three different polypeptide chains of the type IV collagen triple helix. Binding of extracellular ligands (outside-in signaling) and intracellular ligands (inside-out signaling) influences the conformation and activity of integrins (Fig. 30.10). Without bound ligands integrins bend over on closely spaced legs held together by interactions of the transmembrane helices. This closed state has a low affinity for extracellular ligands owing to an occluded binding site. Ligand binding inside or outside the cell favors an open state with widely spaced, extended legs holding the head above the membrane. In this conformation the exposed ligand-binding site has the highest affinity for extracellular ligands.
Extracellular Ligands Integrins are more promiscuous than most adhesion receptors, as some bind to several protein ligands, and many matrix molecules bind to more than one integrin. For example, fibronectin binds to at least nine different integrins, and both laminin and von Willebrand factor bind at least five different integrins. This promiscuity reflects common motifs in the ligands. Approximately one-third of matrix ligands for integrins involve the sequence motif RGD or other simple sequences in otherwise unrelated proteins. Even in the open state (legs apart), integrins generally have a low affinity for extracellular ligands. For example, the micromolar Kd for integrin α5β1 binding fibronectin
A
β-propeller
Ig domains αV
Optional I domain I domain
B
Ig-like
EGF domains
β
(Book integrin icon)
α β C
α1, α2, αL, αM, αX (~ 1200 aa)
I domain β-propeller
α3, α5, α6, αIIb, αV (~1000 aa) I domain EGFs
β subunit (~750 aa)
FIGURE 30.9 INTEGRIN ARCHITECTURE. A, Ribbon diagram of integrin αVβ3 based on a crystal structure of the extracellular domain. The I domain is inserted into the sequence of an immunoglobulin- like domain. B, Integrin icon used throughout this book. C, Domain models of integrin polypeptides. Both α-chains and β-chains have single transmembrane segments and cytoplasmic tails that vary in length. All β-chains and some α-chains have an I domain (red) that binds a divalent cation and participates in ligand binding. The seven blades of the α-chain β-propeller domains are shown in orange. The α-chain I domain, if present, is inserted between the second and third of the seven blades of its propeller domain. (A, Based on an atomic model. For reference, see PDB file 1JV2 and Xiong JP, Stehle T, Diefenbach B, et al. Crystal structure of the extracellular segment of integrin αVβ3. Science. 2001;294:339–345. C, Modified from Kuhn K, Eble J. The structural basis of integrin-ligand interactions. Trends Cell Biol. 1994;4:256–261.)
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
B. Open State
A. Low affinity closed state
Ligand binding site
I-domain
β-subunit
β-propeller
C. Open state with ligand bound
α-subunit
534
Talin FERM domain
FIGURE 30.10 CONFORMATIONAL STATES OF INTEGRINS. Drawings based on atomic models derived from crystal structures and electron microscopy. The bent closed conformation observed in all crystal structures is inactive. Binding of either an extracellular ligand to the head or an activated signal transduction protein such as talin to the cytoplasmic domains can favor the open active state. (Modified from Xiao T, Takagi J, Coller BS, et al. Structural basis for allostery in integrins and binding to fibrinogen-mimetic therapeutics. Nature. 2004;432:59–67.)
A
results in rapid association and dissociation, allowing cells to adjust their grip on fibronectin in the matrix as they move through connective tissue. Nonadhesive RGD proteins, such as tenascin (see Fig. 29.16), may modulate these interactions by competing with fibronectin and other ligands for binding integrins.
Intracellular Ligands Cytoplasmic tails of integrins interact directly or indirectly with a remarkable variety of signaling and structural proteins (Fig. 30.11). These interactions are best understood at focal contacts, specialized sites where integrins cluster together to transduce transmembrane signals and link actin filaments across the plasma membrane to the ECM. The adapter proteins talin and vinculin link the cytoplasmic domains of β-integrins to actin filaments at the ends of stress fibers. Paxillin links integrins to signaling proteins, forming a scaffold for Src family tyrosine kinases (see Fig. 25.3) and focal adhesion kinase (an essential tyrosine kinase). Several types of integrins associate laterally, in the plane of the bilayer, with other transmembrane proteins. The best characterized is CD47 (also called integrinassociated protein), an IgCAM with five transmembrane B
GLASS SLIP ECM
PM
Actin
C
EXTRACELLULAR MATRIX Integrins
? VASP
Src Talin
Csk
FAK (focal adhesion kinase) Paxillin Crk (SH3 /SH2 adaptor) Cas (adaptor)
Vinculin
FIGURE 30.11 FOCAL CONTACTS OF EPITHELIAL CELLS WITH THE EXTRACELLULAR MATRIX (ECM). A, Fluorescence micrograph of parts of two vertebrate tissue culture cells with focal contacts labeled with a fluorescent antibody to phosphotyrosine (orange). Actin filament stress fibers are stained green with phalloidin. B, Electron micrograph of a thin section of two focal contacts showing fine connections to the ECM deposited on the surface of the glass coverslip and cross-sections of actin filaments in the cytoplasm. This HeLa (Henrietta Lacks) cell was grown on a glass coverslip, fixed, and cut perpendicular to the substrate. C, Drawing of the interactions of some of the proteins concentrated on the cytoplasmic face of the membrane at focal contacts. For clarity, the actin filament interactions (left) are shown separately from some signaling proteins (right). The rod-shaped dimeric protein talin interacts with the cytoplasmic domains of β-integrins and actin filaments. Vinculin interacts with membrane phospholipids, actin filaments, and talin. An unidentified protein (the question mark) links the adapter protein paxillin to integrins. Paxillin anchors tyrosine kinases (FAK and Src) and, after phosphorylation, the adapter proteins Crk and Cas. (A, Courtesy K. Burridge, University of North Carolina, Chapel Hill. B, Courtesy Pamela Maupin, Johns Hopkins University, Baltimore, MD. From Maupin P, Pollard TD. Improved preservation and staining of HeLa cell actin filaments. J Cell Biol. 1983;96:51–62, copyright the Rockefeller University Press. C, For reference, see Turner C. Paxillin and focal adhesion signaling. Nat Cell Biol. 2000;2:E231–E236; and Critchley DR. Focal adhesions—the cytoskeletal connection. Curr Opin Cell Biol. 2000;12:133–139.)
segments. Binding of the adhesive glycoprotein, thrombospondin, to the extracellular Ig-like domain of CD47 generates a transmembrane signal through trimeric G-proteins that contributes to neutrophil and platelet activation.
Outside-in Signaling From Integrins Integrin binding to matrix ligands initiates signals that modify cellular adhesion, locomotion, and gene expression, with the responses depending on the particular integrin and cell. Extracellular ligands stabilize the open state with wide spacing of the cytoplasmic domains. Presumably this physical change influences the activities of signal transduction proteins associated with the cytoplasmic domains, but the details are not known. Within seconds, the cytoplasmic tyrosine kinases shown in Fig. 30.11 phosphorylate several focal adhesion proteins, including paxillin, tensin, and focal adhesion kinase, which has a central role in transducing these signals. Within a minute, some cells raise their cytoplasmic Ca2+ concentration high enough to initiate many calciumdependent processes (see Chapter 26). Over a period of minutes, ligand binding to integrins also activates Rho-family GTPases that stimulate actin assembly (see Fig. 33.19) and spreading of the cell on ligand-coated surfaces. Other Rho-family GTPases drive contraction of the trailing edge of moving cells (see Fig. 38.6). Integrins cluster together in small “focal complexes” at the leading edge and grow into mature focal contacts (Fig. 30.11A), also called focal adhesions, which anchor actin filament stress fibers to the cell membrane. Contraction of stress fibers applies tension to the focal contacts, which remain stationary as the cell advances past them. Rapid rearrangements of the linker proteins between the integrins and actin serve as a molecular clutch to transmit forces, even as actin assemble and disassembles. A Ca2+-mediated signal inactivates obsolete attachments at the rear of the cell. The adhesiveness of a cell for its substrate (a function of integrin density on the cell, ligand density on the substratum, and their affinity) determines the rate of movement. The maximum rate occurs at intermediate adhesiveness. Rapid association and dissociation of integrins on matrix ligands allow cells to rearrange their hold on the matrix as they move. After several hours of integrin engagement, activation of the Ras/mitogen-activated protein kinase pathway (see Fig. 27.6) turns on expression of selected genes. These changes in gene expression contribute to cellular differentiation during development. Integrins allow cells to include the ECM as a signaling input along with other stimuli operating through different receptors. Inside-Out Signaling to Integrins Cells fine-tune their interactions with matrix molecules by regulating the activity of cell-surface integrins. For
CHAPTER 30 n Cellular Adhesion
535
example, integrins on white blood cells (Fig. 30.13) and platelets (Fig. 30.14) require “inside-out” activation before they can bind their extracellular ligands. Integrin activation also regulates cellular interactions during development. Cytoplasmic proteins, talin and kindlins, activate integrins by binding the cytoplasmic tail of the β-integrin and separating the two transmembrane domains. One pathway downstream from seven-helix receptors uses a membrane-bound GTPase and an adapter protein to bring together talin and the β-integrin. Some cells can mobilize a reserve pool of integrins stored in cytoplasmic vesicles within minutes. For example, chemoattractants stimulate white blood cells to fuse storage vesicles containing integrins with the plasma membrane (Fig. 30.13). Both intracellular and extracellular ligands can cluster integrins in focal complexes and focal adhesions and increase their activities.
Biological Functions of Integrins With the exception of red blood cells, integrins are present in the plasma membranes of most animal cells. Experiments with neutralizing antibodies, genetic diseases, and experimental gene disruptions revealed the functions of integrin isoforms. For example, many vertebrate cells express β1 and β3 integrins for adhesion to the ECM, so integrin antibodies inhibit cell migration and embryonic development by competing with fibronectin. Like null mutations in the fibronectin gene (see Fig. 29.14), homozygous disruption of the integrin α4 or α5 genes is lethal during development. Only white blood cells express β2-integrins, which they use to bind endothelial cells lining the walls of blood vessels. Other integrins bind adhesion molecules on other cells. For example, mouse sperm bind integrins on the egg membrane during fertilization. Integrins cooperate with adhesion receptors of the IgCAM, mucin, and selectin families to facilitate the adhesion of white blood cells to endothelial cells at sites of inflammation (Fig. 30.13). Other cells supplement the functions of integrins with structurally distinct matrix adhesion proteins, such as muscle dystroglycans and platelet GPIb-IX-V. Integrins also participate in the decision of cells to undergo apoptosis, programmed cell death (see Chapter 46). Normal epithelial cells require anchorage to the basal lamina by β4-integrins to grow and divide. When forced to live in suspension or when dissociated from the matrix by RGD peptides, these cells arrest in the G1 phase of the cell cycle (see Chapter 41) and eventually undergo apoptosis. Loss of contact with the basal lamina may contribute to the terminal differentiation and death of cells in the upper levels of stratified epithelia, such as skin (see Figs. 35.6 and 40.1). Epithelial cancers typically lose this integrin-mediated, anchorage dependence for growth, which is one of the normal limitations on uncontrolled proliferation in inappropriate locations.
536
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
Snake venoms contain small, monomeric RGD proteins that inhibit blood clotting by competing with fibrinogen for binding the integrins that activated platelets use for aggregation. These “disintegrins” are potential inhibitors of the pathological thrombosis that contributes to heart attacks and strokes. Both small-molecule and antibody antagonists for integrins are now used as clinical treatments for heart attacks and stroke.
Selectin Family of Adhesion Receptors White blood cells and platelets use three selectin proteins to interact with vascular endothelial cells and each other. In lymph nodes or at sites of inflammation, selectins snare circulating white blood cells, allowing them to roll over the surface of endothelial cells and eventually to exit the blood (Fig. 30.13). Selectins (Table 30.4) also contribute to adhesion in other systems, including the initial binding of early mammalian embryos to the wall of the mother’s uterus.
The defining feature of selectins is a calcium-dependent lectin domain (Fig. 30.12) that binds O-linked sulfated oligosaccharides containing sialic acid and fucose. The lectin domain sits at the end of a rod-shaped projection composed of complement regulatory domains that is anchored to the plasma membrane by a single transmembrane sequence. Natural ligands for selectins are mucin-like glycoproteins expressed on endothelial and white blood cells. Specific binding to mucins requires selectins to interact with both the oligosaccharide and mucin protein. The affinity is low (millimolar Kds), and not highly selective among oligosaccharides. Interaction with the mucin protein is less well understood, but one or more sulfated tyrosine residues on the leukocyte mucin called Pselectin glycoprotein ligand (PSGL)-1 participate in binding P-selectin. Bonds between selectins and their mucin ligands have high tensile strength (withstanding forces greater than 100 piconewtons [pN]) but form and dissociate rapidly,
TABLE 30.4 Selectin Family (Lec-CAM) of Cell Adhesion Molecules Examples
Structure
Extracellular Ligands
Me2+
Expression
Functions
E-selectin (CD62E, ELAM-1)
Lectin-EGF-6CR-1TM
L-selectin
Ca
Endothelium (regulated)
WBC-endothelium adhesion
L-selectin (CD62L, gp90M)
Lectin-EGF-2CR-1TM
E-selectin, mucins
Ca
Lymphocytes, other WBCs
WBC-endothelium adhesion
P-selectin (CD62P, GMP-140)
Lectin-EGF-9CR-1TM
Mucins
Ca
Endothelium, platelets
WBC-endothelium adhesion
CD, cellular differentiation antigen; CR, complement regulatory domain; EGF, epidermal growth factor; ELAM-1, endothelial-leukocyte adhesion molecule 1; GMP-140, granule membrane protein 140; gp, glycoprotein; Me2+, divalent cation dependence; TM, transmembrane domain; WBC, white blood cell.
A. Leukocyte
B. Endothelium Gly-CAM-1
L-selectin
CD34
MAdCAM-1 N-linked oligosaccharide O-linked oligosaccharide
PSGL-1
Complement regulatory domains
Lectin domain EGF domain
P-selectin
E-selectin
?
FIGURE 30.12 STRUCTURE OF SELECTINS AND THEIR MUCIN LIGANDS. Domain architecture of selectins and mucins exposed on the surfaces of leukocytes (A) and endothelial cells (B). Complement regulatory domains of the selectins are shown in red. EGF, epidermal growth factor; MAdCAM, mucosal addressin cell adhesion molecule; PSGL, P-selectin glycoprotein ligand. (Modified from Rosen SD, Bertozzi CR. The selectins and their ligands. Curr Opin Cell Biol. 1994;6:663–673.)
CHAPTER 30 n Cellular Adhesion
Selectins Chemoattractants Integrins 1. Attachment 2. Rolling 3. Activation 4. Arrest and adhesion strengthening 5. Transendothelial migration Leukocyte Endothelium Basement membrane
BLOOD TISSUE
Chemoattractant source FIGURE 30.13 MIGRATION OF A NEUTROPHIL FROM THE BLOOD TO THE CONNECTIVE TISSUE. Endothelial cells exposed to inflammatory agents like histamine move selectins to their surfaces and snare mucins on neutrophils flowing in the bloodstream (1). As a neutrophil rolls along the surface (2), chemotactic factors and engagement of mucins activate their integrins (3), causing the neutrophil to bind tightly to immunoglobulin cell adhesion molecules (IgCAMs) on the endothelium (4). The neutrophil then migrates between the endothelial cells into the connective tissue (5). (For reference, see Springer T. Traffic signals for lymphocyte and leukocyte emigration: the multi-step paradigm. Cell. 1994;76:301–314.)
on a second time scale. Low forces on these bonds prolong their lifetimes modestly by altering the conformation of the selectin, whereas high forces promote dissociation. Consequently, few selectin-mucin bonds are required to tether white blood cells to the endothelium, whereas the brief lifetime of the bonds allows blood flow to propel the cells with a rolling motion over the surface of the endothelium (Fig. 30.13). Engagement of selectins with mucins stimulates signals in both cells that activate integrins and promote adhesion. The process resembles T-cell activation and includes Src-family kinases and adapter proteins with tyrosine phosphorylation sites. However, the links from the cytoplasmic domains of selectins and mucins to the signaling proteins is not established. Inflammatory mediators regulate selectins in multiple ways. Activation of endothelial cells with histamine or platelets with thrombin causes vesicles storing P-selectin to fuse with the plasma membrane, exposing the selectins on the cell surface. Various inflammatory agents stimulate endothelial cells to synthesize E-selectin and P-selectin. Activation of white blood cells increases the affinity of L-selectin for mucins and later leads to its proteolytic release from the cell surface.
Mucins The extracellular segments of mucins are rich in serine and threonine, which are heavily modified with acidic oligosaccharide chains (Fig. 30.12). Because of their
537
strong negative charge, these proteins extend like rods up to 50 nm from the cell surface. Mucins on endothelial cells or white blood cells interact with complementary selectins on the other cell type. Endothelial mucin CD34 interacts with white blood cell L-selectin, whereas endothelial P-selectin interacts with white blood cell PSGL-1 mucin. These interactions depend on anchoring of the cytoplasmic domain of PSGL-1 to the actin cytoskeleton. Other mucins are displayed on the surface of or secreted by epithelia lining the respiratory and gastrointestinal tracks.
Other Adhesion Receptors Table 30.5 lists a variety of adhesion receptors that fall outside the five main families. See Fig. 25.6 for the CD45 phosphatase, Fig. 31.3 for claudins and Fig. 31.7 for connexins.
Galactosyltransferase The enzyme galactosyltransferase is also an adhesion receptor. This enzyme is a resident protein in the Golgi apparatus where it glycosylates proteins (see Chapter 21). However, the mRNA for galactosyltransferase has two alternative initiation sites, one of which adds 13 amino acids to the cytoplasmic, N-terminus of this transmembrane protein. The longer enzyme moves to the cell surface rather than being retained in the Golgi apparatus. On the cell surface, the enzyme can bind oligosaccharides that terminate in N-acetylglucosamine, which is found on both cell surface and matrix proteins. The complex of transferase and ligand oligosaccharide is stable, because the galactose-nucleotide substrate added to the oligosaccharide in the Golgi apparatus is not available outside the cell to complete the reaction. During fertilization, a surface galactosyltransferase mediates the initial contact of mouse sperm with the matrix surrounding the egg (called the zona pellucida). This association induces secretion of the contents of the sperm acrosomal vesicle, including an enzyme that destroys the transferase binding site on the matrix so that the sperm can proceed through the zona to fuse with the egg. The enzyme is present on the surface of many cells that migrate during embryogenesis and may contribute to their interactions with the matrix. Adhesion Receptors With Leucine-Rich Repeats (GPIb-IX-V) The platelet receptor for the adhesive glycoprotein called von Willebrand factor (Fig. 30.14) is a disulfidebonded complex of four transmembrane polypeptides: GPIbα, GPIbβ, GPIX, and GPV. Leucine-rich repeats at the end of a long stalk bind von Willebrand factor (see Fig. 28.6 for other receptors with leucine-rich repeats).
538
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
TABLE 30.5 Other Cell Adhesion Molecules Examples
Structure
Extracellular Ligands
Intracellular Ligands
CD44
Link protein-1TM
Hyaluronan
Expression
Functions
Ankyrin
Lymphocytes
Adhesion to endothelium
CD44E
Link protein-HS/ CS-1TM
Fibronectin, hyaluronan
No
Many epithelial cells
Adhesion to matrix
Claudins
Four TM
Claudins
No
ZO-1, ZO-2, ZO-3, cingulin
Epithelia
Tight junctions
Connexins
Multispan, hexamer
Self
No
ZO-1
Epithelia, muscle, nerve
Gap junctions
Dystroglycans
Multisubunit, TM
Laminin, agrin
Ca
Dystrophin
Muscle
Adhesion, synapse formation
Galactosyltransferase
Galactose transferase1TM
N-acetylglucosamine
No
? Actin filaments
Many cells, including sperm
Adhesion to cells and matrix
Glypican
4HS-GPI anchor
Fibronectin,
No
None
Endothelium, smooth muscle, epithelium
Adhesion to matrix
GPIB-IX
7 leucine-rich-1TM
von Willebrand factor
Filamin, actin
Platelets, endothelium
Adhesion
LCA (CD45)
50 kD-1TM-tyrosine phosphatase
WBCs
Tyrosine phosphatase
Mucins (CD34, CD43)
Sialylated oligosaccharide-1TM
Epithelia, leukocytes
Intercellular adhesion
Selectins
Me2+
No
CD, cellular differentiation antigen; CS, chondroitin sulfate; GPI, glycosylphosphatidylinositol; HS, heparan sulfate; LCA, leukocyte common antigen; Me2+, divalent cation dependence; TM, transmembrane domain; WBC, white blood cell.
Platelets bind to von Willebrand factor to initiate the repair of damaged blood vessels. This interaction also generates an intracellular signal that enhances affinity of integrin αIIbβ3 for fibrinogen and reorganizes the cytoskeleton.
Dystroglycan/Sarcoglycan Complex In muscles, a complex of transmembrane glycoproteins links a network of dystrophin and actin filaments on the inside of the plasma membrane to two proteins of the extracellular basal lamina, α2 laminin and agrin (see Fig. 39.17). These protein associations stabilize the muscle plasma membrane from inside and outside, similar to the actin-spectrin network of red blood cells (see Fig. 13.11). Genetic defects or deficiencies in dystrophin, transmembrane linker proteins of the dystroglycan/sarcoglycan complex, or α2-laminin cause muscular dystrophy in humans, most likely owing to the mechanical instability of the membrane, leading to cellular damage and eventual atrophy of the muscle. Chapter 39 provides details on their role in muscle function and disease. Nonmuscle cells in other tissues express many of these proteins (or their homologs), where they may contribute to adhesion to the ECM. Some pathogens use the dystroglycan complex to bind their cellular targets. Arenavirus, the cause of Lassa fever, binds directly to α-dystroglycan, and the leprosy bacterium binds laminin-2.
Examples of Dynamic Adhesion Adhesion of Leukocytes to Endothelial Cells Movement of white blood cells from blood into connective tissue illustrates how cells integrate the activities of selectins, mucins, integrins, IgCAMs, and chemoattractant receptors. Infection or inflammation in connective tissue attracts lymphocytes as well as neutrophils and monocytes, the main phagocytes circulating in blood (see Fig. 28.7). Blood cell precursors use a similar mechanism to enter lymphoid organs and bone marrow. In the absence of inflammation, neutrophils flow over endothelial cells without binding, because the appropriate pairs of adhesion molecules are not exposed or activated. Infection or other inflammation in nearby tissues causes neutrophils to bind to the vascular endothelium and to move out of the blood into the tissue. Neutrophils adhere to the endothelium in three sequential but overlapping steps (Fig. 30.13): 1. Locally generated inflammatory molecules, including histamine (secreted by mast cells), bind to seven-helix receptors on endothelial cells and stimulate fusion of cytoplasmic vesicles (called Weibel-Palade bodies) with the plasma membrane. This exposes P-selectin, formerly stored in the vesicle membranes, on the cell surface facing the blood. Selectins bind mucins that are constitutively exposed on the surface of neutrophils, tethering them to the surface. The bonds form and break rapidly, allowing the neutrophil
CHAPTER 30 n Cellular Adhesion
A
B
C
D
Platelet
Damage exposes basal lamina
Activated platelets secrete ADP
Activated platelets aggregate over defect
Endothelium
Platelet binds
Basal lamina
ADP
539
E
ADP
F. Resting platelet
G. Activated platelet
Fibrinogen Inactive fibrinogen receptor (integrin αIIbβ3)
Three independent stimuli activate platelets
GP1B
(3) ADP activates 7-helix receptors
T
E EL
AT
N
TI
ES
PL
R
M
IU
EL
O
D
C
T
TA
IN
TH
EN
L
SA
(2) Thrombin activates 7-helix receptors
vWF vWF binds to collagen
ADP secretion
A
IN
M LA
BA
Active collagen receptor (integrin α2β1) 7-helix ADPreceptor
Proteins not to scale
Vesicle containing ADP
G
Platelets aggregate when fibrinogen crosslinks
αIIbβ3 integrin (active) binds fibrinogen αIIbβ3 integrin (inactive) Two pathways activate αIIbβ3
GPIb binds to vWF
(1) Integrin α2β1 binds collagen in basal lamina
7-helix thrombin receptor FIGURE 30.14 PLATELET ACTIVATION AND AGGREGATION AT A DEFECT IN THE ENDOTHELIUM. A–D, Steps in platelet activation and aggregation. E, Electron micrograph of a thin section of a platelet adhering to the basal lamina through a tiny defect in the endothelium. F, Resting platelets circulate in the blood without interacting with the intact endothelium lining the vessel. G, Platelets are activated in three ways. (1) Binding of α2β1 integrins to collagen results in firm adhesion. Where the basal lamina is exposed, von Willebrand factor (vWF) binds the collagen. (2) Platelet GPIb-IX binds weakly to von Willebrand factor, allowing platelets to adhere to the exposed matrix. (3) Thrombin activates seven-helix receptors. These interactions stimulate secretion of adenosine diphosphate (ADP), which binds seven-helix receptors and activates the αIIbβ3 integrins; then αIIbβ3 integrins bind dimeric fibrinogen and aggregate platelets together. Platelet proteins are not to scale.
to roll along the surface of the endothelium at rates greater than 10 µm/s as the blood flow pushes them along. 2. Two signaling pathways (seven-helix receptors for chemotactic factors and P-selectin glycoprotein ligand [PSGL]) activate leukocyte integrins from inside the cell (Fig. 30.10). Activation of approximately 10% of the neutrophil integrins increases their affinity for their ligand by 200-fold, making the third step possible. 3. Activated integrins bind tightly to IgCAMs on the surface of endothelial cells, immobilizing the leukocyte despite the force of the blood flow. Within 2 minutes, the leukocyte opens the tight junctions between endothelial cells (see Chapter 31) and squeezes into connective tissue toward the source of the chemoattractant. The leukocyte and endothelial cells interact closely during this passage, because they share a self-associating IgCAM called platelet endothelial cell adhesion molecule (PECAM).
Defects in either the weak or strong interactions compromise the movement of leukocytes into connective tissue, increasing the risk of acute and chronic infections. One type of human leukocyte adhesion deficiency is caused by a genetic defect in fucose metabolism that interferes with the synthesis of a carbohydrate ligand on leukocytes that binds endothelial selectins. Cells cannot roll, so they fail to initiate the emigration process. A genetic deficiency of β2-integrins causes a second type of leukocyte adhesion deficiency. White blood cells that lack β2-integrins roll on the endothelium through the selectin mechanism but do not bind tightly enough to migrate out of the circulation. Consequently, these individuals are susceptible to bacterial infections. On the other hand, neutrophils also generate reactive oxygen species that can damage tissues at sites of inflammation or at sites that are temporarily deprived of oxygen. Thus, movement of white blood cells into tissues contributes to damage that occurs when blood
540
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
flow is restored to an ischemic tissue. Therefore adhesion proteins might be targeted therapeutically to mitigate damage after heart attacks or severe frostbite. A similar mechanism and a partially overlapping set of receptors attract blood monocytes and eosinophils to sites of inflammation. Once they are in connective tissue, interactions of monocyte integrins with matrix molecules trigger the expression of genes required for differentiation into macrophages (see Fig. 28.7). Lymphocytes (see Fig. 28.9) patrol the body, circulating from the blood through organs to lymphoid tissues and through the lymphatic circulation back to the blood. This “recirculation” requires lymphocytes to recognize endothelial cells in organs and specific lymphoid tissues where they exit from the blood. Lymphocytes use L-selectin, three different mucin-like proteins, and α4β2 integrins to bind to these target endothelial cells. Lymphocytes from mice that lack L-selectin do not roll on endothelial cells or accumulate in lymph nodes.
Platelet Activation and Adhesion Platelets aggregate at sites where damage to vascular endothelial cells exposes the underlying basal lamina (Fig. 30.14). This process requires the coordinated activity of a variety of receptors, including integrins, leucine-rich repeat adhesion proteins, and seven-helix receptors. These reactions prevent bleeding and bruising, but inappropriate activation of platelets produces clots in blood vessels, causing heart attacks and strokes. To understand the good effects and avert the bad, investigators have studied platelet activation and adhesion in great detail. Resting platelets have a low tendency to aggregate, even though they circulate in a sea of ligands, including fibrinogen and the adhesive glycoprotein von Willebrand factor. Multiple mechanisms limit the reactivity of resting platelets. First, the major integrin, αIIbβ3, has a low affinity (Kd ≫ µM) for its plasma ligand, fibrinogen. Similarly, the GPIb-IX-V complex has a low affinity for soluble von Willebrand factor. Third, the endothelium masks potential ligands, collagen, and von Willebrand factor in the basal lamina. The concentrations of soluble activators, such as adenosine diphosphate (ADP) and thrombin, are low under physiological conditions. Damage to the endothelium usually initiates platelet activation by exposing platelets to von Willebrand factor and collagen in the basal lamina. Under conditions of high shear, GPIb-IX-V interacts strongly with von Willebrand factor bound to basal lamina collagen. This interaction transiently tethers platelets to the basal lamina and favors binding of integrin α2β1 to collagen. Exposure to soluble agonists such as ADP or thrombin also activates platelets and promotes their aggregation. Within seconds of activation, platelet αIIbβ3 integrins
convert to a high-affinity state (Kd < µM) and bind tightly to fibrinogen. Dimeric fibrinogen links platelets into aggregates. Agonists activate platelet αIIbβ3 integrins through three different pathways: 1. Collagen binding to α2β1 integrin directly stimulates platelets to activate αIIbβ3 integrins, secrete ADP, and synthesize the lipid second messenger thromboxane A2 (see Fig. 26.9). 2. Damage to blood vessels activates the blood-clotting proteolytic enzyme thrombin, which binds two related seven-helix receptors and signals through trimeric G-proteins (see Fig. 25.9) to activate integrin αIIbβ3. 3. von Willebrand factor binding to the platelet receptor GPIb-IX-V activates αIIbβ3 integrins. Two additional mechanisms augment all these responses. Activated platelets secrete ADP, which binds two types of seven-helix receptors that amplify the response to thrombin. Aggregation of platelets by binding dimeric fibrinogen further stimulates their response to ADP and thrombin. Platelet aggregation is disadvantageous in the normal circulation, so several mechanisms actively inhibit platelet activation. Endothelial cells produce both nitric oxide and an eicosanoid, prostacyclin (PGI2), which inhibit platelet activation (see Fig. 26.9). Nitric oxide acts through cyclic guanosine monophosphate (cGMP), and PGI2 acts through cyclic adenosine monophosphate (cAMP; see Fig. 26.1). Antibodies and small molecule drugs that inhibit αIIbβ3 are used to treat heart attacks. The most common human bleeding disorder is von Willebrand disease, caused by mutations in von Willebrand factor or its receptor, the GPIbα subunit of GPIb-IX-V. Some mutations reduce the concentration of the factor in blood or reduce the affinity of the factor for its receptor. Remarkably, mutations in either the factor or receptor that increase their affinity for each other also cause bleeding. These high-affinity interactions cause platelets to aggregate and be removed from the blood. Loss-of-function mutations in GPIbα cause the human bleeding disorder called Bernard-Soulier syndrome. Individuals with Glanzmann thrombasthenia bleed abnormally because αIIbβ3 integrin is absent or defective, and their platelets do not aggregate.
Self-Avoidance in the Nervous System Surprisingly, neurons use adhesion proteins to avoid forming synapses with themselves by repelling axons and dendrites from the same cell (Fig. 30.15). Insects use a family of DSCAM1 IgCAMs for this self-avoidance. Alternative splicing of the pre-mRNA in each cell generates a subset of the many thousands of different DSCAM1 isoforms. These large IgCAMs have 10 Igdomains and six FN-III domains that make homophilic interactions between three Ig-domains of each protein.
CHAPTER 30 n Cellular Adhesion
541
N-cadherins and other adhesion proteins to specify synaptic connections. A point mutation in one protocadherin gene is a common cause of human deafness and blindness.
SELECTED READINGS Axon Cell body
100 µm
FIGURE 30.15 SELF-AVOIDANCE OF NEURITES OF CLASS IV SENSORY NEURONS OF A DROSOPHILA LARVA. Stack of fluorescence micrographs ~20 µm thick of neurons expressing a mouse transmembrane protein called CD8 tagged with GFP. False coloring is used to distinguish the cells. The central neuron is red. (Courtesy Sujoy Ganguly and Jonathan Howard, Yale University, New Haven, CT.)
Contacts between neurites with the same isoforms cause repulsion. Self-avoidance of vertebrate neurons depends on a family of 48 cadherins called protocadherins. Alternative splicing of transcripts from three gene clusters produces a unique mixture of protocadherins on each neuron. These protocadherins form random dimers in the membrane and the extracellular domains bind homophilically to like protocadherins. When a process from a neuron contacts another part of itself, extensive interactions between like protocadherins somehow create a signal that repels the neurite. Weaker interactions between mixtures of protocadherins on different cells do not generate a repulsive signal and allow
Campbell ID, Humphries MJ. Integrin structure, activation, and interactions. Cold Spring Harb Perspect Biol. 2011;3:a004994. Case LB, Waterman CM. Integration of actin dynamics and cell adhesion by a three-dimensional, mechanosensitive molecular clutch. Nat Cell Biol. 2015;17:955-963. Chen WV, Maniatis T. Clustered protocadherins. Development. 2013;140:3297-3302. Constantin B. Dystrophin complex functions as a scaffold for signalling proteins. Biochim Biophys Acta. 2014;1838:635-642. Gérard C, Goldbeter A. Dynamics of the mammalian cell cycle in physiological and pathological conditions. Wiley Interdiscip Rev Syst Biol Med. 2016;8:140-156. Gumbiner BM. Regulation of cadherin-mediated adhesion in morphogenesis. Nat Rev Mol Cell Biol. 2005;6:622-634. Harwood A, Coates JC. A prehistory of cell adhesion. Curr Opin Cell Biol. 2004;16:470-476. Hernández AR, Klein AM, Kirschner MW. Kinetic responses of β-catenin specify the sites of Wnt control. Science. 2012;338: 1337-1340. Livne A, Geiger B. The inner workings of stress fibers—from contractile machinery to focal adhesions and back. J Cell Sci. 2016;129: 1293-1304. Logan CY, Nusse R. The Wnt signaling pathway in development and disease. Annu Rev Cell Dev Biol. 2004;20:781-810. McEver RP. Selectins: initiators of leucocyte adhesion and signalling at the vascular wall. Cardiovasc Res. 2015;107:331-339. McEver RP, Zhu C. Rolling cell adhesion. Annu Rev Cell Dev Biol. 2010;26:363-396. Roycroft A, Mayor R. Molecular basis of contact inhibition of locomotion. Cell Mol Life Sci. 2016;73:1119-1130. Rosen SD. Ligands for L-selectin: Homing, inflammation, and beyond. Annu Rev Immunol. 2004;22:129-156. Samanta D, Almo SC. Nectin family of cell-adhesion molecules: structural and molecular aspects of function and specificity. Cell Mol Life Sci. 2015;72:645-658. Shattil SJ, Kim C, Ginsberg MH. The final steps of integrin activation: the end game. Nat Rev Mol Cell Biol. 2010;11:288-300. Yu FX, Zhao B, Guan KL. Hippo Pathway in Organ Size Control, Tissue Homeostasis, and Cancer. Cell. 2015;163:811-828. Zipursky SL, Grueber WB. The molecular basis of self-avoidance. Annu Rev Neurosci. 2013;36:547-568.
This page intentionally left blank
CHAPTER
31
Intercellular Junctions T
he mechanical integrity of animal tissues such as epithelia, nerves, and muscles depends on the ability of the cells to interact with each other and the extracellular matrix. Plasma membrane specializations, called cellular junctions, mediate these interactions. Physical connec tions from the extracellular matrix or adjacent cells through these junctions and the associated cytoskeletal filaments inside cells impart mechanical strength to tissues. Investigation of junctions began when microscopists and physiologists recognized that epithelial and muscle cells adhere to each other and the underlying extracel lular matrix. They also discovered that some epithelia form a tight barrier between the luminal surface and the underlying tissue spaces. The physical basis of these interactions became clear during the 1960s, when elec tron micrographs of thin sections of vertebrate tissues revealed four types of intercellular junctions that connect the plasma membranes of adjacent cells (Table 31.1 and Fig. 31.1) and two types of junctions to bind to the extracellular matrix. Subsequent research established the molecular architecture of these junctions, each based on a different transmembrane protein: Adherens junctions: Transmembrane proteins called cadherins (see Fig. 30.5) link neighboring cells and connect to actin filaments in the cytoplasm. Desmosomes: Another type of cadherin links cells together and connects to cytoplasmic intermediate filaments. Tight junctions: Transmembrane proteins called claudins join the plasma membranes of two cells to create a barrier that limits diffusion of ions and solutes between the cells and molecules between apical and basolateral domains of the plasma membrane. Gap junctions: Transmembrane proteins called con nexins form channels for small molecules to move between the cytoplasms of neighboring cells.
Hemidesmosomes: Integrins (see Fig. 30.9) connect cytoplasmic intermediate filaments to the basal lamina across the plasma membrane. Focal adhesions: Integrins associated with cytoplasmic actin filaments adhere to the extracellular matrix. Each tissue uses a selection of junctions suited to its physiological functions. Columnar epithelial cells in the intestine interact with their neighbors using all four types of intercellular junctions (Fig. 31.1B–D). Belt-like tight junctions and adherens junctions encircle the apex of the cell. Desmosomes and gap junctions form patchlike lateral connections between the cells. Hemidesmo somes anchor the cells to the basal lamina. Stratified epithelial cells in the skin (Fig. 31.1A) use desmosomes and intermediate filaments (Fig. 31.1B) to resist mechani cal forces but also interact via claudins and adherens junctions. Desmosomes and adherens junctions link muscle cells to the surrounding basal lamina (see Fig. 29.17C). Gap junctions connect heart and smooth muscle cells, but not skeletal muscle cells. Most nerve cells communicate chemically, but some use gap junc tions for electrical communication. Invertebrate animals assemble junctions from homolo gous proteins but with different organization than the junctional complex of vertebrate epithelia. Insect epithe lia have apical adherens junctions and more basal “septate junctions” built from claudins and cytoplasmic proteins with sequence homology to the tight junction proteins ZO-1 and ZO-2. Nematode epithelia have one type of junction with adherens functions and claudins.
Tight Junctions Tight junctions form a belt-like adhesive seal that selec tively limits the diffusion of water, ions, and larger solutes between epithelial cells (Fig. 31.2). This allows epithelia to separate the interior of the body from the 543
544
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
TABLE 31.1 Molecular Components of Cell-Cell and Cell-Matrix Junctions Junction
Target Molecule
Adhesive Protein
Cytoplasmic Proteins
Cytoskeletal Filaments
Claudin
ZO-1, ZO-2, cingulin, spectrin
Actin
Connexin
ZO-1, drebrin
Actin
Sealing of the Extracellular Space Tight junction
Claudin
Communication Between Cells Gap junction
Connexin
Adhesion to Other Cells Zonula adherens
Cadherin
Cadherin
Catenins, plakoglobin
Actin
Desmosome
Desmoglein Desmocollin
Desmoglein Desmocollin
Plakoglobin, desmoplakin
Intermediate
Adhesion to the Extracellular Matrix Hemidesmosome
Laminin
Integrin
Plectin, BP 180
Intermediate
Focal contact
Fibronectin
Integrin
Talin, vinculin, α-actinin
Actin
A. Skin
B. Epithelium
C. Epithelium
Desmosome
Zonula occludens (tight junction)
D. Epithelium
Zonula occludens (tight junction)
Zonula adherens
Zonula adherens Macula adherens (desmosome) Gap junction
Macula adherens (desmosome)
Hemidesmosome Basal lamina
Gap junction
FIGURE 31.1 LIGHT AND ELECTRON MICROGRAPHS OF JUNCTIONS. A, Desmosomes. Left, Light micrograph of a section of skin showing numerous desmosomes as pink dots between the cells. Right, Electron micrograph of a thin section of skin showing desmosomes. B, Light micrograph of a section of intestinal epithelium stained with hematoxylin and eosin, showing the junctional complex (also called “terminal bars”) as bright pink dots between the cells near their apex, just below the microvilli of the brush border. C, Electron micrograph of a thin section of intestinal epithelial cells, showing the junctional complex consisting of a belt-like tight junction (also called the zonula occludens), a belt-like adherens junction (also called the zonula adherens), and desmosomes (also called the macula adherens), all in their characteristic relation to each other. The circumferential tight junction seals the extracellular space. The zonula adherens is anchored to the actin cytoskeleton. Desmosomes are attached to cytoplasmic intermediate filaments. D, Drawing showing the position of the junctional complex in the cell and the locations of gap junctions, basal lamina, and hemidesmosomes. (A, Courtesy Don W. Fawcett, Harvard Medical School, Boston, MA. C, Courtesy Marilyn Farquhar, University of California, San Diego.)
CHAPTER 31 n Intercellular Junctions
external world. Tight junctions also define the boundary between the biochemically distinct apical and basolateral domains of the plasma membrane of polarized epithelial cells. Tight junctions were first recognized in electron micrographs of thin sections as places where the plasma membranes of adjacent cells appear to fuse together in one or more contacts (Fig. 31.2). Freeze-fracture images revealed that these contacts correspond to continuous strands of intramembranous particles that form a branch ing network in the plane of the lipid bilayer.
A
B
C
Plasma membrane
Strands of claudins Intercellular space
FIGURE 31.2 EPITHELIAL TIGHT JUNCTIONS. A, Electron micrograph of a thin section of endothelial cells, showing a point of contact between the plasma membranes at a tight junction (arrow). B, Electron micrograph of a replica of a freeze-fractured cell. This method exposes proteins within the lipid bilayer and reveals strands aligned along the points of contact between the plasma membranes. C, Drawing showing the strands of transmembrane proteins at points of contact. (A, Courtesy George Palade, University of California, San Diego. B, Courtesy Don W. Fawcett, Harvard Medical School, Boston, MA.)
A
N
545
Transmembrane proteins forming the strands observed by freeze-fracture were difficult to identify until investi gators found a monoclonal antibody that bound to the cytoplasmic side of the plasma membrane at tight junc tions. They used this antibody to isolate an integral membrane protein and named it occludin. The amino acid sequence of occludin suggested four transmem brane strands and two hydrophobic extracellular loops. However, mice lacking their single occludin gene survive with normal tight junctions. Subsequently, these scien tists discovered claudins, the main structural proteins of tight junction strands (Fig. 31.3A). Humans have a family of 27 homologous claudin genes that are expressed selectively in various tissues. Claudins consist of four highly conserved transmem brane helices with two extracellular loops that fold into a small β-sheet and single helix. Close associations between claudins form barriers to diffusion in the plane of the membrane and between the cells. • The barrier in the membrane bilayer: Intimate lateral interactions of claudins within the lipid bilayer (Fig. 31.3B) block diffusion of lipids and proteins in the plane of the membrane. This barrier separates differ ent pumps, carriers, receptors and lipids in the apical and basolateral domains of the plasma membrane. • The barrier between cells: The small extracellular domains of claudins interact with their neighbors in intramembrane strands and with claudins in similar strands on adjacent cells to make barriers interrupted by rows of pores. These pores block diffusion of solutes larger than approximately 1 nm in diameter, but selectively allow the passage of small cations or anions. Most tight junctions are more permeable to cations than to anions. Permeability in the two direc tions across the junction is identical.
C
Pores ZO-2
Claudin
ZO-1
B
90°
Actin C
FIGURE 31.3 STRUCTURE OF TIGHT JUNCTIONS. A, Ribbon diagram of the transmembrane domains of claudin-15. B, Interactions between the molecules in crystals of claudin-15 that are thought to represent contacts in tight junctions. C, Model of tight junction structure with claudins linking the two membranes together and peripheral protein ZO-1 linking the cytoplasmic tail of claudins to actin filaments. (For reference, see Protein Data Bank [PDB; www.rcsb.org] file 4P70 and Suzuki H, Nishizawa T, Tani K, et al. Crystal structure of a claudin provides insight into the architecture of tight junctions. Science. 2014;344:304–307.)
546
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
Adapter proteins with PDZ protein-interaction domains (see Fig. 25.10) link tight junctions to the cytoskeleton. These adapters are called ZO-1, ZO-2, and ZO-3. They interact with the long C-terminal cytoplasmic tail of claudin and JAM (junctional adhesion molecule) in the membrane. In the cytoplasm they interact with actin fila ments, a small guanosine triphosphatase (GTPase) and other proteins that regulate actin polymerization and the adapter protein cingulin. ZO-2 and cingulin are specific for tight junctions, but ZO-1 also associates with cadherins in adherens junctions and connexins in gap junctions. These two barrier functions of tight junctions set up the two conditions required for many physiological pro cesses. The actions of pumps, carriers, and channels located selectively in the apical or basolateral domains of the plasma membrane allow polarized cells to create dif ferent extracellular environments on the two sides of the epithelium. Maintaining these environments depends on the selective permeability across the epithelium through the extracellular pores of tight junctions. For example, tight junctions are essential for intestinal epithelial cells to take up nutrients from the lumen of the intestine and transport them into the extracellular space beneath the cells (see Fig. 17.2) and for other physiological processes (see Figs. 17.3 and 17-4). Restricting the diffusion of macromolecules across epithelia can regulate some types of signaling. For instance, airway epithelial cells release the growth factor heregulin from the apical surface, but restrict its receptor tyrosine kinase erbB2 (see Fig. 24.5B for a related receptor) to the basolateral surface. Thus, the receptor is activated only if the epithelium is damaged or the tight junctions compromised. Sheets of epithelial cells vary by several orders of magnitude in the quality of the seal created by circum ferential tight junctions between adjacent cells. The tightness of the barrier to diffusion of ions in the extra cellular space determines the electrical resistance across the epithelium and depends on the claudin isoform and the number and continuity of the strands. Epithelia also appear to differ in the ability of water to flow through tight junctions. Extremely tight barriers with many strands and distinct claudin forms are found where epi thelia must maintain high ion gradients, such as in the distal tubules of the kidney, where urine is concentrated. Leaky tight junctions with fewer strands and different claudins are found where ion gradients across epithelia are small but a barrier is required for large solutes, pro teins, and leukocytes (eg, in most blood vessels). Intracellular and extracellular factors regulate the transepithelial barrier established by tight junctions. These regulators include hormones such as vasopressin and aldosterone and cytokines such as tumor necrosis factor (see Fig. 24.9) that act through second messengers (eg, Ca2+ and cyclic adenosine monophosphate [cAMP]; see Chapter 26), and effectors (eg, protein kinases A and C; see Chapter 25). The mechanisms are not yet
well understood, but posttranslational modifications of tight junctions might modulate their assembly. Tension on associated actin filaments may physically open pas sages through tight junctions. The metabolic state of the cell also influences tight junctions; depletion of adenosine triphosphate (ATP) causes tight junctions to leak without destroying the barrier between the apical and basolateral domains of the plasma membrane. White blood cells migrating across epithelia from the blood to the connective tissue, open tight junctions locally without disrupting the tight seal across the epithelium (see Fig. 30.13). A localized increase in cytoplasmic Ca2+ in the epithelial cells is required to open the tight junctions. Several bacterial toxins affect the tight junction barrier. The ZO-toxin of Vibrio cholerae induces diar rhea by loosening tight junctions, independent of the classic cholera toxin, which induces secretion of salt and water. Helicobacter pylori injects a protein toxin into the cells lining the stomach. This toxin disrupts tight junctions, breaking the barrier that protects the underly ing tissues and predisposing to ulcers. Mutations of human claudin genes cause highly selec tive defects in epithelial barriers. One example is reduced ability of the kidney to reabsorb potassium (claudin-16). Another is deafness due to loss of ion gradients in the inner ear (claudin-14).
Gap Junctions The idea that channels might couple cells arose relatively late, because electrophysiological experiments on nerves and skeletal muscles reinforced the widespread belief that cells were autonomous. However, nerve and skeletal muscle cells later turned out to be exceptions to the general principle that cells in animal tissues communi cate with each other by gap junctions. Cells in plant tissues also communicate with each other, but they use direct cytoplasmic connections, called plasmodesmata, rather than gap junctions (Box 31.1 and Fig. 31.4). The first convincing evidence for direct electrical communication between cells came around 1960 from electrophysiological experiments on the synapses between giant axons and the motor neurons that drive the flipper muscles of crayfish. These electrical synapses transmit action potentials (see Fig. 17.6) directly from one cell to the next without the delay required for secretion and reception of a chemical transmitter (see Fig. 17.10). Similar electrical junctions were subsequently found to connect heart muscle cells (see Fig. 39.18). Over the next decade, physiologists used microelec trodes to establish that plasma membrane depolarization of one cell can be transmitted with little resistance to adjacent epithelial cells (Fig. 31.5B), although the ampli tude of the response declined with distance. Similarly, fluorescent molecules, radioactive tracers, and essential
547
CHAPTER 31 n Intercellular Junctions
A
Integral membrane proteins of the ER and plasma membrane
Annulus Desmotubule
ENDOPLASMIC RETICULUM
CELL WALL 100 nm FIGURE 31.4 A PLASMODESMATA CONNECTING TWO PLANT CELLS. The plasma membrane is continuous between the two cells. The space between the tubule of endoplasmic reticulum and the surrounding plasma membrane allows molecules in the cytoplasm to move between the two cells. ER, endoplasmic reticulum.
1
2
3x
4
5 1
2
3
4
5
Halothane
B I1
100 pA
V1
I2
V2
C
10 sec
10 mV
Most cells in plant tissues maintain cytoplasmic continuity with their neighbors through plasmodesmata, channels across the cell wall lined by plasma membrane across the cell wall (Fig. 31.4). These connections form by incom plete cytokinesis, but secondary plasmodesmata can form independently. Plasmodesmata are essential for plant viability. A narrow tubule of modified endoplasmic reticulum (ER) fills most of the pore and is linked to the surround ing plasma membrane by a number of proteins. The protein spokes connecting the ER and plasma membrane are not yet well characterized, but candidates include proteins that participate in membrane contact sites in animals and fungi such as synaptotagmins (see Chapter 21), synaptobrevin (see Chapter 21), lipid exchange proteins (see Fig. 20.17), junctophilins, and stromal interaction molecules (STIMs) (see Fig. 26.12). Molecules smaller than about 1 kD diffuse freely through the narrow cylinder of cytoplasm in plasmodes mata, but larger molecules, even nucleic acids pass selectively through these channels. Constitutive diffusion of small molecules allows exchange of metabolites and hormones between cells. Regulated passage of larger molecules, including double-stranded RNAs and proteins such as transcription factors, allows developmental signals to move between cells and tissues. Specialized viral “movement proteins” allow some viruses to move their whole genomes between cells. Permeability varies among tissues and with physiologi cal states and developmental stages. For example, all cells in plant embryos are connected, whereas cells in some adult tissues are isolated. Reversible deposition of callose and other molecules in the cell wall surrounding plasmo desmata regulates their diameter and permeability. Actin filaments contribute to regulation of the pore size, but the mechanism and signals controlling permeability are not known.
Fluorescein
open close
I1
10 pA
BOX 31.1 Plasmodesmata
I2 2 sec
FIGURE 31.5 GAP JUNCTION PHYSIOLOGY. A, Drawing and fluorescence micrograph, showing the movement of a tracer dye between epithelial cells from the salivary gland of Chironomus. Cell 3 was injected with fluorescein (molecular weight: 330), which spread to adjacent cells via gap junctions. B–C, Electrical recordings from pairs of cells coupled by gap junctions. B, Two cells (1 and 2) were voltageclamped (see the text that describes Fig. 17.6) and subjected alternately to small depolarizing voltage changes (V1, V2). Being electrically coupled, they responded with opposite currents (I1, I2). Transient exposure to the anesthetic halothane (horizontal bar) closes most of the channels, reducing the current in response to depolarization. C, When the cells are held at a constant depolarizing voltage in the presence of halothane, current records reveal the opening and closing of individual gap junction channels as opposite step changes in current. (A, From Lowenstein W. Junctional intercellular communication: the cell-to-cell membrane channel. Physiol Rev. 1981;61:829. B and C, From Eghbali B, Kessler JA, Spray DC. Expression of gap junction channels in communication-incompetent cells. Proc Natl Acad Sci U S A. 1990;87:1328–1331.)
nutrients can pass from the cytoplasm of one cell to the cytoplasm of neighboring cells. Electron microscopy revealed that low-resistance communication between cells is associated with the presence of plasma membrane specializations that were called gap junctions owing to the regular 2- to 4-nm separation of the adjacent cell membranes (Fig. 31.6). Gap junctions are plaques composed of large intercellular channels that connect the cytoplasms of a pair of cells. These plaques exclude other transmembrane pro teins and contain a few to thousands of channels. Half
548
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
A
B
C
D
FIGURE 31.6 LIGHT AND ELECTRON MICROGRAPHS OF GAP JUNCTIONS. A, Thin section of embedded cells, showing the closely apposed membranes of adjacent cells separated by a gap of 2 nm. B, Replica of a freeze-fractured cell, showing an irregular array of particles exposed in the plane of the lipid bilayer. C, Fluorescence micrograph of a gap junction plaque (red and green) between cultured HeLa (Henrietta Lacks) cells expressing connexin-43 with a tetracysteine peptide tag. The cells were first exposed to a green fluorescent dye that binds tightly to the tetracysteine tag and then, after 4 hours of growth without the green dye, the same cells were incubated with a second red fluorescent dye that binds to the tetracysteine tag on newly synthesized connexin-43. The older central part of this plaque is green. The newer peripheral regions of the plaque are red. D, Negative staining of an isolated gap junction reveals the intercellular connexon channels packed together in a regular, two-dimensional array. Each connexon has a central channel filled with stain. (A–B and D, Courtesy Don W. Fawcett, Harvard Medical School, Boston, MA; from the work of N.B. Gilula, Scripps Research Institute, La Jolla, CA. C, Courtesy Mark Ellisman, University of California, San Diego and from Gaietta G, Deernick TJ, Adams SR, et al: Multicolor and electron microscopic imaging of connexin trafficking. Science. 2002;296:503–507.)
channels in each membrane called connexons are each formed from six protein subunits, named connexins (Fig. 31.7).
Structure of Gap Junction Channels A hexagonal ring of connexins forms a central aqueous channel across the lipid bilayer and pairs with a con nexon in an adjacent cell to connect their cytoplasms. Four transmembrane α-helices of each connexin subunit span the lipid bilayer and two loops between the helices form small extracellular domains (Fig. 31.7). Connexons on adjacent cells dock to form a continuous pore between the pair of cells. Tight interactions between the subunits seal the pore and preclude leakage of ions out of either cell. The transmembrane pore begins with a funnel from the cytoplasm that narrows to a diameter of 1.4 nm, larger than the pores of tetrameric S5-P-S6 ion channels (see Fig. 16.5) or pentameric ligand-gated ion channels (see Fig. 16.12). This pore passes hydrophilic molecules up to approximately 1 kD in size, including ions (to establish electrochemical continuity between the cells), second messengers (to establish a common network of information), small peptides, and metabolites (to allow sharing of resources). Connexon hemichannels (the ring of six connexins in one plasma membrane) also open infrequently for the nonspecific release of ions and solutes as large as ATP from the cell.
Connexin Gene Families and Evolution Humans have genes for 21 connexin isoforms, ranging in size from 26 to 60 kD. Connexins are named by molecular weight; for instance, connexin-43 (Cx-43) is the name for the 43-kD isoform. All connexins have the conserved features required to form the connexon channel but variable N- and C-terminal cytoplasmic sequences. The various connexin isoforms make chan nels that differ in their permeability and charge selectivity. Remarkably, gap junction genes seem to have arisen more than once during evolution. Connexins are found exclusively in chordates, while invertebrate gap junc tions are composed of innexins (invertebrate connex ins). Innexins have four transmembrane helices but lack any sequence similarity to connexins and form intercellular junctions from two octameric hemichan nels. Vertebrates have a few genes related to innexins called pannexins. Rather than forming gap junctions, pannexins form plasma membrane channels that release solutes including ATP from the cytoplasm. The ATP activates seven-helix receptors (see Fig. 24.2) and chan nels activated by adenine nucleotides (see Fig. 16.2) as part of local signaling pathways in the immune and nervous systems. Many animals have another channel, calcium homeostasis modulator 1, that shares structural features with connexins, pannexins, and innexins.
CHAPTER 31 n Intercellular Junctions
A
B
D H1
Conserved regions
H2
N
EXTRA– CELLULAR
INTRA– CELLULAR
38 – 56 aa
H3 18 – 195 aa H4
C
CYTOPLASM
EXTRACELLULAR
C
INTRA– CELLULAR
549
N-terminal helix plug
E
Open
Negatively charged patches
N-terminal helix plug
Closed
FIGURE 31.7 STRUCTURE OF THE GAP JUNCTION CONNEXON CHANNEL. A, Drawing of gap junction connexons forming channels between the cytoplasms of adjacent cells. B, Transmembrane topology of connexins. Four α-helices cross the lipid bilayer. Conserved residues (maroon) form the transmembrane and extracellular loops are required for channel assembly. Cytoplasmic loops between helix 2 and helix 3 and the C-terminal tails vary in length among connexin isoforms. Removal of the C-terminal tail from connexin-43 alters its gating properties. C, Diagram showing how the N-terminal α-helices of Cx26 may form a plug that blocks the pore of the closed channel. D–E, Crystal structure of the connexin 26 the gap junction channel. C–D, Top and side views of a ribbon diagram with each of the six subunits a different color. The two extracellular loops of each subunit associate with the other half channel to span the 4-nm gap between the membranes. The upper panel of D shows as a space-filling model cut through the middle of the transmembrane pore. (For reference, see PDB file 2ZW3 and Maeda S, Nakagawa S, Suga M, et al. Structure of the connexin 26 gap junction channel at 3.5 A resolution. Nature. 2009;458:597–602.)
However, the lack of sequence homology suggests that this gene may have arisen separately.
Assembly of Gap Junctions Connexin proteins turn over on a time scale of several hours and are replaced by new connexons that assemble in vesicles along the secretory pathway. New connexons add around the periphery of gap junction plaques in the plasma membrane and old connexons are removed from the middle of plaques (Fig. 31.6C). Many gap junctions are composed of one connexin isoform and pass molecules equally well in both direc tions. However, some connexons assemble from mix tures of subunits creating hybrid gap junctions with novel properties. Furthermore, some connexons can pair with a different type of connexon on the neighbor ing cell. Such hybrid channels may pass fluorescent tracers more readily in one direction than the other or react more sensitively to the transjunctional potential of one polarity than the other. This might explain the asym metrical coupling that is sometimes observed between both excitable and nonexcitable cells, such as neuronal gap junctions, which pass action potentials in one direc tion but not the other.
Regulation of Gap Junction Permeability Electrophysiological measurements showed that con nexons alternate between open and closed states (Fig. 31.5C). The structural basis for this gating is not estab lished definitively, but a plug-gating mechanism is plau sible. In this model N-terminal helices can either form a plug that blocks the pore or pull back against the wall of the channel to create the 1.4-nm pore (Fig. 31.7E). This would allow the two connexins to gate the pore independently; both must be open to connect the two cytoplasms. Cytoplasmic loops, C-terminal tails or the extracellular loops may also contribute to gating. The pores of gap junction channels are large enough to pass all common inorganic ions as well as nucleotides, amino acids, and even small peptides and RNAs. The conductance of the open state depends on the connexin isoform and varies from about 30 psec to 300 psec. Given the permeability of gap junctions to relatively large solutes, it is surprising that their conductance is in the same range as narrower ligand- and voltage-gated ion channels. Both the greater length and the arrangement of charged residues lining the channel may contribute to the low conductance of connexons.
550
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
Gap junctional communication depends on both the number of channels and the fraction of those channels that is open or closed. The fraction of open channels is usually less than 1.0; it is approximately 0.2 in heart and as low as 0.01 in one nerve cell that was tested. Many factors regulate the opening and closing of con nexon channels. The transjunctional potential (ie, the potential difference between the coupled cells) gates most connexons, regardless of the plasma membrane potentials of these cells. Like other voltage-gated chan nels, individual transitions are fast, but the response to potential changes, on the scale of seconds, is very slow in comparison with other channels (see Fig. 16.7). Unphysiological concentrations of cytoplasmic Ca2+ (100 to 500 µM) and cytoplasmic acidification also close con nexons. These effects of membrane potential, H+, and Ca2+ allow cells to terminate communication with neigh boring cells that are damaged (depolarizing the plasma membrane and admitting high concentrations of Ca2+) or metabolically compromised (allowing Ca2+ to leak out of intracellular stores and acidifying the cytoplasm). Chemicals also modulate gap junctions. Oleamide, a fatty acid amide produced by the brain, blocks gap junctional communication and induces sleep in animals. Organic alcohols (heptanol and octanol) and general anesthetics (halothane) can also close gap junction chan nels reversibly (Fig. 31.5), but these agents are not spe cific for gap junctions. Signaling pathways control gap junction activity through phosphorylation of numerous sites by several kinases. For example, on a time scale of seconds, cAMP activates protein kinase A, which phosphorylates the C-terminal tail of some connexins, increasing or decreas ing the fraction of open channels (depending on the connexin isoform and the cell type). On a time scale of hours cAMP also promotes the assembly of gap junctions.
Physiological Functions of Gap Junctions Cells in most metazoans communicate by gap junctions. Coupled cells in vertebrates include epithelial cells of the skin, endocrine glands, exocrine glands, gastrointes tinal tract, and renal-urinary tract as well as smooth muscle, cardiac muscle, bone, some neurons, and glial cells. Epithelial cells can coordinate their activities with their neighbors. This is used to synchronize the beats of cilia (see Fig. 38.12C). Fragments of viral proteins can spread from infected cells to neighboring cells, which then become targets for cytotoxic T lymphocytes (see Fig. 28.9). Gap junctions allow osteocytes buried deep in bone to maintain a cellular supply line to acquire nutrients from distant blood vessels (see Fig. 32.4). Passage of action potentials between cardiac and smooth muscle cells sets off waves of contraction (see Fig. 39.18). Electrical synapses between neurons can trans mit action potentials at very high frequencies (>1000
TABLE 31.2 Phenotypes of Humans With Mutations in Gap Junction Subunits Connexin
Phenotype
Cx-26β2
Dominant and recessive mutations with deafness; skin disease
Cx-30β6
Recessive deafness; skin disease
Cx-31β3
Recessive deafness; skin disease
Cx-32β1
Point mutations, defective myelin, peripheral nerve degeneration in X-linked Charcot-MarieTooth disease; deafness
Cx-37α4
Female infertility, defect in communication of granulosa cells with oocyte
Cx-40α5
Partial block of impulse conduction in heart
Cx-43α1
Deafness; many mutations may be lethal as in mice
Cx-46α3
Cataracts in lens of the eye
Cx-50α8
Cataracts in lens of the eye
Note: Mutations are homozygous loss of function mutations unless noted otherwise. The nomenclature used here combines the Cx-“molecular mass in kD” and molecular phylogeny αβ-number systems.
per second). In some parts of the brain, gap junctions also coordinate action potentials in groups of neurons. Even white blood cells may form transient gap junctions with endothelial cells.
Gap Junctions in Disease Point mutations in connexin genes cause remarkably specific defects in humans (Table 31.2), considering that most connexins are expressed in several tissues. Reces sive mutations in the connexin-26 gene are the most common causes of inherited human deafness. As many as 1 in 30 people are carriers, and their mutations may contribute to hearing loss late in life. Connexin-26 par ticipates in the transport of K+ in the epithelia supporting the sensory hair cells in the ear. Patients with one of more than 100 different mutations in the connexin-32 gene can suffer from degeneration of the myelin sheath around axons, an X-linked variant of CharcotMarie-Tooth disease. Many human tissues express connexin-32, but the pathology is confined to myelin. The stability of myelin might depend on intracellular gap junctions between layers of the myelin sheath that provide a pathway between the metabolically active cell body and the deep layers of the sheath near the axon. Defects in myelin membrane proteins cause other forms of Charcot-Marie-Tooth disease.
Adherens Junctions Adherens junctions use homophilic (like-to-like) interac tions of E-cadherins (see Fig. 30.5) to bind epithelial cells to their neighbors. Adherens junctions are essential for viability from the earliest stages of animal embryonic development. In mature epithelia, a belt-like adherens junction, called the zonula adherens, encircles
the cells near their apical surface (Fig. 31.1D) and maintains the physical integrity of the epithelium. Adherens junctions also anchor muscle cells to the extracellular matrix. Adherens junctions can transmit mechanical forces between cells and reinforce tissues, because the cyto plasmic domains of the E-cadherins are linked to the actin cytoskeleton. Adapter proteins connect cadherins to actin filaments and signaling proteins including guanine nucleotide exchange proteins for the Rho-family GTPases that promote actin assembly and force genera tion by myosin (see Fig. 33.19F). The adapter proteins include β-catenin, a related protein called plakoglobin, p120-catenin, and the actin-binding protein α-catenin. Moderate physical forces stabilize this link from the cadherin tail through β-catenin and α-catenin to actin filaments. Adherens junctions are the first connections estab lished within developing sheets of epithelial cells. Contact begins when cadherins on the tips of filopodia engage partner cadherins of the same type on another cell. The contact spreads laterally as more cadherins are recruited along with associated actin filaments, as illus trated by dorsal closure of the ectoderm by Drosophila embryos (see Fig. 38.5). These pioneering adherens junctions eventually allow like cells to associate in epithelial sheets (see Fig. 30.8) and to influence the maturation of the epithelium. Adherens junctions are a prerequisite for the assembly of tight junctions that allow epithelial cells to establish polarity with different proteins and lipids in the apical and basal plasma mem branes. The shapes of cells in epithelial sheets depend on Rho family GTPases and protein kinases associated with the adherens junction, which regulate the assembly and contraction of the associated actin cytoskeleton. The junctions and polarity of the cells determine the orientation of the mitotic spindle and the plane of division. This allows for asymmetrical division of stem cells, such as those at the base of stratified epithelia (see Figs. 35.6 and 41.4).
Desmosomes Desmosomes (desmos = “bound,” soma = “body”) use cadherins to provide strong adhesions reinforced by intermediate filaments between epithelial and muscle cells. In epithelia, these junctions are small, disk-shaped, “spot welds” between adjacent cells. Desmosomes in the heart are more complicated because they are mixed with adherens junctions (see Fig. 39.18). Two families of desmosomal cadherins, named desmogleins and desmocollins, mediate cellular adhesion at desmosomes (see Fig. 30.5 and Table 30.2). The most distal of five extracellular CAD domains interact head to head with CAD1 domains from the partner cells and laterally with other cadherins in a dense
CHAPTER 31 n Intercellular Junctions
551
tangle midway between the two plasma membranes (see Fig. 30.6). Desmosomal cadherins connect to cytoplasmic inter mediate filaments via adapter proteins analogous to those that connect adherens junction cadherins to actin filaments. Two proteins related to β-catenin, plakoglobin and plakophilin, bind to cytoplasmic domain of desmosomal cadherins and form a physical link to desmoplakin, a dimeric protein related to plectin (see Fig. 35.7). The C-terminus of desmoplakin binds directly to the N-terminal, nonhelical domains of epidermal keratin intermediate filaments. Mutations in this part of epidermal keratins cause blistering skin diseases by compromising the integrity of desmosomes (see Fig. 35.6). Although all desmosomes share a common plan, selec tive expression of isoforms of their component proteins give desmosomes unique molecular compositions in various cells. For example, in epidermis, desmoglein-1 and desmocollin-1 are found only in the upper layers, whereas desmoglein-3 is in the basal layers. This explains the pathology in autoimmune blistering diseases. Patients with pemphigus foliaceus make antibodies that react with desmoglein-1 and disrupt desmosomes in the upper layers of the epidermis, whereas patients with pemphigus vulgaris produce autoantibodies to desmoglein-3 that disrupt the basal layers. These antibodies are directly responsible for the disease; transfusion of human auto antibodies into a mouse reproduces the disease. Other organs are spared, owing to the restricted expression of these two isoforms. Mutations in the corresponding desmoglein genes in mice compromise desmosomes and cause skin blisters similar to pemphigus. The development of animal tissues depends on desmosomes. Loss-of-function mutations can lead to mechanical failures. For example, mutations in the plako globin gene can be lethal in mice and humans during embryogenesis, owing to disruption of the heart. Simi larly, mutations in the desmoplakin gene cause skin and cardiac defects that can be fatal. Desmosomal proteins also participate in signal transduction. For example, plakoglobin and desmoglein suppress proliferation and promote differentiation by competing with β-catenin for binding DNA (see Fig. 30.7) and inhibiting the mitogen-activated protein (MAP) kinase pathway. Loss of desmosomes is associated with the spread of epithelial cancer cells.
Adhesion to the Extracellular Matrix: Hemidesmosomes and Focal Contacts Adhesion to the extracellular matrix is fundamentally different from intercellular adhesion because integrins, rather than homophilic interactions of cadherins, provide the transmembrane link between the cytoskeleton and ligands in the extracellular matrix (see Fig. 30.9). At focal
552
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
contacts and related assemblies, transmembrane integ rins link cytoplasmic actin filaments to the extracellular matrix (see Fig. 30.11). Hemidesmosomes are another type of integrin-based adhesive junction that links cytoplasmic intermediate filaments to the basal lamina. The morphologic resem blance of hemidesmosomes to half of a conventional desmosome belies the fact that they are fundamentally different at the molecular level (Fig. 31.8C). Like desmo somes, hemidesmosomes have a dense plaque on the cytoplasmic surface of the plasma membrane that
A. Adherens junction
anchors loops of intermediate filaments. The similarity ends there. The hemidesmosomes of simple epithelia use α6β4 integrin to adhere to laminin-5 in the basal lamina. Plectin (see Fig. 35.7) links the large cytoplasmic domain of β4-integrin to keratin intermediate filaments. More complex hemidesmosomes of stratified epithe lial cells have, in addition to α6β4-integrin, a second transmembrane adhesion protein, type XVII collagen. The type XVII collagen trimer forms an extracellular collagen triple helix (see Fig. 29.1) that anchors the
B. Desmosome
C. Hemidesmosome
Tight junction
Adherens junction
2 nm/mm CELL 1
4 nm/mm CELL 2
CELL 1
2 nm/mm CELL 2
IF
IF Desmoplakin 130 nm
β-catenin
BP230
Plectin
BP180
2 nm/mm Actin
Plakoglobin
α6 Integrin
β4 BPAG2
E-cadherin
Laminin
BASAL LAMINA FIGURE 31.8 COMPARISON OF ADHERENS JUNCTION, DESMOSOME, AND HEMIDESMOSOME. Top, Electron micrographs of thin sections. Bottom, Molecular models. A, Adherens junction. Electron micrograph from the intestinal epithelium. E-cadherins link two cells together. β-Catenin and α-catenin link the cytoplasmic domain of E-cadherin to actin filaments. B, Desmosome. Two types of cadherins—desmoglein and desmocollin—link adjacent cells together. The central dense stratum seen in the micrograph presumably corresponds to the interaction sites of the cadherins, although accessory proteins may participate. Desmoplakin and other accessory proteins link the cadherins and associated plakoglobin (related to catenin) to keratin intermediate filaments. Desmoplakin molecules are shown extended to their full length in the middle drawing, whereas in desmosomes, they must be kinked or folded (as shown in the upper drawing) because the thickness of the desmoplakin layer is half that expected from extended molecules. C, Hemidesmosome. Integrin α6β4 and type XVII collagen (also called BPAG2 [bullous pemphigoid antigen-2]) attach to the basal lamina. Plectin, BP230, and BPAG1 (bullous pemphigoid antigen-1) link the membrane proteins to keratin intermediate filaments. (A–B, Micrographs courtesy Hilda Pasolli and Elaine Fuchs, Rockefeller University, New York and from Perez-Moreno M, Jamora C, Fuchs E. Sticky business: orchestrating cellular signals at adherens junctions. Cell. 2003;112:535–548. C, Micrograph courtesy Jonathan Jones, Northwestern University, Chicago, IL.)
membrane to the basal lamina. In a blistering skin disease called bullous pemphigoid, autoantibodies attack type XVII collagen, so the protein is also called bullous pemphigoid antigen-2, or BPAG2. This clinical observa tion and genetic deletions established that both α6β4integrin and type XVII collagen are required for assembly of stable hemidesmosomes in skin. BPAG1 (bullous pemphigoid antigen-1; also called BP230) is related to plectin and helps connect the integrin to intermediate filaments. Mutations in the genes for any of the hemidesmosome proteins cause blistering skin diseases known as epider molysis bullosa. Pathology can also occur in other tissues that depend on hemidesmosomes, including the cornea, gastrointestinal tract, and muscles. Mutations in keratin genes also cause epidermolysis bullosa (see Fig. 35.6). SELECTED READINGS Broussard JA, Getsios S, Green KJ. Desmosome regulation and signaling in disease. Cell Tissue Res. 2015;360:501-512. Buckley CD, Tan J, Anderson KL, et al. Cell adhesion. The minimal cadherin-catenin complex binds to actin filaments under force. Science. 2014;346:1254211. Evans WH. Cell communication across gap junctions: a historical perspective and current developments. Biochem Soc Trans. 2015; 43:450-459. Gumbiner BM. Regulation of cadherin-mediated adhesion in morpho genesis. Nat Rev Mol Cell Biol. 2005;6:622-634. Johnson JL, Najor NA, Green KJ. Desmosomes: regulators of cellular signaling and adhesion in epidermal health and disease. Cold Spring Harb Perspect Med. 2014;4:a015297. Lee JY. Plasmodesmata: a signaling hub at the cellular boundary. Curr Opin Plant Biol. 2015;27:133-140.
CHAPTER 31 n Intercellular Junctions
553
Nielsen MS, Axelsen LN, Sorgen PL, et al. Gap junctions. Compr Physiol. 2012;2:1981-2035. Niessen CM, Leckband D, Yap AS. Tissue organization by cadherin adhesion molecules: dynamic molecular and cellular mechanisms of morphogenetic regulation. Physiol Rev. 2011;91:691-731. Oshima A. Structure and closure of connexin gap junction channels. FEBS Lett. 2014;588:1230-1237. Oshima A, Matsuzawa T, Murata K, et al. Hexadecameric structure of an invertebrate gap junction channel. J Mol Biol. 2016;428: 1227-1236. Padmanabhan A, Rao MV, Wu Y, et al. Jack of all trades: functional modularity in the adherens junction. Curr Opin Cell Biol. 2015;36:32-40. Powell AM, Sakuma-Oyama Y, Oyama N, et al. Collagen XVII/BP180: A collagenous transmembrane protein component of the dermoepi dermal anchoring complex. Clin Exp Dermatol. 2005;30:682687. Scemes E. Nature of plasmalemmal functional “hemichannels.” Biochim Biophys Acta. 2012;1818:1880-1883. Siebert AP, Ma Z, Grevet JD, et al. Structural and functional similarities of calcium homeostasis modulator 1 (CALHM1) ion channel with connexins, pannexins, and innexins. J Biol Chem. 2013;288: 6140-6153. Stahley SN, Kowalczyk AP. Desmosomes in acquired disease. Cell Tissue Res. 2015;360:439-456. Suzuki H, Nishizawa T, Tani K, et al. Crystal structure of a claudin provides insight into the architecture of tight junctions. Science. 2014;344:304-307. Tilsner J, Nicolas W, Rosado A, et al. Staying tight: plasmodesmal membrane contact sites and the control of cell-to-cell connectivity in plants. Annu Rev Plant Biol. 2016;67:337-364. Van Itallie CM, Anderson JM. Architecture of tight junctions and principles of molecular composition. Semin Cell Dev Biol. 2014;36:157-165. Walko G, Castañón MJ, Wiche G. Molecular architecture and function of the hemidesmosome. Cell Tissue Res. 2015;360:529-544. Yap AS, Gomez GA, Parton RG. Adherens junctions revisualized: organizing cadherins as nanoassemblies. Dev Cell. 2015;35:12-20.
This page intentionally left blank
CHAPTER
32
Connective Tissues A
nimals use different proportions of matrix macromolecules to construct connective tissues with a range of mechanical properties to support their organs. Bone is a stiff, hard solid; blood vessel walls are flexible and elastic; and the vitreous body of the eye is a watery gel. Plant and fungal cell walls are functionally similar to the animal extracellular matrix but are composed of completely different molecules. This chapter begins with a discussion of simple connective tissues then concentrates on cartilage, bone, development of the skeleton, and the mechanisms that repair wounds, finishing with a discussion of the plant cell wall.
Loose Connective Tissue Loose connective tissue consists of a sparse extracellular matrix of hyaluronan and proteoglycans supported by a few collagen fibrils and elastic fibrils. In addition to fibroblasts, the cell population is heterogeneous, including both indigenous and emigrant connective tissue cells (see Fig. 28.1). The loose connective tissue underlying the epithelium in the gastrointestinal tract is a good example of this heterogeneity (Fig. 32.1A), with lymphocytes, plasma cells, macrophages, eosinophils, neutrophils, and mast cells, as well as fibroblasts and occasional fat cells (see Chapter 28 for details on these cells). This variety of defensive cells is appropriate for a location near the lumen of the intestine, which contains microorganisms and potentially toxic materials from the outside world. Loose connective tissue is also found in and around other organs. In the optically transparent vitreous body of the eye, fibroblasts produce a highly hydrated gel of hyaluronan and proteoglycans, supported by a loose network of type II collagen. Few defensive cells are required, as the interior of the eye is sterile.
Dense Connective Tissue Collagen fibers, with or without elastic fibers, make up the bulk of dense connective tissue (Fig. 32.1B). Sparse
A
B Columnar epithelium
Transitional epithelium
LOOSE CT
DENSE CT FIGURE 32.1 CONNECTIVE TISSUES. A, Loose connective tissue (CT) underlying the columnar epithelium of the small intestine. Light micrograph of a section stained with Masson trichrome stain. B, Dense connective tissue (CT) underlying transitional epithelium in the wall of the ureter. Light micrograph of a section stained with hematoxylin-eosin. (Courtesy D.W. Fawcett, Harvard Medical School, Boston, MA.)
fibroblasts are present to manufacture extracellular matrix. Other connective tissue cells are even rarer, as these tissues are not usually exposed to microorganisms. Collagen fibers can be arranged precisely, as in tendons or cornea (see Fig. 29.3), or less so, as in the wall of the intestine or the skin. Tendons consist nearly exclusively of type I collagen fibers, all aligned along the length of the tendon to provide the tensile strength that is required to transmit forces from muscle to bone. The cornea that forms the transparent front surface of the eye is also well organized into orthogonal layers of collagen fibrils. Dense connective tissues can also be elastic. For example, the walls of arteries (see Fig. 29.8) and the dermal layer of skin consist of both collagen and elastic fibers. Energy from each heartbeat stretches the elastic fibers in the walls of arteries. Recoil of these elastic fibers propels blood between heartbeats and affects the blood pressure. 555
556
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
Approximately one in 5000 humans inherits a mutation in a gene for fibrillar collagens type I, type III, or type V, which causes a range of connective tissue defects called Ehlers-Danlos syndrome. Most affected individuals have thin skin and lax joints. Severe mutations lead to rupture of arteries, bowel, or uterus, often with fatal consequences. Ehlers-Danlos syndrome illustrates the importance of these collagens with regard to the integrity of the affected tissues. Inheritance is dominant, as these collagens consist of trimers of three identical subunits. Given one mutant gene, only one in eight (½ × ½ × ½) procollagen molecules is normal.
Cartilage Cartilage (Fig. 32.2) is tough, resilient connective tissue that performs a variety of mechanical roles. It covers the articular surfaces of joints and supports the trachea, other large airways, the nose, and ears. Cartilage also forms the entire skeleton of sharks and the embryonic precursors of many bones in higher vertebrates. The mechanical properties of cartilage are attributable to abundant extracellular matrix consisting of fine collagen fibrils and high concentrations of glycosaminoglycans and proteoglycans (Fig. 32.3). Chondrocytes synthesize and secrete macromolecules for the cartilage matrix, which eventually surrounds them completely. Chondrocytes replenish the
A
matrix as the macromolecules turn over slowly, but their ability to remodel and repair the matrix is limited. No blood vessels penetrate cartilage, owing to production of several inhibitors of endothelial cell growth by chondrocytes. Thus, all nutrients must diffuse into cartilage from the nearest blood vessel in the perichondrium, a dense capsule of fibrous connective tissue that covers the surface of cartilage. This capsule contains mesenchymal stem cells (see Box 41.2 and Fig. 28.1) that are capable of differentiating into chondrocytes. A meshwork of type II collagen fibrils, accounting for approximately 25% of the dry mass, fills the extracellular matrix. These slender collagen fibrils are hard to see even in electron micrographs but are quite stable, with lifetimes estimated to be many years. Fibrils tend to line up parallel to surfaces but otherwise are arranged randomly. Minor collagen type IX crosslinks type II collagen fibrils and collagen type XI binds to the surface of type II fibrils. Expression of type X collagen is restricted to cartilage that is undergoing conversion to bone. The matrix contains several minor adhesive proteins, and other proteins inhibit invasion of blood vessels. Glycosaminoglycans, including hyaluronan, constitute the second major class of matrix macromolecules. Molecules of the proteoglycan aggrecan attach to a hyaluronan backbone like the bristles of a test tube brush, forming so-called megacomplexes (see Fig. 29.13). Aggrecan also binds type II collagen. Highly charged
B. Chondrocyte
Epithelium Perichondrium
Chondrocytes ER
C. Matrix Type II collagen
FIGURE 32.2 CARTILAGE AND CHONDROCYTES. A, Light micrograph of a section of hyaline cartilage in the wall of the respiratory tree stained with periodic acid–Schiff stain and Alcian blue. The cartilage capsule of dense connective tissue (perichondrium) and the columnar epithelium lining the respiratory passage are at the top. Inset, Light micrograph of hyaline cartilage stained with toluidine blue. The proteoglycans in the matrix stain pink. The rough endoplasmic reticulum stains blue. Shrinkage during fixation and embedding creates the artifactual cavity or lacuna around each cell. B, Electron micrograph of a thin section of hyaline cartilage showing chondrocytes embedded in dense extracellular matrix. C, Electron micrograph of cartilage matrix at high magnification. This specimen was rapidly frozen and prepared by freeze-substitution to avoid collapse of the proteoglycans during dehydration and embedding. ER, endoplasmic reticulum. (A, Courtesy D.W. Fawcett and E.D. Hay, Harvard Medical School, Boston, MA. B, Courtesy of E.D. Hay, Harvard Medical School, Boston, MA. C, Courtesy E.B. Hunziker, M. Müller Institute, University of Bern, Switzerland.)
CHAPTER 32 n Connective Tissues
A
+ Water
+ Uncapped bottle compresses
Capped bottle full of water resists compression
B. Hyaluronan megacomplex trapped by collagen attracts water
Aggrecan Hyaluronan Type II collagen
FIGURE 32.3 MACROMOLECULAR STRUCTURE AND MECHANICAL PROPERTIES OF HYALINE CARTILAGE MATRIX. A, Hydrostatic model of the mechanical properties of cartilage. Water trapped in the extracellular matrix resists compression. Neither water alone (in beaker) nor a pliable container (uncapped plastic bottle) resists compression. However, if water fills a capped bottle, it resists compression. B, In the cartilage matrix, flexible strands of type II collagen trap proteoglycans, which attract large amounts of water. Trapped water resists compression because its “container,” the network of collagen fibrils, does not stretch.
glycosaminoglycans fill the extracellular space and attract water, the most abundant component of the matrix. A hydrostatic mechanism allows cartilage to resist deformation (Fig. 32.3). Collagen fibrils provide tensile strength (ie, resistance to stretching) but do not resist compression or bending. Glycosaminoglycans strongly attract water, resulting in an internal swelling pressure that pushes outward against collagen fibrils aligned parallel to the surface of the cartilage. The force of internal hydrostatic swelling pressure balances the force produced by tension on the collagen fibrils. Remarkably, this internally stressed material can resist strong external forces such as those on the articular surfaces of joints. A macroscopic analog is a thin-walled plastic bottle filled with water. One can stand on the bottle provided that it is sealed, whereas neither the empty bottle nor the water could separately support any weight.
Specialized Forms of Cartilage Hyaline cartilage provides mechanical support for the respiratory tree, nose, articular surfaces, and developing bones. Elastic cartilage has abundant elastic fibers in addition to collagen, making the matrix much more elastic than hyaline cartilage. Elastic cartilage supports structures subjected to frequent deformation, including
557
the larynx, epiglottis, and external ear. Fibrocartilage has features of both dense connective tissue (an abundance of thick collagen fibers) and cartilage (a prominent glycosaminoglycan matrix). It is tough and deformable, appropriate for its role in intervertebral disks and insertions of tendons.
Differentiation and Growth of Cartilage Cartilage grows by expansion of the extracellular matrix either from within or on the surface. For surface growth, mesenchymal cells in the perichondrium differentiate into chondrocytes that synthesize and secrete matrix materials. For internal growth, chondrocytes trapped in the matrix divide and manufacture additional matrix, which is sufficiently deformable to allow for internal expansion. Cartilage has a limited capacity to repair damage, but stem cell transplants can help some patients. Many growth factors and their receptors cooperate to influence the differentiation of precursor cells into chondrocytes, the proliferation of chondrocytes, and the production of cartilage matrix molecules. These include Indian hedgehog (Ihh), members of transforming growth factor-β family (TGF-β and bone morphogenetic proteins [BMPs]), multiple fibroblast growth factors (FGFs), parathyroid hormone–related protein (PTHrP), and insulinlike growth factors (IGF-1 and IGF-2). Chondrocytes produce some of these growth factors (TGF-β, FGFs, and IGFs). During development, adjacent tissues can induce cartilage formation by secreting TGF-β and FGF. SOX9 is the key transcription factor mediating expression of cartilage-specific genes. Diseases of Cartilage Cartilage fails in common human diseases, including arthritis and ruptured intervertebral disks. Osteoarthritis, degeneration of cartilage on joint surfaces, is very common in older people and has a complex genetic component attributable to variations in many genes. Rarely, human diseases are caused by mutations in single genes for cartilage proteins, growth factors or growth factor receptors (Appendix 32.1). For example, more than 25 different mutations of the human gene for type II collagen cause disorders of cartilage, ranging in severity from death in utero to dwarfism or osteoarthritis. Mutations in genes for minor cartilage-associated collagens cause a variety of symptoms, including degenerative joint disease. A premature stop codon in chicken aggrecan causes lethal skeletal malformations.
Bone For most vertebrates, bones provide mechanical support and serve as a storage site for calcium. The great strength and light weight of bones are attributable both to the mechanical properties of the extracellular matrix and to efficient overall design, including tubular form and
558
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
A. Gross-cut section
C. H&E stained
Spongy bone Compact bone Blood vessels in Haversian canal
Osteocyte in lacunae
Compact bone
B. Histological section Circumferential lamellae
D. Dry bone Blood vessels Haversian lamellae
Interstitial lamellae
Trabeculae
E. Osteocyte Volkmann's canals
Sharpey's fibers Haversian canal
Calcified matrix Spongy bone
Collagen fibrils
Filopodia in cannaliculus
Compact bone Marrow cavity
Gap junctions between cells
FIGURE 32.4 ORGANIZATION OF LONG BONES. A, Longitudinal section of a shoulder joint of a dried bone specimen. Struts of trabecular spongy bone reinforce compact bone in the cortex. B, A wedge of long bone. Circumferential lamellae form the outer layer just beneath the periosteum (blue) covering the surface. Osteons (Haversian systems) consist of concentric lamellae of calcified matrix and osteocytes arranged around a channel containing one or two capillaries or venules. Interstitial lamellae are fragments of osteons that remain after remodeling (Fig. 32.10). Radial vascular channels connect longitudinal vascular channels to the medullary cavity or periosteum. C, Light micrograph of a cross section stained with hematoxylin and eosin (H&E) showing circumferential lamellae on the left and two Haversian canals. D, Light micrograph of a cross section of dried bone showing a central interstitial lamella surrounded by three osteons. Narrow canaliculi connect the lacunae housing osteocytes. E, An osteocyte surrounded by calcified matrix and extending filopodia into canaliculi. (Micrographs courtesy D.W. Fawcett, Harvard Medical School, Boston, MA.)
lamination (Fig. 32.4). A superficial layer of compact bone surrounds a central cavity that is filled with marrow, fat, or both and is supported by struts of bone arranged precisely along lines of mechanical stress. External surfaces of bones are covered either by dense connective
tissue, called periosteum, or by cartilage at joint surfaces. Two cell types make bone matrix: osteoblasts covering the internal surfaces and osteocytes embedded in the bone. A third cell type, called the osteoclast, degrades bone, recycling matrix components. Blood vessels
CHAPTER 32 n Connective Tissues
penetrate compact bone through a network of channels to supply the central cavity. Although bone is durable and strong, continuous remodeling makes bone much more dynamic than it appears.
Extracellular Matrix of Bone Bone is a composite material consisting of type I collagen fibrils (providing tensile strength) embedded in a matrix of calcium phosphate crystals (providing rigidity) (Fig. 32.4E). The calcium-phosphate crystals are similar to hydroxyapatite [Ca10(PO4)6(OH)2] and make up about two-thirds of the dry weight of bone. Macroscopic analogs of the bone matrix are concrete reinforced by steel rods and fiberglass consisting of a brittle plastic reinforced by glass fibers. Each of these composites is stronger than its separate components. Simple extraction experiments illustrate the contributions of the two components of bone. After removal of calcium phosphate with a calcium chelator, bone is so rubbery that it bends easily. After destruction of collagen by heating, bone is hard but brittle. Fibrils of type I collagen, the dominant organic component of the matrix (Table 32.1), are arranged in sheets or a meshwork. Covalent crosslinks between the collagen molecules in fibrils make them inextensible. The matrix also contains more than 100 minor proteins, including growth factors, proteins that promote hydroxyapatite deposition and adhesive glycoproteins, but few proteoglycans. Cells that make bone lay down type I collagen as the substrate for crystallization of calcium phosphate. Some calcium phosphate crystallizes directly in the collagen
559
matrix. Other crystals form in small “matrix vesicles” that bud from the plasma membranes of osteoblasts and use pumps and carriers to concentrate calcium and phosphate. After being released from these vesicles, tiny crystals associate with collagen fibrils. The crystals grow and eventually fill spaces between the collagen molecules within the fibrils.
Bone Cells Overview A balance among the activities of osteoblasts, osteocytes, and osteoclasts forms, grows, and maintains bones. Osteoblasts and osteocytes produce extracellular matrix and establish conditions for its calcification. Osteoclasts resorb and remodel bone. An imbalance of these opposing cellular activities causes human diseases. Properties of Osteoblasts A monolayer of osteoblasts on the surface of growing bone tissue uses a well-developed secretory pathway to synthesize and secrete the organic components of the matrix (Fig. 32.5). Osteoblasts also act as endocrine cells, secreting growth factors that control the differentiation of osteoclasts (Fig. 32.6), as well as cells in other organs. They also help to form the niche in the bone marrow for hematopoietic stem cells (see Fig. 41.4). Regulation of Osteoblast Development Osteoblasts arise from the same mesenchymal stem cells that give rise to fibroblasts and chondrocytes (see Fig. 28.1). Growth factors control the differentiation from mesenchymal cells. They include Ihh, a subset of BMPs
TABLE 32.1 Bone Proteins Name
Content
Functions
Bone morphogenic proteins
Minor
Transforming growth factor (TGF)-β homologs; cartilage stimulation and bone development and repair
Collagen type I
90%
Forms fibrils in the bone matrix
Osteocalcin
1%–2%
Network of aspartic acid and γ-carboxylated glutamic acid side chains bind hydroxyapatite; promotes calcification; attracts osteoclasts and osteoblasts
Osteonectin
2%
Synthesized in developing and regenerating bone; binds collagen and hydroxyapatite; may nucleate hydroxyapatite crystallization in bone matrix
Osteopontin
Minor
Arginine-glycine-aspartic acid (RGD) sequence; binds osteoclast integrins to bone surface
Proteoglycans
Minor
Decorin, biglycan, osteoadherin; may bind TGF-β
Sialoproteins
2%
RGD sequence; binds osteoclast integrins to bone surface
A Bone
Osteoblasts
Calcified cartilage
B Secretory vesicle Type I collagen
ER Golgi
Calcified matrix
FIGURE 32.5 OSTEOBLASTS. A, Light micrograph of a section of forming bone stained with toluidine blue. Osteoblasts with abundant, blue-stained, rough endoplasmic reticulum lay down bone matrix (light green) on the surface of calcified cartilage (light pink). B, Drawing of osteoblasts. ER, endoplasmic reticulum. (A, Courtesy R. Dintzis and from the work of D. Walker, Johns Hopkins Medical School, Baltimore, MD.)
560
SECTION VIII n Cellular Adhesion and the Extracellular Matrix A. Osteoclast genesis Monocytes
Osteoclast precursors
RANK M-CSF
RANKL
RANKL binding to RANK stimulates precursors to fuse and differentiate into a multinucleated osteoclast OPG blocks RANKL Wnt
Supporting cells, osteoblasts
B. Bone remodeling
Sclerostin
Proteins not to scale
Osteocyte
C. Osteoclast Bone
Osteoclast Calcified cartilage Cathepsin-K secreted Bone
Sealing zone
H+ ATPase and Cl– channels secrete HCl
Sealing zone
Ruffled membrane
Osteoclast FIGURE 32.6 OSTEOCLASTS. A, Formation of a multinucleated osteoclast by fusion of monocytes stimulated by the receptor activator of nuclear factor κB ligand (RANKL), macrophage colony-stimulating factor (M-CSF), and other factors. B, Light micrograph of a section of forming bone stained with toluidine blue showing two osteoclasts degrading bone and calcified cartilage. C, An osteoclast attached to the bone matrix by a sealing zone, forming a resorption cavity (pink). The cell pumps H+ and secretes lysosomal enzymes into this cavity to resorb the surface of the matrix. ATPase, adenosine triphosphatase; OPG, osteoprotegerin (OPG); RANK, receptor activator of nuclear factor κB. (B, Courtesy R. Dintzis and from the work of D. Walker, Johns Hopkins Medical School, Baltimore, MD.)
(see Fig. 24.7), and some Wnts (see Fig. 30.7). Humans with loss-of-function mutations in a Wnt coreceptor have few osteoblasts and low bone density, whereas loss-offunction mutations in a BMP competitor have the opposite effect. On the other hand, osteocytes secrete a protein called sclerostin that blocks the Wnt coreceptor (Fig. 32.6), so loss-of-function mutations of sclerostin strengthen Wnt signals and promote bone formation. Inside osteoblasts, Runx2/Cbfa1 is the master transcription factor controlling the expression of genes that are required to make bone matrix. Runx2/Cbfa1 is part of a network of transcription factors and microRNAs with positive and negative influences on osteoblast differentiation and function. Mouse embryos lacking Runx2/Cbfa1 have no osteoblasts or osteoclasts. They make a cartilage skeleton that never transforms to bone. Humans and mice with just one active Runx2/Cbfa1 gene lack collarbones and experience a delay in the fusion of joints between skull bones. This syndrome is the most common human skeletal defect. Osteocyte Properties Once an osteoblast has enclosed itself within bone matrix, it is called an osteocyte. Long-lived, metabolically active osteocytes are connected to each other by
many long, slender filopodia that run through narrow channels in the matrix (Fig. 32.4D–E). Gap junctions between the processes of osteocytes provide a continuous network of intercellular communication that stretches from cells adjacent to blood vessels to the most deeply embedded osteocyte. Osteocytes can either lay down or resorb matrix in their immediate vicinity. Circulating hormones influence the activity of osteoblasts and osteocytes. In response to the calcium concentration in blood, parathyroid glands secrete parathyroid hormone, which stimulates osteocytes to mobilize calcium from the surrounding matrix. This feedback loop maintains a constant concentration of calcium in the blood. Osteoclast Properties Osteoclasts form by fusion of blood monocytes and resorb bone, as required for growth and remodeling. Osteoclasts are multinucleated giant cells specialized for bone resorption (Fig. 32.6). They attach like a suction cup to the surface of bone. Interactions of a plasma membrane integrin (αVβ3) with bone matrix proteins (osteopontin and sialoprotein) help to create a leakproof compartment on the bone surface. Osteoclasts amplify the plasma membrane lining this closed space, forming
CHAPTER 32 n Connective Tissues
561
a “ruffled border” composed of microvilli enriched with H+-transporting V-type rotary adenosine triphosphatase (ATPase) pumps (see Fig. 14.6) and chloride channels (see Fig. 16.13). The combined activities of the H+ pump and chloride channels allow the cell to secrete hydrochloric acid into the sealed extracellular compartment on the bone surface. This closed space acts like an extracellular lysosome: Acid dissolves calcium phosphate crystals, and secreted proteolytic enzymes, including cathepsin K, digest collagen and other organic components. Degradation products are taken up by endocytosis and transported across the cell in vesicles for secretion on the free surface. Amino acids are reused, but collagen crosslinking groups are not, so they are excreted in the urine, where their concentration is a measure of bone turnover.
Norepinephrine released by sympathetic nerves activates osteoblasts to secrete RANKL. This explains why animals and people that lack leptin or its receptor not only are obese but also have dense bones. Osteoclast growth factors RANKL, TNF, and interleukin-1 mediate excess bone resorption at sites of chronic inflammation in rheumatoid arthritis and gum diseases. Differentiation of osteoclasts is subject to negative regulation by a soluble decoy receptor for RANKL called OPG (osteoprotegerin), which binds RANKL and competes for activation of RANK (Fig. 32.6). Estrogens inhibit osteoclast differentiation by stimulating osteoblasts to produce OPG, so circulating OPG declines in parallel with estrogen levels after menopause. The resulting increase in osteoclasts contributes to bone loss (osteoporosis) in older women.
Osteoclast Formation Bone marrow supporting cells, osteoblasts, and activated T lymphocytes produce two proteins that stimulate blood monocytes to fuse and differentiate into multinucleated osteoclasts (Fig. 32.6). These key factors are macrophage colony-stimulating factor (M-CSF) and RANKL (receptor activator of nuclear factor kappa B [NF-κB] ligand, also called osteoprotegerin ligand [OPGL] or tumor necrosis factor–related activation-induced cytokine [TRANCE]). Both factors are produced locally in bone marrow as transmembrane proteins with the growth factor domain on the cell surface. These proteins control differentiation through binding to their receptors on monocytes by either direct cell-to-cell contact or release of the active domain by proteolytic cleavage. First, M-CSF activates a cytokine receptor (see Fig. 24.6) on monocytes, stimulating a JAK (just another kinase) kinase–signal transducer and activator of transcription (JAK-STAT) pathway (see Fig. 27.9) and turning on expression of genes required for the monocyte to differentiate into a preosteoclast. An important change is the expression of a receptor called RANK (receptor for activation of NF-κB, a member of the tumor necrosis factor [TNF] receptor family; see Fig. 24.9). Once this receptor is expressed, RANKL can activate preosteoclasts through the transcription factor NF-κB (see Fig. 10.21C) to express the proteins required for cell fusion and further differentiation into an osteoclast. Mice that lack RANKL form no osteoclasts, so bone resorption fails. Other growth factors, including TNF itself, contribute to this process by acting directly on osteoclasts, but many stimulators of osteoclast differentiation (eg, parathyroid hormone, vitamin D, leptin) act indirectly by stimulating supporting cells to make RANKL. For example, leptin, a satiety hormone secreted by fat cells, acts on neurons of the hypothalamus in the brain that regulate not only appetite but also bone metabolism indirectly via the sympathetic nervous system.
Formation and Growth of the Skeleton Both genetic and environmental information direct formation of the skeleton. Genetic information predominates in the master plan and initial development of skeletal tissues, as the size and shape of bones are characteristic for each species. Subsequently, environmental information is important in remodeling of the skeleton in response to use. Mutations in genes for structural and informational molecules have provided valuable clues about the genetic blueprint for the skeleton (Appendix 32.1). Genetic information is read out on at least two levels. First, master genetic regulators—including transcription factors encoded by HOX (homeobox) and PAX (paired box) genes—specify the developmental fate of each embryonic segment. Homeoboxes are DNA sequences that encode a family of 60-residue protein domains that bind DNA (see Fig. 15.14). The human genome contains 39 HOX genes arrayed in four linear arrays, similar to those in other animals. HOX genes were discovered in flies as a result of mutations that cause “homeotic conversion,” whereby the fate of one segment is converted into another, such as the substitution of a leg for an antenna. The same thing happens in vertebrates: Mouse embryos express Hoxd-4 in the second cervical (neck) vertebra and more posterior segments. Mutation of Hoxd-4 results in the second cervical vertebra taking on some of the features of the first cervical vertebra. Mutations in other HOX genes cause congenital malformations in humans. The pathways from HOX genes to determinants of three-dimensional architecture are still incompletely understood. Both systematically circulating and locally secreted growth factors control the proliferation and differentiation of the cells of cartilage and bone. Mutations in these factors and their receptors also cause surprisingly specific human skeletal malformations (Appendix 32.1). Circulating growth hormone produced by the pituitary
562
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
gland is a major determinant of skeletal size. Individuals deficient in growth hormone are short in stature. Locally produced growth factors, including BMPs and FGFs and their receptors, control the development and growth of cartilage and bone during embryogenesis, in addition to stimulating repair after fractures. FGF receptors are tyrosine kinases (see Fig. 24.4). BMPs are related in structure and mechanism to TGF-β and are expressed in tissues other than bone and cartilage (see Fig. 24.7). BMPs are part of a system of positive and negative factors that regulates formation of cartilage, bone, and joints. For example, a BMP called GDF-5 specifies the position of joints, but joints form only if noggin protein, an inhibitor of other BMPs, is present.
Embryonic Bone Formation Bone always forms by replacement of preexisting connective tissue. During embryonic development, flat bones, such as the skull and shoulder blades, form from neural crest cell precursors in loose connective tissue (Fig. 32.7). Somehow, information in the genome is read out as the three-dimensional pattern of a skull. Growth factors, vitamins (eg, retinoic acid), and local matrix molecules, such as glycosaminoglycans, all influence the differentiation of these cells into osteoblasts at specific locations in connective tissue. Osteoblasts lay down struts of bone matrix in the loose connective tissue. As new bone is laid down on the surface of these bone spicules, some osteoblasts are trapped and become osteocytes. During embryonic and postnatal development, genetic information precisely controls changes in the size and proportions of flat bones. For example, for the skull to increase in size both externally and internally, osteoclasts on the outer surface lay down new bone at the same rate as osteoclasts resorb old bone inside (Fig. 32.8A). These cellular activities are carefully coordinated to change the proportions of the skull as the individual matures.
A
Long bones, such as the humerus, begin as cartilage models that are replaced by bone (Fig. 32.9). Multiple, genetically programmed factors induce clusters of mesenchymal cells at specific locations to differentiate into chondrocytes that secrete type II collagen and glycosaminoglycans. This produces a miniature cartilaginous version of the adult bone. Bone replaces this cartilage precursor in a series of steps that are coordinated locally by production of growth factors. Perichondrial cells and proliferating chondrocytes secrete PTHrP, which promotes chondrocyte division and growth. In supporting roles, BMPs promote and FGFs inhibit the growth and differentiation of chondrocytes by acting upon populations of cells that express particular receptors for these molecules (Appendix 32.1). Active FGF receptors stimulate STAT transcription factors (see Fig. 27.9) and/or mitogen-activated protein (MAP) kinase pathways (see Fig. 27.6). More mature chondrocytes produce Ihh, which directs the terminal differentiation of neighboring chondrocytes. For a long bone to maintain its shape as it grows in size, deposition and removal of bone tissue must be highly selective. For the shaft to grow in diameter, new bone is laid down on the outer surface by osteoblasts at the same time as old bone is removed inside by osteoclasts (Figs. 32.8B and 32.9). Bones grow longer as a result of interstitial growth of cartilage in the epiphyseal plate and its continual replacement by bone. Chondrocytes contribute to the elongation of bones in two ways: chondrocytes continuously proliferate in one zone and then rapidly increase in mass and swell in the adjacent zone next to forming bone (Fig. 32.9B). The hypertrophic chondrocytes secrete type X collagen and use matrix metalloproteinases to resorb some of their surrounding matrix. They also direct the calcification of the cartilage matrix before undergoing apoptosis or differentiating into osteoblasts. Osteoblasts lay down bone matrix on the surface of the cavities in the calcified cartilage. Hypertrophic cartilage
Osteocyte
B
Bone Osteoclast
Osteoblast Blood vessel Mesenchymal cells FIGURE 32.7 BONE FORMATION BY INTRAMEMBRANOUS OSSIFICATION. A, Light micrograph of a section of forming bone stained with hematoxylin and eosin. Calcified bone matrix is maroon. B, Interpretive drawings. Connective tissue mesenchymal cells differentiate into osteoblasts, which lay down bone matrix (blue). Osteoblasts become trapped as the matrix grows. (A, Courtesy D.W. Fawcett, Harvard Medical School, Boston, MA.)
CHAPTER 32 n Connective Tissues
A
Hyaline cartilage Bone collar
Primary ossification center
Periosteal bud blood vessel New spongy bone Medullary cavity
1 Proliferating chondrocytes
B
Cartilage eroded
2 Hypertrophic chondrocytes
Periosteal bud invades
3 Calcified cartilage apoptosis
Medullary cavity formed
Epiphyses erode
Epiphyses ossify
563
Secondary ossification center
4 Invasion of cartilage with bone deposition
Epiphyseal blood vessel
Articular cartilage Epiphyseal plate (cartilage)
Epiphyseal plate (ossified)
FIGURE 32.8 FORMATION OF A LONG BONE BY REPLACEMENT OF CARTILAGE. A, The shaft grows in diameter as osteoblasts lay down bone (tan) on the outer surface of the primary collar of bone and osteoclasts remove bone from the inner surface to form and maintain the marrow cavity. The bone grows in length by interstitial expansion of the cartilage in the epiphyseal plate and its replacement by bone. B, Light micrograph of a section of an epiphyseal plate stained with toluidine blue. Cartilage growth, differentiation, and replacement by bone occur in several zones. Proliferation of chondrocytes and their production of matrix (pink) containing type II collagen are solely responsible for the longitudinal growth of the bone (1). Hypertrophic chondrocytes enlarge and make type X collagen, as well as matrix metalloproteinases that resorb some of the surrounding matrix (2). Chondrocytes die by apoptosis (see Chapter 46), and the matrix calcifies (3). Blood vessels and osteoblasts move into spaces vacated by chondrocytes and lay down bone (blue) on the surface of calcified cartilage (4). (Micrograph courtesy R. Dintzis and from the work of D. Walker, Johns Hopkins Medical School, Baltimore, MD.)
A. Skull growth
B. Long bone growth
Osteoclast 25 years 6 years Newborn 7 month fetus
Bone deposition (osteoblasts)
Growth plate
Bone removal (osteoclasts)
FIGURE 32.9 BONE GROWTH. A, Light micrograph of a section of skull stained with Mallory’s trichrome stain and an interpretive drawing. The skull expands during fetal development and growth to adulthood as osteoblasts lay down new bone on the outer surface (blue) and osteoclasts resorb bone (pink) on the inner surface. B, Long bones grow entirely by expansion of cartilage in the epiphyseal plate and its replacement by bone (tan), followed by resorption (pink). (A, Courtesy D.W. Fawcett, Harvard Medical School, Boston, MA.)
564
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
ceases to make the factors that inhibit endothelial cell growth, allowing FGF-2, TGF-β, and vascular endothelium growth factor to attract capillaries as part of the transformation of cartilage to bone. Growth of long bones stops at puberty, when high concentrations of estrogen and testosterone stop proliferation of epiphyseal chondrocytes so that bone replaces this cartilage. This closure of the epiphyses throughout the body occurs over several years in a predictable order, so one can judge the maturity of a child by examining epiphyses by radiographic studies. Genetic variations in this process of maturation give rise to differences in stature. Metabolic and endocrine disorders can also affect the timing of epiphyseal closure.
Bone Remodeling Bone is amazingly dynamic and is remodeled continuously in response to stresses. Bone cells and matrix turn over every few years. Reorganization of bone requires two carefully coordinated steps: breakdown of preexisting bone by osteoclasts and replacement with new bone by osteoblasts. More than 100 years ago, Wolff realized that the strength of a bone depends on use. For example, bones of the racquet arm of tennis players are more robust than the bones of their other arm. Thus, mechanical forces on the bones must generate modulatory signals
A Cutting cone
B Osteoclast
Reversal zone
Time
Blood vessel Fibroblast Osteoblasts
that control remodeling. Sensory nerves are involved in some way, but most research has focused on how cells detect fluid flow through canaliculi. Primary cilia have been implicated in both the differentiation of bone cells and their responses to mechanical forces. Part of their function must be in hedgehog signaling (see Chapter 38), but they probably also sense fluid flowing through canaliculi as a result of mechanical force on the bone. Accordingly, some mutations in genes for components of the intraflagellar transport machinery (see Fig. 38.18) cause severe skeletal defects. Formation of the cylindrical units of long bones called osteons is a good example of well-coordinated remodeling. The process involves two steps (Fig. 32.10). First, osteoclasts resorb preexisting bone to form long, cylindrical, resorption channels in the same way that a plumber’s snake clears debris from drain pipes. The second step is slower, as osteoblasts take weeks to fill in these channels by depositing concentric layers of lamellar bone against the walls. They lay down matrix at a rate of about 1 µm of thickness per day until bone completely surrounds the blood vessels trapped in the middle of the newly formed osteon. Because resorption channels cut randomly through the bone, fragments of older osteons are left behind during the remodeling of mature bone. These fragments are called interstitial lamellae.
Forming resorption cavity
C
D
Resorption cavity
Forming Haversian system
Closing cone
Quiescent osteoblast
Completed Haversian system
FIGURE 32.10 BONE REMODELING. A–B, Longitudinal and cross sections of a time line illustrating the formation of an osteon. Osteoclasts cut a cylindrical channel through bone. Osteoblasts follow, laying down bone on the surface of the channel until the matrix surrounds the central blood vessel of the newly formed osteon. C, Steps in the formation of a new osteon. Parts of older osteons are left behind as interstitial lamellae. D, Microradiograph of a cross section of a long bone, illustrating the range of ages of the structures. A section of bone is placed on x-ray film, exposed to x-rays, developed, and examined by light microscopy. Older parts of the bone, such as the interstitial lamellae, are more heavily calcified and therefore absorb more of the x-rays, appearing lighter. Newly formed osteons appear the darkest, as they are the least calcified. Vascular spaces are empty and fully exposed by the x-rays. (A, Modified from Parfitt AM. The action of parathyroid hormone on bone. Metabolism. 1976;25:809–844. D, Courtesy D.W. Fawcett, Harvard Medical School, Boston, MA.)
Resorption may release growth and differentiation factors from the mineralized matrix that provide a local stimulus for the next round of bone formation by new osteoblasts.
Bone Diseases Osteoporosis, a thinning of bones, is common in elderly people as a result of an imbalance of bone resorption over renewal. In the United States, osteoporosis results in 1.5 million painful fractures, costing nearly $20 billion annually. Almost half of women suffer from such a fracture at some time in their lives, typically as estrogen levels decline after menopause. Osteoporosis also occurs at reduced gravitational forces during space flight. The pathogenesis is not understood, but both behavioral (eg, inactivity, poor nutrition, smoking) and multiple genetic factors have modest effects. Among many genetic factors, one might be naturally occurring variants of the nuclear receptor for vitamin D. This receptor is a transcription factor required for vitamin D to stimulate intestinal calcium uptake and calcification of bone. Variations in the genes for type I collagen or bone growth factors may also contribute. Two strategies are used to treat osteoporosis. The first is to reduce bone resorption using with bisphosphonates (pyrophosphate mimics) or injection of either OPG or antibodies to RANKL, which interfere with osteoclast formation. New inhibitors of cathepsin K are also being tested. The other approach is to promote bone formation with vitamin D, estrogen, calcium, or strontium, but these measures are only partially effective. More promising is injection of an analog of parathyroid hormone or antibodies to sclerostin, which promote osteoblast activity. Osteopetrosis is failure of bone resorption, leading to an imbalance of renewal over resorption. This rare disease of osteoclasts is fatal in humans, owing to bone marrow failure. Recessive mutations in the gene for the V-type proton-ATPase pump (60%), two genes for a chloride channel (∼15%), and two genes for proteins involved with the secretory pathway account for most human cases. Naturally occurring or engineered mutations in the genes for essentially any protein required for osteoclast differentiation or function cause osteopetrosis in mice. Osteoclasts are present in bone but fail to function properly. The disease can be cured in humans and mice by transplantation of bone marrow stem cells to replace defective osteoclast precursors, an early example of stem cell therapy. If the mutation is in the gene encoding RANKL, replacement of this growth factor cures the disease. Osteogenesis imperfecta is the name of a variety of congenital fragile bone syndromes. Severely affected fetuses die in utero from multiple broken bones. Mildly affected individuals are born but suffer multiple fractures resulting in skeletal deformities. All of the patients have
CHAPTER 32 n Connective Tissues
565
mutations in the gene for type I collagen. Some are deletions or insertions, which may be mild. Most patients with severe disease have point mutations leading to replacement of a glycine by a larger amino acid. This prevents the zipper-like folding of the collagen triple helix (see Fig. 29.1), even if only one chain is defective per molecule. This poisons assembly and accounts for the dominant phenotype. No one knows why these mutations in type I collagen do not affect other tissues, such as skin, which are rich in type I collagen.
Repair of Wounds and Fractures Healing of minor skin wounds is a familiar occurrence that illustrates the mechanisms controlling the assembly of connective tissue. Repair of connective tissue in the dermis underlying the epithelium proceeds in three stages: formation of a blood clot, assembly of provisional connective tissue, and remodeling of the connective tissue (Fig. 32.11). Tissue damage ruptures blood vessels, releasing blood that clots to stem the hemorrhage and fill the damaged area. Thrombin, a proteolytic enzyme in blood plasma, drives the clotting reaction by cleaving the plasma protein fibrinogen to form fibrin. Fibrin spontaneously polymerizes and is crosslinked to itself and to plasma fibronectin. This provisional extracellular matrix of fibrin and fibronectin provides physical integrity for the clot and an environment for wound repair. Thrombin also activates seven-helix receptors on platelets (see Fig. 30.14), stimulating them to secrete matrix molecules (thrombospondin, fibrinogen, fibronectin, and von Willebrand factor) and growth factors (platelet-derived growth factor [PDGF], TGF-β, and TGFα) that initiate the cellular events required to complete wound repair. Chemotactic factors attract phagocytes from the blood into the wound. These factors include PDGF, chemokines, peptides cleaved from fibrinogen by thrombin, and peptides from any contaminating bacteria. Neutrophils arrive first from the nearby blood vessels, having attached to activated endothelial cells (see Fig. 30.13) and migrated into the connective tissue and clot. They ingest any bacteria. Then monocytes (using a similar mechanism) migrate into the clot and clear foreign material and any dead neutrophils. The environment in a wound promotes transformation of monocytes into macrophages (see Fig. 28.6), which synthesize and secrete cytokines and growth factors that mediate the cellular events that complete the repair process. In this way, platelets, monocytes, and fibroblasts form a relay, passing information from one cell to the next. During the next phase of repair, macrophages, fibroblasts, and capillary endothelial cells migrate into the fibrin clot and reestablish the connective tissue. Endothelial cells form capillary loops that allow blood to flow and
566
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
A
Wound in connective tissue Blood
B
FIGURE 32.11 REPAIR OF A WOUND IN CONNECTIVE TISSUE. A, Wounding removes some tissue and damages blood vessels, releasing blood into the defect. B, Blood forms a clot of fibrin and fibronectin, releasing fibrin peptides, and platelets secrete plateletderived growth factor (PDGF) and transforming growth factor (TGF)-β, all of which attract neutrophils and monocytes. C, Neutrophils ingest any bacteria. Monocytes clean up debris and differentiate into macrophages, which secrete cytokines, attracting fibroblasts and blood vessels. D, Fibroblasts secrete type III collagen and hyaluronan, which, in turn, replace the fibrin clot. E, Fibroblasts remodel the provisional connective tissue with type I collagen, and blood vessels grow back into the new tissue.
Clot of fibrin and fibronectin forms Platelets secrete PDGF and TGF-β Peptides released from fibrin attract neutrophils and monocytes
C
Neutrophils ingest bacteria Monocytes differentiate into macrophages Macrophages secrete cytokines Cytokines attract capillaries and fibroblasts
D Fibroblasts secrete collagen III and hyaluronan, which replace the fibrin coat
E
Provisional matrix is replaced by collagen I
to provide oxygen. Initially, the endothelial cells are attracted by growth factors released by platelets, but macrophages and dissolution of fibrin provide a more sustained supply of chemoattractants and growth factors. Integrin receptors for fibronectin allow fibroblasts to
migrate into the clot. They secrete more fibronectin as they move. Within the clot, PDGF and TGF-β from macrophages stimulate fibroblasts to secrete type III collagen, hyaluronan, SPARC (secreted protein acidic and rich in cysteine), and tenascin. Initially, this loose connective tissue is disorganized and weak. Hyaluronan predominates transiently, but after about five days, it is gradually replaced by proteoglycans and type I collagen. Two events complete the repair of the matrix. First, fibroblasts differentiate into (smooth muscle–like) myofibroblasts, which contract the collagen matrix, closing the edges of the wound. This step is particularly important for large wounds. Second, fibroblasts remodel the provisional connective tissue to restore its original architecture with nearly normal physical strength. This requires resorption of provisional collagen fibrils by metalloproteinases (see Fig. 29.19) and assembly of more robust type I collagen fibrils. While fibroblasts repair the connective tissue, the epithelium bordering the wound spreads by cell division and migration to cover the defect. This process of migration is initiated within hours of wounding. Both the loss of contacts with neighboring cells at the edge of the wound and the release of growth factors in the wound are thought to transform the static epithelial cells into migrating cells. Keratin filaments that predominate in the cytoskeleton of skin epithelial cells are replaced with actin filaments. Hemidesmosomes that anchor the skin epithelial cells to the basal lamina are lost, and the cells migrate over the surface of the underlying matrix, which consists initially of fibrin and fibronectin and later of collagen. As they go, epithelial cells lay down a new basal lamina. Depending on the size of the defect, proliferation of epithelial cells might be required to complete coverage of the surface. When it is covered, the cells begin to differentiate into stratified epithelium. Many parallels exist between repair of a fractured bone and repair of a skin wound. Blood escapes from damaged blood vessels and clots at the fracture site. PDGF released by platelets stimulates mesenchymal cells to proliferate in the surrounding tissue. These
CHAPTER 32 n Connective Tissues
cells migrate into the clot along with blood vessels and macrophages. Stimulated by growth factors released initially by platelets and in a more sustained fashion by macrophages, mesenchymal cells differentiate into chondrocytes and osteoblasts that recapitulate the development of new bone to fill in the defect. Although the bone that is initially produced to join the fractured ends is poorly organized, fractures are mechanically stable within approximately 6 weeks. The fibrin clot is converted directly into bone if the broken bone is immobilized. A cartilage intermediate may form first if the fracture is allowed to move. Over a period of months, remodeling reestablishes the normal pattern of the bone. With time, remodeling can even straighten out bones that are mildly bent at fracture sites. In all of these examples, wound healing is coordinated by a variety of growth factors and cytokines and is supported by the environment provided by the extracellular matrix. For example, PDGF from platelets stimulates the proliferation of fibroblasts and attracts them to the fibrin clot at the site of a wound. TGF-β inhibits fibroblast proliferation but stimulates fibroblasts to make matrix molecules. The actions of cytokines and growth factors depend on the local environment in the matrix. In a fibrin clot, TGF-β binds to its recep tor on cells rather than the matrix. In the normal connective tissue matrix, TGF-β binds to proteoglycans in
A
567
preference to its cell surface receptors, limiting its effects. In a fibrin/fibronectin clot, cellular fibronectin receptors bind the matrix, stimulating production of matrix metalloproteinases that are appropriate for remodeling the matrix. In normal connective tissue with less fibronectin, cells produce less metalloproteinase. The mechanisms that mediate physiological wound repair can also contribute to disease. For example, PDGF that is released from activated platelets in clots at the sites of wounds initiates the cellular events that are required for repair. On the other hand, when the endothelium lining of large arteries is damaged, binding to the exposed basal lamina activates platelets. This stimulates them to release PDGF, which promotes proliferation of fibroblasts and smooth muscle cells in the artery wall, an early step in the development of arteriosclerosis.
Plant Cell Wall The cell walls of land plants are composite materials consisting of cellulose, other polysaccharides, and glycoproteins (Figs. 32.12 and 32.13). Wood and cotton are two familiar examples of cell wall material that is left behind after plant cells have died. Like the extracellular matrix of animals, plant cell walls not only provide mechanical support but also may influence development. Because of these robust cell walls, plant cells are
B
C
Microfibrils
CYTOPLASM OF CELL 1
CELL WALL MIDDLE LAMELLA
CELL WALL
Microtubules CYTOPLASM OF CELL 2
CYTOPLASM
FIGURE 32.12 PLANT CELL WALL. A, Confocal fluorescence micrograph of an Arabidopsis leaf with cell walls stained by the periodic acid–Schiff reaction using Acriflavine as the Schiff reagent. B–C, Electron micrographs of thin sections of cell walls in the root-like appendages of the parasitic weed dodder. B, Two cells are separated by an electron-translucent cell wall consisting of cellulose, xyloglycan, and pectins. The darker area between the two cell walls is the middle lamella, which contains a high concentration of pectins. C, At high magnification, an oblique section through the plasma membrane and cell wall shows cellulose microfibrils aligned roughly parallel to cortical microtubules inside the plasma membrane. (A, Courtesy Steven E. Ruzin, University of California, Berkeley. B–C, Courtesy K.C. Vaughn, U.S. Department of Agriculture, Stoneville, MD.)
568
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
A
C
CYTOPLASM
ECM
Golgi vesicle with matrix glycans
B. Cellulose synthase monomer
Cellulose synthase complex with 6 enzyme trimers
Active site
Cellulose
Matrix glycans
Microfibrils of ~18 cellulose polymers
FIGURE 32.13 CELL WALL SYNTHESIS. A, Confocal fluorescence micrograph of an Arabidopsis hypocotyl epidermal cell expressing tubulin tagged with cyan fluorescent protein (CFP, shown in magenta) and cellulose synthase CESA6 tagged with yellow fluorescent protein (YFP, shown in green). This is a superimposition of five successive images taken at 10-second intervals to show green particles of CESA aligned with the magenta microtubules. B, Ribbon diagram and space-filling module of the crystal structure of a bacterial cellulose synthase showing the eight transmembrane helices, the glycosyltransferase domain in the cytoplasm between helices 4 and 5, and the growing cellulose polymer (white and red) threading across the membrane. C, Schematic showing the biosynthesis of the cell wall. ECM, extracellular matrix. (A, Courtesy R. Gutierrez, J. Lindeboom, and D. Erhardt, Stanford University. For reference, see Paredez AR, Somerville CR, Ehrhardt DW. Visualization of cellulose synthase demonstrates functional association with microtubules. Science. 2006;312:1491–1495. B, For reference, see Protein Data Bank [www.rcsb.org] file 4HG6 and Morgan JL, Strumillo J, Zimmer J. Crystallographic snapshot of cellulose synthesis and membrane translocation. Nature. 2013;493:181–186. C, Modified from Cosgrove DJ. Loosening of plant cell walls by expansins. Nature. 2000;407:321–326.)
not motile. Therefore the morphology of plants is established by the orientation of the cell divisions that occur during their development. Two types of forces act on cell walls. Internally, the vacuole of the plant cell applies a high turgor pressure on the order of one atmosphere. Cell walls also resist a variety of external mechanical forces that tend to deform the cell. The main constituent of cell walls is cellulose, the most abundant biopolymer on earth. It is a long, unbranched polymer of glucose (see Fig. 3.25). Cellulose polymers associate laterally into 5- to 7-nm bundles called microfibrils (Fig. 32.13C). Two types of branched polysaccharides—hemicelluloses and pectins—associate with cellulose microfibrils along with many proteins. Plasma membrane enzymes synthesize cellulose, while the other components come from the secretory pathway and associate with cellulose outside the cell. Products of more than 1000 genes are thought to participate in cell wall synthesis. Plants inherited their genes for cellulose synthases from bacteria. Arabidopsis has genes for approximately 10 different cellulose synthases. These enzymes consist of eight transmembrane helices with a cytoplasmic β-glycosyltransferase domain similar to hyaluronan synthase and chitin synthase (Fig. 32.13B). The active site is exposed to the cytoplasm to provide access to uridine diphosphate (UDP)-glucose that supplies the glucose added to the polymer. The transmembrane helices
form a channel for the cellulose polymer across the membrane. These transmembrane enzymes form a rosette of six particles that are visible by electron microscopy, each particle likely to consist of three enzyme subunits. Outside the cell the cellulose polymers assemble into linear crystals called microfibrils. The number of polymers per microfibril was long thought to be 36, but 18 is now the accepted number. Hydrogen bonds constrain the glucose units to face in alternate directions in planar ribbons (Fig. 3.25A). These ribbons self-assemble laterally into planar crystalline sheets, which stack vertically into paracrystalline bundles that are held together by C-H•••O hydrogen bonds. Cellulose synthesis moves the rosettes of cellulose synthase in the plane of the plasma membrane along paths defined by cytoplasmic microtubules (Fig. 32.13A). Typically, cytoplasmic microtubules, the tracks of cellulose synthases, and the cellulose microfibrils are all aligned like barrel hoops perpendicular to the axis of cellular growth to allow for directed (or anisotropic) expansion of the cell wall. Cellulose synthesis continues without microtubules but it is not so well organized. Newly synthesized microfibrils are deposited between the cell surface and older cell wall components. A large number of glycosyltransferases and other enzymes in the Golgi apparatus synthesize pectins and hemicelluloses, which are transported in vesicles to the
surface for secretion. Pectins are acidic polysaccharides that form a gel between microfibrils and play important roles in cell wall expansion during growth and development. Hemicelluloses are branched polysaccharides that coat microfibrils but are less important than pectins. Primary cell walls, laid down at the time of cellular growth and expansion, mature with the addition of glycoproteins and organic molecules, such as lignins (polymers of phenylpropanoid alcohols and acids), which contribute to the integrity of the “secondary” cell wall. Covalent and noncovalent bonds are thought to link cellulose and these other matrix molecules. The great strength and flexibility of tree branches illustrate the remarkable mechanical properties of mature cell walls. Cellulose microfibrils are flexible and have a tensile strength greater than that of steel, so they do not stretch. For a plant tissue to expand, microfibrils and bonds among wall components must rearrange, processes facilitated in plants and bacteria by matrix proteins called expansins. The mechanism is still being investigated, but expansins may break noncovalent links between the polymers transiently, allowing turgor pressure to expand the volume of the cell. Genetic defects in expansins inhibit the growth of plant tissues and the ripening of some fruits, such as tomatoes. Expansins in grass pollen are one allergen responsible for hay fever. Localized chemical modifications of pectins can also loosen the cell wall and contribute to expansion of plant cells. Little is known about the molecular basis of plant cells adhering to their cell walls. By virtue of their phy sical connection with their product, cellulose synthases offer one means of attachment. Other plasma membrane proteins, including a family of serine/threonine kinases and some proteins with glycosylphosphatidylinositol anchors, may contribute to adhesion by binding cell walls. Integrins are conspicuously missing from plant cells.
CHAPTER 32 n Connective Tissues
569
ACKNOWLEDGMENT We thank David Ehrhardt for his suggestions on revisions of this chapter. SELECTED READINGS Capulli M, Paone R, Rucci N. Osteoblast and osteocyte: games without frontiers. Arch Biochem Biophys. 2014;561:3-12. Charles JF, Aliprantis AO. Osteoclasts: more than “bone eaters.” Trends Mol Med. 2014;20:449-459. Cosgrove DJ. Loosening of plant cell walls by expansins. Nature. 2000; 407:321-326. Georgelis N, Nikolaidis N, Cosgrove DJ. Bacterial expansins and related proteins from the world of microbes. Appl Microbiol Biotechnol. 2015;99:3807-3823. Goldring MB, Berenbaum F. Emerging targets in osteoarthritis therapy. Curr Opin Pharmacol. 2015;22:51-63. Golub EE. Role of matrix vesicles in biomineralization. Biochim Biophys Acta. 2009;1790:1592-1598. Martin TJ. Bone biology and anabolic therapies for bone: current status and future prospects. J Bone Metab. 2014;21:8-20. McFarlane HE, Döring A, Persson S. The cell biology of cellulose synthesis. Annu Rev Plant Biol. 2014;65:69-94. Newman RH, Hill SJ, Harris PJ. Wide-angle x-ray scattering and solidstate nuclear magnetic resonance data combined to test models for cellulose microfibrils in mung bean cell walls. Plant Physiol. 2013; 163:1558-1567. Ornitz DM, Marie PJ. Fibroblast growth factor signaling in skeletal development and disease. Genes Dev. 2015;29:1463-1486. Peaucelle A, Wightman R, Höfte H. The control of growth symmetry breaking in the Arabidopsis hypocotyl. Curr Biol. 2015;25: 1746-1752. Sobacchi C, Schulz A, Coxon FP, Villa A, Helfrich MH. Osteopetrosis: genetics, treatment and new insights into osteoclast function. Nat Rev Endocrinol. 2013;9:522-536. Tao J, Battle KC, Pan H, et al. Energetic basis for the molecular-scale organization of bone. Proc Natl Acad Sci USA. 2015;112:326-331. Yuan X, Serra RA, Yang S. Function and regulation of primary cilia and intraflagellar transport proteins in the skeleton. Ann N Y Acad Sci. 2015;1335:78-99. Zhong R, Ye ZH. Secondary cell walls: biosynthesis, patterned deposition and transcriptional regulation. Plant Cell Physiol. 2015;56: 195-214.
570
SECTION VIII n Cellular Adhesion and the Extracellular Matrix
APPENDIX 32.1
Genetic Defects of Cartilage and Bone Protein
Species
Mutation
Phenotype
Growth Factors BMP-4
Human
Overexpression
Fibrodysplasia progressiva; ectopic bone formation
BMP-5
Mouse
Null
Defective ears and sternum (short ear mutation)
CSF-1
Mouse
Null
Osteopetrosis; reduced osteoclasts (osteopetrotic mutation)
GDF-5 (TGF-β family)
Mouse
Null
Reduced size of long bones; no joints (brachypodism mutation)
Growth hormone
Human
Null
Reduced size of bones
OPG (osteoprotegerin)
Human
Null
Recessive juvenile Paget disease with excess bone remodeling
PTHrP
Human
Null
Reduced chondrocyte growth; epiphyseal plates fused at birth
RANKL
Mouse
Null
Osteopetrosis; no osteoclasts
Sclerostin
Human
Loss of function
Dense bones, van Buchem disease
Wnt1
Human
Loss of function
Osteoporosis
Signal Transduction Components c-Src
Mouse
Null
Osteopetrosis; osteoclasts fail to attach to or degrade bone
Connexin 43
Human
Point mutations
Dominant oculodentodigital dysplasia
FGF receptor 1
Human
Point mutation
Pfeiffer syndrome; cranial synostosis; long bone defects
FGF receptor 2
Human
Point mutation
Jackson-Weiss syndrome; cranial synostosis; long bone defects
FGF receptor 2
Human
Point mutation
Crouzon disease; cranial synostosis
FGF receptor 3
Human
Point mutation
Gain-of-function mutation; achondroplasia; short, wide bones
LRP5 Wnt coreceptor
Human
Loss of function
Osteoporosis-pseudoglioma syndrome
Transcription Factors c-fos
Mouse
Null
Osteopetrosis; no osteoclasts
hoxa-2
Mouse
Null
Deletion of the second branchial arch; duplication of first branchial arch
hoxd-13
Mouse
Null
Deletion fourth sacral derivatives; duplication third sacral derivatives
msx-1
Mouse
Null
Cleft palate
msx-2
Mouse
Null
Craniosynostosis (fusion of skull bones)
Runx2/Cbfa-1
Human Mouse
+/– Null
Dominant skeletal defects (cleidocranial dysplasia) No osteoblasts or bone
SOX9
Human
Point mutations
Dominant cartilage and skeletal defects (campomelic dysplasia)
Collagen and Other Structural Components of Cartilage and Bone Aggrecan
Mouse
Missense
Recessive cartilage deficiency; dwarfism; cleft palate
Cathepsin-K
Mouse
Deletion
Osteopetrosis
CLC7
Human
Point mutations
Osteopetrosis
COL1
Human
Missense, deletions
Dominant osteogenesis imperfecta; fragile bones
COL2
Human
Nonsense
Dominant Stickler syndrome; chondrodysplasia, eye defects
Human
Point mutations
Dominant chondrodysplasia and osteoarthritis of variable severity
Human
Splicing mutation
Defective cartilage with degeneration of knee joint
COL9A2 COL10A1
Human
Point mutations
Dominant Schmid metaphyseal chondrodysplasia with short bones
COL11A2
Human
Exon skipping
Dominant Stickler syndrome; chondrodysplasia, eye defects
Human
Point mutation
Recessive severe chondrodysplasia; deafness; cleft palate
Lysyl hydroxylase
Human
Point mutation
Bruck disease; fragile bones
Perlecan
Mouse
Deletion
Recessive defects in cartilage and bone formation
Proton ATPase
Human
Point mutations
Osteopetrosis
Sulfate transporter
Human
DTDST gene
Recessive cartilage defects; short limbs; joint deformation
ATPase, adenosine triphosphatase; BMP, bone morphogenetic protein; CSF, colony stimulating factor; FGF, fibroblast growth factor; PTHrP, parathyroid hormone–related protein; RANKL, receptor activator of nuclear factor κB ligand; TGF, transforming growth factor.
SECTION
Cytoskeleton and Cellular Motility
IX
This page intentionally left blank
SECTION IX OVERVIEW T
he seven chapters in this section cover the cytoskeleton and cellular motility. These topics are intimately related, because two of the protein polymers constituting the cytoskeleton—the internal scaffolding of the cell—are also tracks for motor proteins that power many cellular movements. Assembly and disassembly of the cytoskeletal polymers also produce some types of cell movements. Most organisms depend on motility to sustain life itself. Without a motile sperm the egg would not be fertilized. Without cellular motility a fertilized egg would not progress past the single-cell stage. Without active changes in cell shape and cellular migrations complex embryos would not form. Without cellular motility white blood cells would neither accumulate at sites of
Actin Ch 33
inflammation nor ingest invading microorganisms. Without active and rapid movements of organelles in axons and large plant cells the peripheral parts of these cells would not be nourished. Without muscle contractions we would be paralyzed and unable to move. Even a yeast, prevented from locomotion by its rigid cell wall, depends on internal movements for cell division and endocytosis. Many prokaryotes use rotary flagella for locomotion. Consequently, an understanding of the basis of cellular motility is central to our understanding of the functioning of all cells and organisms. This section starts with Chapters 33 to 35, which introduce the three cytoskeletal polymers, and Chapter 36, which explains the mechanisms of motor pro teins. Three concluding chapters show how cells use
Microtubules Ch 34
Motors Ch 36
Intermediate filaments Ch 35
Intracellular transport Ch 37
Muscle contraction Ch 39
Cellular motility Ch 38
573
cytoskeletal polymers and motors to produce a vast variety of movements: intracellular movements (Chapter 37); cell shape changes, cellular locomotion, and swimming (Chapter 38); and muscle contraction (Chapter 39). Mitosis and cytokinesis appear in our discussion of the cell cycle (Chapter 44). Actin filaments (Chapter 33) and microtubules (Chapter 34) have much in common, including their evolutionary origins in prokaryotes. Both assemble spontaneously into polymers that are used as tracks by molecular motors. The protein subunits of both polymers bind a nucleoside triphosphate: adenosine triphosphate (ATP) in the case of actin and guanosine triphosphate (GTP) for tubulin. After polymerization, hydrolysis of these bound nucleotides and dissociation of the γ-phosphate destabilize the polymer, much more so in the case of microtubules than in the case of actin filaments. Both polymers can turn over rapidly in cells or remain as stable components. Cells use many proteins to regulate the assembly of these polymers: Some proteins bind to the cytoplasmic pools of the subunit proteins; others initiate the assembly; some stabilize the polymers, others sever or depolymerize them; still others link the polymers together or to other cellular constituents. During divergent evolution from the common ancestor, contemporary organisms came to use actin filaments and microtubules for some of the same functions. For example, microtubules separate chromosomes in eukaryotes, while homologs of actin separate plasmids in bacteria. Actin filaments are tracks for long distance transport in plants, whereas microtubules serve the same purpose in animal nerve cells. Actin filaments and microtubules cooperate with a third polymer called intermediate filaments (Chapter 35) to form a cytoskeleton, which resists deformation and transmits mechanical forces. Microtubules are rigid, hollow reinforcing rods that sustain both compression and tension. These mechanical properties make microtubules useful for supporting asymmetrical cellular processes and for bidirectional traffic generated by the motor proteins kinesin and dynein. Actin filaments are more flexible, so they must be crosslinked into bundles to bear compression forces or support asymmetrical processes. High tensile strength allows actin filaments to bear forces produced by myosins. Intermediate filaments are flexible cables that have considerable tensile strength but little capacity to resist compression. Both intermediate filaments and actin filaments reinforce whole tissues by anchoring cadherins, transmembrane proteins used for cell-to-cell adhesion (see Chapter 30). Intermediate filaments prevent excessive stretching of cells by external forces in multicellular animals. If intermediate filaments are defective, tissues are mechanically fragile.
574
Most movements of eukaryotic cells depend on actin filaments and microtubules. Assembly and disassembly of actin filaments and microtubules produce force for several types of cellular movements (Chapter 37). Actin polymerization drives extension of pseudopods at the leading edge of motile cells. Hydrolysis of ATP bound to actin regulates recycling of subunits rather than being used directly to produce force. Growth of microtubules supports the extension of some asymmetrical cellular processes, including nerve cell processes. Many other cellular movements result from the physical movements of protein motors (Chapter 36) along actin filaments and microtubules in cytoplasm. Different motors move along these two polymers: myosins move on actin filaments, and dyneins and kinesins move along microtubules. These motors use energy released from the hydrolysis of ATP to take nanometer steps along their protein polymer tracks. These small steps apply force and move cargo attached to the motor. The cargo includes membrane-bound organelles, macromolecular complexes, and cytoskeletal polymers. Microtubule motors power most organelle movements in animal cells (Chapter 37), chromosomal movements during mitosis (Chapter 44), and beating of cilia and flagella (Chapter 38). The actin–myosin system is responsible for cytokinesis (see Chapter 44), some organelle movements (especially in plants and fungi [Chapter 37]), and muscle contraction (Chapter 39). Several motility systems do not depend on actin filaments or microtubules (Chapter 38). Nematode sperm use the reversible assembly of another protein to make pseudopods for their movements. Calciumsensitive contractile fibers cause rapid contractions of some protozoa. A proton or sodium ion gradient across the plasma membrane powers the rotary motor that turns bacterial flagella. Although not usually considered to be molecular motors, nucleic acid polymerases and helicases use ATP hydrolysis to move along polymers of DNA or RNA. The ability of actin filaments and microtubules to resist mechanical deformation and to transmit forces from motors allows the cytoskeletal-motility system to determine cell shape and hence the structure of both tissues and whole organisms. Furthermore, the dynamic nature of cytoskeletal polymers allows cells to change shape rapidly, in a time frame of seconds. Active extension of cellular processes and active changes in shape produce asymmetrical cell shapes. Movements of chromosomes during mitosis and organelles in cytoplasm determine the cellular distribution of these components that are otherwise too large to move by diffusion. Together with the extracellular matrix, the shapes of individual cells define the shapes of tissues and organs.
CHAPTER
33
Actin and Actin-Binding Proteins A
ctin filaments form cytoskeletal and motility systems in all eukaryotes (Fig. 33.1). Crosslinked actin filaments resist deformation, transmit forces, and restrict diffusion of organelles. A network of cortical actin filaments excludes organelles (Fig. 33.2C), reinforces the plasma membrane, and restricts the lateral motion of some integral membrane proteins. The cortex varies in thickness from a monolayer of actin filaments in red blood cells (see Fig. 13.11) to more than 1 µm in amoeboid cells (Fig. 33.2C). Like fingers in a glove, bundles of actin filaments support slender protrusions of plasma membrane called microvilli or filopodia (Fig. 33.2B). Microvilli expand the cell surface for transport of nutrients and participate in sensory processes, including hearing. The actin cytoskeleton complements and interacts physically with cytoskeletal structures composed of microtubules
A
B
C
(see Chapter 34) and intermediate filaments (see Chapter 35). Actin contributes to cell movements in two ways. First, polymerization and depolymerization of the network of actin filaments just inside the plasma membrane contribute to the extension of pseudopods, cell locomotion (Fig. 33.2D–E), and phagocytosis (see Fig. 22.3). Second, actin filaments are tracks for movements of the myosin family of motor proteins (see Fig. 36.7). Actin filaments and myosin filaments form the highly ordered, stable contractile apparatus of muscles (Fig. 33.3B; also see Fig. 39.3), as well as the transient contractile ring that helps separate the two daughter cells at the end of mitosis (Fig. 33.3A; also see Fig. 44.24). Myosins also power movements of membranes and other cargo along actin filaments, complementing
D
FIGURE 33.1 FLUORESCENCE MICROGRAPHS ILLUSTRATING THE DISTRIBUTION OF ACTIN FILAMENTS IN CELLS. A, Intestinal epithelial cells stained red with rhodamine-labeled phalloidin, a cyclic peptide that binds actin filaments. Actin filaments are concentrated in a band of microvilli bordering the intestinal lumen. Nuclei are stained blue with DAPI (4,6-diamidino-2-phenylindole). B, Cultured vascular smooth muscle cells. Actin filaments, stained red with a fluorescent antibody, are concentrated in stress fibers and in the cortex around the edges of these cells. C, Maize epidermis with actin filaments stained with rhodamine-labeled phalloidin in the cortex and in cytoplasmic bundles. D, Fission yeast Schizosaccharomyces pombe, stained with rhodamine-labeled phalloidin. Actin filaments are found in patches at the tips of growing cells and in the cleavage furrow of dividing cells. Scale bars are 10 µm. (A, Courtesy C. Rahner, Yale University, New Haven, CT. B, Courtesy I. Herman, Tufts Medical School, Boston, MA. C, Courtesy M. Frank, University of California, San Diego. D, Courtesy W.-L. Lee, Salk Institute, La Jolla, CA.)
575
576
A
SECTION IX n Cytoskeleton and Cellular Motility
B
C E
D
FIGURE 33.2 ELECTRON MICROGRAPHS OF ACTIN FILAMENTS. A, Filaments of purified actin prepared by negative staining. B, A thin section of an intestinal epithelial cell illustrating finger-like microvilli with tightly packed bundles of actin filaments linked to the surrounding plasma membrane by myosin-I. The barbed ends of these filaments (see Fig. 33.8) are located at the tips of the microvilli. C, A thin section of Acanthamoeba showing the actin filament meshwork in the cortex (the gray area here) beneath the plasma membrane. D, Fluorescence micrograph of a cultured fish scale keratocyte fixed while actively migrating toward the top of the figure. Actin filaments are stained blue with phalloidin and myosin II stained red with antibodies. E, Electron micrographs of actively migrating fish scale keratocytes fixed and extracted to remove the plasma membrane and soluble components prior to coating the cytoskeleton with platinum. A meshwork of branched filaments concentrates near the leading edge with longer, unbranched filaments deeper in the cytoplasm. Most filaments are oriented with their barbed ends forward. (A, Courtesy U. Aebi, University of Basel, Switzerland. B, Courtesy M. Mooseker, Yale University, New Haven, CT. D–E, Courtesy T. Svitkina and G. Borisy, University of Wisconsin, Madison.)
organelle movements along microtubules powered by other motors (see Fig. 37.1). Actin, myosin, and accessory proteins form intracellular bundles called stress fibers (Fig. 33.1B) that apply tension between adhesive junctions on the plasma membrane (see Fig. 30.11), where cells attach to each other or to the extracellular matrix. Stress fibers are prominent in tissue culture cells grown on glass or plastic and in endothelial cells lining major arteries. Actin is often the most abundant protein in eukaryotic cells, composing up to 15% of total protein, and the many types of actin-binding proteins may account for another
10% of cellular protein. In muscle, actin and myosin constitute more than 60% of the total protein. Given this abundance, it is curious that actin was discovered in muscle only in the 1940s and in nonmuscle cells in the late 1960s. Since the 1970s, scientists have discovered new actin-binding proteins every year, but the inventory is probably still incomplete. Genetic defects in components of the actin cytoskeletal and motility system cause many human diseases, including muscular dystrophy (see Table 39.3), hereditary fragility of red blood cells (ie, hemolytic anemias, see Fig. 7.11), and hereditary heart diseases called cardiomyopathies (see Table 39.1).
CHAPTER 33 n Actin and Actin-Binding Proteins
Actin Molecule Actin is folded into two domains that are stabilized by an adenine nucleotide lying in between (Fig. 33.4). The polypeptide of 375 residues crosses twice between the two domains, with the N- and C-termini located near each other. The two domains are folded similarly,
A
B
577
suggesting that the actin gene arose by duplication of an ancestral gene. Actin binds adenosine triphosphate (ATP) or adenosine diphosphate (ADP) and a divalent cation, Mg2+, in cells, with nanomolar affinity. Actin binds ATP with higher affinity than ADP, so given the higher concentration of ATP in cells, unpolymerized actin is saturated with ATP. The bound nucleotide exchanges relatively slowly with nucleotide in the medium (Fig. 33.11). Different actin monomer–binding proteins can inhibit or promote nucleotide exchange. Bound nucleotide stabilizes the molecule but is not required for polymerization in vitro. However, ATP-actin and ADP-actin polymerize at different rates. Posttranslational modifications of actins include acetylation of the N-terminus and (in most cases) methylation of histidine-68. In some insect flight muscles, the small protein ubiquitin (see Fig. 23.2) is attached covalently to approximately one in six actin molecules, yielding a 55-kD polypeptide that is incorporated with unmodified actin into filaments. Some invertebrate actins are phosphorylated on tyrosine-211. The functional significance of these modifications is still being investigated.
C
FIGURE 33.3 MICROGRAPHS OF CONTRACTILE BUNDLES OF ACTIN FILAMENTS. A, Fluorescence micrograph of a dividing normal rat kidney cell stained with fluorescein-phalloidin to mark actin filaments in the contractile ring of the cleavage furrow. The drawing illustrates the filaments in the contractile ring. B, Fluorescence micrograph of a myofibril isolated from skeletal muscle and stained with fluorescein-phalloidin to label actin filaments (green) and rhodamineantibody to α-actinin to label Z disks (yellow). C, Electron micrograph of a thin section of skeletal muscle. (A, Micrograph courtesy Y.-L. Wang, University of Massachusetts, Worcester. B, Courtesy V. Fowler, Scripps Research Institute, La Jolla, CA. C, Courtesy H.E. Huxley, Brandeis University, Waltham, MA.)
Evolution of the Actin Family The actin gene is ancient, with roots before the universal common ancestor. The primordial gene apparently encoded a nucleotide-binding protein and gave rise to actin, the glycolytic enzyme hexokinase (see Fig. 3.12) and the heat shock protein Hsc70. The three proteins have similar folds with a central ATP binding site, but different functions. The actin genes diversified extensively in prokaryotes giving rise to families of genes for actins with distinct functions: MreB involved with cell shape, Pointed end
4 2
ATP
3 C
A
1
N
B
Barbed end
FIGURE 33.4 STRUCTURE OF ACTIN. A, Ribbon model showing the polypeptide fold and the location of Mg-ATP (magnesium–adenosine triphosphate), shown as space-filling. Numbers 1 to 4 indicate the four subdomains. B, Surface rendering. ATP is almost completely buried in the cleft between the two lobes of the protein. The barbed end of the molecule (this is a way to describe the polarity of the actin filament; see Fig. 33.8) is at the bottom in this orientation. (For reference, see Protein Data Bank [PDB; www.rcsb.org] file 1ATN and Kabsch W, Mannherz HG, Suck D, et al. Atomic structure of the actin-DNase I complex. Nature. 1990;347:37–44.)
578
SECTION IX n Cytoskeleton and Cellular Motility
FtsA that participates in cytokinesis, ParM that separates plasmids, and more than 30 other actin-like proteins expressed from plasmids or by bacteriophages. The original eukaryotic actin gene came from an archaeal cell related to Lokiarchaeota (see Fig. 2.4B). Eukaryotic actin genes are highly conserved, likely because of constraints imposed by the interactions required to form polymers and numerous regulatory proteins. However, through divergent evolution, the genes encode subtly different proteins, some with novel functions. Most organisms have multiple actin genes, and all known actin isoform diversity arises from multiple genes rather than from alternative splicing of messenger RNAs (mRNAs). Humans have six actin genes; Dictyostelium has more than 10; but budding yeast has only one. Muscle actin genes diverged from cytoplasmic actins in primitive chordates (see Fig. 2.8). To fulfill special developmental functions, plant actin genes diverged among themselves more than animal actin genes. The biochemical similarities of eukaryotic actin isoforms are more impressive than their differences. The sequences of pairs of actins are generally more than 90% identical, even between highly divergent eukaryotes. Humans express β and γ isoforms in nonmuscle cells and four different α and β isoforms in various muscle cells. Many nonmuscle cells express both the β and γ isoforms, but red blood cells use only β-actin. In every case examined, actin isoforms copolymerize in the test tube, so it is remarkable and still unexplained that cells can sort actin isoforms into different structures. For example, β-actin is concentrated near the plasma membrane of cultured cells, whereas γ-actin is concentrated in stress fibers (Fig. 33.5). In muscle, α-actin forms the thin filaments of the contractile apparatus, whereas γ-actin localizes around mitochondria.
Worm
Actin Other families of Arps
Yeast Human Amoeba
Actin-Related Proteins Genes for actin-related proteins (Arps) diverged from actin genes before the last eukaryote common ancestor (Fig. 33.6). Arps share with actin the fold of the polypeptide chain and residues forming the nucleotide-binding site, but fewer than 60% of the overall residues are
FIGURE 33.5 SORTING OF ACTIN ISOFORMS IN CELLS. Fluorescence micrograph of cultured cells doubly stained with fluorescent antibodies specific for β-actin concentrated at the leading edge (orange) and γ-actin concentrated in stress fibers (green). Nuclei are stained blue with DAPI (4,6-diamidino-2-phenylindole). (Courtesy I. Herman, Tufts Medical School, Boston, MA.)
Human Fly
Worm Yeast
Arp1 (Dynactin complex)
Fly Arp53
Yeast ACT3
Amoeba
Fly Arp13E
Fly Worm Yeast
Yeast Amoeba Fly
Arp3
Cow Worm (Arp2/3 complex)
Arp2
FIGURE 33.6 COMPARISON OF ACTIN AND ACTIN-RELATED PROTEINS. Space-filling models of actin and Arps showing residues identical to actin (yellow), conservative substitutions compared with actin (green), nonconservative substitutions (blue), and insertions (red). The phylogenetic tree, based on sequence comparisons, shows divergence from a common ancestor. Arps, Actin-related proteins. (Modified from Mullins RD, Kelleher JF, Pollard TD. Actin’ like actin. Trends Cell Biol. 1996;6:208–212.)
579
CHAPTER 33 n Actin and Actin-Binding Proteins
identical to actin. Divergent surface residues allow Arps to participate in molecular interactions different from actin. Arp1 forms a short filament as part of the dynactin complex that promotes cargo movement by the microtubule motor dynein (see Fig. 37.2). Arp2 and Arp3 are two of seven subunits in a protein complex that nucleates branched actin filaments in the cell cortex (Fig. 33.12). Eight additional types of Arps are widespread in eukaryotes. Several participate in complexes that regulate chromatin structure.
Actin Polymerization Actin filaments are polarized, owing to the uniform orientation of the asymmetrical subunits along the polymer (Fig. 33.7). The subunits, all pointed in the same direction, form a double helix with the subunits in the two strands staggered by half. One end is called the barbed end, the other is called the pointed end. This nomenclature arises from the asymmetrical arrowhead pattern seen when myosin heads bind along the length of actin filaments (Fig. 33.8). Actin self-assembles into filaments by means of a series of bimolecular reactions (Fig. 33.9; see also Fig. 5.6). Actin is isolated from cells as a monomer at low salt concentrations. Physiological concentrations of monovalent and divalent cations bind to low-affinity sites on actin and promote polymerization. In vitro, actin trimers appear to be the nucleus that initiates polymer growth.
The reactions required to form trimers are very unfavorable in comparison with reactions for elongation of polymers larger than trimers. To initiate new filaments, cells use regulatory proteins to overcome these unfavorable nucleation reactions. Actin filaments grow and shrink by the addition and loss of subunits at the two ends of the polymer. The reactions at the two ends have different rate constants (Fig. 33.8). Subunit association is a diffusion-limited reaction (see Chapter 4) at the rapidly growing barbed end and somewhat slower at the pointed end. Subunit dissociation is relatively slow at both ends, between 0.3 and 8 subunits per second. The rates of these reactions depend on the nucleotide bound to the monomer associating with or dissociating from a filament. In the presence of ATP, purified actin assembles almost completely, leaving as monomers the critical concentration of ~0.1 µM ATP-actin. The critical concentration is the monomer concentration at which equal rates of association and dissociation occur, 1.4 s−1 at the barbed end (see Fig. 5.6). The critical concentration
ATP
T
Pointed end growth
1.3
0.8
D
0.3
0.16
K = 0.6 µM
K = 2.0 µM
K = 0.12 µM
K = 2.0 µM
Fitted monomers
Decorated seed
Barbed end growth
100 nm
EM Density
1.4
A
B
8 12
T
A C
D
FIGURE 33.7 STRUCTURE OF THE ACTIN FILAMENT. A, Electron micrograph of a negatively stained actin filament. B, Reconstruction of the actin filament from electron cryomicrographs at 0.66 nm resolution. C, Surface rendering of the molecular model. Subunits in the two long-pitch helices are shown as yellow-orange and blue-green (see Fig. 5.5 for nomenclature). The short pitch helix, including every subunit, follows a yellow-green-orange-blue pattern. D, Scale drawing used throughout this text. (B, Data from Fujii T, Iwane AH, Yanagida T, Namba K. Direct visualization of secondary structures of F-actin by electron cryomicroscopy. Nature. 2010;467:724–728.)
ATP
4 D
B
FIGURE 33.8 ACTIN FILAMENT ELONGATION. A, Electron micrograph of growth from an actin filament “seed” decorated with myosin heads to reveal the polarity. Growth is faster at the barbed end than at the pointed end. B, Rate constants for association (units: µM−1 s−1) and dissociation (units: s−1) for Mg-ATP-actin (T) and MgADP-actin (D) were determined by measuring the rate of elongation at the two ends as a function of monomer concentration. Ratios of the rate constants yield critical concentrations (K, units: µM) for each reaction. Critical concentrations at the two ends are the same for ADP-actin but differ for ATP-actin. ADP, adenosine diphosphate. (A, Courtesy M. Runge, Johns Hopkins Medical School, Baltimore, MD. B, Data from Pollard TD. Rate constants for the reactions of ATP- and ADP-actin with the ends of actin filaments. J Cell Biol. 1986;103:2747–2754.)
580
SECTION IX n Cytoskeleton and Cellular Motility
A. Actin filament nucleation 10
10
10
~106
~103
1
B
P
B. ATP hydrolysis ATP
Barbed end
ADP + Pi
ATP hydrolysis
Phosphate dissociation
ADP
t1/2 = 2 s
Seed
Pointed end
t1/2 = 350 s
FIGURE 33.9 ACTIN FILAMENT NUCLEATION, GROWTH, AND NUCLEOTIDE HYDROLYSIS. A, Nucleation. Formation of dimers and trimers is very unfavorable, owing to rapid dissociation of subunits. Actin trimers are called nuclei, because they initiate the highly favorable elongation reactions. Rate constants estimated by kinetic modeling have units of µM−1 s−1 for association reactions and s−1 for dissociation reactions. B, ATP hydrolysis by a polymer of ATP-actin (yellow subunits) is random and irreversible at a rate of 0.3 s−1, yielding subunits with bound ADP and inorganic phosphate (orange). Phosphate (Pi) dissociates slowly at a rate of 0.002 s−1, converting half of the newly polymerized subunits to ADP-actin (pink) in 6 minutes. ADP bound to polymerized subunits does not exchange with nucleotide in the medium. Phosphate binding is reversible, but the affinity is low, so most subunits dissociate phosphate. (For reference, see Pollard TD, Blanchoin L, Mullins RD. Biophysics of actin filament dynamics in nonmuscle cells. Annu Rev Biophys Biomol Struct. 2000;29: 545–576.)
for ADP-actin is approximately 20 times higher than for ATP-actin. Hydrolysis of bound ATP and dissociation of the γ-phosphate after assembly modify the behavior of actin filaments, including their affinity for regulatory proteins. Following incorporation of an ATP-actin subunit into a filament, bound ATP is hydrolyzed irreversibly to ADP and phosphate with a half-time of 2 s (Fig. 33.9). These ADP-Pi subunits behave much like ATP subunits. Phosphate dissociates slowly and reversibly over several minutes. This yields filaments with a core of subunits with tightly bound ADP. At the millimolar concentrations of phosphate in cytoplasm, phosphate is bound to some ADP-actin subunits. The critical concentrations for ATP-actin differ at the two ends of the filament. This results from differences in the probability of nucleotide hydrolysis and phosphate release at the two ends, which is more likely to expose ADP-subunits at the pointed end. At steady state in the presence of ATP, the actin monomer concentration falls between the critical concentrations at the
two ends. Though the polymer and monomer concentrations remain constant, net addition of subunits at the barbed end and net loss of subunits at the pointed end result in the slow migration of subunits through the polymer from the barbed end to the pointed end. This slow process is called treadmilling. Neither end exhibits rapid fluctuations in length like those of microtubules (see Fig. 34.6).
Actin-Binding Proteins In contrast to the slow turnover of filaments of purified actin at steady state, actin filaments in live cells can polymerize and depolymerize rapidly. This requires the control of more than 60 families of actin-binding proteins (Fig. 33.10 and Appendix 33.1). Broadly, these proteins fall into families that bind monomers, sever filaments, cap filament ends, nucleate filaments, promote polymerization, crosslink filaments, stabilize filaments, or move along filaments. Like actin, genes for many of these actin-binding proteins arose in early eukaryotes and are found in protozoa, yeast, plants, and vertebrates. This section introduces examples of each class of actin-binding protein. However, no actin-binding protein functions in isolation, so the chapter concludes by explaining how ensembles of these proteins work together to regulate actin filament dynamics in cells.
Actin Monomer–Binding Proteins Proteins that bind actin monomers cooperate with capping proteins to maintain a pool of unpolymerized actin in cells and regulate the nucleotide bound to actin. Profilins are abundant proteins, found in all branches of the eukaryotic tree. Cells require three different profilin activities for viability: binding actin monomers, catalyzing the exchange of nucleotides bound to actin (Fig. 33.11) and binding to polyproline sequences on other protein such as formins (Fig. 33.14). Profilins also bind membrane polyphosphoinositides. The nucleotide bound to actin monomers determines the affinity for profilin: highest for nucleotide-free actin monomers, followed by ATP-actin and ADP-actin. Profilins bind to the barbed end of actin monomers. This sterically blocks nucleation and pointed end elongation but not association of the profilin-actin complex with the barbed end of filaments. Profilin dissociates rapidly after the profilinactin complex binds to a barbed end. Found only in vertebrates, β-thymosins are extended peptides of just 43 residues. α-Helices at both ends block the barbed and pointed ends of the actin monomer. They bind ATP-actin monomers with higher affinity than ADPactin monomers and inhibit both actin polymerization and nucleotide exchange. Thymosin-β4 is the most abundant actin-binding protein in some animal cells, where it sequesters most of the unpolymerized actin.
CHAPTER 33 n Actin and Actin-Binding Proteins
Monomer binding
Monomers
Elongation by formin
Polymerization
Elongation and capping
Branching
581
Arp2/3 complex
Crosslinking
Long filament
Bundle
Network
Severing Capped filaments
Profilin-actin on FH1 domain
Annealing
Short filaments
FIGURE 33.10 FAMILIES OF ACTIN-BINDING PROTEINS. Summary of the reactions of regulatory proteins with actin monomers and filaments.
Profilin
ADF/cofilin Thymosin
Profilin
ADP
ATP
ADP
ATP
Inhibiting
Profilin
Promoting
Nucleotide-free
ATP
FIGURE 33.11 REGULATION OF ACTIN NUCLEOTIDE EXCHANGE BY ACTIN-BINDING PROTEINS. The rate-limiting step is dissociation of the bound nucleotide from the nucleotide cleft. ADF/ cofilins and β-thymosins inhibit dissociation of both ATP and ADP. Profilin competes with both inhibitors for binding actin monomers and increases the rates of both nucleotide dissociation and binding. The ability of profilin to promote nucleotide exchange, the higher affinity of actin for ATP than for ADP, and the higher concentration of ATP than ADP in cytoplasm drive the reactions to the right, so essentially all unpolymerized actin in cells has bound ATP.
Members of the ADF/cofilin family bind ADP-actin monomers with higher affinity than ATP-actin and inhi bit nucleotide exchange but not polymerization. As explained below, their main function is to sever ADPactin filaments (Fig. 33.16). These proteins are essential for the viability of many eukaryotes.
Actin Filament Nucleating Proteins Spontaneous nucleation of filaments from actin monomers is intrinsically unfavorable and inhibited by profilin, so cells use proteins to specify when and where new filaments form. The best characterized are Arp2/3 complex, formins and proteins with multiple WH2 (WASp homology 2) domains such as spire. Each of these has a different evolutionary origin, mechanism of action, and physiological functions. Arp2/3 complex consists of Arp2 and Arp3 tightly bound to five novel proteins (Fig. 33.12). Arp2/3 complex
binds to the side of actin filaments and forms branches with the Arps as the first two subunits at the pointed end of the new filament. Growth of the free barbed ends of these branches produces force that pushes the plasma membrane forward at the leading edge of motile cells (Fig. 33.2E). Arp2/3 complex is intrinsically inactive, but is stimulated to initiate a branch by interactions with actin filaments and nucleation-promoting factors such as WASp (Wiskott-Aldrich syndrome protein). The name comes from Wiskott-Aldrich syndrome, a genetic disorder with defects in blood cells causing immunodeficiency and bleeding. Other tissues use a closely related nucleationpromoting factor called N-WASP. The C-terminal VCA (verprolin homology, connecting, and acidic) motifs of WASp activate Arp2/3 complex. The V motif (also called WH2 for WASp homology 2) binds an actin monomer and the CA motif binds to two sites on Arp2/3 complex (Fig. 33.13). These interactions promote the binding of Arp2/3 complex to the side of a filament, bringing about a large conformational change that positions Arp3 and Arp2 to initiate a branch. Signaling molecules overcome intramolecular interactions that autoinhibit WASp and N-WASP (Fig. 33.13); Rho-family GTPases (guanosine triphosphatases) and membrane polyphosphoinositides bind near the middle of the protein, and proteins with SH3 domains bind proline-rich sequences. Additional nucleation promoting factors with VCA motifs regulate Arp2/3 complex in specific cellular locations. For example, Scar/WAVE is active at the leading edge of motile cells, WASH participates in membrane traffic from endosomes and WHAMM is associated with the Golgi apparatus. The small GTPase Rac regulates Scar/WAVE and WASH by freeing them of inhibition by homologous regulatory complexes of four proteins. Spire was discovered in Drosophila as a gene required for the development of eggs and embryos and later found in other metazoans but not fungi or protozoa.
582
SECTION IX n Cytoskeleton and Cellular Motility A. Arp2/3 complex
B. Model of branch
Arp3
p34
p21
4
D2 D1
ArpC2
ArpC4
1
2
2
3 p20
3
Arp3
3
2
4
ArpC5
1
4
3 1
90°
Arp2
ArpC3
ArpC1 Arp2
1
p16 p40
B
n
70˚
tio
1 2 WASp/Scar nucleation promoting factors bind an actin monomer and then Arp2/3 complex
Daughter filament
ga
Arp2/3 complex
B
3
4
on
WASp/ Scar
Binding to mother filament activates polymerization Mother filament
El
C. Branching pathway
P
FIGURE 33.12 NUCLEATION OF BRANCHED ACTIN FILAMENTS BY ARP2/3 COMPLEX. A, Ribbon diagram of the crystal structure of Arp2/3 complex. The seven subunits are color coded and labeled. Numbers label the subdomains of Arp2 and Arp3. B, Model of a branch based on a three-dimensional reconstruction from electron micrographs. C, Steps in branch formation: (1) A nucleation-promoting factor binds an actin monomer. (2) These binary complexes bind two sites on Arp2/3 complex, bringing together two actin subunits with Arp2 and Arp3. (3) The ternary complex binds to the side of an actin filament, completing the activation process. (4) A new daughter filament grows at its barbed end from the side of the older mother filament. B is the barbed end. P is the pointed end. WASH, Wiskott-Aldrich syndrome protein (WASP) and Scar homolog; WHAMM, WASP homologue associated with actin, membranes, and microtubules. (A, For reference, see PDB file 1K8K and Robinson R, Turbedsky K, Kaiser DA, et al. Crystal structure of Arp2/3 complex. Science. 2001;294:1679–1684. B, From Rouiller I, Xu X-P, Amann KJ et al. The structural basis of actin filament branching by the Arp2/3 complex. J Cell Biol. 2008;180:887–895.)
WASp domain map Ligands: Domains:
N
WIP EVH1
PIP2
PIP2 Cdc42 Basic GBD
Arp2/3 Src, Btk Nck, Grb2 Actin complex C Proline-rich V C A
Activated WASp
Cdc42
Actin + Arp2/3 complex GBD
Autoinhibited WASp
Actin filament branch
FIGURE 33.13 WASP (WISKOTT-ALDRICH SYNDROME PROTEIN) NUCLEATION PROMOTING FACTOR. The V (verprolin homology) motif binds an actin monomer, and the CA (connecting and acidic) motifs bind Arp2/3 complex. Intramolecular association of C with the GBD (GTPase-binding domain) autoinhibits the protein. Membrane-bound Rho-family guanosine triphosphatases (GTPases) and polyphosphoinositides compete C from GBD, releasing VCA to interact with actin and Arp2/3 complex. SH3 domain proteins such as Nck also activate WASp by binding the proline-rich domain. The N-terminal EVH1 (Ena-VASP [vasodilator-stimulated protein] homology) domain binds WIP (WASp interacting protein).
Spire proteins have multiple WH2 domains (corresponding to the V domains in Fig. 33.13) that bind formins, actin monomers, and nucleate unbranched filaments. One function is to build a meshwork of actin filaments in the oocyte cytoplasm. Other proteins with WH2 domains also nucleate filaments.
Actin Filament Polymerases Formins not only initiate unbranched actin filaments but remain associated with the elongating barbed end and promote its growth. Filaments produced by formins are incorporated into the contractile ring and bundles in yeast (see Fig. 37.11) and stress fibers in animal cells (Fig. 33.1B). The three formins in fission yeast each assemble actin filaments for specific structures, cytokinetic contractile ring, interphase actin cables, or mating structures. The specific functions of the 15 different mammalian formins are still being investigated. The characteristic feature of formins is a formin homology-2 (FH2) domain (Fig. 33.14) that forms a doughnut-shaped homodimer around the barbed end of an actin filament. Interactions of the FH2 dimer with two actin monomers nucleate a filament. Then the FH2 domain tracks the growing barbed end by faithfully
CHAPTER 33 n Actin and Actin-Binding Proteins
A. Formin domain map
B. Formin FH2 domain model
N
Regulatory domain
FH1
FH2
Lasso
C1205
Linker
GTPases
Profilin
Actin
αC
Knob
C. Formin elongation mechanism FH2 binds two subunits to nucleate filament Actin
D. VASP domain map PKA phosphorylation site Actin monomers Actin filaments Self association C380 EVH2 CC
N
FH1 attaches here
C
αT
αD
583
Profilin & SH3 domains Proline-rich ligands N EVH1 Proline-rich
αM Post
Coiled-coil
E. VASP elongation mechanism
Profilin-actin bound to FH1 transfers rapidly to barbed end; FH2 steps onto new subunit
VASP binds near end of filament; profilin-actin bound to VASP transfers to barbed end (details under investigation)
Profilin
Actin Pointed
FIGURE 33.14 ACTIN FILAMENT POLYMERASES. A–C, Formins. A, Domain structure of a generic, dimeric formin similar to mouse mDia1. Binding of Rho-family guanosine triphosphatases (GTPases) activates formins by disrupting intramolecular associations that autoinhibit the formin homology-2 (FH2) domains. B, Ribbon diagram of the structure of the homodimer of FH2 domains of budding yeast Bni1p. The linker segments extend when the dimer fits around an actin filament. C, Pathway of actin filament nucleation and elongation by formin FH1 and FH2 domains. Multiple polyproline sequences in FH1 bind profilin-actin for rapid transfer to the barbed end of the filament. D, VASP (vasodilator-stimulated phosphoprotein). The EVH1 domain (see Fig. 25.11B) binds polyproline ligands, including vinculin and zyxin, in focal contacts; the proline-rich domain binds profilinactin complexes; the EVH2 domain binds both actin monomers and filaments; and the C-terminal coiled-coil (CC) mediates the formation of VASP tetramers. E, The diagram shows a VASP tetramer delivering profilin-action onto the barbed end of the filament. (B, See PDB file 1UX5.)
stepping onto each new subunit as it is added. The FH2 domain also protects the barbed end from capping proteins. Polyproline sequences in the flexible FH1 domain next to FH2 bind multiple profilin-actin complexes that transfer rapidly to the barbed end, enhancing the rate of elongation. Some formins are autoinhibited by interactions between parts of the polypeptide flanking the FH1-FH2 domains. Rho family GTPases activate formins by overcoming this autoinhibition. An unrelated protein called VASP (vasodilatorstimulated phosphoprotein) also stimulates the growth of actin filament barbed ends and protects them from capping in filopodia and at the leading edge of motile cells (Fig. 33.14D–E). VASP tetramers bind near the barbed end of actin filaments, allowing their proline-rich domains to bind and deliver profilin-actin complexes to the growing barbed end, similar to formins. EVH1 domains allow VASP and related proteins to bind prolinerich ligands in focal adhesions and on the surface of the bacterium Listeria where it stimulates actin filament elongation.
Actin Filament–Capping Proteins Capping proteins bind to and stabilize either the barbed or pointed end of actin filaments (Fig. 33.15). Some capping proteins also stimulate the formation of new filaments and/or sever actin filaments. Gelsolins consist of six domains with similar folds but different sequences and activities. They bind tightly to the sides and barbed ends of actin filaments, blocking
Tropomodulin + tropomyosin
Gelsolin Fragmin / severin Barbed Capping protein
Pointed Arp2/3 complex
Capping protein
Tropomodulin Tropomyosin
FIGURE 33.15 ACTIN FILAMENT–CAPPING PROTEINS. Interactions of capping proteins with the ends of actin filaments. Most of these proteins bind with high affinity to a filament end. Tropomodulin requires tropomyosin for high-affinity binding. (Roberto Dominguez of the University of Pennsylvania created the PDB file used for the spacefilling model of the capped filament with tropomyosin.)
both the association and dissociation of actin subunits. Gelsolin also binds actin dimers, forming a nucleus that grows at the pointed end. Fragmin and severin, are similar in structure and function to the first three domains of gelsolin. Genes for these capping proteins, found widely among eukaryotes, are likely to have duplicated during evolution to give rise to gelsolin genes. Heterodimeric capping proteins consist of two homologous subunits of ~30 kD. They cap barbed ends
584
SECTION IX n Cytoskeleton and Cellular Motility
with high affinity and promote nucleation of new pointed ends by stabilizing small actin oligomers. Heterodimeric capping proteins are found in most eukaryotic cells. In striated muscle, they cap the barbed end of actin filaments in the Z disk (see Fig. 39.5). Several different proteins regulate capping protein by interfering directly or indirectly with its binding to barbed ends. Tropomodulin caps the pointed end of stable actin filaments in muscle, red blood cells, and other animal cells. High-affinity binding to pointed ends requires tropomyosin, an α-helical protein that binds along the length of actin filaments (Fig. 33.15).
Actin Filament–Severing Proteins Four classes of proteins just introduced—ADF/cofilin, some formins, fragmin/severin, and gelsolin—also sever actin filaments into short fragments (Fig. 33.16). ADF/ cofilins bind ADP-actin subunits in filaments and promote severing and depolymerization. A few formin isoforms can sever actin filaments. Ca2+ triggers gelsolin, fragmin, and severin to sever actin filaments. Domain 2 of gelsolin binds to the side of an actin filament, positioning domain 1 to bind between subunits and disrupt the filament. One gelsolin isoform is found inside cells; another is secreted into blood plasma, where it may sever actin filaments released from damaged cells. Proteins That Bind the Side of Actin Filaments Tropomyosin, nebulin, and caldesmon are extended proteins that bind along the sides of actin filaments. Tropomyosin increases the tensile strength of actin filaments in striated muscles, it is also an essential component of the Ca2+-sensitive regulatory machinery that controls the interaction of myosin and actin (see Fig. 39.4). Nebulin is one factor that determines the length of the actin filaments in skeletal muscle (see Chapter 39). Caldesmon, together with tropomyosin and Ca2+calmodulin, regulates interaction of actin and myosin in smooth muscle (see Fig. 39.24) and nonmuscle cells. Actin Filament Crosslinking Proteins Possession of two actin filament-binding sites enables crosslinking proteins (Fig. 33.17) to bridge filaments and to stabilize higher-order assemblies of actin filaments. Some have a greater tendency to crosslink filaments in regular bundles, like those in microvilli (Fig. 33.2B), but depending on protein concentrations and filament lengths, most of these proteins can form both random networks of crosslinked filaments and regular bundles of filaments. Many of these proteins have actin binding domains consisting of two calponin-homology domains. α-Actinin is found in the cortical actin network, at intervals along stress fibers, on the cytoplasmic side of cell adhesion plaques (see Fig. 30.11), and in the Z-disk of striated muscles (see Fig. 39.5). Ca2+ binding to EF hands (see
Polymerized actin subunits: ADP + Pi
Cofilin
ADP
Cofilin binds ADP-actin filament
Severing
Both new ends free
FIGURE 33.16 ACTIN FILAMENT SEVERING BY COFILIN. ADF/cofilins bind to ADP-actin filaments and stabilize a rare, but naturally occurring, tighter helical twist. This destabilizes and severs the filament. The products are two uncapped ends that are available for subunit association and dissociation. Pi, phosphate. (For reference, see Elam WA, Kang H, De la Cruz EM. Biophysics of actin filament severing by cofilin. FEBS Lett. 2013;587:1215–1219.)
Chapter 26) of some α-actinins inhibits binding to actin. Fimbrin and villin (a relative of gelsolin with an extra actin-binding site) stabilize the regular actin filament bundles in microvilli. Filamin crosslinks filaments in the cortex of many cells and also anchors these filaments to an integrin, a plasma membrane receptor for adhesive glycoproteins (see Fig. 30.11). Actin filament crosslinking proteins of the plasma membrane skeleton, such as spectrin (see Fig. 7.11) and dystrophin (see Fig. 39.9), are anchored to integral membrane proteins.
Functional Redundancy of Actin-Binding Proteins The diversity and the apparent redundancy of actinbinding proteins are striking. Why should most organisms retain genes for 60 or more actin-binding proteins if they have such a limited repertoire of functions: monomer binding, nucleation, capping, severing, crosslinking, stabilizing, and motility? Null mutations show that some proteins are essential for normal physiology. These include myosin-II, Arp2/3 complex, profilin, and cofilin in yeast. On the other hand, organisms can survive genetic deletion of some actin-binding proteins, suggesting that parts of the system are redundant. For example, in a laboratory environment, Dictyostelium tolerates the loss of crosslinking proteins (α-actinin or
CHAPTER 33 n Actin and Actin-Binding Proteins
Class I
β-Sheet domain
Fimbrin/plastin
EF hand
Triple-stranded α-helical domain
Actin-binding domain
ABD ABD
585
Class II α-Actinin
2-fold symmetry
ABD ABD
40 nm
Ankyrin
Spectrin/fodrin
20 nm
Dimer β-Subunit α-Subunit
ABD
95–125 nm Dystrophin ABD
Class III ABP-120
Filamin/ABP
ABD
ABD ABD
GPIB/IX binding domain
35 nm ABD
80 nm FIGURE 33.17 Actin filament crosslinking proteins sharing homologous actin-binding domains (red). Crosslinking requires two actin-binding sites, which can be part of one polypeptide (fimbrin) or on different subunits of dimeric proteins (α-actinin, filamin). Dystrophin has a second actin-binding site in the middle of the tail of triple helical repeats. (Modified from Matsudaira P. Modular organization of actin cross-linking proteins. Trends Biochem Sci. 1991;16:87–92.)
ABP-120), severin, or one of two profilin genes with only minor defects in behavior and growth. Humans who lack dystrophin develop and grow normally for a few years but later succumb to muscle wasting (see Table 39.3). Mice without their single gelsolin gene reproduce normally, with only mild defects in platelets and other cells. These observations suggest that each actin-binding protein has a distinct function, conferring a small selective advantage. Multiple proteins sharing overlapping functions make the actin system relatively failsafe so that it is difficult to detect the phenotypic consequences of the loss of particular proteins in mutant animals. Alternatively, these proteins may be retained owing to unknown functions distinct from actin binding.
Actin Dynamics in Live Cells Cellular actin filaments vary widely in stability. Filament lifetimes are seconds in motile cells such as amoebae and white blood cells, tens of minutes in stress fibers and microvilli, and days in red blood cells and striated muscles where capping protein on the barbed end and tropomodulin on the pointed end limit the exchange of subunits and tropomyosin provides mechanical stability (see Figs. 7.11, 33.15, and 39.4). Biochemical and genetic experiments identified the proteins that orchestrate actin filament assembly and
disassembly. Essential features are a pool of unpolymerized actin, mechanisms to initiate and terminate new filaments and control disassembly. Reconstitution of the actin filament comet tail of the intracellular bacterium Listeria (see Fig. 37.12) showed that the minimal requirements for rapid assembly and turnover are actin, ADF/ cofilin, Arp2/3 complex, a capping protein, and profilin, in addition to a nucleation promoting factor on the surface of the bacterium.
Methods to Document Actin Filament Turnover The original approach to observe actin turnover was treatment with drugs that bind actin monomers such as cytochalasin or latrunculin (Box 33.1). Neither disassembles actin filaments directly, but both interfere with assembly from monomers, so their effects reveal if filaments are turning over naturally. Muscle actin filaments are relatively resistant to these drugs, but they disrupt the cortical networks of actin filaments in other cells in seconds (Fig. 33.18). When the drug is removed, the cortical actin network reforms rapidly from the leading edge. This response indicates that many cellular filaments turn over rapidly. Much more has been learned about actin filament turnover by light microscopy of live cells. Although individual actin filaments are not resolved when crowded together in cytoplasm, networks or bundles of actin filaments can be imaged by differential interference
586
SECTION IX n Cytoskeleton and Cellular Motility
BOX 33.1 Tools to Study the Actin System
Actin Filament Destabilizers Sponges synthesize toxins that destabilize actin filaments in cells by sequestering actin monomers (latrunculin A and B) or severing actin filaments (swinholide A). Latrunculins bind in the nucleotide binding cleft of monomers and prevent their polymerization, making them very useful experimentally. Cytochalasins (meaning “cell relaxing”) were so named, because they cause regression of the cleavage furrow during cytokinesis and disrupt actin filaments in cells. These complex organic compounds synthesized by fungi bind in the barbed end groove of actin and inhibit subunit association and dissociation at the barbed ends of filaments. In addition low-affinity binding to actin monomers promotes their dimerization and the hydrolysis of ATP bound to one subunit, converting ATP-actin to ADP-actin. Given the complicated mechanism of action, their effects on cells must be interpreted cautiously. C2 toxin produced by Clostridium botulinum is an enzyme that catalyzes the ADP-ribosylation of cytoplasmic actins on arginine-177. Clostridium perfringens iota toxin does the same to muscle actin. ADP-ribosylated actin polymerizes poorly and caps the barbed end of actin filaments. The ability of these protein toxins to penetrate live cells, cap actin filaments, and alter actin polymerization accounts for
A
B
Control
60 s
their disruption of the actin cytoskeleton in cells and may contribute to their toxicity.
Actin Filament Stabilizers Phallotoxins (such as phalloidin), cyclic peptides synthesized by poisonous mushrooms, and the sponge toxin jasplakinolide both stabilize actin filaments. They bind to filaments between three subunits and reduce the rate of subunit dissociation to near zero at both ends of the polymer. Jasplakinolide can enter cells but phalloidin must be microinjected. They inhibit processes that depend on actin filament turnover, including amoeboid movement. Phallotoxins are toxic to humans, because they interfere with bile secretion. Fluorescent derivatives of phallotoxins are widely used to localize actin filaments in permeabilized cells and tissues (see Fig. 33.1), as well as to quantify polymerized actin in cells and cell extracts.
Inhibitors of Actin-Binding Proteins Small, drug-like molecules are available to inhibit Arp2/3 complex and formins. CK666 blocks the conformational change that activates Arp2/3 complex and inhibits actin filament branch formation in cells. A molecule called SMIFH2 inhibits nucleation and elongation by many different formins.
F-actin
C
D
E
FIGURE 33.18 ACTIN FILAMENT DYNAMICS AT THE LEADING EDGE OF A GIANT GROWTH CONE OF A NEURON ISOLATED FROM THE MOLLUSK APLYSIA. A network of actin filaments forms continuously at the leading edge of the growth cone, moves inward by retrograde flow, and disassembles near the central organelle-rich zone. A–C, Effect of the drug cytochalasin D on the growth cone. A, (left and middle) Differential interference contrast (DIC) micrographs before and 60 s after applying the drug, which disrupts the network of actin filaments at the leading edge. The double-headed arrow marks the zone cleared of actin filaments, stained with rhodamine phalloidin in the right panel. A narrow rim of filaments survives at the leading edge. B, Time series of DIC micrographs at 6-s intervals, showing that retrograde flow of existing filaments (and small beads on the surface) continued toward the cell body after cytochalasin blocked the formation of new filaments at the leading edge. C, Detail of this growth cone after fixing after 60 s and staining with rhodamine-phalloidin. If cytochalasin is removed from a live cell, the actin filament network recovers, beginning near the leading edge. D–E, Fluorescence speckle microscopy of a growth cone injected with a low concentration of Alexa 488 actin monomers. D, Distribution of actin in lamellar regions between radial bundles. E, Vectors showing the velocities of actin speckles. False color indicates velocities ranging from (dark blue) zero to (red) 7 µm per second. (Courtesy Paul Forscher, Yale University, New Haven, CT, and modified from Forscher P, Smith SJ. Actions of cytochalasins on the organization of actin filaments and microtubules in a neuronal growth cone. J Cell Biol. 1988;107:1505–1516, and Yang Q, Zhang XF, Pollard TD, Forscher P. Arp2/3 complex-dependent actin networks constrain myosin II function in driving retrograde actin flow. J Cell Biol. 2012;197:939–956.)
contrast or phase contrast microscopy. For example, networks of actin filaments in nerve growth cones constantly assemble and move away from the leading edge (Fig. 33.18). Fluorescence microscopy is even more informative if actin is labeled with a fluorescent probe. Purified actin can be labeled with a fluorescent dye and microinjected into live cells, where it is incorporated into the actincontaining structures (Fig. 33.18D). Expression of actin tagged with fluorescent protein is more convenient, although the fluorescent protein tag interferes with polymerization by formins. Alternatively, the fluorescent protein can be attached to an actin binding domain to label cellular actin filaments indirectly. Fluorescent actin incorporates quickly into most of the filaments in nonmuscle cells. If low levels of fluorescent actin are used, random incorporation into filaments can result in fluorescent “speckles” that can be used to follow the movements and turnover of these subunits (Fig. 33.18D). Fig. 38.9 illustrates how bleaching or activating fluorescent actin in a live cell can reveal where filaments assemble and disassemble at the leading edge of motile cells.
Pool of Unpolymerized Actin Cells can respond rapidly to stimuli such as chemoattractants by assembling actin filaments where needed, because they have a large pool of unpolymerized actin to grow the new filaments. Roughly half of the actin in the cytoplasm of cells other than muscle is unpolymerized, corresponding to 50 to 100 µM monomers, 500 to 1000 times higher than the critical concentration. The combination of monomer binding to profilin and capping barbed ends allows cells to maintain a large pool of actin subunits ready to elongate any barbed ends created by uncapping, severing, or nucleation. In vertebrate cells, thymosin-β4 augments the effects of profilin by sequestering a fraction of the actin monomers. The concentrations of profilin and thymosin-β4 exceed the concentration of unpolymerized actin, and these proteins bind tightly enough to reduce the free monomer concentration to the micromolar level. Actin monomers bound to profilin or thymosin-β4 do not nucleate new filaments. However, for profilin and thymosin-β4 to maintain a monomer pool, most actin filament barbed ends must be capped as rapid addition of actin-profilin complexes to free barbed ends would quickly deplete the pool of unpolymerized actin. Cells contain enough heterodimeric capping protein (augmented by gelsolin in some cells) to cap the barbed ends of most filaments. Initiation and Termination of Actin Filaments A variety of external agonists and internal signals stimulate the assembly of actin filaments. Examples include the ability of chemoattractants to direct pseudopod formation in amoebas (see Fig. 38.10) and white blood cells (see Fig. 30.13), and the influence of the mitotic spindle
CHAPTER 33 n Actin and Actin-Binding Proteins
587
on the assembly of the cytokinetic contractile ring (see Fig. 44.24). Polymerization depends on creation of barbed ends, which grow rapidly at rates estimated to be 50 to 500 subunits per second, depending on the concentration of actin-profilin. Three mechanisms are thought to create free barbed ends: uncapping, severing, and de novo formation of new barbed ends. In many cases, new barbed ends appear to form de novo. At the leading edge of motile cells, Rho-family GTPases associated with the plasma membrane and polyphosphoinositides activate nucleation promoting factors, which stimulate Arp2/3 complex to form branches with free barbed ends (Figs. 33.2D and 33.12). Formins nucleate filaments for the cleavage furrow and initiate or sustain the growth of actin filaments in filopodia (see Fig. 38.3A). When thrombin activates platelets (see Fig. 30.14), plasma membrane polyphosphoinositides uncap barbed ends by dissociating gelsolin. Dephosphorylation activates ADF/cofilin proteins, which can sever and nucleate filaments, creating free barbed ends. The duration of the growth of a cellular actin filament depends on the nucleation mechanism and the local environment. At the leading edge, new branches nucleated by Arp2/3 complex grow rapidly but transiently, as the concentration of free capping protein is high enough to terminate growth by capping barbed ends in a few seconds. On the other hand, barbed ends growing in association with a formin are protected from capping and grow persistently, as is observed at the tips of filopodia and the barbed ends of actin cables located in the buds of yeast cells.
Actin Filament Turnover and Subunit Recycling Actin filaments are long lived if protected by tropomyosin and capping, as in muscle and stress fibers, but many actin filaments, such as those at the leading edge of motile cells, turn over quickly (see Fig. 38.7). Turnover starts with hydrolysis of ATP bound to the actin subunits and dissociation of the γ-phosphate, reactions that provide a timer to mark older filaments for depolymerization. ADF/cofilin proteins bind ADP-actin subunits in filaments and sever these older filaments (Fig. 33.16). This creates more ends available for dissociation of ADP-actin. Profilin stimulates the exchange of ADP for ATP on actin monomers and restores the pool of ATP-actin monomers bound to profilin available for polymerization. In cells with a high concentration of thymosin-β4, much of the ATP-actin is stored bound to thymosin. Rapid exchange allows actin monomers to move from thymosin-β4 to profilin, ready for polymerization. How Do Cells Organize Actin Assemblies? Cells organize actin filaments in a variety of structures, including cortical networks, microvilli or filopodia, and contractile bundles (Figs. 33.1 to 33.3). Each cell in a population is unique, but all cells of a particular type
588
SECTION IX n Cytoskeleton and Cellular Motility
achieve a similar pattern of organization. The mechanisms appear to depend on expression of an appropriate mixture of actin-binding proteins to assemble particular structures. For example, actin forms bundles similar to microvilli and filopodia when polymerized in the presence of fimbrin and villin, the major crosslinking proteins in microvilli. Overexpression of villin induces cells to extend existing filopodia and form new ones. Thus, the pool of villin and fimbrin and other components may set the number of microvilli. Many actin filament barbed ends are oriented toward the plasma membrane allowing their growth to push on the inside of the membrane (Fig. 33.2). Other barbed ends are anchored in structures including adhesion sites (see Fig. 30.11), the cleavage furrow (Fig. 33.3), and Z-disks of muscles (see Fig. 39.5). These anchors transmit forces on these filaments produced by myosin to the plasma membrane or the rest of the contractile apparatus in muscles. This makes mechanical sense, as actin filaments sustain tension better than compression, and because (with one interesting exception) all known myosins pull filaments in a direction away from the barbed end. The Rho-family GTPases Cdc42, Rac, and Rho regulate the assembly of many actin filaments (Fig. 33.19). A complex of proteins, including Cdc42, in the bud of yeast cells anchors formins, which mediate the continuous assembly of a cable of actin filaments (see Fig. 37.11). The formins anchor the growing barbed ends as the pointed ends extend into the mother cell. These cables are tracks for myosin-V to move cargo into the bud. In motile cells, signals downstream of chemotactic receptors activate Cdc42 and Rac (see Fig. 38.7), which activate nucleation-promoting factors. They, in turn, stimulate Arp2/3 complex to generate the branched filament network that pushes the membrane forward. Listeria use a surface protein, ActA, to activate Arp2/3
A. Control
B. Cdc 42
complex, which generates a “comet tail” of actin filaments to push the bacterium through the cytoplasm (see Fig. 37.12). Proteins that mediate endocytosis activate homologs of WASp and other proteins to assemble actin patches in budding yeast (see Fig. 37.11) and fission yeast (Fig. 33.1D). Physical forces also help organize actin filaments. Tension generated by myosin II contributes to the alignment of actin filaments in stress fibers (Fig. 33.1) and the contractile ring during cytokinesis (Fig. 33.3; also see Fig. 44.24). Rho activates myosin II by stimulating two kinases that phosphorylate its regulatory light chain and inhibit a phosphatase that reverses the light chain phosphorylation (see Fig. 39.24). Crosslinking proteins, such as α-actinin, help maintain the integrity of these bundles under mechanical stress.
Mechanical Properties of Cytoplasm Actin filaments account for many of the mechanical properties of cytoplasm, a complicated, viscoelastic material. Viscoelastic means that cytoplasm can both resist flow, like a viscous liquid (eg, molasses), and store mechanical energy when stretched or compressed, like a spring. The physical properties of actin filaments depend on their lengths and their interactions. At physiological concentrations, purified actin filaments are viscoelastic. At high concentrations, actin filaments also align spontaneously into large parallel arrays called liquid crystals. Crosslinking actin filaments increases both their viscosity and stiffness. Severing actin filaments decreases their viscoelasticity. On the other hand, shorter filaments have a higher tendency to form bundles in the presence of crosslinking proteins, so severing can actually promote the formation of rigid actin filament bundles.
C. Rac
D. Rho
FIGURE 33.19 RHO-FAMILY GTPASES PROMOTE THE ASSEMBLY OF ACTIN-BASED STRUCTURES. Fluorescence micrographs of Swiss 3T3 fibroblasts stained with rhodamine-phalloidin to reveal actin filaments. A, Resting cells. B, Cells microinjected with activated Cdc42 form many filopodia. C, Cells microinjected with activated Rac have a thick cortical network of actin filaments around the periphery. D, Stress fibers anchored at their ends by focal contacts are abundant in cells microinjected with an activated form of Rho. (Courtesy Alan Hall, University of London, United Kingdom.)
CHAPTER 33 n Actin and Actin-Binding Proteins
A. Static
B. Rapid
589
C. Slow
FIGURE 33.20 DYNAMIC CROSSLINKING OF ACTIN FILAMENTS. Rapid binding and dissociation of crosslinking proteins allow networks of actin filaments to resist rapid deformations but to change shape passively when force is applied for a prolonged time. A, Crosslinked network in a static region. B, Crosslinking proteins resist deformation if force is applied rapidly. C, Crosslinking proteins provide little resistance to deformation if force is applied slowly because the crosslinks rearrange faster than the filaments are displaced. (Modified from Pollard TD, Satterwhite L, Cisek L, et al. Actin and myosin biochemistry in relation to cytokinesis. Ann N Y Acad Sci. 1990;582:120–130.)
Many crosslinking proteins, including α-actinin, have low affinities for actin filaments with Kds in the micromolar range. At steady state in vitro, these crosslinking proteins bind to and dissociate from actin filaments on a second or subsecond time scale. Consequently, gels of actin filaments and α-actinin are much more rigid when deformed rapidly than when the deformations are slower (Fig. 33.20), because crosslinks resisting the displacement of the filaments can rearrange if given sufficient time. Dynamic crosslinks between filaments allow actin networks to remodel passively as cells move. Cells also remodel the actin cytoskeleton actively by nucleating, severing or depolymerizing filaments. SELECTED READINGS Allingham JS, Klenchin VA, Rayment I. Actin-targeting natural products: structure, properties and mechanisms of action. Cell Mol Life Sci. 2006;63:2119-2134. Bravo-Cordero JJ, Magalhaes MA, Eddy RJ, Hodgson L, Condeelis J. Functions of cofilin in cell locomotion and invasion. Nat Rev Mol Cell Biol. 2013;14:405-415. Campellone KG, Welch MD. A nucleator arms race: cellular control of actin assembly. Nat Rev Mol Cell Biol. 2010;11:237-251. Chen Z, Borek D, Padrick SB, et al. Structure and control of the actin regulatory WAVE complex. Nature. 2010;468:533-538. Derman AI, Becker EC, Truong BD, et al. Phylogenetic analysis identifies many uncharacterized actin-like proteins (Alps) in bacteria: regulated polymerization, dynamic instability and treadmilling in Alp7A. Mol Microbiol. 2009;73:534-552. Dominguez R, Holmes KC. Actin structure and function. Annu Rev Biophys. 2011;40:169-186.
Edwards M, Zwolak A, Schafer DA, et al. Capping protein regulators fine-tune actin assembly dynamics. Nat Rev Mol Cell Biol. 2014; 15:677-689. Elam WA, Kang H, De la Cruz EM. Biophysics of actin filament severing by cofilin. FEBS Lett. 2013;587:1215-1219. Gunning PW, Ghoshdastider U, Whitaker S, Popp D, Robinson RC. The evolution of compositionally and functionally distinct actin filaments. J Cell Sci. 2015;128:2009-2019. Higgs HN, Peterson KJ. Phylogenetic analysis of the formin homology 2 domain. Mol Biol Cell. 2005;16:1-13. Janmey PA, Weitz DA. Dealing with mechanics: mechanisms of force transduction in cells. Trends Biochem Sci. 2004;29:364-370. Löwe J, van den Ent F, Amos LA. Molecules of the bacterial cytoskeleton. Annu Rev Biophys Biomol Struct. 2004;33:177-198. McCullagh M, Saunders MG, Voth GA. Unraveling the mystery of ATP hydrolysis in actin filaments. J Am Chem Soc. 2014;136: 13053-13058. Nag S, Larsson M, Robinson RC, Burtnick LD. Gelsolin: the tail of a molecular gymnast. Cytoskeleton (Hoboken). 2013;70:360-384. Paul A, Pollard TD. Review of the mechanism of processive actin filament elongation by formins. Cell Motil Cytoskeleton. 2009;66: 606-617. Pollard TD. Actin and actin binding proteins. Cold Spring Harb Perspect Biol. 2016;8:8. Quinlan M, Heuser JE, Kerkhoff E, Mullins RD. Drosophila Spire is an actin filament nucleation factor. Nature. 2005;433:382-388. Rotty JD, Wu C, Bear JE. New insights into the regulation and cellular functions of the ARP2/3 complex. Nat Rev Mol Cell Biol. 2013; 14:7-12. Stossel TP, Condeelis J, Cooley L, et al. Filamins as integrators of cell mechanics and signalling. Nat Rev Mol Cell Biol. 2001;2: 138-145. Svitkina T. Actin cytoskeleton and actin-based motility. In: The Cytoskeleton. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 2016;281-302.
590
SECTION IX n Cytoskeleton and Cellular Motility
APPENDIX 33.1
Classification of Actin-Binding Proteins Distribution
Subunits (N × kD)
Kd Actin Binding
β Thymosins
An
1×5
DNase I
An
1 × 29
Profilin
Eu
Vitamin D–binding protein (Gc globulin)
Protein
Other Ligands
Diseases
0.7 µM monomer
—
—
0.1 nM monomer and pointed end
Ca2+, DNA
—
1 × 13–15
0.1 µM monomer
PIP2, VASP, formins
Amyotrophic lateral sclerosis
An
1 × 58
1 nM monomer
Vitamin D, C5A complement
Obstructive lung diseases
Eu
1 × 15–19
0.1 µM ADP monomer, 0.5 µM ADP filament
PIP2
Cancer, neurodegenerative diseases
Monomer Binding
Small Severing ADF/cofilin (actophorin, depactin, destrin)
Nucleation and Elongation Arp2/3 complex
Eu
1 each of 49, 44, 40, 35, 21, 20, 16
10 nM pointed end 0.5 µM filament side
Profilin, Scar, WASp, cortactin
Psychiatric diseases (?)
Formins
Eu
Dimers range of sizes
Barbed end
Profilin, spire
Glomerular sclerosis
Spire
An
1 × 84
Side
Formins
—
Capping protein (CapZ)
Eu
1 × 32–36(α) + 1 × 28–32(β)
500 kD (multiple genes & splice isoforms)
Calponin homology domain, various other domains
Crosslinks MT to actin and intermediate filaments
Destabilizers
Severing
Stabilizers
Linkers Anchors glycine receptors to MTs
+Tips (Plus End Binding Proteins) APC
Vertebrates, insects
1 × 300 kD
Binds EB1 and β-catenin
Regulates β-catenin; tumor suppressor (colon cancer)
CLASP (Mast/Orbit)
Eukaryotes
? × 165 kD
Binds CLIP-170 and EB1
Regulates MT dynamics in the cell cortex and at kinetochores
CLIP-170
Eukaryotes
170 kD
Binds EB; phosphorylation inhibits MT binding
Binds endosomes to plus ends of MT
Dis1/TOG family (XMAP215, others)
Eukaryotes
215 kD, other variants
EB-1, -2, -3 (Bim1p, Mal3p)
Eukaryotes
30 kD
KMN network
Eukaryotes
Many subunits including KNL-1, Mis12 complex, NCD80 complex
Regulates MT dynamics and spindle pole Binds MT plus ends and APC
Promotes MT assembly Kinetochore MT binding complex
−Tips (Minus End Binding Proteins) 435,000
Eight different subunits
Binds γ-tubulin ring complex
CAMSAP-1, -2, -3 (Patronin)
Metazoa
>200 kD
Calponin homology domain, coiled-coils
Binds near plus ends; stabilizes end
γ-Tubulin ring complex
Eukaryotes
14 × 50 kD γ-tubulin + 8 other proteins
Polymeric lockwasher
Nucleates MT assembly from minus end
Augmin
ATP, adenosine triphosphate; ATPase, adenosine triphosphatase; CAMSAP, calmodulin-regulated spectrin-associated protein; CLIP, cytoplasmic linker protein; EB, end-binding protein; MAP, microtubule-associated protein; MT, microtubule; PKA, protein kinase A.
612
SECTION IX n Cytoskeleton and Cellular Motility
APPENDIX 34.2
Centrosomal Structural Proteins Name
Distribution
Composition (Subunit Size)
Properties
Functions (Diseases)
C-NAP1
Metazoa
2 × 250 kD
Coiled-coil protein
Fibers connecting PCM of two centrioles, regulated by the protein kinase Nek2
Centriolin
Metazoa
2 × 240 kD
Coiled-coil protein
Located in in subdistal appendages (stem cell myeloproliferative disorder)
γ-Tubulin ring complex
Eukaryotes
14 × 50 kD γ-tubulins + eight other proteins
Polymeric lockwasher
Nucleates MT assembly from minus end
Ninein
Metazoa
2 × 236 kD
Coiled-coil protein
Located in in subdistal appendages; MT anchoring (Seckel syndrome; prenatal dwarfism)
Pericentrin (Kendrin)
Animals Plants
2 × 380 kD
Coiled-coil protein
PCM scaffold; binds calmodulin, dynein, γ-tubulin ring complex, kinases, and phosphatases (human autoantigen; microcephalic osteodysplastic primordial dwarfism)
Rootletin
Many eukaryotes
2 × 230 kD
Coiled-coil protein
Fibers connecting PCM of two centrioles and anchoring basal bodies
SAS-4
Many eukaryotes
2 × 92 kD
Coiled-coil protein
Part of centriole scaffold with SAS-6
SAS-6
Many eukaryotes
2 × 74 kD
Coiled-coil protein with globular domains
Forms 9-fold scaffold for centrioles (microcephaly 14)
MT, microtubule; PCM, pericentriolar material.
CHAPTER
35
Intermediate Filaments
I
ntermediate filaments (Fig. 35.1) are strong but flexible polymers that provide mechanical support for metazoan cells. These filaments are composed of many different but homologous proteins. The filaments were named intermediate, because their 10-nm diameters are intermediate between those of the thick and thin filaments in striated muscles, where they were first recognized (see Figs. 39.3 and 39.8). They are not found in plants, fungi, or prokaryotes, although one bacterial species has a coiled-coil protein with some properties of intermediate filaments. Cytoplasmic intermediate filaments, in particular keratin filaments, tend to cluster into wavy A
B
bundles that vary in compactness, forming a branching network between the plasma membrane and the nucleus. Intercellular junctions called desmosomes anchor intermediate filaments to the plasma membrane (see Fig. 31.8B) and thereby transmit mechanical forces between adjacent cells. Hemidesmosomes connect intermediate filaments across the plasma membrane to the extracellular matrix (see Fig. 31.8C). The continuum of intermediate filaments and junctions prevents excessive stretching of cells and gives tissues such as epithelia and heart muscle their mechanical integrity. Skin appendages built from crosslinked D
E
C
FIGURE 35.1 LIGHT AND ELECTRON MICROGRAPHS OF INTERMEDIATE FILAMENTS. A, Fluorescence light micrograph of a cultured fibroblast stained with antibodies to vimentin (red) and nuclear lamins (green) and with DAPI (4,6-diamidino-2-phenylindole) for DNA (blue). B, Fluorescence micrograph of cultured epithelial cells stained with antibodies to keratin intermediate filaments (orange). Desmosomes are stained green. Scale bar is ~10 µm. C, Fluorescence micrograph of crescentin labeled with a red fluorescent dye in the bacterium Caulobacter crescentus. Scale bar is 2 µm. D–E, Electron micrographs of thin sections of a cultured baby hamster kidney cell showing longitudinal (arrows) and cross sections (arrowheads) of vimentin intermediate filaments. (A, Courtesy U. Aebi, Biozentrum, University of Basel, Switzerland, and H. Herrmann, German Cancer Research Center, Heidelberg, Germany. B, Courtesy E. Smith and E. Fuchs, University of Chicago, IL. C, Courtesy M. Cabeen and C. Jacobs-Wagner, Yale University, New Haven, CT. D–E, Courtesy R.D. Goldman, Northwestern University, Chicago, IL.)
613
614
SECTION IX n Cytoskeleton and Cellular Motility
keratin intermediate filaments such as hair and whale baleen, illustrate their flexibility and high tensile strength. Molecular defects in cytoplasmic intermediate filaments or junctions associated with them result in rupture of skin cells and blistering diseases. Defects in lamins associated with the nuclear envelope cause a bewildering array of diseases (see Fig. 9.10).
Structure of Intermediate Filament Subunits Spectroscopic data and X-ray fiber diffraction of materials composed of intermediate filaments, like wool, established the α-helical coiled-coil as their basic structure. However, the proteins forming intermediate filaments can vary greatly in size, in contrast to the uniform sizes of actins and tubulins. Eventually amino acid sequences of several intermediate filament proteins in the 1980s established that they all have a central α-helical segment that forms a rod domain flanked by variable N- and C-terminal end domains (Fig. 35.2). Gene sequences allow grouping of the proteins into five amino acid sequence homology classes (Table 35.1).
Rod 309–355 AA
Head 6–180 AA N
Coil 1A
Coil 1B
L1
Tail 6–1200 AA
Coil 2A
C
Coil 2B
L12
Rod domain of vimentin
IF protein domains Keratin CK 19 Vimentin NFL NFM NFH Lamin A Lamin B
180
312 71
319
83
326
70
331
100
319
100
313 33 35
151 9 55 142 497 607
λ 355 λ 355
276 197
42
FIGURE 35.2 INTERMEDIATE FILAMENT (IF) PROTEINS HAVE A CENTRAL ROD DOMAIN FLANKED BY HEAD AND TAIL DOMAINS OF VARIABLE LENGTHS. The ribbon diagram shows a crystal structure of the vimentin rod domain (see Protein Data Bank [www.rcsb.org] file 1GK7). Rod domains consist of an α-helical coiledcoil of 310 residues and are 46 nm long. Lamins have an additional 42 residues in the rod domain (λ). The residues that are most important for assembly are at the beginning and end of the rod. End domains differ in sequence and size from 6 to 1200 residues. (Model of vimentin courtesy H. Herrmann, German Cancer Research Center, Heidelberg, Germany.)
The characteristic feature of all intermediate filament proteins is a dimeric, parallel, α-helical coiled-coil rod domain that forms the backbone of the filaments. The rod domains of cytoplasmic intermediate filament proteins are approximately 46 nm long. Those of nuclear lamins are 6 nm longer (see Fig. 35.2 and Fig. 35.3A). Like other coiled-coils (see Fig. 3.10), intermediate filament rod domains have a heptad repeat pattern of amino acids with the first and fourth residues providing a continuous row of hydrophobic interactions along the interface of the two α-helices (see Fig. 3.10). The rod domains have two highly conserved sites with interruptions in the coiled-coil termed L1 and L12. Zones of positive and negative charge alternate along the rod. When staggered appropriately, these zones provide complementary electrostatic bonds for assembly into filaments. Approximately 20 highly conserved residues at each end of the rod are essential for filament elongation through headto-tail interactions of dimeric molecules. Studies with mutant proteins show that assembly of filaments depends on both head-to-tail overlaps and lateral associations between rod domains. The N- and C-terminal end domains flanking the rod are largely unstructured and vary considerably in size (see Fig. 35.2). The N-terminal end (“head”) domains are essential for assembly whereas the C-terminal end (“tail”) domains protrude from the filament surface (see Fig. 35.3C) to control filament diameter and/or interact with other cellular components. Each class of intermediate filament molecule forms in a characteristic manner. Keratins are obligate heterodimers of one acidic (class I) and one basic (class II) keratin polypeptide. Vimentin and desmin (class III) and nuclear lamins (class V) form parallel dimers of identical polypeptides (see Fig. 35.3A). The intermediate filament proteins in the nervous system (class IV) form complex mixtures of filaments, and it is not yet clear if they are mainly homodimers or if some are heterodimers. Two molecules of cytoplasmic intermediate filament proteins associate in an antiparallel, half-staggered manner to form stable apolar dimers, sometimes called “tetramers” because they consist of four polypeptides (see Fig. 35.3D). These tetramers are, at least in vitro, the principal intermediates in filament assembly as they further associate laterally and longitudinally.
Evolution of Genes for Intermediate Filament Proteins Well after the genes for intermediate filament proteins were discovered in the higher branches of the animal lineage, whole-genome sequencing established their presence in a wide range of other organisms that descended from the last common eukaryotic ancestor (see Fig. 2.4B). Genes for animal intermediate filament proteins arose in early metazoan cells from genes
CHAPTER 35 n Intermediate Filaments
615
TABLE 35.1 Classification of Intermediate Filament Proteins Based on Rod Domain Sequences Number of Human Genes
Class
Type
Molecule
Distribution
Diseases
I
Acidic keratin
28
40–65 kD, obligate heterodimer with class II
Epithelial cells and their appendages
Blistering skin, corneal dystrophy, brittle hair and nails
II
Basic keratin
26
Desmin GFAP Peripherin
1 1 1
Synemin
1
Epithelial cells and their appendages Muscle cells Glial cells Peripheral > CNS neurons Muscle cells
Similar to class I
III
51–68 kD, obligate heterodimer with class I 53 kD, homopolymers 50 kD, homopolymers 57 kD
Vimentin
1
Mesenchymal cells
Mouse null viable
Neurofilament NFL
1
Neurons
Mouse null viable; human neuropathies
NFM
1
NFH
1
Nestin
1
Obligate heteropolymers with NFM, NFH Obligate heteropolymers with NFL, NFH Obligate heteropolymers with NFL, NFM 230 kD, homopolymers
α-Internexin
1
55 kD, homopolymers
Embryonic neurons, muscle, other cells Embryonic neurons
Lamins
3
7 Isoforms, 62–72 kD, homodimers
Metazoan nuclei, some protozoa
IV
V
190 kD, interacts with other class III IFs 54 kD, homopolymers and heteropolymers
Cardiac and skeletal myopathies Alexander disease; mouse null viable
Neurons Neurons
Mutations a risk factor in amyotrophic lateral sclerosis
Cardiomyopathy, lipodystrophy, one form of Emery-Dreifuss muscular dystrophy, two forms of progeria, plus many others
CNS, central nervous system; GFAP; glial fibrillary acidic protein; IF, intermediate filament; NFH, neurofilament heavy; NFL, neurofilament light; NFM, neurofilament medium. For reference, see Omary MB, Coulombe PA, McLean WH: Intermediate filament proteins and their associated diseases. N Engl J Med. 2004;351:2087– 2100. For current information see: http://www.interfil.org/.
encoding nuclear lamins (see Fig. 14.7). Most metazoans, including chordates, mollusks, insects, and nematodes (see Fig. 2.8), retain genes for lamins. The gene for cytoplasmic intermediate filaments arose from a duplicated lamin gene in an invertebrate organism in the lineage leading to chordates. One copy of the duplicated gene was modified by deletion of the nuclear localization sequence and the CAAX box (a C-terminal prenylation site; see Fig. 13.10). After deletion of the codons for 42 residues (6 heptads) in coil 1B and the immunoglobulin domain in the “tail” in early chordates, further gene duplications and divergence produced the four families of genes for cytoplasmic intermediate filaments of vertebrates (see Table 35.1). The unique functional requirements for each class of intermediate filament protein have conferred strong selective pressure on their genes, so that orthologs are much more similar than the paralogs. For example, human desmin is much more similar to frog desmin than it is to human keratin. The bacterium Caulobacter crescentus has a gene for a coiled-coil protein, crescentin, with some features of intermediate filament proteins (see Fig. 35.1C). However, it lacks some of the highly conserved residues in the rod domain of animal intermediate filament proteins, including those vital for filament elongation. Crescentin is
required for the asymmetrical shape of Caulobacter cells and when expressed in Escherichia coli makes the cells spiral shaped. The origin of this gene is unknown, but lateral transfer from a eukaryote followed by divergence is possible.
Filament Structure and Assembly Intermediate filaments are approximately 10 nm in diameter with wavy profiles in electron micrographs of thin sections of cells (see Fig. 35.1B) or after negative staining of isolated filaments (see Fig. 35.3B). In some cases, such as neurofilaments, parts of the head domains and most of the tail domains project radially from the filament core, forming a type of bottlebrush (see Fig. 35.3C). The most carefully studied intermediate filaments are built from octameric complexes (ie, two laterally associated molecular dimers) that associate end to end to form protofibrils like the strands of a rope (see Fig. 35.3D). In cross section, a standard intermediate filament has up to 16 coiled-coils, but their exact internal arrangement is not known. Because the molecular dimers lack polarity, intermediate filaments are considered to be apolar (ie, both ends of the filament are equivalent; see Fig. 35.3D). This is a striking difference from actin filaments (see Fig. 33.8) and microtubules (see Fig. 34.4), which depend
616
SECTION IX n Cytoskeleton and Cellular Motility
A
B
C
D
~32 coiled-coils in cross section
Apolar assembly unit
Head
Tail Polar coiled-coil dimer 46 nm Dimer of coiled-coil rod domains
Intermediate filament 10-nm diameter
FIGURE 35.3 INTERMEDIATE FILAMENTS ARE CONSTRUCTED LIKE A MULTISTRAND ROPE. A–C, Electron micrographs. A, Metalshadowed lamin molecules consisting of two polypeptides joined by a long α-helical coiled-coil with globular tail domains at the C-terminus. B, Negatively stained vimentin intermediate filaments. C, Rotary shadowed intermediate filament showing radial projections. D, A model for intermediate filament structure. The building blocks are antiparallel complexes of two coiled-coil molecular dimers. The ribbon diagram is a model of the dimer of vimentin rod domains. Assembly occurs via the formation of unit-length filaments, the products of the lateral association of eight antiparallel dimers, which then longitudinally anneal into intermediate filaments. This model is consistent with x-ray fiber diffraction patterns, chemical crosslinking, and other data, but details of the subunit packing remain to be determined. (A and C, Courtesy U. Aebi, University of Basel, Switzerland. B, Courtesy H. Herrmann, German Cancer Research Center, Heidelberg, Germany. D, Top, Data from Steinert P, Marekov LN, Parry DA. Conservation of the structure of keratin intermediate filaments. Biochemistry. 1993;32:10046–10056. D, Bottom, Model courtesy H. Herrmann, German Cancer Research center, Heidelberg, Germany. For reference, see Chernyatina AA, Nicolet S, Aebi U, Herrmann H, Strelkov SV. Atomic structure of the vimentin central α-helical domain and its implications for intermediate filament assembly. Proc Natl Acad Sci U S A. 2012;109:13620–13625.)
on their polarity for many functions, including the unidirectional motion of motor proteins. Furthermore, the number of protofilaments can vary along a single filament, making them much more heterogeneous than actin filaments or microtubules. Intermediate filaments are insoluble under physiological conditions, but can be dissociated in buffers of low ionic strength and high pH. Under physiological conditions isolated subunits spontaneously repolymerize in a few minutes. The first assembly product observed is a “unit-length filament” consisting of eight laterally associated molecular dimers with the length (60 nm) of a molecular dimer. Intermediate filaments grow by longitudinal annealing of unit-length filaments at both ends, in contrast to growth of actin filaments and microtubules by addition of single subunits at their ends. The nucleation mechanism that initiates polymerization and the elongation reactions are still being investigated, but it is clear that no nucleotides or other cofactors are needed for assembly. Most of the head domain is required to assemble intermediate filaments in vitro and in vivo. The tail domain is dispensable for assembly, although more molecular dimers can pack laterally into a filament in its absence.
Intermediate filaments are among the most chemically stable cellular structures, resisting solubilization by extremes of temperature as well as high concentrations of salt and detergents (Fig. 35.4). Nevertheless, intermediate filaments in some cells exchange their subunits within minutes to hours during interphase. For example, if vimentin is labeled with a fluorescent dye and injected into live cells or GFP-vimentin is expressed, fluorescent vimentin incorporates into cytoplasmic filaments. After a spot of fluorescent filaments is photobleached with a laser, the fluorescence recovers over a period of several minutes, indicating that subunits along the length of the filaments exchange with a pool of unpolymerized molecules. (Fig. 38.8 shows a similar experiment with actin.) Although no known motors move on the apolar intermediate filaments, motor proteins move intermediate filaments along microtubules. A spectacular example is found in nerve cells (see Fig. 37.5C).
Posttranslational Modifications Phosphorylation dramatically affects polymer assembly and dynamics of many types of intermediate filaments.
CHAPTER 35 n Intermediate Filaments
The process is complex and incompletely understood, as several different kinases phosphorylate many different sites and these phosphates tend to turn over rapidly. The impact of phosphorylation depends critically on the particular residue modified. The best example of phosphorylation destabilizing an intermediate filament is the breakdown of the nuclear lamina during mitosis (see Fig. 44.6). The mitotic kinase Cdkl-cyclin B phosphorylates two sites immediately
L V
B
A
FIGURE 35.4 INTERMEDIATE FILAMENTS RESIST SOLUBILIZATION WHEN CELLS ARE EXTRACTED. A, A fluorescence micrograph shows the network of vimentin intermediate filaments remaining after extraction of a Chinese hamster ovary (CHO) cell with the detergent Triton X-100, DNase, and a high concentration of salt to remove lipids, DNA, and soluble proteins. B, Gel electrophoresis of these extracted cells reveals that lamins (L) and vimentin (V) are among the few proteins remaining in the detergent-resistant cytoskeletal fraction. (Courtesy R. Goldman, Northwestern University, Chicago, IL.)
A
5 µm
B
617
flanking the rod domains of lamins, disrupting the headto-tail overlap required for the interactions of molecules that mediate filament elongation and lateral association of subunits. Cytoplasmic vimentin filaments also disassemble in some cell types during mitosis (Fig. 35.5B), but the process is more complex. Vimentin lacks the Cdk1-recognition sites immediately flanking the α-helical rod domain and coassembly with another intermediate filament protein, nestin, appears to be a prerequisite for phosphorylation to mediate disassembly. In contrast the organization of keratins changes only subtly during mitosis (Fig. 35.5C). The role of phosphorylation of intermediate filaments during interphase is less clear, but it might influence the structure of the cytoskeleton in response to various signals. Neurofilaments, abundant intermediate filaments in nerve axons and dendrites (see Fig. 35.9), are an exception to the rule that phosphorylation destabilizes intermediate filaments. The most stable neurofilaments are heavily phosphorylated in their large C-terminal tail end domain (see Fig. 35.2), whereas the pool of unpolymerized of NFM (neurofilament medium) and NFH (neurofilament heavy) molecules is not phosphorylated. The NFM and NFH end domains are not essential for assembly, so phosphorylation might influence other functions of neurofilaments. Keratin intermediate filaments in hair are chemically crosslinked to each other and associated with matrix proteins by disulfide bonds and amide bonds between lysines and acidic residues, creating a tough composite material built on the same principles as fiberglass. Beauticians take advantage of these crosslinks to modify the
C
FIGURE 35.5 FLUORESCENCE MICROGRAPHS OF INTERMEDIATE FILAMENTS. A, A cultured fibroblast stained with antibodies to vimentin intermediate filaments (red) and microtubules (green) and fluorescent phalloidin for actin filaments (blue). B, Vimentin intermediate filaments dispersed in mitosis. C, Dividing epithelial cells stained with an antibody to keratin, which remains polymerized during mitosis. (A, Courtesy U. Aebi, Biozentrum, University of Basel, Switzerland, and H. Herrmann, German Cancer Research Center, Heidelberg, Germany. B, Courtesy R.D. Goldman, Northwestern University, Chicago, IL. C, Courtesy H. Herrmann, German Cancer Research Center, Heidelberg, Germany.)
618
SECTION IX n Cytoskeleton and Cellular Motility
shape of hairs during “permanents.” They first reduce disulfide bonds and then reform them after molding the hair into a new shape.
Expression of Intermediate Filaments in Specialized Cells Animal cells express at least one of the three major nuclear lamins, whereas the repertoire of cytoplasmic intermediate filament proteins varies greatly in different cell types (see Table 35.1). Most cells express predominantly one class—or at the most two classes—of cytoplasmic intermediate filament proteins, presumably making use of their unique properties. For example, epithelial cells express class 1 and class 2 keratins, whereas muscle cells express desmin and mesenchymal cells express vimentin. A few cells, such as the basal myoepithelial cells of the mammary gland, express two types of intermediate filament proteins that sort into separate filaments with different distributions in the cytoplasm. Similarly, microinjection or expression of foreign intermediate filament subunits usually (but not invariably) results in correct sorting to the homologous class of filaments. In tissues such as skin and brain, cells express a succession of intermediate filament isoforms as they mature and differentiate. For example, dividing cells at the base of the epidermis of skin express mainly keratins 5 and 14, whereas terminally differentiating cells express keratins 1 and 10 (Fig. 35.6). The switch in keratin
Skin surface Stratum corneum
A. Histology
B. K14
C. K10
expression is associated with a marked increase in filament bundling, a feature that might contribute to the resistance of the surface layers of the skin to chemical dissociation and mechanical rupture. In the nervous system, supporting glial cells express a class III intermediate filament protein, whereas embryonic neurons first express the class IV α-internexin and later express the three other class IV neurofilament isoforms (see Table 35.1). Although the smallest neurofilament isoform (NFL [neurofilament light]) can assemble on its own in vitro, the formation of intermediate filaments in neurons requires NFL and one of the larger isoforms NFM of NFH, which are encoded by distinct genes. Tumors often express the intermediate filament protein that is characteristic of the differentiated cells from which they arose. This is helpful to pathologists in diagnosing poorly differentiated or metastatic cancers. For example, tumors of muscle cells express desmin rather than keratin (expressed in epithelial cells) or vimentin (expressed in mesenchymal cells). This rule is not absolute, as some tumors arising in epithelia turn down the expression of keratin and turn up the expression of vimentin before invading surrounding tissues.
Proteins Associated With Intermediate Filaments A number of proteins bind intermediate filaments and link them to membranes and other cytoskeletal polymers
D. Keratin mutations
E. Hyperkeratosis (K10 mutant)
Autosomal dominant mutation
Granular Spinous (K1 / K10)
Basal (K5 / K14) Dermis
Minor mechanical stress Great mechanical stress
Null mutation
Lysis
FIGURE 35.6 EXPRESSION OF KERATINS AND EFFECTS OF KERATIN MUTATIONS ON THE STRATIFIED SQUAMOUS EPITHELIUM OF SKIN. A, Light micrograph of a section of mouse skin stained with hematoxylin and eosin (H&E). B, Localization of keratin 14 in a section of skin using antibodies and a histochemical procedure that leaves a brown deposit. Proliferating cells in the basal layer express keratin 5 and keratin 14. C, Localization of keratin 10 to differentiating cells in intermediate layers of the epithelium. These cells eventually lose their nuclei and form the surface layers of cornified cells. D, Drawings illustrating the effects of keratin mutations on the structure of the epithelium. Dominant negative keratin mutations affect the assembly of keratin filaments wherever they are expressed. Human patients with epidermolysis bullosa simplex have point mutations in keratin 5 or keratin 14 that disrupt the filaments in the basal cells of the stratified epithelium, causing mechanical fragility and cellular rupture with mild trauma, resulting in blisters. Mutations in keratin 1 or keratin 10 cause cell rupture in the middle layers of the epithelium where they are expressed. Null mutations in keratin genes disrupt the epithelium to a lesser extent than dominant negative point mutations. E, Light micrograph of a histologic section of skin illustrating how a mutation in keratin 10 disrupts cells in the spinous layer of the epithelium and causes hyperkeratosis (excess scaling of surface layers). (A–C and E, Courtesy P. Coulombe, Johns Hopkins University, Baltimore, MD. D, Modified from Fuchs E, Cleveland DW. A structural scaffolding of intermediate filaments in health and disease. Science. 1998;279:514–519. Copyright 1998 American Association for the Advancement of Science.)
CHAPTER 35 n Intermediate Filaments
619
TABLE 35.2 Proteins Associated With Intermediate Filaments Name
Molecule
Distribution
Partners
Diseases
BPAG-1
Multiple splice isoforms (a, b, e, n)
a: Hemidesmosomes b: Muscle, cartilage e: Epithelial hemidesmosomes n: Neurons
IFs, MTs, actin
Autoimmune bullous pemphigoid
Desmoplakin
Two splice isoforms
Desmosomes
IFs; cadherin and other desmosome proteins
Autoimmune pemphigus; genetic striate palmoplantar keratoderma
Plectin
Multiple splice isoforms
Most tissues except neurons
IFs, actin, MTs, spectrin, β4-integrin
Autoimmune pemphigus; genetic epidermolysis bullosa with muscular dystrophy
Ten 37-kD filaggrins cut by proteolysis from profilaggrin
Cornified epithelia
Aggregates keratin
?
Plakins
Epidermal Filaggrin
Lamin Associated LAP1
57–70-kD isoforms
Integral nuclear membrane proteins
Binds lamins to nuclear envelope
LAP2
50 kD
Integral nuclear membrane protein
Binds lamins to nuclear envelope
LBR
73 kD
Integral nuclear membrane protein
Emerin
34 kD
? Peripheral protein of the inner nuclear membrane
Pelger-Huët anomaly; Greenberg skeletal dysplasia Binds actin filaments to the nuclear envelope
Emery-Dreifuss muscular dystrophy
ABD, actin binding domain; IFs, intermediate filaments; MTs, microtubules.
(Table 35.2). Integral membrane proteins, called nuclear envelope transmembrane proteins, anchor nuclear lamins to the nuclear membrane (see Fig. 9.8). Filaggrin mediates the aggregation of keratin filaments in the upper layers of skin. Plakins are giant proteins that link cytoskeletal polymers to each other and to membranes by virtue of binding sites for cytoskeletal polymers and proteins of adhesive junctions. Like several other plakins, plectin has globular domains on both ends of a 200-nm coiledcoil. Binding sites in both globular domains enable plectin to crosslink intermediate filaments to each other, to actin filaments, and microtubules (Fig. 35.7). Recessive mutations in human plectin cause a rare form of muscular dystrophy associated with skin blisters. The null mutation in mice is lethal. Plectin and two other plakins link keratin filaments to three different plasma membrane adhesion proteins at desmosomes and hemidesmosomes (see Fig. 31.8). Similarly, desmoplakin anchors keratin filaments to cadherins at desmosomes (see Fig. 31.8B). At hemidesmosomes plectin 1a links keratin to β4-integrins, while the plakin BPAG1e (bullous pemphigoid antigen 1-e) links keratin filaments to the transmembrane protein, BPAG2 (bullous pemphigoid antigen 2). BPAG1e is one of the many splice forms of the dystonin gene, which is mutated in some patients with neuropathies and one form of epidermolysis bullosa.
Functions of Intermediate Filaments in Cells Intermediate filaments function primarily as flexible intracellular tendons (analogous to nylon rope) that prevent excessive stretching of cells that are subjected to external or internal physical forces. This function is complemented by their interactions with microtubules, actin filaments, and membranes. For example, if a relaxed smooth muscle is stretched, the intracellular network of desmin filaments between cytoplasmic dense bodies and the plasma membrane (see Fig. 39.23) is transformed from a polygonal three-dimensional network into a continuous strap that runs the length of the cell (Fig. 35.8). Up to the point at which this network is taut, the cell offers little resistance to stretching. Beyond this point, the cell strongly resists further stretching. Actin filaments anchored to dense bodies interact with myosin (see Fig. 39.23) to apply contractile force to the network of intermediate filaments. Although the geometry of the network of intermediate filaments is different in striated muscles, the concept is remarkably similar to smooth muscle. Desmin filaments surround the Z disks in addition to forming a looser, longitudinal basket around the myofibrils (see Fig. 39.8). The ends of both skeletal and cardiac muscle cells must be anchored to transmit their contractile
620
SECTION IX n Cytoskeleton and Cellular Motility
C-terminal glopular domain
Coiled-coil
Spectrin repeats
ABD
N*
A
C
Actin β4-Integrin MAP2, tau
B
β4-Integrin Intermediate filaments β dystroglycan
C
D
FIGURE 35.7 PLECTIN STRUCTURE AND ACTIVITIES. A, Domain structure of plectin with the some of its known ligands listed to the right: the N-terminal actin-binding domain and the spectrin repeats are similar to those of α-actinin (see Fig. 33.17); the 170-nm long, α-helical coiled-coil forms dimers; six C-terminal repeats form a large globular domain. B, Electron micrograph of plectin molecules. C, Electron micrograph of an extracted fibroblast reacted with gold-labeled antibodies to plectin. Gold particles (yellow) identify plectin molecules (blue) as linkers between intermediate filaments (orange) and microtubules (red). The specimen was prepared by rotary shadowing. The molecules are pseudocolored for clarity. To visualize these interactions numerous actin filaments were removed by incubation with a gelsolin fragment. D, Drawing of plectin (blue) connecting cytoskeletal polymers to each other. (B, Courtesy G. Wiche, University of Vienna, Austria. C, Courtesy G. Borisy, University of Wisconsin, Madison.)
Passive tension (g)
15
Unstretched
Stretched
10
5
0 14
17
20
23
26
Length (mm) FIGURE 35.8 SMOOTH MUSCLE CELL INTERMEDIATE FILAMENTS FORM AN INTRACELLULAR TENDON THAT RESISTS EXCESSIVE STRETCHING. The graph shows that a relaxed smooth muscle resists stretching very little up to a length of 21 mm. Resistance (passive tension) increases dramatically with further stretching. At short lengths, the three-dimensional network of intermediate filaments and dense bodies is open, offering little resistance to stretching. At the inflection point of the resistance curve, the filaments are extended linearly from one end of the cell to the other and so resist further stretching. (Data from Cooke P, Fay R. Correlation between fiber length, ultrastructure, and the length tension relationship of mammalian smooth muscle. J Cell Biol. 1972;52:105–116.)
forces. This is accomplished by intercellular junctions that combine features of desmosomes or hemidesmosomes (anchoring intermediate filaments) and adherens junctions (anchoring actin filaments). In heart muscle cells, these hybrid junctions are called intercalated disks (see Fig. 39.18).
Keratin intermediate filaments are the major proteins in the epithelial cells of the skin, where they form a dense network connected to numerous desmosomes and hemidesmosomes (see Figs. 35.1 and 35.6). These junctions anchor a physically continuous network of intermediate filaments, imparting mechanical stability to the epithelium. If either the junctions or keratin filaments fail, cells pull apart or rupture, and the skin blisters. Mutations that compromise keratin intermediate filament assembly or the junctions to which they are anchored illustrate the importance of this network. Point mutations near the ends of the keratin rod cause especially severe forms of skin diseases (such as epidermolysis bullosa simplex) characterized by blistering and sensitivity to mechanical stress. Similar mutations engineered in transgenic mice faithfully reproduce the human disease. The expression pattern of the defective keratin determines which epithelial cells are affected. For example, a mutation in the rod domain of keratin 14 or keratin 5 leads to disruption of the basal cells in the epidermis where these keratins are expressed. Similarly, mutations in keratin 10 or keratin 1 cause cellular rupture at higher cell layers in the epidermis where those keratins are found (see Fig. 35.6E). Similarly, mutations in keratin 12 or keratin 3 cause sores on the cornea of the eye, where they are expressed. A mutant keratin can cause disease in heterozygotes that express one normal keratin gene. This is called a dominant negative mutation (or autosomal dominant mutation). Defective keratin subunits assemble
CHAPTER 35 n Intermediate Filaments
621
50 nm
MT
IFs
A
B
FIGURE 35.9 ELECTRON MICROGRAPHS OF INTERMEDIATE FILAMENTS (CALLED NEUROFILAMENTS) IN AXONS OF NERVE CELLS. A, A thin cross section shows clusters of intermediate filaments and microtubules. B, A longitudinal freeze-fracture preparation shows a microtubule (MT [red]) with associated vesicles and many intermediate filaments (IF [orange]). (A, Courtesy P. Eagle, Kings College, London, United Kingdom. B, Courtesy N. Hirokawa, University of Tokyo, Japan.)
imperfectly with normal keratin subunits and thereby compromise the physical integrity and strength of the filaments. The affected epithelial cells can grow, divide, and even form desmosomes with neighboring cells, but they tear apart physically when subjected to the shearing forces that affect the skin during normal activities. Young children are severely affected, but some patients improve with age. They learn to avoid physical trauma to their skin and may also adapt biochemically in some way. In contrast to these dominant negative keratin mutations, complete loss of a keratin subunit by a null mutation can be less severe (see Fig. 35.6D). Mice and humans that lack keratin 14 suffer from milder blistering than do patients with dominant negative point mutations of keratin 14. Mice without functional keratin 8 or keratin 18 genes may die during embryonic development, but a few survive with only modest defects in their colon and liver. Remarkably, mice also survive deletion of both copies of the gene for desmin have only mildly disorganized muscle architecture, although vigorous exercise is fatal. In contrast humans heterozygous for many different desmin mutations may suffer severely from generalized muscle failure. Other desmin mutations cause severe dilated cardiomyopathy requiring heart transplantation. In addition to providing mechanical stability neurofilaments have a second function of equal importance. Once a nerve cell forms synapses (see Figs. 17.9 and 17.10), it produces neurofilaments, apparently to expand the diameter of the axon (Fig. 35.9). This enhances electrical communication in the nervous system, because the velocity of action potentials (see Fig. 17.6) depends
on the diameter of the axon. Japanese quail with a truncation mutation of NFL gene are viable, but the diameters of their axons are smaller than normal and their coordination is defective. Lamins were originally thought to be a simple support network for the nuclear envelope, but they have other important functions. For example, mutations that create toxic fragments of lamins may perturb lamin assembly and thereby interfere with DNA replication. This effect may reflect a role for the lamina in organizing the chromosomal architecture in the interphase nucleus. Most remarkably, more than 400 human mutations in the lamin A/C gene LMNA cause diverse human diseases. These include premature aging (progeria) (see Fig. 9.10), the Emery-Dreifuss form of muscular dystrophy, and multiple disorders of fat tissue and nerves. These tissuespecific deficiencies are remarkable given the ubiquitous expression of lamins A and C in all tissues. ACKNOWLEDGMENTS We thank Ueli Aebi and Harald Herrmann for their detailed suggestions on the revision of this chapter and their contributions of new illustrations. SELECTED READINGS Bouameur JE, Favre B, Borradori L. Plakins, a versatile family of cytolinkers: roles in skin integrity and in human diseases. J Invest Dermatol. 2014;134:885-894. Chernyatina AA, Guzenko D, Strelkov SV. Intermediate filament structure: the bottom-up approach. Curr Opin Cell Biol. 2015;32: 65-72.
622
SECTION IX n Cytoskeleton and Cellular Motility
Chernyatina AA, Nicolet S, Aebi U, Herrmann H, Strelkov SV. Atomic structure of the vimentin central α-helical domain and its implications for intermediate filament assembly. Proc Natl Acad Sci USA. 2012;109:13620-13625. Clemen CS, Herrmann H, Strelkov SV, Schröder R. Desminopathies: pathology and mechanisms. Acta Neuropathol. 2013;125:47-75. Erber A, Riemer D, Bovenschulte M, Weber K. Molecular phylogeny of metazoan intermediate filament proteins. J Mol Evol. 1998;47: 751-762. Helfand BT, Chang L, Goldman RD. The dynamic and motile properties of intermediate filaments. Annu Rev Cell Dev Biol. 2003;19: 445-467. Herrmann H, Aebi U. Intermediate filaments: molecular structure, assembly mechanism, and integration into functionally distinct intracellular scaffolds. Annu Rev Biochem. 2004;73:749-789. Jefferson JJ, Leung CL, Liem RK. Plakins: goliaths that link cell junctions and the cytoskeleton. Nat Rev Mol Cell Biol. 2004;5:542-553. Kirmse R, Portet S, Mücke N, et al. A quantitative kinetic model for the in vitro assembly of intermediate filaments from tetrameric vimentin. J Biol Chem. 2007;282:18563-18572. Köster S, Weitz DA, Goldman RD, Aebi U, Herrmann H. Intermediate filament mechanics in vitro and in the cell: from coiled coils to filaments, fibers and networks. Curr Opin Cell Biol. 2015;32:82-91.
Leung CL, Green KJ, Liem RKH. Plakins: a family of versatile cytolinker proteins. Trends Cell Biol. 2002;12:37-45. Lowery J, Kuczmarski ER, Herrmann H, Goldman RD. Intermediate filaments play a pivotal role in regulating cell architecture and function. J Biol Chem. 2015;290:17145-17153. Moller-Jensen J, Löwe J. Increasing complexity of the bacterial cytoskeleton. Curr Opin Cell Biol. 2005;17:75-81. Nöding B, Herrmann H, Köster S. Direct observation of subunit exchange along mature vimentin intermediate filaments. Biophys J. 2014;107:2923-2931. Omary MB, Coulombe PA, McLean WH. Intermediate filament proteins and their associated diseases. N Engl J Med. 2004;351:2087-2100. Peter A, Stick R. Evolutionary aspects in intermediate filament proteins. Curr Opin Cell Biol. 2015;32:48-55. Szeverenyi I, Cassidy AJ, Chung CW, et al. The Human Intermediate Filament Database: comprehensive information on a gene family involved in many human diseases. Hum Mutat. 2008;29:351-360. Wiche G. Role of plectin in cytoskeleton organization and dynamics. J Cell Sci. 1998;111:2477-2486. Worman HJ, Courvalin J-C. The nuclear lamina and inherited disease. Trends Cell Biol. 2002;12:591-598.
CHAPTER
36
Motor Proteins M
olecular motors use adenosine triphosphate (ATP) hydrolysis to power movements of subcellular com ponents, such as organelles and chromosomes, along the two polarized cytoskeletal fibers: actin filaments and microtubules. No motors are known to move on intermediate filaments. Motor proteins also produce force locally within the network of cytoskeletal poly mers, which transmits these forces to determine the shape of each cell and, ultimately, the architecture of tissues and whole organisms. Chapters 37 to 39 and 44 illustrate how motors move cells and their internal parts. Just three families of motor proteins—myosin, kinesin, and dynein—power most eukaryotic cellular movements (Fig. 36.1 and Table 36.1). During evolution, myosin, kinesin, and Ras family guanosine triphospha tases (GTPases) appear to have shared a common ances tor (Fig. 36.1), whereas dynein is a member of the AAA adenosine triphosphatase (ATPase) family (Box 36.1). Although the ancestral genes appeared in prokary otes, and prokaryotes have homologs of both actin and tubulin, none of these motor proteins has been found in prokaryotes. Over time, gene duplication and diver gence in eukaryotes gave rise to multiple genes for myosin, dynein, and kinesin, each encoding proteins
with specialized functions. Even the slimmed down genome of budding yeast includes genes for five myosins, six kinesins, and one dynein. Table 36.1 lists other protein machines that produce molecular movements during protein and nucleic acid synthesis, proton pumping, and bacterial motility. Motor proteins have two parts: a motor domain that uses ATP hydrolysis to produce movements and a tail that allows the motors to self-associate and/or to bind particular cargo. Within the three families, the tails are more diverse than the motor domains, allowing for spe cialized functions of each motor isoform. All motor proteins are enzymes that convert chemical energy stored in ATP into molecular motion to produce force upon an associated cytoskeletal polymer (Fig. 36.2). If the motor is anchored, the polymer may move. If the polymer is anchored, the motor and any attached cargo may move. If both are anchored, the force stretches elastic elements in the molecules transiently, but nothing moves, and the energy is lost as heat. Cells use all these options (see Chapters 37 to 39). Biochemists originally discovered and purified these motors using enzyme (eg, ATP hydrolysis) or in vitro motility assays (Fig. 36.11). With the prototype motors identified, investigators found further examples and
Primodial NTPase Primodial GTPase
Many contemporary GTPases
Primodial myosin
Primodial kinesin
Many contemporary myosins
Many contemporary kinesins
Primodial AAA ATPase
Dyneins
Other AAAATPases
FIGURE 36.1 Evolution of myosin, kinesin, and dynein adenosine triphosphatase (ATPase) motors from genes that encoded two primordial proteins that bound and hydrolyzed nucleoside triphosphates. Gene duplication and divergence created genes for many contemporary motors.
623
624
SECTION IX n Cytoskeleton and Cellular Motility
TABLE 36.1 Mechanochemical Enzymes and Other Proteins That Produce Movements Families
Track
Direction
Cargo
Energy
Muscle myosin
Actin
Barbed end
Myosin filament
ATP
Myosin II
Actin
Barbed end
Myosin, actin
ATP
Myosin I
Actin
Barbed end
Membranes
ATP
Myosin V
Actin
Barbed end
Organelles
ATP
Myosin VI
Actin
Pointed end
Endocytic vesicles
ATP
Axonemal
Microtubule
Minus end
Microtubules
ATP
Cytoplasmic
Microtubule
Minus end
Membranes, chromosomes
ATP
Kinesin-1
Microtubule
Plus end
Membranes, intermediate filaments
ATP
Kinesin-14
Microtubule
Minus end
? Microtubules
ATP
ATPases Myosins
Dyneins
Kinesins
Other Mechanochemical Systems Polymerases and Helicases Ribosome
mRNA
5′ to 3′
None
GTP
DNA polymerase
DNA
5′ to 3′
None
ATP
RNA polymerase
DNA
5′ to 3′
None
ATP
CMG DNA helicase
DNA
—
DNA
ATP
RNA helicases
RNA
—
RNA
ATP
None
None
Cell, basal body
Ca2+
Actin filaments
None
Barbed end
Membranes
ATP
Microtubules
None
Plus end
Chromosomes
GTP
Worm sperm MSP
None
Not polar
Cytoskeleton
Bacterial flagella
None
Bidirectional
Cell
H+ or Na+ gradient
F-type ATPase
None
Bidirectional
None
H+ or ATP
V-type ATPase pump
None
None
ATP
Conformational System Spasmin/centrin Polymerizing Systems
Rotary Motors
ATP, adenosine triphosphate; ATPase, adenosine triphosphatase; GTP, guanosine triphosphate; mRNA, messenger RNA; MSP, major sperm protein. The terms “barbed” and “pointed” end refer to the appearance of actin filaments decorated with a myosin fragment (Fig. 33.8).
variant isoforms of each by purifying proteins, cloning complementary DNAs (cDNAs), sequencing genomes, or genetic screening.
Myosins Myosins are the only motors that are known to use actin filaments as tracks. Members of the diverse myosin superfamily arose from a common ancestor and share a motor unit called a myosin “head” that produces force on actin filaments (Fig. 36.3). One or two heads are attached to various types of tails that are adapted for diverse purposes, including polymerization into filaments, binding membranes, and interacting with various cargos. Myosin heads consist of two parts. A catalytic domain at the N-terminus of the myosin heavy chain binds and hydrolyzes ATP and interacts with actin filaments. Light chain domains consist of an α-helical extension of the
heavy chain from the catalytic domain associated with one to seven light chains. Light chains are related to calmodulin (see Fig. 3.12), which also serves as a light chain for many myosins.
Myosin Mechanochemistry Myosin was discovered in skeletal muscle and used to establish general principles that apply, with interesting variations, to energy transduction by all myosins. Muscle myosin is responsible for the forceful contraction of skeletal muscle (see Chapter 39). Like other types of myosin-II, it has two heads on a long tail formed from an α-helical coiled-coil. These tails polymerize into bipolar filaments (see Figs. 5.7 and 39.6). The head of muscle myosin was originally isolated as a proteolytic fragment called subfragment-1 (Fig. 36.3). The N-terminal 710 residues of the heavy chain form the globular catalytic domain. The nucleotide binding site in the core of the catalytic domain is formed by a β-sheet
CHAPTER 36 n Motor Proteins
Generic motor with stretched spring Force
Cytoskeletal fiber
Motor ATP
Spring
ADP + Pi Force
Support or cargo
Resulting movement with anchored motor Fiber moves
Support
Resulting movement with anchored fiber
Motor and cargo move Cargo
Result with anchored fiber and anchored motor Force
Force
Support
Spring stretched, force transmitted through fiber to anchoring sites, no movement, energy lost as heat FIGURE 36.2 GENERAL FEATURES OF ATPase MOTORS. Motors bind stably to a support or cargo and transiently to a cytoskeletal fiber (actin filament or microtubule). Energy liberated by adenosine triphosphate (ATP) hydrolysis produces force to stretch an elastic element somewhere in the physical connection between the cargo and the cytoskeletal fiber. The resulting motion depends on whether the force in the spring exceeds the resistance of the fiber or the cargo.
flanked by α-helices with a topology similar to Ras GTPases (see Fig. 4.6) despite little sequence similarity. The γ-phosphate of ATP inserts deeply into the nucleotidebinding site with the adenine exposed on the surface. Actin binds more than 4 nm away from the nucleotide on the other side of the head. A region of the heavy chain called the converter subdomain is attached to the lightchain domain composed of an essential light chain and a regulatory light chain wrapped around and stabilizing a long α-helix formed by the heavy chain. The inter action of light chains with the heavy chain α-helix
625
BOX 36.1 AAA Adenosine Triphosphatases The common ancestor of life on the earth had a gene for a versatile adenosine triphosphate (ATP)-binding domain. Through gene duplication and divergence this progenitor gave rise to the AAA family of adenosine triphosphatases (ATPases) in all branches of the phylogenic tree. Given the remarkable variety of functions of the contemporary proteins, the name “ATPases Associated with Diverse Activities” is apt. The family now includes regulatory subunits of proteasomes (see Fig. 23.8); proteases from prokaryotes, chloroplasts, and mitochondria; Hsp100 protein folding chaperones; dynein microtubule motors (Fig. 36.14); the microtubule severing protein katanin (see Fig. 34.8); activators of origins of replication (including ORC1, 4, and 5 and Mcm-7 [see Fig. 42.8]); clamp loader proteins for DNA polymerase processivity factors (see Fig. 42.12); two proteins required for peroxisome bio genesis (see Table 18.1); and proteins involved in vesicular traffic such as NSF (the N-ethylmaleimide-sensitive factor [see Fig. 21.15]). AAA domains have a common fold with a catalytic site that binds and hydrolyzes ATP. A “Walker A” motif of conserved residues interacts with the β- and γ-phosphates of ATP, and “Walker B” motif residues participate in ATP hydrolysis. Many AAA ATPases form ring-shaped hexamers of identical subunits or up to six different AAA subunits although the dynein heavy chain has six AAA domains in one large polypeptide. Often, an arginine residue from the adjacent subunit in the hexamer inserts into the active site and facilitates conformational changes in response to ATP binding and release of the γ-phosphate.
A
B
Actin-binding site
Active site
Actin-binding site
Active site
ELC
ELC
RLC
RLC
FIGURE 36.3 ATOMIC STRUCTURE OF THE HEAD OF MUSCLE MYOSIN. A, Ribbon drawing of the polypeptide backbones. B, Space-filling model. Heavy chain residues 4–204 (green); heavy chain residues 216–626 (red); heavy chain residues 647–843 (purple): essential light chain (ELC [yellow]); regulatory light chain (RLC [orange]). The myosin light chains consist of two globular domains connected by an α-helix, like calmodulin and troponin C. (For reference, see Protein Data Bank [PDB; www.rcsb.org] file 2MYS.)
626
A
SECTION IX n Cytoskeleton and Cellular Motility
B
C
Beginning of stroke
D
Pointed end
Catalytic domain
End of stroke
FIGURE 36.4 ACTIN FILAMENTS DECORATED WITH MYOSIN HEADS. A, Electron micrograph of frozen-hydrated actin filaments fully occupied with myosin heads. B, Three-dimensional reconstruction from electron micrographs of an actin filament saturated with myosin heads. C, Superimposition of atomic models of the actin filament and one myosin head on the reconstruction of the decorated filament (blue cage-like surface). D, Space-filling atomic model of an actin filament with one attached muscle myosin head showing the light-chain domain in two positions: (1) the end of the power stroke as observed in the absence of ATP (blue), and (2) the postulated beginning of the power stroke (pink) deduced from X-ray structures of isolated heads and spectroscopic studies. The catalytic domain (red) is fixed in one position on actin (yellow). (Courtesy R. Milligan, Scripps Research Institute, La Jolla, CA.)
resembles calmodulin binding to its target proteins (see Fig. 3.12). Myosin heads bind tightly and rigidly to actin fila ments in the absence of ATP. This is called a rigor complex, because it forms in muscle during rigor mortis when ATP is depleted after death. Myosin heads bound along an actin filament form a polarized structure, resem bling a series of arrowheads when viewed from the side (Fig. 36.4). The heads bind at an angle and wrap around the filament. Their orientation defines the barbed and pointed ends of the actin filament (see Fig. 33.8). All known myosins, except myosin-VI, move toward the barbed end of the filament. The atomic structures of the myosin head and actin filament fit nicely into the three-dimensional struc ture of the decorated filament determined by electron microscopy, providing the structural starting point for understanding the mechanics of force production (Fig. 36.4). Each myosin head contacts two adjacent actin subunits.
Actomyosin Adenosine Triphosphatase Cycle Myosin uses energy from ATP hydrolysis to move actin filaments, so an appreciation of the mechanism requires an understanding of the biochemical steps along the reaction pathway. Fig. 36.5A looks intimidating, but working through it one step at a time reveals its logic and simplicity. Note that the mechanism consists of two parallel lines of chemical intermediates. First, consider the bottom line showing the reactions that explain why myosin alone turns over ATP remarkably slowly, at a rate of only approximately 0.02 s−1: Step 1. At physiological concentrations of ATP, myosin binds ATP in less than 1 millisecond. Energy from ATP binding allows a conformational change in the myosin
that can be detected by a change in the intrinsic fluo rescence of the protein itself. Step 2. The enzyme catalyzes the hydrolysis of ATP. This reaction is moderately fast (>100 s−1) and readily revers ible. The equilibrium constant for hydrolysis on the enzyme is near 1, so each ATP is hydrolyzed to adenos ine diphosphate (ADP) and inorganic phosphate and the triphosphate is resynthesized several times before the products dissociate from the enzyme. ATP splitting provides energy for a second conformational change, reflected in a further increase in the fluorescence of the myosin. These conformational changes reorient the converter subdomain and the light chain domain poised to undergo the molecular rearrangements that subsequently produce movement. Step 3. Inorganic phosphate (P) slowly dissociates from the active site (at a rate of approximately 0.02 s−1) by escaping through a narrow “back door” on the far side of the enzyme. This is the rate-limiting step in the pathway. The loss of phosphate is coupled to confor mational changes that return myosin toward its basal state. The phosphate dissociation step has the largest negative free energy change, so it is presumed that energy derived from ATP binding and hydrolysis and stored in conformational changes in the myosin head is used to do work or dissipated as heat at this point in the reaction pathway. Step 4. Once phosphate dissociates, ADP leaves rapidly from the “front door.” To summarize, in the absence of actin filaments, ATP binds rapidly to myosin and is rapidly but reversibly split, and the products slowly dissociate from the active site. The overall cycle of the enzyme is limited by the slow conformational change coupled to phosphate dissocia tion. Energy derived from ATP binding and hydrolysis is
CHAPTER 36 n Motor Proteins
A
Strong
Weak
A–M
A–M*T
A–M**DP
A–MD
1′
2′
3′
2
3
M*T ≥ 1000 s-1
B
Strong
1 M
627
M
Free myosin
4 MD
10 s-1
Myosin bound to actin
4′
M**DP 100 s-1
A–M
1 s-1
0.1 s-1
P
ADP
ADP + Pi
B
ATP
ADP + Pi Pi
Rapid equilibrium free and bound
Phosphate dissociates Light chain domain rotates
ADP dissociates
ATP binding Head dissociates
ATP hydrolysis FIGURE 36.5 MYOSIN ATPase MECHANISMS. A, A diagram of the actomyosin ATPase cycle of striated muscle myosin-II showing the actin filament (A), myosin head (M), ATP (T), adenosine diphosphate (ADP) (D), and inorganic phosphate (P). Transient-state kinetics revealed the major chemical intermediates and the rate constants for their transitions. Arrow sizes are proportional to the rates of the reactions, with secondorder reactions adjusted for physiological concentrations of reactants. One or two asterisks indicate conformational changes in the myosin head induced by ATP binding and hydrolysis. Myosin without nucleotide (M) and myosin with ADP (MD) bind much more tightly to actin filaments than do AMT and AMDP. The weakly bound AMT and AMDP intermediates are in a rapid equilibrium with free MT and MDP. The beige shading shows the main pathway through the reaction. B, The postulated force-producing structural changes in the orientation of the light-chain domain (purple and blue) coupled to the myosin ATPase cycle. (B, Data from R. Vale, University of California, San Francisco, and R. Milligan, Scripps Research Institute, La Jolla, CA.)
used for a conformational change in the myosin head that is dissipated when phosphate dissociates. The upper line in Fig. 36.5A shows myosin associated with an actin filament. The chemical intermediates are the same, but some of the key rate constants differ for the actin-bound and free myosin. Steps 1 and 2 are similar to those of free myosin, but step 3—the dissocia tion of phosphate—is much faster when a head is bound to an actin filament. As a result, myosin bound to actin traverses the ATPase cycle approximately 200 times faster than myosin free in solution, and ATP hydrolysis becomes the rate-limiting step. This effect of actin is referred to as “actin activation of the myosin ATPase.” A practical advantage of this mechanism is that the ATPase cycle is essentially turned off unless the head interacts with an actin filament. Finally, consider the vertical arrows representing tran sitions between bound and free states of each myosin chemical intermediate. All myosin intermediates bind rapidly to actin filaments, but the dissociation rate constants vary over a wide range depending on the
nucleotide that is bound to the active site of the myosin. Myosin with no nucleotide or with bound ADP alone dissociates very slowly and therefore binds tightly to actin filaments. Myosin with bound ATP or ADP+Pi dissociates rapidly from actin, so these states bind actin weakly. One cycle of ATP hydrolysis takes about 50 millisec onds, but a single pathway cannot be drawn through the reaction mechanism of ATP, myosin, and actin owing to the rapid equilibrium of myosin intermediates (MT and MDP) hopping on and off actin filaments on a millisec ond time scale. Starting with AM, ATP binds very rapidly and sets up a rapid, four-way equilibrium including AMT, MT, AMDP, and MDP—the major intermediates during steady-state ATP turnover in muscle (see Chapter 39). Because the products of ATP hydrolysis dissociate much more rapidly from AMDP than from MDP, the favored pathway out of this four-way equilibrium is through AMDP to AMD and back to AM. The overall ATPase rate depends on the actin concentration, which determines the fraction of myosin heads bound to actin in the
628
SECTION IX n Cytoskeleton and Cellular Motility
AMDP state. At the high actin concentrations in cells, a significant fraction of myosin heads is associated with actin (approximately 10% in contracting muscle), but each molecule continues to exchange on and off actin filaments.
Transduction of Chemical Energy Into Molecular Motion Myosin heads produce force during the transition from the AMDP state to the AMD and AM states. Production of force at this step makes sense for two reasons: First, the large free-energy difference between AMDP and AMD provides sufficient energy to produce force; second, the force-producing AMD and AM intermediates bind tightly to actin, so any force between the motor and the actin track is not dissipated. However, for many myosins, including skeletal muscle myosin, these forceproducing states occupy a small fraction of the whole ATPase cycle. The fraction of the time in force producing states is called the duty cycle. ADP dissociates rapidly from AMD, and ATP binds rapidly to AM, dissociating myosin from the actin filament and initiating another ATPase cycle. Fifty years of research using a combination of mechan ical measurements, static atomic structures of myosin heads with various bound nucleotides, and spectro scopic observations of contracting muscle revealed the structural basis for the conversion of free energy into force: a dramatic conformational change in the orienta tion of the light-chain domain associated with phosphate dissociation (Fig. 36.5B). Elegant mechanical experiments measured the size of the mechanical step produced by a myosin during one cycle of ATP hydrolysis. These experiments on live muscles first suggested that each cycle of ATP hydrolysis moves an actin filament approximately 5 to 10 nm relative to myosin. Now one may observe myosin moving single actin filaments by fluorescence microscopy. An array of myosin heads attached to a microscope slide can use ATP hydrolysis to push actin filaments over the surface (Fig. 36.6A–C). Assays with single myosin molecules show that each cycle of ATP hydrolysis can move an actin filament up to 5 to 15 nm and develop a force of about 3 to 7 piconewtons (pN) (Fig. 36.6D). At low ATP concentrations, the interval between the force-producing step and the binding of the next ATP is relatively long, so single steps can be observed. Further insights emerged from biophysical studies of muscle and purified proteins using x-ray diffraction (see Fig. 39.11), electron microscopy, electron spin reso nance spectroscopy, and fluorescence spectroscopy. These experiments showed that the light-chain domain pivots around a fulcrum, the converter subdomain within the catalytic domain, which is stationary relative to the actin filament. For example, spectroscopic probes on
light chains revealed a change in orientation when muscle is activated to contract, whereas probes on the catalytic domain do not rotate. Crystal structures of myosin heads with various bound nucleotides and nucle otide analogs show that the light-chain domain can pivot up to 90 degrees (Fig. 36.4D). The light-chain domain is bent more acutely in the AMT and AMDP intermediates and pivots to a more extended orientation when phos phate dissociates (Fig. 36.3). ADP dissociation extends this rotation of some classes of myosin. Consistent with rotation of the light-chain domain producing movement, the rate of actin filament gliding in an in vitro assay is proportional to the length of the light-chain domain. The observed range of orientations of the light-chain domain relative to the catalytic domain can account for the observed step size of 10 nm for muscle myosin. Some aspects of these conformational changes and their rela tion to phosphate release are similar to Ras family GTPases (see Fig. 4.6). Rotation of the light-chain domain is believed to produce movement indirectly in the sense that forceproducing intermediates stretch elastic elements in the system. This mechanism is represented by a spring in Fig. 36.2. The elastic elements in the myosin-actin complex are most likely to be mainly in the myosin head, with small contributions from the actin and myosin fila ments. Movement of the light-chain domain tensions the spring transiently in the AMD and AM states. Dissociation of ADP and rebinding of ATP to the AM intermediate reverts the system to the rapid equilibrium of mostly dissociated weakly bound intermediates. Any force left in the spring is lost as soon as the head dissociates from the actin filament. The actual motion produced depends on the mechani cal resistance in the system (Fig. 36.2). If both myosin and actin are fixed, elastic elements are stretched for the life of the force-producing states (AMD and AM), and the energy is lost as heat when the head dissociates. This happens when one tries to lift an immovable object. If the resistance is less than the force in the stretched elastic elements, the actin filament moves relative to myosin, as in muscle contraction. The distance moved in each step depends on the resistance, as the spring stops shortening when the forces are balanced.
Myosin Superfamily Eukaryotes have 35 classes of myosin and many other examples of unique myosins in single species (Fig. 36.7). All arose from a gene similar to myosin-I in the last eukaryote common ancestor more than a billion years ago. The primordial gene then gave rise to the gene for myosin-V, so this class is also widespread. Gene duplica tion, divergence, and acquisition of extra domains pro duced many other myosin genes encoding proteins specialized for particular biological functions made pos sible by variations of the mechanochemical ATPase cycle
CHAPTER 36 n Motor Proteins
A
B
629
D
Actin filament
Step
Return Myosin
C
40
ATP
ADP + Pi
Myosin
Distance (nm)
Actin
Events
20
0
–20
GLASS
0
0.5
1.0
1.5
Time (s)
FIGURE 36.6 IN VITRO MOTILITY ASSAYS WITH PURIFIED MUSCLE MYOSIN AND ACTIN FILAMENTS. A–C, Actin filament gliding assays. A, Filaments are labeled with rhodamine-phalloidin to render them visible by light microscopy. ATP hydrolysis by myosin moves actin filaments over the surface with the pointed end leading as the myosins walk toward the barbed end of the filaments. B–C, Drawings of actin filaments moving over myosin heads immobilized on a glass coverslip. D, Measurement of the muscle myosin step size. An actin filament is attached between two plastic beads, which are suspended by laser optical traps. The optical traps move the filament near a myosin molecule on the surface of another bead attached to the microscope slide, allowing a myosin head to attach to the actin filament. When supplied with ATP, a single myosin head can move the actin filament a short distance corresponding to the step size. The graph shows the time course of displacements of the actin filament and attached beads. Brownian motion limits the precision of the measurement of the size of these steps to a range of 5 to 15 nm. The duration of the step depends on the ATP concentration, because ATP dissociates the force-producing AM state, allowing the force of the optical traps to return the beads and the actin filament to their original position. Pi, inorganic phosphate. (A, Courtesy A. Bresnick, Albert Einstein College of Medicine, New York. D, For reference, see Finer JT, Simmons RM, Spudich JA. Single myosin molecule mechanics: piconewton forces and nanometer steps. Nature. 1994;368:113–119.)
and acquisition of diverse tails to interact with cargo. Within a myosin class, the tails are similar to each other, but between classes, tails are diverse in terms of their ability to polymerize and interact with other cellular components including membranes and ribonucleopro tein particles. No organism has genes for all 35 classes of myosin and a few species, including Giardia lamblia, have no myosin genes. Myosin-I is most widespread, but plants and related organisms lost this gene. A primitive myosin-V gene gave rise to plant myosin-VIII and myosin-XI, which move at very high speeds (see Fig. 37.9). Organisms on the branch including amoebas, yeast and animals have genes for myosins types I, II, and V, but myosin genes diversified in animals, so humans have 40 myosin genes from 13 classes. Gene duplications gave rise to multiple
isoforms within most classes of myosin. For instance, the vertebrate smooth muscle myosin gene arose from dupli cation of a gene for a cytoplasmic myosin-II. Establishing the biological functions of the various myosins has been challenging. Biochemical characteriza tion of cargo and localization in cells provide some clues, but genetic or biochemical knockouts often have mild effects, probably owing to overlapping functions of the myosins and the capacity of some cells to adapt to their loss, at least under laboratory conditions. Myosin-I was the first “unconventional myosin” discovered—unconventional in the sense that it differed from the type II myosin originally isolated from skeletal muscle. These myosins have one head and short tails with various types of domains, including a basic domain with affinity for acidic phospholipids. The presence of
630 Class
SECTION IX n Cytoskeleton and Cellular Motility Example
Heavy chain domains Head
I. Dictyostelium MyoB I. Bovine BB myosin-I
Distribution
Absent
Most eukaryotes, stramenopiles, others
Plants, ampicomplexa
++ Coiled-coil
II. Chicken muscle myosin-II III. Drosophila ninaC long
Architecture
Membrane binding GPQ SH3 IQ motif
Kinase
Amoebas, fungi, animals Plants, others Arthropods, chordates
++
V. Chicken myosin V/dilute
Fungi, plants, others
Amoebas, fungi, animals Plants, stramenopiles
VI. Porcine myosin VI
Animals
Fungi, plants, others
VII. Human myosin VIIA
Animals
Fungi, plants, others
VIII. Arabidopsis ATM1
Algae, plants
Fungi, plants, others
Animals
Fungi, plants, others
TH4
Talin
IX. Human myosin IXb pH domains
X. Bovine myosin X
Deuterostomes, Cnidaria Other animals, others
XI. Arabidopsis MYA1
Algae, plants
Fungi, animals, others
XII. C. elegans Myo12
Nemotodes only
All others
Ampicomplexa only
All others
XIV. Toxoplasma gondii myosin A
+
= 100 amino acids
FIGURE 36.7 THE MYOSIN FAMILY. Drawing of myosin heavy chain domains and molecular models of myosin isoforms showing catalytic domains (rose); IQ motifs, light-chain–binding sites (rose bars); basic domains with affinity for membrane lipids (violet); SH3 (Src homology 3) domains (dark green); coiled-coil (orange); kinase domain (light blue); and pleckstrin homology domain (blue). (For reference, see Odronitz F, Kollmar M. Drawing the tree of eukaryotic life based on the analysis of 2,269 manually annotated myosins from 328 species. Genome Biol. 2007;8:R196. See also Myosin Home Page, available at http://www.mrc-lmb.cam.ac.uk/myosin/myosin.html.)
an Src homology 3 (SH3) domain (see Fig. 25.10) allows some type I myosins to bind proline-rich sequences in other proteins. Those with an actin filament–binding domain separate from the motor domain can crosslink actin filaments. With duty cycles of less than 10%, mul tiple myosin heads must work together in concert to move membranes. Mutations show that myosin-I partici pates in endocytosis, as expected from its concentration at sites of phagocytosis and macropinocytosis. In micro villi of intestinal epithelial cells, myosin-I links actin fila ments laterally to the plasma membrane (see Fig. 33.2B). Heavy-chain phosphorylation activates myosin-I from lower eukaryotes, whereas calcium binding to calmodu lin light chains regulates myosin-I from the intestinal brush border. The myosin-II class includes various muscle and cytoplasmic myosins that also have two heads, two IQ motifs, and long coiled-coil tails. Assembly of tails into bipolar filaments (see Fig. 5.7) allows myosin-II to pull together oppositely polarized actin filaments during muscle contraction (see Chapter 39) and cytokinesis (see Fig. 44.24). As in smooth muscle (see Fig. 39.23), phos phorylation of the regulatory light chain activates myosin-II in animal nonmuscle cells. In addition, phos phorylation of the heavy chain regulates the enzyme activity and/or polymerization of some myosin-IIs. Myosin-V moves pigment granules, ribonucleopro tein particles and other cellular components (see Fig. 37.11). A long light-chain domain with seven IQ motifs
allows myosin-V to take long steps along the actin fila ment (Fig. 36.8). These steps are processive, because slow ADP dissociation from the AMD intermediate allows time for the other head to take a long step and bind an actin subunit 36 nm beyond the first head toward the barbed end of the filament. Mechanical strain after the step may modestly increase the rate of ADP dissociation from the trailing head. This cooperation between the heads initiates ATP binding and the next ATPase cycle, as the motor walks deliberately along the filament. These features make myosin-V a valuable model for the lever arm for movements of the whole myosin family. Myosin-VI arose in metazoan cells and is the only myosin known to move toward the pointed end of actin filaments. Unique features of the converter domain result in the lever arm swinging opposite to the conventional direction. The lever arm is a long, single α-helix begin ning with a single IQ-motif associated with calmodulin. Lacking coiled-coil, myosin-VI is a monomer unless adapter proteins bring together the C-terminal globular cargo-binding domains of two molecules. These dimers can take huge steps of approximately 30 nm, but myosin-VI can also act as a tether, because the forceproducing AMD and AM states occupy a large fraction of the ATPase cycle, owing to slow ADP dissociation from AMD state and slow ATP binding to AM. These features allow myosin-VI to move endocytic vesicles from the plasma membrane into the cytoplasm and to contribute to the formation of autophagosomes. Myosin-VII and
CHAPTER 36 n Motor Proteins
A
Strong
Weak
A–M
A–M*T
M
M*T ≥ 1000 s-1
B
Pointed
Strong A–M**DP
A–MD
M**DP 100 s-1
10 s-1
A–M
MD 1 s-1
M
0.1 s-1
ADP ATP
ADP
ADP ADP
631
ADP
ADP
ADP
ADP-Pi
Barbed
ADP
Force produced by ATP binds trailing Trailing head steps New leading head leading head head which forward 72 nm and binds actin dissociates ADP dissociates hydrolyzes ATP from trailing head Pi dissociates FIGURE 36.8 MYOSIN-V MECHANISM. A, ATPase cycle with ADP release as the rate-limiting step rather than phosphate dissociation as for muscle myosin (see Fig. 36.5A). B, Relationship of mechanical steps to the ATPase cycle. Shown are actin filament (A), myosin head (M), ATP (T), ADP (D), and inorganic phosphate (Pi). (For reference, see De La Cruz EM, Ostap EM. Relating biochemistry and function in the myosin superfamily. Curr Opin Cell Biol. 2004;16:61–67. For a movie of myosin-V stepping on actin filaments, see Kodera N, Yamamoto D, Ishikawa R, et al. Video imaging of walking myosin V by high-speed atomic force microscopy. Nature. 2010;468:72–76.)
myosin-X are also monomers with single chain α-helices as lever arms. Myosin mutations cause human diseases. Loss-of-func tion mutations in the genes for myosins-IIA, -IIIA, -VI, -VIIA, and -XV cause deafness and vestibular dysfunction. Mutations in the genes for cardiac muscle myosin heavy and light chains are responsible for many cases of car diomyopathies (see Table 39.1).
Microtubule Motors The kinesin and dynein families of molecular motors use energy from ATP hydrolysis to move vesicles, membranebound organelles, chromosomes, and other cargo along microtubules (see Fig. 37.1). Dynein also powers bend ing motions of eukaryotic flagella and cilia (see Fig. 38.14). Dyneins move themselves and any cargo toward the minus end of microtubules. Most kinesins move in the opposite direction, toward the plus end, but some kinesin family members are minus-end-directed motors and others promote microtubule disassembly.
Like myosins, microtubule motors have heads with ATPase activity and tails that interact with cargo.
Kinesins Kinesin-1 is a processive motor that moves cargo, such as an organelle, continuously toward the plus end of a microtubule. The two heads are attached to an α-helical coiled-coil tail, much like myosin-II, except both the heads and coiled-coil are smaller (Fig. 36.9). Each head, consisting of approximately 340 residues, is a motor unit that binds microtubules and catalyzes ATP hydrolysis. Light chains associated with the C-terminal bifurcation of the tail bind cargo molecules (Fig. 36.9E). Because the kinesin head is less than half the size of a myosin head and because the proteins lack appreciable sequence homology, the atomic structure of kinesin-1 (Fig. 36.9C) revealed a major surprise: The small kinesin head is folded like the core of the catalytic domain of myosin! In fact, this core, which consists of a central, mixed β-sheet flanked by helices, is similar to the
632 A
SECTION IX n Cytoskeleton and Cellular Motility
N
B
D. Structural overlap
1
E. Kinesin light chain with cargo peptide
α3 β4
Myosin
Head
Kinesin
Tail
Tail
Coiled-coil
Neck
β6 β7 β3 β8 β1
α2 α1
β2
Cargo peptide
C. Kinesin structure ATP
C
955 Light chains
FIGURE 36.9 STRUCTURE OF KINESINS. A, Domain architecture of the polypeptide sequence of the heavy chain of kinesin-1. B, Sketch of kinesin-1 showing two heads and the coiled-coil tail with light chains bound at the distal end. C, Ribbon diagram of the polypeptide backbone of the kinesin head showing ATP as a space-filling model (green), the neck-linker residues (red), and the proximal part of the coiled-coil tail. D, Superimposition of the core of the kinesin-1 head on the catalytic domain of myosin showing the structural homology of the proteins. The detailed ribbon diagram shows only the homologous elements of secondary structure. The overview (right) shows kinesin-1 (blue) superimposed on the structure of the whole head of skeletal muscle myosin (pink). E, Ribbon model of a kinesin-1 light chain (purple) with a bound cargo peptide (green) with a DWED motif. (C, For reference, see PDB file 3KIN and Sack S, Muller J, Marx A, et al. X-ray structure of motor and neck domains from rat brain kinesin. Biochemistry. 1997;36:16155–16165. D, Superimposed ribbon diagrams courtesy of R. Vale, University of California, San Francisco. E, For reference, see PDB file 3ZFW and Pernigo S, Lamprecht A, Steiner RA, et al. Structural basis for kinesin-1: cargo recognition. Science. 2013;340:356–359.)
considerably smaller Ras family GTPases (see Fig. 4.6). This provided strong evidence that all three families of nucleoside triphosphatases evolved from a common ancestor (Fig. 36.1). ATP binds to a site on kinesin that is homologous to the guanosine triphosphate (GTP)binding site of Ras, but the enzyme mechanisms differ in important ways. The microtubule-binding site is some distance from the ATP-binding site (Fig. 36.10).
Kinesin Mechanochemistry Single kinesin-1 heads, produced experimentally from truncated cDNAs, traverse a microtubule-stimulated ATPase cycle much like myosin (Fig. 36.11A). Both bind and hydrolyze ATP rapidly followed by slower release of phosphate and ADP. However, in contrast to muscle myosin, kinesin heads may remain bound to the micro tubule through multiple cycles of ATP hydrolysis rather than dissociating when bound to ATP or ADP-Pi. In vitro motility assays (Fig. 36.12) revealed that a two-headed kinesin-1 can move along a single (or two parallel) microtubule protofilaments for long distances at 0.8 µm/s. The motor makes discrete steps of 8 nm, the spacing of successive tubulin dimers in a microtu bule. Each step takes 10 milliseconds when kinesin is moving at full speed. This step is remarkably large for the small ( 100 s-1
300 s-1
Mt–KDP
Mt–KD
KDP
KD
100 s-1
10 s-1
633
1 s-1
Mt–K
0.1 s-1
(+)
B
ADP
ADP 0
α β
0
ADP
ATP ATP
ADP + Pi
Pi
ADP
ADP
(–)
Trailing head weakly associates with MT
ATP binds leading head Trailing head rotates
New trailing head hydrolyzes ATP New leading head binds MT and dissociates ADP
Pi dissociates from trailing head weakening head's binding to MT
FIGURE 36.11 KINESIN-1 ATPase MECHANISM. A, A diagram of the kinesin-microtubule ATPase cycle for a single kinesin-1 head showing the kinesin (K), microtubule (Mt), ATP (T), ADP (D), and inorganic phosphate (P). Arrow sizes are proportional to the rates of the reactions, with second-order reactions adjusted for physiological concentrations of reactants. The beige shading shows two pathways through the reaction, one along the top line without dissociation, and the other with dissociation from the microtubule. B, Hand-over-hand, processive stepping of kinesin along a microtubule. The empty, microtubule-bound head binds and hydrolyzes ATP resulting in a conformational change favoring docking of its neck-linker (green), thereby moving forward the detached head with its undocked (pink) neck-linker. Binding of the new leading head to the microtubule causes a conformation change that dissociates ADP. Dissociation of the γ-phosphate from the trailing head results in its dissociation from the microtubule. This returns the heads to their original condition, but with the motor advanced 8 nm with the heads in the opposite chemical states. Pi, inorganic phosphate. (B, Data from R. Vale, University of California, San Francisco, and R. Milligan, Scripps Research Institute, La Jolla, CA. For reference, see Cao L, Wang W, Jiang Q, et al. The structure of apo-kinesin bound to tubulin links the nucleotide cycle to movement. Nat Commun. 2014;5:5364.)
Processive movement depends on the ability of kinesin to remain associated with the microtubule through more than a hundred cycles of ATP hydrolysis. This is made possible by cooperation between the two heads that ensures at least one head is bound to the microtubule throughout the ATPase cycle (Fig. 36.11B). Reciprocal affinities for nucleotide and microtubules allow the two heads to alternate between microtubule binding and dissociation. For example, if kinesin-1 with ADP bound to both heads is mixed with microtubules, one head binds a microtubule and rapidly dissociates its ADP, leaving the other head dissociated with bound ADP. Binding and hydrolysis of ATP on the open site of the head associated with the microtubule, drives a con formational change that repositions the trailing head forward so it can pass the bound head and bind the next tubulin dimer toward the plus end of the microtubule. Association of the new leading head with the microtu bule promotes dissociation of its bound ADP. Dissocia tion of the γ-phosphate from the new trailing head
weakens its affinity for the microtubule. Its detachment from the microtubule with bound ADP brings the cycle back to its starting point with kinesin having advanced 8 nm (Fig. 36.12). Experiments with single kinesin-1 molecules labeled with fluorescent dyes showed that they alternate steps on the right and left sides of the microtubule like a gymnast walking on a balance beam. The mechanism of stepping is postulated to be the docking and undocking of a segment of the kinesin-1 heavy chain linking the motor domain to the coiled-coil neck and tail (Fig. 36.10). With ATP bound the motor domain has a conformation that favors association of the “neck-linker” peptide, as in the X-ray structure of dimeric kinesin (Fig. 36.9C). When either no nucleotide or ADP is bound, the conformation of the motor domain releases the neck-linker peptide.
Kinesin Superfamily The last eukaryotic common ancestor had only a single myosin but already had at least 11 families of kinesins
634
A
SECTION IX n Cytoskeleton and Cellular Motility
B
Microtubule movement
C 96 80
(+)
Kinesin stepping toward (+) end
Distance (nm)
(–)
Bead movement
(–)
64 48 32 16
Kinesin stepping toward (+) end
0
(+)
1
2
3
4
5
Time (s)
FIGURE 36.12 IN VITRO MOTILITY ASSAYS FOR MICROTUBULE MOTORS. A, Gliding assay. Kinesin or dynein that is attached to a microscope slide uses ATP hydrolysis to move microtubules over the surface. A single kinesin-1 molecule can move a microtubule in this assay. Microtubules can be imaged by video-enhanced differential interference contrast microscopy (see Fig. 34.6) or fluorescence microscopy. B, Bead assay. Kinesin or dynein that is attached to a plastic bead uses ATP hydrolysis to move the bead along a microtubule attached to the microscopic slide. C, Experimental measurement of the kinesin-1 step size using the bead assay. The bead is held in a laser optical trap so that 8-nm steps can be recorded, as a single, two-headed kinesin-1 moves a bead processively along a microtubule, as in B. The position of the bead is recorded with nanometer precision by interferometry. (For reference, see Svoboda K, Schmidt CF, Schnapp BJ, et al. Direct observation of kinesin stepping by optical trapping interferometry. Nature. 1993;365:721–727.)
Kinesin structures N-terminal motor
Example
Kinesin-1
Hs Kif5B
Kinesin-2
Sp Krp85/95
Kinesin-3
Mm Kif1b
Kinesin-4
Xl KIp1
Kinesin-5
Dm Klp61f
Kinesin-7
Hs CENP-E
Heavy chain domains Head
Architecture
Coiled-coil Tail 85 95
Internal motor Kinesin-13
Xl MCAK
C-terminal motor Kinesin-14
Dm Ncd
FIGURE 36.13 KINESIN FAMILY. A, Phylogenetic relationships of some of the kinesins based on the sequences of the motor domains. B, Drawing of kinesin heavy chain domains and molecular models of kinesin isoforms showing the catalytic domain (red), coiled-coil tail (orange), and tail piece (blue). (Data from R. Case and R. Vale, University of California, San Francisco. For reference, see Lawrence CJ, Dawe RK, Christie KR, et al. Standardized kinesin nomenclature. J Cell Biol. 2004;167:19–22; and Dagenbach EM, Endow SA. A new kinesin tree. J Cell Sci. 2004;117:3–7. See also the Kinesin Home Page at https://labs.cellbio.duke.edu/kinesin.)
with motor domains associated with a variety of coiledcoil stalks and tails (Fig. 36.13 and Table 36.2). Thus the microtubule system was much more developed than the actin system at this point in evolution. Contempo rary eukaryotes have genes for 17 families of kinesins
(often with multiple isoforms) plus a few more para logs identified in isolated species. On the other hand, many eukaryotes have lost one or more kinesin genes; for example, amoebas lack kinesin-2 and alveolates lack kinesin-7.
CHAPTER 36 n Motor Proteins
635
TABLE 36.2 Kinesin Superfamily: Classification and Examples of Kinesin-Family Motor Proteins Examples
Subunits (kD)
Velocity (µm s−1)
Functions
Kinesin-1
Human KHC
2 × 110, 2 × 70
+0.9
Organelle movement
Kinesin-2
Urchin KRP85/95
1 × 79, 1 × 84, 1 × 115
+0.4
Organelle movement
Kinesin-3
Mouse KIF1B
1 × 130
+0.7
Mitochondria movement
Kinesin-4
Xenopus Kp11
2 × 139
+0.2
Chromosome movement
Kinesin-5
Fly KLP61F
4 × 121
+0.04
Pole separation, mitosis
Kinesin-7
Human CENP-E
2 × 340
+0.1
Kinetochore-microtubule binding
MCAK
2 × 83
—
Microtubule disassembly
Fly Ncd
2 × 78
−0.2
Mitotic/meiotic spindle
Class N-Terminal Motor
Internal Motor Kinesin-13 C-Terminal Motor Kinesin-14
Modified from Vale RD, Fletterick RJ. The design plan of kinesin motors. Annu Rev Cell Dev Biol. 1997;13:745–777. More data on kinesins are available at the Kinesin Home Page, https://labs.cellbio.duke.edu/kinesin.
All members of the kinesin family have similar motor domains attached to a variety of tails that interact with cargo (Fig. 36.13). Motor domains are generally found at the N-terminus, but may be located in the middle (kinesin-13) or at the C-terminus (kinesin-14). Regardless of their location, motor domains have similar structures. Most kinesins are dimeric, with two polypeptides joined in a coiled-coil. Most are homodimers, but kinesin-2 not only forms homodimers but also heterotri mers consisting of two different polypeptides with motor domains plus another large subunit. Tetrameric kinesin-V molecules bind to two microtubules with opposite polarities and move them apart. Most kinesins move toward the plus end of the micro tubule, but C-terminal kinesin-14 motors move toward the minus end. The reverse direction of kinesin-14 move ment is not explained by either the architecture of the motor domain, the ATPase mechanism, which is similar to kinesin-1, or the attachment of the N-terminus of the motor domain to the stalk. Instead, the proximal part of the Ncd coiled-coil stalk rotates approximately 70 degrees toward the minus end of the microtubule when ATP binds to the active site. Kinesins transport a variety of cargo, including chro mosomes and organelles, along microtubules. Kinesin-1 moves membrane vesicles toward the plus end of micro tubules away from the centrosome and along axons of neurons (see Fig. 37.1). The light chain links the end of the kinesin-1 tail to cargo proteins (Fig. 36.9E). Kinesin-2 is the motor for anterograde intraflagellar transport, movement of particles toward the tip of the axoneme in cilia and flagella (see Fig. 38.18). It also transports a variety of cargos over long distances in the cytoplasm. Kinesin-4 motors (also called chromokinesins) bind both DNA and microtubules. In neurons, they move cargo along axons. In dividing cells they participate in the formation of condensed mitotic
chromosomes (see Fig. 44.7). Kinesin-7 (originally called CENP-E) concentrates at kinetochores where it helps move the chromosome toward the middle of the mitotic spindle prometaphase (see Fig. 44.5). Bipolar kinesin-5 motors form an antiparallel tetramer of two dimeric kine sins that bridge a pair of oppositely polarized microtu bules and push apart the poles of the mitotic spindle (see Fig. 44.7). Both kinesin-8 and kinesin-13 use cycles of ATP hydro lysis to remove tubulin dimers from the ends of micro tubules. Kinesin-8 motors to the plus end where it works, whereas kinesin-14 lacks motor activity but can diffuse on the microtubule surface to reach either end. This depolymerizing activity is important for mitosis. Most kinesins appear to be constitutively active, but intramolecular interactions autoinhibit kinesin-1. The tail folds back and binds between the heads, shutting off the motor. Adapter proteins compete the tail from the head, freeing the motor to be active.
Dyneins Dynein microtubule-based motors are AAA ATPases (Box 36.1), so their evolutionary origin differs from myosins and kinesins (Fig. 36.1). Most AAA ATPases consist of six separate ATPase domains, but in dynein these domains (AAA1–AAA6) are concatenated in a giant heavy chain of nearly 500 kD rather than separate polypeptides (Fig. 36.14A). Cytoplasmic dynein is a dimer of two heavy chains plus accessory polypeptides that bring the total molecular weight to approximately 1.4 MDa. The N-terminal quarter of the heavy chain forms a tail that interacts with intermediate chains, light intermediate chains, dimers of light chains, and cargo molecules (Fig. 36.14B). A linker domain connects the tail to AAA1. A segment of the dynein heavy chain within AAA4 forms an antiparallel coiled-coil stalk with a small microtubulebinding domain at the tip.
636 A
N
SECTION IX n Cytoskeleton and Cellular Motility
B. Whole dynein molecule
0
DHC dimerization domain DIC WD40
Tail
D
Linker AAA1
Tail
AAA5
LC8
2
Tctex
“Post power stroke” ADP in AAA1
AAA2
Linker 1
“Pre power stroke” ATP analog in AAA1
AAA6
DLIC
1388
AAA-ATPase domains
C
AAA3
Buttress of AAA5
AAA4
Motors 3 4
5
Coiled-coil stalk Microtubule binding site
6 C
Coiled-coil stalk of AAA4
Microtubule binding domain of AAA4 4730
FIGURE 36.14 DYNEIN STRUCTURE. A, Domain organization of a dynein heavy chain showing the six AAA ATPase modules and two sequences that form an antiparallel coiled-coil stalk with an ATP-sensitive microtubule-binding site at the tip. The first AAA domain is the catalytic site. AAA domains 2, 3, and 4 bind ATP, but hydrolysis is not coupled to movement. AAA domains 5 and 6 do not bind ATP. B, Model for cytoplasmic dynein-1 based on crystal structures of motor domains and reconstructions of electron micrographs of the dynactin complex. C, Ribbon diagram of crystal structures of cytoplasmic dynein-1 motor domains (Apo) without nucleotide bound to AAA1. The linker domain is purple. AAA domains are color coded as in A, with the stalk and microtubule binding domain extending from AAA3. D, Space-filling models of dynein with (ADP-Vo) with ADP and the phosphate analog vanadate bound to AAA1 the “pre power stroke” state and ADP bound to AAA1 the “post power stroke state.” The AAA hexamer is more compact with bound ADP-Vo. These two structures differ in the conformations of the AAA hexamer, linker domain and stalk. (For reference, see EMData Bank files 2861 and 2862 and Schmidt H, Zalyte R, Urnavicius L, et al. Structure of human cytoplasmic dynein-2 primed for its power stroke. Nature. 2015;518:435–438; and Urnavicius L, Zhang K, Diamant AG, et al. The structure of the dynactin complex and its interaction with dynein. Science. 2015;347:1441–1446.)
Dynein Mechanochemistry Sufficient information is available from enzyme kinetics, crystal structures, and motility assays to construct a mechanochemical cycle for dynein interacting with a microtubule (Fig. 36.15A). The AAA1 domain binds and hydrolyzes ATP during force-producing interactions with microtubules. Full motor function requires ADP or ATP binding to AAA domains 2 to 4, but ATP hydrolysis by these domains is not coupled directly to motility. AAA domains 5 and 6 do not bind nucleotides. The dynein ATPase cycle of AAA1 resembles the actomyosin ATPase mechanism in broad outline. When AAA1 is free of nucleotide, dynein binds tightly to a microtubule at a site between the α- and β-tubulin subunits with the stalk pointing toward the minus end of the polymer. ATP binding to AAA1 causes compaction of the whole AAA hexamer and produces two important conformational changes (Fig. 36.14). First, the “buttress” on AAA5 communicates the conformational change in the hexamer to the stalk by displacing the stalk helices relative to each other. This, in turn, changes the confor mation of the microtubule-binding domain more than 20 nm distant from the ATP binding site and reduces its affinity for the microtubule. Thus, dynein-ATP dissoci ates from the microtubule. Second, the linker domain bends in the middle, moving to the “pre power stroke state.” After ATP hydrolysis, dynein-ADP-Pi also binds
weakly to microtubules, but phosphate dissociation during one of the transient interactions with a microtu bule reverses both conformational changes produced by ATP. The linker domain straightens out, producing a power stroke that moves the tail and any associated cargo (including the other subunit of a dynein dimer) toward the minus end of the microtubule. In addition, the affinity for microtubules increases, allowing trans mission of force from the microtubule to the cargo. Binding to a microtubule also stimulates the rate of ADP dissociation from dynein approximately 10-fold, from approximately 3 s−1 to approximately 33 s−1, restarting the ATPase cycle. Yeast dynein dimers can walk processively toward the minus end of a microtubule in in vitro motility assays. The mechanism likely involves steps by the two motor domains on adjacent protofilaments of the microtubule. The size of the mechanical step associated with each ATP hydrolysis in most often 8 nm, but cytoplasmic dynein can take larger steps up to 24 nm when the load is low. Yeast dynein produces a force of 5 to 7 pN, similar to kinesin.
Dynein Superfamily Dynein genes are ancient, arising well before the last common eukaryotic ancestor (see Fig. 2.4B), but they were lost multiple times during evolution, so neither red
CHAPTER 36 n Motor Proteins
Apo
S
ATP
ADP•P
W
Strong
ADP
W
Mt–DyT
?
S
Strong
P Mt–Dy
Apo
S
Weak
637
Mt–DyDP
30 s-1
Mt–DyD
Mt–Dy
P Dy
DyT
DyDP
ATP
ADP•P
DyD
3 s-1
Dy
≥ 1000 s-1 10 s-1
100 s-1
1 s-1 0.1 s-1
FIGURE 36.15 DYNEIN-MICROTUBULE ATPase MECHANISM. Chemical pathway and structures. Arrow sizes are proportional to the rates of the reactions, with second-order reactions adjusted for physiological concentrations of reactants. The beige shading shows the main pathway through the reaction. D, ADP; Dy, dynein; Mt, microtubule; P, inorganic phosphate; T, ATP. The drawings are interpretations of the intermediates in the cycle based on crystal structures. (Modified from Cianfrocco MA, DeSantis ME, Leschziner AE, et al. Mechanism and regulation of cytoplasmic dynein. Annu Rev Cell Dev Biol. 2015;31:83–108.)
algae nor flowering plants now have dynein genes. Animals have two genes for cytoplasmic dynein heavy chains and multiple isoforms of intermediate, light inter mediate, and light chains. Alternative splicing, especially of intermediate chains, further increases the complexity. Cytoplasmic dyneins are dimeric proteins (Fig. 36.14B). Cytoplasmic dynein-2 moves cargo for intrafla gellar transport (see Fig. 38.18). Cytoplasmic dynein-1 has diverse functions around the entire cell cycle. During interphase, dynein-1 transports organelles, RNAs, and some viruses toward the minus ends of microtubules (see Fig. 37.2) generally moving cargo toward the cell center where the centrosome and Golgi apparatus are located. In neurons, dynein-1 moves cargo inside axons toward the cell body (see Fig. 37.3). During mitosis, dynein-1 in the cell cortex and bound to kinetochores of chromosomes applies forces to microtubules and helps to position the mitotic spindle (see Fig. 44.7). Given these diverse activities, it is not surprising that dynein mutations cause serious phenotypes and contrib ute to disease. A null mutation in the gene for a mouse cytoplasmic dynein-1 heavy chain leaves the Golgi appa ratus dispersed throughout the cytoplasm and is lethal during embryogenesis. A temperature-sensitive mutation in Caenorhabditis elegans dynein-1 causes defects in mitosis at the restrictive temperature. Mutations of dynein-1 and associated proteins contribute to human neurodegenerative diseases including some cases of Par kinson disease and spinal muscular atrophy. Accessory proteins regulate the enzyme activity and movements of cytoplasmic dyneins. Transport by mammalian dynein depends on the dynactin complex
(see Fig. 37.2), a huge complex of 23 subunits (11 dif ferent proteins) with a total molecular weight of approx imately 1.2 MDa. Dynactin and an adapter protein not only link cytoplasmic dynein-1 to cargo, but also stimu late its motor activity, perhaps by positioning the two motors in a productive way. A complex of proteins (Lis-1 and Nudel/NudE) increases dynein’s affinity for microtubules, slows the motor, and may increase force production. Nudel/NudE binds intermediate chains and tethers Lis-1 to dynein. Lis-1, a β-propeller protein, interacts directly with the AAA hexamer and sterically blocks a movement of the linker domain that is required for microtubule release. Genetic experiments suggest that Lis1 is a dynein acti vator, but how tight binding to microtubules activates dynein remains unclear. Loss-of-function mutations of the Lis1 genes interfere with development of the cerebral cortex, which lacks gyres and is smooth in affected humans. Humans have 14 genes for axonemal dyneins, which consist of one to three heavy chains. In axonemes of cilia and flagella, at least seven different dynein isoforms bind to unique sites on the outer doublets (see Fig. 38.14). Calcium and a cyclic adenosine monophosphate (cAMP)– dependent protein kinase (see Fig. 25.3D) regulate dynein in cilia and flagella. ACKNOWLEDGMENTS We thank Andrew Carter, Erika Holzbaur, Samara ReckPeterson, and Lee Sweeney for their suggestions on revi sions to this chapter.
638
SECTION IX n Cytoskeleton and Cellular Motility
SELECTED READINGS Bhabha G, Johnson GT, Schroeder CM, et al. How dynein moves along microtubules. Trends Biochem Sci. 2016;41:94-105. Bloemink MJ, Geeves MA. Shaking the myosin family tree: biochemical kinetics defines four types of myosin motor. Semin Cell Dev Biol. 2011;22:961-967. Buss F, Spudich G, Kendrick-Jones J. Myosin VI: Cellular functions and motor properties. Annu Rev Cell Dev Biol. 2004;20:649-676. Carter AP, Diamant AG, Urnavicius L. How dynein and dynactin trans port cargos: a structural perspective. Curr Opin Struct Biol. 2016; 37:62-70. Cianfrocco MA, DeSantis ME, Leschziner AE, et al. Mechanism and regulation of cytoplasmic dynein. Annu Rev Cell Dev Biol. 2015; 31:83-108. Cross RA, McAinsh A. Prime movers: the mechanochemistry of mitotic kinesins. Nat Rev Mol Cell Biol. 2014;15:257-271. De La Cruz EM, Ostap EM. Relating biochemistry and function in the myosin superfamily. Curr Opin Cell Biol. 2004;16:61-67. Erzberger JP, Berger JM. Evolutionary relationships and structural mechanisms of AAA+ proteins. Annu Rev Biophys Biomol Struct. 2006;35:93-114. Fu MM, Holzbaur EL. Integrated regulation of motor-driven organelle transport by scaffolding proteins. Trends Cell Biol. 2014;24: 564-574. Geeves MA, Holmes KC. Structural mechanism of muscle contraction. Annu Rev Biochem. 1999;68:687-728. Greenberg MJ, Ostap EM. Regulation and control of myosin-I by the motor and light chain-binding domains. Trends Cell Biol. 2013; 23:81-89. Hammer JA 3rd, Sellers JR. Walking to work: roles for class V myosins as cargo transporters. Nat Rev Mol Cell Biol. 2011;13:13-26. Hartman MA, Finan D, Sivaramakrishnan S, et al. Principles of uncon ventional myosin function and targeting. Annu Rev Cell Dev Biol. 2011;27:133-155.
Hartman MA, Spudich JA. The myosin superfamily at a glance. J Cell Sci. 2012;125:1627-1632. King SM. Integrated control of axonemal dynein AAA(+) motors. J Struct Biol. 2012;179:222-228. Kull FJ, Endow SA. Force generation by kinesin and myosin cytoskeletal motor proteins. J Cell Sci. 2013;126:9-19. Milic B, Andreasson JO, Hancock WO, et al. Kinesin processivity is gated by phosphate release. Proc Natl Acad Sci USA. 2014;111: 14136-14140. Odronitz F, Kollmar M. Drawing the tree of eukaryotic life based on the analysis of 2,269 manually annotated myosins from 328 species. Genome Biol. 2007;8:R196. Preller M, Manstein DJ. Myosin structure, allostery, and mechanochemistry. Structure. 2013;21:1911-1922. Roberts AJ, Kon T, Knight PJ, et al. Functions and mechanics of dynein motor proteins. Nat Rev Mol Cell Biol. 2013;14:713-726. Schmidt H, Zalyte R, Urnavicius L, et al. Structure of human cytoplas mic dynein-2 primed for its power stroke. Nature. 2015;518:435438. For video of conformational change see . Scholey JM. Kinesin-2: a family of heterotrimeric and homodimeric motors with diverse intracellular transport functions. Annu Rev Cell Dev Biol. 2013;29:443-469. Sun Y, Goldman YE. Lever-arm mechanics of processive myosins. Biophys J. 2011;101:1-11. Tumbarello DA, Kendrick-Jones J, Buss F. Myosin VI and its cargo adaptors-linking endocytosis and autophagy. J Cell Sci. 2013;126: 2561-2570. Walczak CE, Gayek S, Ohi R. Microtubule-depolymerizing kinesins. Annu Rev Cell Dev Biol. 2013;29:417-441. Wickstead B, Gull K, Richards TA. Patterns of kinesin evolution reveal a complex ancestral eukaryote with a multifunctional cytoskeleton. BMC Evol Biol. 2010;10:110.
CHAPTER
37
Intracellular Motility Virtually every component inside living cells moves to
some extent (Fig. 37.1), but the magnitude and velocity of these movements vary by orders of magnitude (Table 37.1). At one extreme, the bulk cytoplasm of algae and giant amoebas streams tens of micrometers per second, but most cytoplasm is generally less dynamic. The cytoplasmic network of cytoskeletal polymers has a pore size of less than 50 nm (see Fig. 1.13). This allows small molecules and macromolecules to diffuse essentially unimpaired. Particles larger than the pores must be transported actively. For example, lysosomes, mitochondria, secretory vesicles, and endosomes all move actively in cytoplasm, frequently between the centrosome and the cell periphery. Similarly, messenger RNA (mRNA) moves from its site of synthesis in the nucleus through nuclear pores into the cytoplasm and then may be carried actively to specific parts of the cell. Intracellular pathogenic bacteria and viruses take advantage of the host cell’s actin system to propel themselves through the cytoplasm. Virus particles can also move along microtubules. Two ancient mechanisms (Fig. 37.1C) account for most intracellular movements in eukaryotes. Transport along microtubules by kinesin or dynein predominates in animal cells. Transport along actin filaments by myosin is more important in plants and fungi. Specialized isoforms of myosin, kinesin, and dynein are dedicated to particular movements. Experiments with drugs that depolymerize or stabilize actin filaments or microtubules (see Boxes 33.1 and 34.1) originally identified the cytoskeletal polymers that support various biological movements. Discovering the motors was more challenging, given multiple genes for myosin, kinesin, and dynein in most organisms and a limited choice of pharmacologic agents (Box 37.1). Even loss of function mutations and depletion by RNA interference (RNAi; see Fig. 11.12) have limitations, because some motors are essential for viability while others contribute redundantly to transport. For example, dynein contributes to mitosis, but Drosophila tissue culture cells depleted of dynein can
complete mitosis, because other motors take over allowing mitosis to proceed, albeit more slowly. The diversity of cargo, motors, and tracks poses a traffic control challenge, which cells manage by specifying the organization of microtubules and actin filaments, matching appropriate motors to specific cargo and regulating motor activity. Most cargos associate with more than one type of motor, so their activities must be coordinated. For example, some cargos with two motors reverse their direction either along the way or after reaching their destination. Local or global signaling pathways influence many of these choices, allowing cells to adjust transport to adapt to environmental conditions. This chapter takes a broad view across biology, highlighting the wide range of mechanisms that produce intracellular transport. In most cases, transport involves single organelles on an actin filament or microtubule, but organelle transport can result in bulk movement of the cytoplasm. The chapter also considers mechanisms that produce special types of intracellular movements: cytoplasmic contractions generated by myosin and actin filaments that propel cytoplasmic streaming; and polymerization and depolymerization of microtubules and actin filaments. Chapters on membrane traffic (see Chapters 20 to 22) and mitosis (see Chapter 44) cover more examples of intracellular movements.
Rapid Intracellular Movements on Microtubules Organelles in most eukaryotic cells can move along linear microtubule tracks at relatively high velocities, on the order of 1 µm s−1 (Table 37.1) at least part of the time. The physical organization of microtubules determines the patterns of these movements (Fig. 37.1; also see Fig. 34.2). In cells with a radial arrangement of microtubules (Fig. 37.1) organelles and other cargos associated with kinesin-1 move toward microtubule plus ends located at the periphery of cells. Cargo associated with 639
640
SECTION IX n Cytoskeleton and Cellular Motility
B
A
Nitella streaming Neuron Axon
Fibroblast
C
Synapse
Myosin Kinesin
Dynein
FIGURE 37.1 MECHANISMS OF INTRACELLULAR MOVEMENT. A, Fibroblasts and neurons move organelles bidirectionally along microtubules (red). B, The green alga Nitella moves cytoplasmic organelles along bundles of actin filaments located in the cell cortex. C, Microtubule-based and actin filament–based motors.
TABLE 37.1 Velocities of Intracellular Movements Velocity (µm s−1)
Mechanism
Anterograde fast axonal transport, squid
1
Individual kinesin motors
Retrograde fast axonal transport, squid
2
Individual dynein motors
Chromosome movement in anaphase of mitosis
0.003–0.2
Motors plus depolymerization
Endoplasmic reticulum sliding, newt cell
0.1
Individual kinesin motors
Slow axonal transport, rat nerves
0.002–0.1 net 1 (intermittent)
Motors on microtubules
0.1
Microtubule polymerization
Cytoplasmic streaming, Nitella
60
Myosin motors on tracks
Cytoplasmic streaming, Physarum
500
Actin–myosin contraction
0.5
Actin polymerization
System Microtubule Motors
Microtubule Polymerization Endoplasmic reticulum tip elongation, Newt cell Actin-Myosin Motors
Actin Polymerization Actin-propelled comet, Listeria
dynein moves in the opposite direction. These motors are processive (see Chapter 36), so they can work alone or in small numbers. Intraflagellar transport of proteins in cilia and flagella (see Fig. 38.18) shares many features with the movements of organelles. The chemistry of individual microtubules subtly influences the delivery of cargo to specific locations, as some motors favor microtubules composed of particular tubulin isoforms or with posttranslational modifications. For example, kinesin-1 favors acetylated microtubules,
and CENP-E favors tyrosinated microtubules in the center of the mitotic spindle. In addition, microtubule-associated proteins, such as tau, can influence the choice of tracks by reducing the run lengths of motors under some circumstances.
Matching Microtubule Motors With Cargo Pairing motors with appropriate cargo and regulation of motor activity control the traffic along microtubules. Coupling can be as simple as kinesin binding to a
641
CHAPTER 37 n Intracellular Motility
p25/p27
A Pointed end
B
p62
Cargo
C
Dynactin
Lysosome Arl8-GTP
BICD2
Arp11 -actin Arp1 Filament
Shoulder p150Glued/p135 p50, p24
PH
WD
KLC
Kinesin-1 Anterograde movement
(+) (–)
Bottom protofilament Barbed end
N
Dyenin (ADP)
Arp1
SKIP
RU
Dynein (ADP.Pi)
Microtubule
Top protofilament Capping protein
(–)
(+)
FIGURE 37.2 ATTACHMENT OF CYTOPLASMIC MICROTUBULE MOTORS TO MEMBRANES. A, Ribbon diagram of the structure of the core of the dynactin complex determined by electron cryomicroscopy. The short filament of Arp1 is capped by capping protein on the barbed end and by Arp11 on the pointed end. The shoulder is formed by part of the p150 molecule, most of which was disordered in these samples. B, Drawing showing dynein linked to a vesicle by dynactin complex and moving toward the minus end of a microtubule. C, Coupling of kinesin-1 to a lysosome. Guanosine triphosphatase (GTPase) Arl8 on the lysosome membrane engages the adapter protein SKIP (SifA-kinesin interacting protein), which binds to the kinesin light chain as shown in Fig. 36.9E. (A, From Urnavicius L, Zhang K, Diamant AG, et al. The structure of the dynactin complex and its interaction with dynein. Science. 2015;347:1441–1446. B, From Cianfrocco MA, DeSantis ME, Leschziner AE, ReckPeterson SL. Mechanism and regulation of cytoplasmic dynein. Annu Rev Cell Dev Biol. 2015;31:83–108. C, Based on a drawing from RosaFerreira C, Munro S. Arl8 and SKIP act together to link lysosomes to kinesin-1. Dev Cell. 2011;21:1171–1178.)
BOX 37.1 Tools for Studying Motor Proteins A small compound named monastrol inhibits kinesin-5, resulting in monopolar mitotic spindles that fail to segregate the chromosomes. Higher-affinity inhibitors of kinesin-5 are being tested for cancer therapy. The small molecule blebbistatin inhibits cytoplasmic and skeletal muscle myosin-II (but not smooth muscle myosin-II) and blocks cytokinesis. Vanadate and ultraviolet light can inactivate dynein. Vanadate binds to the γ-phosphate site of dynein–adenosine diphosphate (ADP), but it binds similarly to other adenosine triphosphatases (ATPases), so it is not specific. However, ultraviolet light has a novel effect on the dynein–ADP–vanadate complex: It cleaves and inactivates the dynein heavy chain.
transmembrane protein on a cargo vesicle, but usually involves adapter proteins that recognize a molecule on the surface of the cargo and one or more motors. Dynein associates with the dynactin complex (Fig. 37.2B). About a dozen other known adapter proteins also interact with the dynactin complex, kinesins or both to direct traffic to specific locations (Table 37.2). Local conditions in the cell regulate both coupling of motors to cargo and the activity of the motors. Other proteins tether some organelles to immobilize them and prevent their transport. The 1 mDa dynactin complex consists of a short filament of actin-related protein Arp1 capped on its barbed end by capping protein and on its pointed end by the most divergent actin-related protein, Arp11 (Fig. 37.2A).
Other subunits at the pointed end serve as cargo adaptors. Association of the dimeric tail of dynein with the dynactin complex and adapter proteins activates its motor activity. The 150-kD dynactin subunit (p150Glued) includes a long extension that links the Arp1 filament to both an intermediate chain of dynein and a microtubule. The two dynein motor domains take steps of variable size as they pull the cargo along the microtubule. Mutations of p150Glued cause developmental defects in the eye and brain of Drosophila plus some forms of Parkinson disease and motor neuron degeneration in humans. Most proteins known to link organelles and ribonucleoprotein particles to microtubule motors interact with both dynein and kinesins, but some are specific for one or the other (Table 37.2). These interactions often activate or inhibit motor activity, coupling physical association with transport. For example, some cargo adapters overcome autoinhibitory interactions between the heads and tails of kinesins-1, -2, and -3 by binding to the kinesin tail or light chains. Other than their linker and regulatory activities, adapter proteins have little in common, so they likely had independent evolutionary origins. The following sample illustrates the diversity of these adapters. Motors and adapter proteins are targeted to cargo membranes by integral membrane proteins, peripheral proteins or lipids. Kinesin binds directly to a few transmembrane proteins, so no adapter is required. For example, calsyntenin and alcalpha on certain neuronal vesicles have short sequences with acidic residues flanking a key tryptophan that bind kinesin light chains
642
SECTION IX n Cytoskeleton and Cellular Motility
TABLE 37.2 Adapters for Microtubule Motors Adapter Protein
Links to Motors
Links to Cargos
Cargo
Destination
Regulation
Disease Connection
Dynein Only Endosomes, mitochondria, Golgi membranes
Mutations cause neuron degeneration
Dynactin complex
Dynein IC, p150Glued, other adapters listed here
RILP (Rab7interacting lysosomal protein)
Dynactin
Rab7, spectrin
Late endosomes
Trans-Golgi network
Rab7 GTPase
SNX6 (sorting nexin 6)
Dynactin
Other retromer proteins
Endosome vesicles with recycling receptors
Trans-Golgi network
PI4P on transGolgi membranes releases dynactin
Kinesin Only DENN/MADD (Rab3 guanine nucleotide exchange factor)
Kinesin-3
Synaptic vesicles with Rab3
Synapse
Rab3 GTPase
Mint1 (Munc18interacting protein)
Kinesin-2 tail
Vesicles with NMDA glutamate receptors
Synapse
Phosphorylation by CaMKII
SKIP (SifA-kinesin interacting protein)
Kinesin-1 LC
Lysosomes
Periphery
Arl8 GTPase
Arl8 GTP → lysosome
Dynein and Kinesin Hook
Dynein, kinesin-3
Early endosomes
Bidirectional
Huntingtin
Dynein IC, HAP1 → dynactin, kinesin-1 HC and LC
HAP40 → Rab5, Optineurin → myosin-VI
Endosomes, lysosomes, autophagosomes, vesicles with APP or GABA receptors, mRNA
Bidirectional
Phosphorylation of Huntingtin and motors
JIP1 (JNK interacting protein)
Either dynactin or kinesin-1 stalk, tail & LC with aid of FEZ1
JNK, vesicles with amyloid precursor protein
Synaptic vesicles, autophagosomes, mitochondria, vesicles with APP
Bidirectional
Phosphorylation by JNK favors kinesin-1 binding and transport to plus end
JIP3/JIP4 (JNK interacting protein)
Dynactin, kinesin-1 tail and LC, JIP1
JNK, Arf6
Synaptic vesicles to synapse, endosomes with nerve growth factor to cell body
La
Dynein IC, kinesin-1 HC
RNPs
Ribonucleoprotein particles
Bidirectional
SUMOylation may reverse direction at periphery
Milton/TRAK
Dynactin, kinesin-1 HC
Miro/RhoT2 transmembrane GTPase
Mitochondria
Bidirectional
Local concentrations of ATP and Ca2+
Traffic defects and massive neuronal degeneration in Huntington disease
Phosphorylation of DIC
Parkinson disease (?)
Data from Fu MM, Holzbaur EL: Integrated regulation of motor-driven organelle transport by scaffolding proteins. Trends Cell Biol 2014;24:564–574. APP, amyloid precursor protein; ATP, adenosine triphosphate; CaMKII, calcium/calmodulin-dependent serine protein kinase; FEZ1, fasciculation and elongation protein zeta 1; GABA, γ-aminobutyric acid; GTP, guanosine triphosphate; GTPase, guanosine triphosphatase; HC, heavy chain; IC, intermediate chain; JNK, c-Jun N-terminal kinase; LC, light chain; mRNA, messenger RNA; NMDA, N-methyl-D-aspartate; RNP, ribonucleoprotein; PI4P, phosphatidylinositol 4-phosphate.
CHAPTER 37 n Intracellular Motility
(see Fig. 36.9E). In other cases an adapter protein links a transmembrane protein to a motor. For instance, a transmembrane protein called Miro on the surface of mitochondria interacts with adapter protein Milton/ TRAK1, which binds both the kinesin-1 heavy chain and dynactin for bidirectional transport in axons. Guanosine triphosphatases (GTPases) on organelle membranes interact with other adapter proteins. For instance, different GTPases target kinesin and dynein to lysosomes. Lysosomes with Arl8-GTP (guanosine triphosphate) bind the adapter SKIP (SifA-kinesin interacting protein). An acidic/tryptophan sequence in SKIP interacts with kinesin light chains (Fig. 37.2C) during transport to the periphery. This interaction with SKIP also turns on the kinesin motor by overcoming the autoinhibitory interaction of the motor domains with the tail. Similarly, vesicles containing transmembrane receptors (mannose-6-phosphate receptor, for example) bud from late endosomes with the GTPase Rab7 on their surface. Rab7-GTP binds the adapter protein RILP (Rab7-interacting lysosomal protein), which anchors dynactin and dynein for vesicle transport to the transGolgi network (see Fig. 22.12). Phosphatidylinositol 4-phosphate (see Fig. 26.7) on the Golgi membranes binds sorting nexin 6 (part of the retromer complex; see Fig. 22.12) and releases dynactin and dynein. Many cargo particles associate with both a kinesin and dynein. In this case, regulatory mechanisms responsive to local conditions must coordinate their activities to achieve net transport. One example is the adapter protein La, which links ribonucleoprotein particles to both dynein and kinesin. After being carried to the cell periphery by kinesin, covalent modification of La with a SUMO protein inactivates kinesin, allowing for reverse transport back to the cell center by dynein. Phosphorylation regulates the adapter protein JIP1 (JNK interacting protein-1), which binds either kinesin or dynein. Phosphorylated JIP1 binds and activates kinesin, favoring anterograde transport. Mitochondria move both directions in axons, 0 sec 2
643
but some stop moving in regions with limited adenosine triphosphate (ATP). A local high concentration of cytoplasmic calcium binds to EF-hands (Ca2+ binding site in calmodulin consisting of α-helices E and F) of the anchor protein Miro. Calcium binding turns off both kinesin and dynein, keeping mitochondria in place until ATP concentrations return to normal. Another protein anchors stationary mitochondria to microtubules. Faulty axonal transport resulting in clumps of vesicles is observed in human degenerative diseases of the nervous system including Alzheimer and Parkinson diseases, but cause-and-effect relationships are still under investigation. One intriguing part of this story is that the transmembrane amyloid precursor protein not only binds directly to kinesin-1 light chain, but also is cleaved in Alzheimer disease to produce amyloid-β peptide—a toxic peptide that is implicated in neuronal death. Huntingtin, the giant (350 kD) adapter protein mutated in Huntington disease, normally participates in bidirectional axonal transport of multiple cargos (Table 37.2). Other membrane-associated scaffold proteins, including, JIP-1 and JIP-3, bind kinases from the mitogen-activated protein (MAP) kinase cascade (see Fig. 27.6) in addition to kinesin-1 light chains.
Fast Axonal Transport Analysis of microtubule-based movements is particularly favorable in axons of neurons, because axons are long (up to 1 m) but narrow, the microtubules have a uniform polarity, and organelles move at steady rates in both directions. Furthermore, nerve cells contain high concentrations of microtubules and microtubule motors; indeed, cytoplasmic tubulin, cytoplasmic dynein, and kinesin were all originally isolated from brain. High-contrast light microscopy of living axons reveals that most membrane-bound organelles move either toward (anterograde) or away from (retrograde) the end of the axon (Fig. 37.3) with some pauses and even 3 sec
5 sec 1
2
2 1
1
4
A
3
4
3
3
4
B
FIGURE 37.3 FAST TRANSPORT IN CYTOPLASM ISOLATED FROM SQUID GIANT AXONS. A, Three frames from a series of videoenhanced differential interference contrast micrographs show movement of organelles in both the anterograde (rightwards) and retrograde (leftwards) directions. Four large organelles are marked with numbers and colored green at zero time, blue at 3 seconds, and red at 5 seconds. Movement (arrows in right panel) is from the white to the black number. The original video record shows hundreds of smaller organelles moving steadily in either an anterograde or a retrograde direction at 1 to 2 µm/s. B, Electron micrograph of a thin section showing vesicles associated with microtubules in axoplasm. (A, Courtesy S. Brady, University of Texas Southwestern Medical School, Dallas. B, Courtesy R.H. Miller, Case Western Reserve Medical School, Cleveland, OH.)
644
SECTION IX n Cytoskeleton and Cellular Motility
A
B Proximal
Ligature
Distal
FIGURE 37.4 ELECTRON MICROGRAPHS SHOWING THE RESULT OF NERVE LIGATION. These are longitudinal sections of axons surrounded by a darkly stained myelin sheath. A, The cytoplasm proximal to the ligation demonstrates the accumulation of vesicles and mitochondria, which were being transported toward the nerve terminal to the right. B, The cytoplasm distal to the ligation shows the accumulation of lysosomes, multivesicular bodies, and mitochondria, which were being transported toward the cell body to the left. (B, From Hirokawa N, Sato-Yoshitake R, Yoshida T, et al. Brain dynein (MAP1C) localizes on both anterogradely and retrogradely transported membranous organelles in vivo. J Cell Biol. 1990;111:1027–1037.)
occasional changes of direction. Retrograde movements (2.5 µm s−1 or 22 cm/day) are faster than anterograde movements (0.5 µm s−1 or 4 cm/day). These rates are remarkable given the densely packed microtubules, intermediate filaments, and vesicles in the cytoplasm (Fig. 37.4; see also Fig. 35.9). At these rates, a round trip from a nerve cell body in the spinal cord of a human to the foot and back takes only 3 weeks. If a 0.1-µm vesicle were the size of a small car, it would move anterogradely at 50 miles per hour and retrogradely at 250 miles per hour. In the axons of vertebrate neurons, mitochondria move back and forth in both directions. Their net movement toward the nerve terminal or cell body depends on physiological conditions. Biochemical reconstitution showed that the plus-end motors of the kinesin family are responsible for anterograde movements toward the nerve terminal and the minus-end motor dynein is responsible for movement in the retrograde direction. Accordingly, kinesin mutations in flies result in paralysis of the back half of larvae, because transport fails in the longest axons. Mutations in three different kinesin genes also cause human nerve degeneration. Classic nerve ligation experiments revealed the cargo carried in each direction by fast transport (Fig. 37.4). Different organelles pile up on either side of a mechanical constriction that blocks transport. Small, round, and tubular vesicles, including components of synaptic vesicles (see Fig. 17.8), accumulate on the side near the cell body. Fast anterograde transport carries these cargoes from the cell body toward the end of the axon, where they enter the cycle of synaptic vesicle turnover (see
Figs. 17.8 and 17.9). Endosomes and multivesicular bodies moving by fast retrograde transport pile up on the distal side of the constriction. Autophagic vesicles also move primarily in the retrograde direction. Retrograde transport can also move signals from nerve terminals to the cell body. For example, kinesin carries vesicles with the nerve growth factor TrkA receptor tyrosine kinase (see Fig. 24.4) to nerve terminals where it joins the plasma membrane. When activated by nerve growth factor, TrkA is taken up by endocytosis for transport by dynein back to the perinuclear region. Once the vesicle reaches the cell body TrkA activates the MAP kinase pathway to regulate cell growth (see Fig. 27.6). Herpes virus and rabies virus also use dynein for long-distance transport on microtubules from the terminals of sensory nerves to the cell body, where viral DNA enters the nucleus for replication. Many questions remain regarding the control of microtubule motors during fast transport, including how kinesins remain active and dynein remains inactive during the long trip out to the nerve terminal, how cytoplasmic dynein on retrograde cargo is activated locally at the nerve terminal and kept active during movement to the cell body, and how defects in fast transport may contribute to neurodegenerative diseases.
Slow Transport of Cytoskeletal Polymers and Associated Proteins in Axons Many neuronal proteins move slowly from their site of synthesis in the cell body toward the ends of axons and dendrites. This transport is essential, as most protein
CHAPTER 37 n Intracellular Motility
A. Radiolabeling pulse-chase protocol
B. Photobleaching
Inject radiolabeled amino acids into eye where neurons synthesize them into proteins
Inject fluorescently labeled tubulin
645
C. Fluorescence microscopy of a neurofilament moving through a bleached zone in an axon Anterograde 0 Bleached zone 56
Photobleach discrete zone Wave of labeled proteins moves along axons toward brain
64 Time (sec)
Uniformly fluorescent axon
72
80 Bleached zone 88
Wave of labeled proteins progresses 1-2 mm/day
96 Zone remains stationary
FIGURE 37.5 EXPERIMENTS ON SLOW AXONAL TRANSPORT. A, Pulse-chase experiment. Radioactive amino acids are injected into the eye of an experimental animal. In the nerve cell body, radioactive tracer is incorporated into proteins, which are transported along the axon. Some proteins are incorporated into stationary structures and are left along the way. B, Photobleaching experiment. A cultured nerve cell is injected with tubulin labeled with a fluorescent dye. Tubulin fills the cytoplasm and axon as it grows out. A section of the axon is then bleached with a strong pulse of light. This bleached zone is stationary over a period of minutes. C, Fluorescence micrographs of the axon of a cultured rat neuron showing rapid transport of a short neurofilament labeled with subunits fused to green fluorescence protein (GFP). Note the photobleached region (bracket) and the ends of the moving neurofilament (arrows). The neurofilament moves rapidly into the bleached region, but the bleached region does not move because most of the neurofilaments are stationary. Scale bar is 5 µm. (A–B, Modified from Cleveland DW, Hoffman PN. Slow axonal transport models come full circle. Cell. 1991;67:453–456. C, From Wang L, Brown A. Rapid intermittent movement of axonal neurofilaments observed by fluorescence photobleaching. Mol Biol Cell. 2001;12:3257–3267, 2001.)
synthesis occurs in the cell body, whereas more than 99% of cell volume can be in axons and dendrites. If nerve cells were smaller, shorter lived or less asymmetrical, we might not even notice such slow movements. Labeling proteins with radioactive amino acids during their synthesis in the cell body allows for tracking their movements along axons (Fig. 37.5A). Proteins that are moved by slow axonal transport are classified into two groups based on their velocities. Tubulin, intermediate filament proteins, and spectrin, which compose the “slow component–a,” move exceedingly slowly, approximately 0.1 to 1.0 mm per day (or 1 to 10 nm s−1). In a human, these molecules take more than 3 months to travel from their site of synthesis in the spinal cord to the foot. “Slow component–b” moves approximately 10 times faster and includes 10 times more protein than slow component–a. It is a heterogeneous mixture of
many proteins, including clathrin, glycolytic enzymes, and actin. Defining the mechanism of slow transport was challenging because various experimental approaches yielded apparently conflicting results. Radioactive labeling showed that the moving proteins are spread out and diluted as they move away from the cell body (Fig. 37.5A). Photobleaching of fluorescent tubulin and actin in axons of cultured neurons demonstrated that the bulk of those cytoskeletal polymers is stationary (Fig. 37.5B). This puzzle was solved for slow component–a by imaging single fluorescent intermediate filaments in axons of live nerve cells. These filaments are stationary most of the time (up to 99%), but occasionally, they move rapidly (0.2 to 2 µm s−1) for up to 20 µm. Most of these intermittent movements are away from the cell
646
SECTION IX n Cytoskeleton and Cellular Motility
body, accounting for the net anterograde movement. These rapid but intermittent movements depend on microtubules and are presumably driven by kinesin, although how the motor is linked to the filaments is not known. A few intermediate filaments move in the retrograde direction explaining why waves spread in pulse– chase experiments (Fig. 37.5A). The movements of whole microtubules are similar to those of intermediate filaments. Thus, what appears to be slow bulk transport is caused by fast but intermittent movement of cargos. Slow component–b proteins seem move by alternatively hitchhiking transiently on some moving structures and diffusing locally between movements.
A
C
00:00
00:00
00:35
01:00
00:42
02:00
00:49
02:10
02:06
02:49
03:02
03:30
03:30
04:00
03:58
04:10
04:40
04:40
00:00
Other Microtubule-Dependent Movements Other cells use the same molecular mechanisms as neurons to move organelles in the cytoplasm. Secretory vesicles use plus–end motors to move from the Golgi apparatus to the plasma membrane (see Fig. 21.1). Endosomes use dynein to move from the plasma membrane toward the centrosome. The intracellular distributions of the endoplasmic reticulum (ER) depend on microtubules. Strands of the ER align with microtubules in cultured cells (Fig. 37.6). This codistribution is achieved in two ways: (a) Motors transport strands of ER bidirectionally on microtubules, and (b) other strands of the ER attach to the plus end of microtubules and ride the microtubule tip as it grows and shrinks during dynamic instability. The +TIP EB1 (see Fig. 34.8) interacts with transmembrane protein STIM1 to couple the ER to growing microtubules. STIM1 has another role in maintaining a supply of Ca2+ inside the ER (see Fig. 26.12). This is the best example of movement of an organelle driven by forces generated directly by microtubule assembly. Microtubules and motors also distribute other organelles inside cells. The concentration of the Golgi apparatus near the centrosome depends on microtubules, because dynein motors transport Golgi vesicles toward the minus ends of the microtubules nucleated at centrosomes (see Fig. 21.20). Mitochondria move bidirectionally on microtubules in animal cells but depend on actin filaments in yeast. Thus, not only the dynamics of the organelles but also the overall organization of a cell depend on microtubules and associated motors. As a result, cellular architecture is determined actively, not passively. The intracellular distribution of nucleoprotein complexes also depends on active movements. The most obvious example is the movement of chromosomes during mitosis, which depends on microtubule assembly and microtubule motors (see Fig. 44.7). Microtubules and motors can also produce asymmetrical distributions of mRNAs in cells. The adapter protein La links kinesins and dynein to specific ribonucleoprotein particles for
B
06:24
12:48
19:15
FIGURE 37.6 TWO MODES OF MICROTUBULE-DEPENDENT MOVEMENT OF THE ENDOPLASMIC RETICULUM IN A NEWT EPITHELIAL CELL. The cell was microinjected with rhodaminelabeled tubulin, which incorporates into microtubules, and a lipophilic green-fluorescent dye (DiOC6 [3,3′-dihexyloxacarbocyanine iodide]), which labels endoplasmic reticulum (ER). Time series are indicated in minutes and seconds. Scale bar for all panels is 5 µM. A, This column of fluorescence micrographs illustrates the dynamics of microtubules (red) and ER (green), over a period of 19 minutes. Note the strand of ER moving away from the leading edge (arrowheads). B, Time course of the movement of a strand of ER toward the end of a microtubule, followed by retraction. This type of movement is thought to be driven by a kinesin motor attached to the tip of the elongating membrane (arrowhead). C, Time course of the movement of a strand of ER attached to the tip of a growing microtubule (arrowhead), followed by retraction of the membrane along the microtubule. (Courtesy C. Waterman-Storer and E.D. Salmon, University of North Carolina, Chapel Hill. For reference, see Waterman-Storer C, Salmon ED. Endoplasmic reticulum membrane tubules. Curr Biol. 1998;8:798–806.)
CHAPTER 37 n Intracellular Motility
(–)
647
Nucleus RNA granules 0
42
51
63
75
FIGURE 37.8 TRANSPORT OF AN ISOLATED CHROMOSOME ON A SHORTENING MICROTUBULE IN VITRO. A microtubule was grown from brain tubulin nucleated by a basal body in the extracted carcass of a ciliate, Tetrahymena. A chromosome (arrow) was added as a test cargo and captured by the end of the microtubule. When the concentration of tubulin was reduced, the microtubule shortened, carrying along the chromosome attached to its tip. This transport occurs in the absence of adenosine triphosphate (ATP) or guanosine triphosphate (GTP). (Courtesy J.R. McIntosh, University of Colorado, Boulder.)
Disperse RNA
Branch
Membrane Immobile granules
(+)
Mobile granules Disperse RNA FIGURE 37.7 TRANSPORT OF MESSENGER RNA (mRNA) FOR MYELIN BASIC PROTEIN IN A CULTURED OLIGODENDROCYTE, A GLIAL CELL ISOLATED FROM BRAIN. mRNA synthesized in the cell body (or, in this case, labeled with a fluorescent dye and microinjected into the cell body) is packaged with proteins in a ribonucleoprotein particle, transported from the cell center along microtubules at a steady rate of 0.2 µm s−1, and released at the periphery, where it moves randomly at 1 µm/s. (Modified from Ainger K, Avossa D, Morgan F, et al. Transport and localization of exogenous myelin basic protein mRNA microinjected into oligodendrocytes. J Cell Biol. 1993;123:431–441.)
long distance transport to sites of local protein synthesis (Fig. 37.7). For example, localization of certain mRNAs in fly oocytes helps establish the polarity of the embryos.
Intracellular Movements Driven by Microtubule Polymerization Microtubule polymerization and depolymerization have long been known to play a central role in the assembly of the mitotic apparatus and the movement of chromosomes (see Fig. 44.7), as well as the establishment of cellular asymmetry (see Fig. 34.2). Polymerizing microtubules can exert substantial forces, but strong forces buckle microtubules longer than approximately 10 µm. Consequently, microtubule pushing mechanisms work best over short distances, such as positioning the ER (Fig. 37.6), the nucleus in fission yeast cells (see Fig. 6.3D) and the mitotic spindle in budding yeast cells (see Fig. 34.2D). Remarkably, the end of a depolymerizing microtubule can also pull on attached cargo. An in vitro proofof-principle experiment (Fig. 37.8) showed that chromosomes could ride on the end of a depolymerizing microtubule, even in the absence of ATP or GTP. Linker proteins associated with the kinetochore region of the
chromosome make multiple weak bonds with the walls of the microtubule end (see Figs. 8.21 and 8.22). These interactions rearrange rapidly enough to maintain attachment, even as tubulin subunits dissociate from the end.
Bulk Movement of Cytoplasm Driven by Actin and Myosin Bulk streaming of cytoplasm is most spectacular in plant cells (Fig. 37.1B). Although confined within rigid walls, plant cell cytoplasm streams vigorously at very high velocities (up to 60 µm s−1). At this rate, cytoplasm moves 5 m/day! Such cytoplasmic streaming is best understood in the giant cells of the green alga Nitella. Streaming occurs continuously in a thin layer of cytoplasm between the large central vacuole and chloroplasts immobilized in the cortex. On each side of the cell, a zone of stationary cytoplasm separates streams moving in opposite directions. The physiological function of this streaming is not clearly understood. Bulk streaming in Nitella is brought about by movement of ER along tracks consisting of bundles of polarized actin filaments associated with chloroplasts (Fig. 37.9C). All the actin filaments in these bundles have the same polarity, and cytoplasm streams toward their barbed ends. In Nitella extracts, membrane vesicles move along actin filament bundles at the same high velocities as the cytoplasmic streaming. A type XI myosin (see Fig. 36.7) pulls the ER along these cortical actin tracks, dragging along other cytoplasmic components, including organelles and soluble molecules. This myosin moves nearly 10 times faster than the fastest muscle contraction, apparently by taking large steps rapidly and by the cooperation of several motors working rapidly on the same membrane. A completely different actomyosin mechanism produces equally spectacular cytoplasmic streaming in the acellular slime mold Physarum. In these giant, multinucleated cells, cytoplasm flows back and forth rhythmically at high velocities through tubular channels (Fig. 37.10). Cycles of contraction and relaxation of
648
SECTION IX n Cytoskeleton and Cellular Motility
A
B
D
C
E
F
FIGURE 37.9 CYTOPLASMIC STREAMING IN THE GREEN ALGA NITELLA. A, A pair of differential interference contrast (DIC) light micrographs showing the movement of organelles in cytoplasm. Note the strand of endoplasmic reticulum (ER [arrow]). B, Time series of DIC light micrographs showing movement of a vesicle isolated from Nitella along a bundle of actin filaments isolated from Nitella. C, Scanning electron micrographs of the cortex isolated from Nitella showing the bundles of actin filaments associated with chloroplasts. D, Transmission electron micrographs of a freeze-fracture preparation (upper) and thin section (lower) showing ER associated with actin filament bundles. E, Freeze-fracture preparation of a vesicle associated with an actin filament bundle. F, Movement of ER along actin filament bundles dragging along bulk cytoplasm. (Courtesy B. Kachar, National Institutes of Health. For reference, see Kachar B, Reese T. The mechanism of cytoplasmic streaming in characean algal cells: sliding of endoplasmic reticulum along actin filaments. J Cell Biol. 1988;106:1545–1552.)
cortical actin filament networks push relatively fluid endoplasm back and forth in a manner akin to squeezing a toothpaste tube. Myosin-II is thought to generate the cortical contraction, as it is present in high concentration in this cell and can contract actin filament gels in vitro. (This was the first nonmuscle myosin to be purified in the late 1960s.) Cortical contractions similar to those of Physarum are used by giant amoebas for cell locomotion (see Fig. 38.1), and by other cells for cytokinesis (see Fig. 44.23), and movements of some embryonic tissues (see Fig. 38.5).
Actin-Based Movements of Organelles in Other Cells Like Nitella, budding yeast cells transport vesicles along bundles of actin filaments from the mother to the bud
(Fig. 37.11), although the movements of these solitary vesicles do not produce cytoplasmic streaming. Myosin-V is the motor (see Fig. 36.7), so vesicles fail to move from mother to bud in null mutants of myosin-V genes. Myosin-V also transports certain mRNAs along actin filament cables from the mother to the bud, where they determine cell fate. Similarly myosin-VII and myosin-X transport cytoplasmic and membrane proteins in filopodia of animal cells. Animal cells generally use extended microtubules for long-distance movements and shorter actin filaments for local transport of vesicles and RNAs. Fish skin cells use both systems to change color: dynein aggregates and kinesin disperses pigment granules called melanophores along radial microtubule tracks. Myosin-V contributes by moving dispersed melanophores laterally between microtubules.
CHAPTER 37 n Intracellular Motility
649
M
A
B
C
D
M 5 cm Balance-pressure (cm of water)
A
B
24
FIGURE 37.11 Fluorescence micrographs (A–D) showing actin filament bundles and patches at various stages in the cell cycle of the budding yeast Saccharomyces cerevisiae. Myosin-V uses these actin filament bundles to deliver vesicles (including the vacuole), certain messenger RNAs (mRNAs), and at least one enzyme (chitin synthase) from the mother to the bud. (Courtesy J.A. Cooper, Washington University, St. Louis, MO.)
20 16 12 8 4 0 -4 -8 -12
5
10
Time (min)
C FIGURE 37.10 CYTOPLASMIC STREAMING IN THE ACELLULAR SLIME MOLD PHYSARUM POLYCEPHALUM. A, Photograph of Physarum polycephalum, a giant multinucleated single cell growing in a baking dish. B, Blur photomicrograph made with polarization optics by taking a time exposure showing the bulk streaming of the endoplasm in a cytoplasmic strand (long arrow). M, Mucus. C, Time course of pressure changes produced by shuttle streaming of cytoplasm through a strand. (B, From Nakajima H. The mechanochemical system behind streaming in Physarum. In Allen RD, Kamiya N, eds. Primitive Motile Systems in Cell Biology. New York: Academic Press; 1964:111–123. C, For reference, see Kamiya N. The mechanism of cytoplasmic movement in a myxomycete plasmodium. Symp Soc Exp Biol. 1968;22:199–214.)
The direction of movement depends on the motor and the organization of the actin filaments. Although many actin filaments in animal cells are not uniformly polarized, those in the cortex tend to have their barbed ends near the plasma membrane, allowing for local directional transport. For example, myosin-V moves recycling endosomes toward the plasma membrane. Similarly, myosin-V transports pigment granules called melanosomes within and between cells in the skin. In both cases a small GTPase and an adapter protein link melanosomes to the tail of myosin-V. Mutations of myosin-V, the GTPase, or the adapter protein in mice and humans cause not only pigmentation defects but also neurological problems owning to faulty mRNA transport in neurons. Other adapter proteins link myosin-VI to newly internalized endosomes with various types of receptors for transport away from the plasma membrane.
Movements Driven by Actin Polymerization Some intracellular pathogenic bacteria, including Listeria and Shigella, use actin polymerization to move through the cytoplasm of their animal cell hosts at about 0.5 µm s−1 (Fig. 37.12A). These bacteria hijack the machinery normally used to move the leading edge of motile cells to polymerize a comet tail of actin filaments that pushes the bacterium forward. One end of the bacterium has a concentration of proteins that directly (Listeria) or indirectly (Shigella) activate Arp2/3 complex to polymerize a network of branched actin filaments (see Fig. 33.12). Growth of this network at its trailing end pushes the bacterial cell forward. The comet tail of crosslinked actin filaments is stationary and depolymerizes distally at the same rate at which it grows next to the bacterium, so it remains a constant length. Vaccinia viruses attached to the outer surface of animal cells move by a related mechanism. They use transmembrane proteins to activate the cytoplasmic actin assembly system to drive their movements at one stage in its life cycle (Fig. 37.12B). Placement of a plastic bead coated with adhesion proteins on the plasma membrane of some animal cells can induce similar propulsive actin comet tails in the cytoplasm. Fungal and animal cells use Arp2/3 complex to assemble actin filaments at sites of clathrin-mediated endocytosis. Polymerization of these filaments assists with vesicle separation from the plasma membrane. Nucleation promoting factors associated with some endosomes stimulate Arp2/3 complex to assemble actin filament comets and move similar to Listeria.
650
SECTION IX n Cytoskeleton and Cellular Motility
A. Listeria
B. Vaccinia virus
5 µm
FIGURE 37.12 Fluorescence micrographs of actin filament comet tails in animal epithelial cells infected with the bacterium Listeria monocytogenes (A) or vaccinia virus (B). Both pathogens are stained green with fluorescent antibodies. They use host cell proteins to assemble a crosslinked network of actin filaments shaped like a comet tail. Actin filaments are stained red with rhodamine-phalloidin. A, The comet tail pushes Listeria in a PtK cell through the cytoplasm and into projections of the plasma membrane at the edge of the cell. B, When the replicated vaccinia viruses in this HeLa cell reach the cell surface 8 hours after infection, they activate Arp2/3 complex to assemble a cytoplasmic comet tail of actin filaments that are thought to enhance the spread of the virus from cell to cell. Actin-based motility of vaccinia virus depends on tyrosine phosphorylation of a viral transmembrane protein A36R that remains inserted in the plasma membrane. (A, Courtesy K. Skoble, D. Portnoy, and M. Welch, University of California, Berkeley. B, Courtesy T.P. Newsome and M. Way, Cancer Research UK, London. For reference, see Frischknecht F, Moreau V, Rottger S, et al. Actin-based motility of vaccinia virus mimics receptor tyrosine kinase signaling. Nature. 1999;401:926–929.)
ACKNOWLEDGMENTS We thank Anthony Brown, Erika Holzbaur, and Margaret Titus for suggestions on revising this chapter. SELECTED READINGS Brown A. Slow axonal transport. Encyclopedia of Neuroscience, 2009. Update. October 28, 2014. Reference Module in Biomedical Sciences. Available at , 2014. Bullock SL. Messengers, motors and mysteries: sorting of eukaryotic mRNAs by cytoskeletal transport. Biochem Soc Trans. 2011;39: 1161-1165. Carter AP, Diamant AG, Urnavicius L. How dynein and dynactin transport cargos: a structural perspective. Curr Opin Struct Biol. 2016; 37:62-70. Cossart P, Pizarro-Cerdá J, Lecuit M. Invasion of mammalian cells by Listeria monocytogenes: Functional mimicry to subvert cellular functions. Trends Cell Biol. 2003;13:23-31. Encalada SE, Goldstein LS. Biophysical challenges to axonal transport: motor-cargo deficiencies and neurodegeneration. Annu Rev Biophys. 2014;43:141-169. Frank DJ, Noguchi T, Miller KG. Myosin VI: A structural role in actin organization important for protein and organelle localization and trafficking. Curr Opin Cell Biol. 2004;16:189-194. Franker MA, Hoogenraad CC. Microtubule-based transport-basic mechanisms, traffic rules and role in neurological pathogenesis. J Cell Sci. 2013;126:2319-2329.
Fu MM, Holzbaur EL. Integrated regulation of motor-driven organelle transport by scaffolding proteins. Trends Cell Biol. 2014;24: 564-574. Hammer JA 3rd, Sellers JR. Walking to work: roles for class V myosins as cargo transporters. Nat Rev Mol Cell Biol. 2011;13:13-26. Hancock WO. Bidirectional cargo transport: moving beyond tug of war. Nat Rev Mol Cell Biol. 2014;15:615-628. Hartman MA, Finan D, Sivaramakrishnan S, et al. Principles of unconventional myosin function and targeting. Annu Rev Cell Dev Biol. 2011;27:133-155. Hirokawa N, Niwa S, Tanaka Y. Molecular motors in neurons: transport mechanisms and roles in brain function, development, and disease. Neuron. 2010;68:610-638. Lopez de Heredia M, Jansen R-P. mRNA localization and the cytoskeleton. Curr Opin Cell Biol. 2004;16:80-85. Mooren OL, Galletta BJ, Cooper JA. Roles for actin assembly in endocytosis. Annu Rev Biochem. 2012;81:661-686. Pruyne D, Legesse-Miller A, Gao L, et al. Mechanisms of polarized growth and organelle segregation in yeast. Annu Rev Cell Dev Biol. 2004;20:559-591. Saxton WM, Hollenbeck PJ. The axonal transport of mitochondria. J Cell Sci. 2012;125:2095-2104. Shimmen T, Yokota E. Cytoplasmic streaming in plants. Curr Opin Cell Biol. 2004;16:68-72. Titus MA. Myosin-driven intracellular transport. In: Pollard TD, Goldman R, eds. The Cytoskeleton. Cold Spring Harbor Press; 2016;265-280. Urnavicius L, Zhang K, Diamant AG, et al. The structure of the dynactin complex and its interaction with dynein. Science. 2015;347: 1441-1446.
CHAPTER
38
Cellular Motility
C
ells move at rates that range over four orders of magnitude (Fig. 38.1 and Table 38.1). At one extreme, ciliates, bacteria, and sperm swim rapidly through water, and giant amoebas crawl rapidly over solid substrates. At the other extreme, fungal, algal, and plant cells with rigid cell walls are immobile. However, even some plant cells move, such as pollen, which extends
E. coli
Velocity (µm/sec) 10-2 10-1 100 101 102 103
Sperm Tetrahymena
Amoeba proteus
White blood cell
Pollen tube Fibroblast
Nerve
FIGURE 38.1 VELOCITIES OF MOVING CELLS SPAN MORE THAN FOUR ORDERS OF MAGNITUDE. Scale drawings of cells with a range of velocities. E. coli, Escherichia coli.
tubular pseudopods. Most cells, including white blood cells, nerve growth cones, and fibroblasts move at intermediate rates. Cells produce forces for motility in many different ways, most commonly using the same four mechanisms that produce intracellular movements (see Chapter 37): contraction of actin–myosin networks, movement of motors on microtubules, reversible assembly of actin filaments, or reversible assembly of microtubules. These mechanisms often complement each other, even where movement depends mainly on one system. For example, microtubules contribute to actin-based pseudopod extension by helping to specify the polarity of the cell. This chapter compares these common mechanisms with a few more unusual, but informative mechanisms: contraction of calcium-sensitive fibers of ciliates, reversible assembly of novel cytoskeletal polymers of nematode sperm, and rotation of bacterial flagellar motors. Most cells possess all proteins required for cellular motility, so the striking variation in their rates of movement arises from differences in the abundance, organization, and activities of this machinery. For example, both nonmotile yeasts and contractile muscle cells contain actin, myosin-II, heterodimeric capping protein, α-actinin, and tropomyosin. Yeasts use these proteins for cytokinesis (see Fig. 44.25), while muscle assembles high concentrations of similar proteins into sarcomeres (see Figs. 39.2 and 39.3) for powerful, fast contractions.
Cell Shape Changes Produced by Extension of Surface Processes Alteration of cellular shape can be most simply brought about by assembly of new cytoskeletal polymers or by rearrangement of preexisting assemblies of actin filaments or microtubules. 651
652
SECTION IX n Cytoskeleton and Cellular Motility
TABLE 38.1 Velocities of Cellular Movements System
Unitary Velocity (µm s−1)
Summed Velocity (µm s−1)
Motile Mechanism
Striated muscle contraction (biceps)
5–10
4–8 × 105
Actin-myosin adenosine triphosphatase
Filopodium extension, Thyone sperm
10
10
Actin polymerization
Pseudopod extension, fibroblast
0.02
0.02
Actin polymerization
Pseudopod extension, human neutrophil
0.1
0.1
Actin polymerization
Pseudopod extension, Amoeba proteus
?
10
Actin-myosin ATPase
Pseudopod extension, nematode sperm
1
1
Assembly of major sperm protein
Retraction of axopodium, heliozoan
>100
>100
Disassembly of microtubules
Spasmoneme contraction, Vorticella
?
23,000
Calcium-induced conformational change
Swimming, Escherichia coli
25
Flagellum powered by rotary motor
Swimming, sea urchin sperm
15
Microtubule-dynein ATPase
Note: Unitary velocity refers to a single molecular unit. Summed velocity is the overall motion of the cell.
EGG JELLY
A Egg stimulates Actin polymerization secretion of extends the acrosomal contents acrosomal process
Filopodia
B
Bundle of actin filaments
Packet of actin and profilin Nucleus Axoneme FIGURE 38.2 THYONE SPERM ACROSOMAL PROCESS. Actin polymerization drives the growth of the acrosomal process of the sperm of the sea slug, Thyone. The acrosome (red) is a membranebound secretory vesicle, which fuses with the plasma membrane and releases its hydrolytic enzymes prior to growth of the acrosomal process. When the acrosomal process reaches the egg, the plasma membranes of the two cells fuse. (Based on the work of L. Tilney, University of Pennsylvania, Philadelphia.)
Studies of echinoderm sperm revealed that actin polymerization drives the formation of cell surface projections called filopodia. To fertilize the egg the sperm extends a long filopodium to penetrate the protective jelly surrounding the egg (Fig. 38.2). Actin subunits for this acrosomal process are stored with profilin (see Figs. 1.4 and 33.11) in a novel concentrated packet near the nucleus. Contact with an egg stimulates actin filaments to polymerize, starting from a dense structure near the nucleus. Addition of subunits to the distal (barbed) end of growing filaments drives the elongation of the process and the surrounding membrane at a rate of 5 to 10 µm s−1 (an astounding maximum of 3700 subunits per second!). Actin subunits diffuse rapidly enough from their storage site to drive this rapid elongation. The number of filaments declines from approximately 150
Ruffles
10 µm FIGURE 38.3 FILOPODIA. A, Fluorescence micrograph of the edge of a mouse NIH 3T3 cell expressing formin mDia2 and activated Rif, a Rho-family guanosine triphosphatase (GTPase) that activates mDia2. mDia2 concentrated at the tips of filopodia is stained red with fluorescent antibodies. B, Scanning electron micrograph of mouse macrophages spreading on a glass slide, illustrating the flat peripheral lamellae, wave-like “ruffles” on the upper surface, and finger-like filopodia. (A, From Pellegrin S, Mellor H. The Rho family GTPase Rif induces filopodia through mDia2. Curr Biol. 2005;15:129–133. B, From Chitu V, Pixley FJ, Macaluso F, et al. The PCH family member MAYP/PSTPIP2 directly regulates F-actin bundling and enhances filopodia formation and motility in macrophages. Mol Biol Cell. 2005;16: 2947–2959.)
near the base to less than 20 at the tip of the process, presumably from capping. The molecules (if any) guiding the growth of the filaments are not known. Bundles of actin filaments support filopodia of a more modest size on many other cells. Filopodia on macrophages (Fig. 38.3), nerve growth cones (Fig. 38.6A), fibroblasts, and epithelial cells grow much more slowly and depend on formins and vasodilator-stimulated protein (VASP) at their tips (Fig. 38.3) to guide barbedend assembly of actin filaments crosslinked by a protein called fascin. Microvilli of the brush border of epithelial cells (see Fig. 33.2) are short, stable filopodia. The
CHAPTER 38 n Cellular Motility
A
B
653
C
FIGURE 38.4 DYNAMIC CELL SURFACE PROJECTIONS SUPPORTED BY MICROTUBULES. A–B, Drawings of the radiolarian Echinospherium (a protozoan) showing projections called axopodia, which capture prey and draw them toward the cell body by transport on surface of the projection or collapse of the projection. C, Electron micrograph of a thin section across an axopodium, showing the double spiral array of microtubules. (Courtesy L. Tilney, University of Pennsylvania, Philadelphia.)
A. Epithelial folding
Apical contraction folds the epithelium
B. Neural tube formation
Apex
Basal lamina
Base
C. Drosophila ectoderm contraction
FIGURE 38.5 ACTOMYOSIN CONTRACTIONS MOLD THE SHAPE OF EPITHELIA DURING EMBRYONIC DEVELOPMENT. A, Folding of a planar epithelium into a tube. B, Formation of the neural tube by contraction of the apical pole of columnar epithelial cells resulting in shape change and invagination of the epithelium. C, Contraction around the margin of the ectoderm pulls this epithelium over the surface of a Drosophila embryo. The scanning electron micrographs (SEMs [left and right]) show the steps in the dorsal closure of the epithelium. The time series of fluorescence micrographs (center) shows live embryos expressing an actin-binding fragment of the protein moesin, fused to green fluorescent protein. (C, SEMs courtesy Thom Kaufman, Indiana University, Bloomington [see his movie “Fly Morph-o-genesis” at http://www.sdbonline.org/archive/ dbcinema/kaufman/kaufman.html]; light micrographs courtesy D. Kiehart, Duke University, Durham, NC. For reference, see Kiehart DP, Galbraith CG, Edwards KA, et al. Multiple forces contribute to cell sheet morphogenesis for dorsal closure in Drosophila. J Cell Biol. 2000;149: 471–490.)
bundles of actin filaments supporting microvilli are crosslinked to each other by fimbrin and villin and to the plasma membrane by myosin-I. Synthesis of these accessory proteins triggers assembly of microvilli on epithelial cells as well as cells that normally have few microvilli.
Some embryonic cells grow long filopodia to contact cells micrometers distant. These filopodia create physical contacts that allow the cells to send or receive signals that influence developmental decisions. Long filopodia between some pairs of cells even make gap junctions (see Fig. 31.6).
654
SECTION IX n Cytoskeleton and Cellular Motility
Lamellum
Filopodia
Axo n
A. Growth cone
B
Leading edge
Keratocyte
C
Leading edge
Cytoplast
Continuous gliding
Protrusion & adhesion
Tail retraction
Repeated cycles
FIGURE 38.6 MOTILITY BY PSEUDOPOD EXTENSION. A, Phase-contrast micrographs of a cultured nerve cell’s growth cone at 1-minute intervals. The growth cone extends filopodia and fills in the space between with an actin-filled lamella. B, Gliding movements of a fish epidermal keratocyte and a keratocyte cytoplast, a cell fragment consisting of the leading edge with most of the cell body including the nucleus removed. Differential interference contrast micrographs at 15-second intervals were superimposed. Drawing shows the pattern of movement. C, Phasecontrast micrograph of a keratocyte on glass. This cell moved toward the upper right using cycles of expansion of the broad leading lamella and retraction of the trailing edge from the surface as shown by the drawings. (A, Courtesy D. Bray, University of Cambridge, United Kingdom. B, From Pollard TD, Borisy GG. Cellular motility driven by assembly and disassembly of actin filaments. Cell. 2003;112:453–465. C, Courtesy J. Lee, University of Connecticut, Storrs.)
A group of protists called heliozoans, named for their similarity to a cartoon of the sun, are a rare example of using microtubules instead of actin filaments to extend, support, and retract long, thin processes bounded by the plasma membrane (Fig. 38.4). Microtubules in these axopodia are crosslinked into a precise geometrical array accounting for the rigidity of these long processes. Axopodia capture prey organisms and transport them toward the cell body for phagocytosis by two mechanisms. Some move along the surface of axopodia (similar to intraflagellar transport; Fig. 38.18). Other axopodia collapse owing to rapid depolymerization of the microtubules and drag the prey with them. Ca2+ influx appears to trigger depolymerization of the microtubules, but the details of the mechanism are not known.
Cell Shape Changes Produced by Contraction Cells can change shape by localized or oriented cytoplasmic contractions. Muscle contraction (Chapter 39) and cytokinesis (Chapter 44) are the best examples, but
contractions also remodel many embryonic tissues. Localized contractions at the base or apex of cells in a planar epithelium cause evaginations or invaginations (Fig. 38.5) such as those that form the neural tube (see Fig. 30.8) and glands that bud from the gastrointestinal tract and branches of the respiratory tract. Closure of the epidermis over a Drosophila embryo also requires contraction of a circumferential ring of cells (Fig. 38.5C). Tension generated by myosin-II and actin filaments deforms each cell and, collectively, the whole epithelium. Similarly, contraction of a ring of actin filaments associated with the zonula adherens of intestinal epithelial cells helps regulate the permeability of the tight junctions that seal sheets of epithelial cells (see Fig. 31.2).
Locomotion by Pseudopod Extension The ability to crawl over solid substrates or through extracellular matrix is essential for many cells. Perhaps the most spectacular example is the slowly moving growth cone of a nerve axon (Fig. 38.6A). Although moving less than 50 nm s−1, growth cones navigate
CHAPTER 38 n Cellular Motility
655
E. Growing filaments push membrane forward Extracellular stimuli F. Capping protein terminates elongation
D.
dP an is ys rol yd P-h AT
B. Signals activate WASp/Scar proteins
70˚
El on ga tio n
G.
Arp2/3 complex
C. WASp/Scars and existing filaments activate Arp2/3 complex to form a branch
PAK
i
n tio cia so dis
A. Pool of ATP-actin bound to profilin
H. ADF/cofilin severs ADP-actin filaments that depolymerize
J. LIM-kinase inhibits ADF / cofilin
ADF / cofilin Profilin
Profilin/Actin complexes I. Profilin catalyzes exchange of ADP for ATP FIGURE 38.7 A MODEL FOR ACTIN FILAMENT ASSEMBLY AND DISASSEMBLY AT THE LEADING EDGE. The reactions are separated in space for clarity but actually occur together along the leading edge. A, Cells contain a large pool of unpolymerized actin bound to profilin. B, Stimulation of cell surface receptors produces activated Rho-family guanosine triphosphatases (GTPases) and other signals that activate Wiskott-Aldrich syndrome protein (WASp)/Scar proteins. C, These proteins, in turn, activate nucleation of new actin filaments by Arp2/3 complex on the side of existing filaments. D, The new filaments grow at their barbed ends until they are capped (see F). E, Growing filaments push the plasma membrane forward. F, Capping protein terminates elongation. G, Polymerized adenosine triphosphate (ATP)-actin (yellow) hydrolyzes the bound adenosine triphosphate (ATP) to adenosine diphosphate (ADP) and inorganic phosphate (Pi) (orange), followed by slow dissociation of phosphate yielding ADP-actin (red). H, ADF/cofilins bind and sever ADP-actin filaments and that disassemble, releasing ADP-actin. I, Profilin promotes the exchange of ADP for ATP, restoring the pool of unpolymerized ATP-actin bound to profilin. J, Some of the same stimuli that initiate polymerization can also stabilize filaments when LIM-kinase phosphorylates actin-depolymerizing factor (ADF)/cofilins, inhibiting their depolymerizing activity. Inset, Electron micrograph of the branched network of actin filaments at the leading edge. PAK, p21-activated kinase. (Modified from Pollard TD, Blanchoin L, Mullins RD. Biophysics of actin filament dynamics in nonmuscle cells. Annu Rev Biophys Biomol Struct. 2000;29:545–576. Inset, Courtesy T. Svitkina and G. Borisy, Northwestern University, Evanston, IL.)
precisely over distances ranging from micrometers to meters to establish all the connections in the human nervous system, which consists of billions of neurons and approximately 1.6 million kilometers of cellular processes. Some epithelial cells (Fig. 38.6B) and white blood cells move much faster, about 0.5 µm s−1. These movements enable epithelial cells to cover wounds and allow leukocytes to move from the blood circulation to sites of inflammation (see Fig. 30.13) where they engulf microorganisms by phagocytosis (see Fig. 22.3). During vertebrate embryogenesis, neural crest cells migrate long distances through connective tissues before differentiating into pigment cells and sympathetic neurons. Fibroblasts lay down collagen fibrils as they move through the extracellular matrix (see Fig. 29.4). The best-characterized motile system is animal and amoeboid cells moving on a flat surface such as a microscope slide that may be coated with extracellular matrix molecules. These cells use actin polymerization to
extend the leading edge, then adhere to the underlying substrate, and (if the whole cell is to move) release and retract any attachments of its tail to the substrate. Animal cells move in other environments, such as threedimensional extracellular matrix, using variations of this theme as described at the end of the section.
Lamellar Motility on Flat Surfaces by Pseudopod Extension Pseudopods that lead the way in cell migration are filled with a dense network of branched actin filaments with their fast-growing barbed ends generally facing out towards the plasma membrane (Fig. 38.7). The leading lamella is autonomous for locomotion, and moves normally after amputation from the rest of the cell (Fig. 38.6B). Generally, the leading lamella is very flat, on the order of 0.25 µm thick, but some cells extend the lamellae up from the substrate into a wave-like fold called a ruffle (Fig. 38.3). Microtubules help maintain the polarized
656
SECTION IX n Cytoskeleton and Cellular Motility
A 0s
24 s
77 s
133 s
274 s
5 µm
B
4s
48 s
81 s
10 µm
FIGURE 38.8 DOCUMENTATION OF ACTIN FILAMENT DYNAMICS AT THE LEADING EDGE WITH FLUORESCENT ACTINS. A, Fluorescence photobleaching experiment with a stationary cell. Fluorescent actin is injected into a cultured epithelial cell and allowed to incorporate into filaments. A laser pulse bleaches some of the fluorescent actin, leaving a dark spot (arrow) that reveals movement of the filaments toward the cell center. The spot recovers fluorescence as diffusing fluorescent actin assembles new filaments in the bleached zone. B, Caged fluorescent actin experiment with a motile cell. Fluorescent dye bound to actin is masked with a chemical group preventing fluorescence. After incorporation into actin filaments of a fish keratocyte (see Fig. 33.2E), dyes in one area of the cell are uncaged with a light pulse (arrow), and red fluorescence is followed with time. Fluorescent actin filaments are stationary with respect to the substrate as the cell moves forward (upward). The fluorescent spot of marked filaments fades with time, owing to depolymerization and dispersal of the fluorescent subunits. (A, From Wang Y-L. Exchange of actin subunits at the leading edge of living fibroblasts: possible role of treadmilling. J Cell Biol. 1985;101:597–602, copyright The Rockefeller University Press. B, From Theriot JA, Mitchison TJ. Actin microfilament dynamics in locomoting cells. Nature. 1991;352:126–131.)
shape that is required for persistent directional locomotion, but are not required for pseudopod extension. Actin filaments assemble continuously near the leading edge of pseudopods and turn over rapidly deeper in the cytoplasm (Fig. 38.8). Thus, the inhibitor cytochalasin stops actin polymerization at the leading edge (see Fig. 33.18). Purified actin was labeled with a fluorescent dye and microinjected into live cells, where it incorporated into actin-containing structures, including the cortical network, pseudopods, stress fibers, and surface microspikes. Photobleaching (Fig. 38.8A) or photoactivating the fluorescent actin (Fig. 38.8B) showed that actin assembles at the leading edge and then turns over. Alternatively, if a low concentration of fluorescent actin is injected it will incorporate irregularly into filaments, producing tiny “speckles” of fluorescence that can be tracked over time to reveal sites of assembly and disassembly (see Fig. 33.18D–E). Arp2/3 complex (see Fig. 33.12) and formins cooperate with VASP (see Fig. 33.14) to initiate and elongate actin filaments at the leading edge of motile cells. Chemotactic stimuli (Fig. 38.10) or intrinsic signals
transduced by Rho-family guanosine triphosphatases (GTPases), membrane polyphosphoinositides, and proteins with SH3 (Src homology 3) domains activate Wiskott-Aldrich syndrome protein (WASp)/Scar proteins (see Fig. 33.13), which promote the formation of actin filament branches by Arp2/3 complex. The pool of unpolymerized actin maintained by profilin (and thymosin-β4, where it is present) drives the elongation of actin filament branches at 50 to 500 subunits per second. Growing filaments are generally oriented toward the leading edge and push against the inside of the plasma membrane with forces in the piconewton (pN) range. These forces bend the filaments, which favors branching on their convex surface, facing the leading edge. Heterodimeric capping protein (see Fig. 33.15) terminates elongation of these branches before they grow longer than 1 µm. Longer filaments are less effective at pushing, since they buckle under piconewton forces. A similar mechanism assembles actin filaments in comet tails on intracellular bacteria and viruses (see Fig. 37.12) and endocytic actin patches in yeast (see Figs. 6.3 and 37.11). The recycling of actin and accessory proteins is essential for thousands of rounds of assembly as the cell moves forward. Actin-depolymerizing factor (ADF)/ cofilins (see Fig. 33.16) bind and sever aged adenosine diphosphate (ADP)-actin filaments located away from the leading edge. The fragments may be capped, in which case they slowly depolymerize at the pointed ends. This recycling mechanism is so efficient that the branched network is disassembled in seconds. Formins assemble long unbranched filaments (see Fig. 33.2E) in the flat region of the cell just behind the branched network at the leading edge. Some of these filaments form bundles that protrude as filopodia beyond the leading edge (Fig. 38.3), where VASP (see Fig. 33.14) promotes their elongation. The tropomyosin binds along the sides of these long filaments, protecting them from severing by ADF/cofilins. Actin filament crosslinking proteins stabilize pseudopods. Human melanoma cells that lack the crosslinking protein filamin (see Fig. 33.17) form unstable pseudopods all around their peripheries and locomote abnormally (Fig. 38.9). These tumor cells recover their normal behavior if provided with filamin. Similarly, Dictyostelium cells that lack a homolog of filamin form fewer pseudopods.
Influence of the Substrate on Lamellar Motility The growing actin network at the leading edge will either push the membrane forward or slip backward depending on how well it is connected to the substrate across the plasma membrane. In highly motile cells such as epithelial cells from fish scales (Fig. 38.6B) transmembrane adhesion proteins anchor the actin filament network to the substrate, so the polymerization results in forward motion (Fig. 38.8B). In stationary cells (Fig. 38.8A;
CHAPTER 38 n Cellular Motility
A
B
Migration
FIGURE 38.9 CONTRIBUTION OF THE ACTIN FILAMENT CROSSLINKING PROTEIN FILAMIN TO THE STABILITY OF THE LEADING EDGE OF HUMAN MELANOMA CELLS. Pairs of phasecontrast light micrographs, taken at different times, of living cells grown in serum-containing medium on a plastic surface. A, Melanoma cells expressing filamin have normal leading lamella. B, Melanoma cells lacking filamin form spherical blebs around their margins and migrate very little. (Courtesy C. Cunningham and T.P. Stossel, Harvard Medical School, Boston, MA. For reference, see Cunningham C, Gorlin JB, Kwiatkowski DJ, et al. Actin-binding protein requirement for cortical stability and efficient locomotion. Science. 1992;255:325–327.)
also see Fig. 33.18D–E), actin polymerizes at the edge of the cell causing the entire network to move en masse away from the membrane, a phenomenon called retrograde flow. Fibroblasts are an intermediate state, in which actin polymerization produces some forward movement but also considerable retrograde flow. Movement driven by adhesion requires a compromise. Adhesion must be strong enough for the internal forces to propel the cell forward but not so strong that it prevents movement. Transmembrane adhesion proteins such as integrins (see Fig. 30.9) both anchor the cell and transmit the presence of their ligands and the stiffness of the environment to Rho-family GTPases that regulate actin assembly and myosin contractility. Rapidly reversible binding of integrins and other adhesion proteins to extracellular matrix molecules, such as fibronectin, allows adhesion without immobilization. Rapidly moving white blood cells attach weakly and transiently, whereas slowly moving fibroblasts form longer-lasting focal contacts (see Fig. 30.11). Cells tend to move up gradients of adhesiveness but stop if adhesion is too strong. This graded response to adhesion allows neural crest cells to migrate preferentially through regions of embryonic connective tissue marked by adhesive proteins.
Tail Retraction and Other Roles for Myosin in Motility Growth cones of neurons draw out a long process from a stationary cell body, but most cells must break adhesions at their trailing edge to advance. Adherent, slowly moving cells such as fibroblasts exert significant tension
657
on the underlying substrate, when myosin pulls on the bundles of actin filaments associated with focal contacts (see Fig. 30.11). When tension overcomes the attachments, the rear of the cell shortens elastically and then contracts further (Fig. 38.6C). Myosin also contributes to the retrograde flow of actin filaments in the zone between the leading edge and the cell body.
Other Modes of Motility Cells that move in two-dimensions by pseudopod extension and tail retraction repurpose the same proteins to behave differently in other environments. In the extracellular matrix of connective tissues (see Chapter 29) cells must squeeze between three-dimensional networks of fibers using forces produced by myosin to generate hydrostatic pressure to push forward and to pull the relatively stiff nucleus forward. In other circumstances cells can expand by forming blebs of plasma membrane lacking networks of actin filaments. Nerve growth cones (Fig. 38.6A) extend filopodia and then fill the spaces in between them with a lamella containing new, branched actin filaments. Superfast giant amoebas (Fig. 38.1) use myosin to generate contractions in the cortex or the front of the pseudopod to drive the bulk streaming of cytoplasm into advancing pseudopods. Chemotaxis of Motile Cells Extracellular chemical clues direct locomotion by influencing the formation and persistence of pseudopods. Movement toward a positive signal is called chemotaxis. The best-characterized example is the attraction of Dictyostelium to cyclic adenosine monophosphate (cAMP) (Fig. 38.10), the extracellular chemical that these amoebas use to communicate as they form colonies before making spores. Remarkably, these cells can sense a gradient of cAMP corresponding to a concentration difference of less than 2% along their length. This small difference is amplified into strong internal signals that control motility. Binding of cAMP to seven-helix receptors in the plasma membrane activates trimeric G-proteins inside the cell (see Fig. 25.9). The G-proteins activate pathways that regulate the activity of enzymes that control the concentration of the lipid second messenger phosphatidylinositol 3,4,5-trisphosphate (PIP3) in the plasma membrane: phosphatidylinositol 3-kinase (PI3K) synthesizes PIP3 and PTEN (phosphatase and tensin homolog) degrades it. (See Fig. 26.7 for details on polyphosphoinositides.) A fast positive pathway that is sensitive to local receptor occupancy and a slower global negative signal that is proportional to total receptor occupancy produce a gradient of PI3K activity inside the cell that is highest near the external source of cAMP. These pathways concentrate PTEN on the plasma membrane away from the source of cAMP. This complementary regulation of the kinase and phosphatase creates an internal gradient of PIP3 three to seven times steeper
658
SECTION IX n Cytoskeleton and Cellular Motility
A. Time course of turning toward gradient of cAMP
0
15
30
45
60
75
90
105
B. Amplification of the shallow external gradient of cAMP into a sharp biochemical gradient inside the cell:
135
150
C. Concentration of plasma membrane PIP3 toward the source of cAMP
Source of cAMP
Shallow gradient of cAMP outside
120
Membrane Receptor
Gradient of PIP3 on inside of plasma membrane
0 sec Fast, local excitation
8 sec Slow, global inhibition
20 sec Response
FIGURE 38.10 CHEMOTAXIS OF A DICTYOSTELIUM AMOEBA TOWARD CYCLIC ADENOSINE MONOPHOSPHATE. A, Live cell attracted to cyclic adenosine monophosphate (cAMP) (gold) released from a micropipette. A time series of differential interference micrographs shows the rapid formation of a new pseudopod and reorientation of the direction of movement when the position of the micropipette is moved at the 60-second time point. B, Cells have a uniform distribution of cAMP receptors (yellow and red dots) over their surface. A shallow gradient of cAMP activates these seven-helix receptors (red), which activate a trimeric G-protein and phosphatidylinositol 3-kinase, an enzyme that rapidly converts phosphatidylinositol 4-phosphate (PIP2) to phosphatidylinositol 3,4,5-trisphosphate (PIP3). On a slower time scale, the active G-protein activates phosphatase and tensin homolog (PTEN), a PIP3 phosphatase, throughout the cell. The combination of these two signals creates a steep gradient of PIP3 across the cell. C, Fluorescence micrograph of a cell exposed to a point source of cAMP (yellow). A green fluorescence protein (GFP)-pleckstrin homology (PH) domain fusion protein (green) inside the cell binds to PIP3 on the inside of the plasma membrane, revealing the steep gradient of PIP3. (A, Courtesy Susan Lee and Richard Firtel, University of California, San Diego. B, Modified from a sketch by Pablo Iglesias, Johns Hopkins University, Baltimore, MD. C, Courtesy Pablo Iglesias, Johns Hopkins University, Baltimore, MD. For reference, see Janetopoulos C, Ma L, Devreotes PN, Iglesias PA. Chemoattractant-induced phosphatidylinositol 3,4,5 trisphosphate accumulation is spatially amplified and adapts, independent of the actin cytoskeleton. Proc Natl Acad Sci U S A. 2004;101:8951–8956.)
than the external gradient of cAMP. Transduction of this internal gradient of PIP3 into motility requires Rho-family GTPases and formation of new actin filaments. Local polymerization and crosslinking of these actin filaments expand the cortex facing the source of cAMP into a new pseudopod and move the cell toward the cAMP. Leukocytes are attracted to chemokines and bacterial metabolites at sites of infection (see Fig. 30.13), especially small peptides derived from the N-termini of bacterial proteins, such as N-formyl-methionineleucine-phenylalanine (referred to as FMLP in the scientific literature). Similar to Dictyostelium, activation of seven-helix receptors and trimeric G-proteins amplifies shallow external gradients of FMLP into steeper internal gradients of PIP3 and other signals that control pseudopod formation. Negative signals also influence pseudopod persistence and the direction of motility. A classic example is the negative effect of contact with another cell. Loss of
contact inhibition of motility by tumor cells contributes to their tendency to migrate among other cells and spread throughout the body.
Growth Cone Guidance: A Model for Regulation of Motility Growth cones of embryonic nerve cells use a combination of positive and negative cues to navigate with high reliability to precisely the right location to create a synapse (Fig. 38.11). This combinatorial strategy is much more complex than the simple chemoattraction of Dictyostelium to cAMP, as expected for the more complicated task of connecting billions of neurons to each other or specific muscle cells. Cues for growth cone guidance come from soluble factors and cell surface molecules, each with a specific receptor on the growth cone. As in other systems, extracellular matrix molecules provide the substrate for growth cone movements. Precisely positioned expression of cue molecules and their
CHAPTER 38 n Cellular Motility
A
Two commissures per segment
Longitudinal axon bundles
Midline
B
Neurons with ipsilateral projections
Netrin gradient Slit in matrix repels growth cones with Robo1 High Robo1 repels growth cone from midline Frazzled receptor (DCC) for netrin attracts growth cone to midline Comm on glial cells downregulates Robo1 expression and allows growth cone to pass midline High Robo1 drives growth cone out of midline and prevents recrossing
FIGURE 38.11 DROSOPHILA GROWTH CONE GUIDANCE. A, Light micrograph of a filleted embryo showing the nerve cord stained brown with an axon marker. The axons of approximately 90% of neurons cross the midline a single time in a transverse nerve bundle called a commissure before running longitudinally in fascicles on each side of the midline. B, Drawing showing the ligands and receptors that guide growth cones across the midline and prevent their return to the ipsilateral (original) side. Frazzled receptors for netrin attract the growth cone to the midline where Comm downregulates the activity of Robo1, a repulsive receptor for Slit, allowing axons to cross the midline. (A, Courtesy John Thomas, Salk Institute, La Jolla, CA.)
receptors guides growth cones along a staggering number of different pathways, as illustrated by the following well-characterized examples. Localized cells in the nervous system, such as those in the floor plate of the developing spinal cord, secrete soluble guidance proteins such as netrin. Gradients of netrin provide long-range guidance for growth cones of cells that possess netrin receptors. Some netrin
659
receptors, such as members of the DCC family, steer the growth cone toward a netrin source by activating actin polymerization by VASP. Other netrin receptors repel the growth cone from netrin. Growth cones without these receptors are insensitive to this cue. Slit, a large extracellular matrix protein, repels growth cones with Slit receptors, which are immunoglobulin cell adhesion molecules (IgCAMs) called Robo1, Robo2, and Robo3. Mutations in the genes for these receptors cause growth cones to ignore Slit. IgCAM cell surface adhesion proteins (see Fig. 30.3), such as fasciculin II, prompt growing axons to bundle together in bundles called fascicles by homophilic interactions. Growth cones can be attracted out of these bundles to particular targets, such as muscle cells, that secrete chemoattractants or proteins that antagonize fasciculin II adhesion. The effects of mutations in genes for receptors and their ligands revealed how growth cones in Drosophila embryos navigate using multiple guidance cues (Fig. 38.11). Growth cones of neurons on one side of the nerve cord migrate across the midline to the opposite side and then navigate faithfully to their targets. Netrins secreted by cells at the midline attract growth cones expressing the DCC (Frazzled) netrin receptor. However, midline cells also secrete high levels of the matrix protein Slit, which repels growth cones. Growth cones cross the midline by downregulating the slit receptor. Once across the midline, the growth cones upregulate the slit receptor, so they never cross back to the side of origin. Local cues alert particular growth cones of motor neurons to branch off of fascicles to innervate individual muscle cells. Path finding by capillaries uses some of the same guidance mechanisms to grow blood vessels.
Eukaryotic Cilia and Flagella Axonemes built from microtubules and powered by dynein produce the beating of cilia and flagella (Fig. 38.12). These exceedingly complex structures are remarkably ancient. More than 1 billion years ago the last common eukaryotic ancestor (see Fig. 2.4) had motile flagella with all the essential features of human cilia and flagella, as evidenced by motile axonemes in most branches of eukaryotes. However, most fungi and plants lost the genes for axonemal proteins. Animal cells make three types of axonemes, all with nine outer doublet microtubules anchored by a basal body at the cortex of the cell and surrounded by plasma membrane (Fig. 38.12B). Motile axonemes have dynein motors, a central pair of single microtubules and radial spokes (Figs. 38.14 and 38.15). Rotating nodal cilia (see below for the biological context) have dynein arms but no central pair or radial spokes. Immobile primary cilia lack dynein, central pairs and radial spokes. This section begins with motile cilia and flagella.
660
SECTION IX n Cytoskeleton and Cellular Motility
B. Cilia
A. Flagella
Ciliary motion
Fluid motion
No motion
Cell motion Effective stroke
Flagellar motion
Recovery stroke Cell motion
Beating
Rotary
Primary
C. Cilia metachronal wave
Adapted from P. Satir
9 + 2 + dynein + radial spokes
9 + 0 + dynein
9+0
FIGURE 38.12 BEATING PATTERNS OF CILIA AND FLAGELLA. A, Waves of a sperm flagellum. B, Behavior of three types of cilia with cross sections of their axonemes below. Left, Ciliary power and recovery strokes of a beating cilium. Middle, Rotary cilium. Right, Primary cilium. C, Coordinated beating of cilia on the surface of an epithelium. (Modified from a drawing by P. Satir, Albert Einstein College of Medicine, New York, NY.)
Properties of Cilia and Flagella Cilia and flagella are distinguished from each other by their beating patterns (Fig. 38.12), but are nearly identical in structure. In fact, the flagella of the green alga Chlamydomonas can alternate between propagating waves typical of flagella and the oar-like rowing motion of cilia. So one can use the terms cilia and flagella interchangeably. Subtle differences in the mechanism that converts the dynein-powered sliding of the axonemal microtubules into movements determine which beating pattern is produced. Both cilia and flagella can propel cells as they cycle rapidly, beating up to 100 times per second. Propagation of bends along the length of individual flagella pushes a cell such as sperm forward (Figs. 38.1 and 38.15). Coordinated beating of many cilia can also move large cells (Fig. 38.13A). Reversal of the direction of the power stroke allows a unicellular ciliate to swim forward or backward. Alternatively, if the cell is immobilized, like the epithelial cells lining an animal respiratory tract or forming the embryonic skin (Fig. 38.13B), coordinated beating of cilia propels fluid and particles over their apical surface. Ctenophores fuse the membranes of many cilia together to make macrocilia that propel the organism in a manner similar to that of fins. Structure of the Axoneme The nine outer doublet microtubules of all axonemes consist of one complete A-microtubule with the usual 13 protofilaments, bearing an incomplete B-microtubule composed of 10 protofilaments attached to its side by distinct junctional structures on either side. The filamentous protein tektin associates with one protofilament in the wall of the A-microtubule. The central pair are typical 13-protofilament microtubules. The plus ends of all axonemal microtubules are at the distal tip. More than 200 accessory proteins associate with the 9 + 2 microtubules (Fig. 38.14A), making axonemes stiff
A
10 µm
B
FIGURE 38.13 IMAGES OF CILIA. A, Scanning electron micrograph of Paramecium showing waves of effective strokes passing regularly over the cell surface from one end to the other to keep the cell moving steadily forward. B, Fluorescence micrograph of cilia (green, stained with fluorescent antibodies to tubulin) on epithelial cells of frog skin also stained with Alexa 647-phalloidin for actin filaments. (A, Courtesy T. Hamasaki, Albert Einstein College of Medicine, New York, NY. From Lieberman SJ, Hamasaki T, Satir P. Ultrastructure and motion analysis of permeabilized Paramecium. Cell Motil Cytoskeleton. 9:73–84, 1988. B, Courtesy Brian J Mitchell of Northwestern University. From Werner ME, Hwang P, Huisman F, et al. Actin and microtubules drive differential aspects of planar cell polarity in multiciliated cells. J Cell Biol. 2011;195:19–26.)
but elastic. Pioneering genetic analyses and more recent proteomic studies established the locations of many of these polypeptides, such as the 17 proteins that make up the radial spokes between the central sheath and the outer doublets and the 11 protein subunits of the nexin-dynein regulatory complex that links outer doublets to each other. A long coiled-coil protein acts as a molecular ruler to specify the longitudinal positions of
CHAPTER 38 n Cellular Motility
A **
Large
* ** *
E Bt
*
*
ODA
*
*
(–)
*
D
*
* *
*
*
B
At
B-tubule
A
A-tubule
(+) N-RDC
24 nm
IDA
*
Spoke 24 nm
Small
B
661
C. Cilium with axoneme viewed from tip
Outer dynein arm
32 nm
Outer doublet microtubule Membrane Radial spoke Central sheath
Inner dynein arm
40 nm
S1
S2
96 nm
S3
S4
Central singlet tubule Link to membrane FIGURE 38.14 COMPOSITION AND STRUCTURE OF THE AXONEME. A, Two-dimensional gel electrophoresis separating more than 100 polypeptides of the axoneme of Chlamydomonas. Marked polypeptides (blue asterisks) are components of radial spokes. B, Electron micrograph of a thin cross section of a ciliary axoneme stained with tannic acid. C, Drawing of a cross section of a cilium. D, Three-dimensional reconstruction of one outer doublet and radial spoke based on electron microscopy tomography showing inner (IDA) and outer dynein arms (ODA) and a nexin-dynein regulatory complex (N-DRC; blue). E, A short section of an outer doublet. In this example, the outer arm dyneins have two heads. In some species, they have three heads. The dimensions indicate the longitudinal spacing between dynein arms and radial spokes. (A, Courtesy B. Huang, Scripps Research Institute, La Jolla, CA. B, Courtesy R. Linck, University of Minnesota, Minneapolis. D, From Song K, Awata J, Tritschler D, Bower R, et al. In situ localization of N and C termini of subunits of the flagellar nexin-dynein regulatory complex (N-DRC) using SNAP tag and cryo-electron tomography. J Biol Chem. 2015;290:5341–5353. E, From a Chlamydomonas axoneme modified from Amos LA, Amos WB. Molecules of the Cytoskeleton. New York: Guilford Press; 1991.)
many of these proteins with a 96 nm periodicity along the outer doublet microtubules. Central pair microtubules are connected by a bridge and decorated by elaborate projections. Mutations of these genes compromise axonemal function in experimental animals and human patients. A family of axonemal dyneins bound to outer doublets generates force for movement. Each dynein consists of a large heavy chain forming a globular AAA ATPase (adenosine triphosphatase) domain and a tail anchored to an A-tubule by light and intermediate chains. A thin stalk projecting from the catalytic domain exerts force on the adjacent B-tubule during part of the ATPase cycle (see Figs. 36.14 and 36.15). Like other axonemal proteins the pattern of dynein arms repeats every 96 nm along the A-tubule of each outer doublet (Fig. 38.14D). The outer row of dynein arms consists of four copies of three-headed molecules in each repeat. The inner dynein arms are more complicated: seven different dynein heavy chains form six arms with one head plus one arm with two heads in each repeat.
Mechanism of Axoneme Bending Dynein-powered sliding of outer doublets relative to each other bends axonemes. Sliding was first inferred from electron micrographs of the distal tips of
microtubules in bent cilia. Later, sliding was observed directly by loosening connections between outer doublets with proteolytic enzymes and then adding adenosine triphosphate (ATP) to allow dynein to push the microtubules past each other (Fig. 38.15B). Sliding can be followed in axonemes stripped of their membrane by marking outer doublets with small gold beads (Fig. 38.15A). As outer doublets slide past each other, the relative positions of the beads change. Dynein attached to one doublet “walks” toward the base of the adjacent microtubule, pushing its neighbor toward the tip of the axoneme. Biochemical extraction or genetic deletion of specific dynein isoforms alters the frequency and waveform of axonemal bending. Inner dynein arms are required for flagellar beating, and deletion of even a single type of inner-arm dynein can alter the waveform. Outer dynein arms are not essential but influence the beat frequency and add power to the inner arms. Humans with Kartagener syndrome lack visible dynein arms and have immotile sperm and cilia. As a result, affected males are infertile, and both men and women have serious respiratory infections, owing to poor clearance of bacteria and other foreign matter from the lungs. The mechanism of beating is intrinsic to the axoneme. Thus, sperm tail axonemes swim normally when provided
662
SECTION IX n Cytoskeleton and Cellular Motility
A
B
C
FIGURE 38.15 SLIDING MOVEMENTS OF OUTER DOUBLETS OF AXONEMES. A–B, Time series of darkfield light micrographs. A, Sea urchin sperm extracted with the detergent Triton X-100 and reactivated with ATP. Gold microbeads attached to two different outer doublets allow the visualization of their displacement as the tail bends. B, Fragment of a sea urchin flagellar axoneme treated with trypsin. The addition of ATP results in outer doublets sliding past each other out of the ends of the axonemal fragment. C, Electron micrograph of two outer doublets that have slid past each other in an experiment similar to that in panel B. (A, From Brokaw CJ. Microtubule sliding in swimming sperm flagella. J Cell Biol. 1991;114:1201–1215, copyright The Rockefeller University Press. B, Courtesy Ian Gibbons, University of California, Berkeley. For more information, see Summers KE, Gibbons I. ATP-induced sliding of tubules in trypsin-treated flagella of sea-urchin sperm. Proc Natl Acad Sci U S A. 1971;68:3092–3096. C, Courtesy P. Satir, Albert Einstein College of Medicine, New York, NY. For more information, see Sale WS, Satir P. Direction of active sliding of microtubules in Tetrahymena cilia. Proc Natl Acad Sci U S A. 1977;74:2045–2049.)
with ATP, even without the plasma membrane or soluble cytoplasmic components (Fig. 38.15A). Experiments with these demembranated sperm models revealed that the dynein ATPase activity is tightly coupled to movement. The beat frequency is proportional to ATPase activity, regardless of whether the frequency is limited by increasing the viscosity of the medium or the enzyme activity is limited by decreasing the ATP concentration. The bending that produces the bending waves of flagella or the power and recovery strokes of cilia results from local variation in the rate of sliding of pairs of outer doublet microtubules along the length of an axoneme. Coordination of these events is still being investigated, but at least two factors are involved. Mutations show that the central pair and radial spokes help coordinate the activity of the dyneins around the circumference of the axoneme as it bends. Mechanical constraints are also required to convert microtubule sliding into local bending. Destruction of the links between outer doublets frees them to slide past each other rather than bending the axoneme. Although axonemes function autonomously, signal transduction pathways regulate their activities. Phototaxis of Chlamydomonas is a particularly clear example of how fluctuations in intracellular Ca2+ can modify flagellar activity. The release of Ca2+ affects the two flagella of the organism differentially and allows a cell to steer
toward or away from light (Fig. 38.16). Ciliates also have mechanosensitive channels that depolarize the plasma membrane when the organism collides with something. Depolarization opens voltage-sensitive plasma membrane Ca2+ channels, admitting Ca2+ into the cell. This reverses the direction of ciliary beat. Both calcium and cAMPdependent phosphorylation of outer-arm dynein can change the beat frequency (all the way to zero) or alter the waveform.
Basal Bodies and Axoneme Formation The mother centriole matures into a basal body that anchors and templates the axoneme (Fig. 38.17; see also Fig. 34.3B). After migrating to the cortex during interphase, its distal appendages dock on the plasma membrane and the nine outer doublet microtubules of the axoneme grow directly from the nine outer triplet microtubules of the basal body. This differs from microtubule nucleation in the pericentriolar material during interphase (see Fig. 34.15). Some protozoa use basal bodies as centrioles during mitosis. Multiciliated cells form a basal body for each axoneme de novo. Basal bodies are typically anchored by protein fibers called rootlets. Cilia seem to function normally in mice lacking the major rootlet protein, but are unstable over the long term. Mutations of genes for basal body proteins compromise axonemal function in experimental animals and cause human diseases.
CHAPTER 38 n Cellular Motility
A
A. Normal forward swimming
B. Phototaxis
D. Photoshock
hν
hν
C Ca2+
Ca2+
663
B
Central singlet microtubule Membrane Outer doublet microtubules
Matrix
Triplet microtubules
C. Normal swimming parallel to new light direction
Cartwheel stucture
FIGURE 38.16 CHLAMYDOMONAS PHOTOTAXIS. A, Normal swimming toward the light using a cilia-like rowing motion of the flagella. Absorption of light by a sensory rhodopsin (related to sensory rhodopsins in Archaea) in the eyespot keeps the cell oriented. B, Moderate-intensity light from the side causes Ca2+ to enter the cytoplasm from outside the cell. The two flagella react differently, causing the cell to turn toward the light. C, Once the cell is reoriented, the flagella beat equally, and the cell swims toward the light. D, Highintensity light releases a high concentration of Ca2+ and causes transient wave-like motion of the flagella. This backward swimming allows the cell to reorient and to swim away from the light.
FIGURE 38.17 BASAL BODIES. A, Electron micrograph of a thin cross section of a basal body. B, Electron micrograph of thin longitudinal section of basal bodies and proximal axonemes of cilia. C, Drawings of cross sections and a three-dimensional (3D) drawing of the basal body and proximal flagella of Chlamydomonas. In the 3D drawing, the near side outer doublets are cut away to reveal the central pair microtubules. (A–B, Courtesy D.W. Fawcett, Harvard Medical School, Boston, MA. C, Modified from Amos LA, Amos WB. Molecules of the Cytoskeleton. New York: Guilford Press; 1991.)
664
SECTION IX n Cytoskeleton and Cellular Motility
A
C
Turnaround zone
Axonemes grow at distal tips
Anterograde IFT particle
Regenerating, short flagella Tagged tubulin
D
Kinesin 2
2 hours
Mate
Dynein
Fusing cell
Fused cell
Retrograde IFT particle
B
Cut
Cell body FIGURE 38.18 FLAGELLAR GROWTH AND INTRAFLAGELLAR TRANSPORT. A, Incorporation of protein subunits at the tip of growing Chlamydomonas flagella is revealed by an experiment involving the fusion of two cells, one expressing tubulin with an epitope tag that reacts with a specific antibody and the other regenerating its flagella. As is shown in the fluorescence micrograph, tagged tubulin is incorporated only at the distal tips of the growing flagella. Cells with paralyzed flagella made this experiment more convenient. B, Time course of regeneration of Chlamydomonas flagellum following amputation of one flagellum. The surviving flagellum shortens transiently before both grow out together. C, Electron micrographs of thin sections of Chlamydomonas flagella showing intraflagellar transport particles (arrows). D, Model for intraflagellar transport (IFT). (A, Courtesy K. Johnson, Haverford College, Haverford, PA. Inset, From Johnson KA, Rosenbaum JL. Polarity of flagellar assembly in Chlamydomonas. J Cell Biol. 1992;119:1605–1611, copyright The Rockefeller University Press. B, Based on the work of J. Rosenbaum, Yale University, New Haven, CT. C, Courtesy Joel Rosenbaum, Yale University, New Haven, CT.)
Some organisms regenerate flagella if they are severed from the cell (Fig. 38.18A–B). Absence of the flagellum activates expression of genes required to supply subunits for regrowth of the axoneme. In approximately 1 hour, the cell regrows a replacement flagellum, and the genes are turned off. Even more remarkably, if only one of the two flagella is lost, the remaining flagellum shortens rapidly to provide components required to make two half-length flagella (Fig. 38.18B). Then protein synthesis slowly provides additional subunits to restore both flagella to full length. Axonemes grow at their tips by incorporation of subunits synthesized in the cytoplasm (Fig. 38.18A). A process called intraflagellar transport (IFT) (Fig. 38.18C–D) carries individual proteins and subassemblies such as radial spokes to the growing tip. These cytoplasmic cargo proteins bind one of two IFT complexes, which associate 1 : 1 to form larger “trains” visible of electron microscopy (Fig. 38.18C). Kinesin-2 moves IFT trains toward the tip of the axoneme along the outer doublets just beneath the plasma membrane. Cytoplasmic dynein 1b transports particles back toward the cell
body. A separate complex, called a BBSome, associates with IFT trains and transports transmembrane proteins (including signaling receptors) bidirectionally along microtubules of the underlying axoneme. Movements of transmembrane proteins allows Chlamydomonas to glide on surfaces including the flagellum of a partner cell during mating. Because this motion does not require beating of the axoneme, the mechanism may represent an early stage in the evolution of flagella. IFT is remarkably similar to fast axonal transport (see Figs. 37.1 and 37.3) but on a smaller scale. Cargo proteins are loaded onto IFT complexes at the base of the cilium and then must pass through filters located just above the basal body. These filters have a cutoff of approximately 50 kD but pass much larger IFT complexes. Proteins, including the Ran GTPase, importins, and proteins of the nuclear pore complex (see Chapter 9), participate in this filtration system. After transport cargo proteins dissociate at the flagellar tip. Phosphorylation of kinesin at the tip of the axoneme may reverse transport for the return trip to the cell body. Trains are full of tubulin and other axonemal proteins in growing cilia; they keep moving
CHAPTER 38 n Cellular Motility
bidirectionally but are largely empty when cilia are not growing.
Rotary Cilia Single cilia on the epithelial cells of the “ventral node” of vertebrate embryos are required for the asymmetric location of some internal organs, such as the heart and liver, on opposite sides of the body. These nodal cilia lack the central pair microtubules and radial spokes. Rather than beating, the activity of the dynein arms causes the tip of the cilia to rotate clockwise. This rotary motion propels the extracellular fluid carrying certain growth factors toward the left side of the embryo. The absence of this flow explains why patients with Kartagener syndrome and mice missing a single dynein heavy chain have an equal chance of having their internal organs, such as heart and liver, positioned normally or on the opposite side, a condition called situs inversus. Rotary cilia may provide clues about an intermediate stage in the evolution of axonemes. Primary Cilia Except for blood cells, differentiated cells in vertebrate tissues produce a single primary cilium by growth of an axoneme from their older mother centriole (Fig. 38.19).
Golgi
Flagellum with 9 + 0 axoneme
665
The axonemes lack the central pair, and most lack dynein, so they are immotile (Fig. 38.12B). Primary cilia are sensory organelles for both chemicals and mechanical forces. For example, odorant receptors of nematode olfactory neurons concentrate in the membranes of primary cilia. Rod and cone photoreceptors in the eye are modified cilia with a basal body and a vestigial axoneme (see Fig. 27.2). Primary cilia on the epithelial cells of kidney tubules act as flow sensors, admitting Ca2+ into the cilium through mechanosensitive channels in the plasma membrane when bent. Many receptors concentrate in primary cilia including those for developmental morphogens such as sonic hedgehog and Wnt, growth factors including platelet-derived growth factor, and hormones like somatostatin. Some ligands activate local Ca2+ release into the cilium.
Ciliopathies Genetic deficiencies in the assembly of cilia or IFT result in a remarkably wide range of human disease syndromes, known collectively as ciliopathies. The underlying mutations are found in more than 20 genes encoding proteins for axonemes, basal bodies, and IFT. The defects can appear in virtually any organ, illustrating the diverse functions of primary and motile cilia. An early example was polycystic kidney disease, the most common cause of kidney failure. The most frequent mutations are in genes for a Trp-family calcium channel, but other patients have mutations in genes for IFT proteins. Kidney epithelial cells form abnormal cysts rather than tubules, perhaps as a result of abnormal cell division. Other ciliopathy mutations cause defects in the central and peripheral nervous systems, olfactory neurons, ear, liver, retina, skeleton, and reproductive organs. Patients with some ciliopathy syndromes are obese or have extra digits. Box 38.1 discusses exotic eukaryotic motility systems.
Bacterial Flagella Centriole
Golgi
FIGURE 38.19 PRIMARY CILIUM. Electron micrograph of a thin section of a mesenchymal cell with a primary cilium assembled from one of the two centrioles, which serves as the basal body. (From Fawcett DW. The Cell. Philadelphia, PA: WB Saunders; 1981.)
Bacteria use a reversible, high-speed, rotary motor driven by H+ or Na+ gradients to power their flagella (Figs. 38.24 and 38.25). Bacterial flagella differ in every respect from eukaryotic cilia and flagella. The bacterial flagellum is an extracellular protein wire (see Fig. 5.8), not a cytoskeletal structure like an axoneme inside the plasma membrane. Bacteria with multiple flagella are more common than those with single flagellum. The principles derived from studies of Escherichia coli and a few other bacteria apply generally, although other species exhibit many variations on this theme. A motor, embedded in the plasma membrane, turns the bacterial flagellum either clockwise or counterclockwise (viewed from the tip of the flagellum) like the propeller of a motorboat. In contrast to a motorboat, moving bacteria have no momentum, so they stop in a fraction of a nanometer if the motor stops. When
666
SECTION IX n Cytoskeleton and Cellular Motility
BOX 38.1 Exotic Eukaryotic Motility Systems In contrast to conventional motility systems used by eukaryotic cells discussed in the main text, a few eukaryotes evolved exotic motility systems, four examples of which are described here.
A Preformed Actin Filament Spring Sperm of the horseshoe crab, Limulus, use a novel acrosomal process to fertilize an egg (Fig. 38.20). They preassemble a coiled bundle of actin filaments crosslinked by a protein called scruin. This bundle is a tightly coiled spring. An encounter with an egg stimulates rearrangement of the crosslinks, causing the actin bundle to unwind. Uncoiling drives the bundle through a channel in the nucleus followed by extension of a process surrounded by plasma membrane that literally screws its way through the egg jelly to fuse with the egg plasma membrane.
Calcium-Sensitive Contractile Fibers The ciliate Vorticella avoids predators by contracting a stalk that anchors the cell to leaves or other supports (Fig. 38.21). The contractile fibril, called a spasmoneme, contracts faster than any muscle. Ca2+ released from tubular membranes associated with the spasmoneme triggers contractions, when it binds to spasmin, a calmodulin-like protein that forms 3-nm filaments. Ca2+ binding changes the conformation of spasmin and results in rapid shortening, because many spasmin subunits are assembled in series. The
B
A. Limulus
Egg stimulates secretion of acrosome and uncoiling of actin filament bundle Acrosome
Actin bundle extends acrosomal process
spasmoneme relaxes when Ca2+ dissociates. Energy for contraction is supplied indirectly when ATP-driven pumps create a Ca2+ gradient between the lumen of the membrane system and cytoplasm. Movement of Ca2+ down this gradient drives contraction. Proteins similar to spasmin are found in other ciliates, algae, fungi, and animals, where they are called centrin or caltractin. These calmodulin-like proteins form fibrils that anchor centrosomes and the basal bodies of cilia and flagella. Mutations that inactivate caltractin in algae or yeast compromise the functions of the microtubule organizers (centrosomes or spindle pole bodies; see Figs. 34.15 and 34.20) used for mitosis.
Major Sperm Protein, an Actin Substitute in Nematode Sperm Nematode sperm use amoeboid movements to find an egg rather than swimming with flagella like other sperm (Fig. 38.22). The behavior of these sperm is so similar to a small amoeba cell that anyone would have guessed that it is based on the assembly of actin filaments. However, actin is a minor protein in nematode sperm. Instead, sperm pseudopods are filled with 10-nm wide, apolar filaments assembled from dimers of a 14-kD protein with an immunoglobulin-like fold called major sperm protein. Proteins in the cytoplasm and associated with the plasma membrane guide the assembly of the filaments, which function remarkably like actin, even though they have no bound nucleotide and no known associated motor protein. The 10-nm filaments assemble at the leading edge of the pseudopod and remain stationary with respect to the substrate as the expanding pseudopod
A
C
C
Nucleus Coiled bundle of actin filaments
B
Axoneme
FIGURE 38.20 LIMULUS SPERM ACROSOMAL PROCESS. A, Uncoiling of a bundle of actin filaments extends the acrosomal process of the sperm of the horseshoe crab, Limulus. B–C, Electron micrograph of the actin filament bundle from the acrosomal process of Limulus and a three-dimensional reconstruction of one filament (yellow) decorated with crosslinking proteins (green). (Based on the work of L. Tilney, University of Pennsylvania, Philadelphia. B–C, Courtesy W. Chiu, Baylor College of Medicine, Houston, TX.)
FIGURE 38.21 CALCIUM-SENSITIVE CONTRACTILE FIBERS. A–B, Light micrographs of a group of vorticellid protozoa suspended from the bottom of a leaf, taken before (A) and after (B) contraction of their spasmonemes. C, Electron micrograph of a thin section of contractile fibers and tubular membranes that store and release calcium. (Courtesy W.B. Amos, MRC Laboratory of Molecular Biology, Cambridge, United Kingdom.)
CHAPTER 38 n Cellular Motility
667
BOX 38.1 Exotic Eukaryotic Motility Systems—cont’d A
B
C
E
D
F
Cell movement
pH 7.0
pH 6.8
FIGURE 38.22 MOTILITY OF NEMATODE SPERM. A, Scanning electron micrograph of an amoeboid sperm showing the anterior pseudopod and trailing cell body. B–C, Time series of differential interference contrast light micrographs showing movement of a live sperm by assembly of a network of fibers at the leading edge. Arrows mark the same point in the network, which is stationary with respect to the substrate. D, Transmission electron micrograph of an extracted sperm showing the fibers. E, Atomic model of a short segment of the sperm filaments consisting of a polymer of major sperm protein (MSP). F, Cycle of MSP assembly at the leading edge and disassembly at the cell body. (Courtesy T. Roberts, Florida State University, Tallahassee, and M. Stewart, MRC Laboratory of Molecular Biology, Cambridge, United Kingdom.)
advances. Filament bundles depolymerize at the interface between the pseudopod and the spherical cell body. A pH gradient promotes assembly of major sperm protein at the front and disassembly at the rear of the pseudopod. This highly efficient motility system is still unknown in other parts of the phylogenetic tree.
A
B
C
Axostyles, Specialized Microtubular Organelles Some protozoa use dynein to generate beating movements of large arrays of cytoplasmic microtubules called axostyles (Fig. 38.23). The mechanism seems to be similar to an axoneme, although the organization clearly differs. Crosslinking structures hold together sheets of singlet microtubules, which slide past each other as a result of the action of dynein motors on adjacent sheets. Coordinated beats of the axostyle distort the whole organism, allowing it to wiggle about.
8 µm FIGURE 38.23 MOTILE AXOSTYLE OF SACCINOBACULUS, A PROTOZOAN PARASITE OF TERMITES. The twisting motions of this intracellular assembly of microtubules cause the whole parasite to twist and turn in the gut of termites. A, Polarization light micrograph of an isolated axostyle. B, Drawing of part of the axostyle showing the arrangement of sheets of crosslinked microtubules. C, Transmission electron micrograph of a cross section of the axostyle showing microtubules crosslinked into sheets with dynein arms between the sheets. (Courtesy R. Linck, University of Minnesota, Minneapolis. From Woodrum D, Linck R. Structural basis of motility in the microtubular axostyle. J Cell Biol. 1980;87: 404–414, copyright The Rockefeller University Press.)
668
SECTION IX n Cytoskeleton and Cellular Motility
A
Flagella rotating counterclockwise at (60 – 270 Hz) form a bundle that propels the cell
B
Clockwise rotation during tumble (100 Hz)
C
Tethered cell (20 – 50 Hz)
D
Bead on polyhook (170 Hz)
FIGURE 38.24 DIFFERENT MANIFESTATIONS OF THE ROTATION OF FLAGELLA. A, If the flagella rotate counterclockwise, they form a bundle that propels the cell forward. B, If one or more flagella rotate clockwise, the bundle falls apart and the cell tumbles in one place. C, If a flagellum is tethered to a surface, the bacterium rotates. D, If the flagellar filament is replaced by an elongated hook region with an attached bead, the bead rotates. (Modified from Schuster SD, Khan S. The bacterial flagellar motor. Annu Rev Biophys Biomol Struct. 1994;23:509–539.)
multiple flagella are present, counterclockwise rotation forms a bundle. Four flagella propel E. coli 30 µm s−1, a velocity of 15 cell lengths per second, equivalent to 400 miles per hour if the bacterium were the size of an automobile. When one or more flagella reverse their direction and rotate clockwise, the bundle flies apart, and the cell tumbles in one place. Figs. 27.12 and 27.13 explain how chemotactic stimuli control the probability of clockwise rotation, favoring steady runs toward nutrients and allowing for more frequent tumbles to change direction to avoid harm. Assays for rotation of single flagella provide insights about the mechanism of flagellar motion (Fig. 38.24). When a flagellum is attached to a glass slide by means of antibodies to the flagellar filament, the bacterium rotates, providing decisive evidence for rotation of flagella.
Similarly, beads attached to short flagella are observed to rotate. The rotational speed depends on the resistance. The motor of a single immobilized flagellum can rotate a whole E. coli 10 to 50 times per second, whereas in some species unloaded motors rotate up to 1600 times per second (100,000 rpm)! The rotary engine driving the flagellar filament is constructed from a rotor and stator. The cylindrical basal body on the end of the filament rotates inside the stator, a ring of stationary proteins embedded in the plasma membrane and anchored to the peptidoglycan layer (Fig. 38.25). Genetic screens for motility mutants identified all the protein components of the motor, and their functions were defined by analysis of the behavior of these mutants. Most of these proteins are present in isolated basal bodies. The functional units of the stator consist of four MotA subunits and two MotB subunits. MotA has four hydrophobic segments that are believed to be transmembrane helices and a substantial cytoplasmic domain. MotB has one transmembrane segment and a large periplasmic domain anchored to the peptidoglycan layer. Flagella and basal bodies assemble but are immotile in cells lacking either of these transmembrane proteins. If the missing protein is replaced by initiating its biosynthesis, the paralyzed flagellum begins to turn, increasing its speed of rotation in a stepwise fashion, as 10 to 12 independent, torque-producing MotA4MotB2 units are added one after another. The energy to turn the motor comes from protons (or, in some bacteria, Na+ ions) that move down an electrochemical gradient from outside the bacterium through the MotA4MotB2 units to the cytoplasm. Transfer of one proton across the membrane provides approximately the same energy as the hydrolysis of an ATP. Pumps driven by light, oxidation, or ATP hydrolysis (see Table 14.1) generate the proton gradient. MotA is part of the proton channel, because mutations in its gene inhibit both flagellar rotation and proton permeability. The MotB transmembrane helix has a conserved aspartic acid residue that interacts with the proton crossing the membrane. The mechanism producing rotation is still under investigation, but it involves interaction of the cytoplasmic domain of MotA with FilG subunits on the top of the C-ring. The transfer of a proton across the plasma membrane results in a conformational change in MotA that moves the C-ring. Roughly 1000 protons cross the membrane for each rotation, corresponding to two protons for each tiny rotational step. Proton transfer is tightly coupled to rotation of the basal body, and the efficiency is near 100%.
CHAPTER 38 n Cellular Motility
669
Up to 2500 nm
Cap
Filament Stator complex
Hook
D MotB
L-ring Outer membrane Peptidoglycan
B
P-ring
C
MS-ring motor
Cytoplasmic membrane
Cytoplasmic structure: C-ring switch complex
A
Fli G Fli N Fli M
MotA
MotB MotA
E MotA
FIGURE 38.25 BACTERIAL ROTARY MOTOR. A, Drawing of the flagellar filament and rotary motor. B–C, Three-dimensional reconstruction and schematic cross section of basal body of the flagellar motor from Escherichia coli. D, Details of the stator, with schematic diagrams of the transmembrane and cytoplasmic parts of MotA4MotB2 hexamer. E, Electron micrograph of a freeze-fractured bacterium illustrating the ring of intramembranous particles comprising the stator of MotA4 MotB2 hexamers. (A, Modified from Schuster SD, Khan S. The bacterial flagellar motor. Annu Rev Biophys Biomol Struct. 1994;23:509–539. B–C, From Zhao X, Norris SJ, Liu J. Molecular architecture of the bacterial flagellar motor in cells. Biochemistry. 2014;53:4323–433. D, Stator unit with a ribbon diagram of the C-terminal domain of MotB. Schematic diagrams of MotA and MotB based on Kojima S, Imada K, Sakuma M, et al. Stator assembly and activation mechanism of the flagellar motor by the periplasmic region of MotB. Mol Microbiol. 2009;73:710–718. E, Courtesy S. Khan, Albert Einstein College of Medicine, New York, NY.)
SELECTED READINGS Blanchoin L, Boujemaa-Paterski R, Sykes C, et al. Actin dynamics, architecture, and mechanics in cell motility. Physiol Rev. 2014;94: 235-263. Carmeliet P, Tessier-Lavigne M. Common mechanisms of nerve and blood vessel wiring. Nature. 2005;436:193-2000. Condeelis J, Singer RH, Segall JE. The great escape: When cancer cells hijack the genes for chemotaxis and motility. Annu Rev Cell Dev Biol. 2005;21:695-718. Daniels DR. Effect of capping protein on a growing filopodium. Biophys J. 2010;98:1139-1148. Devreotes P, Horwitz AR. Signaling networks that regulate cell migration. Cold Spring Harb Perspect Biol. 2015;3:a005959. Gambardella L, Vermeren S. Molecular players in neutrophil chemotaxis—focus on PI3K and small GTPases. J Leukoc Biol. 2013;94:603-612. Goetz SC, Anderson KV. The primary cilium: a signalling centre during vertebrate development. Nat Rev Genet. 2010;11:331-344. Kim S, Dynlacht BD. Assembling a primary cilium. Curr Opin Cell Biol. 2013;25:506-511. Kolodkin AL, Tessier-Lavigne M. Mechanisms and molecules of neuronal wiring: a primer. Cold Spring Harb Perspect Biol. 2011;3:a001727. Lechtreck KF. IFT-cargo interactions and protein transport in cilia. Trends Biochem Sci. 2015;40:765-778. Levin M. Left-right asymmetry in embryonic development: A comprehensive review. Mech Dev. 2005;122:3-25. Lin J, Okada K, Raytchev M, et al. Structural mechanism of the dynein power stroke. Nat Cell Biol. 2014;16:479-485.
Martin AC, Goldstein B. Apical constriction: themes and variations on a cellular mechanism driving morphogenesis. Development. 2014; 141:1987-1998. Mizuno N, Taschner M, Engel BD, et al. Structural studies of ciliary components. J Mol Biol. 2012;422:163-180. Mogilner A, Rubinstein B. The physics of filopodial protrusion. Biophys J. 2005;89:782-795. Moriyama Y, Okamoto H, Asai H. Rubber-like elasticity and volume changes in the isolated spasmoneme of giant Zoothamnium sp. under Ca2+-induced contraction. Biophys J. 1999;76:993-1000. Petrie RJ, Yamada KM. Fibroblasts lead the way: a unified view of 3D cell motility. Trends Cell Biol. 2015;25:666-674. Pigino G, Ishikawa T. Axonemal radial spokes: 3D structure, function and assembly. Bioarchitecture. 2012;2:50-58. Pollard TD, Borisy GG. Cellular motility driven by assembly and disassembly of actin filaments. Cell. 2003;112:453-465. Reiter JF, Blacque OE, Leroux MR. The base of the cilium: roles for transition fibres and the transition zone in ciliary formation, maintenance and compartmentalization. EMBO Rep. 2012;13:608-618. Ridge KD. Algal rhodopsins: Phototaxis receptors found at last. Curr Biol. 2002;12:R588-R590. Ridley AJ, Schwartz MA, Burridge K, et al. Cell migration: Integrating signals from front to back. Science. 2003;302:1704-1709. Rørth P. Reach out and touch someone. Science. 2014;343:848-849. Smith HE. Nematode sperm motility. WormBook. 2014;4:1-15. Witman G. Chlamydomonas phototaxis. Trends Cell Biol. 1993;3: 403-408. Zhao X, Norris SJ, Liu J. Molecular architecture of the bacterial flagellar motor in cells. Biochemistry. 2014;53:4323-4433.
This page intentionally left blank
CHAPTER
39
Muscles M
uscles use actin and myosin to generate powerful, unidirectional movements (Fig. 39.1). The molecular strategies are specialized versions of those used by other cells to produce contractions, to adhere to each other and the extracellular matrix, and to control their activity. Vertebrates have three types of specialized contractile cells: smooth muscle, skeletal muscle, and cardiac muscle. These muscles have much in common, but differ in their activation mechanisms, arrangement of contractile filaments, and energy supplies. The nervous system controls the timing, force, and speed of skeletal muscle contraction over a wide range. Cardiac muscle generates its own rhythmic contractions that spread through the heart in a highly reproducible fashion. Neurotransmitters, acting like hormones, regulate the force and frequency of heartbeats over a narrow range. Nerves, hormones, and intrinsic signals control the activity of smooth muscles, which contract slowly but maintain tension very efficiently. This chapter explains the molecular and cellular basis for these distinctive physiological properties of the three types of muscle.
Skeletal Muscle Skeletal muscle cells (also called muscle fibers in the physiological literature) are among the largest cells of vertebrates. During development, mesenchymal stem cells give rise to progenitor cells with a single nucleus called myoblasts. A family of master transcription factors, including MyoD and myogenin, coordinates the expression of specialized muscle proteins. As they differentiate, myoblasts fuse and elongate to form muscle cells with multiple nuclei and lengths ranging from millimeters to tens of centimeters. The number of muscle cells is determined genetically and is relatively stable throughout life even as the size of the cells varies with the level of exercise and nutrition.
A basal lamina (see Fig. 29.17C) surrounds and supports each muscle cell. At the ends of each cell, actin thin filaments are anchored to the plasma membrane at myotendinous junctions, which are similar to adherens junctions (see Fig. 31.8). There, integrins spanning the membrane link actin filaments to the basal lamina and to collagen fibrils of tendons. These physical connections transmit contractile force to the skeleton. Mature muscles harbor small numbers of long-lived stem cells (see Fig. 41.14) with the potential to repair damage. They are called satellite cells because they are located inside the basal lamina next to the muscle cells. Some of these cells are capable of both self-renewal and, when the muscle is injured, producing progeny that can differentiate into myoblasts that can fuse with each other or existing muscle cells to repopulate the muscle. These features have made satellite cells a focus of research to treat degenerative diseases of muscle.
Organization of the Actomyosin Apparatus Skeletal muscle cells are optimized for rapid, forceful contractions. Accordingly, they have a massive concentration of highly ordered contractile units composed of actin, myosin, and associated proteins (Fig. 39.2). Actin and myosin filaments are organized into sarcomeres, aligned contractile units that give the cells a striped appearance in the microscope. For this reason, they are called striated muscles. Myosin uses adenosine triphosphate (ATP) hydrolysis to power contraction, which results from myosin-powered sliding of actin-based thin filaments past myosin-containing thick filaments. Speed of contraction is achieved by linking many sarcomeres in series. Power (force) is achieved by linking multiple sarcomeres in parallel. Nerve impulses stimulate a transient rise in cytoplasmic calcium that activates the contractile proteins. Interdigitation of thick, bipolar, myosin filaments and thin actin filaments in the sarcomeres of living muscle 671
672
SECTION IX n Cytoskeleton and Cellular Motility
A. Skeletal muscle
Section of sarcomere
Myofibril
B. Cardiac muscle
Muscle cell
C. Smooth muscle
Muscle FIGURE 39.2 CONTRACTILE APPARATUS OF STRIATED MUSCLES. The contractile unit is the sarcomere, an interdigitating array of thick and thin filaments. Sarcomeres are arranged end to end into long, rod-shaped myofibrils that run the length of the cell. Mitochondria and smooth endoplasmic reticulum separate myofibrils, which can readily be isolated for functional and biochemical studies.
FIGURE 39.1 LIGHT MICROGRAPHS AND INTERPRETIVE DRAWINGS OF HISTOLOGIC SECTIONS OF SKELETAL, CARDIAC, AND SMOOTH MUSCLES. A, Skeletal muscle cells are shaped like cylinders and may be up to 50 cm long. Multiple nuclei are located at the periphery near the plasma membrane. Striations are seen in the inset, a longitudinal section at high magnification. B, Cardiac muscle cells are striated and have one or two nuclei. Adhesive junctions called intercalated disks (bright pink vertical bars in the longitudinal section, top left arrows) bind these short cells together end to end. C, Smooth muscle cells are spindle shaped with homogeneous cytoplasm and single nuclei.
cells is so precise (Fig. 39.3) that it yields an X-ray diffraction pattern (see Fig. 39.11) revealing the spacing of the filaments and the helical repeats of their subunits to a resolution of about 3 nm. Z disks at both ends of the sarcomere anchor the barbed ends of the actin filaments, so their pointed ends are near the center of the sarcomere. Myosin heads project from the surface of thick filaments, whereas their tails form the filament backbone. Thick and thin filaments overlap, with the
myosin heads only a few nanometers away from adjacent actin filaments. The alignment and interdigitation of the filaments facilitate the sliding interactions required to produce contraction. An important, simplifying architectural feature is that sarcomeres are symmetrical about their middles (Fig. 39.3). Consequently, the polarity of myosin relative to the actin filaments is the same in both halves of the sarcomere, allowing the same force-generating mechanism to work at both ends of the bipolar myosin filaments. Sarcomeres are organized end to end into long, rodshaped assemblies called myofibrils (Fig. 39.2) that retain their contractility even after isolation from muscle. Thin Filaments Thin filaments are a polymer of actin with tightly bound regulatory proteins troponin and tropomyosin (Fig. 39.4). When the cytoplasmic Ca2+ concentration is low, troponin and tropomyosin inhibit the actin-activated adenosine triphosphatase (ATPase) of myosin. Tropomyosin, a 40-nm long coiled-coil of two α-helical polypeptides (see Fig. 3.10), binds laterally to seven
CHAPTER 39 n Muscles
A
A
Tropomyosin
C
B
Tropomodulin
Troponin
Barbed
Pointed
D
B
C
A band Z disk
I band
N
M line
C
E
673
Myosin polarity
Actin polarity
Myosin polarity
FIGURE 39.3 ELECTRON MICROGRAPHS AND DRAWINGS OF SARCOMERES. A, Longitudinal thin section showing the array of thin filaments anchored to Z disks and overlapping bipolar thick filaments crosslinked in the middle at the M line. B, Longitudinal freezefractured, etched, and shadowed sarcomere showing myosin cross-bridges attached to thin filaments near the bare zone in the center (right) of a sarcomere. C–D, Cross-sections of insect flight muscle and vertebrate skeletal muscle showing the double hexagonal arrays of thick and thin filaments. E, Drawings indicating the polarity of the thick and thin filaments. (A and C, Courtesy H.E. Huxley, Brandeis University, Waltham, MA. B and D, Courtesy J. Heuser, Washington University, St. Louis, MO.)
contiguous actin subunits as well as head to tail to neighboring tropomyosins, forming a continuous strand along the whole thin filament. Troponin (TN) consists of three different subunits called TNC, TNI, and TNT (see Table 39.1). TNT anchors one troponin complex to each tropomyosin coiled-coil. TNC is a dumbbell-shaped protein with four EF-hand motifs to bind divalent cations similar to calmodulin (see Fig. 3.12 and Chapter 26). In resting muscle, the C-terminal globular domain of TNC binds two Mg2+ ions and an α-helix of TNI, while the lowaffinity sites in the N-terminal globular domain of TNC
Troponin C alone
C N
Troponin C on TNI peptide
FIGURE 39.4 THIN FILAMENT STRUCTURE. A, Threedimensional reconstruction from electron micrographs of a thin filament from vertebrate skeletal muscle showing actin and the position of tropomyosin in relaxed muscle. B, Drawing of a model of a thin filament from active muscle based on reconstructions of electron micrographs and crystal structures of troponin and tropomodulin. Each troponin–tropomyosin unit is associated with seven actin subunits. C, Ribbon diagrams of the atomic structures of troponin C, free and bound to a troponin I peptide. Two divalent cation-binding EF hands are found at each end, separated by a long α-helix. In cells, two highaffinity sites at the C-terminal end are permanently occupied with Mg2+. Two low-affinity sites at the other end are unoccupied in relaxed muscle but bind Ca2+ when muscle is activated. (A, Based on Protein Data Bank [www.rcsb.org] file 1AX2, created by Roberto Dominguez, University of Pennsylvania.)
are empty. Ca2+ binding to the low-affinity sites (two in fast skeletal muscle; one in slow muscle) during muscle activation exposes a new binding site for TNI. The resulting conformational change in TNI allows tropomyosin to move away from the myosin-binding sites on the actin filament. A protein meshwork in the Z disk anchors the barbed end of each thin filament (Fig. 39.5). Some crosslinks between actin filaments consist of α-actinin, a short rod with actin-binding sites on each end (see Fig. 33.17). At least a half dozen structural proteins stabilize the Z disk through interactions with α-actinin, actin, and titin. Some of these proteins also have signaling functions. Proteins cap both ends of thin filaments. Cap-Z, the muscle isoform of capping protein (see Fig. 33.15), binds the barbed ends of thin filaments with high affinity, limiting actin subunit addition or loss. Tropomodulin associates with both tropomyosin and actin to cap and stabilize the pointed end of the thin filament (Fig. 39.4B). Tropomyosin and a gigantic filamentous protein, nebulin, stabilize thin filaments laterally. Nebulin consists of 185 imperfect repeats of a 35-amino-acid motif
674
SECTION IX n Cytoskeleton and Cellular Motility
A. Longitudinal section Z disk
B. Cross section
A
B
C
D C
M line
E
Bare zone
FIGURE 39.5 Z DISK STRUCTURE. A–B, Electron micrographs of thin sections perpendicular to and in the plane of the Z disk. C, Three-dimensional reconstruction, based on electron micrographs of the Z disk, showing the network of protein crosslinks that anchor the barbed ends of the yellow actin filaments. (Courtesy J. Deatherage, National Institutes of Health, Bethesda, MD and modified from Cheng NQ, Deatherage JF. Three dimensional reconstruction of the Z disk of sectioned bee flight muscle. J Cell Biol. 1989;108:1761–1774.)
that interact with each actin subunit, tropomyosin, and troponin along the length of thin filaments. Interactions with tropomodulin and Z disk proteins anchor nebulin at the two ends of the thin filament. These interactions influence the length of thin filaments, but nebulin does not act simply as a molecular ruler. Thick Filaments The self-assembly of myosin II (see Fig. 5.7) establishes the bipolar architecture of striated muscle thick filaments (Fig. 39.6). Some features of thick filaments are invariant, such as a superhelical backbone consisting of myosin tails, a surface array of myosin heads, the 14.3-nm stagger between rows of heads, and a central bare zone formed by antiparallel packing of tails. Phosphorylation of the myosin regulatory light chains favors release of the myosin heads from the surface of the filament in preparation for engagement with actin. Filaments may
FIGURE 39.6 STRUCTURE OF BIPOLAR THICK FILAMENTS. A, Electron micrograph of a thick filament isolated directly from skeletal muscle and prepared by negative staining. A single myosin molecule is shown at the same magnification at the lower left. The myosin tails form the backbone of the thick filament and allow the two myosin heads to swing out from the side (see the enlarged inset on the right). B, Reconstruction from cryoelectron micrographs of part of a tarantula skeletal muscle thick filament with the bare zone (not shown) to the right. The tails form the backbone of the filament and the myosin heads are folded back toward the bare zone. Space-filling models of the two heads of one myosin molecule are superimposed on the reconstruction. The catalytic domains are green and blue; the light chains are pink, orange, yellow, and tan. C, Cross section of vertebrate skeletal muscle showing the double hexagonal arrays of thick and thin filaments. D, Electron micrograph of a highly stretched sarcomere with the M line in the middle. Note the nine faint stripes of myosin-binding protein C along both halves of the think filaments. E, Drawing of protein links between thick filaments in the M line. (A, Courtesy John Trinick, University of Bristol, United Kingdom. For reference, see Knight P, Trinick J. Structure of the myosin projections on native thick filaments from vertebrate skeletal muscle. J Mol Biol. 1984;177:461–482. B, Courtesy J. Woodhead and R. Craig, University of Massachusetts Medical School, Worcester, MA. For reference, see Woodhead JL, Zhao FQ, Craig R, et al. Atomic model of a myosin filament in the relaxed state. Nature. 2005;436:1195–1199. C, Courtesy J. Heuser, Washington University, St. Louis, MO. D, Courtesy H.E. Huxley, Brandeis University, Waltham, MA.)
vary in length, diameter, and organization of the helical array of heads in various species. Invertebrate thick filaments have a core of paramyosin, a second coiled-coil protein, which is not found in vertebrates. Accessory proteins stabilize thick filaments in striated muscles (Table 39.1). Myosin-binding protein C, with
its fibronectin III and immunoglobulin domains, forms nine stripes along both halves of the thick filament (Fig. 39.6D) and also interacts with thin filaments. It finetunes the interactions of myosin heads with both filaments, maintaining the relaxed state but also favoring crossbridge formation when the muscle is activated. The “M line” in the center of the sarcomere of most types of muscle (Fig. 39.6D) is a three-dimensional array of protein crosslinks. These elastic crosslinks maintain the precise registration of thick filaments but also allow the filaments to move apart as the sarcomere shortens while maintaining a constant volume. At least three structural proteins and the enzyme MM-creatine phosphokinase (which transfers phosphate from creatine-phosphate to adenosine diphosphate [ADP]) are located in the M line. Titin Filaments Titin, the largest protein encoded by the human genome, forms a third array of filaments lying parallel to the thin and thick filaments and connecting the Z disk to the thick filaments and the M line (Fig. 39.7). Three titins flank each half thick filament. Although titin is the third most abundant protein in muscle, it is hard to preserve for electron microscopy, so these diaphanous filaments escaped notice for years. Each filament is a single polypeptide named after a mythological giant owing to its remarkable size: several splice isoforms consist of 27,000 to 33,000 amino acids folded into a linear array of up to 300 fibronectin III and immunoglobulin (Ig) domains measuring more than 1.2 µm long. The elasticity of titin molecules provides passive resistance to stretching of relaxed muscle. Titin connections to the Z disk and thick filaments provide physical continuity from one sarcomere to the next and keep the thick filaments centered in the sarcomere during contraction. Differential splicing creates titin isoforms that differ in stiffness for various types of muscle. If titin molecules are broken experimentally, thick filaments slide out of register toward one Z disk during contraction. Two features provide the elasticity during short (~0.3 µm per titin), physiological stretching (Fig. 39.7). The irregular chain of Ig domains in the I band straightens out, and a segment of the polypeptide rich in proline, glutamic acid, valine, and lysine (the PEVK domain) partly unfolds. This decreases entropy and provides the energy for elastic recoil (an entropic spring; see also Fig. 29.11). Extreme stretching unfolds Ig domains one by one. Titin not only responds to phosphorylation but also binds signaling proteins that influence the performance of muscle in the short and long term. Intermediate Filaments Desmin intermediate filaments (see Chapter 35) help align the sarcomeres laterally (Fig. 39.8) by linking each Z disk to its neighbors and to specialized attachment sites on the plasma membrane called costameres. In
675
CHAPTER 39 n Muscles
PEVK
Titin PEVK Ig domains
PEVK unfolds
Chain of Ig domains stretched Ig domains unfold
Sarcomere length = 2.4 µm
Stretched to 3.2 µm (physiological stretch length) Stretched to 3.6 µm
FIGURE 39.7 TITIN FILAMENTS. Top panel, Electron micrographs of single, isolated titin molecules prepared by heavy metal shadowing. Titin molecules are long enough to extend from the Z disk to the M line. Middle panel, Drawing of a sarcomere, to the same scale as the electron micrograph, with the thick filaments removed from the bottom half to illustrate how titin molecules anchor thick filaments to the Z disk and extend to the M line. Bottom panel, Drawing illustrating a model for the elasticity of titin. Modest stretches within the physiological range reversibly extend the chain of immunoglobulin (Ig) domains in the I-band and the PEVK domain. Extreme extension can unfold Ig domains. (Modified from Reif M, Gautel M, Oesterhelt F, et al. Reversible unfolding of individual titin immunoglobulin domains by AFM. Science. 1997;276:1090–1092. For reference, see Leake MC, Wilson D, Gautel M, Simmons RM. The elasticity of single titin molecules using a two-bead optical tweezers assay. Biophys J. 2004;87:1112–1135.)
Intermediate filaments
Z disk
M line
FIGURE 39.8 DESMIN INTERMEDIATE FILAMENTS IN SKELETAL MUSCLE. Desmin filaments connect Z disks laterally to each other and to the plasma membrane at specializations called costameres. (Modified from Lazarides E. Intermediate filaments as mechanical integrators of cellular space. Nature. 1980;283:249–256.)
676
SECTION IX n Cytoskeleton and Cellular Motility
A. Relaxed and stretched b
B. Contracted
0
Crossbridges
A
Force
Force
a
c
Sarcomere length
0
C
Velocity
C. Rigor showing crossbridges a
b
FIGURE 39.9 SLIDING FILAMENTS. Electron micrographs and interpretive drawings of longitudinal sections of a sarcomere from a relaxed muscle (A) and a contracted skeletal muscle (B). The lengths of the thin and thick filaments are constant as the sarcomere shortens, demonstrating that the filaments slide past each other during contraction. C, Crossbridges between thick and thin filaments from a muscle in rigor. (Micrographs courtesy H.E. Huxley, Brandeis University, Waltham, MA.)
addition to desmin, costameres contain clathrin plus several cytoskeletal proteins (vinculin, talin, spectrin, and ankyrin) found in focal contacts and adherens junctions of nonmuscle cells (see Figs. 30.11 and 31.8). Desmin mutations in humans cause disorganization of myofibrils, resulting in generalized muscle failure.
Molecular Basis of Skeletal Muscle Contraction Sliding Filament Mechanism The key to understanding muscle contraction was the discovery that thick and thin filaments maintain constant lengths and slide past each other as sarcomeres (and the muscle) shorten (Fig. 39.9). About the same time, it was discovered that crossbridges (now recognized to be myosin heads) can connect actin and myosin filaments, and that tension produced during contraction is proportional to the overlap of actin and myosin filaments (Fig. 39.10). Supported by biochemical and ultrastructural evidence for actin–myosin interaction, these pioneering observations led to the theory that crossbridges between the thick and thin filaments produce the force for contraction. Sixty years of research on crossbridges have yielded a detailed picture of the chemistry and molecular mechanics underlying the force-producing reactions. A review of the steps of the actomyosin–ATPase cycle (see Fig. 36.5) is helpful in understanding the contraction mechanism. Three different physiological states reveal information about crossbridge mechanisms. Relaxed. One extreme is relaxed muscle. When the concentration of cytoplasmic Ca2+ is low, tropomyosin
B
c
FIGURE 39.10 PHYSIOLOGICAL PROPERTIES OF SKELETAL MUSCLE. A, Dependence of maximum tension on the length of the sarcomeres. B, Interpretive drawings. Each relates to a point on A. C, Relationship of force and velocity during muscle contraction. (A, For reference, see Gordon AM, Huxley AF, Julian F. The variation in isometric tension with sarcomere length in vertebrate muscle fibres. J Physiol. 1964;171:28P–30P. C, From Ruch TC, Patton HD (eds). Physiology and Biophysics, 19th ed. Philadelphia: WB Saunders; 1965.)
and troponin inhibit the interaction of myosin heads with actin filaments, resulting in few myosin heads being bound. Lacking long-lived physical connections between the filaments, muscle offers little resistance to passive stretching. X-ray diffraction (Fig. 39.11) shows that the myosin heads (with bound ATP or ADP and phosphate) are closely associated with the backbone of thick filaments and arranged in a helical array determined by the thick filament structure (Fig. 39.6B). Rigor. The other extreme occurs after death. Depletion of ATP allows all myosin heads to bind tightly to actin filaments (Figs. 39.3B and 39.9C). By X-ray diffraction, the myosin heads bound to actin filaments contribute to the strength of the reflections from the actin filament helix. Strong physical connections between the filaments prevent stretching, making the muscle stiff (hence the term rigor mortis). This extreme condition illustrates what happens structurally and mechanically when all the crossbridges engage actin filaments. Contracting. The most interesting, but most complicated, state is actively contracting muscle. Myosin heads “walk” along actin filaments toward their barbed ends, pulling Z disks toward the center of the sarcomere. Thousands of sarcomeres shorten in series, causing the whole muscle to shorten. ATP is consumed and force is produced. The thick filament
CHAPTER 39 n Muscles
A
B. Relaxed
677
C. Contracting
Myosin head helix
Actin helix
Heads along thick filament
Actin helix
Dancing heads
FIGURE 39.11 CROSSBRIDGE DYNAMICS REVEALED BY X-RAY DIFFRACTION PATTERNS OF WHOLE MUSCLE. A, Electron micrograph showing the orientation of the muscle in the x-ray beam in B and C. B–C, Fiber diffraction patterns from relaxed and contracting skeletal muscles with interpretive drawings of crossbridges in each state. Reflections from myosin heads arranged on the thick filament helix are strong in relaxed muscle. Reflections from the actin helix are stronger than the thick filament helix in contraction. The myosin and actin reflections are each labeled in only one of four equivalent quadrants. During contraction, a few myosin heads attach transiently to actin, increasing the strength of the actin helix reflections, but most are disordered. (Micrograph and x-ray patterns courtesy H.E. Huxley, Brandeis University, Waltham, MA.)
helical pattern is very weak by X-ray diffraction (Fig. 39.11C). Actin reflections are stronger than relaxed muscle but not as strong as rigor. Disordered myosin heads are distributed between the thick and thin filaments as each one dances asynchronously on and off of actin filaments. Most myosin heads in contracting muscle have bound ATP or ADP-Pi (adenosine diphosphate–inorganic phosphate), allowing them to exchange rapidly among the four “weakly bound” states illustrated in Fig. 36.5. During some of the transient interactions of myosin–ADP-Pi with actin, phosphate dissociates from myosin, and the lightchain domain rapidly reorients (see Figs. 36.4 and 36.5). This stretches elastic elements in the myosin heads and both thick and thin filaments. Energy in these elastic elements can be used over a period of milliseconds to displace the actin filament relative to the crossbridge and contract the muscle. When ADP dissociates from the actin–myosin–ADP intermediate, ATP rapidly binds to the actin–myosin complex, dissociating the crossbridge and starting a new ATPase cycle. Relationship of Crossbridge Behavior to the Mechanical Properties of Muscle Normally, each sarcomere shortens less than 1 µm. However, the whole muscle shortens macroscopically, because it has thousands of sarcomeres in series. For
example, a human biceps muscle 20 cm long has approximately 80,000 sarcomeres in series from end to end. When each contracts 0.25 µm, the muscle shortens 2 cm. Because the system maintains a constant volume, each sarcomere and the whole muscle increase in diameter as they shorten. Although the individual filaments slide past each other relatively slowly (about 2 µm s−1 in both halves of each sarcomere), muscles contract rapidly because the motion of each sarcomere in the series is added together. In our example, without resistance, the biceps contracts approximately 3 cm in 100 ms during which most crossbridges pull just once. Crossbridge behavior explains why the velocity of muscle contractions of an active muscle depends on the external load (Fig. 39.10). When opposed by no load the molecular motion stored in elastic elements of each crossbridge is largely converted into movement of actin filaments relative to myosin filaments and contraction velocity is maximal. Under these conditions, the filaments in muscle slide past each other at a rate of about 5 µm s−1, the same speed observed for free actin filaments moving over myosin heads in vitro (see Fig. 36.6). For this rapid sliding to occur, myosin heads that do not produce force must not impede movement. If bound tightly to actin, they would interfere mechanically with rapid sliding. This is avoided by the rapid equilibrium of the myosin intermediates between being bound to actin and being free. Transient interactions of myosin heads
678
SECTION IX n Cytoskeleton and Cellular Motility
bound to ATP or ADP-Pi with actin do not produce force or retard sliding driven by force-producing crossbridges. Muscle produces maximum force when the contraction rate is zero (Fig. 39.10). The conformational change in the myosin head stretches elastic elements in the crossbridge, but the force cannot overcome the resistance from the load on the muscle. Consequently, the filaments do not slide, and energy stored in each stretched elastic element is lost as heat when the crossbridge dissociates at the end of the ATPase cycle. The maximum force depends on the numbers of sarcomeres in parallel, that is, the cross-sectional area of the muscle. This explains why muscles respond to strengthening exercises by growing in diameter.
Regulation of Skeletal Muscle Contraction Although skeletal muscle cells have only two states— inactive (relaxed) or active (contracting)—skeletal muscles produce a wide range of contractions, varying from slow and delicate to rapid and forceful. These graded contractions are achieved by varying the number of muscle cells activated by voluntary or reflex signals from the nervous system (Fig. 39.12). Control of Skeletal Muscle by Motor Neurons Neural stimuli that activate skeletal muscles arise in two ways (Fig. 39.12). In organisms with well-developed central nervous systems, most neural signals that activate skeletal muscles result from conscious decisions, providing voluntary control over skeletal muscles. Other signals
Sensory neuron Brain
Dorsal root
Ventral root Spinal cord
Motor neuron
Sensory nerve Motor nerve
Neuromuscular junction
Skeletal muscle cell
Sensory muscle spindle cell
FIGURE 39.12 INNERVATION OF SKELETAL MUSCLE. Motor neurons in the spinal cord stimulate one or (usually) more skeletal muscle cells. Two neural pathways control motor neurons. Some stimuli come from neurons in higher centers of the brain. This pathway provides voluntary control over muscle contraction. Other stimuli come through local reflex circuits from sensory detectors, including muscle spindle cells. These signals help coordinate muscle contraction in response to changing forces on and within the muscle.
result from reflex responses to stimulation of sensory nerves. Specialized muscle cells innervated with both motor and sensory nerves function as stretch receptors, relaying information about length and tension back to the spinal cord, where reflexes coordinate the motor neuron output. Neural inputs from both sources converge on motor neurons located in the brainstem and spinal cord of vertebrates. Axons of these motor neurons branch in a muscle to contact one or more muscle cells. A motor neuron together with its target muscle cells forms a motor unit. In the most precisely controlled muscles, such as the extraocular muscles in the eye, some motor neurons innervate single muscle cells. The contractile activity of a muscle is graded in terms of the speed and force of the contraction, so individual muscles can produce both delicate and powerful movements. Nerve stimulation determines the contractile force in two ways: (a) The number of active motor units determines how many muscle cells produce force, and (b) the rate of stimulation adjusts the force produced by active cells. Every time a muscle cell is stimulated, all the sarcomeres are activated, but the force that they produce increases as the rate of stimulation increases, up to a maximum of approximately 200 stimuli per second. The shortening velocity of an active muscle depends on the ratio between force produced and the resistance (Fig. 39.10C). If a large force or high velocity of contraction is required, many motor units are called into action. To sustain contraction, motor nerves fire repeatedly. By varying the number of active cells in a muscle and the rate of stimulation, the nervous system sets the force required for a particular movement. Synaptic Transmission at Neuromuscular Junctions The terminal branch of each motor neuron axon forms a large synapse called the motor end plate or neuromuscular junction on the muscle surface (see Fig. 17.9). These nerve endings are filled with synaptic vesicles containing the neurotransmitter acetylcholine. Arrival of an action potential at the nerve terminal stimulates fusion of synaptic vesicles with the nerve plasma membrane, releasing acetylcholine into the cleft between nerve and muscle. In less than a millisecond, acetylcholine diffuses across the extracellular space and binds to acetylcholine receptors concentrated in the adjacent muscle plasma membrane. Acetylcholine binding opens the receptor cation channel, initiating a new action potential that spreads over the muscle cell plasma membrane and intracellularly into the T tubules. Coupling Action Potentials to Contraction The plasma membrane of skeletal muscle cells, like that of nerve cells, is excitable (see Fig. 17.6) but, unlike that in nerves, it invaginates deeply to form T tubules that run across the entire cell (Fig. 39.13). Depending on the species and type of striated muscle (skeletal
CHAPTER 39 n Muscles
A
Plasma membrane
Entrance to T tubule
T tubule Smooth endoplasmic reticulum
B C. Skeletal muscle T tubule
Smooth ER
D. Cardiac muscle
Smooth ER
T tubule
FIGURE 39.13 PLASMA MEMBRANE SPECIALIZATIONS OF STRIATED MUSCLES. A–B, Electron micrographs of thin sections of fish skeletal muscle showing plasma membrane invaginations called T tubules, which cross the whole muscle cell and associate closely with smooth endoplasmic reticulum (ER). The complex of a T tubule with smooth ER on both sides is called a triad. A, A longitudinal section of two T tubules. B, A cross section of a T tubule flanked on two sides by smooth ER. Foot processes, consisting of voltage-sensitive calcium channels in the T tubule paired with calcium release channels in the ER, connect the T tubule to the smooth ER (see Fig. 39.14 for molecular details). C–D, The three-dimensional arrangement of T tubules and smooth ER relative to the sarcomeres in skeletal and cardiac muscle. (A–B, Courtesy C. Franzini-Armstrong and K. Porter, University of Pennsylvania.)
679
versus cardiac), T tubules may be located either at the level of the Z disks or at the thick filament ends. Inside the muscle cell, T tubules interact extensively with the smooth endoplasmic reticulum (SER) surrounding each myofibril. Historically, this SER was called sarcoplasmic reticulum. Terminal cisternae of SER are closely associated with T tubules at foot processes that can be visualized by electron microscopy. Together, T tubules and SER constitute a signal-transducing apparatus that converts depolarizations of the plasma membrane into a spike of cytoplasmic Ca2+ to trigger contraction (Fig. 39.14). An action potential moving through a T tubule triggers the release of Ca2+ from SER into the cytoplasm (Fig. 39.14). Ca2+ binding to troponin allows myosin to interact with the thin filament, initiating contraction. This signal transduction process is called excitation– contraction coupling. Three transmembrane proteins located in the T tubule and the terminal cisternae of the SER cooperate to generate the transient Ca2+ signal (Fig. 39.14). 1. A voltage-sensitive calcium channel (see Chapter 16) senses action potentials in the T tubule. These channels are called dihydropyridine (DHP) receptors, owing to their affinity for this class of drugs. The actual Ca2+ channel of DHP receptors is not essential for skeletal muscle contraction, as external Ca2+ is not required for contraction in the short term. 2. Ca2+ release channels (see Fig. 26.13), concentrated in the terminal cisternae of SER, release Ca2+ into the cytoplasm. A drug called ryanodine binds these channels and inhibits Ca2+ release. Every second ryanodine receptor is connected to cytoplasmic loops of four DHP receptors, forming bridges called foot processes between the T tubule and the endoplasmic reticulum (Fig. 39.13B). 3. The P-type calcium-ATPase (see Fig. 14.7) actively pumps Ca2+ from cytoplasm into the endoplasmic reticulum against a concentration gradient greater than 104. Inside the SER several low-affinity, highcapacity Ca2+-binding proteins buffer the millimolar concentration of Ca2+. For example, numerous carboxyl groups on the surface of calsequestrin bind Ca2+ with a millimolar Kd. This rapidly reversible reaction increases the Ca2+ storage capacity of endoplasmic reticulum without sacrificing the speed of Ca2+ release. Accessory subunits anchor calsequestrin to Ca2+ release channels, ensuring a local supply of Ca2+ for release into cytoplasm when muscle is activated. An action potential in a T tubule results in a transient rise in cytoplasmic Ca2+, from 0.1 µM to about 2 µM (Fig. 39.15), as follows. The action potential causes a short-lived conformational change in the DHP receptors that is transmitted mechanically to associated ryanodine receptor Ca2+ release channels. Many Ca2+ channels open transiently, allowing Ca2+ to diffuse down the steep
680
SECTION IX n Cytoskeleton and Cellular Motility
A. Skeletal muscle
MYOFIBRIL
B. Cardiac muscle
∆V SER
T TUBULE
Ca2+
Voltagesensitive channel
∆V SER
Ca2+
Ca2+
Ca2+ Ca2+
Ca2+ Ca2+
Ca2+
Ca2+ Ca2+
Ca2+
ATP ADP
Ca2+
Ca2+
ADP ATP Ca2+
Ca2+ Ca2+
Ca2+ Ca2+
Ca2+
Ca2+
Ca2+
Ca2+
Ca2+
Ca2+
Ca2+
Ca2+
Ca2+ Ca2+
Ca2+
Ca2+ Ca2+
Ca2+ Ca2+
Ca2+ Ca2+
Ca2+
Ca2+
∆V
ADP ATP Ca2+
Ca2+
∆V
Ca2+
ADP ATP Ca2+
FIGURE 39.14 MECHANISM OF CALCIUM RELEASE IN SKELETAL AND CARDIAC MUSCLES. Both muscle types use voltage-sensitive calcium channels in the T tubule membranes and calcium release channels in the smooth endoplasmic reticulum (SER). A, Direct coupling in skeletal muscle. An action potential in the T tubule (ΔV) activates the voltage sensor (turning from gray to blue). This direct contact opens the calcium release channel (turning from gray to pink). Cytoplasmic Ca2+ levels rise only briefly because calcium–ATPase (adenosine triphosphatase) pumps Ca2+ back into the lumen of the SER. B, Calcium-induced Ca2+ release in cardiac muscle. An action potential opens the voltage-sensitive Ca2+ channel in the T tubule, releasing Ca2+ into the cytoplasm. This Ca2+ opens the calcium release channel in the SER. ADP, adenosine diphosphate; ATP, adenosine triphosphate.
concentration gradient from the SER lumen to the cytoplasm. Physical connections between ryanodine receptors may spread their activation laterally, ensuring synchronous activation of a patch of channels. After a single action potential, the free Ca2+ in the cytoplasm rises for only a few milliseconds for three reasons. First, Ca2+ release channels close quickly. Second, cytoplasmic Ca2+ binds to troponin C and other proteins. Third, Ca2+ pumps efficiently transport cytoplasmic Ca2+ back into the lumen of the SER, even before the muscle develops maximum force. Ca2+ pumps are continuously active, keeping the cytoplasmic Ca2+ concentration low.
This is why repeated action potentials are required to prolong the rise in cytoplasmic Ca2+ (Fig. 39.15B). Transduction of the Calcium Spike Into Contraction Troponin–tropomyosin on thin filaments cooperates with myosin to turn on contraction in response to a Ca2+ spike. At rest, two Ca2+-binding sites of fast skeletal muscle troponin C are largely unoccupied (owing to their low affinity for Ca2+ and the low Ca2+ concentration). As a result, the troponin–tropomyosin complex partially blocks the binding site for myosin heads on actin (Fig. 39.16). This prevents most of the weak-binding
CHAPTER 39 n Muscles
681
A. Twitch
Force
Ca2+
Stimulus Force Ca2+
Ca2+
Force
Force
Relax ed
Stimulus
Partia lly
Active
B. Tetanus
active
Milliseconds
Ca2+
Milliseconds
A
FIGURE 39.15 CA2+ TRIGGERS CONTRACTION OF SKELETAL MUSCLE. In these experiments, the Ca2+-sensitive protein aequorin was injected into live muscle cells to provide a signal for the cytoplasmic Ca2+ concentration. A, Single stimulus. Cytoplasmic Ca2+ concentration increases transiently, followed by a short contraction. This brief contraction persists after cytoplasmic Ca2+ decreases to the resting level. B, Multiple stimuli. Each stimulus releases a new pulse of Ca2+, prolonging the contraction in so-called tetanus. (For reference, see Ridgway EB, Ashley CC. Calcium transients in single muscle fibers. Biochem Biophys Res Commun. 1967;29:229–234.)
Myosinbinding site
Relaxed Partially active
B myosin intermediates with ATP or ADP-Pi in the active site from binding the thin filament. When released into cytoplasm, Ca2+ binds troponin C, causing a conformational change that creates a binding site for a helical region of TNI. This interaction attracts the C-terminus of TNI away from actin and tropomyosin, allowing a small shift in the position of tropomyosin on the thin filament. This shift increases the probability that myosin-ADP-Pi heads will bind to the thin filament, dissociating their bound Pi and producing force. Activation is cooperative for three reasons: Ca2+ must occupy both binding sites on troponin C, the effects of Ca2+ binding and myosin binding are transmitted to neighboring tropomyosins through their end-to-end attachments, and every myosin that binds accentuates the response. This cooperativity makes the on–off switch respond very sharply to a relatively small, 10- to 20-fold change in the cytoplasmic Ca2+ concentration. The efficiency of this switch is underscored by the fact that the energy consumption of a muscle cell increases more than 1000-fold when it is activated. Activation of slow skeletal muscle (Table 39.3) and cardiac muscle is less cooperative, as their troponin C has only one Ca2+-binding site. Note that the muscle produces force well after the cytoplasmic Ca2+ concentration returns to resting levels (Fig. 39.15). The Ca2+-sensitive switch is sharp but relatively slow owing to the slow response of thin filaments
Active
FIGURE 39.16 THIN FILAMENT ACTIVATION MECHANISM. Reconstructions from electron micrographs showing a short segment of thin filament (A) and a cross section of a thin filament (B). Ca2+ binding to troponin C partially activates the filament by moving tropomyosin away from its lateral position in relaxed muscle, where it overlaps the myosin-binding site on actin (red). Myosin binding to the partially activated filament shoves tropomyosin further out of the way into the active position. (Data from W. Lehman, Boston University, MA.)
to Ca2+ binding. Ca2+ binds troponin C rapidly (milliseconds) but dissociates slowly (tens of milliseconds). Thus, after the Ca2+ spike saturates troponin C and the thin filament turns on, the muscle remains active even after free Ca2+ has returned to resting levels. Force declines slowly as Ca2+ dissociates from troponin C and returns to the SER without raising the cytoplasmic Ca2+ concentration. A single action potential produces a short contractile “twitch” (Fig. 39.15). Maximum contractile force is produced by a series of closely spaced action potentials, leading to a sustained rise in cytoplasmic Ca2+ and prolonged activation of actomyosin. The extended contraction is called tetanus. Regulation by Myosin Light Chains The participation of skeletal muscle myosin light chains in the regulation of contraction varies among species. The skeletal muscles of mollusks are one extreme. Their
682
SECTION IX n Cytoskeleton and Cellular Motility
TABLE 39.1 Sarcomere Proteins of Vertebrate Striated Muscles Name
Size (kD)
Domains
Functions
Disease Manifestations
Myosin
2 × 200
ATPase, coiled-coil
Motor, backbone of thick filament
HCM, HF, arrhythmias
Regulatory light chain
2 × 19
EF-hands
Stabilizes lever arm
HCM, arrhythmias
Essential light chain
2 × 18 or 25
EF-hands
Stabilizes lever arm
HCM, arrhythmias
Myosin-binding protein C
141
Ig, FNIII
Stabilizes thick filament
HCM, DCM, arrhythmias
Titin
3700
FNIII, IgC2, kinase
Elastic connection from Z disk to M line
DCM, HF, muscular dystrophy
Thick Filament
M Line MM-creatine phosphokinase
2 × 43
M protein
165
IgC2, FNIII
M-line fast skeletal muscle
Myomesin (skelemin)
185
IgC2, FNIII
Link M-disk to desmin
None yet known
Glycolytic enzyme
Thin Filament Actin
43
Thin filaments backbone
DCM, HF, myopathies
Tropomyosin
2 × 35
Coiled-coil
Blocks myosin binding to actin filament
HCM, DCM, arrhythmias, myopathies
Troponin C
18
EF-hands
Calcium-binding
DCM likely
Troponin I
21
Inhibitory component
HCM, arrhythmias, myopathies
Troponin T
31
Tropomyosin binding
HCM, DCM, arrhythmias, myopathies
Tropomodulin
43
Caps actin filament pointed end
None yet known
Nebulin
500–900
185 × 35 residues
Binds thin filament
Nemaline myopathy
α-Actinin
2 × 100
Actin-binding, spectrin repeats
Crosslinks thin filaments in the Z disk
Focal segmental glomerulosclerosis
CapZ
31 + 32
Caps actin filament barbed end
None yet known
Desmin
2 × 53.5
Anchors Z disk
DCM, myopathy
Z Disk
Intermediate filament
DCM, dilated cardiomyopathy; EF, calcium-binding helices E and F of calmodulin; FN, fibronectin; HCM, hypertrophic cardiomyopathy; HF, heart failure; Ig, immunoglobulin; MLCK, myosin light-chain kinase.
myosin light chains bind Ca2+ and provide the main on– off switch for contraction. When the Ca2+ concentration is low in resting muscle, no Ca2+ binds to light chains, and the actin–myosin ATPase is off. Ca2+ that is released during activation binds to the light chains, turning on the ATPase and contraction. At the other extreme, the light chains of vertebrate skeletal muscle myosin do not bind Ca2+, but their phosphorylation modulates contractile activity by increasing force production at suboptimal Ca2+ concentrations. Horseshoe crab skeletal muscle uses a dual system: Ca2+ binding to troponin–tropomyosin on thin filaments and Ca2+-regulated phosphorylation of myosin light chains both stimulate contraction.
Specialized Skeletal Muscle Cells All skeletal muscle cells are built on the same principles, but vertebrates actually have several different types of skeletal muscle cells, each with distinct isoforms of contractile protein and metabolic enzymes. The six myosin heavy chains and three actin isoforms are coded by different genes. In contrast, alternative splicing of one
TABLE 39.2 Muscle Cell Types Physiological Type
Myosin Type
Mitochondria
Fatigue
Fast, white
Fast
Few
Rapid
Intermediate
Fast
Medium
Medium
Fast, red
Fast
Many
Slow
Slow, red
Slow
Many
Slow
primary transcript (see Fig. 16.6) creates more than 50 isoforms of troponin T. Mutations in the genes for myosin, actin, desmin, nebulin, tropomyosin, troponin-I, and troponin-T can each cause defects in human skeletal muscles (Table 39.1). Physiological properties, such as the speed of contraction and the rate of fatigue, provide criteria for classifying muscle cells (Table 39.2). The isoforms of myosin (and probably the other contractile proteins) determine the speed of contraction, whereas the content of mitochondria and the oxygen-carrying protein myoglobin
CHAPTER 39 n Muscles
determines the endurance and overall color of the muscle. White muscle cells depend largely on glycolysis to supply ATP, accounting for their rapid fatigue compared with red muscle cells, which are specialized for oxidative metabolism with abundant mitochondria and myoglobin. Some muscles consist of only fast-twitch white muscle cells or slow-twitch red muscle cells, but most muscles are mixtures of two or more cell types. For example, in chickens, the leg muscles that are responsible for supporting the body, walking, and maintaining balance over long periods of time are rich in red muscle cells (“dark” meat). On the other hand, the chicken breast muscles, used for energetic flapping of the wings for short periods, are mainly white muscle cells (“light” meat). Remarkably, the pattern of nerve stimulation determines the muscle cell type by controlling which genes are expressed (and, presumably, how the troponin T messenger RNA is processed). This was demonstrated by transplanting motor nerves between fast and slow muscles. Over a period of weeks, slow isoforms replaced fast isoforms and vice versa. Even more surprising, the same result is achieved by stimulating muscles electrically with fast or slow patterns of impulses. Chronic low-level stimulation biases gene expression toward the proteins that are found in slow muscle cells. Calcium and calmodulin provide one prominent link between activity and gene expression. The concentration of active calmodulin tracks with the pattern of stimulation, because Ca2+ is released in the cytoplasm each time a muscle contracts. Among other things, calciumcalmodulin activates protein phosphatase PP2b (calcineurin; see Fig. 25.6), which dephosphorylates transcription factors (see Fig. 10.21). These activated transcription factors move into the nucleus and help to establish a transcription program that turns on expression of proteins found in slow muscles. These include contractile proteins and enzymes for oxidative metabolism. The proportions of slow and fast muscle cells are determined genetically, so world-class sprinters (with a high proportion of fast, white fibers) and marathoners (with a high proportion of slow, red fibers) are born with advantages for their specialties. Training can lead to hypertrophy of specific muscle cell types and improved performance. Endurance training also leads to an increased proportion of slow cells. Without training, muscle strength declines with age; cell number remains constant, but each cell decreases in size. Structural Proteins of the Plasma Membrane: Defects in Muscular Dystrophies In addition to providing a permeability barrier, the plasma membrane of the muscle cell must maintain its integrity while being subjected to years of forceful contractions. Occasional breaches of the membrane are inevitable, so muscle cells also depend on a repair
683
process that reseals holes. If membrane damage exceeds the repair capacity, muscle cells degenerate locally (segmental necrosis) or globally. Cell death beyond the ability of muscle stem cells to regenerate the tissue results in muscular dystrophy. The proteins that stabilize muscle membranes were discovered in the late 1980s, when mutations in the dystrophin gene on the X-chromosome were linked to Duchenne muscular dystrophy, the most common human form of the disease. Dystrophin is an enormous member of the α-actinin superfamily of actinbinding proteins (see Fig. 33.17). The dystroglycan– sarcoglycan complex (Fig. 39.17 and Table 39.3) was found when it copurified with dystrophin after solubilizing the membrane with detergents. More than 40 proteins are required to maintain the integrity of the plasma membrane as shown by mutations that cause muscular dystrophies (Table 39.3). Disease-causing mutations in genes for proteins of the dystroglycan–sarcoglycan complex typically lead to secondary loss of the other proteins in the complex. O-linked glycosylation by Golgi apparatus glycosyltransferases is
A
Laminin
B Sarcoglycan complex Sarcospan Plasma membrane Dystrophin
C
Dystroglycan complex Syntrophin complex
Actin filament
FIGURE 39.17 DYSTROPHIN AND ASSOCIATED PROTEINS STABILIZE THE PLASMA MEMBRANE OF SKELETAL MUSCLE. A–B, Fluorescent antibody staining of cross sections of human skeletal muscle showing the localization of dystrophin at the plasma membrane of a normal individual (A) and its absence in an individual with Duchenne muscular dystrophy (B). C, Model of the transmembrane complex of proteins that links dystrophin and actin filaments in cytoplasm to laminin in the basal lamina outside the cell. (A–B, Courtesy L. Kunkel, Harvard Medical School, Boston, MA. C, Based on a drawing by K. Amann and J. Ervasti, University of Wisconsin, Madison.)
684
SECTION IX n Cytoskeleton and Cellular Motility
TABLE 39.3 Proteins Required to Stabilize and Repair Muscle Plasma Membranes Protein
Partners/Functions
Expression
Inheritance, Diseases
Dystrophin
β-Dystroglycan, actin
Muscle, brain
X-linked, DMD, BMD
Utrophin
β-Dystroglycan, actin
Muscle, other tissues
α-Syntrophins
Dystrophin
Muscle > other tissues
None detected in humans
β-Syntrophins
Dystrophin, utrophin
Muscle > other tissues
None detected in humans
Caveolin-3
Cholesterol
Muscle
AD, LGMD
α-Dystroglycan
Laminin, agrin
Many tissues
Embryonic lethal
β-Dystroglycan
Dystrophin, utrophin
Many tissues
Embryonic lethal
α-Sarcoglycan
Sarcoglycans, biglycan
Muscle
AR, LGMD, cardiomyopathy
β-Sarcoglycan
Sarcoglycans
Muscle
AR, LGMD
γ-Sarcoglycan
Sarcoglycans, biglycan
Muscle
AR, LGMD
Integrin α7
Laminin
Many tissues
AR, CMD
Collagen VI α1, α2, α3
Biglycan
Muscle, other tissues
AD, Bethlem myopathy, Ulrich syndrome
α2-Laminin
α-Dystroglycan
Muscle, other tissues
AR, CMD, dy/dy mouse
Agrin
α-Dystroglycan, AChR
Muscle
Null mouse perinatal lethal
Titin
Myosin, Z-disk
Muscle
AR, LGMD, tibial MD
Myotilin
α-actinin, Z-disk
Muscle
AR, LGMD
Membrane Skeleton
Transmembrane Proteins
Extracellular Matrix
Sarcomeric Proteins
Golgi Enzymes That Process Membrane and ECM Proteins Fukutin
Glycosyltransferase
Many tissues
AR, Fukuyama CMD
LARGE
Glycosyltransferase
Many tissues
AR, CMD
POMGnT1
Glycosyltransferase
Many tissues
AR, Muscle eye brain disease
POMTi
O-mannosyltransferase
Many tissues
AR, Walker-Warburg syndrome
Muscle
AR, LGMD, Miyoshi myopathy
Membrane Repair Machinery Dysferlin Nuclear Envelope Proteins Emerin
Lamins, actin
All cells
XR, Emery-Dreifuss MD
Lamin A/C
Nuclear envelope
All cells
AD/AR, LGMD, Emery-Dreifuss MD
AD, autosomal dominant; AR, autosomal recessive; BMD, Becker muscular dystrophy; CMD, childhood muscular dystrophy; DMD, Duchenne muscular dystrophy; dy, dystrophia muscularis; ECM, extracellular matrix; LGMD, limb-girdle muscular dystrophy; MD, muscular dystrophy.
required for dystroglycan to bind to its extracellular ligands including α2-laminin in the basal lamina. Proteins in cytoplasmic vesicles are used to repair damaged plasma membranes. The mechanical activity of muscle cells might make them more sensitive than other cells to deficiencies in proteins that support the nuclear envelope (lamin A/C and emerin). Some of these mutations also affect the nervous system. Other than the X-linked dystrophin mutations, mutations causing muscular dystrophies are usually autosomal recessive. About one in several thousand humans develops some form of muscular dystrophy, because they inherit mutations in both copies of one of the sensitive genes. The mechanism of disease in muscular dystrophies is similar to that in hereditary spherocytosis, in which deficiencies of the membrane skeleton make red blood cells susceptible to mechanical damage (see Fig.
13.11). The age of onset and clinical features of inherited muscular dystrophies depend on the molecular defect. Patients with severe defects develop progressive muscle weakness as children. Ultimately, failure of respiratory muscles is fatal. Dystroglycans and a dystrophin homolog, utrophin, participate in clustering acetylcholine receptors at the neuromuscular junction, the chemical synapse between motor neurons and skeletal muscle (see Fig. 17.9). When, during development, a motor neuron contacts the surface of its target muscle cell, the neuron secretes a proteoglycan called agrin, which is incorporated into the adjacent basal lamina. Agrin binds dystroglycan and a receptor tyrosine kinase in the muscle plasma membrane, which position associated acetylcholine receptors at the site where they receive acetylcholine secreted by the nerve in response to an action potential.
CHAPTER 39 n Muscles
685
Cell A
Cell B
Sarcomere
Intercalated disk
Gap junction
FIGURE 39.18 ELECTRON MICROGRAPHS OF A LONGITUDINAL SECTION OF TWO CARDIAC MUSCLE CELLS. Sarcomeres are similar to skeletal muscle. Intercalated disks anchor neighboring cells together, and gap junctions couple the cells electrically. (Courtesy D.W. Fawcett, Harvard Medical School, Boston, MA.)
Cardiac Muscle To maintain the circulation of blood, heart muscle is specialized for repetitive (~100,000 times per day), fatigue-free contractions driven at regular intervals by action potentials from specialized pace-making cells within the heart. Gap junctions allow these action potentials to spread from one muscle cell to the next. Unlike skeletal muscle, the heart does not regenerate after injury, so efforts are being made to reprogram cardiac cells to recapitulate their normal development.
Contractile Apparatus of Cardiac Muscle The architecture of the sarcomeres is very similar in cardiac and skeletal muscle, but cardiac muscle has more mitochondria, larger T tubules, and less SER (Fig. 39.18). The human atrium has one major myosin isoform. Two ventricular myosin isoforms (one shared with slow skeletal muscle) differ in ATPase activity and speed of contraction. Humans almost exclusively express only one of these isoforms. Myosin-binding protein C binds at intervals along the backbone of the thick filaments, interacts with actin, and modulates the myosin cross-bridges. The thin filaments are composed of a cardiac isoform of αactin, tropomyosin, troponin, and a smaller version of nebulin called nebulette. When the cells are damaged by a heart attack or other disease, these proteins leak into the blood. Cardiac isoforms TNI and TNT in blood are a sensitive measure for cellular damage. Short, modestly branched, cardiac muscle cells have centrally located nuclei and squared-off ends (Fig. 39.1) where cadherins (see Fig. 30.5) attach neighboring cells to each other at specialized adhesive junctions called intercalated disks (Fig. 39.18). These junctions have features of both adherens junctions (links to
Atrioventricular node
Atria
Sinoatrial node
L R Ventricles
Bundle of His Purkinje fibers FIGURE 39.19 ACTIVATION OF CARDIAC CONTRACTION. An action potential (yellow arrows) starts at the sinoatrial node and travels through atrial muscle cells to the atrioventricular node. After a short delay at the atrioventricular node, the action potential spreads through the interventricular septum in modified cardiac muscle cells, called Purkinje fibers, and then through muscle cells to the whole ventricle. The action potential follows the same path each time, giving rise to electrical signals that can be detected on the body surface by electrocardiogram. Damage during myocardial infarctions changes the electrocardiographic pattern and may cause arrhythmias.
actin filaments) and desmosomes (links to intermediate filaments).
Pacemaker Cells Intrinsically excitable pacemaker cells in the sinoatrial node drive rhythmic contractions of the heart (Fig. 39.19). The membrane potential of these cells drifts spontaneously toward threshold, setting off action potentials about once each second (Box 39.1). Each
686
SECTION IX n Cytoskeleton and Cellular Motility
BOX 39.1 Action Potentials in Cardiac Pacemaker Cells Spontaneous action potentials of cardiac pacemaker cells in the sinoatrial node are more complicated than those of nerves (Fig. 39.20). Seven different plasma membrane channels determine their frequency: 1. Voltage-gated Na+ channels. As in nerves, these channels rapidly activate at membrane potentials above threshold and then rapidly inactivate.
0
1 µm
[Ca2+] i
-30 -45 7
0 0
A
2
3 1 100
4
7
2
200
3 1 300
-70
4
Voltage (mV)
-15
400
Time (msec)
Na+
1 Voltage-gated Na+-channel
Ca2+
2 T-type voltage-gated Ca2+-channel
Ca2+
3 L-type voltage-gated Ca2+-channel
K+
4 Voltage-gated delayed-rectifier K+-channel (HERG)
K+
5 Kir3.1 inwardrectifier K+-channel
K+
6 Kir6.2 inwardrectifier K+-channel
K+
Na+
7 Nonselective K+/ Na+-channel
B FIGURE 39.20 MECHANISM OF SPONTANEOUS CARDIAC PACEMAKER ACTION POTENTIALS. Sequential activation and inactivation of seven different plasma membrane channels account for the time course of action potentials and cytoplasmic Ca2+ transients in pacemaker cells of the sinoatrial node. A, Time courses of the fluctuations in membrane potential (orange) and cytoplasmic Ca2+ con centration (blue). Colored boxes indicate when the various channels enumerated in part B are open. Kir3.1 and Kir6.2 are inactive under these conditions. B, Channels contributing to pacemaker activity.
action potential spreads from cell to cell through gap junctions (see Fig. 31.6), activating all cells in the atrium within a few hundred milliseconds. After a brief delay in the atrioventricular node, the action potential and contraction spreads from cell to cell through the ventricle. This highly reproducible pattern of electrical
2. T-type, voltage-gated Ca2+ channels. These lowconductance channels activate transiently at membrane potentials more negative than Na+ channels, about −70 mV. 3. L-type, voltage-gated Ca2+ channels. These highconductance channels slowly activate and inactivate when the membrane depolarizes to about −40 mV. Sympathetic nerve stimulation sensitizes these channels to membrane depolarization. Drugs called dihydropyridines block these channels. 4. Delayed-rectifier, voltage-gated K+ channels. As in nerves, these HERG channels activate and inactivate slowly in response to membrane depolarization. Sympathetic nerves stimulate these channels. 5. Kir3.1 inward-rectifier K+ channels. These channels conduct K+ over a limited range of membrane potential, between about −30 and −80 mV. Parasympathetic nerve stimulation activates these channels. 6. Kir6.2 inward-rectifier K+ channels. Normal levels of cytoplasmic ATP inhibit these channels. Depletion of cytoplasmic ATP activates these channels. 7. Nonselective K+/Na+ channels. Repolarization of the membrane activates these HCN (hyperpolarizationactivated cyclic nucleotide-gated) channels, which slowly depolarize the membrane. Acting together, these channels produce a spontaneous cycle of pacemaker action potentials. At the threshold potential (about −40 mV), voltage-gated Na+ channels open synchronously and rapidly depolarize the membrane. As they inactivate, L-type Ca2+ channels open, prolonging the depolarization and admitting Ca2+; this, in turn, triggers contraction by releasing more Ca2+ from internal stores (Fig. 39.14B). As these Ca2+ channels slowly inactivate, delayed-rectifier K+ channels open and drive the membrane potential toward EK, the K+ equilibrium potential. As the membrane potential reaches a minimum, delayed-rectifier K+ channels inactivate, but the two Kir channels open. In the absence of other channel activity, the membrane potential would remain near EK, but the nonselective Na+/K+ channels open and the membrane slowly depolarizes, drifting toward threshold. T-type Ca2+ channels contribute to the slow, spontaneous depolarization. At threshold, the cycle repeats. Channels similar to those in the sinoatrial node generates action potentials in cardiac muscle cells and stimulate contraction. Cardiac muscle cells can generate spontaneous action potentials, but they have fewer T-type Ca2+ channels and more Kir K+ channels, so the rate of spontaneous action potentials is lower than that of pacemaker cells. Except in disease, pacemaker cells drive action potentials throughout the rest of the heart. Later in life, cells in the pulmonary veins can also generate action potentials. This is one cause of atrial fibrillation, a common arrhythmia in older humans.
activity can be recorded on the surface of the body as an electrocardiogram. Mutations in six different human ion channel genes including voltage-gated sodium channels (SCN5A) and potassium channels (HERG, KVLAT1, and minK) cause disorders of cardiac muscle electrophysiology. These
CHAPTER 39 n Muscles
A. Sympathetic stimulus Norepinephrine β-Adrenergic receptor
D Gαs
T Gαs Activates
Inhibits
687
B. Parasympathetic stimulus Ca2+
Adenylyl cyclase
L-type Ca-channel
K+ Acetylcholine Kir 3.1 Muscarinic Ach receptor K-channel
Gβγ + GαTι
ATP
GαDιβγ
Inactive cAMP Active PKA PKA
FIGURE 39.21 Regulation of the rate of cardiac pacemaker cells by sympathetic (A) and parasympathetic (B) nerves. GTP-Gαs stimulates adenylyl cyclase. GTP-Gαi inhibits adenylyl cyclase. D is GDP associated with G protein α-subunits; T is GTP. Ach, acetylcholine; cAMP, cyclic adenosine monophosphate; GDP, guanosine diphosphate; GTP, guanosine triphosphate; PKA, protein kinase A.
inherited diseases are called long-QT syndrome because the interval between the initial depolarization of the muscle cells and their relaxation is prolonged. This change predisposes the person to abnormal cardiac rhythms that are potentially fatal.
Excitation Contraction Coupling As in skeletal muscle, plasma membrane action potentials stimulate cardiac muscle cells to contract by releasing cytoplasmic Ca2+ to activate troponintropomyosin (Fig. 39.14B). Calcium ATPase pumps (see Fig. 14.7) in SER and plasma membrane maintain the low cytoplasmic Ca2+ concentration with some help from plasma membrane Na+/Ca2+ antiporters. In both cardiac and skeletal muscle action potentials activate L-type voltage-sensitive Ca2+ channels (DHP receptors) in T tubules, but subsequent events differ as revealed by a requirement for extracellular Ca2+ in heart but not skeletal muscle (Fig. 39.14). Rather than interacting directly with ryanodine receptors as in skeletal muscle, the active cardiac L-type voltage-sensitive Ca2+ channels admit extracellular Ca2+. This opens nearby ryanodine receptors in the SER, releasing a flood of Ca2+ to trigger contraction. This is called calcium-induced calcium release. Excitation-contraction coupling can be defective when heart muscle cells grow larger in response to abnormal demands, such as high blood pressure. The defect may be explained by growth separating T tubules from SER, either physically or functionally, thereby decreasing the probability that Ca2+ entering through DHP receptors will trigger Ca2+ release from the endoplasmic reticulum. Seven-Helix Receptors and Trimeric G-Proteins Regulate Heart Rate and Contractility Motor nerves do not stimulate cardiac muscle directly. Instead, molecules secreted by autonomic nerves and the adrenal gland modulate the rate and force of
contraction (Fig. 39.21). Acetylcholine from parasympathetic nerves slows the heartbeat, whereas norepinephrine from sympathetic nerves and epinephrine from the adrenal gland speed the rate and increase the strength of contraction (Fig. 39.21). These neurotransmitters target channels indirectly by activating two different seven-helix receptors and their associated trimeric G-proteins (see Fig. 25.9). The resting rate reflects a compromise in the competition between these two inputs. Norepinephrine and epinephrine increase the heart rate and force of contraction by modulating L-type Ca2+ channels (Fig. 39.21A). Both bind β-adrenergic receptors that activate trimeric G proteins, which stimulate the production of cyclic adenosine monophosphate (cAMP) by adenylyl cyclase (see Fig. 26.2). cAMP activates protein kinase A (PKA) (see Fig. 25.3) that phosphorylates three types of channels. Phosphorylated L-type, voltage-gated Ca2+ channels are more likely to open in response to membrane depolarization and admit more Ca2+ to activate the contractile machinery more fully. Phosphorylated nonspecific HCN (hyperpolarizationactivated cyclic nucleotide-gated) cation channels push the membrane potential toward threshold. Both increase the frequency of action potentials and the heart rate. Phosphorylated delayed-rectifier K+ channels are more active in repolarizing the membrane, so they prevent activated Ca2+ channels from prolonging the action potential. Phosphorylation of phospholamban, the regulatory subunit of the calcium pump in the endoplasmic reticulum, increases its activity, which speeds relaxation These changes allows heart muscle cells to keep up with stimuli generated at a higher rate from pacemaker cells. PKA also targets sarcomeric proteins: phosphorylation of troponin-I and myosin-binding protein C each increases the rate of crossbridge cycling and the strength of contraction. Acetylcholine released by parasympathetic nerves activates Kir3.1 inward-rectifier K+ channels that
688
SECTION IX n Cytoskeleton and Cellular Motility
slow the heartbeat (Fig. 39.21B). Acetylcholine binds seven-helix receptors, called muscarinic acetylcholine receptors (because they bind muscarine. These are distinct from pentameric nicotinic acetylcholine receptors (see Fig. 16.12). Active muscarinic receptors catalyze the exchange of guanosine diphosphate (GDP) for guanosine triphosphate (GTP) on the Gαi subunit of a trimeric G protein, releasing the Gβγ subunits to activate Kir3.1/3.4 channels. When open, these channels reduce the rate at which the membrane potential drifts toward threshold. The free GTP-Gαi subunits inhibit cAMP production and reduce Ca2+ channel phosphorylation. This decreases the probability that Ca2+ channels are open, contributing to a lowering of the heart rate. If the energy supply of the heart is compromised, ATP levels fall. This activates Kir6.2 channels, which reduce the rate of spontaneous depolarization and lower the heart rate until ATP levels are restored.
Therapeutic Effect of Digitalis in Congestive Heart Failure In congestive heart failure, cardiac contraction fails to produce enough force to maintain adequate circulation of blood. Digitalis from the foxglove plant was the first drug to treat heart failure. Digitalis and related compounds strengthen cardiac contraction indirectly by inhibiting the α2 isoform of Na+K+-ATPase in the plasma membrane (Fig. 39.22). Reduced sodium pump activity lowers the Na+ gradient across the membrane, providing less driving force for Na+/Ca2+ antiporters to exchange extracellular Na+ for cytoplasmic Ca2+. The slightly higher steady-state concentration of Ca2+ in cytoplasm strengthens contraction.
3 Na+
3 Na+ 2 K+ ATP ADP
3 Na+
Ca2+
3 Na+
Ca2+
Na+ /
Ca2+ Antiporter Na+ K+–ATPase (target of cardiac glycosides)
FIGURE 39.22 CHEMIOSMOTIC CYCLE THAT HELPS CLEAR CA2+ FROM THE CYTOPLASM OF CARDIAC MUSCLE CELLS. Na+K+-ATPase pumps create a Na+ gradient (purple triangle) to drive the Na+/Ca2+ antiporter to transport Ca2+ up its concentration gradient (blue triangle) out of the cell. Cardiac glycosides such as digitalis inhibit the cardiac isoform of Na+K+-ATPase, raising the concentration of cytoplasmic Ca2+ and strengthening cardiac contraction. K+ channels (left) allow K+ to circulate.
Molecular Basis of Inherited Heart Diseases Because the heart is so vital to survival, relatively minor molecular defects command attention in humans. Nearly 1% of individuals carry an inherited or de novo mutation in a gene for a sarcomeric protein that compromises cardiac function. The most common mutations are in the genes for myosin heavy chain, myosin-binding protein C, and titin, but virtually every sarcomeric protein is affected (Table 39.1). Long, noncoding RNAs participate in regulating the stress response that leads to cardiomyopathies. Patients are typically heterozygous for these mutations (ie, they are dominant negative mutations). In hypertrophic cardiomyopathies, the heart attempts to compensate for abnormal contractility (either increased or decreased depending on the mutation) through hypertrophy, but the thickened heart wall compromises cardiac relaxation and refilling the chambers with blood. In dilated cardiomyopathies, the ventricles swell and their walls thin. Both types are associated with heart failure and abnormal cardiac rhythms that can be fatal. The rate of progress of these diseases depends on not only the particular mutation but also other factors that vary from person to person. Individuals with defects in myosin-binding protein C develop hypertrophy in their fifties but can live normal life spans. By contrast, those with defects in troponin T can be affected as teenagers and die of arrhythmias in their twenties. These severe mutations of cardiac contractile proteins account for about half of the deaths of apparently healthy young athletes. Smooth Muscle Contractile Apparatus Smooth muscle cells are specialized for slow, powerful, efficient contractions under the control of a variety of involuntary mechanisms. Smooth muscle cells are generally confined to internal organs, such as blood vessels (where they regulate blood pressure), the gastrointestinal tract (where they move food through the intestines), and the respiratory system (where they control the diameters of the air passages; their excessive contraction contributes to asthma and other allergic reactions). The cytoplasm of spindle-shaped smooth muscle cells (Fig. 39.1) appears homogeneous by light microscopy, because the contractile proteins are not organized in regular arrays like sarcomeres of skeletal and cardiac muscle. A basal lamina and variable amounts of collagen and elastic fibers surround each cell. Smooth muscle cells rarely divide in adults, but they are capable of responding to local physiological conditions by remodeling their structure, as in atherosclerosis and hypertension. In terms of organization and biochemistry, smooth muscle cells (Fig. 39.23) resemble nonmuscle cells more than skeletal or cardiac muscle. For example, the gene for smooth muscle myosin arose relatively recently from a cytoplasmic myosin II gene. These myosins also share
CHAPTER 39 n Muscles
689
Dense plaque on plasma membrane
Dense body in cytoplasm Myosin filament Intermediate filament Actin filament
C A A
D D
B
Dense body Myosin Actin IF
E E FIGURE 39.23 CONTRACTILE APPARATUS OF SMOOTH MUSCLE. A, Electron micrograph of a thin cross section of two smooth muscle cells. B–C, Organization of the contractile units, which stretch across the cell between plasma membrane attachment plaques. Contractile units consist of myosin filaments connecting thin filaments attached to a dense body or plasma membrane plaque. D, High-power electron micrograph showing a dense body and cross sections of three types of filaments. E, Electron micrograph of a longitudinal section of an extracted vascular smooth muscle cell illustrating associations of actin filaments and intermediate filaments (IF) with dense bodies, and myosin filaments interacting with actin filaments. (A, D, and E, Courtesy A.V. Somlyo and A.P. Somlyo, University of Virginia, Charlottesville. For reference, see Somlyo AP, Devine CE, Somlyo AV, Rice RV. Filament organization in vertebrate smooth muscle. Philos Trans R Soc Lond B Biol Sci. 1973;265:223–229; Bond M, Somlyo AV. Dense bodies and actin polarity in vertebrate smooth muscle. J Cell Biol. 1982;95:403–413.)
the same regulatory light chain. Long myosin thick filaments are interspersed among the thin filaments, but not in a regular way as in striated muscles. Thin filaments are composed of actin and tropomyosin, along with two regulatory proteins, caldesmon and calponin, rather than troponin. Thin filaments are arranged obliquely in the cell, some with their barbed ends attached to dense plaques on the plasma membrane, others to dense bodies in the cytoplasm. Like Z disks in striated muscles, dense bodies anchor desmin intermediate filaments, forming a continuous, inextensible, internal “tendon”
running from end to end of the cell, preventing excess stretching (see Fig. 35.8). Smooth muscle cells contract like a concertina (Fig. 39.24), because tension generated by myosin and actin is applied to discrete spots on the plasma membrane. This compression can be seen in light micrographs as irregular cells with “corkscrew” nuclei. Given that smooth muscle cells have less myosin than striated muscle cells do, it is remarkable that they develop the same force. This is explained by two factors. First, the force-generating unit, the myosin filament, is larger in
690
SECTION IX n Cytoskeleton and Cellular Motility Relaxed
Contracted
Stimulus
Ca2+
Force
Myosin-LC-P Ca2+
A
Seconds
Stimulus
Trimeric G-protein
Calmodulin Ca2+ Ca-CM
B
–Ca2+ MLCK active
IP3 Myosin inactive
Phospholipase C
RhoGTP
MLCK inactive
P
P Myosin active
Myosin-LC Phosphatase active Phosphatase
Rho-kinase P-MLCPhosphatase inactive
FIGURE 39.24 ACTIVATION OF SMOOTH MUSCLE CONTRACTION. A, The spindle-shaped smooth muscle cell develops accordion pleats as it contracts, owing to the attachment of the actin filaments at intervals along the plasma membrane. The graph shows the time course of activation, consisting of the release of Ca2+ into the cytoplasm, phosphorylation of myosin regulatory light chains, and then the slow development of force. Myosin light-chain phosphorylation (LC-P) is required to initiate, but not to sustain, the contraction of smooth muscle. B, Biochemical pathways controlling phosphorylation of myosin regulatory light chains. Receptor stimulation leads to production of IP3 (inositol 1,4,5-triphosphate) by phospholipase C and release of Ca2+ into cytoplasm. Ca2+ binds calmodulin (CM), which activates myosin light-chain kinase (MLCK) by binding the kinase’s autoinhibitory peptide and displacing it from the active site. Active MLCK phosphorylates activating sites on the regulatory light chain. Light-chain phosphatase reverses phosphorylation of myosin. Activation of the small GTPase (guanosine triphosphatase) Rho with GTP stimulates Rho-kinase, which phosphorylates and inactivates lightchain phosphatase. This makes the system more sensitive to Ca2+ levels, as light-chain phosphorylation is prolonged. LC, light chain; P-MLC-, phosphorylated myosin light chain. (A, Modified from the work of K. Kamm and J. Stull, University of Texas Southwestern Medical School, Dallas.)
smooth muscle than in skeletal muscle. Deploying a given amount of myosin in large, thick filaments in a long sarcomere produces more force than does the same myosin in smaller filaments arranged in a series of short sarcomeres. Second, individual smooth muscle myosin molecules produce a larger force than skeletal muscle myosin, at least in vitro assays.
Regulation of Smooth Muscle Contraction A wide range of stimuli trigger smooth muscle contraction, but they all seem to act through seven-helix receptors coupled to trimeric G-proteins. Hormones stimulate contraction of the uterus, whereas motor nerves stimulate intrinsic eye muscles that close the pupil. Depending on the particular smooth muscle, Ca2+ for contraction enters the cytoplasm through either voltage-dependent calcium channels in the plasma membrane or IP3 (inositol 1,4,5-triphosphate) receptor Ca2+ release channels in the SER (Fig. 39.24). Drugs that block plasma membrane calcium channels can distinguish these two pathways experimentally. In intestines, parasympathetic nerves release acetylcholine to stimulate seven-helix muscarinic receptors (Fig. 39.21). Associated trimeric G-proteins activate cation channels that depolarize the plasma membrane and allow Ca2+ to enter through voltage-sensitive calcium channels. Consequently, calcium channel blockers strongly inhibit activation of gut smooth muscle. Gap junctions couple gut smooth muscle cells, allowing excitation to spread from cell to cell. At the other end of the spectrum, vascular smooth muscle depends on IP3 to release Ca2+ from intracellular stores rather than depending on Ca2+ from outside the cell. Following stimulation, intracellular Ca2+ increases rapidly but transiently, declining to a value above resting level as the receptors desensitize (see Fig. 24.3). Ca2+ pumps in both SER and plasma membrane clear the cytoplasm of Ca2+ so that Ca2+ levels decrease to resting levels and the muscle eventually relaxes when the activating stimulus declines. Relaxing agents, acting through cyclic guanosine monophosphate (cGMP) or cAMP (see Fig. 26.1), promote clearance of cytoplasmic Ca2+. Epinephrine relaxes smooth muscles of the respiratory system by another mechanism. Stimulation of β-adrenergic receptors activates potassium channels that hyperpolarize the plasma membrane and reduce Ca2+ entry. This approach is widely used to treat asthma. After a considerable delay (>200 ms) following the Ca2+ spike, contractile force develops slowly. The delay is attributable to the time required for a sequence of three biochemical reactions: Ca2+ binding to calmodulin, calcium-calmodulin activation of myosin lightchain kinase (see Fig. 25.4), and phosphorylation of myosin regulatory light chains, turning on the myosinactin ATPase cycle (Fig. 39.24). Unphosphorylated myosin-II from smooth muscle and vertebrate nonmuscle cells is inactive. Phosphorylation of myosin light chains is required to initiate but not maintain contraction, so slowly cycling, unphosphorylated myosins maintain peak force with little expenditure of energy. Regulation of unphosphorylated crossbridges is not well understood, but they appear to be activated cooperatively by a small population of phosphorylated myosin heads. Caldesmon, a
calcium-calmodulin-binding protein associated with tropomyosin on actin filaments, may contribute to activation and/or allow myosin heads to cycle very slowly even in the presence of ATP. The sensitivity of light-chain phosphorylation to Ca2+ depends on a parallel signaling pathway that partially inhibits myosin phosphatase, thus increasing the number of phosphorylated myosin cross-bridges and force at any given Ca2+ concentration (Fig. 39.24). Receptors coupled to trimeric G-proteins activate the small guanosine triphosphatase (GTPase) RhoA, which stimulates a protein kinase that inhibits myosin lightchain phosphatase. ACKNOWLEDGMENTS We thank Lee Sweeney and John Solaro for their suggestions on revisions to this chapter. SELECTED READINGS Agarkova I, Perriard J-C. The M-band: an elastic web that crosslinks thick filaments in the center of the sarcomere. Trends Cell Biol. 2005;15:477-485. Alexander MR, Owens GK. Epigenetic control of smooth muscle cell differentiation and phenotypic switching in vascular development and disease. Annu Rev Physiol. 2012;74:13-40. Bolton TB, Prestwich SA, Zholos AV, Gordienko DV. Excitationcontraction coupling in gastrointestinal and other smooth muscles. Annu Rev Physiol. 1999;61:85-115. Butler T, Paul J, Europe-Finner N, Smith R, Chan EC. Role of serinethreonine phosphoprotein phosphatases in smooth muscle contractility. Am J Physiol Cell Physiol. 2013;304:C485-C504. Cahill TJ, Ashrafian H, Watkins H. Genetic cardiomyopathies causing heart failure. Circ Res. 2013;113:660-675. Clark KA, McElhinny AS, Beckerle MC, Gregorio CC. Striated muscle cytoarchitecture: an intricate web of form and function. Annu Rev Cell Dev Biol. 2002;18:637-706. Doles JD, Olwin BB. Muscle stem cells on the edge. Curr Opin Genet Dev. 2015;34:24-28. Franzini-Armstrong C, Protasi F, Ramesh V. Comparative ultrastructure of Ca2+ release units in skeletal and cardiac muscle. Ann N Y Acad Sci. 1998;853:20-30. Gao N, Huang J, He W, et al. Signaling through myosin light chain kinase in smooth muscles. J Biol Chem. 2013;288:7596-7605. Geeves MA, Holmes KC. Structural mechanism of muscle contraction. Annu Rev Biochem. 1999;68:687-728. Gokhin DS, Fowler VM. A two-segment model for thin filament architecture in skeletal muscle. Nat Rev Mol Cell Biol. 2013;14: 113-119. Gordon AM, Homsher E, Regnier M. Regulation of contraction in striated muscle. Physiol Rev. 2000;80:853-924. Guiraud S, Aartsma-Rus A, Vieira NM, et al. The pathogenesis and therapy of muscular dystrophies. Annu Rev Genomics Hum Genet. 2015;16:281-308.
CHAPTER 39 n Muscles
691
Han R, Campbell KP. Dysferlin and muscle membrane repair. Curr Opin Cell Biol. 2007;19:409-416. Hidalgo C, Granzier H. Tuning the molecular giant titin through phosphorylation: role in health and disease. Trends Cardiovasc Med. 2013;23:165-171. Hinson JT, Chopra A, Nafissi N, et al. Titin mutations in iPS cells define sarcomere insufficiency as a cause of dilated cardiomyopathy. Science. 2015;349:982-986. Hwang PM, Sykes BD. Targeting the sarcomere to correct muscle function. Nat Rev Drug Discov. 2015;14:313-328. Hwang JH, Zorzato F, Clarke NF, Treves S. Mapping domains and mutations on the skeletal muscle ryanodine receptor channel. Trends Mol Med. 2012;18:644-657. Lehrer SS, Geeves MA. The myosin-activated thin filament regulatory state, M−-open: a link to hypertrophic cardiomyopathy (HCM). J Muscle Res Cell Motil. 2014;35:153-160. Marx SO, Marks AR. Dysfunctional ryanodine receptors in the heart: new insights into complex cardiovascular diseases. J Mol Cell Cardiol. 2013;58:225-231. Moss RL, Fitzsimons DP, Ralphe JC. Cardiac MyBP-C regulates the rate and force of contraction in mammalian myocardium. Circ Res. 2015;116:183-192. Murthy KS. Signaling for contraction and relaxation in smooth muscle of the gut. Annu Rev Physiol. 2006;68:345-374. Myhre JL, Pilgrim D. A Titan but not necessarily a ruler: assessing the role of titin during thick filament patterning and assembly. Anat Rec (Hoboken). 2014;297:1604-1614. Rao JN, Madasu Y, Dominguez R. Mechanism of actin filament pointedend capping by tropomodulin. Science. 2014;345:463-467. Somlyo AP, Somlyo AV. Ca2+-sensitivity of smooth and non-muscle myosin II: modulation by G Proteins, kinases and myosin phosphatase. Physiol Rev. 2003;83:1325-1358. Spudich JA. Hypertrophic and dilated cardiomyopathy: four decades of basic research on muscle lead to potential therapeutic approaches to these devastating genetic diseases. Biophys J. 2014;106: 1236-1249. Takamori M. Structure of the neuromuscular junction: function and cooperative mechanisms in the synapse. Ann N Y Acad Sci. 2012;1274:14-23. Takeda S, Yamashita A, Maeda K, Maeda Y. Structure of the core domain of human cardiac troponin in the Ca2+-saturated form. Nature. 2003;424:35-41. Teekakirikul P, Padera RF, Seidman JG, Seidman CE. Hypertrophic cardiomyopathy: translating cellular cross talk into therapeutics. J Cell Biol. 2012;199:417-421. Tskhovrebova L, Trinick J. Making muscle elastic: the structural basis of myomesin stretching. PLoS Biol. 2012;10:e1001264. von der Ecken J, Müller M, Lehman W, et al. Structure of the F-actintropomyosin complex. Nature. 2014;519:114-117. Wang YX, Dumont NA, Rudnicki MA. Muscle stem cells at a glance. J Cell Sci. 2014;127:4543-4548. Wehrens XH, Lehnart SE, Marks AR. Intracellular calcium release and cardiac disease. Annu Rev Physiol. 2005;67:69-98.
This page intentionally left blank
SECTION
Cell Cycle
X
This page intentionally left blank
SECTION X OVERVIEW T
his last section of the book draws together principles from previous chapters to explain some of the rules that govern the lifestyles of cells. Cells exhibit a remarkable diversity in their patterns of growth, proliferation, and death. For example, some human cells (neurons) are born around the time of birth and live until the person dies—more than 100 years in a few cases. The fate of other cells is to live for only a day or two (eg, cells in the gut lining). Many differentiated cells form by elaborate pathways that employ a carefully choreographed series of cues from within the cell and from its neighbors. Other cells, such as many in the immune system, are spawned in excess, followed by selection of the few with correctly rearranged genes or with productive connections to partner cells. The unlucky majority of their siblings whose differentiation did not go so well ultimately commit suicide. Very different strategies maintain populations of cells. Long-lived cells divide seldom, if at all. In contrast, the cells that are involved in producing the gut lining grow and divide at top speed. Most human cells differentiate
to carry out specific functions and then no longer proliferate. How do cells decide whether to proliferate, to stop proliferating and differentiate, or to die? This section answers these and other questions. Chapter 40 begins the section with an introduction to the language of the cell cycle. The cell cycle is driven by changing states of the cytoplasm created by shifting balances of protein phosphorylation, dephosphorylation, and degradation machinery. For the cell cycle, the key kinases are cyclin-dependent kinases (Cdks), which require an associated cyclin subunit for activity. Cdks are also regulated by phosphorylation and by additional protein cofactors that bind and inactivate them. Cdks are usually stable, but cyclin levels fluctuate, owing to targeted destruction at particular points in the cell cycle. In fact, targeted proteolytic destruction by the proteasome is a key aspect of cell-cycle control. Each cell-cycle phase is characterized by the activity of one or more E3 ubiquitin ligases. Each of these targets particular proteins for destruction by decorating them with chains of ubiquitin, a protein that was introduced
Mitosis Ch 44
Meiosis Ch 45
G2 phase and control of entry into mitosis Ch 43
S phase and DNA replication Ch 42
Programmed cell death Ch 46
Active CDK Introduction to the cell cycle Ch 40
Normal cell arrested at restriction point STOP
G1 phase and the regulation of cell proliferation Ch 41
Rb
Cancer cell passes restriction point Rb
GO
Danger
695
in Chapter 23. This sequential destruction of key factors gives cell-cycle transitions their irreversible character. The chapters that follow explain how the cell-cycle machinery controls each step in the proliferation and differentiation of cells. Chapter 41 begins with newly born cells in the G1 phase of the cell cycle. These cells must decide whether to commit themselves to a round of proliferation or to withdraw from the proliferation rat race and enter a nondividing differentiated state called G0. Cells that will proliferate must first pass a control point known as the restriction point. This is a control circuit that determines whether internal and external conditions are suitable for proliferation. Malfunctions of the restriction point lead to one of the most terrifying perturbations of the cell cycle: cancer. If DNA damage is detected during G1 phase, a checkpoint halts cell cycle progression until either the damage is repaired or the cell dies. The chapter includes a section on stem cells and concludes by considering the role of one of the most famous cell-cycle proteins, p53, in cell-cycle control. Cells that decide to proliferate must replicate their DNA in a timely and accurate manner. Chapter 42 explains the mechanism of DNA replication during the S phase, including the selection of sites on DNA to initiate replication, the enzymes that copy the DNA, the regulation of replication by the cell-cycle machinery, the organization of replicating chromosomes within the nucleus, and the checkpoints that help the cells to cope with various problems that they encounter along the way. Chapter 43 discusses the G2 phase, during which cells conduct a final “cockpit check” before embarking on the irreversible process of division. This is also the last point in the cell cycle at which the genome is scanned for damage so that it can be repaired before division. A checkpoint restrains cells from entering into mitosis if damaged DNA is detected, and the chapter briefly explains the major pathways of DNA repair. Chapter 44 describes mitosis, certainly the most dramatic and complex program in the cell cycle. Mitosis
696
has been studied since the 1800s, but technical advances have considerably advanced our understanding of how it is accomplished at the molecular level. Division requires wholesale reorganization of cellular structures, including chromosome condensation and the assembly of the mitotic spindle. In many cells, the nuclear envelope breaks down. Once the chromosomes are all attached to the microtubules of the mitotic spindle (yet another important checkpoint here), they are separated equally and form two daughter nuclei. Finally, cytokinesis separates the two daughter cells. Chapter 45 considers meiosis, a specialized form of division that produces the gametes required for sexual reproduction. In this division, DNA recombination is key to segregation of the chromosomes. A number of arcane terms are used to describe the specialized structures and processes involved. The chapter then explains how problems with meiosis can lead to genetic diseases and how studies of chromosome segregation in yeast led to an understanding of why birth defects become more prevalent as human mothers age. Chapter 46 closes the book with a discussion of what happens when cells commit suicide by apoptosis, necroptosis, and autophagy. This is not, strictly speaking, a cell-cycle event but instead represents several alternative pathways, each with its own machinery and signaling systems. Apoptosis sometimes results when it all “runs off the rails” and cells receive insults from which they cannot recover. But cell death is not always bad: apoptosis is an essential part of development of metazoan organisms, homeostasis of their organs and tissues, and can be a last-ditch defense against viral infection. Malfunctions of apoptotic pathways can lead to cancer. The concepts that are discussed in this section of the book build on the ideas in earlier sections. Cells are wonderfully complex systems whose behavior is driven by the laws of chemistry and physics. A major challenge for cell biology in the future is to devise molecular explanations for the complex behaviors exhibited in this closing section of our book.
CHAPTER
40
Introduction to the Cell Cycle The cell cycle is the series of events that leads to the
duplication and division of a cell. Research on the molecular events of cell-cycle control revealed that variations of similar mechanisms operate the cell cycles of all eukaryotes from yeasts to humans. Furthermore, the components that regulate cell growth and division also play key roles in the cessation of cell division that is required for cells to differentiate. Control of the cell cycle is of major importance to human health because cancer is usually caused by perturbations of cell-cycle regulation. Based on 2010 to 2012 data, 40% of Americans will develop cancer during their lifetime. Although animal cells have a wide variety of specialized cell cycles, the cells in the stratified epithelium that forms skin illustrate the most common types of cell cycles (Fig. 40.1). The basal layer of the epithelium is composed of stem cells that divide only occasionally (see Box 41.2). They can activate the cell cycle on demand and then return to a nondividing state. Stem cell populations can replenish themselves by symmetrical division, but when specific signals induce them to proliferate, usually one daughter cell remains a stem cell and the other enters a pool of rapidly dividing cells. The dividing cells populate the upper layers of the epithelium, stop dividing, and gradually differentiate into the specialized cells that cover the surface. Like stem cells, fibroblasts of the connective tissue (see Fig. 28.2) typically are in a nondividing state, but they can be stimulated to enter the cell cycle following wounding or other stimuli (see Fig. 32.11). In the most extreme case, the nervous system contains a few stem cells and a few dividing glial cells, but most neurons, once differentiated, can live for more than 100 years without dividing again.
Principles of Cell-Cycle Regulation The goal of the cell cycle in most cases is to produce two daughter cells that are accurate copies of the parent
(Fig. 40.2). The cell cycle integrates a continuous growth cycle (an increase in cell mass) with a discontinuous division or chromosome cycle (the replication and partitioning of the genome into two daughter cells). The chromosome cycle is driven by a sequence of enzymatic cascades that produce a sequence of discrete biochemical “states” of the cytoplasm. Progress through the cell cycle is ratchet-like and irreversible because each new state arises not only by expression or activation of a new cohort of activities, but also by destruction or inactivation of key activities characteristic of the preceding state. Later sections of this chapter explain these mechanisms.
Phases of the Cell Cycle In describing the cell cycle, it is convenient to divide the process into several phases. Recognition of these phases began in 1882, when Flemming named the process of nuclear division mitosis (from the Greek mito, or “thread”) after he first observed the condensed chromosomes. Mitosis was a clear cell cycle landmark, and the rest of the cell cycle between mitoses was called interphase (Box 40.1). Once DNA was recognized as the agent of heredity in the 1940s, it was deduced that it must be duplicated at some time during interphase so that daughter cells can each receive a full complement of genetic material. In 1953, a key experiment identified the relationship between the timing of DNA synthesis and the mitotic cycle (Fig. 40.3). This defined the four cell-cycle phases as they are known today (see Fig. 40.2). Each cell is born at the completion of the M phase, which includes mitosis, the partitioning of the chromosomes and other cellular components, and cytokinesis, the division of the cytoplasm. The chromosomal DNA is replicated during S phase (synthetic phase). The remaining two phases are gaps between mitosis and the S phase. The G1 phase (first gap phase) is the interval 697
698
SECTION X n Cell Cycle
Death
Final stage of differentiation in skin
Cessation of cycling Terminal differentiation
Rapid cell cycles Expansion of population
Infrequent cell cycles Activation
A
B
Renewal of stem cell population
Stem cell
FIGURE 40.1 CELL CYCLES IN A STRATIFIED EPITHELIUM. A, Light micrograph of a section of skin, a stratified squamous epithelium, stained with hematoxylin and eosin (H&E). B, Diagram showing the different types of cell cycles at the various levels of this epithelium.
A. Cell-cycle details (not to scale)
B. Cycle phases in cultured cell
Mitosis M
Check for damaged or unduplicated DNA
Check for chromosome attachment to mitotic spindle
Cytokinesis
ENLARGED VIEW OF CHROMOSOME
G1
10
S
7.5
G2
3.5
M
1.0 22
G0
Growth G 1 in mass Cohesion established in S phase
Length (hours)
Generation time
DNA
G2
Cell-cycle phase
C. Time-scaled diagram (times in hours) 21
22 0
M Check for DNA damage
G2 17.5
G1 Check for DNA damage or stalled replication forks
S Chromosome duplication
Restriction point: pass only if environmental conditions favorable Centrosome duplication starts
S 10
FIGURE 40.2 INTRODUCTION TO THE CELL-CYCLE PHASES. A, Diagrams of cellular morphology and chromosome structure across the cell cycle. B, Length of cell-cycle phases in cultured cells. C, Time scale of cell-cycle phases.
CHAPTER 40 n Introduction to the Cell Cycle
699
BOX 40.1 Selected Key Terms M phase: Cell division, comprising mitosis, when a fully grown cell segregates the replicated chromosomes to opposite ends of a molecular scaffold, termed the spindle, and cytokinesis, when the cell cleaves between the separated chromosomes to produce two daughter cells. In general, each daughter cell receives a complement of genetic material and organelles identical to that of the parent cell. Interphase: The portion of the cell cycle when cells grow and replicate their DNA. Interphase has three sections. The G1 (first gap) phase is the interval between mitosis and the onset of DNA replication. The S (synthetic) phase is the time when DNA is replicated. The G2 (second gap) phase is the interval between the termination of DNA replication and the onset of mitosis. In multicellular organisms, many differentiated cells no longer actively divide. These nondividing cells (which may physiologically be extremely active) are in the G0 phase, a branch of the G1 phase. Checkpoints: Biochemical circuits that regulate cell-cycle transitions in response to the physiological condition of the cell and signals from its environment. Checkpoints detect damage to the DNA due to external agents or problems that arise during DNA replication and trigger the DNA damage response. Other checkpoints detect problems that arise during attachment of chromosomes to the spindle.
between mitosis and the start of DNA replication. The G2 phase (second gap phase) is the interval between the completion of DNA replication and mitosis. All cycling cells have an M phase and an S phase. The G1 and G2 phases vary in length and are very short in some early embryos. The following sections describe the stages of the cell cycle, starting just after the birth of the cell.
G1 Phase G1 is typically the longest and most variable cell-cycle phase. When cells are “born” at cytokinesis, they are roughly half the size they were before mitosis, and during G1, they grow back toward an optimal size. During this time, many activities involved in cell-cycle progression are repressed so that the cell cannot initiate a new round of proliferation. This repressive control system is called the restriction point. If the supply of nutrients is poor or if cells receive an antiproliferative stimulus such as a signal to embark on terminal differentiation, they delay their progress through the cell cycle in G1 or exit the cycle to enter G0 (see “G0 and Growth Control” below). However, if appropriate positive stimuli are received, cells overcome the restriction point block and trigger a program of gene expression that commits them to a new cycle of DNA replication and cell division. Cancer cells often have defects in restriction point
Nucleus
Cell synthesizing DNA
Add 32P, incubate briefly, then wash out free 32P
Photographic emulsion
Cell exposes photographic emulsion VIEW IN MICROSCOPE
20% of cells turn the emulsion black FIGURE 40.3 DISCOVERY OF CELL-CYCLE STAGES. To determine whether cells synthesize DNA during a defined portion of the cell cycle or constantly throughout the entire cycle (as is the case in bacteria, for example), Howard and Pelc fed a radioactive component of DNA (32P) to onion root tip cells, spread the cells in a thin layer on a microscope slide, washed away the 32P that had not become incorporated into DNA, and overlayered the slide with photographic emulsion. After incubation in the dark, the emulsion was developed like film and examined with a light microscope. The nuclei of cells active in replicating their DNA incorporated the radioactive 32P into DNA and exposed the photographic emulsion above them. Two possible outcomes were predicted. If cells synthesized DNA constantly during interphase, then all cells would incorporate the radioactive label. Conversely, if each cell synthesized DNA only during a discrete portion of the cell cycle, then only cells engaged in active replication during the period of exposure to 32P would expose the photographic emulsion. When the slides were examined, 20% of the interphase cell nuclei were labeled, proving that cells synthesize DNA only during a discrete portion of interphase. Mitotic cells were unlabeled. Assuming that the cells traverse the cycle at a more or less constant rate, it was possible to calculate the length of the synthetic phase. Overall, the time between successive divisions—the generation time—was approximately 30 hours in the root tip cells. If approximately 20% of the cells were labeled, then approximately 20% of the 30-hour generation time must be spent in DNA synthesis. Thus, 0.2 × 30, or 6 hours, was spent in replication. (Data from Pelc HA Sr. Synthesis of DNA in normal and irradiated cells and its relation to chromosome breakage. Heredity Suppl. 1953;6:261–273.)
700
SECTION X n Cell Cycle
control and continue to grow and attempt to divide even in the absence of appropriate environmental signals.
G0 and Growth Control Most cells of multicellular organisms differentiate to carry out specialized functions and no longer divide. Such cells are considered to be in the G0 phase. Cells often enter G0 directly as they exit their last mitosis. G0 cells are not dormant; indeed, they are often actively engaged in protein synthesis and secretion, and they may be highly motile. Many G0 cells have a nonmotile primary cilium, which is an important sensory organelle (see Fig. 38.19). The G0 phase is not necessarily permanent. In some specialized cases, G0 cells may be recruited to reenter the cell cycle in response to specific stimuli. Cell-cycle reentry involves changes in gene expression and protein stability and disassembly of the primary cilium, if present. This process must be highly regulated, as the uncontrolled proliferation of cells in a multicellular organism can lead to cancer. S Phase Chromosomes of higher eukaryotes are so large that replication of the DNA must be initiated at many different sites, termed origins of replication. In budding yeast, the approximately 400 origins are spaced an average of 30,000 base pairs apart. An average human chromosome contains about 150 × 106 base pairs of DNA, approximately 10 times the size of the entire budding yeast genome, so many more origins are required. Each region of the chromosome that is replicated from a single origin is referred to as a replicon. Groups of neighboring replicons cluster in topologically associating domains (TADs) (see Chapter 8). Proliferating diploid cells replicate their DNA once, and only once, each cell cycle. Each origin of replication is prepared for replication by the formation of a prereplication complex during G1 (a process that is referred to as licensing). As each origin “fires” during S phase, the prereplication complex is dismantled and cannot be reassembled until the next G1 phase. This ensures that each origin fires only once per cell cycle. The cyclic nature of origin licensing is driven at least in part by fluctuations in the activity of cyclin-dependent kinases and protein destruction machinery (discussed later). During replication, the duplicated DNA molecules, called sister chromatids, become linked to each other by a protein complex called cohesin (see Fig. 8.18). This pairing of sister chromatids is important for their orderly segregation later in mitosis (see Fig. 44.16). G2 Phase In most cells of metazoans, G2 is a relatively brief period during which key enzymatic activities that will trigger the entry into mitosis gradually accumulate and are converted to active forms. When these activities reach a
critical threshold level, the cell enters mitosis. Along the way, the chromatin and cytoskeleton are prepared for the dramatic structural changes that will occur during mitosis. If damaged DNA is detected during G2, a checkpoint activates the DNA damage response and delays entry of the cell into mitosis.
M Phase During M phase (mitosis and the subsequent cytokinesis), chromosomes and cytoplasm are partitioned into two daughter cells. Mitosis is normally divided into five discrete phases. Prophase is defined by the onset of chromosome condensation and is actually the final part of G2 phase. TADs disassemble inside the intact nucleus and mitotic chromosomes begin to form their characteristic array of loops (see Chapter 8). In the cytoplasm, a dramatic change in the dynamic properties of the microtubules decreases their half-lives from approximately 10 minutes to approximately 30 seconds. The duplicated centrosomes (centrioles and associated pericentriolar material in animal cells; see Fig. 34.14) separate and form the two poles of the mitotic spindle. Prometaphase begins when the nuclear envelope breaks down (in higher eukaryotes) and chromosomes begin to attach randomly to microtubules emanating from the two poles of the forming mitotic spindle. Other microtubules originate on chromosomes and within the mitotic spindle. As both kinetochores on a pair of sister chromatids attach to microtubules from opposite spindle poles, the pair of chromatids slowly moves to a point midway between the poles. When all chromosomes are properly attached, the cell is said to be in metaphase. The exit from mitosis begins at anaphase with abrupt separation of the two sister chromatids from one another. Most cohesion molecules linking sister chromatids are removed without cleavage during prophase in a process initiated by mitosis-specific phosphorylation. Proteolytic cleavage of the remaining cohesin molecules triggers the metaphase-anaphase transition. During anaphase, the separated sister chromatids move to the two spindle poles (anaphase A), which themselves move apart (anaphase B). As the chromatids approach the spindle poles, the nuclear envelope reforms on the surface of the chromatin. At this point, the cell is said to be in telophase. Finally, during telophase, a contractile ring of actin and myosin assembles as a circumferential belt at the cortex midway between spindle poles and constricts the equator of the cell. The separation of the two daughter cells from one another is called cytokinesis.
Control of Cell-Cycle Progression Control networks and checkpoints regulate progression of the cell cycle. Checkpoints are biochemical
CHAPTER 40 n Introduction to the Cell Cycle
circuits that detect external or internal stimuli and send appropriate signals to the cell-cycle system. The restriction point in G1 phase is a control network that integrates the physiological state of the cell with its environment, including input from other cells and interactions with the surrounding extracellular matrix. Cells must receive appropriate growth stimuli from their environment to progress past this point in the G1 phase; if not they may live on without dying or commit suicide by apoptosis (see Chapter 46). DNA damage checkpoints operate throughout interphase. If damage is detected, the DNA damage response initiates a cascade of events that blocks cellcycle progression and can also trigger cell death by apoptosis. Problems with DNA replication generally produce single-stranded DNA and activate the DNA damage response. This response stabilizes stalled replication forks so that they can be repaired. During mitosis, the spindle assembly checkpoint delays the onset of chromosome segregation until all chromosomes are attached properly to the mitotic spindle. The DNA damage response regulates cell-cycle progression in a three-tier pathway (Fig. 40.4). First sensors detect DNA damage. These sensors activate transducers, which include both protein kinases and transcriptional activators. The transducers act on effectors that ultimately block cell-cycle progression and may also fulfill other functions. Two key protein kinases, ataxiatelangiectasia mutated (ATM) and ataxia-telangiectasia and Rad9 related (ATR), lie at the head of the pathway and may also act as sensors of DNA damage. They activate two transducer kinases, Chk1 and Chk2, as well as a transcription factor called p53 that induces the expression of a cohort of genes that halt cell-cycle progression by inhibiting cyclin-dependent kinases as well as genes
Problems at replication forks
DNA breaks
ATM dimer
ATM monomer
Chk2 kinase
ATR + cofactors bind ssDNA
p53 transcription factor
Chk1 kinase
Cell cycle proteins DNA repair proteins Apoptosis
Cell-cycle arrest
DNA repair
Genotoxic stress
Sensors
Transducers
Effectors Responses
FIGURE 40.4 ELEMENTS OF THE DNA DAMAGE CHECKPOINT AND RESPONSE SYSTEM. ATM, ataxia-telangiectasia mutated; ATR, ataxia-telangiectasia and Rad9 related; ssDNA, singlestranded DNA.
701
that trigger cell death by apoptosis. Chapters 41 and 43 discuss these proteins in detail.
Biochemical Basis of Cell-Cycle Transitions Transitions between cell-cycle phases are triggered by a network of protein kinases and phosphatases that is linked to the discontinuous events of the chromosome cycle by the periodic accumulation, modification, and destruction of several key components. This section provides a general introduction to the most important components of this network.
Cyclin-Dependent Kinases Genetic analysis of the cell cycle in the fission yeast Schizosaccharomyces pombe identified a gene called cell division cycle–2+ (cdc2+) that is essential for cellcycle progression during both the G1 → S and G2 → M transitions (Box 40.2). The product of this gene, a protein kinase of 34,000 Da originally called p34cdc2, is the prototype for a family of protein kinases that is crucial for cell-cycle progression in all eukaryotes. This mechanism of cell-cycle control is so well conserved that a human homolog of p34cdc2 can replace the yeast protein, restoring a normal cell cycle to a cdc2 mutant yeast. Boxes 40.3 and 40.4 present a number of the key experiments and experimental systems that led to the identification of the molecules that drive the cell cycle. Humans have more than 10 distinct protein kinases related to p34cdc2, although only a few are involved in cell-cycle control. To be active, each enzyme must associate with a regulatory subunit called a cyclin. Thus, they have been termed cyclin-dependent kinases (Cdks). p34cdc2, now termed Cdk1, seems to function primarily in the regulation of the G2 → M transition in animal cells. Cdk2 (plus Cdk4 and Cdk6 in some cell types) is involved in passage of the restriction point during G1. Cdk2 also contributes to the G2 → M transition, although Cdk1 is the only Cdk absolutely essential for this step (Appendix 40.1). Cdk7 is important for activation of other Cdks, and also appears to participate in transcribing RNA and repairing damaged DNA. Other Cdks participate in diverse processes ranging from transcriptional regulation to neuronal differentiation. Surprisingly, fibroblasts from mice that lack Cdk2, Cdk4, or Cdk6 are viable; other Cdks can drive the cell cycle if necessary. The mice suffer developmental difficulties because those genes are needed for the differentiation of particular cell types. Cyclins Cdks require cyclin binding for catalytic activity (Fig. 40.13). Cyclins were discovered in rapidly dividing invertebrate embryos as proteins that accumulate gradually during interphase and are abruptly destroyed during mitosis (see Fig. 40.11). This process of cyclic accumulation and destruction inspired their name.
702
SECTION X n Cell Cycle
BOX 40.2 Use of Genetics to Study the Cell Cycle Studies of the distantly related budding and fission yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe (see Fig. 2.8) were important for understanding the cell cycle for several reasons. First, the proteins that control the cell cycle are remarkably conserved between yeasts and mammals. Second, both yeast genomes are small, simplifying the discovery of important gene products. Third, genetic analysis is straightforward, as both yeasts can grow as haploids, and both efficiently incorporate cloned DNA into their chromosomes by homologous recombination. These two yeasts evolved very different strategies for cell division. Budding yeasts divide by assembling a single bud on the surface of the cell every cell cycle. Fission yeasts divide by fission across the center of an elongated cell. A useful feature of using yeast to study the cell cycle is that the stage of the cell cycle is revealed by the cellular morphology in the light microscope. For budding yeast, unbudded cells are in G1, cells with buds smaller than the mother cell are in S phase, and cells whose buds are similar in size to the mother cell are in G2 or M. For fission yeast, cell length provides a yardstick for estimating cell-cycle position. The cell cycles of both yeasts differ from those of animal cells. In budding yeast, much of the 90-minute cell cycle is spent in G1. Thus, the networks controlling the G1 → S transition are particularly amenable to study. In contrast, a fission yeast spends most of its 2-hour cell cycle in G2. S phase follows separation of sister chromatids and occurs prior to cytokinesis. Thus, the control of the G2 → M transition is readily studied in fission yeast. During mitosis, the nuclear envelopes of both yeasts remain intact, so chromosomes segregate on a spindle inside the nucleus. Genetic studies revealed that the yeast cell cycle is a dependent pathway whereby events in the cycle occur normally only after earlier processes are completed. The cell cycle can be modeled as a line of dominoes, each domino corresponding to the action of a gene product that is essential for cell-cycle progression (Fig. 40.5) and the nth domino Model of the cell cycle as a simple dependent pathway
falling only when knocked down by the (n-1)th domino. According to the model, mutations in genes that are essential for cell-cycle progression cause an entire culture of yeast to accumulate at a single point in the cell cycle (the point at which the defective gene product first becomes essential). This is referred to as the arrest point. Fig. 40.5 shows this by including a “mutant” domino that does not fall over when struck by the upstream domino. Mutants that meet this criterion are called cell division cycle mutants or CDC mutants. Genetic screens for CDC mutants have identified many important genes involved in cell-cycle control. Because CDC genes are essential for cell-cycle progression, it is impossible to propagate strains of yeast carrying CDC mutants unless the mutants have a conditional lethal phenotype. The most commonly used conditional lethal mutations are temperature sensitive (ts). Many yeast temperature-sensitive mutants are viable at 23°C (the permissive temperature), but cease dividing at 36°C (the restrictive temperature). Temperature-sensitive proteins often have an altered amino acid sequence, but occasionally, the lack of a gene product can cause a ts phenotype. More recently, the use of auxin-inducible degrons (see Chapters 6 and 23) has enabled experimenters to study the consequences of depleting an essential protein from yeast in a matter of minutes. Fission yeasts with CDC mutants affecting the entry into mitosis have distinctive morphologies. Cells mutant in Wee1 (a kinase that keeps cyclin-dependent kinase–1 [Cdk1] inactive prior to mitosis) enter mitosis prematurely and are shorter than normal (Fig. 40.6B). In contrast, cells lacking Cdc25 (a phosphatase that counteracts Wee1 and activates Cdk1) are unable to undergo mitosis but continue their growth cycle, therefore becoming greatly elongated (Fig. 40.6C). This simple morphologic assay allowed straightforward classification of yeast CDC genes into those that stimulate progression through mitosis and those that retard entry into mitosis. A. Wild type
B. Wee1 mutant C. Cdc25 mutant
Wild type
CDC mutant
FIGURE 40.5 THE CELL CYCLE MAY BE MODELED AS A SIMPLE DEPENDENT PATHWAY. A cell division cycle (CDC) mutation can block further progression along the pathway at a characteristic point in the cell cycle.
FIGURE 40.6 FLUORESCENCE MICROGRAPHS OF FISSION YEAST CELLS ILLUSTRATING PHENOTYPES OF CELL-CYCLE MUTATIONS. Cell walls and nuclei are stained. A, Wild-type cells. B, A wee1 mutation that accelerates entry into mitosis at the restrictive temperature. C, A cdc25 mutation that delays entry into mitosis at the restrictive temperature. (Courtesy H. Ohkura, Wellcome Trust Institute for Cell Biology, University of Edinburgh, United Kingdom.)
CHAPTER 40 n Introduction to the Cell Cycle
703
BOX 40.3 Studies of the Cell Cycle in Vitro Amphibian oocytes and eggs are storehouses of most components needed for cell-cycle progression. Oocytes are arrested in G2 until a surge of the hormone progesterone causes them to “mature” into eggs, which are then naturally arrested in metaphase of the second meiotic division (see Chapter 45). After fertilization, the embryo of the South African clawed frog (Xenopus laevis) undergoes a rapid burst of cell divisions. An initial cell cycle 90 minutes long is followed by a rapid succession of 11 cleavages spaced only 30 minutes apart to produce an embryo of 4096 cells (Fig. 40.7). Thirty minutes per cycle is insufficient to transcribe and translate all the genes needed to make the daughter cells that are produced at each division. The frog solves this problem 1 Cleavage 90 minutes
A
11 Cleavages 30 minutes apart
B
C Somatic cell enlarged ×10
FIGURE 40.7 CLEAVAGES SUBDIVIDE THE EGG DURING XENOPUS EARLY DEVELOPMENT. A, Fertilized egg. B, Two-cell stage. C, Multicellular embryo. Compare size of somatic cell and egg.
A. Tightly packed eggs
B. Eggs crushed by centrifugation
C. Added sperm nucleus with membrane removed
by making oocytes extremely large (~500,000 times the volume of a typical somatic cell) and storing within them vast stockpiles of the structural components needed to make cells. As a result, only DNA and a very few proteins need be synthesized during early embryonic divisions. In addition to structural components, many factors that regulate normal cell-cycle progression are also stockpiled in oocytes. These features make Xenopus oocytes an excellent source of material for biochemical analysis of the cell cycle. Remarkably, it is possible to make cell-free extracts from Xenopus eggs that progress through the cell cycle in vitro (Fig. 40.8). Nuclei from G1 cells or haploid sperm nuclei, when added to these extracts, efficiently replicate their DNA and proceed through the cell cycle into mitosis, complete with chromosome condensation, nuclear envelope breakdown, chromosome alignment on a spindle, and anaphase segregation of sister chromatids without any additions to the tube. Because these events occur in a cell-free milieu, they are readily accessible to biochemical manipulation. For example, antibodies and other proteins can be added to the extracts, and their effect on the cell cycle can readily be determined. Thus, the Xenopus extract system offers a powerful tool for testing the role of various proteins in the cell cycle in higher eukaryotes. D. Reassembly of nucleus
E. DNA replication
F. Mitosis
Lipid Centrifuge hard Xenopus eggs
Extract
Pellet FIGURE 40.8 USE OF XENOPUS EGG EXTRACTS TO STUDY THE CELL CYCLE. A–B, Making a Xenopus egg extract that is competent to carry out cell-cycle oscillations in vitro. C–F, A cycling Xenopus extract undergoes alternating S and M phases. G1 and G2 phases are minimal (as they are during early development of the frog).
Humans have at least 16 different cyclin proteins that range in size from 35 to 130 kD. The highly conserved cyclin box domain, which docks with the Cdk partners, is the defining structural feature of these proteins. Only a handful of cyclin isoforms are involved in cell-cycle control. Of those that are, some function during G1 phase, others during G2 phase, and still others during M phase.
Positive Regulation of Cyclin-Dependent Kinase Structure and Function Cdks monomers are intrinsically inactive, so they depend on activation by cyclins and are regulated by positive and
negative controls. Like other eukaryotic protein kinases (see Fig. 25.3), Cdks have a bilobed structure with the active site in a deep cleft between a small N-terminal and larger C-terminal domain. Monomeric Cdks have a flexible T loop that blocks the mouth of the catalytic pocket. In addition, a short α-helix is oriented such that a glutamic acid required for adenosine triphosphate (ATP) hydrolysis points away from the catalytic cleft. As a result, ATP bound by the monomeric kinase cannot transfer its α-phosphate to protein substrates (see Fig. 40.13A). Several different mechanisms regulate Cdk activity (Fig. 40.14). On one hand, cyclin binding and
704
SECTION X n Cell Cycle
BOX 40.4 Discovery of Factors Essential for Cell-Cycle Progression The best early evidence for the existence of positive inducers of cell-cycle transitions in mammals was obtained in cell fusion experiments. When cultured cells in S phase were fused with cells in G1, the G1 nuclei initiated DNA replication shortly thereafter. In contrast, if S phase cells were fused with G2 cells, the G2 nuclei did not rereplicate their DNA until after passing through mitosis. The most dramatic results were obtained when mitotic cells were fused with interphase cells. This caused the interphase cells to enter into mitosis abruptly (as judged by nuclear envelope breakdown and chromosome condensation). The phenomenon was termed premature chromosome condensation (PCC). The mitotic inducer could work in any cell-cycle phase (Fig. 40.9). If mitotic cells were fused with cells in G1 phase, interphase chromosomes condensed into long, single filaments. If the interphase cell was in G2 phase, the duplicated chromosomes appeared as double filaments. If the interphase cell was in the S phase, the partially replicated chromosomes condensed into a complex pattern of single and double condensed regions separated by regions of decondensed chromatin corresponding to sites where DNA was actively replicating at the time of fusion. Working independently, developmental biologists who were interested in the control of cell division during early development in frogs also discovered an activity that could cause interphase cells to enter the M phase. They used a micropipette to extract a tiny bit of cytoplasm from a mature egg that was arrested in metaphase of meiosis II and inject it into oocytes (which are in G2 phase). The oocytes rapidly entered M phase, with concomitant chromosome condensation and nuclear envelope disassembly (Fig. 40.10). This stimulation to enter M phase is called maturation, and the unknown factor present in the egg cytoplasm that induced oocyte maturation was termed MPF, or maturation-promoting factor (now often referred to as M phase–promoting factor). It was realized early on that
A. M–G1 fusion
MPF might be related to the inducer of mitosis detected in the PCC experiments. In fact, extracts from mitotic tissue culture cells could induce meiotic maturation when injected into oocytes. Similar extracts from cells in other phases of the cell cycle did not cause the G2/M phase transition in oocytes. Other cell biologists studying protein synthesis in starfish and sea urchin embryos noticed a curious protein that seemed to accumulate across the cell cycle but was then destroyed during mitosis. They were well aware of the work on MPF, and immediately suspected that their protein, which they called cyclin, might be somehow involved in MPF activity (Fig. 40.11). In a third line of investigation, geneticists working on yeasts realized that the cell cycle could be dissected through the isolation of cell division cycle (CDC) mutants (see Box 40.2). The analysis of the cell cycle with these mutants dominated cell-cycle research to such an extent that many human genes that are important in cell-cycle control bear the CDC name if they are related to well-characterized yeast genes. The best-known genes to emerge from this analysis were Cdc2 (Cdk1) and Cdc25, both of which were determined genetically to encode proteins that actively promote the G2/M transition. Other genes, such as Wee1, were found to encode activities that act as antagonists that inhibit the G2/M transition. When eventually purified from Xenopus eggs (Fig. 40.12), active MPF consisted primarily of two polypeptides of 32,000 and 45,000 Da. The smaller component of MPF is the Xenopus equivalent of the fission yeast Cdc2 gene product (now known as Cdk1). The larger component of Xenopus MPF is a B-type cyclin. Only 15 years later was it recognized that fully functional MPF also requires Greatwall kinase and its small substrates to inhibit protein phosphatase PP2AB55δ and give Cdk activity a chance to take off and trigger mitotic entry.
B. M–S fusion
C. M–G2 fusion
5 µm
FIGURE 40.9 FUSION WITH MITOTIC CELLS CAUSES INTERPHASE CELLS TO ENTER MITOSIS PREMATURELY, NO MATTER WHERE THEY ARE IN THE CELL CYCLE. The resulting prematurely condensed chromosomes are single threads if the interphase cell was in G1 phase (A), double threads if the cell was in G2 phase (C) and a complex mixture of both interspersed with uncondensed regions if the cell was in S phase (B). M, mitosis. (From Hanks SK, Gollin SM, Rao PN, et al. Cell cycle-specific changes in the ultrastructural organization of prematurely condensed chromosomes. Chromosoma. 1983;88:333–342.)
705
CHAPTER 40 n Introduction to the Cell Cycle
BOX 40.4 Discovery of Factors Essential for Cell-Cycle Progression—cont’d
A
B
Meiotic spindle
Nucleus
C
Nuclear disassembly and chromosome condensation
Suck out Inject it cytoplasm into oocyte Egg
D Meiotic spindle
Oocyte enters M phase
Fully grown oocyte
Egg
FIGURE 40.10 THE EXPERIMENT THAT IDENTIFIED MATURATION-PROMOTING FACTOR. A, The box shows the meiotic spindle in a Xenopus egg arrested in metaphase II of meiosis. B, The box shows the interphase nucleus in a mature oocyte. Following injection of MPF, the nucleus disassembles, mitotic chromosomes form (C), and the cell assembles a meiotic spindle (D). Disassembly of the oocyte nucleus and entry into M phase is called maturation, and the factor triggering this event was named maturation-promoting factor (MPF).
A B
97
Molecular weight (kDa)
Intensity of bands A and B
75
A. SDS gel of proteins in purified MPF
B 50
A
67
43
Cyclin B
25
30
Cleavage index
Induction of M phase (%) 0
0 0
1
2
Time (hours)
FIGURE 40.11 EXPERIMENTAL IDENTIFICATION OF A CYCLIN. Newly synthesized proteins (labeled with 35S-methionine) in fertilized sea urchin eggs were separated by sodium dodecylsulfate (SDS) polyacrylamide gel electrophoresis. It was noted that the protein labeled A (which was named cyclin) first accumulated, was greatly reduced at the metaphase/anaphase transition, and then began to accumulate again. Protein B, which is not involved in cell-cycle regulation, accumulated progressively over this time. “Cleavage index” refers to the percentage of dividing cells observed in the microscope at varying times after fertilization. (From Evans T, Rosenthal ET, Youngblom J, et al. Cyclin: a protein specified by maternal mRNA in sea urchin eggs that is destroyed at each cleavage division. Cell. 1983;33:389–396.)
phosphorylation of the T loop stimulate enzyme activity. On the other, Cdks are inhibited by phosphorylation of residues adjacent to the ATP-binding site and binding of inhibitory proteins. Cyclin binding causes the T loop to retract away from the mouth of the catalytic pocket (see Fig. 40.13B). In addition, the secondary structure of the N-terminal
Cdk 1 0
0
0
50 85 25
0
0
0
B. H1 kinase activity Fraction number 6
9
12
15
FIGURE 40.12 PURIFICATION OF MATURATION-PROMOTING FACTOR (MPF). A, Sodium dodecylsulfate (SDS) polyacrylamide gel electrophoresis of fractions from the final step of the purification. The numbers at the bottom show the percentage of oocytes that entered M phase when a portion of each column fraction was injected (the classical MPF assay). The roughly 32-kD band is Cdk1 (p34cdc2). The roughly 45-kD band is cyclin B. B, Assay of the ability of the column fractions to phosphorylate histone Hl. This is a standard assay for active Cdk enzymes. (Modified from Lohka M, Hayes MK, Maller JL. Purification of maturation-promoting factor, an intracellular regulator of early mitotic events. Proc Natl Acad Sci U S A. 1988;85:3009–3013.)
domain is altered, allowing the bound ATP to assume a conformation suitable for reaction with substrates. Despite these changes, the Cdk–cyclin complex has low catalytic activity. Full activation of most Cdks requires a kinase called Cdk-activating kinase (CAK), which phosphorylates a threonine in the T loop of the Cdk (this threonine gives the loop its name). In
706
SECTION X n Cell Cycle A. Cdk2
B. Cdk2–cyclin A
N lobe PSTAIR helix
C. Active Cdk2–cyclin A
Active site ATP
T loop C lobe
Cyclin A
Phosphorylation site
FIGURE 40.13 ATOMIC STRUCTURES OF CYCLIN-DEPENDENT KINASES. A, Cdk2. The PSTAIR helix, found in most Cdks, is named after a sequence of six amino acids that binds to cyclins (one letter code). (For reference, see Protein Data Bank [PDB; www.rcsb.org] file 1DM2.) B, Cdk2–cyclin A (kinase at basal activity level). (PDB file 1FIN.) C, Cdk2–cyclin (kinase fully active following phosphorylation of threonine160). (PDB file 1JST.) ATP, adenosine triphosphate.
vertebrates, CAK is composed of Cdk7-cyclin H. The phosphorylated threonine fits into a charged pocket on the surface of the enzyme, flattening the T loop back even farther from the mouth of the catalytic pocket (see Figs. 40.13C and 40.14A). This stimulates the catalytic activity up to 300-fold, in part because the flattened T loop forms part of the substrate-binding surface. Threonine phosphorylation also stabilizes the association of Cdk2 with cyclin A. In addition to their cyclin partner, Cdk1 and Cdk2 bind an additional small Cdc kinase subunit (Cks) protein to their C-terminal domain, away from the active site. Bound Cks enables the kinase to better hold onto its substrates and increases the efficiency with which Cdks can phosphorylate substrates at multiple sites (a hallmark of Cdk target phosphorylation).
Negative Regulation of Cyclin-Dependent Kinase Structure and Function At least two mechanisms slow or stop the cell cycle by inactivating Cdks (see Fig. 40.14). During G2 phase, the protein kinases Myt1 and Wee1 hold Cdk1 in check by phosphorylating threonine14 and tyrosine15 in the roof of the ATP-binding site. These phosphates interfere with ATP binding and hydrolysis. Threonine14 and tyrosine15 are accessible to the regulatory kinases only following cyclin binding, so this phosphorylation of Cdks depends, at least in part, on the availability of cyclins. In mammals, three Cdc25 phosphatases (see Fig. 25.5) reverse these inhibitory phosphorylations. Cdc25A regulates both the G1 → S and G2 → M transitions and is essential for life of the cell. Ccd25B is dispensable for mitosis, but it is essential for the production of gametes in meiosis. Cdc25C is a target of the G2 DNA damage checkpoint that prevents cells from undergoing mitosis with damaged DNA (see Fig. 43.11), but cells can survive without it. A parallel mechanism for inactivating Cdks involves binding of subunits from two families of small inhibitory
A. Kinase activation
B. Inactive forms
INK4
Cyclin
Cdk Cyclin cannot bind
Inactive kinase Cyclin
INK4
p27
Interference with ATP use
CAK phosphorylation ATP cannot bind Wee1 and Myt1 phosphorylation Cdc25 phosphatase dephosphorylation Active kinase
Interference with ATP use
FIGURE 40.14 POSITIVE AND NEGATIVE REGULATION OF CYCLIN-DEPENDENT KINASES. A, Pathway of activation by cyclin binding and phosphorylation. B, Pathways of inactivation by inhibitor binding and phosphorylation. When INK4 binds, twisting of the Cdk upper lobe blocks cyclin binding or interferes with ATP hydrolysis. When p27 binds, a loop insinuates into the upper lobe of the Cdk and blocks ATP binding. (For reference, see PDB files 1B17 [Cdk2-INK4], 1FIN and 1BI7 [Cdk2-INK4-cyclin], and 1JSU [Cdk2-p27-cyclin A].) CAK, Cdk-activating kinase.
proteins called the cyclin-dependent kinase inhibitors (CKIs) and inhibitors of Cdk4 (INK4) (for their names, see Appendix 40.1). When activated in the DNA damage response pathway, p53 turns on transcription of the CKI p21, which inhibits Cdk-cyclin A. CKI p27Kip1 inactivates complexes of Cdk2 and cyclin A by having a protein loop invade the N-terminal domain of the Cdk, disrupting its structure and competing with ATP for binding to the active site (see Fig. 40.14B). Members of the INK4 family preferentially inactivate Cdk4 and Cdk6 in two ways (see Fig. 40.14B). First, binding to monomeric Cdk distorts the orientation of the N- and C-terminal lobes, so cyclin D does not bind. INK4 family inhibitors also inhibit preformed Cdk4/6– cyclin D complexes by binding the Cdk and distorting the ATP-binding site so that the kinase uses ATP much less efficiently. Cdk inhibitors are important for growth regulation during the G1 and G0 phases of the cell cycle (see Chapter 41). They also play a critical role in the cell-cycle arrest that occurs in response to DNA damage and to antiproliferative signals. Mutations in the INK4 locus are strongly linked to cancer.
Role of Phosphatases in Counter-Balancing Cdk Activity Two important phosphatases counter-balance Cdk activity in mitosis. Just as Cdks exhibit cyclic behavior and are activated in mitosis, these counteracting phosphatases must also be cyclic—but they are inhibited in mitosis. Protein phosphatase 1 (PP1), associates with numerous targets on chromosomes and the mitotic apparatus, and is highly active during mitotic exit. Many target proteins have a simple loop with the sequence RVS/TF that inserts into a groove on the enzyme. During mitosis, Cdk phosphorylation of adjacent sites often blocks this interaction with substrates. Cdk1-cyclin B also phosphorylates the PP1 catalytic subunit during mitosis, thereby inactivating the enzyme. PP2A is more directly involved in Cdk regulation. It is a trimeric enzyme with catalytic, scaffolding, and regulatory subunits. The latter include B55α-δ and B56α-ε (see Appendix 40.1). PP2A-B55δ is largely responsible for removing phosphates added by Cdks. Consequently PP2A-B55δ must be inhibited to allow Cdks to drive the cell into mitosis. The Greatwall (Gwl) kinase regulates PP2A. Gwl is unusual, because 500 amino acids are inserted into its large lobe roughly adjacent to the T loop (see Fig. 40.13). Phosphorylation by Gwl allows two small proteins, Arpp19 (cyclic adenosine monophosphate [cAMP]-regulated phosphoprotein 19) and ENSA (α-endosulfine), to bind and inhibit PP2A-B55δ. Thus, Gwl confers the necessary cyclic behavior on PP2A. Understanding how Gwl is turned on and off is important for developing a full
CHAPTER 40 n Introduction to the Cell Cycle
707
picture of mitotic regulation, and this is a subject of active study.
Role of Protein Destruction in Cell-Cycle Control During mitosis active Cdk1–cyclin B–Cks phosphorylates key substrates leading to dramatic reorganization of the cell and, ultimately, to separation of sister chromatids on the mitotic spindle. Once chromatids are separated, the cell must return to a state with low levels of Cdk activity so that nuclear envelope reassembly, spindle disassembly, and cytokinesis can occur. Exit from mitosis requires Cdk inactivation by the ubiquitin-directed proteolytic machinery. The destruction of A- and B-type cyclins inactivates Cdk1 and Cdk2. This allows PP1, PP2A-B55δ, and other phosphatases to reverse the action of Cdks and bring mitosis to a close. Ubiquitylation also results in proteolysis of a protein called securin, which regulates the onset of sister chromatid separation at anaphase. Ubiquitin-mediated destruction of cyclins involves a cascade of three enzymes described in Chapter 23 (see Fig. 23.3). First, an E1 enzyme (ubiquitin-activating enzyme) activates the small protein ubiquitin by forming a thioester bond between the ubiquitin C-terminus and a cysteine on the enzyme. Activated ubiquitin is next transferred to another thioester bond on an E2 enzyme (ubiquitin-conjugating enzyme). The E2 often cooperates with an E3, which is important for imparting substrate specificity, to transfer ubiquitin to the ε-amino group of a lysine on a target protein. The resulting polyubiquitinated proteins are usually targets for destruction by the cylindrical 26S proteasome (see Fig. 23.8). This large multienzyme complex functions like a cytoplasmic garbage disposal, grinding target proteins down to short peptides and spitting out intact ubiquitin monomers for reuse in further rounds of protein degradation. Its role was originally thought to be the removal of damaged proteins from the cytoplasm; however, it is now recognized as a central factor in cellcycle control. The key E3 ligase regulating cyclin proteolysis is a large (15-subunit) complex called the anaphase-promoting complex/cyclosome (APC/C) (Fig. 40.15). The APC/C is inactive during the S and G2 phases of the cell cycle. Phosphorylation by Cdk1-cyclin B-Cks1 and binding of the protein coactivator Cdc20 activate the APC/C in early mitosis. APC/CCdc20 then triggers the metaphase– anaphase transition. An important checkpoint, the spindle assembly checkpoint, regulates APC/CCdc20 during mitosis, keeping it inactive until all kinetochores are productively attached to spindle microtubules. The checkpoint effector is the mitotic checkpoint complex whose formation is triggered by unattached kinetochores. This
708
SECTION X n Cell Cycle SCFSkp2, SCFβ–TrCP
APC/CCdc20 (active)
APC/CCdh1 (active)
Cdc20
Cdh1
APC
SCFSkp2
Cdh1
E2
Degron
Cdk cyclin P M A T P M
G1
S
G2
P M A T P M
APC/CCdh1 surface and backbone
G1
FIGURE 40.15 TWO FORMS OF THE ANAPHASE-PROMOTING COMPLEX/CYCLOSOME (APC/C) CONTROL THE CELL CYCLE. At the metaphase-anaphase transition, the APC/C with associated Cdc20 triggers the onset of anaphase by signaling the degradation of securin and cyclin B. During mitosis Cdh1 phosphorylated by Cdk1–cyclin B is unable to bind the APC/C, so APC/CCdh1 activity is low. As Cdk1–cyclin B activity declines in anaphase, Cdh1 binds the APC/C and APC/CCdh1 drives the exit from mitosis into G1. APC/CCdh1 remains active throughout G1 but is inactivated following synthesis of the specific inhibitor Emi1. After the onset of S phase, SCF (shown here adding a ubiquitin chain to a docked substrate) directs the degradation of cell-cycle substrates such as p27Kip1, following their phosphorylation by protein kinases.
complex inhibits APC/CCdc20 by acting as a competitive substrate. As a result, cyclin B and securin are stable until the checkpoint is satisfied. A few substrates, including cyclin A, bind directly to the APC/C without requiring Cdc20, so they are marked for degradation even when the checkpoint is active. Exiting from mitosis and allowing the G1 cell to prepare chromatin for DNA replication (see Chapter 42) requires low Cdk activity and destruction of Cdc20. This is accomplished late in mitosis by Cdh1, a different co-activator of the APC/C. Phosphorylation of Cdh1 by Cdks blocks its binding to the APC/C early in mitosis. Thus, APC/CCdh1 forms only after cyclin levels (and therefore Cdk activity) decline late in mitosis. As cells pass from G1 into S phase, a newly synthesized inhibitory protein, Emi1, binds to and inactivates APC/CCdh1. This allows cyclins to accumulate during S and G2. Remarkably, APC/CCdh1 is also involved with regulating the activity of synapses in nondividing neurons. After the G1 → S transition and throughout the remainder of interphase members of a different family of E3 activities called SCF regulate the levels of proteins that control Cdk and other cell cycle factors. SCF is named after three of its four subunits: Skp1, cullin, and F-box protein (Fig. 40.16). SCF is a molecular toolbox built on a bow-shaped scaffold formed by the cullin subunit. The fourth subunit, Rbx1, binds near the C-terminus of cullin and uses a protein motif called a RING finger to dock to a ubiquitin-linked E2 enzyme. Skp1 binds to the other end of the cullin, where it provides a docking site for an F-box protein that recognizes and binds the substrate. (The F-box got its name because it was first discovered in cyclin F.) Humans have 78 F-box proteins, giving SCF enormous versatility. Two examples shown in Fig. 40.16 are Skp2, which targets the Cdk inhibitor p27 helping to drive the G1 → S transition, and β-Trcp, which targets Cdc25A, Wee1, and Emi1.
A. Structure of SCFSkp2-E2 complex Skp 2
~50 Å
Skp 1
E2
RbX 1
Cullin 1
B. Some F-box proteins
C. Their target substrates
Skp 2
F
E2F-1: Cell-cycle regulator F box
p27Kip1: Cdk2 inhibitor
Skp 1 β-TrCP1 Cdc25A: Cdk1 activator F
Wee1: Cdk1 inhibitor Emi1: APC/CCdh1 inhibitor Skp 1
β-Catenin: Cell-proliferation regulator
FIGURE 40.16 STRUCTURE AND FUNCTION OF SCF. A, Structure of SCFSkp2. Left, SCF recognizes target proteins through its F-box subunit (Skp2 in this case). Right, Ubiquitin is then transferred from an E2 enzyme. The whole is assembled on a rigid bow-like scaffold composed of the cullin subunit. Structures of F-box proteins Skp2 and β-Trcp (B) and a list of several of their known target proteins (C).
SCF is fundamentally different from the APC/C, because it is constitutively active. However, it ubiquitylates substrates only after they have been phosphorylated, often by Cdks. This feature links SCF activity to the cell cycle.
CHAPTER 40 n Introduction to the Cell Cycle
2
Increasing Cdk activity
1
4
Cdk2-cyclin E
P M A T P M
Increasing APC activity
3
G1
5
1
2
3
709
4
Cdk1-cyclin B Cdk2-cyclin A
S
APC/CCdh1
G2
P M A T P M
G1
APC/CCdc20 Emi1 SCFSkp2, SCFβ–TrCP PP2AB55δ, PP1
FIGURE 40.17 DIAGRAM SHOWING THE CHANGING STATES OF THE CYTOPLASM AS CELLS TRAVERSE THE CELL CYCLE. Between G2 and G1 are shown the various stages of mitosis: P, prophase; PM, prometaphase; M, metaphase; A, anaphase; T, telophase. The states of the cytoplasm discussed in the text are shown as green arrows across the top. Increased Cdk activity is shown as peaks upwards from the central bar. Increasing anaphase-promoting complex/cyclosome (APC/C) activity is shown as a mirror image, with peaks going down from the central bar. APC, anaphase-promoting complex; PP1, protein phosphatase 1; PP2A, protein phosphatase 2A.
Changing States of the Cytoplasm During the Cell Cycle The cell cycle is characterized by five discrete physiological states of the cytoplasm (Fig. 40.17). Changing levels of Cdk activity drive transitions between these states, sometimes counteracted and sometimes reinforced by targeted proteolysis. 1. Early mitosis. Cdk2-cyclin A peaks in prophase, followed by Cdk1-cyclin B. When the nuclear envelope breaks down, the APC/CCdc20 starts degrading cyclin A, but the spindle checkpoint inhibits destruc tion of other substrates. When the last chromosome has attached correctly to spindle microtubules (metaphase), the checkpoint is satisfied, and APC/ CCdc20 starts to degrade cyclin B and securin, an inhibitor of a key protease called separase. Their degradation continues throughout metaphase. 2. Anaphase and mitotic exit. When securin levels fall below a critical threshold, active separase cleaves a key component of the cohesin ring (see Fig. 8.18). This triggers sister chromatid separation. Cyclin destruction continues throughout anaphase and telophase, and falling Cdk1 activity allows the formation of APC/CCdh1, which marks Cdc20 for destruction along with the remaining B-type cyclins. SCFβ-Trcp destruction of Emi1 allows APC/CCdh1 to be active when it forms. 3. G1 phase. APC/CCdh1 and Cdk inhibitors of the CKI and Ink4 families cooperate to inhibit Cdk activity.
Low Cdk activity is required for cytokinesis, spindle disassembly, chromosome decondensation, nuclear envelope reassembly, reactivation of transcription, reassembly of the Golgi apparatus, and assembly of prereplication complexes on the chromosomes. 4. G1–S phase transition. Growth signals from the environment promote the transcription of Cyclin E. If the levels pass a critical threshold, a burst of Cyclin E-Cdk2 activity allows the cell to pass the restriction point, leading to synthesis of proteins required for DNA replication and cell-cycle progression. Cdk phosphorylation targets the CKI peptides for destruction by SCF allowing Cdk2 to become activated. In addition, APC/CCdh1 is inactivated by newly synthesized Emi1. 5. S–G2 phase. Cdk activity remains high throughout the remainder of the cell cycle, and SCF continues to degrade selected proteins tagged by Cdk phosphorylation. SCFβ-Trcp destruction of Cdc25A keeps Cdk1 inactive, preventing a premature entry into mitosis. The APC/C remains inactive, allowing mitotic cyclins to accumulate. It is not known what ultimately triggers entry into mitosis, but an important factor may be a switch in the specificity of SCFβ-Trcp, which spares Cdc25A and instead degrades the Cdk-inhibitory kinase Wee1. Although this sounds complicated, the underlying principles are actually quite straightforward. The following chapters discuss the cell-cycle transitions in greater
710
SECTION X n Cell Cycle
detail and show how the process is modulated in response to a changing environment. ACKNOWLEDGMENTS We thank Tim Hunt, David Morgan, and Jonathon Pines for their suggestions on revisions to this chapter. SELECTED READINGS Bartek J, Lukas J. DNA damage checkpoints: from initiation to recovery or adaptation. Curr Opin Cell Biol. 2007;19:238-245. Brown JS, Jackson SP. Ubiquitylation, neddylation and the DNA damage response. Open Biol. 2015;5:150018. Craney A, Rape M. Dynamic regulation of ubiquitin-dependent cell cycle control. Curr Opin Cell Biol. 2013;25:704-710. Hartwell LH, Weinert TA. Checkpoints: Controls that ensure the order of cell cycle events. Science. 1989;246:629-634.
Hunt T. On the regulation of protein phosphatase 2A and its role in controlling entry into and exit from mitosis. Adv Biol Regul. 2013; 53:173-178. Lorca T, Castro A. The Greatwall kinase: a new pathway in the control of the cell cycle. Oncogene. 2013;32:537-543. Morgan DO. The Cell Cycle: Principles of Control. London: New Science Press; 2007: 297p. Nasmyth K. A prize for proliferation. Cell. 2001;107:689-701. Nurse P. A long twentieth century of the cell cycle and beyond. Cell. 2000;100:71-78. Primorac I, Musacchio A. Panta rhei: the APC/C at steady state. J Cell Biol. 2013;201:177-189. Qian J, Winkler C, Bollen M. 4D-networking by mitotic phosphatases. Curr Opin Cell Biol. 2013;25:697-703. Stukenberg PT, Burke DJ. Connecting the microtubule attachment status of each kinetochore to cell cycle arrest through the spindle assembly checkpoint. Chromosoma. 2015;124:463-480. Wieser S, Pines J. The biochemistry of mitosis. Cold Spring Harb Perspect Biol. 2015;7:a015776.
APPENDIX 40.1
Inventory of the Enzymes of the Cell-Cycle Engine Cyclin-Dependent Kinases and Their Cyclin Partners Kinase
Cyclin (+ Other) Partner
Function
Cdk1 (p34cdc2)
A B1, B2 (Xenopus has 5 B-type cyclins) Cdk1–cyclin B binds Cks1 (Cdc kinase subunit)
Mammals: triggers G2 → M transition. Yeasts: triggers G1 → S and G2 → M transitions. Cyclin A is synthesized in S and destroyed starting at prometaphase. Cyclins B are synthesized in S/G2 and destroyed following the completion of chromosome attachment to the spindle. Cyclins A1, B3 function preferentially in meiosis.
Cdk2
A, E
Triggers G1 → S transition. Can be replaced by other Cdks in mouse.
Cdk4, Cdk6
D1–D3
Phosphorylation of the retinoblastoma susceptibility protein (pRb) in G1. Triggers passage of the restriction point and cyclin E synthesis in some cell types. Extracellular growth factors control synthesis of D cyclins. Can be replaced by other Cdks in mouse.
Cdk5
CDK5R1 or CDK5R2
Neuronal differentiation, sensory pathways.
Cdk7 (CAK)
H; also binds assembly factor MAT1
Cdk activation by phosphorylation of the T loop. Also in TFIIH, important for regulation of RNA polymerase II transcription and DNA repair.
Cdk8
C
Regulation of RNA polymerase II transcription.
Cdk9
T
Regulation of RNA polymerase II transcription.
Inhibitor
Cdk Substrates
Function
CKI: p21Cip1/Waf1 most Cdk-cyclin complexes
Most Cdk-cyclin complexes
Induced by p53 tumor suppresser. Cell-cycle arrest after DNA damage. Binds PCNA (proliferating cell nuclear antigen; see Chapter 42) and inhibits DNA synthesis. Promotes cell cycle arrest in senescence and terminal differentiation. At low levels, may help assemble active Cdk-cyclin complexes.
CKI: p27Kip1
Most Cdk-cyclin complexes
Cell cycle arrest in response to growth suppressers like TGF-β and in contact inhibition and differentiation.
CKI: p57Kip2
Most Cdk-cyclin complexes
Important in development of the palate.
INK4: p15Ink4b
Cdk4, Cdk6
Cell-cycle arrest in response to transforming growth factor (TGF)-β. Altered in many cancers.
INK4: p16Ink4a
Cdk4, Cdk6
Cooperates with the retinoblastoma susceptibility protein (pRb) in growth regulation. Cell-cycle arrest in senescence. Altered in a high percentage of human cancers. This gene overlaps the gene for p19ARF, an important regulator of the p53 tumor-suppresser protein.
INK4: p18Ink4c
Cdk4, Cdk6
Cell-cycle arrest in response to growth suppressers.
INK4: p19Ink4d
Cdk4, Cdk6
Cell-cycle arrest in response to growth suppressers.
Cyclin Inhibitors
CHAPTER 40 n Introduction to the Cell Cycle
711
APPENDIX 40.1
Inventory of the Enzymes of the Cell-Cycle Engine—cont’d Other Components Enzyme
Substrates
Functions
Wee1 kinase
Cdk1 Y15
Nuclear kinase. Inhibits Cdk1-cyclin B in G2.
Myt1 kinase
Cdk1 T14 + Y15
Cytoplasmic kinase. Inhibits Cdk1-cyclin B in G2.
Greatwall (Gwl) kinase (MASTL in humans)
Arpp19, ENSA (α-endosulfine)
Phosphorylated Arpp19 and ENSA inhibit PP2a, allowing active Cdk1cyclin B to accumulate and trigger mitotic entry
Cdc25A phosphatase
Cdk1 T14, Y15
Promotes G1 → S transition and G2 → M transition. Essential for life of the cell.
Cdc25B phosphatase
Cdk1 T14, Y15
Promotes G2 → M transition. Essential in meiosis.
Cdc25C phosphatase
Cdk1 T14, Y15
Promotes G2 → M transition. Dephosphorylates Cdk1 complexed to cyclins A, B at T14 and Y15. Not essential for life.
PP2A phosphatase
Many proteins phosphorylated by Cdk1-cyclin B
Regulated by ENSA/Greatwall. With its targeting subunits B55α-δ and B56α-ε it regulates many activities during mitotic exit and cytokinesis.
PP1 phosphatase
Many targets
Associates with many “targeting subunits,” which can be, for example, regulatory proteins, such as RepoMan or can be structural subunits of the kinetochore (see Chapter 8). It is inactivated by Cdk phosphorylation during mitosis, but has a key role in mitotic exit.
APC/CCDC20
Cyclin B, securin many others
E3 ubiquitin ligase active during M. Requires high Cdk activity to function. Destruction of cyclins and other substrates essential for exit from mitosis. Contains 15 subunits plus the specificity factor Cdc20.
APC/CCdh1
Cyclins A, B, many others
E3 ubiquitin ligase active during G1. Requires low Cdk activity to function. Keeps Cdk activity low in G1 through cyclin proteolysis. Contains 15 subunits plus the specificity factor Cdh1.
SCF
Cyclin E, many others
Class of E3 ubiquitin ligases containing Skp1 + cullin + Rbx1 + an F-box protein. Humans have 78 F-box proteins (Caenorhabditis elegans has more than 300), acting as specificity factors for substrates phosphorylated at specific sites, including cyclin E and Cdk inhibitors.
This page intentionally left blank
CHAPTER
41
G1 Phase and Regulation of Cell Proliferation D
uring the G1 phase of the cell cycle, each cell makes a key decision: whether to continue through another cycle and divide or to remain in a nondividing state either temporarily or permanently. During development of metazoans, cells exit the cell cycle as the first step toward forming differentiated tissues. In adults, strict regulation of the timing and location of cell proliferation is critical to avoid cancer. Cells enter G1 phase at the end of a proliferation cycle, after completing mitosis. To be free to decide whether to proliferate or differentiate, the cell must inactivate the remnants of the proliferation machinery from the preceding cell cycle. This is initiated in late M-phase by inactivating cyclin-dependent kinases (Cdks [see Chapter 40]) via proteolytic destruction of their cyclin subunits. This continues in G1 phase and is accompanied by synthesis and stabilization of Cdk-inhibitory proteins. The absence of Cdk activity activates a regulatory network that represses the transcription of many genes that promote cell-cycle progression. While this repressive network is active, the cell cannot proceed through the cell cycle. The repression can be switched off if the cell is stimulated by specific signals from the surrounding medium, extracellular matrix, and other cells (see Chapters 27 and 30). If these signals are diffusible substances, they are known as mitogens. Mitogens can trigger another round of DNA replication and mitosis, but first, the cell must pass a major decision point in G1 called the restriction point (Fig. 41.1). In metazoans, many cells cease cycling, either temporarily or permanently, exiting the cell cycle into a state known as G0 (Fig. 41.1). This frequently accompanies their acquisition of specialized, differentiated characteristics. Occasionally, it is desirable in tissues for cells in G0 to reenter the cell cycle to replace lost cells. Specialized cells called stem cells fulfill this role in tissue maintenance. This chapter describes how cells decide whether to exit from the cell cycle into the G0 phase, how they return
to the cycle from G0, and how they regulate their progress through the G1 phase. It also considers some of the points at which defects in cell cycle control lead to cancer.
G0 Phase and Growth Control Cells stop cycling in three ways. First, they may receive external signals instructing them to withdraw from the cell cycle, enter into G0, and differentiate, as discussed later. Second, cells may find themselves in an environment with insufficient mitogens to drive proliferation. Under these conditions many cell types enter a transient nondividing state known as quiescence, while they wait for conditions to improve. Third, cells that have suffered DNA damage because of a loss of cell-cycle control undergo senescence, or may, in some cases, commit suicide by apoptosis (see Chapter 46). Senescence is a permanent nondividing state that is physiologically distinct from G0 and from which cells normally cannot exit. One physiological signal that can lead to senescence is a critical shortening of the telomere regions of the chromosomes in cells
G2
M
Lack of mitogens Antiproliferation signals (eg, contact inhibition, TGF-β telomere damage) G0
S
Restriction point: Regulates G1 progression to S phase Blocks cell-cycle progression G1 unless nutrients & mitogens are continuously present
Proliferation signals Nutrients
FIGURE 41.1 CELL CYCLE SHOWING MAJOR LANDMARKS IN THE G1 PHASE. TGF-β, transforming growth factor–β.
713
714
SECTION X n Cell Cycle
Contact inhibition
TUMOR
p27 p21 p16 p15
Differentiation signals Senescence R Blood vessels
Cdk2–cyclin E–p21 Cdk2–cyclin E–p27 (inactive) Cdk2–cyclin E p27 Cdk4–cyclin D–p27
TGF-β
Cdk4-p15 Cdk4-p16 (inactive)
Cyclin D (degraded)
Normal epithelial cells Various signals stimulate synthesis of Cdk inhibitors
Cdks inactive Hypophosphorylated Rb binds E2F and inactivates genes required for cell-cycle progression
FIGURE 41.3 HOW EXTERNAL STIMULI ACT ON CDK INHIBITORS TO CAUSE CELLS TO ENTER A NONDIVIDING G0 STATE. TGF-β, transforming growth factor–β. NORMAL FIGURE 41.2 DISRUPTION OF NORMAL TISSUE ARCHITECTURE BY UNCONTROLLED PROLIFERATION OF CANCER CELLS. Lower right, Normal thyroid tissue. Upper left, A thyroid tumor with loss of the normal gland structure. (Courtesy Clara Sambade, IPATIMUP, Porto, Portugal.)
that have divided more than a critical number of times. Overexpression of telomerase (see Fig. 7.14) can, when combined with suitable mitogenic stimuli, prevent cells from undergoing senescence in tissue culture. Most cells in multicellular organisms are differentiated (adapted to carry out specialized functions) and no longer divide. They typically form specialized tissues, each with a distinctive structural organization that is important for its function. Unscheduled cell division can severely disrupt the organization of such tissues (Fig. 41.2). Accordingly, tissues strictly regulate both the location and the frequency of cell division. In most tissues, divisions normally occur at a low rate, producing new cells in numbers just sufficient to replace those that die. Under special circumstances, however, such as in response to wounding (see Fig. 32.11), the rate of cell division may increase dramatically. This highlights an important constraint on cell-cycle control in multicellular organisms: To make organized tissues, cells must exit from the cell cycle, but some cells must also retain the ability to reenter the active cell cycle when needed to repair injuries or replace worn-out cells. Cells that stop cycling to differentiate are said to have left the cycle and entered a nonproliferating state called G0 (Fig. 41.1). G0 may last hours or days, or even for the life of the organism, as it does for most neurons. It is important to note that nondividing cells are not dormant: G0 cells can be biochemically very active and continue to expend large amounts of energy for many
BOX 41.1 Cdk Inhibitor Scorecard The regulation of G1 progression requires the action of two families of Cdk inhibitory proteins (see Chapter 40). Cyclin-dependent kinase inhibitors (CKIs), which include p21Cip1 and p27Kip1, usually inhibit all Cdks and block cellcycle progression, but under special circumstances, they can actually activate Cdk4/6-cyclin D and promote cell proliferation during G1. Cell-cycle blocks imposed by CKIs tend to be temporary. INK inhibitors, which include p15Ink4B and p16Ink4a, are specialized at inhibiting Cdk4/6cyclin D. These can promote a profound and often permanent cell-cycle arrest by activating the Rb pathway of gene repression.
ongoing processes. Because of turnover, all cells must continuously synthesize housekeeping proteins. They must also expend energy to maintain intracellular pH and ionic composition and to power intracellular motility. In addition, many specialized G0 cells consume large amounts of energy to synthesize and secrete protein products and generate action potentials. Energy metabolism is particularly dramatic in muscle cells that are responsible for all body movements. Thus, most G0 cells should be regarded as active cells that just happen no longer to be engaged in cell division. Transforming growth factor–β (TGF-β) is an example of an external signal that arrests progress through the cycle and regulates differentiation and tissue morphogenesis (Fig. 41.3). TGF-β stimulates a receptor serine/ threonine kinase that activates SMAD transcription factors (see Fig. 27.10). SMADs suppress Cdk-4 synthesis and increase expression of CKI (cyclin-dependent kinase inhibitor) and Ink4 class Cdk inhibitors (Box 41.1). p15Ink4B preferentially inactivates Cdk4–cyclin D and Cdk6-cyclin D complexes. It also displaces CKI class
CHAPTER 41 n G1 Phase and Regulation of Cell Proliferation
inhibitors from the Cdk4–cyclin D, permitting them to transfer to Cdk2–cyclin E complexes in the nucleus. This further inhibits cell-cycle progression. The CKI p27Kip1 helps arrest the cell cycle of normal cells when they become crowded by neighboring cells (contact inhibition; see ahead, Fig. 41.11) or if their environment lacks mitogens. Genetic analysis in mice revealed that p27Kip1 also regulates cell-cycle progression during development. Mice lacking p27Kip1 are 30% larger than their normal littermates by several weeks of age. This is at least partly because cells in many organs undergo extra rounds of cell division. An analogous mechanism limits proliferation during the differentiation of muscle. The transcription factor MyoD drives expression of the CKI inhibitor p21Cip1, which helps arrest proliferation and start muscle differentiation (Fig. 41.3). p21Cip1 stops cell-cycle progression in at least two ways. First, it binds Cdk–cyclin complexes and stops them from promoting cell-cycle progression. It also blocks DNA replication by inhibiting the DNA replication factor proliferating cell nuclear antigen (PCNA [see Chapter 42]) that is required for DNA polymerase δ activity.
715
Both p21Cip1 and p16Ink4a contribute to the permanent cell-cycle arrest of senescent cells. p16Ink4a activates inhibitory proteins of the retinoblastoma protein (Rb) and E2F families (see later) that bind to promoters and recruit histone methyltransferases, forming heterochromatin that permanently inactivates genes required for proliferation (see Fig. 8.7). Once cells exit the cycle, multiple redundant pathways block reentry by reinforcing the primary inhibition of Cdk activity. In addition, a specialized histone variant H1o replaces histone H1 in G0 cells, resulting in more condensed chromatin. This represses transcription generally. However, not all gene expression is suppressed in differentiated cells, many of which synthesize large amounts of specific proteins (eg, digestive enzymes secreted by the pancreas).
Moving Into and Out of G0: Stem Cells Stem cells are professionals at moving back and forth between G0 and proliferative cell cycles. One of their roles is to replace worn-out parts of tissues as differentiated cells age or die as a result of various misadventures. Box 41.2 provides a brief introduction to stem cells.
BOX 41.2 Stem Cells in Mammals The defining feature of stem cells is their capacity to selfrenew while producing daughter cells with the capacity to differentiate into more specialized cells under the control of intrinsic and environmental cues. Stem cells play a key role in the development of multicellular organisms in addition to providing cells for the renewal and regeneration of adult tissues. Each multicellular organism begins as a single cell (a fertilized egg) with a genome encoding the information required to produce an adult. The divisions prior to implantation in the uterine wall produce a small group of epiblast cells that go on to form the embryo. The other cells that are produced at this stage are specialized to support the embryo. Epiblast cells are termed pluripotent because their progeny can form all the specialized cells of the adult. Although the pluripotent cells disappear as the embryo develops, epiblast cells can be propagated and maintained permanently in culture without losing their pluripotency if optimal conditions are provided. These cell lines are called embryonic stem cells (ES cells). Most adult tissues set aside a few tissue stem cells that have the capacity to renew themselves and to produce daughter cells that differentiate into a limited range of specialized cells (see Figs. 28.1, 28.4, and 40.1). Adult stem cells have diverse patterns of cell-cycle regulation. Some tissue stem cells continue the cell cycle throughout life. For example, epithelial stem cells give rise to mature cells that continuously replace the skin and the lining of the gastrointestinal tract. Hematopoietic stem cells in bone marrow give rise to both short-lived and long-lived differentiated blood cells. In other organs, such as liver and skeletal muscle,
tissue stem cells are held in reserve unless the tissue is damaged, when they produce daughter cells to repair the damage. Stem cells are present even in organs that have a limited capacity for renewal and regeneration, such as the nervous system. The potential for regeneration from stem cells has stimulated research to find ways of using embryonic or tissue stem cells to repair damaged or diseased organs in human patients. Stem cells have also been useful for production of transgenic animals for scientific research (eg, knockout mice).
Discovery and Defining Features of Stem Cells Pioneering work on blood cell development (see Fig. 28.4) established the existence of stem cells and defined many of the concepts that apply to all types of stem cells. The key experiment was to inject bone marrow cells from a normal mouse into a mouse that had been irradiated to kill all the cells that produce blood cells. Transplantation of bone marrow cells rescued the irradiated mice from death from anemia, bleeding, and infections. The transplanted bone marrow contained precursor cells that formed colonies of proliferating cells that regenerated the full range of blood cells. The blood-forming colonies in the spleen, each of which formed from a single stem cell, contained either one or, infrequently, several types of differentiating blood cells. This experimental system first revealed the existence of several different types of progenitor cells in bone marrow with the dual capacity to proliferate and to give rise to more differentiated cells (see Fig. 28.4). These committed progenitor cells have a limited proliferation capacity and can Continued
716
SECTION X n Cell Cycle
BOX 41.2 Stem Cells in Mammals—cont’d give rise to only to specific subsets of blood cells, such as red blood cells, platelets, granulocytes, or lymphocytes. In contrast, there are a very few multipotent hematopoietic stem cells in bone marrow. They replenish the pool of committed progenitor cells, ultimately acting as a source of all types of blood cells, while maintaining themselves throughout the life of the individual. Antibodies for surface markers can now be used to distinguish and purify the various types of hematopoietic progenitors as well as stem cells from mice and humans. Once separated from the far more numerous mature and differentiating cells in bone marrow, stem cells can be used for transplantation into patients with bone marrow defects. Most multipotent hematopoietic stem cells are in the G0 phase of the cell cycle. A low level of metabolic activity is thought to contribute to their longevity, which can potentially exceed the life span of the individual. When stimulated by demand for more blood cells, growth factors drive multipotent stem cells into a cell cycle that culminates in an asymmetrical division. One daughter cell is another multipotent stem cell. The second daughter cell enters the proliferating pool of blood cell precursors as a committed progenitor cell. Committed progenitor cells and their progeny proliferate vigorously and differentiate into mature blood cells. An adult human produces more than one million blood cells every second. Cytokines and other growth factors regulate proliferation and differentiation at every stage of blood cell production. The later stages are best understood. For example, the cytokine erythropoietin acts through a kinase-coupled receptor to activate a cytoplasmic transcription factor that stimulates the proliferation and differentiation of the red blood cell lineage. Other cytokines guide the differentiation of granulocytes and monocytes. Hematopoietic stem cells respond to the same families of growth factors that control other aspects of development, including Wnts (see Fig. 30.7), Notch (see Chapter 24), fibroblast growth factor (see Fig. 24.4), and insulin-like growth factor (see Fig. 24.4). However, too little is known about these regulatory mechanisms to grow hematopoietic stem cells in the laboratory.
Properties of Adult Stem Cells Years of detailed analysis in the laboratory and clinic established hematopoietic stem cells as a model for stem cells in other tissues. General features include the capacity for selfrenewal and the production of daughters that proliferate and differentiate. This dichotomy can be achieved by asymmetrical cell divisions guided by the same types of internal cues that control unequal divisions of cells in early embryos (Fig. 41.4). Symmetrical divisions yielding two daughter stem cells can also expand the numbers of stem cells during growth to maturity and during regeneration of damaged tissues. Stem cells depend on local environmental cues to maintain their status as stem cells. These special environments, called stem cell niches, are created by tissue cells and the extracellular matrix. Niche cells anchor stem cells with adherens junctions and provide cell surface and secreted proteins that activate the signaling pathways that regulate the cell
Niche cell
Stem cell
Adherens junction Asymmetrical Stem cell division + Committed cell
Symmetrical 2 stem cells division
Differentiated progeny
FIGURE 41.4 TWO PATTERNS OF STEM CELL DIVISION. Asymmetrical divisions create two daughter cells: a stem cell that remains associated with its niche cell to maintain the pool of stem cells and one that is committed to multiply and produce differentiated progeny. Symmetrical divisions produce two stem cells to expand the pool of stem cells.
cycle of the stem cell. Some of these factors stimulate division; others inhibit differentiation. The niches occupied by germ cells and neural stem cells from invertebrates are particularly well characterized. During asymmetrical divisions of these stem cells, the renewed stem cell stays behind in the niche, while the daughter that is destined to differentiate into an egg, sperm, or neuron is released. In bone marrow, osteoblasts (see Fig. 32.5) and endothelial cells (see Fig. 30.13) provide niches for hematopoietic stem cells.
Epidermal Stem Cells Skin is an example of a continuously renewing organ with a considerable capacity for regeneration (see Fig. 40.1). Multipotential and committed stem cells contribute to both renewal and regeneration. Committed stem cells reside in the basal layer of the epidermis. Asymmetrical cell divisions oriented at right angles to the basal lamina produce two daughter cells. The daughter touching the basal lamina carries on as the stem cell. The apical daughter cell divides multiple times and differentiates into an ascending column of cells, forming the superficial layers of the epidermis (see Fig. 35.6). Multipotent stem cells associated with hair follicles give rise to all the cells of the hair follicle and also serve as a reserve for the committed epidermal stem cells in the event of injury (Fig. 41.5).
Skeletal Muscle Stem Cells Small numbers of stem cells reside in a niche sandwiched between the basal lamina and the giant multinucleated muscle cells. If the muscle is damaged, these quiescent “satellite cells” multiply and produce muscle cells that regenerate the tissue. Positive signals for proliferation and differentiation come through receptor tyrosine kinases and the mitogenactivated protein (MAP) kinase pathway (see Fig. 27.6) and other pathways. Restraining signals are provided by myostatin, a member of the transforming growth factor (TGF)-β family (see Fig. 27.10). Inactivation of the myostatin pathway results in massive enlargement of muscles in mice and humans. Muscles are capable of regenerating multiple times,
CHAPTER 41 n G1 Phase and Regulation of Cell Proliferation
717
Epidermis Ba
sa
ye r Ha ir s haf t
BOX 41.2 Stem Cells in Mammals—cont’d
l la
Dermis
Sebaceous gland
Bulge
A
B
C
Hair bulb
FIGURE 41.5 STEM CELLS FROM SKIN. Multipotent stem cells of the skin reside in the hair follicle bulge (green cells in A, diagram in C). They move up and repair the epidermis during wound healing, and they move down and generate new hair growth during the hair cycle. B, Depicts a Nude mouse grafted with the cultured cell progeny of a single “bulge” stem cell and displaying a large tuft of hair, all derived from that single stem cell. (A and C, From Fuchs E, Tumbar T, Guasch G. Socializing with the neighbors: stem cells and their niche. Cell. 2004;116:769–778. B, From Blanpain C, Lowry WE, Geoghegan A, et al. Self-renewal, multipotency, and the existence of two cell populations within an epithelial stem cell niche. Cell. 2004;118:635–648.)
so the stem cell population renews itself during regeneration or is augmented by stem cells that migrate through the blood from bone marrow or other tissues.
Cancer Stem Cells Stem cells may play a role in cancer, acting as a source for proliferating cells that make up the bulk of the tumor. If true, this concept helps explain why it is relatively easy to reduce the size of tumors by targeting dividing cells but difficult to completely eliminate residual tumor stem cells, which may divide less frequently. It is thought that the spread of cancer from the primary tumor to other tissues (metastasis) requires circulating cancer cells to find locations (niches) where they can establish themselves as stem cells.
Meristematic Stem Cells in Plants The growth of plants depends on carefully orchestrated proliferation and differentiation of cells derived from stem cells called meristems. Through asymmetrical divisions, these relatively inactive cells give rise to daughters that proliferate at the tips of shoots and roots. The proliferating cells differentiate into specialized tissues such as flowers, while the stem cells maintain a pool of slowly replicating cells in a special niche.
Use of Stem Cells to Make Transgenic Animals Because embryonic stem cells grow in culture, they can be manipulated experimentally. They can be transfected with DNA, and if the proper sequences are present, this DNA can replace a region of the endogenous chromosome by homologous recombination. If the modified embryonic stem cells are subsequently injected into developing embryos at the blastocyst stage, they are, with low frequency, able to colonize the cell population that will produce germ cells. When such chimeric embryos grow to adulthood, a proportion of their gametes will carry a chromosome with the modification engineered in the embryonic stem cells. Furthermore, this chromosome will now be inherited by all progeny of that embryo, giving rise to a line of transgenic animals. This method is widely used in research to knock out genes by
designing the original DNA construct so that when it enters the chromosome by homologous chromosome, a critical region of a target gene is deleted or disrupted. The use of knockout mice has revolutionized the study of developmental biology by allowing investigators to determine the function of specific genes in living animals. Note the distinction between transgenic animals and reproductive cloning. “Cloned” animals are produced by introducing a somatic cell nucleus into an enucleated egg. Experiments first in frogs and later in mammals, such as Dolly the sheep, established that egg cytoplasm can reprogram gene expression of differentiated cell nuclei, and enable the development of a cloned animal. Transfer of nuclei from lymphocytes and olfactory neurons has been used to derive healthy adult mice. This approach involves the reversal of epigenetic changes in the nucleus that drove the differentiation of the adult cell. The molecular mechanisms underlying reprogramming are not well understood and the success rate at obtaining healthy animals through nuclear cloning is very low. This cloning does not involve the use of stem cells, but embryonic stem cells can be derived from the blastocyst stage of the cloned embryos.
Therapeutic Applications of Stem Cells Where committed stem cells can be isolated from an adult organ, it is now possible to regenerate damaged tissues by transplanting these stem cells from patients themselves or from donors. The best example is transplantation of bone marrow stem cells to treat patients whose bone marrow has been damaged by cancer, chemotherapy, or other disease. Adverse immunologic reactions are a challenge for transplants from donors other than an identical twin. On one hand, the immune system of the recipient can reject the transplanted cells. On the other hand, lymphocytes contaminating the donor stem cells can mount an immunologic attack on the recipient. Using purified hematopoietic stem cells (ideally, the patient’s own stem cells) rather than mixed bone marrow cells avoids this problem. Knowing how to expand hematopoietic stem cells in vitro would be helpful. This approach is already used for treating burns with epidermal stem cells. Normal skin is used as a source of committed Continued
718
SECTION X n Cell Cycle
BOX 41.2 Stem Cells in Mammals—cont’d skin stem cells, which are multiplied in culture and used to regenerate all the layers of the skin. Stem cells might be used to regenerate other damaged tissues, including the insulin-producing cells that are lost in Type I diabetes, but appropriate stem cells are not available for many organs, including the pancreas, brain, and heart. Indeed, even with appropriate stem cells in hand, much remains to be learned about how to grow them and then direct them to differentiate into mature tissues. Embryonic stem cells can potentially supply all cells necessary to replace any damaged tissue, but sources of human embryonic stem cells are limited, and acquiring them from early embryos discarded by fertility clinics is unacceptable to some people. Patient-specific (autologous) embryonic stem cells can be derived via somatic cell nuclear transfer (cloning) followed by expansion of epiblast cells from the embryo. However, the production of such “artificial” human embryos is also highly controversial. Adult stem cells are an alternative to embryonic stem cells. This approach has the advantage that stem cells can be isolated from bone marrow, blood, and skin by using antibodies that recognize specific surface protein “markers”; however, these specialized stem cells normally do not produce differentiated cells for regeneration of other tissues. An alternative method to generate pluripotent stem cells is based on the transient reintroduction into a differentiated cell of a group of specific transcription factors commonly expressed in stem cells. These include Sox2, Oct4, Klf4, and
Myc. When these factors are ectopically expressed in differentiated cells, they can induce the expression of proteins required for pluripotency, as well as factors required to change the epigenetic landscape of the cell. This results in a stable perpetuation of pluripotency. These induced pluripotent stem (iPS) cells, offer the promise that they can be differentiated into any desired specific cell type for medical applications. This procedure potentially eliminates the problem of the availability of autologous stem cells as many cell types can be reprogrammed to iPS cells. This technology is rapidly developing, as researchers attempt to improve the generation of functional cell types from iPS cells When injected into immunodeficient mice, iPS cells make tumors containing all cell types, called teratocarcinomas. This clearly shows that the iPS cells are pluripotent. They can also generate all cell types in tissue culture when appropriate growth factors necessary for self-renewal are removed from culture media. However, this differentiation is difficult to control. We cannot yet generate specific fully functional cell types necessary for therapies with high purity from pluripotent stem cells, except for a few cell types such as retinal pigment epithelial (RPE) cells. iPS cell-derived RPE cells were transplanted for the treatment of macular degeneration in 2014, and appeared to be successful in blocking the degeneration. However, as of 2016, this was the only clinical trial performed with iPS cell-derived cells. Investigation on how to control iPS cell differentiation is one of the main obstacles to be overcome for regenerative medicine.
Reentry Into the Cell Cycle Levels of expression
Cells in the G0 phase may reenter the growth cycle in response to specific stimulation by mitogens, often induced by injury or normal cell turnover. Cultured fibroblasts are favored for laboratory studies of this process, as they readily enter a quiescent state mimicking G0 phase when deprived of serum (ie, mitogens and growth factors) and rapidly reenter the cell cycle when serum is restored. This response reproduces that found in wounded tissues. When a living tissue is wounded (see Fig. 32.11), fibroblasts are exposed to serum that has leaked from damaged blood vessels. In response, they divide and colonize the wound, where they lay down new extracellular matrix to repair the damage. Serum stimulates three waves of gene expression in cultured quiescent fibroblasts (Fig. 41.6). The first includes more than 100 “immediate early” genes. These include transcription factors of the Jun, fos, myc, and zinc finger families (see Chapter 10) that activate numerous downstream genes required for cell growth and division. Other immediate early genes encode tissue remodeling factors, cytokines (growth factors), extracellular matrix components (fibronectin), plasma membrane adhesion proteins (integrins), and cytoskeletal proteins (actin, tropomyosin, vimentin), as well as
Immediate early transcription factors
Delayed early genes Immediate early structural proteins Immediate early tissue repair proteins
0
1
2
3
4
Time (hours)
5
6 Cdk inhibitors
Addition of serum or growth factor
FIGURE 41.6 PATTERNS OF EXPRESSION OF IMMEDIATE AND DELAYED EARLY GENES DURING THE RETURN OF GROWTH-ARRESTED FIBROBLASTS FROM G0 TO ACTIVE PROLIFERATION AND THE CELL CYCLE.
activities involved in angiogenesis (blood vessel formation), inflammation, and coagulation. These proteins facilitate the movement of fibroblasts into wounds and initiate the repair of tissue damage. Expression of a second wave of “delayed early” genes encoding a variety of proteins that are required for cell growth and proliferation, including cyclin D, precedes the onset of the S phase. Genes activated after the onset of the S phase are referred to as “late” genes. Both
CHAPTER 41 n G1 Phase and Regulation of Cell Proliferation
delayed early and late gene transcription require synthesis of the transcription factors encoded by immediate early genes. These waves of transcription in response to mitogens enable the G0 cells to pass through a “gate” and reenter the active cell cycle. This gate is analogous to the restriction point, a critical aspect of G1 control that regulates the proliferation of all normal cells.
The Restriction Point: A Critical G1 Decision Point All eukaryotes have a mechanism that operates during the G1 phase to ensure that cells duplicate their genome only when the environment is supportive and the chromosomes are undamaged. Healthy yeast cells do not embark on a round of DNA replication and division until they reach an appropriate minimum size (actually, they probably measure their ribosome content and ongoing rate of protein synthesis). This is important because after cell division, the daughter cell (bud) is smaller than the mother and needs more time to grow before it divides if the population is to maintain a constant cell size. Whether dividing mammalian cells also monitor their size is not yet settled. The influence of cell size on the division cycle was first demonstrated in an elegant microsurgery experiment (Fig. 41.7). Two Amoeba proteus cells were grown under identical conditions in parallel cultures. Each day, a portion of the cytoplasm was amputated from one amoeba, and the other was left untouched as a control. Under those circumstances, the cell that suffered the amputations did not divide for 20 days. During this time, the control amoeba divided 11 times. When the
Amoeba
719
amputations were stopped, the amoeba that had been operated on divided within 38 hours. The interpretation of this experiment was that the repeated amputations prevented the experimental amoeba from ever attaining a size sufficient to turn on the division program. Evidence suggests that some types of human cells have a similar size control while others do not. An essential aspect of growth control during the G1 phase involves monitoring the external environment for nutrient availability and for signals to proliferate (mitogenic signals) coming from other cells and from the extracellular matrix. In a classic experiment, when three flasks containing populations of cultured cells were starved by deprivation of amino acids, serum, or phosphate respectively, they all stopped cycling in G1. When the missing ingredients were restored, cells in all three flasks resumed the cell cycle and entered the S phase at about the same time. This was surprising because amino acids are needed to make protein, serum provides growth factors and mitogens, and phosphate is needed for synthesis of DNA and phospholipids (needed to make membranes). This experiment was interpreted as evidence that all three types of starvation caused cells to arrest at a single point in the G1 phase, termed the restriction point. The restriction point is now defined as the point after which the cell cycle will proceed even if mitogenic factors are withdrawn (Fig. 41.8). This supremely important aspect of cell-cycle control prevents cells from dividing at inappropriate times and in inappropriate places. Defects in restriction point control are among the most common causes of cancer. Genetic analysis also revealed a point in the G1 phase after which budding yeast cells appear to be committed to completion of the cycle. Cells that are starved for
A. Control
Nucleus
B. Experiment
FIGURE 41.7 A MICROSURGERY EXPERIMENT DEMONSTRATES THAT AMOEBAE WILL NOT DIVIDE IF THEY ARE PREVENTED FROM ATTAINING A SUFFICIENT SIZE. A, Control cell continues to divide. B, Experimental cell does not divide. (For reference, see Prescott DM. Relation between cell growth and cell division. II: The effect of cell size on cell growth rate and generation time in Amoeba proteus. Exp Cell Res. 1956;11:86–98.)
720
SECTION X n Cell Cycle Restriction point
M
G1 Cell continues to cycle only if extracellular signals are received
S Cell committed to cycle
FIGURE 41.8 RESTRICTION POINT. During late G1, cells assess external and internal stimuli and decide whether to commit to a further round of DNA replication and division.
nutrients arrest at, or just prior to, this point, termed START. The mammalian restriction point resembles yeast START in a number of aspects, but they are not exactly equivalent, owing to differences between animal and yeast cell cycles.
Regulation of Cell Proliferation by the Restriction Point The restriction point is a molecular “gate” that regulates the expression of genes required for cell-cycle progression. The gate is based on proteins that are related to the Rb susceptibility protein and a family of essential transcription factors known as E2F. In brief, E2F is a master transcriptional regulator that activates many of the genes whose products drive DNA replication and cell-cycle progression. Rb regulates the cell cycle by binding to E2F and converting it into a repressor of those same cell-cycle genes. When Rb is bound to E2F, the cell is said to be arrested at the restriction point. Escape from this arrest involves Cdk activation and subsequent Rb phosphorylation. This releases E2F, which then drives cell-cycle progression. An alternative way to pass the restriction point gate depends on a potent regulator called Myc, which is discussed separately later. Mammals have three Rb-related proteins (pRb, p107, and p130) and approximately eight E2F family members. These together constitute a complex multifunctional network. This chapter refers to the families generically as Rb and E2F. Some E2F proteins form a heterodimer with one of three DP (differentiation-regulated transcription factor-1 protein) family members. This dimer associates with the promoter region of E2F target cell-cycle genes (Fig. 41.9A). Rb binding converts E2F/DP from a transcriptional activator to the Rb/E2F/DP repressor. Rb also recruits histone deacetylases, enzymes that remove acetyl groups from a wide range of proteins, including the aminoterminal tails of histones (see Fig. 8.7). This causes compaction of chromatin structure and represses genes required for cell-cycle progression. Nonproliferating cells have low Cdk activity in G1 before the restriction point. First, cyclin D messenger RNA (mRNA) levels are low, so little protein is made. Second, any cyclin D that is made is retained in the cytoplasm, where it is phosphorylated by glycogen synthase
kinase (GSK)-β (see Fig. 30.7) and marked by SCF (Skp, Cullin, F-box containing complex) for degradation (see Fig. 40.16). Rb/E2F/DP represses the expression of the genes for cyclins E and A required for Cdk2 activation. Furthermore, high levels of the CKI class Cdk inhibitor p27Kip1 inhibit any Cdk2–cyclin E or Cdk2–cyclin A that happens to be present (see Fig. 40.14). Signals from mitogens and the extracellular matrix open the restriction point gate. Stimulation of receptor tyrosine kinases (see Chapters 25 and 27) or integrins (see Chapter 30) activates Ras and the mitogen-activated protein (MAP) kinase/extracellular signal–regulated kinase (ERK) cascade (see Fig. 27.6). The output of this cascade stimulates transcription of D-type cyclins (Figs. 41.9 and 41.10) and also inactivates GSK. This allows cyclin D to accumulate in nuclei. Nuclear cyclin D binds to and activates Cdk4 and Cdk6 (referred to hereafter as Cdk4/6–cyclin D), producing an initial pulse of Cdk activity that is later amplified by Cdk2–cyclin E and Cdk2–cyclin A. Cdk activity in early G1 is regulated by adjusting the relative levels of the three D-type cyclins as well as the levels of Ink4 and CKI inhibitors. This regulation of cyclin D levels and Cdk inhibitors provides the crucial link between extracellular mitogens and the cell cycle. Mitogens also stimulate transcription of the CKI class Cdk inhibitor p27Kip1. This protein actually activates Cdk4/6–cyclin D complexes in two ways. First, the receptor associated tyrosine kinases Jak or Src phosphorylate p27, inducing a structural change within the Cdk4/6–cyclin D-p27Kip1 complex that activates the kinase. This links mitogen signaling to Cdk-4/6 activation. It also promotes the nuclear import of Cdk4/6– cyclin D. This both leads to full activation of Cdks by Cdk-activating kinase, a nuclear enzyme (see Chapter 40), and increases the stability of cyclin D. All of this depends on the continuous presence of mitogenic signals; if these cease, then cyclin D stability rapidly declines again, since following dephosphorylation, p27 and p21 again act as Cdk4/6–cyclin D inhibitors. Cdks push the cell past the restriction point by phosphorylating Rb, causing it to dissociate from E2F (Fig. 41.9B; see also Chapter 40 and Appendix 40.1). The E2F/ DP heterodimer remains bound to promoter regions and now potently activates, rather than represses, the transcription of genes that stimulate cell proliferation. The proteins produced synthesize DNA (DNA polymerase α, accessory factors, and enzymes that synthesize nucleotide precursors; see Chapter 42), promote cell-cycle progression (cyclins E and A, Cdk1, and Cdc25), and regulate cell-cycle progression (pRb, p107, Emi1). The chain of events as mitogens break the blockade on cell-cycle progression imposed by Rb involves a positive feedback loop as follows. Cdk4/6–cyclin D complexes begin to phosphorylate Rb. This releases some E2F and permits the initial expression of genes that
CHAPTER 41 n G1 Phase and Regulation of Cell Proliferation
A. Absence of mitogens Tyrosine kinase receptor
External signals
Seven-helix receptors
Ras
G1
Nucleosome
S Rb
E2F/DP
Seven-helix receptors
MEK ERK G1
4/6 D Cdk 4/6 cyclin D
E2F/DP
AC
AC
AC
AC
G2
M
S Histone deacetylase
Rb
S
FIGURE 41.10 HOW GROWTH FACTORS REGULATE CDK4/6 ACTIVITY: THE ROLE OF D-TYPE CYCLINS AND P21. Mitogeninduced tyrosine phosphorylation of p27 determines if it assembles active or inactive CDK4/6–cyclin D complexes–linking assembly of active kinase to mitogens.
cAMP PKA Synthesis and stability of cyclin D, etc
Raf
Enters nucleus
Phosphorylates Rb
G1
Steroid receptors
p27-P
Passage of restriction point
B. Mitogens present
M
+ Cdk 4/6 Kinase phosphorylates p27 tyrosine on p27
Cdk 4/6–cyclin D–p27-P (active kinase)
Histone N-terminal tails
Histone deacetylase
Tyrosine kinase receptor
Cyclin D (stable)
Gene off
Histone deacetylation results in chromatin compaction and repression of transcription
Ras
Cyclin D Transcription
Restriction point
M
721
RNA pol II
AC
AC
Cell-cycle genes (cyclins A, E, Cdk1) DNA replication genes
AC
AC
AC
AC
Acetylated "open" chromatin favors transcription FIGURE 41.9 REGULATION OF CELL-CYCLE PROGRESSION BY THE E2F/DP/RB COMPLEX. A, The Rb/E2F/DP complex recruits histone deacetylases (see Chapter 8) and represses specific genes that are required for cell-cycle progression. This blocks cellcycle progression at the restriction point. B, Phosphorylation of Rb by Cdks alleviates this block and permits passage of the restriction point. cAMP, cyclic adenosine monophosphate; MEK, mitogen-activated protein kinase kinase; PKA, protein kinase A.
encode cyclin E, cyclin A, and CDC25A. Active Cdk2cyclin E can initiate p27Kip1 degradation, permitting the rapid accumulation of active Cdks. The restriction point probably is passed here, and from this point on Cdks remain active until cyclin destruction in late mitosis. Cdk2–cyclin E participates in a second wave of Rb phosphorylation on many sites, leading to the wholesale liberation of E2F/DP and a surge in transcription of genes that trigger the onset of DNA replication (S phase entry) and promote progression through the cell cycle. As the cell cycle proceeds, Rb phosphorylation is maintained first by Cdk2–cyclin A and then later by Cdk1–cyclin B
until the exit from mitosis. Rb is dephosphorylated at the mitosis-G0 or G1 transition. This enables it once again to bind E2F and close the restriction point gate to exit from the next G1. The transcriptional regulator Myc drives an alternative pathway for G1 exit that is also stabilized by mitogenic signals. Association of myc with one partner activates the transcription of cyclins E and D2. Association with a different partner downregulates the transcription of Cdk inhibitors of both the CKI and INK classes. These activities partly explain why Myc can act as an oncogene—a protein that helps transform normal cells into cancer cells (explained further later).
Restriction Point and Cancer Cancer is a complex class of diseases in which genetic changes within clones of cells lead to production of cell populations whose uncontrolled growth can disrupt tissue function and ultimately kill the individual. Two in five Americans will be affected by cancer during their lifetimes. This sounds very high, but considering the number of cell cycles that are required to produce a human composed of approximately 1014 cells, and considering the over 1 million cell divisions that occur per second in a healthy adult, the disease is actually remarkably rare on a per cell basis. This is at least partly because multiple genetic alterations are required to transform a normal cell into a cancer cell. Furthermore, the cell cycle is highly regulated by a web of negative feedback pathways that hold in check activities driving cellular
722
SECTION X n Cell Cycle
Cells traversing the cell cycle
Normal
Contact inhibition: Signaling from cadherin-based adherens junctions stops cells from cycling. They arrest in G1.
Cancer
Transformed cells continue to cycle
FIGURE 41.11 LOSS OF GROWTH CONTROL IN TRANSFORMED CELLS.
proliferation. In fact, many cancer-causing mutations that disturb growth control pathways are actually deleterious in normal cells and cause them to undergo senescence or commit suicide by apoptosis (see Chapter 46). Dysregulation of cell proliferation in the G1 phase causes most types of cancer. This is readily seen in the laboratory when cells are grown on plastic tissue culture dishes. Most normal cells proliferate until they cover the surface completely, forming a monolayer. When the monolayer is confluent (ie, when cells are touched by other cells on all sides), signaling initiated by cadherin proteins (see Fig. 30.7) causes cells to express p27Kip1 and to arrest their cell-cycle progression in G1. This is called contact inhibition of growth (see Chapter 30, in the section titled “Cadherin Family of Adhesion Receptors”). Cancer cells lack this control, so they keep proliferating and piling up on top of one another as long as nutrient and mitogen supplies last (Fig. 41.11). Cells that lose this aspect of growth regulation are said to be transformed. Malfunction of the restriction point is an extremely common contributor to transformation. Indeed, one or more components of the p16/cyclin D/Cdk-4/Rb system are mutated in most human cancers. In addition, several cancer-causing viruses, such as simian virus 40 (SV40), papillomaviruses, and adenovirus, make proteins that facilitate the G1 → S transition by binding Rb and liberating E2F. Most cancer cells have abnormalities in the activities of two classes of genes. Oncogenes are genes whose inappropriate activation can cause oncogenic (cancerous) transformation of cells. The protein products of most oncogenes regulate cellular growth and proliferation. They are typically components of signal trans duction pathways that are controlled by feedback mechanisms. Tumor suppressors are genes whose inactivation can lead to cancerous transformation. Their
protein products typically inhibit products of oncogenes or negatively regulate cell proliferation. Several genes that are involved in restriction point control can act as either oncogenes or tumor suppressors. More than 100 oncogenes have been identified thus far. Most normally function in signal transduction pathways that lie downstream of mitogens that stimulate cell-cycle progression. Their inappropriate activation can mimic the effects of persistent mitogenic stimulation, thereby uncoupling cells from normal environmental controls and leading to uncontrolled proliferation and cancer. For example, Ras activates the MAP/ERK kinase cascade and accumulation of cyclin D (Fig. 41.9). Ras genes are mutated in approximately 15% of human cancers. Inappropriate activation of Ras tricks the cell into thinking that it is receiving mitogenic signals, leading it to express cyclin D, phosphorylate Rb, and proliferate. Fortunately, in most cells, mutations that result in uncontrolled proliferation usually lead to DNA damage (oncogenic stress) and activate a protective pathway leading to senescence. Other proteins involved in restriction point control can also act as oncogenes if hyperactivated. These include E2F1, cyclin D (overexpressed in 50% of breast cancers), and Cdk4. In each case, activation of the protein causes inappropriate transcription of genes promoting cell-cycle progression, bypassing the restriction point, and leading to uncontrolled cell cycles and cancerous transformation (Fig. 41.12). Rb is one of the best-characterized tumor-suppressor genes. As discussed earlier, a primary function of Rb is to block cell-cycle progression until sustained mitogenic stimulation results in its inactivation. It is therefore not surprising that loss of Rb can lead to inappropriate cellcycle progression and cancer. Rare individuals who inherit one defective Rb gene tend to develop retinoblastomas as children and osteosarcomas as adults. The cancer arises when the “good” allele is inactivated in a proliferating cell (this is called a somatic mutation). Such cancers are rare and occur only later in life in individuals who inherit two good Rb genes, as two independent somatic mutations (two “hits”) are required in the same proliferating cell. Homozygous loss of Rb is lethal during embryogenesis. This is partly because under some circumstances, the unleashed E2F can act as a potent inducer of apoptotic cell death (see Chapter 46). P16Ink4a is another important tumor suppressor involved in G1 growth control. Normally, it suppresses Cdk4/6 activity in nondividing cells (see next section; also see Chapter 40), thereby reinforcing the ability of Rb to maintain the growth arrest of G1 cells (Fig. 41.12). Mutations in the p16Ink4a gene are very common in cancer, but this is partly because this gene is fascinatingly complex (Fig. 41.16). Mutations in other INK4 Cdk inhibitors and the CKI protein p27Kip1 are also found in cancer, although less frequently.
723
Transcription prevented
=
Transcription of cyclins A, E, Cdk 1 DNA replication genes
=
Normal cell arrested at restriction point
+1
STOP
No external signals
External signals
Rb
Cdk 4/6–cyclin D (active)
Normal cell passes restriction point GO
Rb
+1
Rb mutant cell passes restriction point No external signals
GO
Rb
Danger +1
No external signals Active oncogene mimics external signals
Cdk 4/6–cyclin D (active)
GO
Danger +1
Cdk 4/6–p16Ink4a (inactive) External signals may be present
Differentiated cell expressing p16 does not cycle STOP
Rb
*p16Ink4a Cdk 4/6
External signals may be present
Differentiated cell mutant for p16 passes restriction point
Cdk 4/6–cyclin D (active) Rb
GO
Nutrients Mitogens Cdk2–cyclin E Cdk4/6–cyclin D
M
G1
S
Restriction point
Emi1 SCFSkp2 SCFβ-TrCP
APC/CCdh1
Degraded proteins:
Cyclins
Cdc25A
p27Kip1
FIGURE 41.13 PROTEOLYTIC ACTIVITIES IN G1 PHASE.
Cell with active oncogene passes restriction point Rb
Increasing ubiquitin ligase activity
KEY
Increasing Cdk activity
CHAPTER 41 n G1 Phase and Regulation of Cell Proliferation
Danger +1
FIGURE 41.12 HOW ACTIVATED ONCOGENES OR MUTATIONS IN THE RB OR P16 TUMOR SUPPRESSOR PROTEINS CAN LEAD TO ABNORMAL PASSAGE OF THE RESTRICTION POINT AND CANCER.
Proteolysis and G1 Cell-Cycle Progression Just as controlled destruction of proteins is key to the transition of cells from mitosis to the G1 phase (see Chapter 40), proteolysis also fulfills several key roles during progression from G1 into the S phase (Fig. 41.13). For example, when Cdk2–cyclin E is activated following cyclin D synthesis, it phosphorylates its inhibitor p27Kip1.
Phosphorylated p27Kip1 is recognized by a specific E3 ubiquitin ligase called SCFSkp2 (see Figs. 40.15 and 40.16). The resulting destruction of p27Kip1 permits a burst of Cdk2-cyclin E activation in a positive feedback loop that rapidly amplifies Cdk activity at the initiation of the S phase. Later in the S phase, phosphorylation of the DP subunit of E2F causes its dissociation from DNA, recognition by SCF, and destruction. This is essential to complete S phase. SCF also targets cyclins D1 and E for destruction, the former when mitogens are limiting and the latter following autophosphorylation during progression through the S phase. Progression throughout G1 requires the activity of a second E3 ubiquitin ligase known as the anaphasepromoting complex or cyclosome (APC/C). A specific cofactor, Cdc20, activates the APC/C to trigger the metaphase-to-anaphase transition (see Chapter 40). As the cell leaves mitosis, Cdh1 replaces Cdc20 and targets many cell-cycle regulatory proteins for degradation, including Cdc20, A- and B-type cyclins, and factors involved in DNA replication. Destruction of these target proteins during mitotic exit and G1 (Fig. 41.13) likely contributes to the requirement for transcription and de novo synthesis of these proteins at the moment cells decide whether to enter or not a new cycle of genome replication.
Integrity of Cellular DNA Monitored by a G1/S Checkpoint The S phase is a point of no return in the history of any dividing cell. Given the semiconservative mechanism of DNA replication, whereby existing DNA strands serve as templates for newly synthesized strands, any DNA defect that passes unnoticed through the S phase becomes
724
SECTION X n Cell Cycle Undamaged DNA
(ATM)2 inactive
ATR not signaling
E2F drives expression of cell cycle genes
p53 present in low amounts in cytoplasm
dsDNA breaks
(ATM)2
E2F active
ssDNA
ATM active Mdm2 inactive
ATR active localized to damage site p53 active in nucleus
Chk1 kinase Cdc25A
Normal cell-cycle progression
Apoptosis
p21
Degraded
Stable cellcycle arrest
Rapid cellcycle arrest
FIGURE 41.14 THE G1/S CHECKPOINT AND DNA DAMAGE RESPONSE.
perpetuated as a mutation that is transmitted to all future progeny of the cell. Furthermore, any single-stranded nick in DNA may become a full-fledged chromosome break if present during replication. To avoid these problems, cells have a quality control mechanism to block entry into the S phase if damaged DNA is detected. This quality control mechanism involves a checkpoint that operates throughout the G1 and S phases (Fig. 41.14). Checkpoints are biochemical circuits superimposed on the normal cell cycle. When activated, the G1/S checkpoint triggers a DNA damage response that blocks cell-cycle progression. The block may be temporary but, in some cases, checkpoint activation leads to senescence or cell death by apoptosis. Sensor proteins detect DNA damage and activate this checkpoint. In the subsequent DNA damage response, the sensor proteins activate protein kinases and a key transcriptional regulator that block cell-cycle progression (see Fig. 40.4). The DNA damage response has fast and slow components: The former is analogous to applying the brakes in a car; the latter is analogous to removing the wheels and putting it up on blocks. Both components start with the protein kinases, ATM and ATR (see Fig. 40.4). ATM and ATR are related to the lipid kinase phosphatidylinositol 3-kinase (see Fig. 26.7), but their only known substrates are proteins (see Chapter 40). People lacking ATM have the disease ataxia-telangiectasia, which is characterized by immunodeficiency, photosensitivity, cerebellar degeneration, and an elevated incidence of leukemias and lymphomas. Loss of ATR is fatal. DNA damage that disrupts ongoing DNA replication and produces single-stranded DNA activates ATR, which phosphorylates and activates a downstream kinase called Chk1. Chk1 targets the essential phosphatase CDC25A (Fig. 41.14), marking it for destruction. Because CDC25A is required to remove inhibitory phosphate groups from inactive Cdks, its destruction applies a rapid brake to cell-cycle progression.
DNA double-strand breaks vigorously activate ATM, which directly and indirectly stabilizes and activates a critical tumor suppressor, p53. This transcription factor has been called the “guardian of the genome,” because in response to DNA damage it also applies a rapid brake to cell cycle progression. In some cases this arrest leads to senescence, a permanent cessation of the cell cycle (putting the car up on blocks). p53 is very powerful medicine for the cell cycle and must be carefully regulated by a ubiquitin ligase (E3) called Mdm2 (mouse double-minute 2; the human ortholog is Hdm2) that keeps p53 levels low when the cell cycle is running normally (Fig. 41.15A). Both p53 and Mdm2 protein shuttle in and out of the nucleus (see Chapter 9). When the two proteins associate in the cytoplasm, Mdm2 promotes rapid degradation of p53 by the ubiquitin/proteasome system (see Chapter 23). Because p53 directly stimulates expression of Mdm2, a feedback loop keeps levels of p53 low. Loss of the Mdm2 gene in mice is lethal unless the p53 gene is also lost. Both p53 and Mdm2 are phosphorylated following DNA damage (Fig. 41.15B). These phosphorylations prevent Mdm2 from binding, so p53 is stabilized, and its concentration in the nucleus increases dramatically. The phosphorylations also make p53 a more potent transcriptional activator. The result is a burst of transcription of p53-regulated genes. p53 is rapidly activated in response to DNA damage resulting from hyperproliferation of cells following loss of restriction point control (oncogenic stress). It also responds to DNA damage induced by the environment. If the damage is rapidly repaired, cells continue to cycle, but if the damage is too severe, p53 induces senescence or apoptotic cell death (see next paragraph). In the case of oncogenic stress following loss of restriction point control (eg, by Ras mutations) E2F stimulates the expression of the tumor suppressor protein p19Arf (alternate reading frame). This binds and sequesters Mdm2 in the nucleolus (Fig. 41.15C), allowing p53
CHAPTER 41 n G1 Phase and Regulation of Cell Proliferation
A. Healthy cell
B. Irradiation
C. Oncogene activation
Mdm2 ubiquitin ligase directs destruction of p53 in cytoplasm
ATM phosphorylates and activates p53, which blocks Mdm2 binding
First the irradiation pathway is activated, then phosphorylation of E2F promotes transcription of p19Arf, which sequesters Mdm2 in nucleoli, releasing active p53
Nucleolus
DNA damage ATM
p53-Mdm2
p
p53-Mdm2 Ub Ub Ub p53-Mdm2
5 3 + Mdm2
p53 levels low throughout cell
p19Arf Mdm2
E2F
p53 activated
Mdm2 cannot bind
p53
725
p19Arf Mdm2 p53 activated p53
Active nuclear p53 drives expression Active nuclear p53 drives expression of proteins that arrest the cell cycle of proteins that arrest the cell cycle and promote cell death (apoptosis) and promote cell death (apoptosis) p53 activated Mdm2
Active p53 also drives expression of the ubiquitin ligase Mdm2, providing negative feedback on p53
FIGURE 41.15 P53 REGULATION AND THE DNA DAMAGE CHECKPOINT IN G1. A, Healthy cell. B, After irradiation, Mdm2 (mouse double-minute 2) can no longer bind p53, which accumulates in active form in the nucleus. C, After oncogene activation, Mdm2 is sequestered in the nucleolus, and active p53 accumulates in the nucleus. Activated p53 can induce either cell-cycle arrest or cell death.
to accumulate in the nucleoplasm. There, it activates transcription of the CKI p21, stopping the cell cycle. If this arrest is prolonged, p16INK4A is induced, activating Rb and leading to permanent cell cycle arrest (senescence). p53 activation can also induce apoptosis by activating transcription of genes for proapoptotic proteins, including Bax, BH3-domain proteins, Puma, CD95 (Fas/ Apo1), and Apaf-1 (Fig. 41.15; also discussed in Chapter 46). The decision whether to induce senescence or death is very complex and may be regulated by different posttranslational modifications of p53. One way or the other, p53 serves its function as guardian, as the outcome is that aberrantly proliferating cells are either permanently silenced or removed, and the body is protected. p53 is mutated or deleted in about half of all human cancers. Families that carry a mutated p53 allele have Li-Fraumeni syndrome with an elevated risk of cancers. Mice lacking p53 are viable but a defective G1 DNA damage checkpoint results in cancers while young. This illustrates a common theme that in many cases checkpoint components are not essential for life as long as nothing untoward occurs. Checkpoints exist primarily to deal with problems that arise during cell-cycle progression. However, the elevated cancer rates in Li-Fraumeni syndrome patients indicate that although p53 is not essential for the passage of every cell cycle, it is essential for long-term genetic stability and for maintaining a proper balance among cell proliferation, differentiation, and death during the lifetime of a mammal. The p19Arf protein (in humans, the protein is smaller and so is called p14Arf) is extremely unusual, as it is encoded in a common gene with the Cdk inhibitor
A. One gene, two promoters 1α
1β
B. Two key proteins
2
Rb
Ink4a Arf
Ink4a/Arf
p16Ink4a
Cdk4-cyclin D1
3
p19Arf
Mdm2
p53
FIGURE 41.16 DUAL CONTROL OF G1 PROGRESSION BY THE P16INK4A/P19ARF GENE. This gene encodes two completely different proteins that are key to avoiding cancer. A, The intron–exon structure of the p16Ink4A/p19Arf gene. B, p16Ink4A and p19Arf negatively regulate the restriction point via Rb and the DNA damage checkpoint via p53, respectively.
p16Ink4a (Fig. 41.16). Despite having different promoters (that respond to different stimuli), the two genes not only overlap but also share a common exon. Nevertheless, the two proteins have no common amino acid sequences because the shared exons are read in different frames in the mature mRNAs (mRNAs) for the two proteins. Thus, the p16Ink4a/p14Arf locus encodes two vital protective factors with different jobs. It is not surprising that mutations in this key locus are found in between 25% and 70% of human cancers.
G1 Regulation: A Matter of Life and Death To commit to a new cycle of proliferation, cells must pass through the restriction point gate. The key to this
726
SECTION X n Cell Cycle
gate is the phosphorylation of Rb by Cdks, so signals such as mitogens that activate Cdks set up a feedback loop that promotes passage of the gate. Of course, in the real world, accidents happen, and the G1/S checkpoint and DNA damage response provide a way to block cell-cycle progression even in the presence of growth factors and mitogens. The complex G1 regulatory networks have a potential impact on all of us. If they are disrupted by mutations or damage, the result is cancer. In fact, very few cancers have intact restriction point control networks. ACKNOWLEDGMENTS We thank Jiri Bartek, Ludger Hengst, Keisuke Kaji, and Marcos Malumbres for their suggestions on revisions to this chapter. SELECTED READINGS Blanpain C, Fuchs E. Epidermal stem cells of the skin. Annu Rev Cell Dev Biol. 2006;22:339-373. Bryder D, Rossi DJ, Weissman IL. Hematopoietic stem cells: The paradigmatic tissue-specific stem cell. Am J Pathol. 2006;169:338-346. Cardozo T, Pagano M. The SCF ubiquitin ligase: Insights into a molecular machine. Nat Rev Mol Cell Biol. 2004;5:739-751. Chandler H, Peters G. Stressing the cell cycle in senescence and aging. Curr Opin Cell Biol. 2013;25:765-771. Chen HZ, Tsai SY, Leone G. Emerging roles of E2Fs in cancer: an exit from cell cycle control. Nat Rev Cancer. 2009;9:785-797. Cheung TH, Rando TA. Molecular regulation of stem cell quiescence. Nat Rev Mol Cell Biol. 2013;14:329-340.
Childs BG, Baker DJ, Kirkland JL, Campisi J, van Deursen JM. Senescence and apoptosis: dueling or complementary cell fates? EMBO Rep. 2014;15:1139-1153. Clevers H. The intestinal crypt, a prototype stem cell compartment. Cell. 2013;154:274-284. Dick FA, Rubin SM. Molecular mechanisms underlying RB protein function. Nat Rev Mol Cell Biol. 2013;14:297-306. Fuchs E, Tumbar T, Gausch G. Socializing with the neighbors: Stem cells and their niche. Cell. 2004;116:769-778. Goldstein M, Kastan MB. The DNA damage response: implications for tumor responses to radiation and chemotherapy. Annu Rev Med. 2015;66:129-143. Jackson SP, Bartek J. The DNA-damage response in human biology and disease. Nature. 2009;461:1071-1078. Johnson A, Skotheim JM. Start and the restriction point. Curr Opin Cell Biol. 2013;25:717-723. Rando TA. Stem cells, ageing and the quest for immortality. Nature. 2006;441:1080-1086. Sage J. The retinoblastoma tumor suppressor and stem cell biology. Genes Dev. 2012;26:1409-1420. Scadden DT. The stem-cell niche as an entity of action. Nature. 2006;441:1075-1079. Sherr CJ. The INK4a/ARF network in tumour suppression. Nat Rev Mol Cell Biol. 2001;2:731-737. Shi X, Garry DJ. Muscle stem cells in development, regeneration and disease. Genes Dev. 2006;20:1692-1708. Silverman JS, Skaar JR, Pagano M. SCF ubiquitin ligases in the maintenance of genome stability. Trends Biochem Sci. 2012;37:66-73. Sperka T, Wang J, Rudolph KL. DNA damage checkpoints in stem cells, ageing and cancer. Nat Rev Mol Cell Biol. 2012;13:579-590. Takahashi K, Yamanaka S. A decade of transcription factor-mediated reprogramming to pluripotency. Nat Rev Mol Cell Biol. 2016;17: 183-193. Veit B. Stem cell signalling networks in plants. Plant Mol Biol. 2006; 60:793-810.
CHAPTER
42
S Phase and DNA Replication A
ccurate replication of DNA, which is crucial for cellular propagation and survival, occurs during the S phase (DNA synthesis phase) of the cell cycle. This chapter begins with a brief primer on the events of replication and then discusses its regulation. Next, the chapter covers the proteins at origins of replication that ensure that each region of DNA is replicated once and only once per cell cycle. It closes by discussing how the structure of the nucleus influences replication.
DNA Replication: A Primer One of the most exciting byproducts of the WatsonCrick model for the structure of DNA was a predicted mechanism for DNA replication. Because DNA strand pairing is determined by complementary base pairing, it was logical to propose the existence of “DNA polymerases,” enzymes that would move along a single strand of DNA, recognize each base in turn, and insert the proper complementary base at the end of the growing chain. Thus, one might have surmised that only a single enzyme was required for DNA synthesis. In fact, DNA replication in cells involves a complex macromolecular machine. In the chemical reaction of DNA replication, the 3′ hydroxyl at the end of the growing DNA strand makes a nucleophilic attack on the α-phosphate of the incoming nucleoside triphosphate to form a phosphodiester bond. This incorporates the nucleotide into the growing chain and releases pyrophosphate (Fig. 42.1). This reaction requires the presence of a template strand of DNA that specifies, via base pairing, which of the four nucleoside triphosphates is added to the growing molecule. The exact site on the chromosomal DNA where replication begins is termed the origin of bidirectional replication. As the term bidirectional implies, two sets of DNA replication machinery head off in opposite directions from the origin. At the replication fork, one parental DNA molecule splits into two daughters (Fig. 42.2). The
protein complex associated with the fork that is actively replicating the DNA is known as the replisome. Accumulating evidence suggests that the replisome is stationary at the fork as it “reels in” replicating DNA rather than moving along the DNA like a train on a track. Replicating DNA
Mechanism of chain elongation O O H2C
P
O O– O
Base 1
H H O O H2C
P
O
O– O
Base 2
H H O O H2C
Growing DNA chain
H H H
P
O
O– Base 3
O
H H
H H
OH Nucleophilic attack
Free 3'-hydroxyl end
H H H
O
H O
O O H2C
P
O
P
O
P
O– O–
O–
O– Base 4
O
H H OH
Entering dNTP
H H H
FIGURE 42.1 MECHANISM OF DNA POLYMERIZATION. A 3′ OH group at the end of a growing DNA chain makes a nucleophilic attack on the α-phosphate of a triphosphate precursor in the active site of polymerase (enzyme not shown here). dNTP, deoxynucleoside triphosphate.
727
728
SECTION X n Cell Cycle
Lagging strand polymerase complex Polymerase/ primase
Lagging strand
5' 3'
Origin of bidirectional replication
Nascent DNA (Okazaki fragment)
RNA primer
CMG Helicase
RPA Replication fork
3' 5'
Replisome
5'
Nascent DNA
3'
Leading strand
Leading strand polymerase complex
FIGURE 42.2 KEY COMPONENTS AND EVENTS AT THE REPLICATION FORK. All enzymes are closely associated in the replisome complex, but are shown separate here for clarity.
The bidirectional nature of DNA replication causes a fundamental problem, as the chemical reaction of DNA synthesis invariably proceeds in a 5′ to 3′ direction. Replication of the so-called leading strand, in which DNA polymerase ε moves in a 3′ to 5′ direction along the template (laying down nascent DNA in a 5′ to 3′ direction) poses no problems; the polymerase simply chases behind the replication fork (Fig. 42.2). However, the other template strand faces in the opposite direction, apparently requiring a DNA polymerase to synthesize DNA in the wrong direction as the replication fork progresses away from the origin (ie, adding nucleotides in a 3′ to 5′ direction). No DNA polymerase with this polarity has been found. Instead, this lagging strand replicates in a series of short segments. Every time the DNA strands have been peeled apart (unwound) by 200 nucleotides or so, probably corresponding to the unwinding of a single nucleosome, a polymerase/ primase complex (Figs. 42.2 and 42.12) initiates DNA synthesis on the lagging strand, and DNA polymerase δ runs away from the fork, back toward the replication origin, again synthesizing nascent DNA in a 5′ to 3′ direction. Thus, lagging strand synthesis proceeds in bursts in a direction opposite to the overall direction of fork movement. Synthesis of each lagging strand fragment stops when DNA polymerase runs into the 5′ end of the previous fragment, displacing the RNA primer with which it was initiated (see later). Thus, the lagging strand is copied in a highly discontinuous fashion into short fragments known as Okazaki fragments (named after their discoverer [Fig. 42.2]). Fig. 42.12 describes the enzymes and events at the replication fork in greater detail.
Origins of Replication Bacteria such as Escherichia coli replicate their circular chromosomes using two replication forks starting from a single origin of replication (Fig. 42.3A), but eukaryotes must use multiple origins of replication to duplicate
A. E. coli chromosome Origin oriC
500,000 bp
B. Portion of eukaryotic chromosome Active origin
Dormant origins
Active origin
Dormant origins
130,000 bp FIGURE 42.3 A, The Escherichia coli chromosome is a simple replicon with a single origin of replication. In cells, this chromosome has a complex, highly supercoiled structure. B, Eukaryotic chromosomes have multiple origins of replication, most of which remain dormant unless needed.
their large genomes during the S phase, which can be as short as a few minutes in some early embryos. These numerous origins are distributed along the chromosome: up to 600 to 700 in budding yeast and more than 100,000 in human cells. The origins are distributed so that all the DNA is replicated in the available time, and many more origins are prepared than are actually needed. How is the “firing” of all these origins orchestrated so that each is used no more than once per S phase? Cells manage this problem by a mechanism termed licensing, which ensures that each segment of DNA is replicated just once per cell cycle. Replication of the origin removes the license, which cannot normally be renewed until the cell has completely traversed the cycle and has passed through mitosis. The portion of chromosomal DNA replicated by the two bidirectional forks initiated at a single origin is
CHAPTER 42 n S Phase and DNA Replication
termed a replicon. The classic replicon is the E. coli chromosome (which is 4 × 106 base pairs [bp] in size) with a single genetically defined replicator site called oriC (Fig. 42.3). An initiator protein (product of the E. coli DnaA gene [Fig. 42.13]) binds to this origin and either directly or indirectly promotes melting of the DNA duplex, giving the replication machinery access to two single strands of DNA. Other factors unwind the DNA, leading to the full assembly of the replisome, which powers a wave of DNA replication proceeding outward in both directions along the DNA (a replication “bubble”) at approximately 750 to 1250 bases per second. An average human chromosome contains approximately 150 × 106 bp of DNA. Because the replication machinery in mammals moves only approximately 20 to 40 bases per second (partly reflecting the fact that the DNA is packaged into chromatin and partly reflecting the slower speed of the eukaryotic replisome), it would take up to 2000 hours to replicate this length of DNA from a single origin. In most human cells the S phase takes approximately 10 hours. This means that at least 25 to 125 origins of replication are required to replicate an average chromosome in the allotted time. In fact, origins of replication are much more closely spaced than this. It is estimated that 30,000 to 50,000 origins of replication “fire” during each cell cycle to replicate the entire human genome, but many more dormant origins are licensed for use if necessary. These dormant origins are essential for resolving replication stress (see later). To explain the events at origins of replication, the budding yeast Saccharomyces cerevisiae serves as a good example. Its DNA replication is better understood than that of any other eukaryote and although its origins of replication are specialized, the proteins that act on them are conserved across metazoa.
Replication Origins in Saccharomyces cerevisiae Approximately 400 origins of replication participate in replicating the budding yeast genome. A major breakthrough in understanding DNA replication in S. cerevisiae was the identification of short (100 to 150 bp) segments of DNA that act as replication origins in vivo when cloned into a yeast plasmid (circular DNA molecule). These autonomously replicating sequences (or ARS elements) allow yeast plasmids to replicate in parallel with the cellular chromosomes (Fig. 42.4). ARS elements are often, although not always, bona fide replication origins in their native chromosomal context. Replication always initiates within ARS elements, but not all ARS elements act as origins of DNA replication in every cell cycle. Yeast replication origins are spaced approximately every 30,000 bp, with a maximum separation of approximately 130,000 bp. Even this longest interval should replicate easily within the 30 minutes available during
729
A
Plasmid
Selectable marker gene
Introduce into yeast cells
Grow cells under selective conditions
Introduce into yeast cells
Grow cells under selective conditions
B ARS
Plasmid
Selectable marker gene
FIGURE 42.4 PLASMID ASSAY FOR IDENTIFICATION OF AN AUTONOMOUSLY REPLICATING SEQUENCE (ARS) ELEMENT (ORIGIN OF DNA REPLICATION) IN BUDDING YEAST. The plasmid has a selectable marker gene (eg, a gene required for the synthesis of an essential amino acid) plus (in panel B) an ARS element. This plasmid is transferred into growing yeast cells that are defective in the marker gene carried by the plasmid; these cells are then plated out on agar medium that lacks the essential amino acid. A, A plasmid lacking an ARS fails to replicate and is lost from the cells. These cells cannot grow into colonies on plates that lack the essential amino acid. B, If the plasmid contains an ARS element, it replicates along with the chromosomal DNA and is maintained in the population. These cells grow into colonies in the absence of the essential amino acid.
the S phase. Because the number of origins exceeds the number required to replicate the genome within the allotted time, some origins need not “fire” every cell cycle. The probability that any given origin will be used in a given cell cycle ranges from less than 0.2 to greater than 0.9. It is important to note that replication by a passing fork coming from an adjacent origin inactivates dormant origins. This prevents re-replication of genomic regions during the cell cycle. The ARS element does two things to establish an origin of replication. First, its conserved sequences act as binding sites for a protein complex that marks it as a potential origin. Second, it has nearby sequences that are readily induced to unwind by separating the basepaired strands. Budding yeast ARS elements share a common DNA sequence motif called the ARS core consensus sequence: 5′-(A/T)TTTAT(A/G)TTT(A/T)-3′ (Fig. 42.5). Single base mutations at several locations within this sequence completely inactivate ARS activity. Other, less well-conserved DNA sequences also contribute to the activity of the ARS as a replication origin. One of these, termed B1, together with the ARS core, forms the bind ing site for a complex of six proteins termed the origin
730
SECTION X n Cell Cycle
ORC complex ABF-1
B3
OBR (origin of bidirectional replication) B2 DNA unwinding element
ATP
B1
EdU
PCNA (t + 30 min)
Merge
90° ATP
ARS core A A A T T T A TG T T TT T
FIGURE 42.5 ORGANIZATION OF THE AUTONOMOUSLY REPLICATING SEQUENCE (ARS)-1 ELEMENT. The ORC (origin recognition complex) binds to the ARS core sequence plus element B1. B2 is a sequence that can readily be induced to unwind. The OBR (origin of bidirectional replication) is the site where DNA synthesis actually begins. B3 is a binding site for an auxiliary factor called ABF-1 that is both a transcriptional activator and an activator of the ARS element. ATP, adenosine triphosphate. (For reference, see Protein Data Bank [PDB; www.rcsb.org] file 4X6C.)
recognition complex (ORC [see later section]). The DNA unwinding element is another short sequence (B2) located a bit further along the DNA. DNA synthesis begins at an origin of bidirectional replication midway between the ORC binding site and the DNA unwinding element. ORC was identified by its ability to bind the 11-bp ARS core sequence (Fig. 42.5). This binding has two noteworthy features. First, the subunits of the ORC complex are AAA adenosine triphosphatases (ATPases; see Box 36.1) and adenosine triphosphate (ATP) hydrolysis is required for ORC to bind ARS DNA. Second, in yeast, the ORC complex remains bound to the origins of replication across the entire cell cycle. Thus, something other than the presence of ORC must regulate the periodic activation of origins in the S phase (Fig. 42.14). In some metazoan cells, ORC behavior is more complex—for example, the largest subunit, Orc1, is degraded during part of the cell cycle. ARS elements often contain binding sites for other sequence-specific DNA binding proteins, including transcription factors. For example, a transcription factor called ARS-binding factor 1 (ABF-1) binds to the B3 sequence within the ARS1 element (Fig. 42.5). Deletion of the ABF-1 binding site only slightly reduces the ability of ARS1 to act as a replication origin in vivo and other transcription factors can substitute for ABF-1. In addition to their role in DNA replication, several ORC subunits also regulate heterochromatin formation and transcription (see Chapters 8 and 10). This crosstalk between the machinery used for transcription and DNA replication may explain why regions of chromosomes with actively transcribed genes typically replicate early in the S phase (see the next section below). In some metazoan cells, the Orc6 subunit also functions during cytokinesis, apparently via interactions with cytoskeletal filament proteins independent of its role at replication origins.
5 µm
FIGURE 42.6 SUPERRESOLUTION VIEW OF ACTIVE REPLICONS IN A HELA (HENRIETTA LACKS) CELL. EdU (red) was used to label sites of active replication for 15 minutes followed by washing out for 15 minutes. Then antibody recognizing proliferating cell nuclear antigen (PCNA) was used to stain all active replicons in the cell. Thousands of replicons can be seen, and it is clear that in the 30 minutes between the labelling of the red and green channels, many new origins of replication have been activated. Scale bar = 5 µm. OMX superresolution. (Microscopy by Vadim Chagin and Cristina Cardoso, Technical University of Darmstadt, Germany.)
Replication Origins in Mammalian Cells Less is known about the structure and function of mammalian origins of DNA replication than about ARS elements in budding yeast. Attempts to develop a mammalian equivalent to the yeast ARS assay had few successes. Over the years approximately 30 metazoan origins of replication were identified by painstaking methods, but more recently, high throughput methods, including the sequencing of short nascent strands of replicating DNA have identified thousands of mammalian replication origins. Superresolution microscopy can now resolve the thousands of replicons active at any one time in a human cell (Fig. 42.6). Overall, the chromatin landscape influences mammalian origins. Important factors include DNA sequence, DNA modifications, chromatin structure, and nuclear organization. Although high throughput sequencing revealed no sequences as specific as the yeast ARS elements, many origins are associated with short regions of G-rich DNA. These G-rich regions can form specialized structures that have fewer nucleosomes and are therefore more accessible to the DNA replication machinery. Indeed, replication origins tend to be located near gene promoters, where the density of nucleosomes is low. Such actively transcribed regions of the genome are packaged into euchromatin with modified histones and tend to replicate earlier during S phase than regions of the genome that are not transcribed. A typical mammalian replicon encompasses about 130 kb and contains approximately four or five licensed origins-only, one of which is typically used. The first replicon to be mapped lies just downstream of the hamster gene for dihydrofolate reductase, an enzyme that is essential for biosynthesis of thymidine. Investigators selected cells with this chromosomal region amplified as hundreds or even thousands of copies (Fig. 42.7) and looked for the first regions of the amplified DNA to
CHAPTER 42 n S Phase and DNA Replication
Mapping a cellular replication origin Add high concentration of methotrexate Cultured cells
731
CYTOPLASM Cdt1
Cells die Geminin degraded Cdks inactive
Add gradually increasing concentration of methotrexate over many generations
Cells live in high concentration of methotrexate
Prereplication complex ORC Cdc6
Cdt1
Geminin
Cdk phosphorylation triggers other CMG components to assemble
Mcm2-7 double hexamer 5’
3’ 3’
Cdc45
5’
GINS
Loop domain containing dihydrofolate reductase gene Normal chromosome
Chromosome with amplified domain
Base of domain
Potential OBRs (origins of bidirectional replication)
DHFR gene
Dormant Dormant origins Active origins origin
–30
–20
–10
0
10
Dormant Active origins origin
γ
β
DHFR gene
20
30
40
Number of bases (kb)
Restriction point
NUCLEUS
Next gene 50
60
70
Initiation zone FIGURE 42.7 MAP OF DNA REPLICATION LANDSCAPE NEAR THE DIHYDROFOLATE REDUCTASE (DHFR) GENE. Normal cells are killed by exposure to methotrexate, but it is possible to select resistant cell lines by growing them in progressively increasing concentrations of the drug, selecting at each stage for cells that survive. Use of this procedure on hamster cells resulted in a cell line that contains approximately 1000 copies of a 230,000-bp chromosomal domain containing the DHFR gene. This region of DNA is replicated using origins found within a 55,000-bp region adjacent to the DHFR gene. Most initiation occurs at two specific origins, called β and γ, but other dormant origins are scattered throughout the entire 55,000-bp region.
replicate. They found that DNA replication can initiate with low efficiency at a number of sites distributed across a broad region of approximately 55,000 bp. Two of these sites, termed Ori-β and Ori-γ (Fig. 42.7) account for approximately 20% of all initiation in the region. The emerging view is that the replication machinery is highly conserved between budding yeasts and vertebrates but that the location of replication origins is more flexible in vertebrates. This might reflect both the differing chromatin landscapes across metazoan genomes (see Chapter 8) and the diverse range of cell cycles required to make a complex metazoan.
Origin Licensing and Assembly of the Prereplication Complex To preserve the integrity of the genome, each origin of replication must “fire” no more than once per cell cycle. We now have a reasonable understanding of the various
G1
CMG helicases unwind DNA S
FIGURE 42.8 COMPONENTS OF THE PREREPLICATION COMPLEX BEFORE AND AFTER THE INITIATION OF DNA REPLICATION. CMG, Cdc45, Mcm2–7, and GINS (Go-Ichi-Ni-San). (For reference, see PDB files 3JA8 [for CMG helicase with Cdk45 bound], 3JC5 [for Mcm2–7], and 5F9R [for ORC].)
solutions to this problem that have been reached by yeasts and vertebrates. Recall that yeast ORC is stably bound to replication origins throughout the cell cycle. However, ORC is not the trigger for DNA replication. Rather, it acts as a “landing pad” for assembly of a prereplication complex of other proteins that initiates DNA replication. Formation of the prereplication complex “licenses” each origin for a single initiation event as follows. During late anaphase or very early G1 phase, several proteins, including Cdc6 and Cdt1, bind to the ORC complex at origins of replication (Table 42.1). Before the onset of the S phase ORC-Cdc6-Cdt1 loads a double hexamer of Mcm proteins (minichromosome maintenance) on the origin DNA (Fig. 42.8). Mcm proteins were identified in a screen for genes of budding yeast that are required for the stability of small artificial chromosomes. Six of these Mcm genes encode a structurally related group of proteins, termed Mcm2–7, that is required for DNA replication. ORC-Cdc6–Cdt1 uses ATP hydrolysis to crack open two hexameric Mcm2-7 rings so they can wrap around DNA in a head-to-head orientation. Replication starts when each Mcm2–7 hexamer associates with the GINS (Go-Ichi-Ni-San; 5-1-2-3 in Japanese) complex and the Cdc45 protein to form the replicative CMG helicase (Cdc45-Mcm-GINS). This helicase forms the core of the replisome and uses ATP hydrolysis to separate DNA strands (Fig. 42.8). As the two CMG helicases move away from each other, the origin is converted from a licensed to an unlicensed state. Because origin licensing by Mcm2–7 loading can only occur prior to entry into S phase, origins only fire once per cell cycle. In mammals, licensing occurs in the early G1 phase before passage of the restriction point (see Chapter 41). At the exit from mitosis, destruction of cyclins and synthesis of inhibitory proteins inactivates Cdks. This creates a window of time between anaphase and the restriction
732
SECTION X n Cell Cycle
TABLE 42.1 Biochemical Activities Required for Replication of DNA in Eukaryotes Activity
Name of Protein
Origin recognition
ORC (origin recognition complex; five of six subunits are AAA ATPases)
Prereplication complex
Cdc6 (recruits Mcm2–7) Cdt1 (recruits Mcm2–7)
Origin activation
Cdc7-Dbf4 Cdk (in human, Cdk2-cyclin A) phosphorylation of Sld2/RecQ4 (yeast/human names) and Sld3/treslin recruits Dbp11/TopBP1 and initiates CMG assembly to trigger the actual start of replication
DNA unwinding (helicase)
CMG helicase is made up of: • Mcm2–7 (assemble double hexamer before other CMG components) • Cdc45 and the GINS complex (4 proteins)
Stabilization of single-stranded DNA
RPA (binds single-stranded DNA)
Replicative polymerases
DNA polymerase α/primase (no editing function) starts synthesis of the leading strand and each Okazaki fragment DNA polymerase δ replicates the lagging strand DNA polymerase ε replicates the leading strand (both δ and ε have 3′–5′ exonuclease editing capability)
Processivity factor
PCNA (ring-shaped clamp that slides along the DNA. Keeps polymerases δ and ε attached to the template strand so that they make longer chains; coordination of cell-cycle control and replication; role in repair)
PCNA loader
RF-C (Binds primer: template junction. AAA ATPase. Loading factor for PCNA, important for polymerase switch)
Closing Factors Removal of RNA primer
Fen1 5′–3′ exonuclease Dna2 helicase and endonuclease RNase H
Ligation of discontinuous DNA fragments
DNA ligase I
Releasing superhelical tension
DNA topoisomerase I
Disentangling daughter strands
DNA topoisomerase II
ATPase, adenosine triphosphatase; CMG, Cdc45, Mcm2–7, and GINS; GINS, Go-Ichi-Ni-San; PCNA, proliferating cell nuclear antigen; RPA, replication protein A.
point for licensing replication origins (Fig. 42.9). Mammalian cells pass the restriction point when the levels of Cdk2–cyclin E, and subsequently, Cdk2–cyclin A rise (see Fig. 40.17). This prevents the relicensing of replication origins until after the next mitosis. In yeasts, the single Cdk that is complexed with B-type cyclins inhibits origin licensing. Experimental inactivation of CdkCdc2 during the G2 phase in the fission yeast Schizosaccharomyces pombe demonstrated the importance of Cdk activation: Cells lacking CdkCdc2 activity loaded Mcm2–7 onto already replicated DNA and then carried out further rounds of “illegal” DNA replication without division. Cdk regulation of origin licensing during G1 phase actually provides one explanation for oncogenic stress (see Chapter 41), in which inappropriate activation of oncogenes leads to cell-cycle arrest followed by death or senescence. Inappropriate oncogene activation can lead to premature stabilization of cyclins D and cyclin E, resulting in premature activation of Cdks. This, in turn, can lead to an insufficient number of replication origins being licensed, thereby inhibiting the cell’s ability to deal with replication stress (see later) and its associated DNA damage.
Different cell types use various combinations of several different pathways to downregulate the activity of the licensing system once cells enter S phase. In metazoans, the most important pathways reduce Cdt1 activity by degradation in S and G2 phases. After the cell passes the restriction point, Cdks phosphorylate Cdt1. This promotes its ubiquitylation by the SCFSkp2 and degradation. In addition, the key replisome component proliferating cell nuclear antigen (PCNA) recruits another ubiquitin ligase that causes Cdt1 degradation during S phase. In some cell types, Cdks phosphorylate Cdc6 and Orc1, leading to their ubiquitylation and degradation. Vertebrates use a protein called geminin as an alternative regulator of origin licensing by Cdt1 (Fig. 42.8). Geminin binds to Cdt1 and prevents it from loading Mcm proteins onto DNA. The anaphase-promoting complex/ cyclosome (APC/C; see Fig. 40.15) triggers degradation of geminin, keeping its concentration very low from anaphase through the late G1 phase, allowing prereplication complexes to assemble. Accumulation of geminin starting in the S phase inhibits the assembly of new prereplication complexes until after the next mitosis. Yeasts control origin licensing without geminin.
CHAPTER 42 n S Phase and DNA Replication
HUMAN CERVICAL EPITHELIUM STAINED WITH ANTI-MCM5
Increased Cdk activity
Cdk2-cyclin E
A. Normal Cdk2-cyclin A
Origin licensing
733
B. Low-grade lesion
C. High-grade lesion
Cdk4/6-cyclin D G1
S
Increased APC activity
Restriction point
Geminin Emi1 SCFSkp2 SCFβ-TrCP
APC/CCdh1
APC/CCdh1 Geminin Cdc6 Cyclins
SCFβ-TrCP Emi1 Cdc25A
SCFSkp2 p27Kip1 Orc1
Other SCF Cyclin D DP1 Cdt1
FIGURE 42.9 PROTEIN DEGRADATION REGULATES DNA REPLICATION. Degradation of geminin, Cdc6, Cdc25A, and cyclins during the G1 phase keeps Cdk activity low and allows prereplication complex formation. Degradation of p27Kip1 and inactivation of the APC/CCdh1 by Emi1 allows the activation of Cdks to levels sufficient for the initiation of the S phase. Once cells enter the S phase, the G1/S regulatory machinery (cyclins D and E and the E2F cofactor DP1) is degraded. Degradation of Cdt1 and accumulation of geminin block reassembly of prereplication complexes. APC, anaphase-promoting complex; SCF, Skp2 (S-phase kinase-associated protein), cullin, and F-box proteins.
A third way to regulate origin licensing is to sequester molecules required to assemble the prereplication complex in the cytoplasm following the onset of S phase. This was first suggested by studies of DNA replication in Xenopus egg extracts (see Fig. 40.8) in which intact nuclei replicate their DNA only once but re-replicate if the nuclear membrane is disrupted. In different cell types, nuclear exclusion of ORC, Cdc6, Cdt1, or Mcm2–7 all contribute to limiting licensing during S and G2 phases. Components of the prereplication complex are absent from nondividing differentiated cells. In fact, detection of these proteins with antibodies in cells from cervical smears can be used as a sensitive method for the early detection of cancer cells (Fig. 42.10).
Signals That Start Replication A classic experiment (Fig. 42.11) demonstrated that (a) a cytoplasmic inducer triggers the transition into the S phase and (b) this inducer triggers DNA replication in a G1 nucleus but not in a G2 nucleus. The inducer is a combination of protein kinases, including Cdk–cyclin pairs, as well as a specialized kinase, Cdc7-Dbf4. In mammals, Cdk2–cyclin E activity peaks at the G1/S transition (Fig. 42.9) and phosphorylates pRb, thereby opening the restriction point “gate.” This allows the E2F/DP dimer to function as a transcription factor and stimulate
FIGURE 42.10 SECTIONS OF HUMAN CERVIX STAINED WITH ANTIBODIES TO MCM5. A, Normal G0 cells in this stratified epithelium lack Mcm5 and other replication proteins. B–C, Cancer cells express Mcm5 at higher levels as they become more malignant. Bound antibodies were detected with peroxidase coupled to a secondary antibody (brown reaction product) and lightly counterstained with hematoxylin (blue). (Modified from Williams GH, Romanowski P, Morris L, et al. Improved cervical smear assessment using antibodies against proteins that regulate DNA replication. Proc Natl Acad Sci U S A. 1998;95:14932–14937.)
the transcription of genes involved in DNA replication (see Fig. 41.9). In addition to cyclin E itself, E2F drives expression of cyclin A, Cdc25A, enzymes required for synthesis of DNA precursors (dihydrofolate reductase, thymidine kinase, and thymidylate synthase), originbinding proteins (Cdc6, Orc1, Cdt1 and its inhibitor geminin), and two components of the replication machinery (DNA polymerase α and PCNA; see Fig. 42.9). In the S phase, the SCFSkp2 ubiquitin ligase complex targets the Cdk inhibitor p27Kip1 for destruction by proteasomes (see Fig. 40.16). SCF gets its name from three of its components: Skp2, cullin, and F-box proteins (see Fig. 40.16). Skp2, which is short for “S-phase kinaseassociated protein,” was identified in a complex with Cdk2–cyclin A. SCF recognizes and ubiquitylates many of its substrates only after they are phosphorylated by Cdk2cyclin A. Cdk2–cyclin A also targets E2F/DP and cyclin E for degradation when cells enter the S phase (Fig. 42.9). For DNA replication to start, the paired strands of the double helix must be separated so DNA polymerase can recognize the bases and begin synthesizing the daughter strand. The multi-subunit CMG DNA helicase uses ATP hydrolysis to peel apart the paired strands of the DNA double helix. Prior to S phase each origin of replication
734
SECTION X n Cell Cycle
A
+
Cell in S phase
Heterokaryon (cell with two nuclei)
100
% G1 cells that enter S phase
B
Cell in G1 phase
Fusion of S and G1 cells G1 cell alone Advancement due to inducer
50
0 0
4
8
12
16
Hours after fusion FIGURE 42.11 CELL FUSION EXPERIMENT REVEALS THE EXISTENCE OF A POSITIVE INDUCER OF THE S PHASE. A, Synchronized cells in different stages of the cycle were fused to yield two nuclei in a single cytoplasm. B, If the fusion involved G1 and S cells, the G1 nucleus was induced to enter the S phase sooner than expected. If the fusion involved S and G2 cells, the G2 nucleus failed to rereplicate its DNA (not shown). (Modified from Rao PN, Johnson RT: Mammalian cell fusion: studies on the regulation of DNA synthesis and mitosis. Nature. 1970;225:159–164.)
is licensed by assembly of a double hexamer of Mcm2–7 complexes. Cdc7 kinase with its associated regulatory subunit Dbf4 phosphorylates the Mcm complex. Next, Cdks phosphorylate other proteins that recruit GINS and Cdc45 to the phosphorylated Mcm2–7 hexamer, thereby producing the CMG helicase (Table 41.1 and Fig. 42.12A). Subsequent activation of the helicase initiates replication.
Mechanism of DNA Synthesis Replication initiates when the activated CMG helicase starts to separate the DNA strands, which move outward in both directions as the replisome assembles at the origin of bidirectional replication. The newly unwound DNA binds the single-strand DNA-binding protein, RPA (replication protein A), ensuring that the separated strands do not base-pair with one another again (Fig. 42.12B). Table 42.1 describes several other proteins that also bind at this time. Box 42.1 provides an introduction to DNA replication in E. coli. The separated DNA strands are ready for replication, but DNA synthesis always involves addition of an incoming nucleoside triphosphate to a free 3′ OH group at the terminus of a preexisting nascent polynucleotide (Fig. 42.1). In the absence of a nascent DNA chain, how does DNA polymerase get started? This problem is solved on the lagging strand by a DNA-dependent RNA polymerase
called a primase, which, like other RNA polymerases, can initiate synthesis de novo without the need for a 3′OH group. In eukaryotes, all DNA chains are started by a complex of DNA polymerase α and a primase subunit, collectively known as Pol α/Primase. Primase synthesizes an RNA chain of approximately seven nucleotides to which DNA polymerase α adds another 20 to 25 nucleotides of so-called initiator DNA (iDNA) (Fig. 42.12D–E). These initiating reactions are potentially risky, because DNA polymerase α lacks proofreading ability. To avoid errors created by mismatching of an incoming base, the RNA primer and most or all the iDNA laid down by Pol α/Primase are subsequently replaced. Once Pol α/Primase has done its job, two further essential factors act to complete replisome assembly. A pentameric protein complex called replication factor C (RFC) binds the 3′ end of the iDNA. RFC uses energy from ATP hydrolysis to load the trimeric protein PCNA onto the DNA (Fig. 42.12E–G). The PCNA trimer is doughnut-shaped, and when the DNA is inserted into its central hole, it is topologically locked onto the DNA. RFC binding and PCNA loading displace Pol α/Primase from the DNA, and PCNA then recruits DNA polymerases δ and ε to the DNA. With PCNA acting as a sliding platform, DNA is reeled through the replisome and polymerase ε synthesizes DNA continuously on the leading strand (Fig. 42.12G). On the lagging strand, polymerase δ synthesizes DNA in bursts of approximately 250 bp, each initiated by Pol α/Primase and roughly correlating with the passage of one nucleosome by the replication fork. The replisome contains a chromatin remodeling activity that removes nucleosomes from the DNA ahead of the fork and replaces them after the fork passes. Both polymerases δ and ε have associated exonuclease activities. This enables them to proofread the newly synthesized DNA and correct most mistakes that they have made. The combination of selecting the correct nucleotides and efficient correction of mistakes may explain the amazing fidelity of DNA replication, with typically only one error per 109 bp polymerized. Locally, the final steps of DNA replication are removal of the RNA primer (and probably iDNA) and ligation of adjacent stretches of newly synthesized DNA. Removal of the primer can be accomplished in two ways (Fig. 42.12H). On one hand, an RNA exonuclease called RNase H can chew in from the 5′ end of the primer. However, this enzyme cannot remove the last ribonucleotide that is joined to iDNA. That requires the nucleases Fen1 (Flap endonuclease) or Dna2. Fen1 requires a helicase to peel the RNA (and possibly the iDNA) away from the template, creating a sort of flap. Fen1 then cleaves at the junction where the flap is anchored to the DNA template, removing the oligomer of unwanted nucleotides in one step. Alternatively, the Dna2 nuclease can do the whole job itself because it also has helicase
CHAPTER 42 n S Phase and DNA Replication
F. RFC binds iDNA, evicting polymerase α / primase, loading PCNA
A. Activation of origin Cdc7 Cyclin A
Dbf4
Cdk2 RFC
ORC Mcm2-7 double hexamer
B. Binding of Cdc45, GINS and RPA 5’
3’
Cdc45 GINS
ATP
PCNA
Polymerase α and primase leave
Cdc6 and Cdt1 leave
RPA
735
G. PCNA recruits polymerase ε to leading strand Polymerase ε
Active CMG helicase
C. Cdc45 recruits polymerase α / primase
H. Processive DNA synthesis starts RFC
PCNA
Nascent DNA Primase Polymerase α
D. Synthesis of RNA primer
I. Primer and i-DNA removal RNase H
RNA primer
Fen1
E. Synthesis of i-DNA
5’
RNA primer
i-DNA
J. DNA synthesis completed by polymerase ε followed by ligation of strands; p97/Cdc48 removes Mcm2-7 and replisome disassembles p97/Cdc48
3’
FIGURE 42.12 MAIN EVENTS OF DNA REPLICATION ON THE LEADING STRAND. For a more detailed description, including events on the lagging strand, see the text. (For reference, see PDB files 1A76 [for Fen1], 1SXJ [for RFC/PCNA], 2ZXX [for Cdt1], and 3CF2 [for p97/Cdc48].)
activity. Most of these maturation/processing enzymes are recruited by interacting with the PCNA trimer. The individual interactions are transient while PCNA remains as a stable loading platform. Following removal of initiator RNA, the Pol δ/PCNA complex extends the upstream nascent chain until it runs into the downstream 5′ end created by Fen1. DNA ligase I then joins the two stretches of DNA together (Fig. 42.12I). When the DNA is ligated, the replisome must be disassembled. This is an active process triggered by ubiquitylation of the Mcm7 subunit of the CMG helicase. The AAA-ATPase p97/Cdc48 (which also extracts proteins from the endoplasmic reticulum in the endoplasmic
reticulum-associated degradation [ERAD] pathway; see Chapter 23) recognizes this ubiquitin and then actively separates the Mcm2–7 hexamer from Cdc45 and GINS, causing the replisome to fall apart.
Higher-Order Organization of DNA Replication in the Nucleus The term S phase gives the impression that all DNA replicates more or less synchronously, but this is far from true. At any given time during the S phase, only 10% to 15% of the replicons actively synthesize DNA. Some replicate early, others late. This pattern of replication is not random; some segments of DNA consistently
736
SECTION X n Cell Cycle
BOX 42.1 DNA Replication in Escherichia coli The DNA replication system of Escherichia coli has been reconstituted entirely from purified components. Analysis of this system reveals many similarities with eukaryotic replication, indicating that this process is highly conserved. E. coli DNA replication can be subdivided into three phases: initiation, elongation, and termination. Thus far, at least 28 polypeptides are known to be involved.
Initiation E. coli chromosomal DNA replication initiates within a 245-bp region, termed oriC. This region contains four 9-bp binding sites for the E. coli initiator protein, DnaA. Nearby are three repeats of a 13-bp A/T-rich sequence. oriC also contains specific binding sites for two small histone-like proteins called HU and IHF. Replication is initiated with the cooperative binding of 10 to 20 DnaA monomers to their specific binding sites (Fig. 42.13). To be active, these monomers must each have bound ATP. Binding of DnaA permits unwinding of the DNA at the 13-bp repeats, in a reaction A
R1 13-mers
R2 R3
R4
oriC DNA
DnaA sites DnaA + ATP HU or IHF
B DnaC ATP DnaB•DnaC + DnaB complex DnaC
C SSB ATP ADP
DnaA
D
FIGURE 42.13 FACTORS INVOLVED IN THE INITIATION OF DNA REPLICATION IN ESCHERICHIA COLI. A, DNA sequences at OriC. B, Unwinding of the origin. C, Binding of helicase. D, The template, now ready for binding of DNA polymerase. ADP, adenosine diphosphate; ATP, adenosine triphosphate; SSB, single-stranded DNA binding protein. (Modified from Baker TA, Wickner SH. Genetics and enzymology of DNA replication in Escherichia coli. Annu Rev Genet. 1992;26:447–477.)
replicate early in the S phase, whereas others consistently replicate late in the S phase. Up to 1000 sites of replication, called replication foci are active at any one time during the S phase in a mammalian cell nucleus (Figs. 42.6 and 42.14B–C and E–F). Given that each of these replication foci is active for only about 30 minutes out of the 8- to 10-hour S phase, a cell will replicate DNA at approximately 10,000
that requires the histone-like proteins. Next, DnaC binds to DnaB and escorts it to the unwound DNA. DnaB is the key helicase that will drive DNA replication by unwinding the double helix, but it binds DNA poorly on its own in the absence of its DnaC escort. Once DnaB has docked onto the DNA, DnaC is released, and the helicase can then start to unwind the DNA, provided that ATP, SSB, and DNA gyrase are present. SSB is a single-stranded DNA binding protein that stabilizes the unwound DNA, and DNA gyrase is a topoisomerase (see Chapter 8) that removes the twist that is generated when the two strands of the double helix are separated.
Elongation As in eukaryotes, E. coli DNA replication involves a leading strand, with the daughter DNA synthesized as a single continuous molecule, as well as a lagging strand, with the DNA synthesized as discontinuous Okazaki fragments. All daughter strands are started by an RNA primase that deposits primers of 11 ± 1 nucleotides. The enzyme that actually synthesizes the DNA is the polymerase III holoenzyme, which has at least 10 subunits. This contains polymerase and proofreading subunits and is held to the DNA by a doughnutlike “sliding clamp” (β). The β is loaded onto the DNA by a pentameric complex in a process that requires ATP. The parallel with PCNA and RFC in eukaryotes is striking. Activities specific for the lagging strand include RNase H, which removes the RNA primers; DNA polymerase I, which fills in the gaps left behind by primer removal; and DNA ligase, which links the Okazaki fragments together. DNA replication in E. coli is significantly faster than it is in eukaryotes, with the fork moving at a rate of approximately 1000 bp per second. This higher speed is presumed to be at least partially attributable to the absence of nucleosomes on the bacterial chromosome, but the eukaryotic enzymes may also be intrinsically slower.
Termination A specialized termination zone is found on the circular E. coli chromosome opposite oriC. This zone contains binding sites called ter sites, to which the ter binding protein binds. This protein appears to block the movement of DNA helicases, such as DnaB, thereby stalling the DNA replication fork. Following termination of replication, a specialized topoisomerase, the product of the parC and parE genes, is required to separate the daughter chromosomes from one another.
of these foci. There are roughly 60,000 origins in a mammalian cell, so each replication focus represents five or six replication origins that are activated coordinately. The replication foci correspond to structural domains of chromosomes that are now called TADs (topologically associating domains) (see Fig. 8.13). The function of each TAD, measured by replication timing, chromatin composition and transcriptional output, can change
CHAPTER 42 n S Phase and DNA Replication
A. BrdUTP O–
O– P O O O– Incorporated P O into cellular DNA O O Br by replication NH O– P O O CH N O 2 O H H H H Fed to cells OH H
B
C
D
Fluorescein-dUTP Bodipy-TR-dUTP at same time
Fluorescein-dUTP Bodipy-TR-dUTP 3 hrs later
5 µm
E
5 µm IdU CIdU 4 hrs later
F
5 µm CIdU added for 2 min IdU added 4 hrs later for 5 min
737
FIGURE 42.14 VISUALIZATION OF DNA REPLICATION WITHIN THE NUCLEUS. A, BrdUTP is introduced into DNA in place of dTTP to label newly replicated DNA. The incorporated BrdU (bromodeoxyuridine) molecules are detected with fluorescent antibodies. B, In a related approach, green-dUTP (deoxyuridine triphosphate) and red-dUTP, when added together, show the many sites of DNA replication in a cell nucleus. Because both UTP (uridine triphosphate) analogs are incorporated simultaneously into the DNA, the sites of replication appear yellow. C, Green-dUTP is followed by red-dUTP added 3 hours later. The later sites of DNA replication show very little overlap with the earlier sites. D, Mitotic chromosome from a cell that was labeled early in the S phase with IdU (green), and then 4 hours later with CldU (red). The late-replicating and early replicating regions of the chromosome are segregated into discrete bands. E, CldU (green) added early in the S phase and IdU (red) added 4 hours later show relatively little overlap. F, CldU (green) added early in the S phase and IdU (red) added 6 hours later show no overlap. The large red blocks of labeling seen with the IdU are characteristic of the pattern of replicating heterochromatin seen late in the S phase. Bodipy-TR-dUTP, a red fluorescent form of dUTP; BrdUTP, bromodeoxyuridine triphosphate; CldU, chlorinedUTP; Fluorescein-dUTP, a green fluorescent form of dUTP; IdU, iodine-dUTP. All are used in place of dTTP (deoxythymidine triphosphate) in DNA synthesis. (B–C, Courtesy P.R. Cook, University of Oxford, United Kingdom. From Manders EMM, Kimura H, Cook PR. Direct imaging of DNA in living cells reveals the dynamics of chromosome formation. J Cell Biol. 1999;144:813–821, copyright The Rockefeller University Press. D, Courtesy A.I. Lamond, University of Dundee, United Kingdom. From Ferreira J, Paolella G, Ramos C, et al. Spatial organization of large-scale chromatin domains in the nucleus: a magnified view of single chromosome territories. J Cell Biol. 1997;139:1597–1610, copyright The Rockefeller University Press. E–F, From Ma H, Samarabandu J, Devdhar RS, et al. Spatial and temporal dynamics of DNA replication sites in mammalian cells. J Cell Biol. 1998;143:1415–1425. Copyright 1998 The Rockefeller University Press.)
5 µm CIdU added for 2 min IdU added 6 hrs later for 5 min
coordinately during differentiation. Thus, TADs with similar function tend to end up close to each other forming larger compartments—for example, euchromatin and heterochromatin. Evidence for the replication of clusters of replicons within the nucleus was first obtained by fiber autoradiography. Cells were fed radioactive precursors for DNA synthesis and then examined by electron microscopy. (For an explanation of this technique, see Fig. 40.3.) Clusters of replicating DNA regions were observed. Several methods are available to observe the spatial distribution of DNA replication during the S phase. One can employ a pulse of nucleotide base analogs that are incorporated into DNA and later detected by fluorescence microscopy. For example, thymidine analogs such as BrdU (bromodeoxyuridine triphosphate), IdU (iododeoxyuridine), and CldU (chlorodeoxyuridine) can be detected in fixed cells by reaction with fluorescent antibodies (BrdU is called by its correct name BrdUTP in Fig. 42.14). Alternatively analogs labeled with a
fluorescent dye allow the newly replicated DNA to be observed directly in living cells. The time at which each replicon fires during the S phase can be seen clearly by synchronizing cells at the beginning of the S phase, releasing them from cell-cycle arrest, and then exposing them to BrdU at various times thereafter. This experiment reveals very distinctive patterns of DNA synthesis occurring at different times during the S phase (Fig. 42.14B–C and E–F). Early on, euchromatin replicates throughout the nucleus. Later, replicating regions are concentrated around nucleoli and other areas of constitutive and facultative heterochromatin (see Chapter 8). Toward the end of the S phase, replication is largely concentrated in blocks of heterochromatin. These observations show that DNA replication occurs throughout the nucleus, wherever DNA is located. DNA does not move to a small number of discrete sites to be replicated (as was once thought). The timing of replication of particular replication origins was studied in detail in budding yeast using a modification of the BrdU labeling method. First, a
738
SECTION X n Cell Cycle
procedure was developed whereby all cells in a population entered S phase synchronously. Next, the shift in the density of the DNA following BrdU incorporation was used to distinguish DNA that had replicated from DNA that had not (Fig. 42.15). Incorporation of BrdU into DNA makes the newly synthesized daughter DNA strand heavier, allowing its separation from the parental DNA by centrifugation on a cesium chloride density gradient (see Chapter 6). It was then relatively simple to take DNA probes from different regions of the chromosome and determine when each replicated (changed its A
1
2
3 * Origin
4
5
1
2
3
4
5
density) during the S phase. This experiment demonstrated that each ARS element replicates at a characteristic time during the S phase. More recently, genome-wide maps of replication timing have been constructed by isolating newly replicated DNA at various times during S phase and identifying the chromosome regions involved by high-throughput DNA sequencing. The most striking aspect of these patterns of DNA synthesis is their reproducibility from one cell cycle to the next. For example, regions of DNA labeled early in the S phase overlap little or not at all with DNA labeled = Restriction enzyme cutting site
Initiation
1
H:L
3
2
4
5
Replication
H:H H:L
G force
H:H
3 1
2
4
Heavy: Heavy:Heavy + Heavy:Light Heavy
Heavy: Light
Isolated DNA digested with restriction enzyme
Heavy:Light + Heavy: Heavy:Heavy Heavy
B
C
D
1,5 50
3 2,4
0 HL
Top
Density
3
1,5
50
0 HH
Bottom
Centromeres replicate early
100
% Replicated
100
% In H:L peak
100
Amount of DNA
5
a b
50
Telomeres replicate late
0 0
2
4
6
8
Time of replication (min)
0
14
28
42
56
Time in S phase (min)
FIGURE 42.15 MEASUREMENT OF THE TIME OF REPLICATION OF PARTICULAR CHROMOSOMAL REGIONS IN SACCHAROMYCES CEREVISIAE. A–C, This protocol is based on a classic density shift experiment of Messelson and Stahl that proved that DNA replication is semiconservative. S. cerevisiae cells are grown for several generations in a medium containing 13C and 15N so that their DNA is fully substituted with heavy isotopes. At the beginning of the experiment, the cells are synchronized so that they enter the S phase in a single wave. At the same time, the heavy (H) isotope medium is removed and replaced with “light medium” (L) containing 12C and 14N. At various times after the initiation of the S phase, aliquots of cells are removed, and the DNA is isolated. The DNA is then cleaved with restriction enzymes so that the chromosomes are cut into many fragments. DNA from each time point is then subjected to CsCl density gradient centrifugation. When any local region of DNA is replicated, its density alters from heavy/heavy to heavy/light. After very short incubations with light isotopes, only DNA near the origin of replication will be heavy/light; all other DNA will be heavy/heavy. These two populations of molecules are separated from one another by density gradient centrifugation. To examine the timing of replication of a specific gene, a cloned segment of DNA corresponding to the region of interest is used to probe (by DNA hybridization) the heavy/heavy and heavy/light peaks from each gradient. The time of replication of each locus is the time at which the restriction fragment being detected by DNA hybridization moves from the heavy/heavy peak to the heavy/light peak. The numbers in panels B and C refer to the numbered regions of the chromosomes shown in A. D, Data from a replication timing experiment show that in budding yeast, centromeres replicate early in the S phase and telomeres replicate late. To generate curve a, fractions from a gradient like that shown in panel B were hybridized to a cloned centromere region. To generate curve b, fractions from the same gradient were hybridized to a cloned telomere region probe. Note that in mammalian cells, centromeres replicate late and telomeres replicate earlier. (Based on the work of the laboratory of B.J. Brewer and W.L. Fangman. For reference, see Meselson M, Stahl FW. The replication of DNA in Escherichia coli. Proc Natl Acad Sci U S A. 1958;44:671–682.)
CHAPTER 42 n S Phase and DNA Replication
3 hours later (Fig. 42.14C and E). However, DNA labeled at corresponding points of the S phase in two successive cell cycles superimposes almost entirely. This strongly suggests that particular TADs initiate DNA synthesis during preferred “windows” during S phase. At least three mechanisms may explain the sequence of replication patterns for different chromosomal regions: 1. Local chromatin structures, established as a result of gene expression influence the time of replication, particularly in metazoans. Thus, transcriptionally active loci (where transcription factors are already bound) have a head start over other regions of the chromosomes, permitting them to initiate DNA replication first. In general, euchromatin (which has low DNA methylation and high histone acetylation) replicates first, followed by facultative heterochromatin (including the inactive X) and, finally, the constitutive pericentric heterochromatin (which has heavy DNA methylation and low histone acetylation). This mechanism can explain why a particular locus may replicate at different times in two cell types. For example, one origin in mammalian genomes is located just upstream of the DNA encoding the β-globin gene (encoding a protein subunit of hemoglobin). This region of more than 200-kb of DNA replicates early in erythroid cells that express the β-globin gene, but later in other cells where the gene is inactive. 2. The higher-order packing of chromatin in the nucleus may influence when origins replicate. It is likely that replication origins fire in a sequence that is imposed on them at the same time that the TAD organization of interphase chromosomes is reestablished during mitotic exit and early G1 phase. 3. Limiting amounts of certain essential proteins may contribute to the sequential replication of different chromatin domains. These proteins accumulate on replisomes assembled on early-firing replicons and only become available to later-firing replicons
dsDNA breaks
Unreplicated DNA
739
following completion of DNA synthesis at the early replicons and disassembly of those replisomes.
Replication Stress and the Intra-S Checkpoint The textbook-smooth DNA double-helix is actually littered with problems and obstacles, including DNA damage, ribonucleotides inserted by mistake into the DNA, and awkward secondary structures caused, for example, by base-pairing within single DNA strands as opposed to between strands. In addition, if cell-cycle regulation is perturbed, cells can simply run out of either the deoxynucleotide triphosphate (dNTP) precursors that they need to synthesize DNA or a number of key proteins that are also present in very low amounts. The replication fork may stall as result of these problems and often needs help to complete replication. An intra-S checkpoint responds to stalled replication forks, probably as a result of detecting excessive singlestranded DNA (Fig. 42.16). Replication forks stall if they encounter a damaged DNA base, bases that the replisome cannot “read” or a DNA secondary structure that it cannot unfold. DNA damage can result from ultraviolet light, mutagens or chemicals such as aldehydes produced as a by-product of ethanol metabolism. Replication blockage leads to a condition known as replication stress in which single-stranded DNA accumulates. Replication stress with stalled forks can also result from inappropriate activation of genes that promote cell cycle progression (oncogenic stress; see Chapter 41). Possible reasons include insufficient licensing of origins during the G1 phase, depletion of nucleoside triphosphate pools, or insufficient pools of essential replication factors. Surprisingly, cells make a number of essential components—including the dNTPs—in amounts just sufficient for replication, so there is little margin for error.
RPA
DNA damage *
RPA
(ATM)2
2x ATM active
ATR actived by excess ssDNA stabilizes stalled replication forks
Stalled fork
Fork approaches from activated dormant origin
Chk1/Chk2 Cdc25A Degraded
Block initiation of new replication foci Block cell cycle progression
FIGURE 42.16 REPLICATION STRESS AND THE INTRA-S CHECKPOINT. If DNA breaks are detected, the ATM (ataxia-telangiectasia mutated) kinase activates downstream kinases Chk1 and Chk2, leading to phosphorylation of Cdc25A and its subsequent ubiquitin tagging and degradation. This blocks the initiation of new replication forks as well as cell-cycle progression more generally. If DNA persists in unreplicated form, or if replication forks stall, the ATR (ataxia-telangiectasia and Rad3–related) kinase activates a similar downstream response. ATR stabilizes stalled replication forks to give nearby dormant origins time to fire and replicate the DNA downstream of the block.
740
SECTION X n Cell Cycle
Transcription increases
Processing of mRNA increases Primary transcript
Histone gene
Mature mRNA exported to cytoplasm
Stability of the mRNA (on free ribosomes) increases Degrading enzymes
3' ends of pre-mRNA remain behind NUCLEUS
FIGURE 42.17 THREE MECHANISMS THAT ELEVATE HISTONE EXPRESSION DURING THE S PHASE.
In conditions of replication stress, the ATM (ataxiatelangiectasia mutated) and ATR (ataxia-telangiectasia and Rad3–related) kinases activate the DNA damage response (see Box 43.1). The activities of these kinases and their downstream effectors result in the degradation of Cdc25A, the phosphatase that triggers entry of the cell into mitosis. The resulting inactivation of Cdks during the S phase prevents replication initiation within other unreplicated domains. A key aspect of this response unique to S phase is that ATR-activated by the presence of excessive levels of single-stranded DNA associated with RPA protects the stalled forks from disassembly, known as replication fork collapse. This response is critical, because replication forks with unwound and nicked DNA molecules can turn into breaks if the fork disassembles before replication is complete and all the DNA is ligated. Error-prone mechanisms tend to repair DNA damage that arises during replication stress. The resulting mutations are thought to contribute to oncogenic transformation. Of course, stalled replication forks must not only be protected, but they must also somehow be rescued. At moderate levels of replication stress, ATR activation blocks the activation of new replication foci, but allows dormant origins to fire near stalled forks in replication domains that are already active. This results in the arrival of a converging fork on the downstream side of the damaged DNA, leaving only a very small gap that can be repaired by a specialized DNA polymerase in a process known as translesion synthesis. Thus, in the real world where replication forks routinely encounter problems that cause them to stall, a pool of dormant licensed replication origins is essential for integrity of the genome.
Synthesis of the Histone Proteins Chromatin contains approximately equal masses of DNA and core histones. Human cells require about 62 × 106 copies of each core histone, assuming a genome size of 6.2 × 109 bp and 200 bp per nucleosome. Because approximately 90% of histone transcription occurs during the S phase, enormous amounts of these proteins
are made during a relatively brief period. Histone synthesis apparently keeps pace, in part, because there are approximately 40 sets of histone genes. Synthesis of histones during the S phase is tightly coupled to ongoing DNA replication and a controlled supply of histones is essential for normal replication. If replication is blocked either by addition of drugs or by temperature-sensitive mutants, histone synthesis declines abruptly shortly thereafter, possibly because nucleosome deposition appears to be linked to lagging strand synthesis. Indeed, the chaperone that assembles histones into nucleosomes is associated with the replisome. This link between histone synthesis and DNA replication appears to involve at least three processes (Fig. 42.17). First, transcription of the histone genes rises threefold to fivefold as cells enter the S phase. Each histone gene has a cell-cycle-responsive element in its promoter to which a transcription factor binds specifically during the S phase. Second, the processing of histone messenger RNAs (mRNAs) increases six- to 10-fold as cells enter the S phase. Histone mRNAs are not polyadenylated, and the primary transcripts are considerably longer than the mature forms. Processing of the 3′ end of histone premRNAs involves the U7 small nuclear ribonucleoprotein (snRNP) (see Chapter 11), a portion of which recognizes histone mRNA and base-pairs with it during processing. Cell-cycle-dependent regulation of processing appears to involve changes in the accessibility of the necessary portion of U7 small nuclear RNA (snRNA). This region is inaccessible in G0 cells but becomes accessible when cells that reenter the cycle and begin the S phase. The mechanism for this change in RNA conformation is not known. Third, changes in the stability of the mRNA also regulate histone synthesis. Normally, the level of histone mRNA on free polysomes drops rapidly by approximately 35-fold as cells enter the G2 phase. If DNA synthesis is interrupted during the S phase, a region at the 3′ end of the mature message targets the mRNA for degradation. If this region is removed from the 3′ terminus of the histone mRNA, the normal link between ongoing
replication and mRNA stability is lost. Furthermore, this sequence, transposed onto the 3′ terminus of a globin mRNA, renders that mRNA sensitive to degradation if DNA synthesis is blocked. Degradation of histone mRNA requires ongoing protein synthesis, and it has been speculated that histones themselves participate in the control. As discussed in Chapter 8, specialized variant forms of histones are synthesized and inserted into the chromatin outside of S phase. These histones are encoded by mRNAs with introns and normal poly(A) tails and are therefore not processed by the specific S phase– associated pathway (see Chapter 11). Their insertion into chromatin is typically correlated with RNA transcription rather than DNA replication and involves specific chromatin-remodeling factors.
Other Events of the S Phase Although the bulk of attention on the S phase focuses on the duplication of the chromosomes, centrosomes also duplicate at this time. Duplication of the centrosomes is essential for stability of the genome, because they set up the poles of the mitotic spindle that is responsible for accurate partitioning of the replicated chromosomes. Interestingly, Orc1 functions independent of DNA replication with cyclin A to limit centriole duplication to once per cell cycle. (See Fig. 34.17 for a discussion of centrosome duplication.) With the completion of DNA replication and duplication of the centrosomes, the cell is ready to divide. As the levels of Cdk activity rise toward the threshold that is sufficient to trigger mitotic entry and other factors necessary for mitosis accumulate, the cell continues to screen the integrity of the DNA to ensure that the
CHAPTER 42 n S Phase and DNA Replication
741
genome has been replicated completely and that no harmful DNA damage is present. These checks, together with other ongoing preparations for mitosis, are the principal events of the G2 phase (see Chapter 43). ACKNOWLEDGMENTS We thank Hiro Araki, Julian Blow, Cristina Cardoso, David Gilbert, and Julian Sale for suggestions on revisions to this chapter. SELECTED READINGS Boos D, Frigola J, Diffley JF. Activation of the replicative DNA helicase: breaking up is hard to do. Curr Opin Cell Biol. 2012;24:423-430. Costa A, Hood IV, Berger JM. Mechanisms for initiating cellular DNA replication. Annu Rev Biochem. 2013;82:25-54. Deegan TD, Diffley JF. MCM: one ring to rule them all. Curr Opin Struct Biol. 2016;37:145-151. Hills SA, Diffley JFX. DNA replication and oncogene-induced replicative stress. Curr Biol. 2014;24:R435-R444. Labib K. How do Cdc7 and cyclin-dependent kinases trigger the initiation of chromosome replication in eukaryotic cells? Genes Dev. 2010;24:1208-1219. McIntosh D, Blow JJ. Dormant origins, the licensing checkpoint, and the response to replicative stresses. Cold Spring Harb Perspect Biol. 2012;4:a012955. Méchali M, Yoshida K, Coulombe P, Pasero P. Genetic and epigenetic determinants of DNA replication origins, position and activation. Curr Opin Genet Dev. 2013;23:124-131. O’Donnell M, Langston L, Stillman B. Principles and concepts of DNA replication in bacteria, archaea, and eukarya. Cold Spring Harb Perspect Biol. 2013;5:a010108. O’Donnell M, Li H. The eukaryotic replisome goes under the microscope. Curr Biol. 2016;26:R247-R256. Rhind N, Gilbert DM. DNA replication timing. Cold Spring Harb Perspect Biol. 2013;5:a010132. Zeman MK, Cimprich KA. Causes and consequences of replication stress. Nat Cell Biol. 2014;16:2-9.
This page intentionally left blank
CHAPTER
43
G2 Phase, Responses to DNA Damage, and Control of Entry Into Mitosis T
multifaceted regulation of Cdk1 (see Fig. 40.14) includes binding of cyclin cofactors, inhibition and activation by phosphorylation, binding of inhibitory molecules, changes in subcellular localization and changes in the activities of competing protein phosphatases (Fig. 43.1). These various factors are finely balanced, until a positive feedback loop enables the cells to make an abrupt and decisive entry into mitosis. Mammals have at least three B-type cyclins: B1, B2, and B3. Cyclin B1 is essential for triggering the G2/M transition. Cyclin B1, newly synthesized during the latter part of the cell cycle, binds Cdk1 and shuttles it in and out of the nucleus. Importin β carries the Cdk1–cyclin B1 complex into the nucleus, and then Crm1 rapidly exports it back to the cytoplasm (see Chapter 9). Cdk1– cyclin B2 associates with the Golgi apparatus during interphase and might function in Golgi disassembly during mitosis (see Fig. 44.4). Cyclin B3 may function only during meiosis in mammals.
he G2 phase is the gap between the completion of DNA replication and the onset of mitosis. This chapter begins with the biochemical basis for the G2/mitosis (M) transition and discusses how the G2/M checkpoint delays this transition if DNA damage is detected. Finally, the chapter introduces the major pathways that cells use to repair damaged DNA.
Enzymology of the G2/Mitosis Transition The transition from the G2 phase into mitosis is the most profound morphologic and physiological change that occurs during the life of a proliferating cell. Entry into mitosis is controlled by a network of stimulatory and inhibitory protein kinases and phosphatases, presided over by Cdk1–cyclin B1. (Chapter 40 introduced the components involved in the G2/M transition.) Cdk1, the driving force for entry into mitosis, is present at a constant level throughout the cell cycle. The Wee1/Myt1 phosphorylates Cdk1 on T14 & Y15 (inhibitory)
Y15 Cyclin B1 Cyclin B1 Cyclin B1 14 Wee1 Myt1 T T161 T161 Cdk1 Cdk1 Cdk1 Cdk1 Cdk1 Inactive Inactive Inactive Active Inactive kinase kinase kinase Cdc25A,B,C kinase kinase CAK Cyclin B1 CAK phosphorylates Cdc25 Cyclin B1 degraded Cdk1 on T161 (stimulatory) dephosphorylates T14 and Y15 S
G2
M
G1
FIGURE 43.1 REGULATION OF CDK1 BY CYCLIN BINDING AND PROTEIN PHOSPHORYLATION FROM THE LATE S PHASE THROUGH MID-MITOSIS. (For reference, see Protein Data Bank [PDB; www.rcsb.org] file 1X8B and Squire CJ, Dickson JM, Ivanovic I, et al. Structure of human Wee1A kinase: kinase domain complexed with inhibitor PD0407824. Structure. 2005;13:541–550.)
743
744
SECTION X n Cell Cycle
As cells approach the G2/M transition, phosphorylation simultaneously activates and inhibits Cdk1 bound to cyclin A or B. Cdk-activating kinase (CAK-actually Cdk7– cyclin H), located in the nucleus, phosphorylates Cdk1 on T161 (Fig. 43.1). This triggers a refolding of the active site cleft that allows the enzyme to bind substrates (see Fig. 40.13). This phosphorylation activates Cdk1-cyclin kinases in the nucleus (Fig. 43.1). At the same time, Wee1 and Myt1 kinases phosphorylate T14 and Y15 to turn the enzyme off. Wee1 in the nucleus phosphorylates Cdk1 on Y15 adjacent to the adenosine triphosphate (ATP)-binding site. This stops the kinase from using ATP to phosphorylate substrates. While Cdk-cyclin complexes are in the cytoplasm, they are phosphorylated on T14 and Y15 by Myt1 kinase associated with the Golgi apparatus and endoplasmic reticulum. The role of Wee1 as a mitotic inhibitor is clearly demonstrated in Schizosaccharomyces pombe: overexpression of Wee1 delays or prevents the entry of cells
into mitosis (Fig. 43.2). Together, Wee1 and Myt1 ensure that Cdk1 remains inactive as it shuttles into and out of the nucleus (Fig. 43.3). This combination of stimulatory and inhibitory phosphorylations holds Cdk1–cyclin complexes poised for a burst of activation. That burst occurs when one of three Cdc25 protein phosphatases removes the inhibitory phosphates from T14 and Y15 (Figs. 43.1 and 43.8). Cdc25s are “dualspecificity” protein phosphatases (see Fig. 25.5) that remove phosphates from serine (S), threonine (T), and tyrosine (Y) residues. Of the three Cdc25 phosphatases, only Cdc25A is indispensable for life. It functions at both the G1/S and G2/M transitions, whereas Cdc25B and Cdc25C have roles only in the G2/M transition. Because Cdc25 phosphatases trigger mitotic entry, their regulation is both important and elaborate. Cdc25A and Cdc25C are held inactive during interphase by mechanisms involving stimulatory and inhibitory
B. Wee1+
A. Wee1
C. Wee1+ (3x)
D. Wee1+ (5x)
FIGURE 43.2 EFFECT OF CHANGING CELLULAR LEVELS OF WEE1 PROTEIN ON CELL-CYCLE PROGRESSION IN FISSION YEAST. A, Cells that lack functional Wee1 protein enter mitosis too soon in the cell cycle and are smaller than wild-type cells (B). C–D, Cells that express excess Wee1 protein are too effective at inactivating Cdk1 and are severely delayed in their ability to enter mitosis (hence their larger size). (From Russell P, Nurse P. Negative regulation of mitosis by Wee1+, a gene encoding a protein kinase homolog. Cell. 1987;49:559–567.)
AT Cyclin B1 Myt1 docking (Active) site
14-3-3 Cdc25C (Inactive)
ADP
Y15 Cyclin B1 T161 Cdk1 NES (Inactive)
T14
Cyclin B1 Myt1 docking site (Inactive) blocked
CYTOPLASM NUCLEUS
14-3-3
14-3-3 14-3-3
Cdc25C (Inactive)
Cdc25A (Inactive/unstable) S phase
Wee1 (Active)
AT
ADP
Y15 Cyclin B1 T14 T161 Cdk1 NES (Inactive)
Wee1 (Inactive)
Cyclin B1 T161
Cdk1 (Active)
NES
Cyclin B1 nuclear export signal inactivated G2 phase
Mitosis
FIGURE 43.3 CDK REGULATION IN INTERPHASE AND MITOSIS. Inhibitory phosphorylation and shuttling of components between the nucleus and cytoplasm regulate Cdk1 activity in interphase and mitosis. Cdc25, which would remove the inhibitory phosphates, is held inactive until its activation by Polo kinase and Cdk1-cyclin B stimulates mitotic entry (not shown here). ADP, adenosine diphosphate; ATP, adenosine triphosphate; NES, nuclear export sequence. (Based on an original figure by Helen Piwnica-Worms. For reference, see PDB file 1YWT.)
CHAPTER 43 n G2 Phase, Responses to DNA Damage, and Control of Entry Into Mitosis
Interphase
Mitosis
Low Cdc25C activity, cytoplasmic 14-3-3 Cdc25C
745
Phosphorylated S216 binds 14-3-3 protein
14-3-3
Enhanced Cdc25C activity, nuclear, S216 phosphorylation blocked Y15 Cyclin B1 Cdc25C T14 T161 Cdk1 (Inactive)
Cyclin B1 T161
Cdk1 (Active)
Cdc25A Cdc25A
Cyclin B1/Cdk1 docking site
Low Cdc25A protein levels due to proteolysis Reduced interactions with Cdk1-cyclin B1 due to 14-3-3 binding at the Cdc25 C-terminus
Cdc25A accumulates due to increased stability Enhanced interaction with Cdk1-cyclin B1 complexes due to loss of 14-3-3 binding Enhanced activity due to mitotic-specific phosphorylation
FIGURE 43.4 REGULATION OF CDC25A AND CDC25C ACTIVITY IN INTERPHASE AND MITOSIS. (Based on an original figure by Helen Piwnica-Worms.)
phosphorylation, binding to a 14-3-3 adapter protein, alterations in their subcellular localization, and ubiquitinmediated proteolysis (Fig. 43.4). Phosphorylation on a serine residue of Cdc25 creates a binding site for a 14-3-3 adapter protein that interacts with the phosphoserine and several flanking amino acids (see Fig. 25.10). Association with 14-3-3 interferes with Cdc25A binding to Cdk1–cyclin B1. Phosphorylation at other sites also targets Cdc25A for ubiquitination by SCFβTrCP and destruction by proteasomes. Cdc25C has a nuclear export sequence (see Fig. 9.17), so Cdc25C bound to 14-3-3 is primarily cytoplasmic (Fig. 43.3). Following the theme of phosphorylation having both positive and negative effects, full activation of Cdc25s requires phosphorylation of other residues in the aminoterminal region. A protein kinase called Polo (see next paragraph) initiates this phosphorylation of Cdc25, followed by more extensive phosphorylation by Cdk1–cyclin B1, the substrate of Cdc25. This action of Cdk1–cyclin B1 on Cdc25 creates a powerful positive feedback amplification loop that provides a burst of Cdk activity and triggers entry into mitosis (Fig. 43.7). A molecular trigger is required to start the amplification cycle in which Cdk1–cyclin B1 and Cdc25 activate each other. Members of the Polo family of protein kinases are candidates for this role, by activating Cdc25. These kinases possess amino acid motifs called “polo boxes” that bind to protein partners after they have been “primed” by phosphorylation by another kinase. Interestingly, the most common “priming kinases” for polo are Cdk–cyclin pairs. This creates another positive amplification loop that allows for additional levels of control and rapid activation of Cdk– cyclin complexes. In addition to activating Cdks by phosphorylating Cdc25, polo family kinases are involved in a variety of mitotic events, including formation of a bipolar spindle, cytokinesis, and passage through certain cellcycle checkpoints.
For cells to enter and complete mitosis, protein phosphatase 2A (PP2A, a protein serine/threonine phosphatase; see Fig. 25.6) must be inactivated. PP2A removes the phosphates that Cdk1 puts on its target proteins. If PP2A is not inhibited, cells enter mitosis, but they then slip back into interphase. PP2A inhibition involves three steps. Newly active Cdk1 phosphorylates a protein kinase called Greatwall (a name from Drosophila genetics). Greatwall phosphorylates a small target protein that then binds to the substrate-binding pocket of PP2A, acting as a competitive inhibitor. Cdk1 activation at the onset of anaphase eventually leads to dephosphorylation of the small protein and activation of PP2A, thereby allowing cells to exit mitosis.
Changes in Subcellular Localization at the G2/M Transition Late in G2, phosphorylation inactivates the nuclear export signal of cyclin B1 (see Chapter 9). As a result, Cdk1–cyclin B1 accumulates in the nucleus within 5 minutes (Figs. 43.3 and 43.5). Cdc25C also stops shuttling at the G2/M transition, probably as a result of Polo kinase phosphorylation. The Cdk1–cyclin B that accumulates in the nucleus is thought to be already active, so the reason for this rapid nuclear localization is not entirely clear. The concentration of Cdk–cyclin B complexes together with Cdc25A and Cdc25C in the confined volume of the nucleus may contribute to the final burst of Cdk1–cyclin B1 activation.
Cdk1 Activity and the Initiation of Prophase Cdk2–cyclin A plays a critical role during the S phase (see Chapter 42), but also helps trigger the G2/M transition. Cdk–cyclin A activity peaks at G2/M, before the peak of Cdk1–cyclin B1 activity, and inactivation of
746
SECTION X n Cell Cycle
cyclin A in Drosophila or mammalian cultured cells arrests the cell cycle in G2. Furthermore, G2 cells enter mitosis prematurely if injected with active Cdk2–cyclin A complexes just completing S phase. Finally, microinjection of a selective inhibitor of Cdk–cyclin A causes prophase cells to return rapidly to interphase; chromosomes decondense, rounded prophase cells flatten, and the interphase microtubule network returns. Cdk activity regulates several events during the G2-toprophase transition. In the cytoplasm, the half-life of microtubules drops dramatically from approximately 10 minutes to approximately 30 seconds late in G2 (see Table 44.1). This, coupled with an enhanced ability of centrosomes to initiate microtubule polymerization, completely transforms the organization of the microtubule cytoskeleton. Centrosomes take on the appearance of spindle poles and migrate apart over the surface of the nucleus. While this is happening, chromatin begins to condense in the nucleus. A protein complex called condensin is required for this condensation to begin during prophase (see Fig. 8.18). Cdks activate condensin by phosphorylation of two of its subunits in late G2. G2 phase
Prophase
These events occur while most Cdk1–cyclin B1 is in the cytoplasm. It appears most likely, therefore, that Cdk1–cyclin A triggers at least the nuclear events of prophase (Fig. 43.6). Commitment to mitosis appears to be irreversible only after Cdk1–cyclin B1 enters the nucleus.
Recap of the Main Events of the G2/M Transition Synthesis of cyclin B1 in the latter portion of the S and G2 phases leads to assembly of Cdk1–cyclin B heterodimers that shuttle into and out of the nucleus, spending most of their time in the cytoplasm associated with microtubules. In late G2 phase, Cdk1–cyclin A activity initiates mitotic prophase, beginning with changes in microtubule dynamics and chromosome condensation. Several events trigger entry into the active phase of mitosis. Cdc25A accumulates and no longer binds 14-3-3 proteins, allowing it to interact more effectively with Cdk1–cyclin B. The phosphoserine-binding site for 14-3-3 proteins on Cdc25C is dephosphorylated, allowing Cdc25C to accumulate in the nucleus. In addition, phosphorylation of cyclin B1 blocks its export from the nucleus and promotes its import, thus causing Cdk1– cyclin B1 to accumulate rapidly in the nucleus. The inhibitory kinase Wee1 is also phosphorylated, causing its activity to drop. Cdc25A and Cdc25C activate Cdk1– cyclin B1 by removing inhibitory phosphates on T14 and Y15. This starts in the cytoplasm and then may be stimulated as the proteins concentrate in the nucleus. There, the action of Cdk1–cyclin B1 on the nuclear lamina triggers nuclear envelope breakdown and drives the cell into mitosis. Fig. 43.7 highlights the positive feedback loop between Cdc25 and Cdk1–cyclin B that drives the cell into mitosis.
Metaphase
FIGURE 43.5 CYCLIN B1 LOCALIZATION DURING MITOSIS. Cyclin B1 moves rapidly from the cytoplasm into the nucleus at the onset of prophase and subsequently associates with the spindle during mitosis. (Courtesy Christina Karlsson and Jonathon Pines, Wellcome/CRC Institute, Cambridge, UK.)
Cyclin B1
Cyclin B1
Cyclin B1
Cdk1 CYTOPLASM
Cdk1
G2/M transition events: Microtubule stability drops Centrosomes migrate apart Chromosomes condense Kinetochore assembly starts
Shuttling
NUCLEUS
Cyclin B1
Cyclin B1 Cdk1
Cdk1 Cyclin A
Cdk1
Cdk1
Cyclin A Cdk2 G1
Cyclin A
Cdk2
Cyclin B1 degraded
Cyclin A degraded
Cdk1 S
G2
Prophase
M
FIGURE 43.6 CDK1 REGULATION. Locations and patterns of activation of Cdk1 complexed to cyclin A versus cyclin B across the cell cycle.
CHAPTER 43 n G2 Phase, Responses to DNA Damage, and Control of Entry Into Mitosis
747
Cell-cycle arrest in Rad9+ and rad9 cells with DNA damage
Cdc25 phosphatase dephosphorylation
cdc13tsRad9+
cdc13tsRad9-
Wee1 and Myt1 phosphorylation Inactive kinase Activated by Polo kinase and amplified by positive feedback from Cdk1–cyclin B S
T161 Active Cdk1 CAK protein kinase Myt1, Cdc25 Wee1 protein protein phosphatase kinases
Cells on plates
Bud G2
M
FIGURE 43.7 SUMMARY OF THE CDK1 FEEDBACK REGULATION MECHANISM AT THE G2/M TRANSITION.
Why did such an elaborate system evolve to regulate the G2/M transition? The answer appears to lie in the exquisite sensitivity provided by the interlocking network of stimulatory and inhibitory activities. On the one hand, this network ensures a rapid, almost explosive, final transition into mitosis. On the other, it provides a number of ways to delay the G2/M transition if the cell detects damage to chromosomes. Attempting mitosis with chromosomal damage can lead to cell death or contribute to cancer.
G2/M Checkpoint Separation of sister chromatids during mitosis is a potential danger point for a cell. After DNA is replicated each chromosome consists of paired sister chromatids held together by cohesin. Therefore, if the DNA is damaged, the cell can use information present in the undamaged chromatid to guide the repair process. However, once sisters separate, this corrective mechanism can no longer operate. In addition, if a cell enters mitosis before completing replication of its chromosomes, attempts to separate sister chromatids damage the chromosomes. To minimize these hazards, a checkpoint operates in the G2 phase to block mitotic entry if DNA is damaged or DNA replication is incomplete. Just as DNA damage can arrest the cell cycle in G1 phase, damaged or unreplicated DNA also halts the cell cycle temporarily in the G2 phase. Interestingly, the G1 checkpoint—which can be activated by a single DNA break in human cells—is more sensitive than the G2/M checkpoint, which requires 10 to 20 breaks to block cell-cycle progression. The G2/M checkpoint may be less sensitive then the G1 checkpoint, because G2 cells are already primed to enter mitosis. Consequently, human cells can enter mitosis with limited amounts of damaged or unreplicated DNA. These problem regions can be
DNA damage
pe ild-ty
e
otyp
phen
W
Nucleus
G2 arrest No delay Phe n Cells otype o f k with eep divi rad9 mu tan fragm ding ente and d t d nu i clei e FIGURE 43.8 GENETIC IDENTIFICATION OF THE G2/M CHECKPOINT. Cells defective in the G2/M checkpoint (Rad9 mutants of budding yeast) cannot delay their entry into mitosis in the presence of damaged DNA and therefore divide themselves to death. Budding yeast Rad9 is an ortholog of the ATM (ataxia-telangiectasia mutated) adapter protein 53BP1 (see text). Confusingly, budding yeast Rad9 is not related to fission yeast Rad9, which gives the 9-1-1 complex its name (see text). (Courtesy Ted Weinert, University of Arizona, Tucson.)
detected and repaired in the daughter cells after division (see later). Studies of radiation-induced G2 delay in budding yeast identified a major cell-cycle checkpoint that is sensitive to the status of the cellular DNA. Cells defective in this checkpoint are more sensitive than wild-type cells to radiation injury because they continue to divide, despite the presence of broken or otherwise damaged chromosomes (Fig. 43.8). The cells die, presumably from chromosomal defects or loss. In metazoans, the G2/M checkpoint delays entry into mitosis until the damage is either fixed, triggers cell suicide by apoptosis, or causes cells to enter a nonproliferating (senescent) state. The checkpoint works by modulating the activities of the components that control the G2/M transition.
DNA Damage Response Considering that it is the blueprint for life, DNA is remarkably accident-prone, and organisms have an elaborate network of mechanisms to repair those accidents. This complex network is known as the DDR
748
SECTION X n Cell Cycle
(DNA damage response). This section discusses a few key components of the DDR, and Box 43.1 (see later) explains several mechanisms that repair damaged DNA. The minimal machinery of a DNA damage checkpoint (see Fig. 40.4) involves SENSORS that detect DNA damage, TRANSDUCERS (usually protein kinases) that produce a
γH2AX BRCA1
G2 nucleus
G1 nucleus
FIGURE 43.9 INDUCTION OF DNA DAMAGE FOCI USING A HIGH-ENERGY LASER. Laser-induced DNA damage results in activation of the DDR with production of γH2AX (red) at damage sites followed by binding of other factors, including BRCA1 (green). DNA is blue. The left cell is not yellow because BRCA1 is not present in G1phase cells. Micrograph taken 30 minutes after induction of damage. (Micrograph courtesy Martin Mistrik and Jiri Bartek.)
biochemical signal as a result of the detected damage, and EFFECTORS (both protein kinases and transcriptional activators) that coordinate repair pathways to block cellcycle progression and repair the damage. When DNA is damaged, several proteins concentrate at the damage site within seconds to minutes, forming a focus that activates the G2/M checkpoint and repairs the damage. The use of high-energy lasers to “draw” patterns of DNA damage onto cell nuclei revolutionized our understanding of the DDR by allowing a minute-byminute mapping of the events during focus formation and leading to DNA repair (Fig. 43.9). If a single strand of DNA is broken, the SENSOR enzyme poly(adenosine diphosphate [ADP]-ribose) polymerase, binds to the broken end and immediately starts polymerizing a chain of ADP-ribose (which it makes from nicotine adenine dinucleotide [NAD]). Several proteins bind to poly(ADP-ribose) chains and recruit the critical TRANSDUCER ATM kinase (ataxia-telangiectasia mutated) to the break together with its partner NBS1, a subunit of the MRN complex (Fig. 43.10A). ATM is usually present in the nucleus as an inactive dimer. Interactions with MRN separate the dimer into ATM monomers that are activated by MRN docked to the broken end of a DNA molecule. The MRN complex is a SENSOR that detects doublestrand DNA breaks. In this case, the NBS1 subunit of MRN recruits ATM kinase to the break site. MRN is also a nuclease with a key role in repairing those breaks. Its job is to chew back (resect) one DNA strand in a 5′ → 3′ direction at the site of a DNA break, leaving a stretch
A
DNA damage
B DNA breaks
ATM dimer (inactive)
DNA strand with damage removed
MRN complex binds to DNA break
RPA binds ssDNA
AT ADP ATM autophosphorylation
ATM monomer (active)
ATRIP
Active ATM binds MRN complex
ATR
ATRIP binds to RPA
ATRIP
γ-H2AX phosphorylated
ATR
ATR active
FIGURE 43.10 MECHANISMS FOR LOCALIZING ATM AND ATR TO SITES OF DNA DAMAGE. ATR, ataxia-telangiectasia and Rad3 related; ssDNA, single-stranded DNA.
CHAPTER 43 n G2 Phase, Responses to DNA Damage, and Control of Entry Into Mitosis
of single-stranded DNA that becomes coated with RPA, the protein that also coats single-stranded DNA at replication forks (Fig. 42.12). RPA then attracts a protein called ATRIP plus its partner, the second key TRANSDUCER, ATR kinase (ataxia-telangiectasia and Rad3 related). ATM and ATR stay bound to the chromatin, which is remodeled by reactions involving protein phosphorylation, acetylation/deacetylation, methylation, and ubiquitylation. The response is very complex. For example, several hundred phosphorylation events have been detected during the DDR. These chromatin changes are likely part of a response to suppress transcription so that RNA polymerase does not collide with the damage and cause more problems. The histone modifications associated with the DDR generally tend to create “open” chromatin that is more accessible to the DNA repair machinery. Repair foci are often found just adjacent to heterochromatin, and indeed, DNA repair appears to be more efficient in euchromatin than heterochromatin. ATR tends to stay on the single-stranded DNA close to the break, but ATM spreads along the chromatin through a cycle of reactions involving its most famous substrate, the specialized histone isoform H2AX (Figs. 43.9 and 43.11). H2AX phosphorylated by ATM is known as γ-H2AX. γ-H2AX forms very rapidly in a focus of modified chromatin immediately surrounding the damage. This amplifies and spreads the response to damage, because γ-H2AX binds a SENSOR protein, MDC1 that recruits more ATM and repair factors. The response spreads along the chromatin as ATM creates more γH2AX, so that after a few minutes the γ-H2AX focus covers about a megabase of DNA. MDC1 has two BRCT domains that bind the phosphorylated site on γ-H2AX. BRCT domains were discovered in BRCA1, one of the first genes whose mutations were linked to familial breast cancer. BRCA1 is mutated in 90% of families where inherited predisposition to
749
ovarian cancer coexists with breast cancer. BRCA1 is a very large, complex protein with numerous roles in DNA repair that are still being determined. ATM (Table 43.1) is dispensable for life, though people lacking it have ataxia-telangiectasia—a disorder associated with degeneration of cerebellar neurons, dilation of blood vessels, a very high predisposition for cancer, and a number of other symptoms. The NBN gene is mutated in humans with Nijmegen breakage syndrome, a rare inherited disorder featuring chromosomal instability and a predisposition to cancer. ATR is essential for life, presumably because it has a key role in ensuring that replication forks are stabilized until all the single-stranded DNA created during S phase is replicated. ATM/ATR
H2AX
Chk1 activated
Chk2 activated
γ-H2AX
MDM2 sequestered
p53 stabilized
Chromatin effects: Cohesin binding Focus formation Recruitment of repair machinery
Cdc25A inhibited p21 14-3-3σ degraded transcribed transcribed Wee1 activated Cdk1–cyclin A inhibited Cdk1–cyclin B1 inhibited G2
S
Other kinases
Cdc25C inhibited sequestered
M
G1
FIGURE 43.11 THE G2/M CHECKPOINT BLOCKS THE G2/M TRANSITION FOLLOWING ACTIVATION OF ATM AND/OR ATR BY DNA DAMAGE. Dotted lines show activities that are switched off by the checkpoint. The dashed line between ATM/ATR and the kinase that inhibits Cdc25C indicates that this pathway is not yet known. (Based on an original figure by Helen Piwnica-Worms.)
TABLE 43.1 Key DNA Repair Gene Defects Associated With Human Disease Human Disease
Pathway
Defective Genes*
Ataxia-telangiectasia (AT)
Checkpoint
ATM
Seckel syndrome
Checkpoint
ATR
Xeroderma pigmentosum
NER
XP-A, XP-B, XP-C, XP-D, DDB-2, XP-F, XP-G, POLH
Cockayne syndrome
NER
XP-B, XP-D, CSA, CSB
Trichothiodystrophy
NER
XP-D
Hereditary nonpolyposis colon cancer
MMR
MSH2, PMS2, MLH1
AT-like disorder
DSB repair
MRE11
Nijmegen breakage syndrome
DSB repair
NBN
Breast cancer predisposition
HR
BRCA1, BRCA2
LIG4 syndrome
NHEJ
LIGIV
Severe combined immune deficiency
NHEJ
ARTEMIS
ATM, ataxia-telangiectasia mutated; ATR, ataxia-telangiectasia and Rad3 related; DSB, double-strand break; HR, homologous recombination; MMR, mismatch repair; NER, nucleotide excision repair; NHEJ, nonhomologous end joining. *This list is an outline. For an updated list of the human DNA repair genes known to date, the reader is referred to http://sciencepark.mdanderson.org/labs/ wood/.
750
SECTION X n Cell Cycle
Table 43.1 shows only a few of the diseases associated with mutations in proteins involved in repairing DNA breaks. (See Box 43.1 for a discussion of the major DNA repair pathways.) It is likely that new components of this pathway remain to be identified.
From the DNA Damage Response to the G2/M Checkpoint ATM and ATR at damage foci cooperate with adapters to phosphorylate and activate the two key EFFECTOR kinases
BOX 43.1 DNA Repair in Vertebrates* Every human cell experiences approximately 105 DNA damage events each day. These are mostly repaired accurately, so only about 100 mutations are passed on in each new human generation. In cancer, the spontaneous mutation frequency can be at least 200-fold higher. Cell division must not occur with inaccurately replicated or damaged genomes, as this may cause cell death or heritable mutation. A number of systems have evolved to repair particular forms of DNA damage sustained by the cell (Fig. 43.12). These
repair mechanisms act in concert with the apoptotic machinery to ensure that the cell will die if the DNA damage cannot be repaired (see Chapter 46). DNA damage checkpoints are a critical component of the cellular response to DNA damage (see Fig. 40.4), as they impose a delay in the cell cycle during which cells have a chance to repair their genomes. Promising new strategies for cancer treatment include use of agents that actually cause DNA damage, with the goal of overwhelming the already defective defenses in cancer cells, and causing them to activate cell death pathways.
Base Excision Repair A. Type of DNA B. Repair lesion pathway Base damage: deamination of cytosine to uracil
U
Base excision repair Base damage: methylation of guanosine
G
A
C
T
T
3' OH
G
C
Nucleotide Excision Repair
Incorporation of mismatched base
Mismatch repair
Formation of UV photoproduct: 6-4 thymidine dimer
Nucleotide excision repair
Double-strand break
Double-strand break repair • Homologous recombinational repair • Nonhomologous end joining
3' OH
Bases in DNA can become oxidized, reduced, alkylated, or deaminated owing to endogenous activities or environmental stress. Damaged bases are cut away from the DNA sugar– phosphate backbone by a damage-recognizing glycosylase, leaving an abasic site (Fig. 43.13A). Abasic sites, which also can be generated directly by DNA damage, are then removed by cleavage of the sugar–phosphate backbone mediated by certain glycosylases and endonucleases. The missing sequence is then reconstructed from its complementary strand by DNA polymerase β, with DNA ligase III-Xrcc1 completing the repair by sealing the gaps in the backbone.
DNA double helix expanded
FIGURE 43.12 INTRODUCTION TO DNA REPAIR. Examples of DNA damage and the repair pathways that respond to different types of lesion.
Bulky DNA adducts caused by chemical agents or environmental stress (particularly UV radiation from sunlight) are excised by a complex, though well-understood, reaction (Fig. 43.13B). Defects in nucleotide excision repair genes cause the human genetic disease xeroderma pigmentosum (XP), which is characterized by hypersensitivity to sunlight and predisposition to skin cancer. Eight proteins encoded by genes mutated in xeroderma pigmentosum (Table 43.1) take part in nucleotide excision repair, providing one of the best examples in which human genetics has helped to unravel a complicated biological process. Recognition of the DNA lesion involves the heterotrimeric replication protein RPA, XPA, and XPC, and the nine-subunit transcription factor TFIIH, which contains XPB and XPD. ATP-dependent unwinding of the DNA by XPB and XPD forms a preincision complex. XPG, which replaces XPC in the complex, makes an incision six to nine bases 3′ of the damaged base, and XPF-Ercc1 cuts 20 to 25 bases 5′ of the damage site. This releases a short single-stranded DNA fragment containing the damaged DNA. After excision, DNA polymerases δ or ε fill in the gap by copying the undamaged strand. Prokaryotes have a similar system of adduct recognition, removal, and repair involving the UvrA, UvrB, and UvrC proteins; however, the enzymes that are involved are not conserved between kingdoms.
CHAPTER 43 n G2 Phase, Responses to DNA Damage, and Control of Entry Into Mitosis
751
BOX 43.1 DNA Repair in Vertebrates*—cont’d A. Base excision repair
B. Nucleotide excision repair
DNA damage
Glycosylase removes damaged base
DNA damage RPA, XPA, XPC Recognition TFIIH (contains XPB and XPD)
Abasic site
DNA polymerase β DNA ligase III-Xrcc1
C. Mismatch repair Mismatch
Recognition MSH2/MSH6 dimer Nick
ATP-dependent unwinding by RPA, XPA, XPC, TFIIH Sliding clamp XPB, XPD searches for nick Msh2/Msh6 dimer plus (Okazaki frag.) Mlh1/Pms2 dimer form sliding clamp Preincision complex forms May travel thousands of bases
XPF-Ercc1 nicks 20–25 bp to 5'
XPG nicks 6–9 bp to 3'
DNA polymerase δ, ε
Exonuclease degrades back Exo1 to mismatch
DNA polymerase δ, ε
FIGURE 43.13 PATHWAYS FOR THE REPAIR OF BASE DAMAGE, BULKY ADDUCTS SUCH AS THYMIDINE DIMERS FORMED BY ULTRAVIOLET LIGHT, OR MISMATCHED BASES. For detailed descriptions, see the text. The inset in panel A shows human 3-methyladenine DNA glycosylase complexed to DNA. This enzyme scans the DNA for bases that are not strongly H-bonded, uses its “finger” to swing them up into the pocket for scanning, and, if they are damaged, catalyzes excision of that base. (Inset illustration by Graham Johnson [www.fivth.com] for the Howard Hughes Medical Institute, copyright 2004, all rights reserved. For reference, see PDB file 1BNK and Lau AY, Scharer OD, Samson L, et al. Crystal structure of a human alkylbase-DNA repair enzyme complexed to DNA: mechanisms for nucleotide flipping and base excision. Cell. 1998;95:249–258.)
Mismatch Repair
Double-Strand Break Repair
Errors in DNA replication missed by the proofreading activity of the DNA polymerase are recognized by a dimer consisting of the MSH2 and MSH6 proteins. When a mismatch is detected, this heterodimer undergoes an ATP-dependent transition to a sliding clamp and recruits a second heterodimer, consisting of MLH1 and PMS2 (Fig. 43.13C). To distinguish between the original (“correct”) sequence and the newly synthesized DNA strand, this sliding clamp complex can then translocate along the DNA until a break is reached, such as that found between Okazaki fragments. The broken strand is therefore identified as the newly synthesized DNA strand. The mismatch repair complex then recruits the exonuclease EXO1 and degrades the newly synthesized DNA strand all the way back to the misincorporated base. The resultant long, single-stranded region is stabilized by binding of RPA and eventually filled in by the replicative DNA polymerases δ and/or ε. A second type of “mismatch” involves mistaken incorporation of ribonucleotides into DNA by either DNA polymerases δ and ε. Specialized ribonuclease H (RNAse H) enzymes detect and cleave DNA/RNA hybrids and remove the ribonucleotides.
DNA double-strand breaks can be caused by ionizing radiation or radiomimetic drugs or arise spontaneously after replication. They are particularly hazardous forms of damage, as they carry the risk of losing chromosomal material or, if misrepaired, causing chromosomal translocations. Two major pathways repair double-strand breaks. Homologous recombinational repair uses undamaged DNA as a template for the accurate repair of double-strand breaks, this sequence usually being derived from the sister chromatid after replication. Nonhomologous end joining (NHEJ) repairs doublestrand breaks with no requirement for homology. NHEJ is the predominant activity that repairs double-strand breaks in the G1 and early S phase, whereas homologous recombination becomes more important in the late S and G2 phase. Homologous recombinational repair is normally extremely accurate, but NHEJ often introduces errors, as it can join both related and unrelated DNA ends together. Both pathways require the activity of the MRN protein complex (Mre11/RAD50/Nbs1), which localizes to DNA double-strand breaks and is also found at telomeres. The exonuclease activity of this complex resects (chews back) broken DNA ends Continued
752
SECTION X n Cell Cycle
BOX 43.1 DNA Repair in Vertebrates*—cont’d A number of proteins, including BRCA1 and BRCA2, control the activity of mammalian Rad51 in homologous recombinational repair. Inactivation of BRCA1 and BRCA2 predisposes to cancer. The helicase Rad54 facilitates strand invasion, when the single-stranded region forces its way into the complementary DNA duplex on the undamaged sister chromatid. Following invasion of the recombining DNA strands, polymerase activity extends the DNA beyond the site of the double-strand break, leading to the formation of a Holliday junction (Fig. 43.14B). Resolution of the Holliday junction
to provide single-stranded DNA substrates for the repair systems. These mechanisms repair also repair double-strand breaks during genome editing (see Fig. 6.16). The key protein required for homologous recombinational repair in cells is Rad51, the eukaryotic homologue of Escherichia coli RecA (Fig. 43.14A). Rad51 forms an extended nucleoprotein filament on single-stranded DNA, replacing RPA, which acts like a placeholder when singlestranded DNA is produced. Rad51 catalyses the search for homologous sequences, strand pairing, and strand exchange. A. Homologous recombinational repair
C. Nonhomologous end joining DNA break
Double-strand break
5' to 3' resection 5'
Binds ends Recruits DNA- Ku70/ dependent Ku80 ring protein kinase
MRN complex Rad50/Mre11/Nbs1
MRN complex Aligns broken ends Other repair factors
3'
3' 5'
5'
3'
Initial strand invasion
Rad51
3'
DNA ligase IV/ Ends sealed Xrcc4
5'
New DNA synthesis
B. Holliday junction structure 3'
Second end capture, synthesis, ligation
5'
5'
= Resolvase cutting sites
3'
Holliday junction
Holliday junction resolution 3' 5' 3'
3' 3'
5'
+
5' 3'
5' 5'
3'
Repair complete 5'
FIGURE 43.14 PATHWAYS FOR THE REPAIR OF DNA DOUBLE-STRAND BREAKS. A double-strand break is recognized by the MRN complex, which recruits ATM (ataxia-telangiectasia mutated). The break is then bound by repair factors involved in either the homologous recombination or nonhomologous end-joining pathway of DNA repair. A, Homologous recombination pathway of DNA repair. The MRN complex in conjunction with other nucleases chews back (resects) the DNA at a break, leaving a single-stranded overhang that is stabilized by RPA (not shown). This recruits ATR (ataxia-telangiectasia and Rad3 related) to the damage site. It is believed that the MRN complex also plays a role in keeping the broken ends close to one another. Next, RAD51 forms a nucleoprotein filament on the single-stranded DNA, displacing the RPA. The Rad51 nucleoprotein filament then initiates homology searching and repairs the DNA break by inserting the extended single-stranded DNA into homologous sequences (usually on the sister chromatid [blue]) and allowing homologous recombination and DNA repair/resynthesis to occur. Capture of the second single-stranded DNA end allows the formation of a joint molecule with a double Holliday junction. Resolution of this Holliday junction structure results in accurate, templated repair of the double-strand break. B, A Holliday junction formed by four complementary oligonucleotides complexed to the enzyme Cre (not shown). The Holliday junction is a dynamic structure (arrows) that can migrate along the DNA. C, Nonhomologous end-joining pathway of DNA repair. This pathway is initiated by break recognition by the Ku70/Ku80 heterodimer, which recruits DNA-PK and tethers the broken ends. The breaks are then processed in a reaction involving the MRN complex and other repair factors. DNA-PK’s precise role is not yet entirely clear. Next, DNA ligase IV/XRCC4 is recruited to the processed double-strand break, which is ligated back together. (B, For reference, see PDB files 3CRX and 2CRX and Gopaul DN, Guo F, Van Duyne GD: Structure of the Holliday junction intermediate in Cre-loxP site-specific recombination. EMBO J 1998;17: 4175–4187.)
CHAPTER 43 n G2 Phase, Responses to DNA Damage, and Control of Entry Into Mitosis
753
BOX 43.1 DNA Repair in Vertebrates*—cont’d and filling-in of the repaired DNA sequences results in complete repair of the lesion. One of two known human Holliday junction resolvases is the Gen1 nuclease, a relative of the Fen1 flap endonuclease that functions during DNA replication (see Fig. 42.12). (The other more complex resolvase is the product of four genes.) Cells must have one or the other of these enzymes or they die during mitosis because of an inability to separate DNA strands that have undergone repair in the previous cell cycle. Note that Fig. 43.14 shows only a subset of the proteins involved in homologous recombinational repair. It is likely that new factors remain to be identified. NHEJ is initiated at a DNA double-strand break by binding of the Ku70 and Ku80 heterodimer as a ring that binds the
catalytic subunit of the DNA-dependent protein kinase, stimulating other repair factors and aligning the broken ends of the DNA (Fig. 43.14C). A complex of the protein XRCC4 with DNA ligase IV seals the ends of the broken DNA. NHEJ is also necessary for V(D)J recombination and therefore for the development of the immune system (see Fig. 28.10). Given the importance of accurate transmission of the genetic material, deficiencies in DNA repair and checkpoint genes are associated with a number of diseases (Table 43.1). Note that several DNA repair activities are essential for life, so their inactivation has not been described in any human diseases.
*By Ciaran Morrison, National University of Ireland, Galway.
Chk1 and Chk2. Phosphorylation of the ATM adapter, 53BP1 (p53-binding protein 1), recruits Chk2 for activation by ATM. The trimeric 9-1-1 complex is required for Chk1 activation by ATR. This complex, which gets its name from its subunits Rad9, Hus1, and Rad1, resembles proliferating cell nuclear antigen (PCNA), the doughnut-shaped processivity factor that is indispensable during DNA replication (see Fig. 42.12). PCNA is loaded onto DNA by replication factor C (RFC, a pentameric AAA-ATPase) and anchors DNA polymerases and other factors to DNA. A similar ATPase composed of one special subunit, Rad17, plus the four small subunits of RFC loads the 9-1-1 complex onto DNA at or near sites of damage. Mutants in those four RFC subunits are defective in G2/M checkpoint control in yeasts, Drosophila and Caenorhabditis elegans. RPA stimulates loading of the 9-1-1 complex at damage sites, making it specific for regions of single-stranded DNA. Phosphorylation releases Chk1 and Chk2 from chromatin, so they can diffuse throughout the nucleus and cell to implement the DDR. They also trigger the G2/M checkpoint response by altering the cell cycle machinery and inducing the transcription of key EFFECTORS. In some cases their actions trigger cell death by apoptosis. Activation of Chk1 is important to establish the G2/M checkpoint response because Chk1 phosphorylates the Cdc25A and Cdc25C protein phosphatases thereby blocking cell-cycle progression (Fig. 43.11). Phosphorylation produces binding sites for a 14-3-3 protein that blocks Cdc25A from activating Cdk1–cyclin B. Chk1 phosphorylation also targets Cdc25A for ubiquitinmediated proteolysis ensuring that levels of Cdc25A remain low. The transcription factor p53 (see Fig. 41.15), another EFFECTOR of the G2/M checkpoint, is phosphorylated and activated following DNA damage (Fig. 43.11). Activated
p53 stimulates transcription of the Cdk inhibitor p21. Although p21 is best known for promoting cell-cycle arrest in G1 cells as part of the G1 DNA damage checkpoint, it can also act in G2. Expression of p21 is an effective way of blocking the initiation of prophase, because it inhibits Cdk1–cyclin A approximately 100-fold better than it inhibits Cdk1–cyclin B1. Active p53 also drives the expression of 14-3-3σ, an adapter protein that interferes with shuttling of Cdk1– cyclin B1 between the nucleus and cytoplasm (Fig. 43.11). Binding of 14-3-3σ maintains the Wee1 inhibitory kinase in a more active state, ensuring that the Cdk1– cyclin B1 complex remains inactive. Disruption of the gene for 14-3-3σ is fatal for cells with DNA damage. Instead of activating their G2/M checkpoint, they enter an aberrant state with characteristics of both mitosis and apoptosis, and then die. Typically, G2/M checkpoint activation has one of three outcomes. If DNA damage is so extensive that it cannot be repaired, the cell either enters a non-proliferating state known as senescence or commits suicide by apoptosis (see Chapter 46). Less-serious damage can be repaired by one of the systems described in Box 43.1.
Transition to Mitosis The complex web of stimulatory and inhibitory activities in the G2 phase poises Cdk1–cyclin B in a state ready for the explosive burst of activation that triggers the G2/M transition. These complex regulatory pathways are the basis of the G2/M checkpoint control that prevents cells from segregating their chromosomes if genomic DNA cannot meet stringent quality control standards. Eventually, however, if all goes well, Cdk1–cyclin A and Cdk1– cyclin B1 are activated, and the cell embarks on mitosis, probably the most dramatic event of its life.
754
SECTION X n Cell Cycle
ACKNOWLEDGMENTS We thank Anton Gartner, Ciaran Morrison, and Ashok Venkitaraman for their suggestions on revisions to this chapter. SELECTED READINGS Ciccia A, Elledge SJ. The DNA damage response: making it safe to play with knives. Mol Cell. 2010;40:179-204. Hartwell LH, Weinert TA. Checkpoints: Controls that ensure the order of cell cycle events. Science. 1989;246:629-634. Jackson SP, Bartek J. The DNA-damage response in human biology and disease. Nature. 2009;461:1071-1078. Mehta A, Haber JE. Sources of DNA double-strand breaks and models of recombinational DNA repair. Cold Spring Harb Perspect Biol. 2014;6:a016428. Melo J, Toczyski D. A unified view of the DNA-damage checkpoint. Curr Opin Cell Biol. 2002;14:237-245.
Morgan DO. The Cell Cycle: Principles of Control. London: New Science Press; 2007. Parrilla-Castellar ER, Arlander SJ, Karnitz L. Dial 9-1-1 for DNA damage: the Rad9-Hus1-Rad1 (9-1-1) clamp complex. DNA Repair (Amst). 2004;3:1009-1014. Paull TT. Mechanisms of ATM activation. Annu Rev Biochem. 2015;84: 711-738. Pearl LH, Schierz AC, Ward SE, et al. Therapeutic opportunities within the DNA damage response. Nat Rev Cancer. 2015;15:166-180. Polo SE, Jackson SP. Dynamics of DNA damage response proteins at DNA breaks: a focus on protein modifications. Genes Dev. 2011;25: 409-433. Smits VA, Medema RH. Checking out the G(2)/M transition. Biochim Biophys Acta. 2001;1519:1-12. Wieser S, Pines J. The biochemistry of mitosis. Cold Spring Harb Perspect Biol. 2015;7:a015776. Wood RD, Mitchell M, Sgouros J, et al. Human DNA repair genes. Science. 2001;291:1284-1289.
CHAPTER
44
Mitosis and Cytokinesis
M
itosis is the division of a somatic cell (a vegetative cell in yeast) into two daughter cells. The daughters are usually identical copies of the parent cell, but the process can be asymmetrical. For example, division of some stem cells gives rise to one stem cell and another daughter cell that goes on to mature into a differentiated cell. See Box 41.2 for examples. Traditionally, mitotic events are subdivided into six phases: prophase, prometaphase, metaphase, anaphase, telophase, and cytokinesis (Fig. 44.1). The dramatic reorganization of both the nucleus and cytoplasm during the mitotic phases is brought about by activation of a number of protein kinases, including Cdk1–cyclin B–cks (abbreviated here as “Cdk1 kinase”; see Chapter 40). After activation by Cdc25 phosphatase, Cdk1 kinase accumulates in the nucleus, where it joins Cdk1–cyclin A–cks, which was activated somewhat earlier (see Fig. 43.6). These two Cdk1 kinase complexes operate as both master controllers and workhorses that directly phosphorylate many proteins whose functional and structural status is altered during mitosis. Their progressive inactivation following the correct attachment of the chromosomes to spindle microtubules drives the orderly exit of cells from mitosis. Mitosis is an ancient process, and a number of variations emerged during eukaryotic evolution. Many singlecelled eukaryotes, including yeast and slime molds, undergo a closed mitosis, in which spindle formation and chromosome segregation occur within an intact nuclear envelope to which the spindle poles are anchored. This chapter focuses on open mitosis, as used by most plants and animals, in which the nuclear envelope disassembles before the chromosomes segregate. Fig. 44.2 summarizes some of the important events during the various mitotic phases.
Prophase
Early prometaphase CDK activity high
Metaphase
Late prometaphase
Anaphase
Telophase APC/C activity high
Early cytokinesis
Late cytokinesis
FIGURE 44.1 OVERVIEW OF THE PHASES OF MITOSIS. APC/C, anaphase-promoting complex/cyclosome.
Prophase Prophase, the transition from G2 into mitosis, begins with the first visible condensation of the chromosomes and disassembly of the nucleolus (Fig. 44.3). In the 755
756
SECTION X n Cell Cycle A. Interphase
B. Prophase
C. Prometaphase
D. Metaphase
Nucleus Centrosome Chromosomes NE
Microtubules Centrosomes separate Chromosomes condense
E. Anaphase A
Begins with nuclear envelope (NE) break-down Chromosomes attach to spindle
Chromosomes align on spindle equator
G. Telophase
H. Cytokinesis
F. Anaphase B
NE CS Midbody CF
CF
CS remnant
CS Pole Sister chromatids separate and move to poles
Organized central spindle (CS) assembles Poles (arrows) separate Cleavage Furrow (CF) assembles
Cleavage furrow (CF) constricts Nuclear envelope (NE) reassembles
Chromosomes decondense Interphase microtubule network reforms Daughter cells separate
FIGURE 44.2 KEY EVENTS OF MITOSIS. A–C, Prophase–prometaphase: Cdk1 kinase triggers condensation of replicated sister chromatids, disassembly of the nuclear envelope and Golgi, and a dramatic reorganization of the cytoskeleton. As the barrier between the chromosomes and cytoplasm is abolished, microtubules contact the condensed chromosomes and attach at the kinetochores (see Fig. 8.21). Interaction of kinetochores with dynamic microtubules culminates with the chromosomes aligned at the midplane of the bipolar spindle. D–F, Metaphase–anaphase: Once all chromosomes achieve a bipolar attachment to the spindle, an inhibitory signal is silenced, leading to activation of a proteolytic network that destroys proteins responsible for holding sister chromatids together and also inactivates Cdk1 by destroying its cyclin B cofactor (see Fig. 40.15). Sister chromatids separate and move toward opposite spindle poles, which themselves move apart. G–H, Telophase–cytokinesis: Targeting of nuclear envelope components back to the surface of the chromatids leads to the reformation of two daughter nuclei. In most cells, the two daughter nuclei and the surrounding cytoplasm are partitioned by cytokinesis. (Micrographs courtesy William C. Earnshaw.)
cytoplasm, the interphase network of long microtubules centered on a single centrosome (see Fig. 34.18) is converted into two radial arrays of short microtubules called asters. Most types of intermediate filaments disassemble, the Golgi apparatus and endoplasmic reticulum fragment, and both endocytosis and exocytosis are curtailed.
Nuclear Changes in Prophase Chromosome condensation, the landmark event at the onset of prophase, often begins in isolated patches of chromatin at the nuclear periphery. Later, chromosome condense into two threads termed sister chromatids that are closely paired along their entire lengths. Although chromosome condensation was first observed more than
a century ago, the biochemical mechanism remains a mystery. Protein kinases trigger mitotic chromosome condensation and onset of condensation is correlated with phosphorylation of histones H1 by Cdk1 kinase and H3 by Aurora-B protein kinase. However, chromosomes still condense when both of these phosphorylation events are blocked. It is possible that a combination of histone modifications promotes mitotic chromatin condensation (see Fig. 8.3). Two pentameric protein complexes, condensin I and condensin II are major regulators of mitotic chromosome architecture. These complexes share the SMC2 and SMC4 (structural maintenance of chromosomes) ABC adenosine triphosphatases (ATPases), but have two different sets of three auxiliary proteins (see Fig. 8.18).
CHAPTER 44 n Mitosis and Cytokinesis
A. Prophase Phosphorylated condensin enters nucleus Histone H3 phosphorylation begins Cell surface markers internalized Intracellular membrane networks remodeled
Chromosome condensation begins Duplicated centrioles begin to separate Microtubule halflife decreases and asters form Cell begins to round up
757
B
DNA Microtubules Centrosomes
FIGURE 44.3 INTRODUCTION TO PROPHASE. A, Summary of the major events of prophase. B, Distribution of DNA (blue), microtubules (red), and γ-tubulin (centrosomes [green]) in a prophase human cell. (B, Images were recorded by Dr. Melpomeni Platani on the University of Dundee’s School of Life Sciences Imaging Facility OMX 3DSIM Microscope and stored and processed in OMERO.)
Condensins are required to disassemble the topologically associating domains (TAD) of interphase chromatin (see Fig. 8.13) and promote the formation of the linear array of loops that characterizes mitotic chromosomes (see Fig. 8.14). The decisive experiment was to target SMC2 for rapid degradation, with the result that no wellorganized mitotic chromosomes formed. The chromatin compacted but was less organized. Slower depletion of condensins by RNA interference (RNAi) or conditional gene disruption gave less-dramatic results. Condensin complexes can encircle DNA and also promote its supercoiling in vitro, but how these activities help them to orchestrate the changes in chromatin architecture is not known. Condensin II is nuclear during interphase, and has an important role in prophase chromosome formation following its activation by phosphorylation. Condensin I acts both in prophase and prometaphase in further compacting the chromosomes. DNA topoisomerase II and other scaffold proteins also function during mitotic chromosome formation, but condensin appears to play the major role.
Cytoplasmic Changes in Prophase Most of the cytoskeleton reorganizes during prophase. The microtubule array changes from an extensive network permeating the cytoplasm into two dense, radial arrays of short, dynamic microtubules around the duplicated centrosomes (see Fig. 34.17). Each of these asters eventually becomes one pole of the mitotic spindle. During prophase, the two asters usually migrate apart across the surface of the nuclear envelope, signaling the start of spindle assembly (Fig. 44.3). Mitotic microtubules behave like interphase microtubules in many ways (see Chapter 34). They are mostly nucleated at their minus ends, they grow by addition of tubulin subunits at their plus ends, and they undergo random catastrophes during which they rapidly shorten. To a large extent, the prophase changes in microtubule organization can be explained by two simple biochemical changes: (a) increased microtubule-nucleating activity of centrosomes and (b) altered dynamic instability
TABLE 44.1 Comparison of Microtubule Dynamics in Interphase and Mitotic Newt Lung Cells Parameter
Interphase
Mitosis
Elongation rate
7 µm/min
14 µm/min
Elongation time before catastrophe
71 s
60 s
Shortening rate
17 µm/min
17 µm/min
Probability of rescue from catastrophe*
0.046/s
0
Length
100 µm
14 µm
*Most cellular microtubules grow constantly by addition of subunits to their free ends but they occasionally stop growing and begin shrinking rapidly (a “catastrophe”). Unless shrinking is reversed (a “rescue”), the microtubule completely disappears. Data from Gliksman NR, Skibbens RV, Salmon ED: How the transition frequencies of microtubule dynamic instability regulate microtubule dynamics in interphase and mitosis. Mol Biol Cell. 1993;4:1035–1050.
properties of the microtubules (Table 44.1; also see Chapter 34). Interphase microtubules have a high probability of recovering from catastrophes, so they grow quite long. Mitotic microtubules grow more rapidly but exist only transiently, because rescues are rare following a catastrophe. Thus, they usually shorten all the way back to the centrosome, with little chance of rescue. These differences in dynamic instability can be reproduced in vitro in mitotic and interphase cellular extracts. They appear to arise, at least in part, from counterbalancing interactions between microtubuleassociated proteins that promote microtubule stability, and kinesin-13 (see Fig. 36.13), which promotes microtubule disassembly. Other cytoskeletal elements that disassemble during prophase include many, but not all, classes of intermediate filaments (including the nuclear lamins) (see Fig. 35.1A) and specialized actin filament structures, such as stress fibers. However, the junctional complexes between adjoining cells are maintained in epithelial cells. As a result of the cytoskeletal reorganization, most cells round up during prophase. This is particularly evident for animal cells that are cultured on a flat
758
SECTION X n Cell Cycle
substrate, but cells in tissues also change their shape dramatically during mitosis. RNA transcription of the chromosomes stops during mitosis except for highly specialized transcription at centromeres. Phosphorylation of components of the transcriptional machinery by Cdk1 kinase appears to be responsible for this shutoff. Cdk1 kinase phosphorylation of ribosomal elongation factor 2a (EF2a) also stops most (but not all) ongoing protein synthesis and assem-
Interphase
Mitosis
Golgi
Ministacks
Golgi stacking proteins
GM130
Cdk1– cyclin B–p9
p115
bly of new ribosomes. Phosphorylation of several nucleolar proteins leads to disassembly of the nucleolus. The Golgi apparatus and endoplasmic reticulum fragment and disperse during prophase (Fig. 44.4). Several kinases, including Cdk1 drives Golgi apparatus disassembly, the first step being fragmentation into smaller ministacks following phosphorylation of Golgi stacking proteins and tethers. Later steps are still being investigated. Many lines of evidence argue that Cdk1 phosphorylation of key components prevents the fusion of transport vesicles back into Golgi stacks (see Chapter 21), the net result being that the Golgi buds away into small vesicles that disperse throughout the mitotic cell cytoplasm. Other evidence suggests that an imbalance of vesicle flow between the Golgi and the endoplasmic reticulum results in the Golgi being absorbed into the endoplasmic reticulum during mitosis. Whatever the mechanism of its disassembly, Golgi reassembly begins again during late anaphase/early telophase, following inactivation of Cdk1 kinase.
Prometaphase
FIGURE 44.4 GOLGI APPARATUS DYNAMICS IN INTERPHASE AND MITOSIS. Disassembly in mitosis is driven by phosphorylation of components blocking fusion of Golgi membranes.
A. Early prometaphase 1
In cells that undergo an open mitosis, prometaphase begins abruptly with disassembly of the nuclear envelope (Fig. 44.5). Microtubules growing outward from the spindle poles penetrate holes in the nuclear envelope, make contact with the chromosomes, and attach to them
C. Late prometaphase 4
2
E. Kinetochore attachments to the spindle Tension
6 3
5
(+) ends
4. Chromosome slides rapidly poleward along microtubule 5. Microtubule from opposite pole is captured by sister kinetochore 6. Chomosome attached to both poles congresses to middle of spindle
1. Nuclear envelope disassembles 2. Microtubules grow and shrink in aster 3. Kinetochore captures microtubule
B
Amphitelic (correct)
D Syntelic (errors)
DNA Microtubules Centrosomes
Merotelic (errors)
FIGURE 44.5 INTRODUCTION TO PROMETAPHASE. A, Summary of the key events of early prometaphase. B, Distribution of DNA (blue), microtubules (red), and γ-tubulin (centrosomes [green]) in early prometaphase human cells. C, Summary of the key events of late prometaphase. D, Distribution of DNA, actin, microtubules, and centrosomes in late prometaphase PtK1 cells. E, Terms used to describe the orientation of kinetochore attachments to the mitotic spindle. (B and D, Images were recorded by Dr. Melpomeni Platani on the University of Dundee’s School of Life Sciences Imaging Facility OMX 3DSIM Microscope and stored and processed in OMERO.)
759
CHAPTER 44 n Mitosis and Cytokinesis
at specialized structures called kinetochores (see Fig. 8.19). Interactions of the two opposing kinetochores of paired sister chromatids with microtubules from opposite poles of the spindle ultimately result in alignment of the chromosomes in a group midway between the poles. An important cell-cycle checkpoint (see Chapter 40) known as the spindle assembly checkpoint (SAC) delays the onset of chromosome segregation until all kinetochores are attached to microtubules.
Nuclear Envelope Disassembly in Prometaphase Nuclear envelope disassembly involves the removal of two membrane bilayers coupled with disassembly of the nuclear pores and the fibrous nuclear lamina meshwork that underlies the inner bilayer (Fig. 44.6). Phosphorylation causes the nucleoporin Nup98 to dissociate from nuclear pores. This removes the permeability barrier between nucleus and cytoplasm. Phosphorylation of other proteins causes the pore to disassemble to soluble subcomplexes. Phosphorylation of the nuclear lamins at two sites flanking the coiled-coil causes the lamina network to disassemble into subunits. Interaction between microtubules and dynein associated with the nuclear envelope can rip holes in the envelope, although this is not required for nuclear envelope disassembly. Nuclear envelope membranes are dispersed in the cytoplasm from prometaphase until telophase (Fig. 44.6), but the mechanism is not settled. Some experiments suggest that the nuclear membranes break up into small vesicles that disperse in the cytoplasm. Other experiments suggest that the nuclear envelope is absorbed into the endoplasmic reticulum, which remains
A. Interphase
B. Mitosis
C
as an extensive tubular (or flattened cisternal network— another source of discussion) throughout mitosis. Further experiments are required to answer this question, and both mechanisms could contribute. Lamin B remains associated with the dispersed nuclear envelope, whereas lamins A and C and many proteins of the nuclear pore complexes disperse as soluble subunits. During prophase, kinetochores transform from nondescript balls of condensed chromatin into structures on the surface of the chromosomes. By early prometaphase, the characteristic trilaminar disk structure (see Fig. 8.19) can be seen. Each sister chromatid has a kinetochore. Sister kinetochores are located on opposite faces of the mitotic chromosome.
Organization of the Mitotic Spindle The mature metaphase spindle is a bilaterally symmetrical structure with centrally located chromosomes flanked by arrays of microtubules radiating from the poles (Fig. 44.7). Three predominant classes of microtubules are present in the metaphase spindle (Fig. 44.12). Kinetochore microtubules have their plus ends embedded in the kinetochore and their minus ends at or near the spindle pole. They characteristically form bundles, called kinetochore fibers, which contain anywhere from 1 microtubule in the budding yeast to more than 200 microtubules in some higher plants. Each human kinetochore binds approximately 20 microtubules. Up to approximately 80% of the approximately 2200 spindle microtubules in humans may be present in kinetochore fibers, but not all microtubules in those fibers stretch all
D. Lamins A and C
E. Lamin B
Endoplasmic reticulum
Nuclear membrane: Outer Inner
Lamina disassembly Nuclear lamina Condensing chromosomes
Interphase
HO
Somatic cells?
O OCH3
Cdk1– cyclin B–p9
Eggs?
PO4– O
PO4– O C
C
G2 / M
G2 / M
PO4
PO4 O
C
Earliest prometaphase
OH
O PO4–
OCH3 OH
C
O O–
O
O– O PO4–
Lamin B
+ Chromosomes
or Vesicles derived from the nuclear envelope
Lamin B dispersed through the endoplasmic reticulum
Metaphase
Lamina
Soluble subunits
Lamina
Subunits on membrane vesicles
FIGURE 44.6 DISASSEMBLY OF THE NUCLEAR ENVELOPE DURING MITOSIS. A–B, Two contrasting models to explain the fate of the nuclear envelope during the transition from interphase to mitosis in a higher eukaryote. C, Micrographs showing solubilization of lamin A fused to green fluorescent protein (GFP) (green) during mitosis. DNA is blue. Scale bar is 10 µm. D–E, Reversible disassembly of lamins A, C, and B is driven by posttranslational modifications of the lamin polypeptides. (C, Courtesy William C. Earnshaw.)
760
SECTION X n Cell Cycle
A. Metaphase–forces balanced, spindle length stable Cytoplasmic dynein moves toward minus (–) end (+) (+)
(–) (–) (–)
(+)
(+) (+)
(–)
(+)
Flux
(+)
Microtubule disassembly at kinetochores plus flux moves sister chromatids toward poles
(+)
Microtubule
B. Anaphase–forces elongating the spindle dominate
(+) (–) (–) (–) (–) (–)
(+)
(+)
(+)
Kinesin-13
Ordered central spindle assembles (–) (–)
(+)
(–) (–) (–)
(–)
(+) (+)
(–)
(+)
(–)
(+)
(–) (+)
NuMA
Inward force from minus-end–directed kinesin-14 and microtubule disassembly at poles by kinesin-13 is balanced by outward force from kinesin-5 plus dynein stretching poles apart
(–)
(+)
Bipolar plus-end–directed kinesin-5 dominates to elongate spindle, pushing poles apart
Microtubule assembly at kinetochores and disassembly at poles causes tubulin subunit flux along kinetochore microtubules FIGURE 44.7 ROLE OF MOTOR PROTEINS IN SPINDLE DYNAMICS. Mitotic spindle structure depends on microtubule assembly/ disassembly plus balanced forces that slide microtubules relative to one another and to pull the poles together or push them apart. A, In metaphase, the structure is at steady state. Forces that tend to elongate the spindle, including cytoplasmic dynein (which moves toward microtubule minus ends, pulling the poles out toward the cell cortex) and bipolar kinesin-5 (which moves toward microtubule plus ends, pushing the poles apart), are counterbalanced by cohesion between sister chromatids and kinesin-14, which moves toward microtubule minus ends (and pulls the poles together) and microtubule disassembly at the spindle poles. Dynein and its associated protein, NuMA (nuclear mitotic apparatus), also help to organize a focused spindle pole. B, In anaphase, sister chromatids separate, the balance of kinesin activity shifts, microtubule disassembly at the poles declines, and the spindle undergoes a dramatic elongation. During anaphase, bipolar kinesin-5, chromokinesin KIF4A, and protein regulated in cytokinesis 1 (PRC1) also have important roles in organizing the central spindle, which is essential for subsequent assembly and function of the cleavage furrow.
the way from the kinetochores to the spindle poles. Interpolar microtubules are distributed throughout the body of the spindle and do not attach to kinetochores. Many interpolar microtubules penetrate between and through the chromosomes and extend for some distance beyond them. Thus, the central spindle contains a large number of interdigitated antiparallel microtubules. Tracking these spindle microtubules by electron microscopy revealed a tendency for the interdigitated microtubules of opposite polarity to pack next to one another. During late anaphase, these antiparallel microtubules bundle to form a structure called the central spindle that plays important roles in signaling during cytokinesis. Astral microtubules project out from the poles and have a role in orienting and positioning the spindle in the cell through interactions with the cell cortex in somatic cells. All microtubules within each aster have the same polarity, with their minus ends proximal to the pole. Each unit of a spindle pole, with its associated kinetochore and interpolar and astral microtubules, is referred to as a half-spindle. Spindle structure is largely determined by a combination of microtubule dynamics plus the action of at least seven different types of kinesins and cytoplasmic dynein (see Chapter 36). These motors often work in opposition to one another. Furthermore, forces exerted by motors can influence the dynamic assembly/disassembly of
microtubules. As a result, the highly dynamic spindle changes shape as a delicate balance of forces shifts between the various motors. For example, inactivating one or more kinesins with drugs or switching a temperature-sensitive mutant to the nonpermissive temperature can cause the spindle to collapse rapidly on itself.
Spindle Assembly In metazoans, spindle assembly starts in prophase with the separation of the asters. In most cells, each aster is organized around a centrosome, consisting of a centriole pair and associated pericentriolar material. γ-Tubulin ring complexes in the pericentriolar material efficiently nucleate microtubules (see Fig. 34.16), so each centrosome acts as a microtubule organizing center (MTOC). By the end of prophase, the spindle consists of two asters linked by a few interpolar microtubules. Cytoplasmic dynein at the cell cortex exerts an outward force separating the centrosomes, whereas kinesin-14 motors (which move toward microtubule minus ends) on the interpolar microtubules exert a counterbalancing force holding the asters together. This balance of forces changes when the nuclear envelope breaks down. Bipolar kinesin-5 motors phosphorylated by Cdk1 kinase concentrate in the central spindle, where they crosslink adjacent antiparallel
CHAPTER 44 n Mitosis and Cytokinesis
interpolar microtubules. Kinesin-5 moves toward the plus ends of microtubules, so when attached to two adjacent antiparallel microtubules, its action will cause them to slide and push the spindle poles apart (Fig. 44.7). However, the two half-spindles do not separate because they are physically linked via the chromosomes, with sister kinetochores attached to opposite spindle poles. Also at this time, the asters mature into focused spindle poles. The pericentriolar material efficiently nucleates the assembly of new microtubules with their minus ends at the pole. Microtubule assembly at two other sites also contributes to spindle morphology. At kinetochores, a complex derived from interphase nuclear pores recruits γ-tubulin. This locally nucleates microtubules that grow by inserting subunits at the kinetochores, pushing the minus ends with γ-tubulin outwards toward the spindle poles. These preformed kinetochore fibers are then incorporated into the spindle. In the central spindle, microtubules are nucleated on the walls of other microtubules by γ-tubulin recruited by the multi-subunit augmin complex. The action of augmin creates branched microtubules throughout the central spindle, contributing to an even fir tree-like distribution of microtubules. The daughter microtubules may be released from their nucleation site, free at both ends, and move toward the spindle poles as part of the flux of tubulin from the center of the spindle toward the centrosomes. The microtubule array focuses at the poles partly because centrosomes tether microtubules, and partly due to the concerted action of various motors and microtubule crosslinking proteins such as dynein and nuclear mitotic apparatus (NuMA) protein. NuMA is released from the nucleus when the nuclear envelope breaks down, and it accumulates near the poles at the minus ends of microtubules. Large cells that lack centrosomes, such as eggs, form spindles by an alternative pathway that also functions in the background in cells with centrosomes (Fig. 44.8). This mechanism hijacks the nuclear trafficking system to
761
enable chromosomes to control the activity of key spindle assembly factors. The nuclear import receptors importin α and β bind these factors, as though they were going to transport them into the nucleus. This blocks their spindle assembly activity. During mitosis, chromosomes bind high levels of RCC1, the guanine exchange factor for Ran (Ran-GEF in Fig. 9.18). This RCC1 creates a gradient of the active guanosine triphosphatase (GTPase) Ran-GTP (guanosine triphosphate) around the chromosomes. During interphase, Ran-GTP is confined to the nucleus, where it releases importins from their cargo. In mitosis the chromosome-associated cytoplasmic Ran-GTP gradient locally liberates the spindle assembly factors including motor proteins and NuMA from sequestration by importins. They then stabilize nearby microtubules and organize them into a bipolar spindle. If centrosomes are removed or destroyed experimentally in cells about to enter mitosis, somatic cells can also use motor proteins to organize microtubules into bipolar spindles that lack asters but are otherwise remarkably normal. Most treated cells manage to complete mitosis successfully but normal mammalian cells then either arrest in the next cell cycle prior to replicating their DNA or commit suicide. Both outcomes depend on the presence of the important tumor suppressor protein p53 (see Fig. 43.11). Thus, centrosomes are not required to form spindles, but they contribute to cell-cycle progression in many cells. This dependence on centrosomes is not universal; Drosophila, for example, can live without centrosomes.
Chromosome Attachment to the Spindle Dynamic microtubules of prometaphase asters scan the cytoplasm effectively “searching” for binding sites that will capture and stabilize their distal plus ends. Captured microtubules are approximately fivefold less likely to depolymerize catastrophically than free microtubules. When catastrophes do occur, the microtubules depolymerize back to the pole, recycling tubulin subunits for incorporation into other, growing microtubules.
Unstable microtubule g TP
radient
Microtubule stabilized by Ran-GTP gradient
Ra n-
G
Stabilized microtubules accumulate on chromosomes
Motors sort microtubules, which bind to kinetochores
Other motors and (–) end-binding proteins organize spindle poles (note absence of asters)
FIGURE 44.8 ASSEMBLY OF A BIPOLAR SPINDLE IN THE ABSENCE OF CENTROSOMES. A gradient of Ran-guanosine triphosphate (GTP) stabilizes microtubules around chromosomes, which contain high concentrations of bound Ran GEF RCC1. This releases spindle assembly factors from importin α and β. Microtubules that accumulate around the chromosomes are sorted, organized, and focused to make poles by motors and (−) end-binding proteins such as NuMA. These spindle poles lack prominent astral microtubules.
762
SECTION X n Cell Cycle
Breakdown of the nuclear envelope makes the condensed chromosomes accessible to the microtubules. Chance encounters allow kinetochores to capture microtubule plus ends. Capture probably involves the nine-component KMN network, which includes the rodshaped Ndc80 complex (see Fig. 8.21) that binds along the sides of microtubules near their plus ends. Another member of the complex, the scaffolding protein Knl1 (its name in vertebrates—the “K” of KMN; see later), anchors Ncd80 in the kinetochore. Historically, it was thought that forces generated by bipolar attachment of the kinetochores of sister chromatids center chromosomes midway between the two spindle poles. This hypothesis was based on the observation that when a kinetochore first attaches to a microtubule, the chromosome moves along the side of that microtubule toward the spindle pole (Fig. 44.9). Subsequent capture of a microtubule emanating from the opposite spindle pole by the sister kinetochore would provide a counterforce pulling the chromosome in the opposite direction. Chromokinesin family motor proteins distributed along the chromosome arms were also thought to contribute to the gradual movement of the
A
chromosome toward the middle of the spindle. These movements are accompanied by coordinated shrinkage of the microtubules at the leading kinetochore and growth of microtubules at the trailing kinetochore. More recent studies revealed that chromosomes attached to only one spindle pole can move away from that pole if the unattached kinetochore associates with the kinetochore fiber of a chromosome already aligned at the spindle equator. In this case, the kinetochore of the mono-oriented chromosome glides toward the equator, where it is more likely to capture microtubules emanating from the opposite pole. This motion of one chromosome along the kinetochore fiber of another chromosome requires the kinesin-7 motor centromere protein E (CENP-E) (see Fig. 36.13) associated with the kinetochore of the moving chromosome. Recognition of a tubulin posttranslational modification leads CENP-E to move the chromosome toward the spindle equator, rather than out into the aster. The attachment of microtubules to kinetochores can be reconstituted in vitro from mixtures of chromosomes, isolated centrosomes, and tubulin subunits. The plus ends of microtubules grow out from centrosomes and
E Spindle pole
Spindle pole Microtubule
B
Distance traveled toward spindle pole (µm)
4
Kinetochore
3 2 1 0
Start of movement
–1 –2 0
20
40
60
80
Time (seconds)
C
D Microtubule
10 µm
Kinetochore
FIGURE 44.9 INITIAL CHROMOSOMAL MOVEMENTS DURING PROMETAPHASE. A–B, Capture of a microtubule by the kinetochore results first in movement along the side of the microtubule toward the pole from which that microtubule originated. These images come from a study in which living cells, observed by differential interference microscopy, were subjected to rapid chemical fixation just after a chromosome had attached to the spindle (arrow). C, Attachment of the chromosome to the spindle was confirmed by indirect immunofluorescence staining for tubulin and thin-section electron microscopy (D). E, The graph shows the movements of the chromosome before and after attachment. (From Rieder CL, Alexander SP, Rupp G. Kinetochores are transported poleward along a single astral microtubule during chromosome attachment to the spindle in newt lung cells. J Cell Biol. 1990;110:81–95, copyright The Rockefeller University Press.)
CHAPTER 44 n Mitosis and Cytokinesis
attach to the chromosomes. Surprisingly, chromosomebound microtubules can either lengthen or shorten at the attached end without detaching from the chromosome. Similar experiments with kinetochores isolated from budding yeast cells showed that kinetochores can remain attached to a shortening microtubule plus end even against an applied force of 9 pN (piconewtons). Physiological levels of tension actually stabilize the attachments of kinetochores to microtubules in vitro, as in vivo. This tethering of kinetochores to disassembling microtubules is essential for chromosome movements during mitosis.
Correcting Errors in Chromosome Attachment to the Spindle The goal of mitosis is to partition the replicated chromosomes accurately between two daughter cells. Therefore, all chromosomes must attach correctly to both spindle poles (known as amphitelic attachment; Fig. 44.5) before being segregated. Three other sorts of attachment are seen: (a) chromosomes with one kinetochore lacking attached microtubules (known as monotelic attachment; this is a normal intermediate), (b) chromosomes with both sister kinetochores attached to the same spindle pole (known as syntelic attachment), and (c) chromosomes with a single kinetochore attached simultaneously to both spindle poles (known as merotelic attachment). Correcting monotelic and syntelic errors takes time, and the SAC (see “Finding Time to Fix A
Borealin
763
Chromosome Attachment Errors: The Spindle Assembly Checkpoint” below) delays mitotic progression to allow the correction process to occur. When syntelic attachments occur, one or both kinetochores must detach for the chromosome to achieve a bipolar orientation. Chromosome attachment to opposite spindle poles is more stable than attachment to a single pole, because the tension generated by bipolar attachment (where forces pull a chromosome simultaneously toward opposite spindle poles) preferentially stabilizes microtubule connections to both kinetochores. Merotelic attachments are more dangerous, as the kinetochore is under tension and the attachments are therefore stable. Merotelic attachments are the most common cause of chromosome segregation errors in cultured mammalian cells. Syntelic and merotelic chromosome attachments are corrected through the action of Aurora B protein kinase, which forms the chromosomal passenger complex (CPC), along with inner centromere protein (INCENP), survivin, and borealin (Fig. 44.10). The other subunits target Aurora B to its various sites of action during mitosis and regulate the kinase activity. The complex concentrates at inner centromeres (the heterochromatin beneath and between the two sister kinetochores) during prometaphase and metaphase. At anaphase onset, the CPC moves to the overlapping interpolar microtubules of the central spindle and to the cell cortex, where the cleavage furrow will form, ultimately winding up in the
B
C
DNA Survivin Microtubules
D
E. CPC
Aurora-B
F
INCENP Borealin Survivin
Histone H3 phosphorylation Kinetochore targets Central spindle targets Cleavage furrow targets FIGURE 44.10 CHROMOSOMAL PASSENGER COMPLEX (CPC) REGULATES MITOTIC EVENTS. The CPC (green) is present at centromeres in prometaphase and metaphase (B), but transfers to the spindle midzone at anaphase (C) and midbody at anaphase (D). E, Diagram of Aurora B protein kinase complexed with INCENP (inner centromere protein), survivin, and borealin with some key targets of the CPC. F, If CPC function is inhibited (in this case by RNA interference [RNAi] depletion of borealin), chromosome attachment errors are common and many chromosomes fail to segregate properly in anaphase. Distribution of DNA (blue), microtubules (red), and survivin–green fluorescent protein (GFP) (green) in human mitotic cells. Inset in A, Distribution of kinetochores (red), and borealin (green) in a prometaphase cell. (A–D, Micrographs by Sally Wheatley and William C. Earnshaw. F and Inset in A, Micrographs by Ana Carvalho, Reto Gassmann, and William C. Earnshaw. Insets in A and F, From Gassmann R, Carvalho A, Henzing AJ, et al. Borealin: a novel chromosomal passenger required for stability of the bipolar mitotic spindle. J Cell Biol. 2004;166:179–191, copyright The Rockefeller University Press. A–C, From Wheatley SP, McNeish IA. Survivin: a protein with dual roles in mitosis and apoptosis. Int Rev Cytol. 2005;247:35–88.)
764
SECTION X n Cell Cycle
intercellular bridge during cytokinesis. The CPC regulates mitotic events from prophase through cytokinesis. Along the way it contributes to the correction of chromosome attachment errors and to the operation of the checkpoint that delays the cell cycle in response to those errors. Aurora B corrects chromosome attachment errors by phosphorylating Ndc80 in the microtubule-binding KMN complex (see Fig. 8.21). Aurora B phosphorylation strongly inhibits Ndc80 binding to microtubules, causing the kinetochore to release attached microtubules. When a chromosome is correctly attached to both spindle poles, tension stretches the kinetochore away from the CPC buried in the chromatin beneath. This can stabilize chromosome–microtubule interactions by preventing the kinase from phosphorylating Ndc80.
Finding Time to Fix Chromosome Attachment Errors: The Spindle Assembly Checkpoint Segregation of replicated chromosomes into daughter cells is extremely accurate. For example, budding yeasts
lose a chromosome only once in 100,000 cell divisions. The frequency of chromosome loss may be 20-fold to 400-fold higher for human cells grown in culture. To achieve even this level of accuracy, most cells delay entry into anaphase until all chromosomes have achieved amphitelic attachment to the spindle. This delay is caused by the spindle assembly checkpoint (SAC), which senses the completion of microtubule binding to kinetochores at metaphase (Fig. 44.11). The spindle checkpoint differs from DNA damage checkpoints in that its default setting is “on” as cells enter mitosis. It is silenced only when every chromosome is properly attached to the spindle. The SAC involves the products of the mitotic arrest– defective (MAD) genes, the budding-uninhibited-bybenzimidazole (BUB) genes, the monopolar spindle (Mps1) kinase and the CPC. The MAD and BUB genes were identified in yeast genetic screens for cells that continued to divide (and die) when the spindle was disassembled by drugs. These genes are conserved from yeast to humans. SAC proteins accumulate at
A
D
C 23 minutes
Time
B Seat belt N
Time
Mad1 • Mad2 Destroy free kinetochore with blast of laser light
23 minutes
APC/C substrate APC/C Cdc20APC/C
Mad2 APC/C
Microtubule
Cdc20APC/C
Checkpoint proteins not on kinetochore Go!
Wa
APC/C MCC
Go
Kinetochore• Mad1 • Mad2
it
C
BubR1
Wait!
Cdc20MCC
MCC Mad2 Proteins not to scale relative to kinetochore
U Ub Ub Ub b
Substrate
FIGURE 44.11 SPINDLE CHECKPOINT. Signaling by unattached kinetochores stops the cell from entering anaphase until all chromosomes have made a proper bipolar spindle attachment. A, As long as there is a chromosome that is not properly attached to the spindle (beige cells), the cell does not enter anaphase. The cell enters anaphase approximately 20 minutes after chromosome attachment is complete (green cells). B, In a cell with a persistently maloriented chromosome, anaphase entry is delayed (beige cells). If the unattached kinetochore is destroyed with a high-powered laser, the cell enters anaphase about 20 minutes later. This proves that the unattached kinetochore sends an inhibitory signal. C, Overview of the spindle assembly checkpoint (see text for details). D, Structure of the Mad1/Mad2 complex.
CHAPTER 44 n Mitosis and Cytokinesis
kinetochores early in mitosis, when the checkpoint is “on” (ie, during prophase or prometaphase), and are gradually displaced as microtubules bind and the kinetochores come under tension. The target of the SAC is the APC/CCdc20, the anaphasepromoting complex/cyclosome (APC/C) ubiquitinprotein ligase (an E3 enzyme; see Figs. 23.3 and 40.16) with its substrate recognition factor Cdc20 bound, APC/ CCdc20 ubiquitylates target proteins to mark them for destruction by proteasomes. Key APC/CCdc20 substrates are proteins that must be degraded for the cell to move from metaphase to anaphase, including cyclin B and securin, an inhibitor of the enzyme that triggers separation of sister chromatids at anaphase (Fig. 44.16). During mitosis, Cdk1 activates the APC/C by phosphorylating an auto-inhibitory loop, allowing Cdc20 to bind. The SAC is an additional regulatory circuit that inactivates APC/CCdc20 until all kinetochores attach to spindle microtubules. Kinetochores without microtubules attract proteins that assemble the mitotic checkpoint complex (MCC), the inhibitor that inactivates APC/CCdc20. Checkpoint activation starts when Aurora B in the CPC activates Mps1 kinase, allowing it to phosphorylate Knl1 in the kinetochore at several sites. Mps1 phosphorylation of Knl1 creates a binding site that results in Mad1 recruitment to the kinetochore. Mad1 then recruits Mad2 to form a stable complex (Fig. 44.11). A loop on Mad2 wraps around Mad1 like a safety belt making the complex particularly stable. This form of Mad 2 is known as “closed” Mad2. Mad1/Mad2 can transiently bind soluble Mad2 molecules (known as “open” Mad2), load them onto Cdc20 in the closed safety belt conformation (this loading probably occurs at kinetochores), then release them to form the soluble MCC of Mad2/Cdc20/ BubR1/Bub3. The MCC associates with the APC/CCdc20, interfering with binding of cyclin B and other key substrates. As each chromosome becomes attached to both poles of the spindle it stops producing MCC. When the last chromosome has achieved a proper attachment, the last source of MCC is extinguished, and entry into anaphase can proceed.
A. Metaphase
765
Silencing of the checkpoint involves several pathways. Protein phosphatase 2A (PP2A) and protein phosphatase 1 (PP1) are recruited to the kinetochore in feedback loops involving the CPC and other checkpoint components. They dephosphorylate Knl1 so that it releases Mad1. When microtubules bind, cytoplasmic dynein motors actively strip checkpoint components from the kinetochore, dragging them away toward the centrosomes. In yeast, access of Mps1 to its target sites on Knl1 is physically blocked when microtubules bind. Exactly how this interaction is regulated in metazoans is still actively studied. In addition, the SAC appears to crosstalk with the DNA damage response (see Chapter 43), since DNA damage response components activate SAC components and vice versa. However, the network of interactions is very complex and details are still being worked out. Experimental inactivation of the spindle checkpoint causes a catastrophic, premature entry into anaphase, regardless of the status of chromosome alignment. This leads to an unequal distribution of sister chromatids and genetic imbalance between daughter cells known as aneuploidy. Yeasts can live without the checkpoint genes, but their loss is lethal for mice, which die early during embryogenesis. Mice heterozygous for various checkpoint components show increased aneuploidy. Humans with mutations in BubR1 have mosaic variegated aneuploidy syndrome (extra copies or loss of various chromosomes in a variety of tissues), which is associated with microcephaly (decreased brain size) and an increased cancer risk.
Metaphase When all the chromosomes have attained amphitelic orientations and moved to positions roughly midway between the two spindle poles, the cell is said to be in metaphase (Fig. 44.12). The compact grouping of chromosomes at the middle of the spindle is referred to as the metaphase plate. In many cells, even though chromosomes remain, on average, balanced at the middle of
Kinetochore microtubules
B
Astral microtubules Interpolar microtubules Cyclin B and securin degraded
Chromosomes oscillate
DNA Microtubules Centrosomes
FIGURE 44.12 INTRODUCTION TO METAPHASE. A, Summary of the major events of metaphase. B, Distribution of DNA (blue), microtubules (red), and gamma tubulin (centrosomes [green]) in a metaphase human cell. (B, Images were recorded by Dr. Melpomeni Platani on the University of Dundee’s School of Life Sciences Imaging Facility OMX 3DSIM Microscope and stored and processed in OMERO.)
766
SECTION X n Cell Cycle
A
E
B
C
D
Spindle pole
F P
AP
2 µm
G
AP
P
2 min Spindle pole
FIGURE 44.13 KINETOCHORE OSCILLATIONS BETWEEN P (POLEWARD) AND AP (AWAY FROM THE POLE) MOVEMENT DURING LATE PROMETAPHASE AND ANAPHASE IN PTK1 (RAT KANGAROO) CELLS. A–D, Images showing the movements of several pairs of sister kinetochores, labeled with green fluorescent protein (GFP)-Cdc20 (green), combined with phase-contrast images of the cell (red). E and G, Highermagnification views of sister kinetochores (marked with dashed lines) in prometaphase and anaphase, respectively. F, Kymograph (collage of images of a vertical strip showing the same two kinetochores at various time points during the movie) showing the movements of these two kinetochores. Movements toward (P) and away from (AP) spindle poles are indicated. P movement involves microtubule shrinkage at the leading kinetochore and microtubule growth at the trailing kinetochore (which is undergoing AP movement away from its associated kinetochore). Spindle poles are near the top and bottom of panels E to G. (Micrographs courtesy E.D. Salmon, University of North Carolina, Chapel Hill.)
the spindle, they jostle one another and undergo numerous small excursions toward one pole or the other throughout metaphase (Fig. 44.13). Metaphase can also be defined biochemically as the time of destruction of cyclin B and securin (Fig. 44.16), because this begins as soon as the last chromosome achieves amphitelic orientation. Degradation of cyclin A begins earlier, at the entry into prometaphase, and is largely complete before metaphase. Loss of securin initiates a process leading to the separation of sister chromatids and the onset of anaphase.
Microtubule Flux Within the Metaphase Spindle Although the average length of the kinetochore microtubules is roughly constant during metaphase, the microtubules change continuously in three ways. First, there is constant net addition of new tubulin subunits (approximately 10 subunits per second) to the plus end of the microtubules, where they are attached to the kinetochore. Second, a comparable number of tubulin subunits is continuously lost from the minus end of the kinetochore tubules at the spindle poles. There fore, tubulin subunits slowly migrate through kinetochore microtubules from the kinetochore to the pole (Fig. 44.14). This subunit flux or treadmilling in kinetochore microtubules is caused by microtubule depolymerization at the poles driven by kinesin-13 family members. In addition, tubulin moves toward the poles as by microtubules are transported towards the pole (many nucleated by the augmin complex) within the kinetochore fiber. All microtubules attached to each kinetochore
A
B
C
0 sec
Fluorescent tubulin speckles move toward poles
10 sec
20 sec P P
P
P
P
P
30 sec
FIGURE 44.14 MICROTUBULE FLUX IN METAPHASE. A, Cells entering mitosis were injected with tubulin subunits modified chemically by attachment of a caged fluorescent dye. This dye becomes fluorescent after being irradiated with UV light. When cells entered metaphase, the spindle was illuminated with a narrow stripe of UV light, activating a narrow band of fluorescent tubulin subunits. With time, these subunits approach the spindle poles (P). Because the length of kinetochore microtubules is constant during this time, the labeled tubulin molecules must migrate along the microtubules toward the pole (arrows). This can occur if new subunits are added to the microtubule at the kinetochore and old subunits are removed at the pole. B, Microtubule flux at metaphase in a Drosophila embryo visualized by fluorescence speckle microscopy. Embryos were injected with very low levels of fluorescent tubulin, which appears as speckles distributed along the microtubules. If a very sensitive camera is used, these speckles can be seen to move toward the poles, reflecting the flux in the underlying microtubules. Scale bar is 5 µm. C, Movement of labeled tubulin speckles toward the spindle poles. (A, Courtesy Arshad Desai and the MBL Cell Division Group, Marine Biology Laboratory, Woods Hole, MA; from Mitchison TJ, Salmon ED. Mitosis: a history of division. Nat Cell Biol. 2001;3:E17–E21. B, Courtesy Paul Maddox and Arshad Desai, University of California, San Diego.)
CHAPTER 44 n Mitosis and Cytokinesis
change coordinately in length during chromosomal oscillations.
Anaphase The separation of sister chromatids at anaphase is one of the most dramatic events of the entire cell cycle (Fig. 44.15). Sister chromatids move to opposite spindle poles (anaphase A), and the poles move apart (anaphase B). Anaphase is also the time when the mitotic spindle activates the cell cortex in preparation for cytokinesis. Two forms of the APC/C (see Fig. 40.15) trigger the transition from metaphase to anaphase by degrading key proteins. APC/CCdc20 targets cyclin B for degradation, causing Cdk activity to fall (see Fig. 40.17). This decline in Cdk activity allows for activation of APC/CCdh1, because Cdh1 phosphorylated by Cdk1 kinase cannot bind to the APC/C. APC/CCdh1 targets polypeptides whose destruction by the proteasome is required for the cell to exit from mitosis and return to interphase. APC/CCdh1 remains active during G1 phase, where it is essential for the licensing of DNA replication origins (see Fig. 42.28).
Biochemical Mechanism of Sister Chromatid Separation Separation of sister chromatids is regulated by the chromosomes themselves, not by the mitotic spindle. Under certain circumstances, sister chromatids can separate in the absence of microtubules, ruling out a requirement for forces from the spindle in the process. Three factors regulate sister chromatid separation: a protein complex known as cohesin, a protease known as separase, and an inhibitor of separase known as securin (Fig. 44.16). This system is conserved from yeast to human. Chapter 8 discusses the functions of cohesin in interphase. Cohesin is a complex of four proteins that resembles the condensin complex (see Fig. 8.18). Like condensin, cohesin has two large subunits from the SMC ATPase family. These proteins, SMC1 and SMC3, are complexed with proteins called Scc1 (which has other names A. Anaphase Cohesin degrades
767
omitted here for simplicity) and Scc3. Additional proteins are required to stabilize the loading of this complex onto DNA. Cells with mutations in cohesin components separate sister chromatids prematurely in mitosis, resulting in chaotic chromosome missegregation. This system is very ancient, as bacteria depend on an SMC-related protein for orderly chromosome segregation. A variety of evidence suggests that cohesin forms a ring with a diameter of 35 nm, large enough to encircle two sister chromatids like a lasso. In yeast, the complex functions only if it binds chromosomes during DNA replication. Cohesin accumulates at preferred sites on the chromosomes, often near centromeres in budding yeast or in regions of heterochromatin in fission yeast. In vertebrates, most cohesin dissociates from the chromosome arms by late metaphase, owing to the action of the protein kinases Plk1 and Aurora B. Importantly, a critical fraction remains associated with heterochromatin flanking centromeres where it is protected from cleavage by shugoshin until the onset of anaphase (see following paragraphs and Chapter 45). Sequential cleavage of two key proteins triggers sister chromatid separation at anaphase. This proteolysis makes anaphase onset an irreversible transition. The first target, securin, inhibits the separase protease. After the last chromosome forms an amphitelic attachment to the spindle, the spindle checkpoint is silenced. This allows APC/CCdc20 to tag securin with ubiquitin, leading to its destruction by proteasomes throughout metaphase. When securin levels fall below a critical threshold, separase is unleashed to cleave the Scc1 subunit of cohesin. Cleavage of Scc1 breaks the cohesin ring, allowing the sister chromatids to separate triggering the onset of anaphase (Fig. 44.16B). Efficient Scc1 cleavage requires that the protein be phosphorylated near its cleavage site. This allows a mode of regulation where shugoshin (Japanese for “guardian spirit”) recruits PP2A to centromeres. PP2A keeps Scc1 dephosphorylated. This inhibits its cleavage and protects cohesin until shugoshin is released following amphitelic attachment of the chromosome. This
Sister chromatids separate
B
Anaphase A: Chromatids approach poles (APC/CCdc20 active) Interdigitated interpolar microtubules bundled by PRC1 and stem body material to form central spindle Anaphase B: Spindle poles migrate apart (APC/C Cdh1active)
DNA Microtubules Centrosomes
FIGURE 44.15 INTRODUCTION TO ANAPHASE. A, Summary of the major events of anaphase. B, Distribution of DNA (blue), microtubules (red), and gamma tubulin (centrosomes [green]) in a mid-anaphase human cell. APC/C, anaphase-promoting complex/cyclosome; PRC1, protein regulated in cytokinesis 1. (B, Images were recorded by Dr. Melpomeni Platani on the University of Dundee’s School of Life Sciences Imaging Facility OMX 3DSIM Microscope and stored and processed in OMERO.)
768
SECTION X n Cell Cycle
Hinge
A. S phase
Replication fork
Scc1
Smc1
Smc3
Cohesin complex
SA1
C
B
D
E
Separase Securin
Chromatin loops
APC/C Sister chromatids separate
Ubiquitin Cohesin Proteasome Sister chromatids
Mitotic chromosome S phase
Early mitosis
Active separase cleaves Scc1 Metaphase/ anaphase transition
Anaphase
FIGURE 44.16 REGULATION OF SISTER CHROMATID PAIRING BY THE COHESIN COMPLEX. A–B, The cohesin complex forms a 35-nm diameter ring that links sister chromatids during DNA replication. At anaphase onset, degradation of its securin inhibitor liberates active separase enzyme. C–E, Separase then cleaves cohesin subunit Scc1, and the two sister chromatids are freed to separate from one another and move toward opposite spindle poles.
mechanism is absolutely essential during meiosis, as without it, it would not be able to segregate homologous chromosomes from each other (see Fig. 45.12). Securin can act as an oncogene in cultured cells and is overexpressed in some human pituitary tumors. Overexpression of securin may disrupt the timing of chromosome segregation, leading to chromosome loss and, ultimately, contributing to cancer progression.
Mitotic Spindle Dynamics and Chromosome Movement During Anaphase Anaphase is dominated by the orderly movement of sister chromatids to opposite spindle poles brought about by the combined action of motor proteins and changes in microtubule length. There are two components to anaphase chromosome movements (Fig. 44.15). Anaphase A, the movement of the sister chromatids to the spindle poles, requires a shortening of the kinetochore fibers. During anaphase B, the spindle elongates, pushing the spindle poles apart. The poles separate partially because of interactions between the antiparallel interpolar microtubules of the central spindle and partially because of intrinsic motility of the asters. Most cells use both components of anaphase, but one component may predominate in relation to the other.
Microtubule disassembly on its own can move chromosomes (see Fig. 37.8). Energy for this movement comes from hydrolysis of GTP bound to assembled tubulin, which is stored in the conformation of the lattice of tubulin subunits. Microtubule protofilaments are straight when growing, but after GTP hydrolysis protofilaments are curved, so they peel back from the ends of shrinking microtubules (see Fig. 34.6). Several kinesin “motors” influence the dynamic instability of the spindle microtubules. Members of the kinesin-13 class, which encircle microtubules near kinetochores and at spindle poles, use adenosine triphosphate (ATP) hydrolysis to remove tubulin dimers and promote microtubule disassembly rather than movement. Kinetochores are remarkable in their ability to hold onto disassembling microtubules. In straight (growing) microtubules, the Ndc80 complex is mostly responsible for microtubule binding. It binds to the interface between α and β tubulin subunits. This interface bends in curved (shrinking) microtubules, so Ndc80 cannot bind. This could allow it to redistribute onto straight sections of the lattice and thereby move away from the curved protofilaments at the disassembling end. In metazoans the Ska complex in the outer kinetochore binds α and β tubulin subunits away from the interface, so it can bind
CHAPTER 44 n Mitosis and Cytokinesis
to curved (disassembling) protofilaments. At yeast kinetochores the Dam1 ring (green in Fig. 8.21) couples the kinetochore to disassembling microtubules. Anaphase A chromosome movement involves a combination of microtubule shortening and translocation of the microtubule lattice that result from flux of tubulin subunits (Fig. 44.14). The contributions of the two mechanisms vary among different cell types. When living vertebrate cells are injected with fluorescently labeled tubulin subunits, the spindle becomes fluorescent (Fig. 44.17). If a laser is used to bleach a narrow zone in the fluorescent tubulin across the spindle between the chromosomes and the pole early in anaphase, the chromosomes approach the bleached zone much faster than the bleached zone approaches the spindle pole. This shows that the chromosomes “eat” their way along the
A
Photobleached zone Time Microtubule disassembly
B
42
70
212
422
92
210
436
5 µm
C
40
FIGURE 44.17 CHROMOSOMES MOVE ON SHRINKING MICROTUBULES DURING ANAPHASE. A, Mitotic cells were injected with a fluorescently labeled tubulin that was incorporated into the spindle. Just after anaphase onset, a laser was used to photobleach a stripe (white) across the spindle near the upper pole. The live cell was monitored over time by fluorescence (B) and phase-contrast (C) microscopy. In this mammalian cell, the chromosomes approach the bleached stripe much faster than the stripe approaches the spindle pole. In other organisms with higher rates of microtubule flux in their spindles, the bleached zone would also move appreciably toward the pole. The numbers are time in seconds. (B–C, From Gorbsky GJ, Sammak PJ, Borisy GG. Microtubule dynamics and chromosome motion visualized in living anaphase cells. J Cell Biol. 1988;106:1185– 1192, copyright The Rockefeller University Press.)
769
kinetochore microtubules toward the pole. In these cells, subunit flux accounts for only 20% to 30% of chromosome movement during anaphase A, and this flux is dispensable for chromosome movement. In Drosophila embryos, in which subunit flux accounts for approximately 90% of anaphase A chromosome movement, the chromosomes catch up with a marked region of the kinetochore fiber slowly, if at all. Anaphase B appears to be triggered at least in part by the inactivation of the minus-end–directed kinesin-14 motors, so that all the net motor force favors spindle elongation. Four factors contribute to overall lengthening of the spindle: release of sister chromatid cohesion, sliding apart of the interdigitated half-spindles, microtubule growth, and intrinsic motility of the poles themselves (Fig. 44.7). During the latter stages of anaphase B, the spindle poles, with their attached kinetochore microtubules, appear to move away from the interpolar microtubules as the spindle lengthens. This movement of the poles involves interaction of the astral microtubules with cytoplasmic dynein molecules anchored at the cell cortex. Anaphase B spindle elongation is accompanied by reorganization of the interpolar microtubules into a highly organized central spindle between the separating chromatids (Fig. 44.15). Within the central spindle, an amorphous dense material called stem body matrix stabilizes bundles of antiparallel microtubules and holds together the two interdigitated half-spindles. Proteins concentrated in the central spindle help regulate cytokinesis. One key factor, PRC1 (protein regulated in cytokinesis 1), is inactive when phosphorylated by Cdk kinase and functions only during anaphase when Cdk activity declines and phosphatases remove the phosphate groups placed on target proteins by Cdks and other mitotic kinases. PRC1 directs the binding of several kinesins to the central spindle. The kinesin KIF4A targets Aurora B kinase to a particular domain of the central spindle, where phosphorylation of key substrates then regulates spindle elongation and cytokinesis. How can protein kinases such as Aurora B continue to function during anaphase while protein phosphatases are removing phosphate groups placed there by Cdks and, indeed, Aurora B during early mitosis? One answer is that the phosphatase activity is highly localized, controlled by specific targeting subunits. Cdk phosphorylation can inhibit targeting subunits such as the exotically named Repo-Man (recruits PP1 onto mitotic chromatin at anaphase) from binding protein phosphatase 1 or localizing to targets, such as chromatin in early mitosis. When Cdk activity drops, Repo-Man (and other similar targeting subunits) is dephosphorylated, and now targets PP1 to chromatin, where it removes phosphates placed there by Aurora B in the CPC. As long as phosphatases are not specifically targeted to the cleavage furrow, Aurora B can continue to control events there during
770
SECTION X n Cell Cycle
mitotic exit by phosphorylating key target proteins required for cytokinesis.
Telophase During telophase, the nuclear envelope reforms on the surface of the separated sister chromatids, which typically cluster in a dense mass near the spindle poles (Fig. 44.18). Some further anaphase B movement may still occur, but the most dramatic change in cellular structure at this time is the constriction of the cleavage furrow and subsequent cytokinesis.
Reassembly of the Nuclear Envelope Nuclear envelope reassembly begins during anaphase and is completed during telophase (Fig. 44.19). As in spindle assembly, Ran-GTP promotes early steps of nuclear envelope assembly at the surface of the chromosomes by releasing key components sequestered by importin β. These include several nuclear pore com ponents, and one of the earliest events in nuclear envelope reassembly involves binding of the nuclear pore
A. Telophase
scaffold protein ELYS to chromatin. ELYS can recognize DNA regions rich in A : T base pairs, so it is likely to bind directly to the DNA. ELYS then recruits other components of the nuclear pore scaffold and nuclear pore trans-membrane proteins. The pore subsequently matures as various peripheral components and elements of the permeability barrier are added. The mechanism of nuclear membrane reassembly is debated. In cells where nuclear membranes fragments into vesicles during mitosis, a Ran-GTP–dependent pathway directs at least two discrete populations of vesicles to chromatin where they fuse to reform the nuclear envelope. In cells where the nuclear membrane is absorbed into the endoplasmic reticulum during mitosis, reassembly involves lateral movements of membrane components within the membrane network and their stabilization at preferred binding sites at the periphery of the chromosomes. Lamin subunits disassembled in prophase are recycled to reassemble at the end of mitosis. Lamina reassembly is triggered by removal of mitosis-specific phosphate groups and methyl-esterification of several COOH side
Cleavage plane specified
Nuclear envelope reassembles around chromosomes
B
Organized central spindle assembles
DNA Microtubules Centrosomes
Poles continue to separate
FIGURE 44.18 INTRODUCTION TO TELOPHASE. A, Summary of the major events of telophase. B, Distribution of DNA (blue), microtubules (red), and γ-tubulin (centrosomes [green]) in a telophase human cell. (B, Images were recorded by Dr. Melpomeni Platani on the University of Dundee’s School of Life Sciences Imaging Facility OMX 3DSIM Microscope and stored and processed in OMERO.)
B
A
0 min
C
8–10 min
333 nm
≥ 25 min
FIGURE 44.19 SCANNING ELECTRON MICROSCOPY OF THE STAGES OF ASSEMBLY OF MEMBRANE VESICLES ON THE SURFACE OF CHROMOSOMES IN A XENOPUS EGG CYTOSOLIC EXTRACT. A cell lysate containing membrane vesicles was added to isolated chromatin from Xenopus sperm, fixed, and then imaged by scanning electron microscopy. Each panel shows the time of incubation prior to fixation. (Micrographs courtesy of K.L. Wilson, Johns Hopkins Medical School, Baltimore, MD. A and C, From Wiese C, Goldberg MW, Allen TD, et al. Nuclear envelope assembly in Xenopus extracts visualized by scanning EM reveals a transport-dependent “envelope smoothing” event. J Cell Sci. 1997;110:1489–1502.)
CHAPTER 44 n Mitosis and Cytokinesis
chains on lamin B (Fig. 44.6). Together with ELYS, B-type lamins are among the earliest components of the nuclear envelope to target to the surface of the chromosomes during mid-anaphase. Either at this time or shortly thereafter, other proteins associated with the inner nuclear membrane, including BAF, LAP2, and lamin B receptor (see Fig. 9.10), join the forming envelope. Later during telophase when nuclear import is reestablished, lamin A enters the reforming nucleus and slowly assembles into the peripheral lamina over several hours in the G1 phase. If lamin transport through nuclear pores is prevented, chromosomes remain highly condensed following cytokinesis, and the cells fail to reenter the next S phase.
A. Early cytokinesis New membrane inserted Actomycin
Actomycin contractile ring forms Midbody begins to form
771
Cytokinesis Cytokinesis divides a mitotic cell into two daughter cells (Fig. 44.20). Cytokinesis depends on signals to specify the cleavage plane (Fig. 44.21), assembly and constriction of the contractile apparatus, specific alterations of the cell membrane, and the final separation (abscission) of the two daughter cells. In animals, protozoa, and most fungi, a contractile ring of actin filaments and myosin-II guides the separation of daughter cells at the end of mitosis (Fig. 44.2). Myosin-II pulls on the ring of actin filaments, applying tension to the plasma membrane, much like contraction of smooth muscle (see Figs. 39.23 and 39.24). Because
B. Late cytokinesis
Chromatin decondenses Nuclear substructures reform Interphase microtubule array reassembles Midbody
C
ESCRT III action leads to separation (abscission) of the two cells FIGURE 44.20 INTRODUCTION TO CYTOKINESIS. A–B, Summary of the major events of cytokinesis. C, Distribution of DNA (blue), microtubules (red), and γ-tubulin (centrosomes [green]) in a human cell undergoing cytokinesis. ESCRT, endosomal sorting complexes required for transport. (C, Images were recorded by Dr. Melpomeni Platani on the University of Dundee’s School of Life Sciences Imaging Facility OMX 3DSIM Microscope and stored and processed in OMERO.)
A. Evidence that the cleavage furrow is positioned midway between asters in eggs
B. An organized central spindle is required for cleavage furrow formation and/or function
TOP VIEW Glass rod pushed down into egg 90°
Sand dollar egg
Microtubules Chromosomes Actin ring
Ectopic furrow
Metaphase 1
Cytokinesis 1
Metaphase 2
Cytokinesis 2
Wild type
Profilin mutant
FIGURE 44.21 IN EGGS, A CLEAVAGE FURROW FORMS MIDWAY BETWEEN SPINDLE ASTERS. IN ANIMAL CELLS, THE CENTRAL SPINDLE IS IMPORTANT. A, A classic experiment in which a sand-dollar egg is caused to adopt a toroid shape. At cytokinesis 2, the egg cleaves into four cells, and a furrow forms between the back sides of the two spindles. (For a description of this and other classic experiments in cytokinesis, see the book by Rappaport in the “Selected Readings” list.) B, Left, A wild-type Drosophila spermatocyte undergoing cytokinesis, with the contractile ring stained in yellow. Right, In a profilin mutant, no central spindle forms, and the cell fails to form a contractile ring. (Micrographs courtesy Professor Maurizio Gatti, University of Rome, Italy. B, From Giansanti MG, Bonaccorsi S, Williams B, et al. Cooperative interactions between the central spindle and the contractile ring during Drosophila cytokinesis. Genes Dev. 1998;12:396–410.)
772
SECTION X n Cell Cycle
the contractile ring is confined to a narrow band of cortex around the equator, it forms a cleavage furrow, constricting the plasma membrane locally like a purse string (Fig. 44.20). Signals from the mitotic spindle and cell cycle machinery control the position of this ring (ie, the relative sizes of the two daughter cells) and the timing of its constriction. Protozoa, animals, fungi, and plants use an evolutionarily conserved set of components to implement different strategies to separate daughter cells. For example, both fission yeast and metazoan cells use signals from polo kinase and a Rho-GTPase to direct the assembly of a contractile ring of actin, myosin-II, and other conserved components, even though the yeast has a closed mitosis and the metazoans have an open mitosis. In animal cells, contractile ring constriction provides the force that remodels the cortex to generate the two daughter cells. In contrast, in yeasts, which have a cell wall, contractile ring constriction is thought to guide the orderly centripetal growth of the cell wall septum, which contributes force to overcome turgor pressure and invaginate the plasma membrane. Plants lack myosin-II, so they divide by targeted fusion of membrane vesicles to build a new cell wall rather than constricting a cleavage furrow (Box 44.1 and Fig. 44.26). These differences reflect the fact that widely divergent eukaryotes use variations of similar themes for cytokinesis. Cytokinesis in prokaryotes is genuinely different, since completely different proteins are involved (Box 44.2 and Fig. 44.27). Although cytokinesis has been studied for more than 100 years, it has posed a number of challenges due to its complexity at the molecular level. For example genetic analysis of fission yeast revealed more than 150 genes that contribute to cytokinesis. RNAi-based protein knockdown and molecular replacement analy sis indicates that similar proteins participate in cytokinesis of Caenorhabditis elegans, Drosophila, and vertebrate tissue culture cells. Cytokinesis research typically employs living cells, although progress is being made toward reconstituting some aspects of the process in cell-free systems.
Signals Regulating the Position of the Cleavage Furrow Elegant experimental data from classic studies on fertilized echinoderm eggs suggest that a cleavage stimulus, emitted by the mitotic spindle, specifies the position of the cleavage furrow midway between the poles and perpendicular to the long axis of the spindle, thereby ensuring that the cleavage process separates the daughter nuclei (Fig. 44.21). In fertilized eggs, the poles, with their large astral arrays of microtubules (see Fig. 6.4B), were regarded as the source of the cleavage stimulus, as furrows can be induced to form midway between two poles, even when no chromosomes are present. In addition, a signal emitted by the bundled microtubules of the
central spindle appeared to modulate the behavior of the furrow signaled by the poles. We now know that the central spindle does emit a positive signal directing a cleavage furrow to form above it, while the poles contribute by focusing that furrow at a point on the cortex midway between them. The molecular nature of the cleavage stimulus is now beginning to be understood in animals. The following is a simplified scenario: 1. During anaphase, overlapping microtubules between the separating chromatids establish an ordered array known as the central spindle. A key protein component of this array is a protein heterodimer known as centralspindlin. Centralspindlin is normally sequestered in the cytoplasm, but phosphorylation by the CPC enables it to target to the central spindle. Drosophila mutants that fail to form a central spindle cannot initiate cytokinesis (Fig. 44.21B). In contrast, C. elegans embryos that lack a central spindle can initiate but not complete the process. 2. One of the components of centralspindlin recruits a GEF (guanine exchange factor; see Figs. 4.6 and 4.7) for the small GTPase RhoA. This Rho-GEF, Ect2, also has a motif for targeting to the inside surface of the equatorial plasma membrane. 3. Membrane associated Ect2 locally activates RhoA, which then stimulates localized actin filament assembly and activation of myosin-II to begin assembly of the contractile ring. Signals from the poles of the mitotic spindle contribute, particularly in large invertebrate embryos, by confining the zone of active RhoA to a narrow equatorial band between the separating sister chromatids.
Assembly and Regulation of the Contractile Ring Exposure of the cell cortex to the cleavage stimulus culminates in the assembly of a contractile ring consisting of a very thin (0.1 to 0.2 µm) array of actin filaments attached to the plasma membrane at many sites around the equator (Fig. 44.22). Polymerization of the actin filaments depends on formins (see Fig. 33.14). Small, bipolar filaments of myosin-II are interdigitated with actin filaments. The plasma membrane adjacent to this actinmyosin ring undergoes alterations in its lipid composition that may help recruit proteins important for the function of the contractile ring. Membrane furrowing requires actin and the motor activity of myosin-II (see Fig. 36.7). In animals, the small GTPase RhoA regulates actin polymerization by formins as well as constriction of the ring. Many other proteins are required for cytokinesis to go to completion. In their absence, furrowing begins, but the cleavage furrows ultimately regress, producing binucleated cells. These supporting proteins include anillin, actin filament crosslinking proteins, the CPC (Fig. 44.10) and the centralspindlin complex, among many others. Anillin helps
CHAPTER 44 n Mitosis and Cytokinesis
A. Early anaphase
B. Late anaphase
773
C
INCENP Myosin II
D Actin pointed ends
INCENP Myosin II
E. Myosin confocal
Actin barbed ends
F
Central optical section
H. Contractile mechanism
G
Actin
I. Equatorial section
Myosin II
J. Grazing saggital section
FIGURE 44.22 ORGANIZATION OF THE CONTRACTILE RING. A, Organization of actin at the cell cortex prior to cytokinesis. B, Distribution of actin and myosin at the start of ring contraction. C, INCENP (inner centromere protein) (red) concentrates at the site where the cleavage furrow will form just before myosin (green). D, INCENP and myosin concentrate in the contractile ring during contraction. E, Confocal micrograph shows the distribution of myosin in an optical cross-section contracting contractile ring. F–G, Dividing invertebrate egg with DNA (blue) and actin (red) in the contractile ring. H, Organization of actin and myosin filaments during cytokinesis. I–J, Electron micrographs showing actin filaments in the contractile ring. Note the thick filaments that are thought to be myosin-II filaments (red arrowheads) and the thinner actin filaments (yellow arrows). (C–D, Courtesy William C. Earnshaw. E, I, and J, Courtesy P. Maupin, Johns Hopkins Medical School, Baltimore, MD. F–G, Courtesy Professor Issei Mabuchi, University of Tokyo, Japan. For reference, see Maupin P, Pollard TD. Arrangement of actin filaments and myosin-like filaments in the contractile ring and actin-like filaments in the mitotic spindle of dividing HeLa cells. J Ultrastruct Res. 1986;94:92–103; Maupin P, Phillips CL, Adelstein RS, et al: Differential localization of myosin-II isozymes in human cultured cells and blood cells. J Cell Sci. 1994;107:3077–3090; and Eckley DM, Ainsztein AM, MacKay AM, et al. Chromosomal proteins and cytokinesis. J Cell Biol. 1997;136:1169–1183.)
keep active myosin-II focused into an organized contractile ring throughout cytokinesis. The CPC and the centralspindlin complex are both required for animal cells to assemble the central spindle. Consequently, if either of the two complexes is eliminated in C. elegans, a contractile ring fails to form. Thus, they appear to contribute to the cleavage stimulus, and indeed, both require microtubules to localize to the site of cleavage furrow formation as originally shown for the cleavage stimulus. In addition, the CPC regulates the timely completion of cytokinesis by blocking premature activation of the ESCRT (endosomal sorting complexes
required for transport) complex, which has a key role in the final separation of daughter cells (see later). In fission yeast, with closed mitosis, the nucleus determines the position of cleavage. Fission yeast assem ble a contractile ring along a well-defined pathway by recruiting proteins from cytoplasmic pools (Fig. 44.23). During interphase, assemblies of proteins called nodes form on the inside of the plasma membrane around the middle of cell. Prior to mitosis, an anillin-like protein leaves the nucleus and joins these nodes. During prophase, myosin-II, a formin and other contractile ring proteins join the nodes. When the formin polymerizes
774
SECTION X n Cell Cycle
Interphase actin patches
-60 Mid1p (anillin-like protein) exits nucleus
Anillin-like Cell wall Formin
Time (min)
-10 Nodes containing anillin, myosin-II and formin assemble around equator 0
SPBs separate
+5
Anaphase A
+10 Anaphase B elongates mitotic spindle
Myosin-II
Profilin Nodes condense into a contractile ring of actin filaments and myosin-II
Contractile ring matures by addition of actin binding proteins
+30 End Anaphase B
+40 Constriction begins
+70 Constriction ends
Signal from cell cycle via SIN pathway triggers constriction of contractile ring and deposition of cell wall material to form a septum
Plasma membrane fusion completes cytokinesis
FIGURE 44.23 CYTOKINESIS IN FISSION YEAST SCHIZOSACCHAROMYCES POMBE. During interphase, microtubules (red) position the nucleus in the middle of the cell. Actin filaments concentrate in small patches (yellow) in the cortex at the two growing ends of the cell (see Fig. 33.1). The mitotic spindle is inside the nucleus, as the nuclear membrane does not break down during mitosis. As the cell enters mitosis, an anillin-like protein moves from the nucleus to the equatorial cortex, where it sets up nodes of proteins, including myosin-II and a formin. The formin grows actin filaments (yellow), and myosin-II pulls the nodes together into a continuous contractile ring. At the end of anaphase a signaling system consisting of a GTPase and three protein kinases (the septation initiation network [SIN]) triggers constriction of the contractile ring and associated synthesis of new cell wall to form a septum. The septum is a three-layered structure, with the primary septum flanked by two secondary septae. Digestion of the primary septum separates the daughter cells. (For reference, see Wu J-Q, Kuhn JR, Kovar DR, et al. Spatial and temporal pathway for assembly and constriction of the contractile ring in fission yeast cytokinesis. Dev Cell. 2004;5:723–734.)
actin filaments, myosin-II pulls the nodes together into a ring around the equator of the cell (Fig. 44.23). Contractile ring assembly in animal cells shares many properties with fission yeast, but is less completely understood. The decline in the activity of cell cycle kinases at the onset of anaphase is part of the trigger, since they inhibit centralspindlin components through metaphase. INCENP and anillin move from the interphase nucleus to the cortex around the cell equator in early anaphase (Fig. 44.22). Formins and profilin polymerize some new actin filaments, but preexisting actin filaments are recruited into the contractile ring from adjacent areas of the cortex. Myosin-II is dispersed throughout the cytoplasm until anaphase, when it concentrates in the cortex, especially around the equator where the furrow forms. The myosin-II is derived from various interphase structures including stress fibers (see Fig. 33.1) that break down during prophase.
Constriction of the Cleavage Furrow Contractile rings of echinoderm eggs produce enough force to invaginate the plasma membrane and form the cleavage furrow, although many details are still being studied. Constriction of the ring probably involves a sliding filament mechanism similar to muscle (see Figs. 39.9 and 39.23), but little is known about how the contractile ring is attached to the plasma membrane. During the early stages of furrowing, contractile rings maintain a constant volume, but then disassemble as they constrict further. The role of myosin-II as the motor for cytokinesis was established by microinjection of inhibitory antibodies into echinoderm embryos and confirmed by genetic inactivation in the slime mold Dictyostelium. Slime mold amoebas lacking the myosin-II heavy chain round up during mitosis and complete nuclear division but cannot
CHAPTER 44 n Mitosis and Cytokinesis
form a normal cleavage furrow. Mutant cells accumulate many nuclei, because the mitotic cycle continues. Mutant cells can divide on a substratum using pseudopods to pull themselves apart into smaller cells. Constriction of the contractile ring is regulated so that it does not begin until after the onset of anaphase B, when sister chromatids are well separated. In fission yeast, a signaling pathway called the septation initiation network (SIN) initiates constriction. Much less is known in other cells.
Abscission As the contractile ring pulls the cell membrane inward, the single cell that entered mitosis is gradually transformed into two daughters joined by a thin intercellular bridge (Fig. 44.20). This process requires a significant net increase in the surface area of the cell. New plasma membrane is inserted adjacent to the leading edge of the furrow. The source of the new membrane appears to be recycling endosomes (see Chapter 22), so addition of membrane to the cleavage furrow is a specialized form of exocytosis. Fusion of vesicles providing the new membrane depends on specific syntaxins, t-SNAREs (soluble N-ethylmaleimide-sensitive factor attachment protein receptors) (see Fig. 21.15) that promote vesicle fusion along the secretory pathway. The plasma membrane in the cleavage furrow has a discrete composition. In budding yeast, this compartment is delineated by rings made from polymers of septins, a family of GTP-binding proteins recruited by anillin. Septins are essential for cytokinesis in Saccharomyces cerevisiae but not fission yeast. In most animal cells, constriction of the cleavage furrow ultimately reduces the cytoplasm to a thin intercellular bridge between the two daughter cells. The intercellular bridge contains a highly ordered, antiparallel array of microtubules derived from the spindle with a dense knob, the midbody, at its center (Fig. 44.20). Isolated midbodies contain more than 160 proteins, with approximately one-third involved in various aspects of membrane trafficking. The midbody is encircled by a dense ring of proteins that includes centralspindlin, anillin, and a centrosomal protein known as Cep55. Interactions between anillin and membrane-associated septin filaments tether the membrane to the ring. Cep55 recruits the ESCRT (endosomal sorting complexes required for transport) III complex, proteins with important roles in vesicle budding events such as formation of multivesicular bodies (see Chapters 22 and 23). In the intercellular bridge, ESCRT III forms a helical filament that spirals around the inner surface of the membrane, becoming more and more constricted as it grows away from the midbody (Fig. 44.24). ESCRT III also recruits factors that disassemble the bundled microtubules in the intercellular bridge, allowing the membrane to constrict further.
775
A
B. Abscission
Midbody ring (centralspindlin, anillin, cep55) ESCRT I and ESCRT II ESCRT III Midbody Vps4
Microtubules
ESCRT I
FIGURE 44.24 ABSCISSION IN ANIMAL CELLS. Proteins associated with overlapping bundles of microtubules form the midbody. The midbody ring links the midbody to the membrane, and nucleates the formation of ESCRT (endosomal sorting complexes required for transport) III filaments that constrict the membrane to divide the daughter cells.
FIGURE 44.25 INCOMPLETE CYTOKINESIS IN A DROSOPHILA EGG CHAMBER LEAVES CELLS JOINED BY RING CANALS. Colocalization of actin (red) and the ring canal protein HtsRC (green) in the ring canals makes them appear yellow. In the Drosophila egg chamber, ring canals connect nurse cells to each other and to the oocyte. Late in oocyte development, nurse cell contraction forces their cytoplasmic contents through the ring canals and into the oocyte. This helps the oocyte gain the stockpile of components that is needed for early development of the fly embryo. (Courtesy Andrew Hudson and Lynn Cooley, Yale University, New Haven, CT.)
Ultimately, disassembly of the ESCRT filaments leads to separation of the two daughter cells—abscission. In some tissues, intercellular bridges remain open as ring canals. After several rounds of nuclear division with incomplete cytokinesis, the network of cells maintains cytoplasmic continuity as each former contractile ring matures into a larger ring canal. During Drosophila oogenesis, four rounds of nuclear division with persistent ring canals creates 15 nurse cells, all in continuity with the oocytes (Fig. 44.25). The cytoplasmic continuity through ring canals allows nurse cells to transfer
776
SECTION X n Cell Cycle
BOX 44.1 Cytokinesis in Plants materials (see Fig. 32.13), move along phragmoplast microtubules to the equator, where they fuse due to the action of cytokinesis-specific soluble N-ethylmaleimide-sensitive factor (NSF) attachment protein receptor (SNARE) proteins (see Fig. 21.15), forming a membrane network that becomes the new plasma membrane and laying down the material that will become the new cell wall. Dynamin-related proteins also participate in shaping the newly forming plasma membrane. Thus, the membrane fusion machinery used for cytokinesis by eukaryotes likely came from the last eukaryotic common ancestor. Actin filaments polymerized by formins and myosin-VIII help position the phragmoplast in the cell. As the zone of newly deposited membrane expands radially, the ring of microtubules surrounding it similarly expands. Eventually, the new membrane reaches the lateral cell periphery, and fusion with the plasma membrane separates the two daughter cells. The cortical division site, not the spindle, determines the site of cleavage. This was shown by centrifuging mitotic cells to displace the spindle from the central location where it initially formed. Late in mitosis, the phragmoplast formed at the midzone of the displaced spindle, but this phragmoplast then migrated to the plane of the preprophase band, where cytokinesis occurred. Since plant cells have cell walls and do not move, the orientation of cleavage planes critically determines the morphology of the organism. The hormone auxin can influence cleavage, giving rise to asymmetric division of daughter cells, but the underlying mechanism is not yet known.
Chromosome segregation is similar in plants and animals, but cytokinesis is very different because plants lack myosin-II and do not form a conventional contractile ring (Fig. 44.26). Myosin-II appeared during evolution in the common ancestor of amoebas, fungi and animals, after branching from plants (see Fig. 2.4B). Plants also lack dynein, so microtubule dynamics in mitosis are regulated by some of the more than 20 different plus-end– and minus-end–directed kinesins that are expressed in mitotic cells. In a further difference from animals, plants also lack centrosomes, and during interphase, microtubules radiate out from the surface of the cell nucleus in all directions. In mitosis, the spindle does not focus to sharp poles at metaphase; instead, it assumes a barrel shape with broad, flat poles. Early in mitosis, a band of microtubules and actin filaments forms around the equator of the cell adjacent to the nucleus. This so-called preprophase band disassembles as cells enter prometaphase. Because the entire cell cortex is covered by a meshwork of actin filaments, disassembly of the preprophase band actually leaves an actinpoor zone in a ring where cytokinesis will ultimately occur. This is called the cortical division site, and it is marked by the tethering of specific kinesin motors. In late anaphase, two nonoverlapping, antiparallel arrays of microtubules form over the central spindle. This structure, the phragmoplast, gradually expands laterally until it makes a mirrorsymmetric double disk of short microtubules oriented parallel to the spindle axis with their plus ends abutting the plane of cell cleavage. Golgi vesicles, containing cell wall
Chromosomes
Cytokinesis in higher plants Early phragmoplast
Late phragmoplast
Cortical actin Preprophase band (microtubules)
Cortical actindepleted zone Prophase
Metaphase
Golgi vesicles
Early cytokinesis
Late cytokinesis
Daughter cells
FIGURE 44.26 CYTOKINESIS IN HIGHER PLANTS. See the text for details.
their cytoplasm into the developing egg, thus greatly increasing its stockpile of proteins and messenger RNAs available for use in early development. In mammals, incomplete cytokinesis in the testis results in ring canals connecting several hundred developing sperm cells.
Exit From Mitosis To exit from mitosis, cells must inactivate the Cdk1 kinase. This reverses the biochemical and structural changes that are characteristic of mitosis and prepares the cell for proliferation in the next cell cycle.
Yeast cells use a signaling pathway to terminate mitosis, promote contraction of the contractile ring, and initiate septation. These pathways, called the mitotic exit network (MEN) in budding yeast and the SIN in fission yeast, involve a small GTPase and protein kinases. Cdk kinase activity suppresses the pathway until anaphase, when Cdk activity drops sharply. The MEN GTPase is associated with one spindle pole body (the yeast version of the centrosome), while its key regulator, a GTP exchange factor, is located in the bud. Elongation of the mitotic spindle during anaphase B moves the GTPase into the bud, where it is activated.
CHAPTER 44 n Mitosis and Cytokinesis
777
BOX 44.2 Cytokinesis in Bacteria The strategy for cytokinesis in bacteria is similar to that in animal cells (Fig. 44.27), but the molecules are completely different. Cleavage of most bacterial cells depends on a ring of the FtsZ protein (filamentous temperature-sensitive; mutants in fts genes cannot divide and make long filaments on cells). This is called the Z ring. FtsZ is the prokaryotic homolog of eukaryotic tubulins, but it assembles into filaments rather than tubules. As for tubulins (see Fig. 34.4), FtsZ polymerization requires bound GTP and hydrolysis of this GTP destabilizes the polymers. Although purified FtsZ forms rings that use energy from GTP hydrolysis to deform lipid vesicles, the main function of the Z ring seems to be to coordinate the assembly of a complex of proteins (divisome) including an actin homolog FtsA and number of transmembrane proteins. The transmembrane proteins synthesize cell wall materials to form the cleavage furrow. The Z ring is positioned at the cell equator of Escherichia coli by the action of three gene products: MinC, MinD, and MinE (minicell mutants divide at inappropriate locations and give birth to tiny cells). MinD is an enzyme that recruits MinC
Min C/D inhibitor
Min E Nucleoid
to the cell cortex, where it inhibits Z-ring formation. MinE is an antagonist of MinC/MinD action. This system works in a truly remarkable way. MinE forms a ring at the cell equator that migrates along the inner surface of the cell membrane until it reaches the end of the cell, at which point it disassembles. The ring then reforms in the center of the cell and sweeps toward the other end of the cell. As it moves, MinE inactivates the MinC/MinD inhibitory complex on the cell cortex. The inhibitory complex rapidly reestablishes itself on the cell cortex behind the moving MinE ring. It takes approximately 2 minutes for each sweep of the MinE ring along half of the cell, and this cycle is repeated continuously until the FtsZ ring assembles at the cell center. Bacillus subtilis uses an alternative mechanism to position the Z ring for cytokinesis. Chloroplasts use a homolog of FtsZ for their division, and FtsZ has been detected in mitochondria of certain primitive eukaryotes. Mitochondria of higher eukaryotes appear to use another GTPase, dynamin, to coordinate their fission (see “Biogenesis of Mitochondria” in Chapter 19).
Cytokinesis in E. coli
FtsZ ring
2 minutes
Zone of minimal Min C/D FIGURE 44.27 CYTOKINESIS IN THE BACTERIUM ESCHERICHIA COLI. See the text for details.
The MEN kinases downstream of the GTPase activate the phosphatase Cdc14p by releasing it from sequestration in the nucleolus. Cdc14p inhibits Cdk kinase activity in two ways: (a) it inhibits the degradation of a Cdk inhibitor protein, and (b) it dephosphorylates Cdh1, which binds the APC/C and triggers the degradation of B-type cyclins and other proteins. Cdc14p also triggers other events during anaphase, including the transfer of chromosomal passenger proteins to the central spindle. In metazoans mitotic exit is triggered by the inactivation of Cdk1 and other mitotic kinases. This transition is irreversible, in part because cyclins and Aurora (and other kinases) are degraded. PP2A and its inhibitory kinase Greatwall (see Chapter 40) replace Cdc14 in mitotic regulation in metazoans. Greatwall activity requires Cdks, so when Cdk activity declines, PP2A is released from inhibition. When directed to targets
throughout cells by their specificity, determining subunits PP2A and PP1 remove many of the phosphates placed on target proteins by the mitotic kinases. Targets include chromatin, where phosphorylation during mitosis had displaced factors involved in both gene activation and repression. Removal of those phosphates allows the interphase regulation of gene expression to resume. Dephosphorylation of other targets allows intermediate filaments to reform, nuclear envelope reassembly plus the resumption of RNA transcription, protein translation, and membrane trafficking. ACKNOWLEDGMENTS We thank David Burgess, Iain Cheeseman, Per Paolo D’Avino, Arshad Desai, Tatsuo Fukagawa, Gary Gorbsky, Karen Oegema, Jonathon Pines, and Graham Warren for
778
SECTION X n Cell Cycle
their suggestions on revisions to this chapter. We thank the Dundee Imaging Facility for access to the OMX and help with microscopy. SELECTED READINGS Carmena M, Wheelock M, Funabiki H, et al. The chromosomal passenger complex (CPC): from easy rider to the godfather of mitosis. Nat Rev Mol Cell Biol. 2012;13:789-803. Collas P, Courvalin J-C. Sorting nuclear membrane proteins at mitosis. Trends Cell Biol. 2000;10:5-8. Glotzer M. Cytokinesis in metazoa and fungi. Cold Spring Harb Perspect Biol. 2016;(in press). Green RA, Paluch E, Oegema K. Cytokinesis in animal cells. Annu Rev Cell Dev Biol. 2012;28:29-58. Haeusser DP, Margolin W. Splitsville: structural and functional insights into the dynamic bacterial Z ring. Nat Rev Microbiol. 2016;14: 305-319.
Jürgens G. Plant cytokinesis: Fission by fusion. Trends Cell Biol. 2005; 15:277-283. McIntosh JR. Mitosis. Cold Spring Harb Perspect Biol. 2016;(in press). Müller S, Jürgens G. Plant cytokinesis—no ring, no constriction but centrifugal construction of the partitioning membrane. Semin Cell Dev Biol. 2016;53:10-18. Nasmyth K, Haering CH. Cohesin: its roles and mechanisms. Annu Rev Genet. 2009;43:525-558. Qian J, Winkler C, Bollen M. 4D-networking by mitotic phosphatases. Curr Opin Cell Biol. 2013;25:697-703. Rappaport R. Cytokinesis in Animal Cells: Developmental and Cell Biology Series. Cambridge, England: Cambridge University Press; 1996. Sánchez-Huertas C, Lüders J. The augmin connection in the geometry of microtubule networks. Curr Biol. 2015;25:R294-R299. Sharp DJ, Rogers GC, Scholey JM. Microtubule motors in mitosis. Nature. 2000;407:41-47. Stukenberg PT, Burke DJ. Connecting the microtubule attachment status of each kinetochore to cell cycle arrest through the spindle assembly checkpoint. Chromosoma. 2015;124:463-480.
CHAPTER
45
Meiosis M
eiosis (from the Greek, meaning “reduction”) is a specialized program of two coupled cell divisions used by eukaryotes to maintain the proper chromosome number for the species during sexual reproduction. It also generates novel combinations of genes. Meiosis is an ancient process that occurs in virtually all eukaryotes, including the animal, fungal, and plant kingdoms, and is thought to have been present in the last eukaryotic common ancestor. Sexually reproducing organisms are typically diploid, with pairs of homologous chromosomes, the two highly similar but nonidentical copies of each chromosome, one inherited from each parent. The number of chromosomes is halved during meiosis to form haploid gametes carrying just one set of chromosomes. The subsequent fusion of male and female gametes restores the diploid chromosome number. This pairing and subsequent separation of homologous chromosomes is made possible by genetic recombination, which occurs during the lengthy and complex prophase of the first meiotic division. Each human somatic cell has 23 pairs of chromosomes (46 in all). Females have 23 homologous pairs, while males have 22 “autosomal” pairs and two different sex chromosomes that share a region of homology known as the pseudoautosomal region. One of each pair is contributed by each parent in the egg and sperm, respectively. The number of chromosome pairs, 23, is known as the haploid chromosome number. In animals, the only haploid cells are gametes (sperm and eggs). At fertilization, haploid gametes fuse to form a zygote, restoring the diploid chromosome number of 46. In plants, the haploid phase is represented by gametophytes, which produce ovules and pollen. In most fungi, such as yeasts, haploid and diploid forms are alternate phases of the life cycle, and both can propagate by mitosis. Meiosis changes the genetic makeup of offspring relative to parents in two ways: the first round of meiotic segregation produces novel combinations of
chromosomes from the two parents in each gamete, and recombination between parental chromosomes produces novel chromosomes. It works like this. During meiosis, one round of DNA replication and two rounds of chromosome segregation reduce the number of chromosomes from 2n to 1n. Each haploid gamete is endowed with a random set of the homologous chromosomes from the two parents. Prior to meiosis I the chromosomes duplicate (just like mitosis) (Fig. 45.1A), but during the first round of segregation (Fig. 45.1C) the duplicated chromatids remain paired and the homologous chromosomes separate randomly between the two daughter cells. Thus each daughter cells ends up with just one of each pair of homologous chromosomes. This differs from mitosis where the duplicated chromatids separate, so both daughter cells get the full set of homologous chromosomes from both parents. During meiosis II the duplicated chromatids separate and are partitioned equally between the two daughter cells. Equally important, homologous chromosomes exchange DNA sequences during meiotic prophase I, generating novel chromosomes. The unique segregational events of meiosis usually occur in the first division, termed meiosis I (Figs. 45.1 and 45.2). Because it culminates in daughter cells carrying just one set of chromosomes instead of two, meiosis I is also known as the reductional division. The second division, meiosis II, is similar in most respects to mitosis: sister chromatids segregate, and the number of chromosomes remains the same (Box 45.1; see also Chapter 44). Meiosis II is called the equational division.
Meiosis: An Essential Process for Sexual Reproduction Sexual reproduction is an important survival strategy that offers organisms an accelerated mechanism for altering the genetic makeup of offspring. Without meiosis, there would be no sex, because fusion of diploid gametes 779
780
SECTION X n Cell Cycle
Preparation for meiosis (This step duplicates each chromatid) Centromere
A a
Fused sister centromeres
B b Premeiotic S phase
Two pairs of homologous chromosomes
Meiotic prophase
Meiotic prophase (Recombination drives pairing of homologous chromosomes and produces novel chromosomes) Recombination nodules A b
Chiasmata
A
a
a
B
B Leptotene stage
Bouquet
Meiosis I
b
Zygotene stage
Pachytene stage
Diplotene stage
Diakinesis stage
Meiosis I (Homologous chromosomes randomly separate from one another producing haploid progeny) Chiasmata A
b
a
B
Metaphase I
A A
b
a
B
b
B
Meiosis I
Anaphase I
a
Meiosis II
Interphase (no S phase)
Meiosis II (Sister chromatids separate producing one or more gametes) A b
B a
Prophase II
Metaphase II
Anaphase II
Haploid gametes
FIGURE 45.1 OVERVIEW OF THE PHASES OF MEIOSIS. Shown are important structures and the outcome of each stage for a homologous pair of metacentric chromosomes (A, a) and a homologous pair of telocentric chromosomes (B, b).
would double the number of chromosomes in the progeny at every generation. Meiosis I produces random combinations of homologous maternal and paternal chromosomes. For each pair of homologs, orientation on the spindle is random during meiosis I (ie, each homolog has two equivalent options for the direction to migrate). Thus, for humans (with 23 pairs of homologous chromosomes), each gamete has one of 223 (more than 8 million) possible complements of maternal and paternal chromosomes. This process does not create new versions of genes, but it guarantees the offspring will have novel combinations of subtly different (due to polymorphisms) chromosomes. Meiosis I also produces novel versions of genes and chromosomes by recombinational exchange of DNA segments between homologs. This occurs because to segregate from one another, each pair of homologous
chromosomes must first find each other. They do this by undergoing reciprocal recombination (crossover) events that then hold them together until anaphase of meiosis I. Chromosomes and the genes they carry vary hugely between individuals. In humans an average genome varies from the “reference genome” (see Chapter 7) at 4 to 5 × 106 sites. These include not only polymorphisms (differences of single base pairs), but also thousands of longer insertions, deletions, and rearrangements. Recombination events that result in a crossover and exchange chromosomal segments produce new chromosomes that are a patchwork of segments from the maternal and paternal homologs. The combined effects of recombination and random assortment of homologs in meiosis I yields a vast number of genetically different gametes. This genetic diversity increases the ability of eukaryotic populations to adapt to changing environmental conditions.
CHAPTER 45 n Meiosis
A. Metaphase I
C. Late anaphase I
Spindle pole Paired sister kinetochores being pulled toward poles
Spindle pole
X X Chiasma
Spindle pole
D. Telophase I
Chiasma Spindle pole
B. Early anaphase I
781
Paired sister kinetochores moving toward poles
Spindle pole X
Spindle pole
Spindle pole
Spindle pole
FIGURE 45.2 FIRST MEIOTIC DIVISION STAGES FROM THE GRASSHOPPER PYRGOMORPHA CONICA (2N IN MALES = 18 AUTOSOMES + 1 X CHROMOSOME). A, Metaphase I. B, Early anaphase I. C, Late anaphase I. D, Telophase I spermatocytes stained with lactopropionic orcein. All chromosomes are telocentric (see Fig. 7.2). Seven bivalents shown in the metaphase I spermatocyte have a single chiasma, whereas the two bivalents at the far right and far left have two chiasmata. The sex chromosome (X) remains unpaired and moves to a single spindle pole. (Courtesy José A. Suja and Julio S. Rufas, Universidad Autónoma de Madrid, Spain.)
The Language of Meiosis Meiosis has a language of its own, characterized by a number of unusual terms, and is easiest to understand by focusing on the essential biological processes that are involved. This reduces the process to only three essential key terms: pairing, homologous recombination, and segregation. This chapter discusses each step in detail, so they are defined only briefly here. Pairing is a two-step alignment of homologous chromosomes with one another in the nucleus. In alignment, corresponding DNA sequences on the homologous chromosome find each other among the billions of base pairs of DNA in the nucleus. In many organisms, early events of recombination drive the homologous pairing process. In the second stage, synapsis, the paired homologous chromosomes become intimately aligned along their entire lengths with one another separated by approximately 100 nm. A specialized scaffolding structure called the synaptonemal complex mediates this process. Homologous recombination results in physical exchange of DNA between homologous chromosomes (a crossover event) and is a key determinant of chromosome behavior during meiotic prophase. Recombination drives the pairing process in many organisms and can occur without synapsis under certain circumstances. Crossover recombination sites are detected by microscopy
as chromatin structures called chiasmata (singular: chiasma, from the Greek, meaning “X-shaped cross”). Segregation of homologous chromosomes in meiosis I differs from the segregation of sister chromatids during mitosis (Box 45.1), because the paternal and maternal homologous chromosomes segregate randomly to the two daughter cells. When homologs orient at the metaphase plate of the meiosis I spindle, centromeres belonging to the two sister chromatids are fused to form a single kinetochore that binds microtubules. Cohesion between chromosome arms distal to chiasmata (ie on the other side from the centromere; Figs. 45.2 and 45.10) keeps homologous chromosomes paired with one another until anaphase of meiosis I, counteracting the bipolar pulling force of the spindle on the homologs (Fig. 45.2). At anaphase I the distal cohesion is released from chromosomes allowing the chiasmata to separate, and the two sister chromatids (at least one of which has undergone a crossover exchange) move as a single unit toward the same spindle pole while the sister chromatids from other parent move to the other daughter cell. As a result, the two daughter cells produced in meiosis I have a haploid number of chromosomes derived randomly from the two parents, each with two sister chromatids. Each of the four daughter cells produced in meiosis II has one sister chromatid for each homologous chromosome (ie, half the number of chromatids as there are chromosomes in somatic cells).
782
SECTION X n Cell Cycle
BOX 45.1 Important Differences Between Meiosis
and Mitosis
Meiosis involves two cell divisions. The two meiotic divisions are preceded by a round of DNA replication. There is no DNA replication between meiosis I and meiosis II. The products of meiosis are haploid. The products of mitosis are diploid. The products of meiosis are genetically different. After recombination and random assortment of homologs in meiosis I, the sister chromatids that segregate in meiosis II are different from each other. In normal mitosis, sister chromatids are identical. Prophase is longer in meiosis I. Proper orientation and segregation of homologous chromosomes is achieved thanks to the pairing, synapsis (synaptonemal complex formation), and recombination that occur in a lengthened prophase during meiosis I. In humans, prophase in mitosis takes an hour, whereas meiotic prophase lasts many days in males and many years in females. Recombination is increased in meiosis. The recombination rate is 100- to 1000-fold higher in prophase I of meiosis than in mitosis. The process has two main consequences: the formation of chiasmata and the introduction of genetic variation. Chiasmata are structures that physically link the homologous chromosomes after crossover and play an essential role in meiotic chromosome segregation. Kinetochore behavior differs in meiosis. During meiosis I, kinetochores of sister chromatids attach to spindle microtubules emanating from the same pole. Homologous kinetochore pairs connect to opposite poles. In mitosis and meiosis II, sister kinetochores attach to spindle microtubules coming from opposite poles. Chromatid cohesion differs in meiosis. Sister chromatid cohesion is essential for orientation of bivalents (paired homologous chromosomes) on the metaphase I spindle. During anaphase of meiosis I, cohesion is destroyed between sister chromatid arms, and chiasmata are released to allow segregation of homologs. Cohesion at sister centromeres persists until the onset of anaphase II, when it is lost to permit segregation of sisters. In prometaphase of meiosis II, sister chromatids are joined only by the centromeres, whereas at the beginning of mitotic prometaphase, sisters are joined all along the arms.
Recombination Although meiotic recombination is similar to the process of homologous recombinational repair of double-strand DNA breaks in somatic cells (review Box 43.1 and Fig. 43.14 as a prelude to studying meiotic recombination), the two processes differ in two respects. First, meiotic cells use a specialized enzyme called Spo11 to create double-strand DNA breaks on purpose. Second, somatic cells with replicated chromosomes usually repair DNA
breaks using the corresponding DNA sequence on a sister chromatid as a template. Meiotic cells usually use a homologous chromosome. The mechanism for this difference in selectivity is not yet fully understood. Spo11, together with essential accessory proteins, generates programmed double-strand DNA breaks early during meiotic prophase (Fig. 45.3). Similar to type II DNA topoisomerases (see Fig. 8.16), Spo11 cleaves both DNA strands in a reaction that produces a covalent linkage between a tyrosine of the enzyme and the cleaved phosphodiester backbone. However, Spo11 does not reseal the breaks; instead it remains attached to one strand of DNA at the broken end. In mice, Spo11 creates about 10-fold more DNA breaks than ultimately recombine to produce reciprocal DNA exchanges between homologous chromosomes or crossovers. Repair of Spo11-mediated DNA double-strand breaks can also result in noncrossover events known as gene conversions (Box 45.2 and Fig. 45.3I–J). DNA double-strand breaks generated by Spo11 are required for the initial lengthwise alignment of homologous chromosomes in many organisms, including mice, plants, and yeast. In mice lacking Spo11, recombination is not initiated, and synapsis, if it occurs at all, is aberrant, often involving nonhomologous chromosomes (Fig. 45.4). Gametes in these mutant mice die by apoptosis early in meiotic prophase. Spo11-induced double-strand breaks are not required for synapsis of homologous chromosomes in the nematode Caenorhabditis elegans and the fruit fly Drosophila melanogaster. How these organisms pair their homologs without recombination is still mysterious. Once the DNA double-strand breaks are produced, the MRN (Mre11/Rad50/Nbs1) endonuclease nicks the single-stranded DNA, releasing Spo11. The DNA ends that lost Spo11 then undergo further processing, as Exo1 exonuclease, chews back the 5′ strands of the double helix (a process called resection) leaving single-stranded DNA tails with 3′ termini (Fig. 45.3C; see also Fig. 43.14). The MRN and Exo1 nucleases also function in somatic DNA repair. Next, the Rad51 and Dmc1 proteins drive a search of the 3′ single-stranded DNA tails for complementary DNA sequences of the other chromosomes. Rad51 and Dmc1 are related to the Escherichia coli RecA protein used for homologous recombination in bacteria. Rad51 and Dmc1 coassemble along 3′ single-stranded DNA tails and use adenosine triphosphate (ATP) hydrolysis to catalyze a strand exchange reaction with an intact homologous DNA duplex. The Rad51-and-Dmc1-decorated nucleoprotein filament disrupts the targeted homologous double helix, displacing one of the two DNA strands. This allows formation of new Watson-Crick base pairs between the invading 3′ single-stranded DNA and the complementary strand of the target DNA. Following strand invasion and exchange, new DNA synthesis
A
5'
3'
3' 5' 3'
5' 3' 5' 3'
5'
Cohesin Paired homologous chromosomes
5' 3'
3' 5' 3'
5'
Spo11 cuts DNA, remaining covalently attached to one strand
Double-strand break
B
Red enzymes are meiosis specific
5' to 3' resection
Spo11 released by MRN complex (Rad50/ Exo1 resects one DNA strand Mre11/Nbs1)
Green enzymes also function in somatic C repair
Initial strand invasion, DNA synthesis
Rad51 + Dmc1
D
E
Second end capture, synthesis, ligation
H
Strand displacement
Formation of double Holliday junction
Strand annealing
I
F Holliday junction
Holliday junction resolution
Synthesis, ligation
= Resolvase cutting sites
J
G +
+
Crossover
Noncrossover
FIGURE 45.3 EVENTS OF RECOMBINATION. Recombination occurs between homologous chromosomes rather than between sister chromatids. A, Paired homologous chromosomes. Sister chromatids are held tightly together by cohesin, shown here schematically as hoops. B, Spo11 makes a double-strand break, remaining attached to the DNA. C, Removal of Spo11 and resection of the break. D, First strand invasion. At this point, the pathway splits in two, one outcome leading to a crossover and the other to a noncrossover. Crossover pathway: E, The second resected strand establishes base-pairing interactions with the displaced DNA strand of its homologous partner. New DNA synthesis fills the gaps. F, The resulting molecule contains a double Holliday junction in which the DNA is fully base-paired (see Fig. 43.14B). If the resolvase (nuclease) cuts the double Holliday junction asymmetrically as shown (ie, one vertical and one horizontal cut), the result is a crossover (G). If the cuts are symmetrical, a noncrossover molecule is produced. Noncrossover pathway: H, In most cases, the invading DNA, strand is ejected prior to stabilization and formation of a double Holliday junction. I, DNA gap-filling and ligation yield a noncrossover chromosome (J).
784
SECTION X n Cell Cycle
BOX 45.2 Brief Overview of Genetic Terminology A comprehensive introduction to the field of genetics is beyond the scope of this text. However, here are a number of terms used by geneticists that will assist in the understanding of the discussion of genetic recombination and its role in meiosis (also see Box 6.1). The genotype of an organism is the combination of genes present on the chromosomes of that organism. The phenotype is the physical manifestation of the action of these gene products (ie, the appearance and macromolecular composition of the organism). In discussing recombination, scientists typically refer to the presence or absence of specific genetic markers. Each genetic marker is a particular DNA sequence in or around a gene that can be monitored by examining the phenotypes of the cells that carry it. A genetic marker might be the presence of a functional gene, a mutation with altered activity, or simply a polymorphism of DNA sequence that has no known functional consequence. A haploid organism has one copy of each chromosome. A diploid organism has two homologous copies of each chromosome. A diploid organism that is homozygous for a particular genetic marker has the same sequence of that particular region of the DNA on both the maternal and paternal homologous chromosomes. A heterozygous organism has different forms of the genetic marker on the two homologous chromosomes. Although the physical events of genetic recombination occur in both homozygotes and heterozygotes, they are most readily detected in the latter. Two genetic markers located on different chromosomes will separate from one another in the anaphase of meiosis I
A
B
FIGURE 45.4 PAIRING OF HOMOLOGOUS CHROMOSOMES IS SEVERELY DISRUPTED IN THE SPO11 MUTANT. Pachytene chromosomes from wild-type mice (A) and mice in which the Spo11 gene has been disrupted (B). SYC3 (axial elements) and centromeres are red. SYCP1 (in transverse filaments, which are seen only when synapsis has occurred) is green. (From Baudat F, Manova K, Yuen JP, et al. Chromosome synapsis defects and sexually dimorphic meiotic progression in mice lacking Spo11. Mol Cell. 2000;6:989–998.)
50% of the time as a result of the random distribution of chromosomes to the two spindle poles. If they are on the same chromosome, they will be linked to one another unless the chromosome undergoes a genetic recombination event between them. The greater the separation of two markers along one chromosome, the more likely it is for such an intervening recombination event to occur. Two types of recombination events occur during meiosis (Fig. 45.3). The first of these—noncrossover events (frequently referred to as gene conversion)—may involve the loss of one or more genetic markers. Noncrossover events are the most common outcome of the programmed doublestrand DNA breaks that occur during leptotene. They are thought to involve the invasion of a double helix by a region of single-stranded DNA with complementary sequence but then ejection of this sequence before assembly of a Holliday junction and completion of recombination. The second type of recombination event—crossing over—involves the physical breakage and reunion of DNA strands on two different chromosomes, typically producing a balanced exchange of DNA sequences. This is what most people think of as recombination. In recombination by crossing over, the makeup of genetic markers remains constant; it is the linkage between different markers that changes. The normal separation of chromosomes or chromatids is referred to as disjunction (disjoining). Mistakes in this separation are referred to as nondisjunction. Nondisjunction in meiosis I and II results in the production of gametes with either too many or too few chromosomes, a condition known as aneuploidy.
restores sequences that may have been lost or damaged at the position of the original DNA double-strand break. Mutants lacking Dmc1 are defective in homologous chromosome pairing and interhomolog recombination. As a result, Dmc1 is thought to facilitate the search for homologous chromosomes as a DNA repair template, rather than sister chromatids as in somatic DNA repair. Rad51p and Dmc1p are found in structures called early recombination nodules that are distributed along the chromosome axes early in meiosis (Fig. 45.9). Dmc1 functions only in meiosis, but Rad51 has other essential functions. It is now believed that noncrossover events arise primarily from recombination intermediates that involve a relatively transient single strand invasion of the homologous chromosome followed by restorative DNA synthesis and disassembly of the joint molecule intermediate. Crossovers, on the other hand, are thought to arise predominantly through a pathway that involves stable branched intermediates known as Holliday junctions (Fig. 45.3F–G; see also Fig. 43.14B), which are then cleaved by resolvases such as Gem1 to form (predominantly) crossover products.
CHAPTER 45 n Meiosis
Most sexually reproducing organisms depend on recombination during meiosis to produce haploid gametes, but fruit flies and yeast have other systems for segregating homologs in meiosis I. These mechanisms, collectively known as achiasmate segregation, allow the segregation of chromosomes that have not undergone crossover recombination. One model for the achiasmate segregation in flies proposes that nonrecombined chromosomes remain paired at the end of meiotic prophase owing to stickiness of heterochromatin and, as a result, segregate properly during anaphase I of meiosis. In a rare but notable example, the spermatocytes of D. melanogaster males do not recombine at all, yet still segregate their chromosomes happily during meiosis I. This might be regarded as a cruel joke of evolution by those students who find all the Greek terms of meiotic nomenclature to be daunting. However, meiosis without recombination is clearly the exception, and in most species meiosis depends on recombination in both males and females.
A. Early leptotene
785
Tracking the Homologous Chromosomes Through the Stages of Meiotic Prophase I Pairing and recombination of homologous chromosomes take place during prophase of meiosis I. Five stages of meiotic prophase are used to describe the process: leptotene, zygotene, pachytene, diplotene, and diakinesis (Fig. 45.1B). Leptotene (from the Greek, meaning “thin ribbon”) starts with the first visible condensation of the chromosomes. Paired sister chromatids become visible as linear arrays of loops flanking a single dense protein-containing axis (Fig. 45.5A–B). This axis consists of proteins that play a role in mitotic chromosome structure as well as proteins specialized for meiotic chromosomes. For example, the cohesin complex with several specialized meiosis-specific subunits is a prominent component of this axial structure (see Fig. 8.18). According to recent models, recombination begins during leptotene with the formation of DNA double-strand breaks. By the end of
C. Early zygotene
E. Pachytene
Sex chromosomes
B. Late leptotene
D. Late zygotene
F. Diplotene Chiasma
FIGURE 45.5 IMMUNOFLUORESCENCE IMAGES OF PROPHASE I SUBSTAGES IN MOUSE SPERMATOCYTES. These images demonstrate the pairing and synapsis of homologous chromosomes revealed by visualizing the synaptonemal complex proteins SYCP3 (a component of the axial elements [red]) and SYCP1 (a component of the transverse filaments that is present only when homologs are synapsed [green]). Centromeres are blue. (Courtesy Paula Cohen, Cornell University, Ithaca, NY.)
786
SECTION X n Cell Cycle
A
C
B
D
E
F
G
H
0 hr
I
2 hr
8 hr
Telomere
Nuclear envelope FIGURE 45.6 CHROMOSOMAL MOVEMENTS DURING EARLY MEIOTIC PROPHASE. A–G, Pairing of homologous chromosomes during leptotene in the ascomycete Sordaria. Scale bar is 1 µm in A–F and 5 µm in G. A–B, In early leptotene, homologous chromosomes (visualized in panels A–F by electron microscope reconstructions of serial-sectioned nuclei) are not yet aligned with one another. C–E, In mid-leptotene, regions of some homologs begin to align. (In panel D, only the telomeres have aligned. In panel E, the pair of homologs is fully aligned.) F, The alignment of homologs is complete by late leptotene. G, The alignment of homologs also can be seen by light microscopy using Spo76-GFP, a component of the chromosome axes. H–I, Stages of formation of the bouquet arrangement in rye. H, Telomeres (green) were detected in nuclei by in situ hybridization (see Fig. 8.10) after 0, 2, and 8 hours in culture. Chromatin is red. I, Three-dimensional models of the nuclei (nuclear periphery [red dots], telomere position [green stars]). (A–G, Modified from Tesse S, Storlazzi A, Kleckner N, et al. Localization and roles of Ski8p protein in Sordaria meiosis and delineation of three mechanistically distinct steps of meiotic homolog juxtaposition. Proc Natl Acad Sci U S A. 2003;100:12865–12870. Copyright 2003 National Academy of Sciences. H–I, Modified from Carlton PM, Cowan CR, Cande WZ. Directed motion of telomeres in the formation of the meiotic bouquet revealed by time course and simulation analysis. Mol Biol Cell. 2003;14:2832–2843.)
leptotene, homologous chromosomes are aligned loosely about 400 nm apart (Fig. 45.6D–G). During leptotene, one or both telomeres attach to the inner surface of the nuclear envelope and move actively around the nuclear surface until they coalesce near the centrosome (spindle pole body in yeasts [Fig. 45.6]). These movements and clustering of telomeres depend on cytoplasmic microtubules. Telomeres are linked to microtubules through a pair of nuclear envelope proteins known as the LINC (linker of nucleoskeleton and cytoskeleton) complex (see Fig. 9.8). Telomere clustering peaks at the leptotene–zygotene transition with the chromosomes radiating into the nuclear interior like a bouquet of flowers, hence the name “bouquet stage.” Bouquet formation is a nearly universal feature of this phase of meiosis and the movements of tethered telomeres help homologs find each other through physical alignment. Thus, telomere clustering per se may not be the goal of this movement. The details vary among different organisms. In fission yeast dynein motors and microtubule dynamics in the cytoplasm move the telomere cluster from one end of the cell to the other every 10 minutes or so. These “horsetail movements” stretch
the chromosomes parallel to each other. In C. elegans special chromosome regions known as “pairing centers” mediate chromosome movement instead of telomeres; in budding yeast, the telomeres are linked to actin instead of microtubules, and D. melanogaster may have lost such a mechanism altogether. During the transition from leptotene to the zygotene (Greek, “yoke ribbon”) stage of prophase, clustering of chromosome ends at the nuclear envelope reaches its peak, with the “bouquet” arrangement of chromosomes. During this stage homologous chromosomes begin to achieve their maximal alignment as well, through the initiation of synapsis (Fig. 45.5C–D). Synapsis involves the assembly of the axial element. This protein scaffold forms part of the synaptonemal complex when pairing is complete. In pachytene (from the Greek, meaning “thick ribbon”), synapsis is complete, with the homologous chromosome axes joined together along their lengths by synaptonemal complexes (Fig. 45.5E). During pachytene, crossover-designated recombination intermediates mature into Holliday junction-containing structures within the context of the full-length synaptonemal
CHAPTER 45 n Meiosis
The earliest pairing events in meiosis involve a tendency of homologous chromosome territories to move together in the nucleus even before leptotene chromosome condensation. The mechanism is unknown. As double-strand breaks created by Spo11 initiate the recombination pathway during leptotene, the condensing homologous chromosomes align with one another at a distance of about 400 nm (Figs. 45.6 and 45.7). Genetic analysis in budding yeast revealed that mutants defective in the earliest stages of recombination are also defective in homolog pairing. Interphase Sister chromatid 1
Replicating DNA Sister chromatid 3 Sister chromatid 2
Sister chromatid 4 Cohesin
Sister chromatids linked by cohesin in premeiotic S phase
Leptotene
Initiation of recombination Pairing of homologous chromosomes
Chromatid axis
Zygotene
Assembling central element of synaptonemal complex
Pachytene
Synapsis Axial (lateral) elements Transverse filaments Central element
Diplotene followed by diakinesis Disassembling synaptonemal complex
Time
complex. The final resolution of these recombination intermediates into crossovers occurs close to the time of synaptonemal complex disassembly, dispersal of the bouquet of chromosomes and exit from pachytene. The crossovers then mature into structures called chiasmata that link homologous chromosomes through meiosis I metaphase. Early in diplotene (from the Greek, meaning “double ribbon”), the synaptonemal complex disassembles, telomeres detach from the nuclear membrane, and chromosomes begin to condense in preparation for division (Fig. 45.5F). The duplicated sister chromatids remain closely associated, and chiasmata hold the homologous chromosomes together, although their axes tend to drift apart in the absence of synaptonemal complex. This part of meiotic prophase may last for days or years, depending on the sex and organism (up to 45 years or more in female humans). Oocytes (immature eggs) actively transcribe their chromosomes during diplotene, as they store up materials for use during the first few divisions of embryonic development. Transcription can be so active that DNA loops are massively coated with nascent RNA transcripts whose associated proteins are visible by light microscopy in oocytes of most animals (except mammals). Chromosomes at this stage are known as lampbrush chromosomes (see Fig. 8.12). Diakinesis (from the Greek, meaning “across movement”) is the prometaphase of meiosis I. Following nuclear envelope breakdown, homologous chromosomes shorten and condense. At metaphase I, the bivalents (pairs of homologous chromosomes) are aligned at a metaphase plate (Figs. 45.1, 45.2, and 45.12). The two homologs (each a pair of tightly linked sister chromatids) are attached to opposite poles of the meiotic spindle, which applies force, attempting to pull them apart. Cohesion of the arms distal to chiasmata resists these pulling forces. The homologs separate and move to opposite spindle poles during anaphase I when the cohesion along the chromosome arms is released. The sister chromatids move together to one pole, because they remain linked by cohesion at their centromeres, where the cohesion complex is protected by a shugoshin protein (see later). After telophase I, cells enter a brief interkinesis during which there is no DNA replication. The second meiotic division is mechanistically similar to mitosis except that the number of chromosomes is reduced by half. Additionally, in the eggs of most female vertebrates, meiosis is arrested at metaphase II until fertilization.
787
Chromatin loops
Chiasma
Pairing and Synapsis in More Detail Pairing describes the side-by-side alignment of homologous chromosomes at a distance. Homologs are paired in nonmeiotic (somatic) cells in a few organisms, such as the fruit fly D. melanogaster, but not in vertebrates.
FIGURE 45.7 CHROMOSOMAL PAIRING IN MEIOTIC PROPHASE. Structural organization of homologous chromosomes and synaptonemal complex during various stages of meiotic prophase.
788
SECTION X n Cell Cycle
The process of homolog alignment almost certainly involves the invasion of neighboring DNA duplexes by single-stranded DNA complexed with Rad51 and Dmc1. Thus, recombination has important roles both in the exchange of genetic material and in the mechanics of chromosome behavior during meiotic prophase. Recombination is probably not the only factor driving homolog pairing, however. Pairing is reduced but not absent in yeast meiotic cells lacking both Rad51 and Dmc1, and homologous chromosomes still pair in some systems that lack recombination (eg, certain D. melanogaster recombination mutants), synaptonemal complex formation (asynaptic mutants in yeast), or both (eg, normal D. melanogaster males). Homolog pairing initiated during leptotene be comes much more intimate during synapsis as the chromosomes are linked by transverse fibers to form the synaptonemal complex. This structure looks roughly like railroad tracks linked by transverse bands (Figs. 45.7 and 45.8). Each of the two outer rails, 90 to 100 nm apart, is the axis of a pair sister chromatids. They are traditionally termed lateral elements, but for the sake of simplicity, we refer to them as axial elements. Thin transverse filaments oriented perpendicular to the axial elements appear to connect homolog axes to each other and to the central element (the “third rail”). Synaptonemal complex formation begins during zygotene at a limited number of sites along the paired homologous chromosomes where recombination
A
events will mature into crossovers. By pachytene, a continuous synaptonemal complex assembles along the full length of the aligned homologous chromosomes (Fig. 45.5C–E). It was once thought that the synaptonemal complex aligns homologous chromosomes in preparation for recombination, but it is now clear that homolog pairing and (in many organisms) the initiation of recombination precedes synapsis. Thus, synapsis is a downstream consequence of early steps in recombination in some well-studied organisms including yeast and mammals. However, under certain artificial circumstances, even nonhomologous chromosomes can undergo synapsis. Another longstanding model proposed that the synaptonemal complex promotes the resolution of crossoverdesignated recombination intermediates. However, analysis of budding yeast mutants missing certain synaptonemal complex proteins indicates that the structure per se is dispensable for the formation of crossovers and that the resulting chiasmata can hold homologous chromosomes paired until anaphase of meiosis I. What then is the function of the synaptonemal complex? One possibility is that it may have a key role in crossover interference (see later), which ensures that crossovers are distributed broadly across the genome. Another interesting possibility is that the synaptonemal complex communicates information about meiotic chromosomes (such as homolog alignment and the formation of crossover-designated recombination
B
C
CE
CE
LE
FIGURE 45.8 ELECTRON MICROGRAPHS OF THE SYNAPTONEMAL COMPLEX. A, Low-magnification view of maize synaptonemal complexes stained with silver. The lateral (LE) and central elements (CE) are clearly seen. B, A negatively stained cricket synaptonemal complex following treatment with deoxyribonuclease (DNase). The central element (CE) and transverse filaments (arrow) are visible. C, A whole mount of a silk moth zygotene chromosome. Cells in meiotic prophase were swollen and then lysed under gentle conditions with detergent. The chromosomes were then centrifuged onto thin carbon films so that they could be examined by electron microscopy. The axial elements are easily seen on this chromosome. Chromatin loops radiate outward from both the unpaired axial elements and the paired lateral elements (where synapsis has occurred). (A, Modified from Gillies CB. Electron microscopy of spread maize pachytene synaptonemal complexes. Chromosoma. 1981;83:575–591. B, Modified from Solari AJ, Moses MJ. The structure of the central region in the synaptonemal complexes of hamster and cricket spermatocytes. J Cell Biol. 1973;56:145–152, copyright the Rockefeller University Press. C, From Rattner JB, Goldsmith M, Hamkalo BA. Chromatin organization during meiotic prophase of Bombyx mori. Chromosoma. 1980;79:215–224.)
CHAPTER 45 n Meiosis
intermediates) to the cell-cycle pathways that control progression through the substages of meiosis.
Synaptonemal Complex Components Both genetic and biochemical approaches have identified components of the synaptonemal complex. The budding yeast protein Zip1 (mammalian SYCP1) comprises the transverse filaments oriented perpendicular to chromosome axes in mature synaptonemal complex, between the axial elements (Fig. 45.9). Mammalian SYCP1 and Zip1 both consist of an extensive coiled-coil flanked by two globular domains but lack amino acid sequence similarity. Altering the length of the Zip1 coiled-coil changes the spacing between axial elements in the synaptonemal complex. Leptotene
SYCP3 + SYCP2, Cohesin
Paired sister chromatids Axial element (chromosome axis)
Time
Chromatin loops Rad51, Dmc1 Early recombination nodule (may be site of noncrossover events) SYCP1 (yeast Zip1)
Zygotene Synaptonemal complex assembles from site of crossover (late recombination nodule)
Zip2/Zip4, Zip3 (yeast synapsis initiation complex) Mlh1/Mlh3, Msh4/Msh5 SYCE2, TEX12
FIGURE 45.9 PROTEINS OF THE SYNAPTONEMAL COMPLEX. Homologous chromosomes and synaptonemal complex showing the locations of some protein constituents.
A
789
Several protein components of the axial elements (sister chromatid axes) have also been identified. One of these, SYCP3, interacts with both the cohesin complex (see Fig. 8.18) and Rad51p and Dmc1p. In SYCP3 knockout mice, the axial elements are much less prominent, and the axis of the condensed chromosome is about twofold longer. Other proteins of the synaptonemal complex, including SYCP1, do not assemble properly, and as a result chromosomes in male germ cells lack chiasmata, are unpaired, and cells die in pachytene/diplotene. Humans with mutations in genes for cohesin subunits lack chiasmata, fail to complete meiosis, and are infertile.
Chiasmata Chiasmata are specialized chromatin structures that link homologous chromosomes together until anaphase I (Figs. 45.1 and 45.10). They form at sites where programmed DNA breaks generated by Spo11 undergo the full recombination pathway to generate crossovers. It is not known how crossover events, which represent exchanges of DNA sequence information, are turned into chiasmata. The ultrastructure of chiasmata remains a mystery, but presumably each chiasma consists of two unperturbed sister chromatid arms intertwined with two recombinant arms in which the DNA molecules and their associated protein structures have been spliced. This DNA complex is held in place on the chromosome by cohesion of the distal sister chromatid arms between the chiasma and the telomeres. Chiasmata too close to telomeres can be unstable, presumably because the short length of sister chromatid arms between them and the telomeres is insufficient for stable cohesion. This can lead to failure of chromosome segregation in meiosis. A single chiasma can link homologous chromosomes together during meiosis I. Humans have 39 such arms
B
C
Microtubules Fused sister kinetochores Paired sister chromatids (maternal homolog) Distal cohesin and Aurora B Chiasma
Arrows point to chiasmata
AIR-2 REC-8
REC-8
Paired sister chromatids (paternal homolog)
FIGURE 45.10 BIVALENTS (PAIRED HOMOLOGOUS CHROMOSOMES) ARE HELD TOGETHER BY CHIASMATA AFTER DISASSEMBLY OF THE SYNAPTONEMAL COMPLEX. A, Three diplotene bivalents from the grasshopper species Chorthippus jucundus are held together by three (left), one (middle), and four (right) chiasmata. The middle cross-shaped bivalent is telocentric; the other two longer bivalents are submetacentric. (For an explanation of the terminology, see Fig. 7.2.) Lactopropionic orcein staining. B, Caenorhabditis elegans chromosomes at metaphase I. Aurora B kinase AIR-2 (red) is located distal to chiasmata. Cohesin subunit REC-8 (green) is all along the chromosomes. C, Explanatory diagram. (A, Courtesy José A. Suja and Julio S. Rufas, Universidad Autónoma de Madrid, Spain. B, Courtesy Josef Loidl, Max Perutz Labs, Vienna Biocenter, Austria.)
790
SECTION X n Cell Cycle
A. Normal
B. Infertile male
SB SB
FIGURE 45.11 ABNORMAL PACHYTENE CHROMOSOMES IN AN INFERTILE MALE. A, Normal pachytene chromosome spread from a testis biopsy showing synaptonemal complexes (red), MLH1 foci (recombination sites [green]), and centromeres (blue). B, Abnormal pachytene spread from an infertile patient containing one synaptonemal complex with an area of asynapsis and one synaptonemal complex with a gap (arrows). SB, sex body (the paired X and Y chromosomes). (Courtesy Renée H. Martin, University of Calgary, Alberta, Canada.)
on the 23 pairs of homologous chromosomes, if one excludes the five acrocentric short arms, which do not normally undergo crossovers. Remarkably, there is typically only one chiasma produced for most arms; human males typically have 46 to 53 chiasmata (Fig. 45.11). Even more remarkably, that single chiasma can hold homologous stably paired for over 40 years in human females, yet still be released on schedule when the oocyte matures into an egg. Only a small fraction of DNA breaks formed by Spo11 mature into full crossovers, because a mechanism called crossover interference decreases the likelihood that DNA breaks near a crossover-designated recombination event will also become crossovers. This interference tends to spread crossovers apart across the genome. If all breaks had an equal probability of forming crossovers, small chromosomes might be left without a crossover in a significant fraction of meiotic nuclei if large chromosomes used all of the structural components necessary to form crossovers and chiasmata. Crossover interference has been defined genetically for almost 100 years, but its mechanism is not certain. Interference may be mediated by the synaptonemal complex. Organisms such as the fission yeast Schizosaccharomyces pombe and the mold Aspergillus nidulans that naturally lack synaptonemal complex also lack interference. Furthermore, the frequency of meiotic recombination is directly proportional to the length of the synapsed chromosome axis (ie, the length of the synaptonemal complex), rather than the actual length of DNA in the chromosome. For example, in human females, the synaptonemal complex is roughly 50% longer than in males, and females undergo recombination about twice as frequently as males. However, other observations suggest that interference may be established before the synaptonemal complex forms.
Cohesion and Chromosomal Movements During Meiosis I Chromosomes in mitosis achieve a dynamic alignment at metaphase as a result of a balance of forces in the spindle. In mitosis, the two kinetochores of the sister chromatids are attached to microtubules emanating from opposite spindle poles, and each chromatid is pulled toward the pole that its kinetochore faces (Fig. 45.12A). In meiosis I, homologs linked by chiasmata (called bivalents) are balanced at the metaphase plate, but the organization differs in three important ways from mitosis. First, the two kinetochores of the sister chromatids are fused and act as a single unit oriented toward one spindle pole. The structure of the meiosis I kinetochore is most easily explained if the two kinetochores are each rotated 90 degrees toward one another relative to their position on mitotic chromosomes and then fused (Fig. 45.12A). In yeast, this coorientation of sister kinetochores requires the presence of a meiosis-specific kinetochore protein— spo13 (meikin in vertebrates)—that associates with sister kinetochores from pachytene until anaphase of meiosis I. Spo13/meikin recruits polo kinase to kinetochores, but the critical kinase substrates are not known. Second, the physical connection between the fused kinetochores is not broken in anaphase I. As a result, at anaphase the two homologs with their paired sister chromatids move in opposite directions. A third major difference between bivalents in meiosis I and mitotic chromosomes is that cohesion of homologous chromosome arms distal to chiasmata (Figs. 45.2 and 45.10) rather than cohesion between sister chromatids at centromeres resists the poleward pulling of the kinetochores at the spindle midzone at metaphase I (Fig. 45.12B). This reflects specialized behavior of the meiotic cohesin complex in which the meiosis-specific
CHAPTER 45 n Meiosis
791
A. Kinetochore movement Kinetochore
90° rotation of kinetochores
Microtubules Sister chromatids Mitosis
Meiosis
Meiosis
B. Centromere behavior in meiosis Paired sister kinetochores Microtubules
Chiasmata
Recombination 2× Homolog pairing
Arms of sisters separate
Paired sister chromatids
Paired homologs Anaphase I Chiasmata released Homologs separate
Metaphase I
Paired sisters Anaphase I
Sister centromeres separate
Metaphase II
Leptotene
Sisters segregate
Anaphase II
FIGURE 45.12 CHROMOSOMAL BEHAVIOR DURING MEIOSIS I AND II. During meiosis I, sister chromatids are tightly paired along their lengths, sister kinetochores are fused, and homologs are held together at the metaphase plate by chiasmata. During anaphase I, loss of cohesion between the arms of sister chromatids releases the chiasmata and allows homologous chromosomes to segregate to opposite spindle poles. During metaphase of meiosis II, sister chromatids are held together only at their centromeres. Release of centromeric cohesion at meiosis II allows the sister chromatids to segregate to opposite spindle poles.
Rec8 and Rad21L proteins replace Scc1 (see Figs. 8.18 and 44.16). After premeiotic DNA replication, the meiotic cohesion complex keeps sister chromatids together all along the arms. The cohesin complex plus synaptonemal complex proteins SYCP3 and SYCP2 are required for assembly of the dense axial elements that extend along the length of the chromosome during synapsis. In mitosis, cohesion is released between sister chromatid arms during prometaphase, but in meiosis I it is retained distal to chiasmata until the onset of anaphase (Figs. 45.7 and 45.12), when Rec8 and Rad21L along the chromosome arms are cleaved by a protease called separase (see Fig. 44.16). Separation of sister chromatid arms allows the chiasmata to resolve (untangle), and the homologous chromosomes segregate to opposite spindle poles. In the meantime, the Rec8 and Rad21L at centromeres are protected from cleavage and continue to hold the
sister chromatid centromeres tightly paired until anaphase of meiosis II. This protection requires a class of proteins called Shugoshins (from the Japanese, meaning “guardian spirit”), which recruit the protein phosphatase 2A (PP2A). Rec8 must be phosphorylated for separase to cleave it efficiently, so the localized phosphatase can block the cleavage reaction. Cleavage of centromeric Rec8 releases sister chromatid cohesion at the onset of anaphase II similar to the release of cohesion during mitosis (see Fig. 44.16).
Behavior of the Sex Chromosomes in Meiosis Of the 46 human chromosomes, two sex chromosomes carry genes that define the sex of the individual. The other 22 pairs of chromosomes are called autosomes. If genetic recombination is required to stabilize homologous chromosomes at the metaphase plate in
792
SECTION X n Cell Cycle
A
C X
2
Y
9
B
Synapsed pseudoautosomal region Y X Condensed unpaired portion of X and Y chromosomes FIGURE 45.13 SEX CHROMOSOMES OF A CHINESE HAMSTER AT PACHYTENE. A–B, The X and Y chromosomes are synapsed at the pseudoautosomal region. Elsewhere, the unpaired chromatin adopts a highly condensed morphology. C, In the same preparation, autosomes are completely synapsed and show a lesser degree of condensation. (From Dresser ME, Moses MJ. Synaptonemal complex karyotyping in spermatocytes of the Chinese hamster (Cricetulus griseus). IV. Light and electron microscopy of synapsis and nucleolar development by silver staining. Chromosoma. 1980;76: 1–22.)
meiosis I, how is this accomplished for the X and Y chromosomes? The answer in most mammals is that the X and Y chromosomes have a short region of homologous sequence (approximately 2.6 million base pairs in humans) that does pair and undergo genetic recombination during meiosis. This pseudoautosomal region must undergo genetic recombination in every meiosis I cell for the X and Y chromosomes to be partitioned correctly. Thus, the X and Y chromosomes act like short homologous chromosomes with large regions of unrelated DNA attached (Fig. 45.13). Unpaired regions of the X and Y chromosomes acquire a distinct chromatin structure during late pachytene.
Cell-Cycle Regulation of Meiotic Events Meiosis employs the full set of functions that regulate the division of somatic cells (see Chapters 40 to 43). However, the peculiarities of the meiotic cell cycle require additional regulation. One major difference from somatic cells is that meiotic chromosomes must undergo recombination and form chiasmata to segregate properly at the first meiotic division. Like somatic cells, meiotic cells have a DNA damage response that arrests cell-cycle progression in the presence of DNA breaks (see Fig. 43.11). In addition, they have a “crossover assurance” checkpoint that can detect the presence of stalled or abnormal recombination intermediates. Such intermediates accumulate if there are problems with the core
recombination enzymes or if the assembly of the synaptonemal complex (required for the completion of recombination) is defective. When such problems are detected, cells arrest in meiotic prophase I. In yeast, this has been called the pachytene checkpoint, as cells arrest late in meiotic prophase with nuclear morphology reminiscent of the pachytene stage. Apoptosis eliminates mammalian germ cells that arrest owing to defects in recombination.
Suppression of DNA Replication Between Meiosis I and Meiosis II Meiosis is unique in that it involves two M phases with no intervening S phase. On exit from meiosis I, Cdk1 kinase is reactivated immediately. This blocks assembly of prereplication complexes (see Fig. 42.8), thereby blocking DNA replication. At least two pathways contribute to reactivation of Cdk1. The first involves downregulation of translation of Wee1 protein kinase in meiosis. Wee1 is a mitotic inhibitor (see Fig. 43.3) that inactivates Cdk1 by phosphorylation at Tyr15. The absence of Wee1 in meiosis I was first observed in Xenopus laevis but this seems to be a universally conserved way of reactivating Cdk1 without an S phase. Ectopic expression of Wee1 in mature X. laevis oocytes prevents reactivation of Cdk1 immediately after the meiosis I division. As a result, the oocytes reenter interphase and replicate their DNA. Meiotic cells also express a specialized isoform of Cdc25, the phosphatase that counteracts Wee1 (see Fig. 43.1).
Metaphase II Arrest and the Mitogen-Activated Protein Kinase Pathway Following their activation and release from the ovary (ovulation), oocytes of many vertebrates arrest in metaphase II of meiosis until they are fertilized. The activity that is responsible for this arrest was discovered in X. laevis eggs arrested in metaphase of meiosis II and is called cytostatic factor (CSF). Injection of cytoplasm containing CSF into one blastomere of a two-cell frog embryo blocks the next cell cycle at metaphase, just like the egg (Fig. 45.14). Therefore, CSF can even block somatic cells indefinitely in mitotic metaphase. CSF activity appears in meiosis II and disappears after fertilization. One active component of CSF is the X. laevis homolog of a well-known viral oncogene, v-mos, the transforming gene of the Moloney murine sarcoma virus, which causes solid tumors in mice. The v-mos gene is a mutated form of the cellular c-mos gene. Vertebrates express c-mos (a mitogen-activated protein [MAP] kinase kinase kinase; see Fig. 27.5) exclusively in oocytes and eggs. Injection of either v-Mos or c-Mos proteins into dividing
CHAPTER 45 n Meiosis
A. Two-cell embryo
793
modification creates a binding site for polo kinase, which then also phosphorylates Emi2. Emi2 phosphorylated by polo kinase is recognized by SCFβTrCP, which ubiquitylates it, marking it for destruction. This results in activation of the APC/C (see Fig. 40.15), termination of the CSF metaphase arrest, and completion of meiosis II.
B
Inject egg cytosol, with activated c-Mos, into one blastomere
Timing of Meiosis in Humans C c-Mos removed by antibody before injection
D. The MAP kinase cascade that arrests eggs in metaphase of meiosis II: c-Mos
MEK
MAPK
p90Rsk
[Emi2]
CSF
FIGURE 45.14 ROLE OF C-MOS IN THE CYTOSTATIC FACTOR PATHWAY. A, Diagram of the experiment that identified c-Mos as an essential component of cytostatic factor (CSF) required for arrest of eggs in meiotic metaphase. One blastomere of an Xenopus laevis embryo at the two-cell stage was injected with cytoplasm from a metaphase-arrested egg containing CSF activity. B, This blastomere (right half of the embryo) remained blocked in metaphase while the left blastomere divided many times. C, The same experiment was performed, but prior to injection, the c-Mos was removed from the egg cytoplasm by absorption with a specific antibody. Both the injected and uninjected blastomeres continued to divide normally. D, The mitogen-activated protein (MAP) kinase (MAPK) pathway leading to metaphase II arrest in vertebrate eggs. (B–C, Micrographs courtesy George Vande Woude, NCI, Frederick, MD. Modified from Sagata N, Watanabe N, Vande Woude GF, et al. The c-Mos proto-oncogene product is a cytostatic factor responsible for meiotic arrest in vertebrate eggs. Nature. 1989;342:512–518.)
blastomeres of early frog embryos arrests the cells at metaphase (Fig. 45.14). These experiments led to a proposal that c-Mos is CSF. CSF arrest requires the MAP kinase (MAPK) signal transduction pathway (see Fig. 27.5). Mos activates the pathway by phosphorylating mitogen activated protein/ extracellular signal-related kinase kinase (MEK), which then activates MAPK. MAPK then activates a downstream kinase called p90Rsk (Fig. 45.14D). Introduction of constitutively active c-Mos or p90Rsk into X. laevis eggs induces metaphase arrest like CSF. However, this is not the whole story, because metaphase arrest is maintained in extracts depleted of p90Rsk. Thus the pathway must include at least one unidentified step beyond p90Rsk. The extra component of CSF is an anaphase-promoting complex/cyclosome (APC/C) inhibitor called Emi2. A burst of cytoplasmic Ca2+ released at fertilization (see Fig. 26.15) activates Ca2+/calmodulin-dependent protein kinase II (CaMKII), which phosphorylates Emi2. This
The fate of cells undergoing meiosis, as well as the timing of meiotic events, differs significantly between human males and females. Males produce approximately 100 million sperm a day in a process called spermatogenesis. This process continues throughout adult life. Spermatogenesis starts with the division of stem cells called spermatogonia and involves eight divisions prior to meiosis. These divisions are unusual in that cytokinesis is incomplete, and the cells remain connected by intercellular bridges. The process could produce up to 256 cells, but usually some cells die and others fail to divide, so a more typical number is around 200 cells arising from the initial stem cell division. When these cells pass through meiosis (they are then referred to as spermatocytes) the final result is approximately 800 postmeiotic spermatids. Spermatids then undergo a complex program of differentiation, resulting in the production of highly specialized spermatozoa. The entire process of spermatogenesis takes approximately 64 days, the bulk of which is spent in meiosis I. Approximately 16 days are spent in pachytene, the longest stage of the meiosis I prophase. In contrast, only about 8 hours are spent in meiosis II. By the twentieth week of fetal life each ovary of a human female contains approximately 3 × 106 primordial follicles, each with an oocyte arrested in the diplotene stage of meiosis. This lengthy arrested stage is referred to as dictyate. Thereafter, arrested primordial follicles are recruited continuously to mature into growing primary follicles. However, successful follicular growth depends on follicle-stimulating hormone (FSH), so before puberty—when the hypothalamus-pituitary-ovarian axis matures—all activated oocytes undergo programmed cell death and degenerate in a process known as atresia. At birth, the ovary contains approximately 1,000,000 germ cells of which approximately 300,000 to 400,000 remain at puberty. Following puberty and in response to high levels of FSH each month, a cohort of follicles with their dictyate-arrested oocytes is activated to complete meiosis and grow. Only one of these activated oocytes matures fully and is shed from the ovary in response to a surge of luteinizing hormone. The other follicles degenerate by atresia. As the successful oocyte is shed from the ovary, it completes meiosis I and is arrested at metaphase of meiosis II by CSF. It remains arrested at this stage until fertilization occurs. By the time of menopause
794
SECTION X n Cell Cycle
First polar body
Metaphase II oocyte
50 µm FIGURE 45.15 POLAR BODY FORMATION. Asymmetric cleavage during mouse oogenesis produces a large oocyte and small polar body. (Modified from Jiao ZX, Xu M, Woodruff TK. Age-associated alteration of oocyte-specific gene expression in polar bodies: potential markers of oocyte competence. Fertil Steril. 2012;98:480–486.)
the ovarian reserve is almost depleted, leaving approximately 1000 primordial follicles in the ovary. In human females, meiosis produces only one mature egg. Both meiotic cell divisions are asymmetrical, producing one large and one very small and short-lived cell. The small cells are called polar bodies (Fig. 45.15).
Meiotic Defects and Human Disease Abnormalities in meiosis are surprisingly common but are not widely observed in human populations because their consequences are extremely severe. In fact, meiotic abnormalities are a leading cause of fetal death, particularly during the first trimester of pregnancy in humans. The two major causes of problems are chromosome nondisjunction during the meiotic divisions and the generation of unbalanced chromosomal rearrangements via faulty recombination. When chromosomes fail to segregate properly in one or both meiotic divisions (nondisjunction), the daughter cells lack the normal haploid complement of chromosomes. Embryos that have gained an entire set of chromosomes are referred to as polyploid. In human embryos, polyploidy is a common type of chromosomal abnormality. It is estimated that 1% to 3% of all conceptions are triploid (69 chromosomes; 23 from one parent and 46 from the other parent). Two-thirds of these arise from two sperm fertilizing one egg (nothing wrong with meiosis there). In other cases, they come from a diploid gamete, the result of a defective meiotic segregation. Very few triploid embryos survive to birth. Most chromosomal abnormalities in human embryos result from the loss or gain of one or more chromosomes during meiosis. This condition is called aneuploidy. Most zygotes that arise from aneuploid gametes die
during fetal development. (Any fetal death is a spontaneous abortion, commonly called a miscarriage.) It is now thought that at least 30% of all conceptions result in spontaneous abortions. Furthermore, more than 60% of those spontaneous abortions are aneuploid. These figures probably underestimate the frequency of meiotic abnormalities. Many spontaneous abortions occur very early during pregnancy and many are never detected at all. Few fetuses lost in the first 4 to 6 weeks of gestation are tested in a laboratory, so their karyotypes are unknown. Meiotic errors involving certain autosomes can produce fetuses that survive to birth. Females with three or more copies of the gene-rich X chromosome survive, because all but one X chromosome are inactivated (silenced) in somatic cells (see Fig. 8.6). Rare individuals with three copies of chromosome 13 or chromosome18 survive to birth; but those who do typically die shortly thereafter. The exception is individuals trisomic for chromosome 21 (a condition that is commonly known as Down syndrome). These individuals have intellectual disability as well as other characteristic phenotypic features, including decreased life expectancy. Why do individuals with Down syndrome survive whereas others affected by aneuploidy do not? Perhaps the very small number (233) of coding genes on chromosome 21 includes none whose dosage is critical for survival. The frequency of certain types of aneuploidy, such as trisomy for chromosome 21, increases with the ages of the mother and father. Only 0.04% of children of 20-yearold mothers old have trisomy 21. This number rises dramatically with maternal age; nearly 5% of the conceptions in mothers 45 years old have trisomy 21 (Fig. 45.16). This maternal age effect is a leading cause of human genetic disease. Some believe that during the many years of arrest of oocytes in meiosis I dictyate, chiasmata joining homologous chromosomes gradually dissociate. A mechanism to explain this might be the progressive loss of cohesion between sister chromatids as the mother ages. There appears to be no mechanism to replace cohesin complexes that are gradually lost in dictyate oocytes. Mice with a mutation in a key subunit of the cohesin complex (Fig. 45.7; see also Fig. 8.18) exhibit a pattern of chromosome nondisjunction with increasing maternal age that looks much like that seen in aging human mothers. Another potential source of errors lies in the mechanism of spindle assembly in oocytes, which does not involve centrosomes and appears to be more prone to errors than in somatic cells (see Chapter 44). Not all cases of human aneuploidy originate from the mother. One of the most common aneuploidies, 45,X (see Table 45.1 for an explanation of nomenclature) involves the loss of the paternal X or Y chromosome 70% to 80% of the time. This aneuploidy accounts for nearly 10% of spontaneous abortions. In addition, about 7% of
CHAPTER 45 n Meiosis
795
TABLE 45.1 Aneuploidies Involving the Sex Chromosomes in Newborn Humans Karyotype
Frequency
Sex
Comments
47,XXY*
1/1000
M
Klinefelter syndrome. Increased height, sterile, a proportion may have some learning difficulties.
47,XYY
1/1000
M
Increased height, generally fertile, typically with chromosomally normal offspring. A proportion may have some learning difficulties.
Other X or Y aneuploidy
1/1350
47,XXX
1/900
F
Increased height, generally fertile, typically with chromosomally normal offspring. A proportion have serious learning difficulties.
45,X
1/4000
F
Turner syndrome. Reduced height, infertile, normal intelligence. Of 45,X embryos, 99% terminate as spontaneous abortions.
Other X or Y aneuploidy
1/2700
Total: 1 in 360 male births
Total: 1 in 580 female births *This number gives the total number of chromosomes, followed by the complement of sex chromosomes. Modified from Nussbaum RL, McInnes RR, Willard HF. Genetics in Medicine. 6th ed. Philadelphia, PA: WB Saunders; 2001:150.
[Table 45.1]) reveal that spontaneous abortion is a highly efficient protective mechanism for the elimination of chromosomal imbalances that arise from errors in meiosis.
Chromosome number 3.0
ACKNOWLEDGMENTS
% Trisomies
16
We thank Abby Dernburg, Jim Haber, Scott Hawley, Amy MacQueen, Adele Marston, and Alberto Pendas for their suggestions on revisions to this chapter.
2.0
21
SELECTED READINGS
1.0
18
0