Molecular and Cellular Aspects of the Serpinopathies and Disorders in Serpin Activity

Molecular and Cellular Aspects of the Seerpinopathies and Disorders in Serpin Activity This page intentionally left ...

Author: Gary A. Silverman | Gary A. Silverman; David A. Lomas

80 downloads 679 Views 8MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Molecular and Cellular Aspects of the Seerpinopathies and

Disorders in Serpin Activity

This page intentionally left blank

Molecular and Cellular Aspects of the Serpinopathies and

Disorders in Serpin Activity

Editors

Gary A. Silverman University of Pittshburgh School of Medicine USA

David A. Lomas University of Cambridge, UK

World Scientific NEW JERSEY . LONDON . SINGAPORE . BEIJING . SHANGHAI . HONG KONG . TAIPEI . CHENNAI

Published by World Scientific Publishing Co. Pte. Ltd. 5 Toh Tuck Link, Singapore 596224 USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601 UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.

MOLECULAR AND CELLULAR ASPECTS OF THE SERPINOPATHIES AND DISORDERS IN SERPIN ACTIVITY Copyright © 2007 by World Scientific Publishing Co. Pte. Ltd. All rights reserved. This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher.

For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA. In this case permission to photocopy is not required from the publisher.

ISBN-13 978-981-256-963-9 ISBN-10 981-256-963-4

Typeset by Stallion Press Email: [email protected]

Printed in Singapore.

JQuek - Molecular and Cellular.pmd

1

3/14/2007, 4:17 PM

Jan. 10, 2007

8:57

SPI-B429

Molecular and Cellular Aspects of the Serpinopathies and Disorders in Serpin Activity

Contents

Editors

ix

Contributors

xi

Serpinopathies Foreword

xxv

Serpinopathies Preface

xxvii

SECTION I

OVERVIEW OF THE SERPIN SUPERFAMILY

Chapter 1

Evolution and Classification of the Serpin Superfamily James A. Irving, Lisa D. Cabrita, Dion Kaiserman, Margaret M. Worrall and James C. Whisstock

1

Chapter 2

Serpin Conformations Mary C. Pearce, Robert N. Pike, Arthur M. Lesk and Stephen P. Bottomley

35

Chapter 3

Mechanisms of Serpin Inhibition Peter G. W. Gettins

67

SECTION II LESSONS FROM MODEL ORGANISMS Chapter 4

Mouse Serpins and Transgenic Studies David J. Askew, Paul Coughlin and Phillip I. Bird

101

Chapter 5

Serpins in Prokaryotes Qingwei Zhang, Ruby Law, Ashley M. Buckle, Lisa Cabrita, Sheena McGowan, James A. Irving, Noel G. Faux, Arthur M. Lesk, Stephen P. Bottomley and James C. Whisstock

131

v

fm

Jan. 10, 2007

8:57

SPI-B429

vi


Contents

Chapter 6

New Lessons from Poxvirus Serpins Peter C. Turner, Priscilla F. McAuliffe, Amy L. MacNeill, and Richard W. Moyer

Chapter 7

The Intracellular Serpins of Caenorhabditis elegans Stephen C. Pak, Yuko. S. Askew and Cliff J. Luke

Chapter 8

Chapter 9

Chapter 10

Drosophila Serpins: Regulatory Cascades in Innate Immunity and Morphogenesis David Gubb, Andrew Robertson, Tim Dafforn, Laurent Troxler and Jean-Marc Reichhart Serpins in a Lepidopteran Insect, Manduca sexta Michael R. Kanost MENT, a Chromatin-Associated Serpin from Avian Blood Cells: Structure and Function Sergei A. Grigoryev, Yaroslava A. Bulynko, Robert N. Pike and James C. Whisstock

163

195

207

229

243

Chapter 11

Uterine Serpins Peter J. Hansen, Saban Tekin and Maria B. Padua

261

Chapter 12

Plant Serpins Jørn Hejgaard and Thomas H. Roberts

279

SECTION III SERPIN PHYSIOLOGY AND PATHOPHYSIOLOGY Chapter 13

Chapter 14

Serpins, Apoptosis and Other Aspects of Cell Death Fiona L. Scott Serpins in Malignancy: Tumor Cell Invasion, Motility and Angiogenesis Philip A. Pemberton, Christian Schem, Nicolai Maass and Ming Zhang

301

337

fm

Jan. 10, 2007

8:57

SPI-B429


Contents

vii

Chapter 15

Serpins in Malignancy: Prognostic Indicators Hiroshi Kato, Yuko S. Askew, Shugo Nawata, Akihiro Murakami, Yoshinori Suminami and Gary Clayman

363

Chapter 16

SERPINB7 (Megsin) Nephropathy Toshio Miyata

383

Chapter 17

Serpin-Protein and Receptor Interactions Toni M. Antalis and Dudley K. Strickland

393

Chapter 18

Serpins and the Metabolic Syndrome Jun Wada

411

Chapter 19

Serpin Interactions with Bacterial Peptidases Jan Potempa, Tomasz Kantyka and Daniel Nelson

425

SECTION IV CONGENITAL DEFICIENCIES AND SERPINOPATHIES Chapter 20

The Serpinopathies and Respiratory Disease David A. Lomas, Didier Belorgey, Elena Miranda, Meera Mallya, Peter Hägglöf, Lynda K. Sharp, Russell L. Phillips, Richard Page, Mark J. Davies and Damian C. Crowther

445

Chapter 21

Liver Disease in α 1 -Antitrypsin Deficiency David H. Perlmutter

483

Chapter 22

Regulation of Hemostasis by Heparin-Binding Serpins Frank C. Church, Robert N. Pike, Douglas M. Tollefsen, Ashley M. Buckle, Angelina V. Ciaccia and Steven T. Olson

Chapter 23

C1 Inhibitor: Structure, Function, Biologic Activity and Angioedema Philip A. Patston and Alvin E. Davis III

509

555

fm

Jan. 10, 2007

8:57

SPI-B429

viii


Contents

Chapter 24

Neuroserpin in Neurological Disease Manuel Yepes and Daniel A. Lawrence

593

Chapter 25

Serpins and Alzheimer’s Disease H. Tonie Wright and Sabina Janciauskiene

619

Index

635

fm

Jan. 10, 2007

8:57

SPI-B429


EDITORS Gray A. Silverman Professor of Pediatrics University of Pittsburgh School of Medicine Chief of Neonatology and Development Biology Chief, UPMC Newborn Medicine Program Magee-Womens Hospital and Children’s Hospital of Pittsburgh 300 Halket St. Pittsburgh, PA 15213, USA David A. Lomas Department of Medicine University of Cambridge Cambridge Institute for Medical Research Wellcome Trust/MRC Building Hills Road, Cambridge, CB2 2XY, UK

ix

fm

Jan. 10, 2007 8:57 SPI-B429



fm

Jan. 10, 2007

8:57

SPI-B429


CONTRIBUTORS Toni M. Antalis Center for Vascular and Inflammatory Diseases The Departments of Physiology and Surgery University of Maryland School of Medicine BioPark One, 800 West Baltimore Street Baltimore, MD 21201, USA David J. Askew UPMC Newborn Medicine Program Department of Pediatrics University of Pittsburgh School of Medicine Children’s Hospital of Pittsburgh The Magee-Womens Research Institute 204 Craft Avenue Pittsburgh, PA 15213, USA Yuko S. Askew UPMC Newborn Medicine Program Department of Pediatrics University of Pittsburgh School of Medicine Children’s Hospital of Pittsburgh The Magee-Womens Research Institute 204 Craft Avenue Pittsburgh, PA 15213, USA Didier Belorgey Department of Medicine University of Cambridge Cambridge Institute for Medical Research Wellcome Trust/MRC Building Hills Road, Cambridge, CB2 2XY, UK xi

fm

Jan. 10, 2007

8:57

SPI-B429

xii


Contributors

Phillip I. Bird Department of Biochemistry and Molecular Biology Monash University, Clayton Campus Melbourne VIC 3800, Australia Stephen P. Bottomley The Protein Crystallography Unit Monash Centre for Synchrotron Science Department of Biochemistry and Molecular Biology Victorian Bioinformatics Consortium ARC Centre for Structural and Functional Microbial Genomics Monash University, Clayton Campus Melbourne VIC 3800, Australia Ashley M. Buckle The Protein Crystallography Unit Monash Centre for Synchrotron Science Department of Biochemistry and Molecular Biology Co-operative Research Centres for Vaccine Technology and Oral Health Sciences Victorian Bioinformatics Consortium ARC Centre for Structural and Functional Microbial Genomics Monash University, Clayton Campus Melbourne VIC 3800, Australia Yaroslava A. Bulynko Department of Biochemistry and Molecular Biology Pennsylvania State University College of Medicine H171, Milton S. Hershey Medical Center PO Box 850, 500 University Drive Hershey, PA 17033, USA Lisa D. Cabrita Department of Chemistry University of Cambridge Lensfield Road, Cambridge, CB2 1EW, UK

fm

Jan. 10, 2007

8:57

SPI-B429


Contributors

Frank C. Church Departments of Pathology and Laboratory Medicine and Pharmacology Carolina Cardiovascular Biology Center Division of Hematology-Oncology/Medicine Campus Box 7035 The University of North Carolina at Chapel Hill, School of Medicine Chapel Hill, NC 27599-7035, USA Angelina V. Ciaccia Departments of Pathology and Laboratory Medicine and Pharmacology Carolina Cardiovascular Biology Center Division of Hematology-Oncology/Medicine Campus Box 7035 The University of North Carolina at Chapel Hill School of Medicine Chapel Hill, NC 27599-7035, USA Gary Clayman The Departments of Head and Neck Surgery and Cancer Biology The University of Texas MD Anderson Cancer Center Houston, TX 77030, USA Paul Coughlin Australian Centre for Blood Diseases Monash University, Clayton Campus Melbourne VIC 3128, Australia Damian C. Crowther Department of Medicine University of Cambridge Cambridge Institute for Medical Research Wellcome Trust/MRC Building Hills Road, Cambridge, CB2 2XY, UK Tim Dafforn Department of Biosciences University of Birmingham Birmingham, B15 2TT, UK

xiii

fm

Jan. 10, 2007

8:57

SPI-B429

xiv


Contributors

Mark J. Davies Department of Medicine University of Cambridge Cambridge Institute for Medical Research Wellcome Trust/MRC Building Hills Road, Cambridge, CB2 2XY, UK Alvin E. Davis III CBR Institute for Biomedical Research Harvard Medical School 800 Huntington Avenue Boston, MA 02115, USA Noel G. Faux The Protein Crystallography Unit Monash Centre for Synchrotron Science Department of Biochemistry and Molecular Biology Victorian Bioinformatics Consortium ARC Centre for Structural and Functional Microbial Genomics Monash University, Clayton Campus Melbourne VIC 3800, Australia Peter G. W. Gettins Department of Biochemistry and Molecular Genetics, M/C 669 University of Illinois at Chicago 900 South Ashland Avenue Chicago, IL 60607, USA Sergei A. Grigoryev Department of Biochemistry and Molecular Biology Pennsylvania State University College of Medicine H171, Milton S. Hershey Medical Center PO Box 850, 500 University Drive Hershey, PA 17033, USA David Gubb CIC bioGUNE, Parque tecnológico de Bizkaia Ed. 801-A, Derio, 48160, Spain

fm

Jan. 10, 2007

8:57

SPI-B429


Contributors

Peter Hägglöf Department of Medicine University of Cambridge Cambridge Institute for Medical Research Wellcome Trust/MRC Building Hills Road, Cambridge, CB2 2XY, UK Peter J. Hansen Department of Animal Sciences University of Florida Gainesville, FL 32611-0910, USA Jørn Hejgaard Biochemistry and Nutrition Group, Biocentrum Technical University of Denmark DK-2800 Lyngby, Denmark James A. Irving Division of Structural Biology Wellcome Trust Centre for Human Genetics Oxford University Roosevelt Drive, Headington, Oxford OX3 7BN, UK Sabina Janciauskiene Department of Medicine University Hospital of Malmo Lund University Wallendberg Lab Ing 46, Plan 2, MAS S-20502 Malmo, Sweden Dion Kaiserman Department of Biochemistry and Molecular Biology Monash University, Clayton Campus Melbourne VIC 3800, Australia Michael R. Kanost Department of Biochemistry and Arthropod Genomics Center Kansas State University Manhattan, KS 66506, USA

xv

fm

Jan. 10, 2007

8:57

SPI-B429


xvi

Contributors

Tomasz Kantyka Department of Microbiology, Faculty of Biotechnology Jagiellonian University Gronostajowa 7, 30-387 Krakow, Poland Hiroshi Kato Life Network Laboratory Ube Medical Creative Center 361-1 Majime Ooaza-Kogushi Ube 755-0067, Japan Ruby Law The Protein Crystallography Unit Monash Centre for Synchrotron Science Department of Biochemistry and Molecular Biology Victorian Bioinformatics Consortium ARC Centre for Structural and Functional Microbial Genomics Monash University, Clayton Campus Melbourne VIC 3800, Australia Daniel A. Lawrence Department of Internal Medicine University of Michigan School of Medicine 7301 MSRB III, 1150 W. Medical Center Drive Ann Arbor, MI, 48109-0644, USA Arthur M. Lesk Department of Biochemistry and Molecular Biology The Huck Institutes of the Life Sciences The Institute for Genomics, Proteomics, and Bioinformatics The Pennsylvania State University University Park, PA 16802, USA

fm

Jan. 10, 2007

8:57

SPI-B429


Contributors

Cliff J. Luke UPMC Newborn Medicine Program Department of Pediatrics University of Pittsburgh School of Medicine Children’s Hospital of Pittsburgh The Magee-Womens Research Institute 204 Craft Avenue Pittsburgh, PA 15213, USA Nicolai Maass Department of Gynecology and Obstetrics University of Schleswig Holstein, Campus Kiel Michaelisstr. 16, 24105 Kiel, Germany Amy L. MacNeill Department of Molecular Genetics and Microbiology University of Florida Gainesville, FL 32610, USA Meera Mallya Department of Medicine University of Cambridge Cambridge Institute for Medical Research Wellcome Trust/MRC Building Hills Road, Cambridge, CB2 2XY, UK Priscilla F. McAuliffe Department of Surgery, College of Medicine University of Florida Gainesville, FL 32610, USA Sheena McGowan The Protein Crystallography Unit Monash Centre for Synchrotron Science Department of Biochemistry and Molecular Biology Monash University, Clayton Campus Melbourne VIC 3800, Australia

xvii

fm

Jan. 10, 2007

8:57

SPI-B429

xviii


Contributors

Elena Miranda Department of Medicine University of Cambridge Cambridge Institute for Medical Research Wellcome Trust/MRC Building Hills Road, Cambridge, CB2 2XY, UK Toshio Miyata Institute of Medical Sciences and Department of Medicine Tokai University School of Medicine Kanagawa 259-1193, Japan Richard W. Moyer Department of Molecular Genetics and Microbiology University of Florida Gainesville, FL 32610, USA Akihiro Murakami Department of Reproduction, Development and Infection Yamaguchi University School of Medicine 1-1-1 Minamikogushi, Ube 755- 8505, Japan Shugo Nawata Department of Reproduction, Development and Infection Yamaguchi University School of Medicine 1-1-1 Minamikogushi, Ube 755- 8505, Japan Daniel Nelson Laboratory of Bacterial Pathogenesis and Immunology Rockefeller University 1230 York Avenue New York, NY 10021, USA Steven T. Olson Center for Molecular Biology of Oral Diseases University of Illinois at Chicago Rm 530E, Dentistry (M/C 860), 801 S. Paulina St. Chicago, IL 60612, USA

fm

Jan. 10, 2007

8:57

SPI-B429


Contributors

Maria B. Padua Department of Animal Sciences University of Florida Gainesville, FL 32611-0910, USA Richard Page Department of Medicine University of Cambridge Cambridge Institute for Medical Research Wellcome Trust/MRC Building Hills Road, Cambridge, CB2 2XY, UK Stephen C. Pak UPMC Newborn Medicine Program Department of Pediatrics University of Pittsburgh School of Medicine Children’s Hospital of Pittsburgh The Magee-Womens Research Institute 204 Craft Avenue Pittsburgh, PA 15213, USA Philip A. Patston Department of Oral Medicine and Diagnostic Services Center for Molecular Biology of Oral Diseases College of Dentistry, University of Illinois at Chicago Chicago, IL 60612, USA Mary C. Pearce Department of Biochemistry and Molecular Biology Monash University, Clayton Campus Melbourne VIC 3800, Australia Philip A. Pemberton Arriva Pharmaceuticals Inc. 2430-B Mariner Square Loop Alameda, CA 94501, USA

xix

fm

Jan. 10, 2007

8:57

SPI-B429

xx


Contributors

David H. Perlmutter Department of Pediatrics, Cell Biology and Physiology University of Pittsburgh School of Medicine Children’s Hospital of Pittsburgh 3705 Fifth Avenue Pittsburgh, PA 15213, USA Russell L. Phillips Department of Medicine University of Cambridge Cambridge Institute for Medical Research Wellcome Trust/MRC Building Hills Road, Cambridge, CB2 2XY, UK Robert N. Pike Department of Biochemistry and Molecular Biology Co-operative Research Centres for Vaccine Technology and Oral Health Sciences The Victorian Bioinformatics Consortium ARC Centre for Structural and Functional Microbial Genomics Monash University, Clayton Campus Melbourne VIC 3800, Australia Jan Potempa Department of Microbiology, Faculty of Biotechnology Jagiellonian University Gronostajowa 7, 30-387 Krakow, Poland Department of Biochemistry and Molecular Biology University of Georgia Life Science Bldg., Athens, GA 30602, USA Jean-Marc Reichhart Institute de Biologie Moléculaire et Cellulaire UPR 9022 du CNRS, 15 rue Rene Descartes F67084 Strasbourg Cedex, France

fm

Jan. 10, 2007

8:57

SPI-B429


Contributors

Thomas H. Roberts Department of Chemistry and Biomolecular Sciences Macquarie University NSW 2109, Australia Andrew Robertson The Wellcome Trust Sanger Institute Wellcome Trust Genome Campus Hinxton Cambridge, CB10 1SA, UK Christian Schem Department of Gynecology and Obstetrics University of Schleswig Holstein, Campus Kiel Michaelisstr. 16, 24105 Kiel, Germany Fiona L. Scott Program in Apoptosis and Cell Death Research The Burnham Institute 10901 North Torrey Pines Rd La Jolla, CA 92037, USA Lynda K. Sharp Department of Medicine University of Cambridge Cambridge Institute for Medical Research Wellcome Trust/MRC Building Hills Road, Cambridge, CB2 2XY, UK Dudley K. Strickland Center for Vascular and Inflammatory Diseases The Departments of Physiology and Surgery University of Maryland School of Medicine BioPark One, 800 West Baltimore Street Baltimore, MD 21201, USA Yoshinori Suminami Division of Obstetrics and Gynecology Onoda Municipal Hospital 1863-1 Higashitakadomari, Onoda 756-0088, Japan

xxi

fm

Jan. 10, 2007

8:57

SPI-B429


xxii

Contributors

Saban Tekin Department of Biology University of Gaziosmanpa¸sa Tokat, Turkey Douglas M. Tollefsen Hematology Division, Department of Medicine Campus Box 8125 Washington University School of Medicine 660 South Euclid Ave. St. Louis, MO 63110, USA Laurent Troxler Institute de Biologie Moléculaire et Cellulaire UPR 9022 du CNRS, 15 rue Rene Descartes F67084 Strasbourg Cedex, France Peter C. Turner Department of Molecular Genetics and Microbiology University of Florida Gainesville, FL 32610, USA Jun Wada Department of Medicine and Clinical Science Okayama University Graduate School of Medicine 2-5-1, Shikata-cho, Okayama 700-8558, Japan James C. Whisstock The Protein Crystallography Unit Monash Centre for Synchrotron Science Department of Biochemistry and Molecular Biology Victorian Bioinformatics Consortium ARC Centre for Structural and Functional Microbial Genomics Monash University, Clayton Campus Melbourne VIC 3800, Australia Margaret M. Worrall UCD School of Biomolecular and Biomedical Science University College Dublin, Belfield, Dublin 4, Ireland

fm

Jan. 10, 2007

8:57

SPI-B429


Contributors

H. Tonie Wright Institute of Structural Biology and Drug Discovery Department of Biochemistry Virginia Commonwealth University 800 E. Leigh St., Suite 212 Richmond, VA 23219-1540, USA Manual Yepes Department of Neurology Center for Neurodegenerative Disease Emory University School of Medicine 615 Whitehead Biomedical Research Building, Room 505J Atlanta, GA 30322, USA Ming Zhang Department of Molecular and Cellular Biology Baylor College of Medicine Houston, TX 77030, USA Qingwei Zhang The Protein Crystallography Unit Monash Centre for Synchrotron Science Department of Biochemistry and Molecular Biology Victorian Bioinformatics Consortium ARC Centre for Structural and Functional Microbial Genomics Monash University, Clayton Campus Melbourne VIC 3800, Australia

xxiii

fm

Jan. 10, 2007 8:57 SPI-B429



fm

Jan. 10, 2007

8:57

SPI-B429


Serpinopathies Foreword Advances in medical knowledge have traditionally focused on archetypal examples, as with Sir William Osler’s dictum in the late nineteenth century, that “He who knows syphilis knows medicine.” Similarly, with the commencement of the molecular era of medicine in the 1960’s, hemoglobin became a paradigm for the way in which genetic variations in the structure of a protein could result in a range of diverse diseases. The power of this model resulted from the solving of the structure of hemoglobin, but what impressed me as a participant at that time was the realization that the structure not only provided a template for studying diseases of hemoglobin, but also for the study of the consequences of variations in the globin family as a whole. An opportunity to apply the same approach, using structurebased homologies, to diseases arising from mutations in other protein families, was initiated by the work of Carl-Bertil Laurell, in Malmo Sweden in 1963. His finding that a mutation commonly present in the plasma protein α1-antitrypsin was associated with lung degeneration, identified the function of this otherwise unknown and atypical protease inhibitor. The subsequent amino acid sequencing of α1-antitrypsin with my colleagues in New Zealand to determine the position of the common mutations in the molecule, also led to the recognition of its close homology with another plasma protease inhibitor, antithrombin in 1979. However, two swallows do not make a summer and the recognition that we were looking at a new protein superfamily came from the additional alignment with ovalbumin by Hunt and Dayhoff in 1980. Interest in this newly identified family was transformed by the determination of the structure of α1-antitrypsin by Huber and colleagues in Martinsried Germany in 1984. Their structure provided a template on which to align other members of the family and because the α1-antitrypsin crystallized in Martinsried was in a protease-cleaved form, the structure also revealed the ability of this family of protease inhibitors to undergo a xxv

fm

Jan. 10, 2007

8:57

SPI-B429

xxvi


Serpinopathies Foreword

profound conformational change. By this time, other members had been added to the family, including the plasma proteins such as antiplasmin, PAI-1, C1-inhibitor, angiotensinogen and the hormone-binding globulins. Each of these was in turn associated with diverse familial disorders and it was soon apparent that most of the underlying mutations were closely sited together in what became recognized as the hinges and mobile regions of the template molecule. In this new family, we had a paradigm to supplant that of hemoglobin, not just as a model of genetic disease, but also as a model of the way that changes in molecular shape and fold could lead to dysfunction and to what became known as conformational diseases. Yet, how could these new ideas and concepts be communicated on the basis of what was denoted by Hunt and Dayhoff as the ovalbumin-antithrombinIII-alpha 1-proteinase inhibitor superfamily? A simpler name was needed. My thoughts immediately turned to the globins and hence to the acronym for this novel family of serine protease inhibitors, the serpins. James Travis who had led studies on the functional and biochemical properties of α1antitrypsin readily agreed to join in proposing the new name, which was soon adopted by the wider field because of its convenience. The prime reason though for the choice of name, the serpins, was to encourage the study of the family as a whole and it has been rewardingly successful in this. The flowering of these studies is seen in this volume, which documents the ubiquitous distribution of the serpins and their diverse functions. It is no exaggeration that the conformational mechanism of the serpins has provided a template for adaptation that has been critical to the evolution of higher organisms. As the authors each show in this book, the clearest evidence of the way this adaptation has allowed tissue-specific functions and modulation comes from the diseases that resulted from mutations in the individual serpins. The diversity of these serpinopathies, as brought together in this volume, covers the whole range of molecular and conformational pathologies. This timely book establishes the basis for what could be a dictum for the 21st century, “He who knows the serpins knows molecular medicine.” Robin Carrell University of Cambridge (June 2006)

fm

Jan. 10, 2007

8:57

SPI-B429


Molecular and Cellular Aspects of the Serpinopathies and Disorders in Serpin Activity Gary A. Silverman and David A. Lomas

Preface In 1980, Lois Hunt and Margaret Dayhoff identified a new protein superfamily containing ovalbumin, antithrombin III, and α1 -proteinase inhibitor.1 Although provisionally named the ovalbumin–antithrombin superfamily, Robin Carrell and James Travis utilized the acronym serpins to describe this expanding superfamily, as the members were predominately serine proteinase inhibitors.2 Since the 1980s, the serpin field has virtually exploded in terms of family size and our knowledge of their roles in biological systems.3–6 Inspection of the nucleotide repositories provides evidence for over 2000 family members, including 37 in humans (http://www.ncbi.nlm. nih.gov/Genbank/index.html; http://www.ensembl.org/index.html). Serpin citations and structures deposited in PubMed (http://www.ncbi.nlm.nih. gov/entrez/query.fcgi) and the PDB (http://www.rcsb.org/pdb/) exceed 43,000 and 80, respectively. Once considered unique to metazoans and most Poxviridae, serpins have now been detected in all domains of life such as Archaea, Prokarya, and Eukarya. Strict conservation of the serpin fold throughout evolution attests to the critical relationship between the potential energy stored in the native structure and the dynamic inhibitory function that distinguishes serpins from the canonical peptidase inhibitors. In essence, serpins are finely tuned molecular machines that undergo a major conformational rearrangement xxvii

fm

Jan. 10, 2007

8:57

SPI-B429

xxviii


G. A. Silverman and D. A. Lomas

to irreversibly trap their target peptidase. For reasons that have yet to be fully appreciated, nature has sought to retain this more complex inhibitory machine relative to the simpler canonical inhibitors and to place serpins at critical checkpoints in a variety of intra- and extra-cellular proteolytic cascades. However, like the moving parts of any finely tuned machine, serpin function can be severely impaired by amino acid variants positioned within any one of the several structural elements that affect protein folding or overall motility. Indeed, similarly placed mutations in the serpin scaffold of different serpin genes have led to the emergence of a new class of conformational disorders, the serpinopathies.7–10 The major goal of this text is to summarize our recent understanding of how serpin mutations and amino acid variants can lead to significant human diseases such as α1 -antitrypsin deficiency, hereditary angioedema, familial presenile dementias, and thrombophilia. While some of the serpin disorders serve as prime examples of conformational disorders (serpinopathies), others are due to the more classical type 1 (decreased steady-state amounts of a functional inhibitor) or type 2 deficiencies (normal steady amounts of protein, but the inhibitor is dysfunctional), or reactive site loop “changein-function” mutations. While most serpins are irreversible inhibitors of serine and cysteine peptidases, several family members serve other functions whose normal and pathological roles in biological systems have yet to be fully elucidated. As a prelude to understanding the pathogenesis and the broad biological implication of serpin disorders, we first present an overview of the serpin superfamily with an emphasis on their distribution, evolution, structure, and inhibitory mechanism. Next, insight into the spectrum of biological systems regulated by serpin gene activity is provided by several prokaryotic and eukaryotic model organisms. The roles that serpins play in defined physiological process such as cell death and tumor invasion are explored in the third section. Finally, lessons from all of these studies are used to better understand the relationships between serpins in human health and disease. The editors are grateful to the contributing authors for providing their collective expertise in bringing this text to fruition. We are also grateful to our wonderful colleagues in the serpin and peptidase community who provided their thoughtful insights into many of the biological and biochemical issues discussed in these pages.

fm

Jan. 10, 2007

8:57

SPI-B429


Serpinopathies Preface

xxix

Acknowledgments The editors and contributing authors dedicate this text to Robin Carrell and James Travis, whose scientific insight and thoughtful mentoring have helped inspire the study of serpin biology.

References 1. Hunt LT and Dayhoff MO (1980) A surprising new protein superfamily containing ovalbumin, antithrombin-III, and α1-proteinase inhibitor. Biochem Biophys Res Commun 95(2):864–871. 2. Carrell R and Travis J (1985) α1-Antitrypsin and the serpins: Variation and countervariation. Trends Biochem Sci 10:20–24. 3. Gettins PG (2002) Serpin structure, mechanism, and function. Chem Rev 102(12):4751–4803. 4. Potempa J, Korzus E and Travis J (1994) The serpin superfamily of proteinase inhibitors: Structure, function, and regulation. J Biol Chem 269(23): 15957–15960. 5. Silverman GA, Bird PI, Carrell RW, et al. (2001) The serpins are an expanding superfamily of structurally similar but funtionally diverse proteins: Evolution, mechanism of inhibition, novel functions, and a revised nomenclature. J Biol Chem 276(36):33293–33296. 6. Silverman GA, Whisstock JC, Askew DJ, et al. (2004) Human clade B serpins (ov-serpins) belong to a cohort of evolutionarily-dispersed intracellular proteinase inhibitor clades that protect cells from promiscuous proteolysis. Cell Mol Life Sci 61:301–325. 7. Davis RL, Shrimpton AE, Holohan PD, et al. (1999) Familial dementia caused by polymerization of mutant neuroserpin. Nature 401(6751):376–379. 8. Lomas DA and Mahadeva R (2002) Alpha1-antitrypsin polymerization and the serpinopathies: Pathobiology and prospects for therapy. J Clin Invest 110(11):1585–1590. 9. Lomas DA and Carrell RW (2002). Serpinopathies and the conformational dementias. Nat Rev Genet 3(10):759–768. 10. Carrell RW and Lomas DA (2002) Alpha1-antitrypsin deficiency–a model for conformational diseases. N Engl J Med 346(1):45–53.

fm


Jan. 10, 2007

8:57

SPI-B429


1 Evolution and Classification of the Serpin Superfamily James A. Irving, Lisa D. Cabrita, Dion Kaiserman, Margaret M. Worrall and James C. Whisstock

1. Introduction In a pioneering study, the sequence similarity between three protein sequences (human antitrypsin, antithrombin, and chicken ovalbumin) established these proteins — two of them protease inhibitors — as a discrete family.1 The importance of this family was highlighted by the work over the preceding 17 years, examining the association of the two human proteins with hereditary disease.2,3 On the basis of their sequence similarity, and the fact that these, and related, proteins were predominantly characterized as serine protease inhibitors, the label of serpin was applied to this family of high-molecular weight protease inhibitors.4 With the discovery of the plant serpin protein, barley Z,5 serpin genes extended into the termini of two early diverging lineages of the eukarya: the higher plants and the amniota. As sequence data were obtained for many new serpin variants in an ever-increasing spectrum of species, the databases were populated with sequences of serpins in flowering plants (magnoliophyta), amniotes (amniota) and viruses,6,7 but failed to fill the genetic gulf between these terminal branches. Key gaps in the species tree of the serpins included simple plants such as the green algae (phylum chlorophyta), early diverging animal lineages such as hydra (phylum cnidaria), fungi, and bacteria and archaea.8 1

ch01

Jan. 10, 2007

8:57

SPI-B429

2


J. A. Irving et al.

More recently, the accumulation of large volumes of genome sequence data has led to the discovery of several serpins from unicellular organisms, such as thermopin from the moderate thermophilic bacterium Thermobifida fusca,9,10 and a serpin from the green alga Chlamydomonas reinhardtii.11 A targeted approach seeking to address the paucity of data directly has also resulted in the discovery of the jellyfish serpin, jellypin.12 Species coverage is inexorably expanding to enable a broader account of the family within an evolutionary context.9,11,13,14 However, the nature and timing of the evolutionary events that occurred at the inception of the serpin fold remain unclear. For example, the sparse distribution of prokaryotic serpins neither favors the possibility of an ancient, ancestral prokaryotic serpin, nor the possibility that serpins appeared after the eukaryote/prokaryote split and were exchanged between prokaryotes and eukaryotes by a lateral gene transfer event.9,11

2. The Challenges of Serpin Classification At the time of writing, a PSI-BLAST search of the non-redundant NCBI database identifies around 1600 full and partial serpin sequences and minor variants (Fig. 1). When the redundancy is minimized by removing sequences with greater than 95% identity, the large number remaining — 820 serpin genes in 189 species — is testament to the broad utility of the serpin fold in many biological contexts. With such a large collection of sequences, it is useful to use some form of classification to partition the dataset into smaller sized relational groups. By far, the most common strategy applied to protein families relies explicitly or implicitly, due to a Trinitarian relationship between sequence, structure and function on evolutionary relationships. Sequence domain databases such as Pfam15 and the “superfamily” level of classification in the structural database SCOP16 reflect evolutionary considerations. Evolutionary relationships are in general a preferable means of classifying proteins over functional similarity for several reasons: (i) Within the context of a chosen scoring system, sequence homology or structural similarity is an intrinsic, objective property of two proteins. (ii) Furthermore, an analysis of protein similarity is based on data whose state of completeness can be determined: it is usually known whether a sequence is complete and whether

ch01

Jan. 10, 2007

8:57

SPI-B429


Evolution and Classification of the Serpin Superfamily

3

Fig. 1. Growth in the number of serpin sequences, taken from the literature1,4,6–8 and from PSI-BLAST searches of the non-redundant database (http://www.ncbi.nlm.nih.gov/ BLAST). Inset graph shows the trend in the number of species in which serpin sequences are found, from the literature and PSI-BLAST searches of SwissPROT releases 11, 19, 24, 38 and 49.118 At 189, the number of species is considerably higher in the non-curated NCBI non-redundant database.

all potential paralogues can be accounted for (i.e. whether it comes from a fully sequenced genome). (iii) In contrast, the functional data collected of a protein will be much more subjective, and reflect factors as varied as how comprehensively it has been studied in the laboratory, the experimental protocols used, and how biological activity has been measured. Across a large dataset, knowledge of function will most likely be subject to marked inconsistency. (iv) Furthermore, the state of knowledge of a protein’s function is likely to change and thus the resulting classifications may quickly become redundant. Nevertheless, one common goal of protein classification is to enable the prediction of function for a group of proteins based on their association with one another.17 The basic unit describing a cluster of evolutionarily related sequences in a phylogenetic tree is known as a clade. There are many examples of clades coinciding with protein function. For example, the phylogeny of solute transporter proteins reflects subgroups of proteins related by mechanism, energy coupling and polarity, and has permitted the development of a detailed classification system.18 Similarly, the biological

ch01

Jan. 10, 2007

8:57

SPI-B429

4


J. A. Irving et al.

role of the members of the kinesin superfamily of microtubule-associated motor proteins was found to be reflected by their deduced evolutionary relationships.19 However, function is a multi-layered concept that includes characteristics such as the mechanism used, reactions catalyzed, binding partners, and localization within an organism.17 Most serpins are protease inhibitors.20 As a consequence, the function of interest usually concerns the specific protease inhibited, ligand bound, or the biological effect achieved. This detailed concept of “function” is poorly defined by the evolutionary relationships between serpins,8 with one exception: in general, the phylogenetic trees can identify direct orthologues as far back as the fish–bird/mammal split (in the order of roughly 450 MYA) where no significant gene expansion has occurred.8,14 One reason that evolutionary groupings fail to predict serpin function is the ease with which a functional transition is achieved. It is well known that the reactive center loop (RCL), the primary factor influencing a serpin’s protease target specificity, is subjected to the greatest adaptive evolutionary pressure in a serpin gene.21 The substitution of a single amino acid in the RCL can completely alter the protease target in vivo. For example, the variant antitrypsin Pittsburgh, a mutant in which the key specificitydetermining methionine residue of this serpin has been substituted with arginine, results in a hemorrhagic disorder22,23 due to the inappropriate inhibition of thrombin. A similarly striking change of activity is seen upon substitution of threonine-345 in the so-called “hinge region” of the RCL by arginine, i.e. the serpin can no longer act as an inhibitor.24 A change in the class of protease inhibited by a serpin is also achieved with only a few amino acid substitutions.25,26 Differences in structural morphology, such as the presence or absence of a secondary structural element, binding loop or active site, can assist in distinguishing between protein subgroups. For example, the variations in mechanism and function that distinguish the cysteine endopeptidase cathepsin L from the exopeptidase cathepsin B, such as pH stability and exopeptidase activity, are reflected in structural differences and readily identifiable sequence motifs.27,28 Considerably greater structural divergence is seen with the calcium-activated protease, calpain.29 These differences coincide with an evolutionary partition between the classes of protease early in the evolution of the eukaryotes.30,31

ch01

Jan. 10, 2007

8:57

SPI-B429



5

In the case of serpins, there are few examples of structural characteristics that have become fixed within sub-families and permit differentiation between sub-families. Figure 2 shows a superposition of serpin structures spanning prokaryotes and eukaryotes with a diverse range of functions, which reveals a relatively immutable core fold that includes most secondary structural elements, surprisingly little variation in length, and a lack of structural element insertion within loop regions. In spite of sequence identities ranging from 25 to almost 90%, the three core sheets and 8-9 alpha

Fig. 2. Stereo figures showing the superposed structures of (A) cleaved serpins and (B) uncleaved serpins, as listed in Table 2. Figure prepared using Molscript.119

ch01

Jan. 10, 2007

8:57

SPI-B429

6


J. A. Irving et al.

helices are present and they superimpose reasonably well. The notable exceptions are the reactive centre loop which can vary in length, conformation and flexibility; the region between the C- and D-helices usually known as the “C–D loop”; N-terminal extensions for some serpins, such as heparin cofactor II;32 and C-terminal extensions, as seen with the thermophilic serpin thermopin.10 Perhaps, the most pronounced correlation between these structural features and evolutionary grouping is found with the ov-serpin subfamily, whose members generally possess a C–D loop, lack N- and C-terminal extensions to the core domain and lack signal sequences.33 Of the structures solved to date, a relatively short C–D loop is only visible in the case of ovalbumin.34 The implication of this degree of structural conservation is that amino acid sequence-based approaches provide the greatest amount of discriminatory information in determining relationships between genes in the serpin superfamily. However, in some cases, the gene structure, defined by the position and phasing of intron/exon boundaries, can discern relationships that are not resolved at an appropriate level of statistical significance by protein sequence alone.14,33,35,36 This has proven to be the case for the mammalian serpins: six serpin clades whose associations were unreliable using a variety of sequence-based phylogenetic approaches8,37 could be reduced to three larger clades based on common gene structures.36 The appearance of serpin genes at the same chromosomal locus has also been seen as evidence of a recent common ancestor.13,14,38 However, as a substantial number of serpin sequences lack information on gene structure and chromosomal position, these characteristics cannot form the basis of a universal serpin classification system in themselves. Hence, the classification convention that was adopted for the serpins is based on evolutionary relationships deduced from their amino acid sequences.39

3. The Phylogeny of the Serpin Superfamily In approaching a phylogenetic reconstruction of a protein family, a choice must be made between a number of approaches that: (i) use amino acid or nucleotide sequence; (ii) treat each sequence as a whole (sequence-based) or each amino acid/nucleotide individually (character-based); (iii) make different assumptions about how readily one amino acid type is substituted

ch01

Jan. 10, 2007

8:57

SPI-B429



7

with another, and whether this differs depending on the position in the sequence; (iv) measure the difference between sequences or amino acids/nucleotides differently; (v) produce trees according to different criteria, e.g. deriving a tree that achieves a minimum overall evolutionary score; and (vi) involve very different degrees of computation time.40 Diverse approaches have been applied to the serpin superfamily. A character-based method that attempts to derive the tree with the least number of “evolutionary steps” between branches and nodes, the “maximum parsimony” approach,41 has been applied to nucleotide data of the whole family7 and to protein data within subfamilies;8 the character-based “maximum likelihood” approach,42 which attempts to find the branching arrangement that produces the most likely tree, has also been used.37 The application of the “neighbor-joining” method, based on computed evolutionary differences between sequences, has been assessed as being in best agreement with other data such as serpin gene structure37 and co-localization at chromosomal loci.8 Nevertheless, the statistically significant phylogenetic groups determined by the various methods are compatible with one another, and differ primarily in the proportion of sequences that can be assigned to those groups. There are 35 known functional human serpin genes. A summary evolutionary tree depicting the established phylogenetic relationships between subfamilies of serpin genes containing at least one human member are shown in Fig. 3; the human sequences are listed in Table 1. The tree is based on a “neighbor-joining” approach,43 with the non-parametric bootstrap technique44 used to eliminate relationships with poor support; consequently, several branches radiate from the base of the tree because their relationships are uncertain.8 Where evidence of a common gene structure36 or chromosomal position13,14 suggests the order of sequence divergence, these are also shown in Fig. 4. As the base of the tree indicates the lowest point at which groupings can be reliably determined, it does not reflect a fixed point in time. However, clades containing human serpin genes notably are restricted to sequences from vertebrate organisms; furthermore, non-vertebrate clades (which are not shown in this tree) lack vertebrate sequences. Hence, the deepest point depicted by the base of the tree post-dates the appearance of the vertebrate lineage, estimated to be approximately 800 MYA.45,46

ch01

Jan. 10, 2007

8:57

SPI-B429

8


J. A. Irving et al.

Fig. 3. A phylogenetic tree illustrating the relationship between the 35 full-length human serpins. Only relationships strongly supported by the data are shown. Black hexagons indicate clades with 100% support in a consensus neighbor-joining tree and ovals indicate subgroups within clades with >90% support;8 black squares indicate >90% support using the partition cluster consensus method;8,120 “G” indicates groups based on gene structure information.37 and “P” indicates branches based on relative positions within chromosomal loci.13,14 Branch lengths are not to scale.

Clades C (antithrombin), D (heparin cofactor II), G (C1-inhibitor), H (heat shock protein 47), and I (neuroserpin) comprise few sequences which are mostly orthologues; hence, the evolutionary relationships between their members are consistent with common functions. The other clades mostly have members with markedly different properties. A dichotomy is apparent between clade B, which comprised of predominantly intracellular serpins, and the other vertebrate clades, which are predominantly populated by serpins with extracellular roles, although there

ch01

Jan. 10, 2007

8:57

SPI-B429


Evolution and Classification of the Serpin Superfamily Table 1 Cladeb A Antitrypsin-like

Division of the serpin superfamily into clades.a Locus 14q32.1

Xq22.2 1q42-q43 B Ovalbumin-like

6p25

18q21.3

C Antithrombin

9

Member SERPIN A13 SERPIN A3 SERPIN A5 SERPIN A4 SERPIN A12 SERPIN A9 SERPIN A11 SERPIN A1 SERPIN A2 SERPIN A6

Kallisatin-like Antichymotrypsin Protein C inhibitor Kallistatin Vaspin Centerin Antiproteinase-like 2 Antitrypsin Antitrypsin-related Corticosteroid-binding globulin SERPIN A10 Protein Z inhibitor SERPIN A7 Thyroxine-binding globulin AGT Angiotensinogen (SERPIN A8)1 SERPIN BP1 Pseudogene SERPIN B1 Monocyte/neutrophil elastase inhibitor SERPIN B9 PI-9 SERPIN B6 PI-6 SERPIN B8 PI-8 SERPIN B10 Bomapin SERPIN B2 Plasminogen activator inhibitor-2 SERPIN B7 Megsin SERPIN B11 Epipin SERPIN B3 Squamous cell carcinoma antigen-1 SERPIN B4 Squamous cell carcinoma antigen-2 SERPIN B13 Hurpin SERPIN B12 Yukopin SERPIN B5 Maspin

1q23-q25.1 SERPIN C1

D Heparin cofactor II 22q11.2

Genec

Common name

SERPIN D1

I I I I I I I I I I I I I

II II II II III III III III III III III III II

Antithrombin

IV

Heparin cofactor II

I (Continued)

ch01

Jan. 10, 2007

8:57

SPI-B429


10

J. A. Irving et al. Table 1

Cladeb E PAI-1/GDN

F PEDF

G C1-inhibitor

(Continued)

Locus

Member

7q21.3-q22

SERPIN E1

Common name

Genec

2q33-q35

Plasminogen activator V inhibitor-1 SERPIN E2 Glia-derived nexin V

17p13.1

SERPIN F1

17pter-p12

SERPIN F2

Pigment epitheliumderived factor Alpha-2-antiplasmin

11q12-q13.1 SERPIN G1 C1-inhibitor

VI VI VI

H Heat shock protein 47 11q13.5

SERPIN H1 Heat-shock protein 47 VII

I

SERPIN I1 SERPIN I2

Neuroserpin

3q26

Neuroserpin Pancpin

V V

a The human serpins are shown, including the chromosomal location, nomenclature and gene structure. For multigene loci, genes are listed in the telomere-to-centromere direction. b Clade designation — as shown in the phylogenetic tree (Fig. 3). c Representative intron/exon patterns are shown in Fig. 4.

are exceptions to this observation. Non-inhibitory serpins do not cluster together and several clades contain serpins that are known to be activated by ligands. 3.1. The nomenclature of the serpin superfamily The International Committee on the Nomenclature of the Serpins has adopted a system of naming that refers to the clade from which a given serpin arises8 and includes a systematic identifier which is applied as the family grows.39 The gene symbol used is “SERPIN”, followed by a letter designating the subfamily (or clade) from which the particular protein arises, the letter “P” if the gene in question is a pseudogene, and a number indicating that particular paralogue. To indicate the species of origin, the three-letter prefix designation of that species by the Committee on Standardization in Human Cytogenetics is used. The 35 human sequences act as “reference sequences” by which orthologues from other organisms are named. Hence, as human antitrypsin is referred to as (HSA)SERPINA1, mouse antitrypsin is referred to as referred to as (MMU)SERPINA1. Where

ch01

Jan. 10, 2007

8:57

SPI-B429



11

Fig. 4. The gene structure of the human serpins are shown grouped into representative “families” numbered as in Table 1: boxed regions indicate the average relative size and alignment of exons, with white and grey shading used for contrast; a solid horizontal bar indicates a large (>20 amino acid) region of gaps in the alignment that occurs within an exon, while gap regions outside exons are blank. Triangles are used to indicate the phase of an exon: white indicates the exon is in frame with the ATG initiation codon; grey indicates a +1 frameshift; and black indicates a +2 frameshift. As members of family VI have different numbers of exons, individual gene structures are shown.

there is no human orthologue — for example, a mouse sequence that is a member of clade A but has no strong evolutionary relationship with a particular human serpin, that gene would have a unique numerical designation greater than the highest numbered human serpin in the clade. Clade A This is the largest clade and contains the antitrypsin-like serpins. These serpins are involved in a diverse range of processes that are either commonly associated with the inhibition of serine proteinases (kallistatin, RASP-1, antitrypsin, and antichymotrypsin) or function in non-inhibitory roles. For instance, antitrypsin has a role in preventing lung damage by inhibiting elastase,47 while protein C inhibitor has an active role in the coagulation pathway.48,49 Non-inhibitory serpins such as thyroxine binding globulin are involved in hormone transport,50 while angiotensinogen has been shown to regulate blood pressure51 and UTMP has a proposed sequestering role in pregnancy.52 Antichymotrypsin has been implicated as having a role in Alzheimer’s disease due to its presence in the plaques of the brains

ch01

Jan. 10, 2007

8:57

SPI-B429

12


J. A. Irving et al.

of Alzheimer’s patients.53 In addition, it has also been shown to interact with Alzheimer’s Aβ peptide,1–42 which is then released during RCL cleavage and able to readily fibrillize.54 The protein currently designated SERPINA13 appears to lack a functional RCL, due to the presence of a premature stop codon, and hence would not be expected to exhibit inhibitory activity, even though transcripts have been detected in the liver.55 A recent study has indicated that this stop codon is not present in homologues from other primate species, and has been acquired since the human-chimpanzee divergence.56 In the mouse, in spite of the overall conservation of gene order at the clade A locus (12F1) with respect to the syntenic region of the human locus (14q32), SERPINA1 and SERPINA3 have undergone significant expansion to yield 5 and 14 homologues, respectively.57,58 There is evidence of functional diversification among these genes.59,60 Clade B The serpins that form part of this clade are often referred to as the ov-serpins, due to the sequence similarity they share with ovalbumin.33 These serpins are inhibitors of serine or cysteine proteases, with the exception of maspin and ovalbumin, which have no defined inhibitory functions. Clade B serpins partake in a myriad of regulatory roles involving inflammation, angiogenesis, apoptosis and fibrinolysis, and collectively, they provide a protective role within the cell.61 Ov-serpins lack an N-terminal signal sequence and mostly exert their biological effect within the cell; however, they have also been detected in the extracellular space. Most also characteristically possess a loop between the C- and D-helices (the “C–D loop”); in PAI-2, this loop affects protein polymerization and cell survival,62,63 while in the avian protein MENT, this loop plays a role in chromatin remodeling.64 Comparison with non-human genomes suggests that the repertoire and position of 13 human clade B serpins have developed by gene duplication and chromosomal breakage from four progenitor genes,14 present at a single locus in the last shared ancestor with amphibians (SERPINs B1, B5, B6, and B12); that this expanded to six at the divergence point of mammals from birds (SERPINs B1, B2, B5, B6, B10, and B12);13,14 and that the mouse orthologues of three genes at the human 6p25 locus (SERPINs B1, B6, and B9) have undergone further propagation to yield more than 15 apparently functional serpins on mouse chromosome 13.65,66

ch01

Jan. 10, 2007

8:57

SPI-B429



13

Clade C Antithrombin is the major anticoagulant serpin and inhibits a number of serine proteinases of the coagulation pathway, even though the principal targets are generally regarded to be thrombin and factor Xa.67 Antithrombin is activated against its target proteases by the binding of heparin to helix D.68 Antithrombin has been found to exhibit antiviral properties,69 as well as being involved in inflammation,70 and antiangiogenesis.71 An analysis of sequence conservation patterns strongly suggests that the mechanism of activation of this protein is conserved in a wide variety of vertebrate species.72 Clade D Heparin cofactor II is a potent inhibitor of thrombin; however, unlike antithrombin, it does not inhibit other proteases of the coagulation system. It possesses an uncharacteristic N-terminal extension which interacts with the anion-binding exosite of thrombin. As with antithrombin (clade C), heparin cofactor II also relies on activation by glycosaminoglycans.32 Clade E Plasminogen activator inhibitor-1 has a defined role in the plasminogen/plasmin system, and thus contributes to a multitude of physiological processes; for example, it is associated with common syndromes such as atherosclerosis, diabetes, and hypertension. It is able to inhibit urokinasetype plasminogen activator and is thus said to have a role in cell migration and tissue remodeling.73 Glia-derived nexin-1 (or protease nexin-1) is a potent inhibitor of thrombin, even though its physiological role has not been properly defined. It is a specific regulator of several proteases and has been implicated in roles within the central nervous system, the vasculature and the extracellular matrix.74,75 Clade F Pigment epithelium derived factor functions both as an inhibitor of angiogenic processes and is a neurotrophic factor.76 In contrast, its closest relative, alpha-2-antiplasmin, regulates fibrinolysis during clot formation.77

ch01

Jan. 10, 2007

8:57

SPI-B429

14


J. A. Irving et al.

Clade G C1-inhibitor regulates the activation of proteases in both the complement and the contact pathways, and its deficiency leads to angioedema. It has been shown that C1-inhibitor is involved at sites of inflammation and also in the migration of leukocytes across the endothelium.78 Clade H Heat shock protein 47 is not a protease inhibitor but a procollagen/collagenspecific binding protein that functions as a molecular chaperone during collagen biosynthesis.79 It has also been implicated in a number of diseases including arteriosclerosis, myocardial infarction and a number of fibroses.80 Clade I Neuroserpin inhibits tissue-type plasminogen activator and has a prominent role in the nervous system, including neurite outgrowth and synaptogenesis. It also has a role in the vascular system and plays a therapeutic role in protecting the brain during ischaemia.81 Familial variants of neuroserpin have been shown to polymerise and aggregate as Collins bodies within neurons, leading to the neurodegenerative disease FENIB.82

4. The Structural Phylogeny of the Serpin Domain By 1993, sufficient serpin structures had been solved to clearly demonstrate the high degree of structural similarity between members. Cleaved ovalbumin, antichymotrypsin, and antitrypsin structures showed an overall root mean square deviation (between Cα atoms) of 0.67–1.71 Å.7 In each case, the overall native fold remained the same, i.e. reactive centre loop protruding from the top of the molecule, three β-sheets, and nine α-helices. In the last few years, there has been an increasingly extensive coverage of the serpin superfamily. Nine of the sixteen evolutionary branches of the serpin family now have one or more representative crystal structures (Table 2). The structural biology of this family is somewhat complicated by the ability of the fold to adopt alternative topologies, including the

ch01

Jan. 10, 2007

8:57

SPI-B429



15

Table 2 Highest-resolution structures of native and cleaved members from the serpin superfamily. Cladea A

Serpin Antitrypsin

Homo sapiens

Antichymotrypsin

Mus musculus Homo sapiens Homo sapiens

Protein C inhibitor B

Species

PDBb

Åc

Ref

Native Cleaved, complex Native Cleaved Cleaved

1qlp 1ezx AB

2.0 2.6

105 87

1yxa 1as4 1lq8 AB

2.1 2.1 2.4

90 106 107

1by7

2.0

85

1hle

1.95

108

Native Cleaved, mutant Native

1ova 1jti

1.95 2.3

34 95

1wz9

2.1

109

Plasminogen activator inhibitor 2 Leukocyte elastase inhibitor Ovalbumin

Gallus gallus

Maspin

Homo sapiens

C

Antithrombin

Homo sapiens

Native Latentd

1t1f 1e05 L

2.75 2.62

— 110

D

Heparin cofactor II

Homo sapiens

Native

1jmj

2.35

32

E

Plasminogen activator inhibitor 1

Homo sapiens

Native Cleaved

1dvm 9pai

2.4 2.7

111 112

F

Pigment epithelium derived factor

Homo sapiens

Native

1imv

2.85

113

I

Neuroserpin

Mus musculus

Cleaved

1jjo

3.06

114

K

Alaserpin (K variant)

Manduca sexta

Native

1sek

2.1

115

N

crmA

Cowpox virus

Cleaved

1f0c

2.26

104

Thermopin

Thermobifida fusca

Native Cleaved

1sng 1mtp

1.76 1.5

92 116

n.d.

a n.d.

Homo sapiens

Form

Equus caballus

Native, mutant Cleaved

— no clade designation.

b The PDB accession of the crystal structure;117

if it has multiple components, the relevant serpin chains are listed. c Resolution of the structure; italics denote structures higher than 2.5 Å resolution. d Latent antithrombin is used in place of the 3.2 Å cleaved antithrombin structure.

ch01

Jan. 10, 2007

8:57

SPI-B429

16


J. A. Irving et al.

metastable native state, the stable cleaved and latent forms, serpin–protease complexes, and loop-sheet polymers.83 In particular, this highlights the flexibility of two important features of the serpin mechanism: the RCL and the Aβ-sheet. In spite of this conformational variation, superpositions of cleaved and uncleaved structures highlight the well-conserved nature of the serpin fold (Fig. 2). This suggests that the fold is well-suited to its biological function and is most likely intolerant to substantial structural changes within the molecule. For example, extending the length of the reactive center loop promotes the formation of an inactive latent conformation84 and mutation of underlying residues in the shutter region can lead to folding defects.85 Conservation of structure may also reflect the fact that serpins exist as “two-in-one” proteins, in that they adopt two different conformations with different packing arrangements.86 Broadly speaking, there are two major conformations of serpin: RCL fully inserted (due to cleavage, transition to latency, or polymerization) and RCL mostly expelled (ranging from partially inserted in the case of nonactivated antithrombin, to fully expelled as seen with maspin). Most serpins are required to achieve a delicate balance between these two states; it is the transition from RCL-expelled to RCL-inserted that permits serpins to act as covalent protease inhibitors.87 Hence, this requires that during the course of evolution, a serpin sequence is under pressure to both favor folding to the active state and maintaining the ability to adopt an energetically favorable RCL-inserted form, without leading to inappropriate conformations such as polymers and the latent form. Indeed, the path of the reactive center loop as it inserts into the central A-β sheet of the molecule is mirrored by the pattern of highly conserved residues among members of the serpin superfamily;8 and regions that play a role in the transition between loop-expelled and loop-incorporated, termed the hinge, the breach and the shutter, appear to be co-evolving at a slower than average rate.88 These residues are shown in Fig. 5 and it can be seen that the association between them is best understood in the context of the cleaved structure of antitrypsin, and not the active, inhibitory, uncleaved form, illustrating the influence of different structural contexts on evolutionary behavior. Figure 6 shows the influence of evolutionary divergence on structural movements within the serpin fold. It is clear from the uncleaved serpin

ch01

Jan. 10, 2007

8:57

SPI-B429



17

Fig. 5. The position of residues identified as conserved in more than 75% of serpin sequences8 (white spheres) and those representing co-evolving sites88 (black spheres) are shown mapped onto the crystal structures of (A) cleaved and (B) uncleaved human antitrypsin (see Table 2 for accessions). Gray spheres indicate residues common to the two datasets. The stereo figure was prepared using Molscript119 and Raster3D.121

ch01

Jan. 10, 2007

8:57

SPI-B429

18


J. A. Irving et al.

Fig. 6.

(Continued)

ch01

Jan. 10, 2007

8:57

SPI-B429



Fig. 6.

(Continued)

19

ch01

Jan. 10, 2007

8:57

SPI-B429

20


J. A. Irving et al.

Fig. 6. A tree depicting the phylogenetic relationship between (A) cleaved and (B) native serpins of known structure, as listed in Table 2. Multiple pairwise superpositions were performed using the DaliLite program.122 At each ancestral node on the tree, the residues constituting the “common core”123 of all structures that descend from that node were calculated from these pairwise superpositions. A “common core” excludes regions that are not consistently structurally aligned. Each cartoon shown represents the “common core” where two or more branches meet; colored lines indicate those regions that are absent from the common core at the next ancestral node (cyan, alpha-helix; red, beta-sheet; green, and loop region). For clarity, the multi-gene clades A/D, B, and E/I are colored blue, red and green, respectively. (C) The structure of native antitrypsin, indicating the order in which regions are lost from the common core when moving from the terminal branch to the most ancestral node; red indicates the first region to be lost, followed by orange, yellow, and green; blue-violet residues are present in the common core at the base of the tree. Structures were drawn using Molscript.119

structures (Fig. 6A and B) that the reactive centre loop is the most mutable element, along with the CD-loop of the clade B serpin, ovalbumin, and the C- and N-termini of thermopin and the N-terminus of heparin cofactor II. This is followed closely by the loop joining helix D to strand 1A, the loop

ch01

Jan. 10, 2007

8:57

SPI-B429



21

joining the F-helix to strand 3A, and the N-terminal portion of the G-helix. Deeper into the tree, more loops joining structural elements are lost, such as the residues between strands 1B and 2B. Finally, the D-helix, E-helix, and G-helix are found to superpose poorly near the base of the tree (Fig. 6C). 4.1. Sequence plasticity in the reactive center loop The reactive center loop (RCL) of serpins has long been recognized as being hypervariable,21 i.e. with a non-synonymous mutation rate above the background rate, and this can be easily seen when examining the multigene clusters that have appeared due to extensive gene duplication in the mouse.58,59 The RCL is uniquely varied in its structure and when comparing one serpin with another, it is evident that it is able to adopt a number of different conformations; for example, the so-called canonical conformation of antitrypsin around the specificity-determining residues89 and the RCL of the closely related murine antichymotrypsin, which is almost perpendicular to that of antitrypsin and is partially inserted into the A-sheet.90 The serpin fold is largely well ordered. The central region of the RCL represents a frequent exception to this and typically makes few contacts with the body of the serpin; thus, it is often not visible in crystal structures due to a high degree of disorder such as seen with plasminogen activator inhibitor-191 and thermopin.92 Presumably, this permits the acquisition of optimal target specificity on the basis of sequence alone, without requiring compensating mutations for the maintenance of structural integrity. As discussed above, this is a two-edged sword; a point mutation in this region can detrimentally alter the biological properties of the serpin, and genetic disease ensues. The N-terminal portion of the RCL is somewhat more constrained in sequence as it becomes incorporated into the A-sheet during the inhibitory process, and is required to contain a string of small polar and aliphatic amino acids known at the “hinge region motif”.93 This is exemplified by ovalbumin, a serpin that does not undergo conformational change.94 With the requirement for a hinge region compatible with the A-sheet insertion discarded, residue 339 of the RCL has undergone a change from threonine to arginine with respect to related serpins. Introducing this mutation into other serpins results in their inability to effectively inhibit proteases, and mutating this residue back to a threonine in ovalbumin permits RCL insertion into the A-sheet.95

ch01

Jan. 10, 2007

8:57

SPI-B429

22


J. A. Irving et al.

Similar plasticity is seen in the inhibition of different protease classes by serpins. While the majority of serpins characterized to date have been shown to be inhibitors of trypsin-like serine proteinases,6,96 the family also has inhibitory members that target other classes of proteinase. The viral crmA protein inhibits caspase-1,97 and squamous cell carcinoma antigen-1 inhibit papain-like cysteine proteinases,98 but minimal changes in the specificitydetermining region of the RCL are required to alter the target protease class.25,26,99

5. Concluding Statements: The Prokaryotic Serpins and the Root of the Serpin Superfamily At the root of the “tree of life” lies the dichotomy between prokaryotic and eukaryotic organisms, and the divergence of plants, animals, and fungi. The order in which these branches diverged at this period in natural history is a well-debated topic. An ancient gene represented in a significant number of prokaryotes and eukaryotes implies universal importance to cellular life; whereas there are a significant number of eukaryotic genes that appeared only after the divergence from prokaryotic ancestors.100 Due to the high throughput sequencing of whole prokaryotic organisms from a diverse range of environments and genetic backgrounds, a host of serpin genes have been identified that suggest the root of the gene tree may lie prior to the divergence of bacteria from archaea.9,11 However, it is clear that serpin genes are not indispensable to cellular life, as there are a large number of fully sequenced prokaryotic organisms, as well as yeasts, lacking a recognizable member of this family.9 Whether this arises due to some other proteins providing an analogous function cannot be definitively known; the primary role of serpins is to inhibit proteases, but it is also clear that the presence of a protease does not predicate the presence of a serpin.9 The prokaryotic organisms known to contain serpin genes include extremophiles (such as Pyrobaculum aerophilum, Thermoanaerobacter tengcongensis, and Thermobifida fusca) and mesophiles. Based on current data, there is no evidence to suggest the capture of eukaryotic serpin genes by bacteria. However, there are interesting examples of bacteria that have an association with an eukaryotic host: Ruminococcus albus, which resides in the foregut of ruminant animals and produces

ch01

Jan. 10, 2007

8:57

SPI-B429



23

cellulolytic enzymes; and Nostoc punctiforme, which lives as an endosymbiont in the fungus Geosiphon pyriformis. It is possible that a shared lineage exists with these organisms but remains undetectable due to gene divergence. Multiple serpin genes have been identified in plants,101 but as yet cognate target proteases have not been identified. It was clear from a study involving serpins from Arabidopsis thaliana and Hordeum vulgare that for plants, the diverse repertoire of serpins evolved from a single progenitor serpin, for each of the species, prior to species divergence.8 It seems likely then that a single serpin gene existed at least as far back as the plant/animal split, and its recruitment by complex, multicellular organisms has led to the large number of genes seen today. Based on the connectivity of secondary structure elements in serpins, two separate studies have countenanced the possible integrity of a C-terminal domain composed solely of the B- and C-sheets and G- and H-helices, and lacking helices A–F and strands s1A–s3A.102,103 It has been suggested on this basis that the serpin fold may have developed from the joining of the two domains near the conserved Trp-194 residue, followed by an insertion event generating two strands of the A-sheet.102 Furthermore, there is structural evidence that the D-helix and the G-helix are dispensable secondary structural components.10,104 It may be that only with the knowledge of the earliest precursors of the serpin fold will the timing of the origin of the serpin superfamily become clear.

References 1. Hunt LT and Dayhoff MO (1980) A surprising new protein superfamily containing ovalbumin, antithrombin-III, and alpha 1-proteinase inhibitor. Biochem Biophys Res Commun 95:864–871. 2. Laurell CB and Eriksson S (1963) The electrophoretic alpha 1-globulin pattern of serum in alpha 1-antitrypsin deficiency. Scand J Clin Lab Invest Suppl 15:132–140. 3. Egeberg O (1965) Inherited antithrombin deficiency causing thrombophilia. Thromb Diath Haemorrh 13:516–541. 4. Carrell RW and Travis J (1985) Alpha-1-antitrypsin and the serpins: Variation and countervariation. Trends Biochem Sci 8:20–24.

ch01

Jan. 10, 2007

8:57

SPI-B429

24


J. A. Irving et al.

5. Hejgaard J, Rasmussen SK, Brandt A and Svendsen I (1985) Sequence homology between barley endosperm protein Z and protease inhibitors of the alpha1-antitrypsin family. FEBS Lett 180:89–94. 6. Huber R and Carrell RW (1989) Implications of the three-dimensional structure of alpha 1-antitrypsin for structure and function of serpins. Biochemistry 28:8951–8966. 7. Marshall CJ (1993) Evolutionary relationships among the serpins. Philos Trans R Soc Lond B Biol Sci 342:101–119. 8. Irving JA, Pike RN, Lesk AM and Whisstock JC (2000) Phylogeny of the serpin superfamily implications of patterns of amino acid conservation for structure and function. Genome Res 10:1845–1864. 9. Irving JA, Steenbakkers PJM, Lesk AM, Op den Camp HJM, Pike RN and Whisstock JCW (2002) Serpins in prokaryotes. Mol Biol Evol 19: 1881–1890. 10. Irving JA, Cabrita LD, Rossjohn J, Pike RN, Bottomley SP and Whisstock JC (2003) The 1.5 Å crystal structure of a prokaryote serpin: Controlling conformational change in a heated environment. Structure 11:387–397. 11. Roberts TH, Hejgaard J, Saunders NF, Cavicchioli R and Curmi PM (2004) Serpins in unicellular Eukarya, Archaea, and Bacteria: Sequence analysis and evolution. J Mol Evol 59:437–447. 12. Cole EB, Miller D, Rometo D, Greenberg RM, Bromme D, Cataltepe S, Pak SC, Mills DR, Silverman GA and Luke CJ (2004) Identification and activity of a lower eukaryotic serine proteinase inhibitor (serpin) from Cyanea capillata: Analysis of a jellyfish serpin, jellypin. Biochemistry 43:11750–11759. 13. Benarafa C and Remold-O’Donnell E (2005) The ovalbumin serpins revisited: Perspective from the chicken genome of clade B serpin evolution in vertebrates. Proc Natl Acad Sci USA 102:11367–11372. 14. Kaiserman D and Bird PI (2005) Analysis of vertebrate genomes suggests a new model for clade B serpin evolution. BMC Genomics 6:167. 15. Bateman A, Coin L, Durbin R, Finn RD, Hollich V, Griffiths-Jones S, Khanna A, Marshall M, Moxon S, Sonnhammer EL, Studholme DJ, Yeats C and Eddy SR (2004) The Pfam protein families database. Nucleic Acids Res 32:D138–D141. 16. Murzin AG, Brenner SE, Hubbard T and Chothia C (1995) SCOP: A structural classification of proteins database for the investigation of sequences and structures. J Mol Biol 247:536–540. 17. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, Davis AP, Dolinski K, Dwight SS, Eppig JT, Harris MA, Hill DP, Issel-Tarver L, Kasarskis A, Lewis S, Matese JC, Richardson JE, Ringwald M, Rubin GM

ch01

Jan. 10, 2007

8:57

SPI-B429



18.

19. 20. 21.

22. 23.

24.

25.

26.

27.

28.

29.

25

and Sherlock G (2000) Gene ontology: Tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 25:25–29. Saier MH Jr, Tran CV and Barabote RD (2006) TCDB: The transporter classification database for membrane transport protein analyses and information. Nucleic Acids Res 34:D181–D186. Dagenbach EM and Endow SA (2004) A new kinesin tree. J Cell Sci 117: 3–7. Gettins PG (2002) Serpin structure, mechanism, and function. Chem Rev 102:4751–4804. Inglis JD and Hill RE (1991) The murine Spi-2 proteinase inhibitor locus: A multigene family with a hypervariable reactive site domain. Embo J 10:255–261. Lewis JH (1978) Antithrombin Pittsburgh: An alpha1-antitrypsin variant causing hemorrhagic disease. Blood 51:129–137. Owen MC, Brennan SO, Lewis JH and Carrell RW (1983) Mutation of antitrypsin to antithrombin alpha 1-antitrypsin Pittsburgh (358 Met leads to Arg), a fatal bleeding disorder. N Engl J Med 309:694–698. Schulze AJ, Huber R, Degryse E, Speck D and Bischoff R (1991) Inhibitory activity and conformational transition of alpha 1-proteinase inhibitor variants. Eur J Biochem 202:1147–1155. Luke C, Schick C, Tsu C, Whisstock JC, Irving JA, Bromme, D, Juliano L, Shi AUGP, Chapman HA and Silverman GA (2000) Simple modifications of the serpin reactive site loop convert SCCA2 into a cysteine protease inhibitor: A critical role for the P3 proline in facilitating RSL cleavage. Biochemistry 39:7081–7091. Irving JA, Pike RN, Dai W, Bromme D, Worrall DM, Silverman GA, Coetzer TH, Dennison C, Bottomley SP and Whisstock JC (2002) Evidence that serpin architecture intrinsically supports papain-like cysteine protease inhibition: Engineering alpha(1)-antitrypsin to inhibit cathepsin proteases. Biochemistry 41:4998–5004. Karrer KM, Peiffer SL and DiTomas ME (1993) Two distinct gene subfamilies within the family of cysteine protease genes. Proc Natl Acad Sci USA 90:3063–3067. Turk B, Turk V and Turk D (1997) Structural and functional aspects of papain-like cysteine proteinases and their protein inhibitors. Biol Chem 378: 141–150. Hosfield CM, Elce JS, Davies PL and Jia Z (1999) Crystal structure of calpain reveals the structural basis for Ca(2+)-dependent protease activity and a novel mode of enzyme activation. Embo J 18:6880–6889.

ch01

Jan. 10, 2007

8:57

SPI-B429

26


J. A. Irving et al.

30. Ward W, Alvarado L, Rawlings ND, Engel JC, Franklin C and McKerrow JH (1997) A primitive enzyme for a primitive cell: The protease required for excystation of Giardia. Cell 89:437–444. 31. Huang X, Czerwinski E and Mellgren RL (2003) Purification and properties of the Dictyostelium calpain-like protein. Cpl. Biochemistry 42:1789–1795. 32. Baglin TP, Carrell RW, Church FC, Esmon CT and Huntington JA (2002) Crystal structures of native and thrombin-complexed heparin cofactor II reveal a multistep allosteric mechanism. Proc Natl Acad Sci USA 99:11079– 11084. 33. Remold-O’Donnell E (1993) The ovalbumin family of serpin proteins. FEBS Lett 315:105–108. 34. Stein PE, Leslie AG, Finch JT and Carrell RW (1991) Crystal structure of uncleaved ovalbumin at 1.95 A resolution. J Mol Biol 221:941–959. 35. Scott FL, Eyre HJ, Lioumi M, Ragoussis J, Irving JA, Sutherland GA and Bird PI (1999) Human ovalbumin serpin evolution: Phylogenic analysis, gene organization, and identification of new PI8-related genes suggest that two interchromosomal and several intrachromosomal duplications generated the gene clusters at 18q21-q23 and 6p25. Genomics 62:490–499. 36. Ragg H, Lokot T, Kamp PB, Atchley WR and Dress A (2001) Vertebrate serpins: Construction of a conflict-free phylogeny by combining exon-intron and diagnostic site analyses. Mol Biol Evol 18:577–584. 37. Atchley WR, Lokot T, Wollenberg K, Dress A and Ragg H (2001) Phylogenetic analyses of amino acid variation in the serpin proteins. Mol Biol Evol 18:1502–1511. 38. Rollini P and Fournier RE (1997) A 370-kb cosmid contig of the serpin gene cluster on human chromosome 14q32.1: Molecular linkage of the genes encoding alpha 1-antichymotrypsin, protein C inhibitor, kallistatin, alpha 1antitrypsin, and corticosteroid-binding globulin. Genomics 46:409–415. 39. Silverman GA, Bird PI, Carrell RW, Church FC, Coughlin, PB, Gettins PG, Irving JA, Lomas DA, Luke CJ, Moyer RW, Pemberton PA, RemoldO’Donnell E, Salvesen GS, Travis J and Whisstock JC (2001) The serpins are an expanding superfamily of structurally similar but functionally diverse proteins. Evolution, mechanism of inhibition, novel functions, and a revised nomenclature. J Biol Chem 276:33293–33296. 40. Nei M and Kumar S (2000) Molecular Evolution and Phylogenetics, 1st Edition. Oxford University Press: Oxford. 41. Eck RV and Dayhoff MO (1966) Atlas of Protein Sequence and Structure, 1st Edition. National Biomedical Research Foundation: Silver Spring, MD. 42. Felsenstein J (1981) Evolutionary trees from DNA sequences: A maximum likelihood approach. J Mol Evol 17:368–376.

ch01

Jan. 10, 2007

8:57

SPI-B429



27

43. Saitou N and Nei M (1987) The neighbor-joining method: A new method for reconstructing phylogenetic trees. Mol Biol Evol 4:406–425. 44. Felsenstein J (1985) Confidence limits on phylogenies: An approach using the bootstrap. Evolution 39:783–791. 45. Gu X (1998) Early metazoan divergence was about 830 million years ago. J Mol Evol 47:369–371. 46. Feng DF, Cho G and Doolittle RF (1997) Determining divergence times with a protein clock: Update and reevaluation. Proc Natl Acad Sci USA 94:13028– 13033. 47. Beatty K, Bieth J and Travis J (1980) Kinetics of association of serine proteinases with native and oxidized alpha-1-proteinase inhibitor and alpha-1antichymotrypsin. J Biol Chem 255:3931–3934. 48. Rezaie AR, Cooper ST, Church FC and Esmon CT (1995) Protein C inhibitor is a potent inhibitor of the thrombin-thrombomodulin complex. J Biol Chem 270:25336–25339. 49. Elisen MG, von dem Borne PA, Bouma BN and Meijers JC (1998) Protein C inhibitor acts as a procoagulant by inhibiting the thrombomodulin-induced activation of protein C in human plasma. Blood 91:1542–1547. 50. Refetoff S, Murata Y, Mori Y, Janssen OE, Takeda K and Hayashi Y (1996) Thyroxine-binding globulin: Organization of the gene and variants. Horm Res 45:128–138. 51. Gardes J, Bouhnik J, Clauser E, Corvol P and Menard J (1982) Role of angiotensinogen in blood pressure homeostasis. Hypertension 4:185–189. 52. McFarlane JR, Foulds LM, O’Connor AE, Phillips DJ, Jenkin G, Hearn MT and de Kretser DM (1999) Uterine milk protein, a novel activinbinding protein, is present in ovine allantoic fluid. Endocrinology 140: 4745–4752. 53. Abraham CR, Selkoe DJ and Potter H (1988) Immunochemical identification of the serine protease inhibitor alpha 1-antichymotrypsin in the brain amyloid deposits of Alzheimer’s disease. Cell 52:487–501. 54. Janciauskiene S, Eriksson S and Wright HT (1996) A specific structural interaction of Alzheimer’s peptide A beta 1-42 with alpha 1-antichymotrypsin. Nat Struct Biol 3:668–671. 55. Marsden MD and Fournier RE (2005) Organization and expression of the human serpin gene cluster at 14q32.1. Front Biosci 10:1768–1778. 56. Hahn Y and Lee B (2006) Human-specific nonsense mutations identified by genome sequence comparisons. Hum Genet 119:169–178. 57. Barbour KW, Wei F, Brannan C, Flotte TR, Baumann H and Berger FG (2002) The murine alpha(1)-proteinase inhibitor gene family: Polymorphism, chromosomal location, and structure. Genomics 80:515–522.

ch01

Jan. 10, 2007

8:57

SPI-B429

28


J. A. Irving et al.

58. Forsyth S, Horvath A and Coughlin P (2003) A review and comparison of the murine alpha1-antitrypsin and alpha1-antichymotrypsin multigene clusters with the human clade A serpins. Genomics 81:336–345. 59. Barbour KW, Goodwin RL, Guillonneau F, Wang Y, Baumann H, and Berger FG (2002) Functional diversification during evolution of the murine alpha(1)proteinase inhibitor family: Role of the hypervariable reactive center loop. Mol Biol Evol 19:718–727. 60. Horvath AJ, Forsyth SL and Coughlin PB (2004) Expression patterns of murine antichymotrypsin-like genes reflect evolutionary divergence at the Serpina3 locus. J Mol Evol 59:488–497. 61. Silverman GA, Whisstock JC, Askew DJ, Pak SC, Luke CJ, Cataltepe S, Irving JA and Bird PI (2004) Human clade B serpins (ov-serpins) belong to a cohort of evolutionarily dispersed intracellular proteinase inhibitor clades that protect cells from promiscuous proteolysis. Cell Mol Life Sci 61: 301–325. 62. Wilczynska M, Lobov S, Ohlsson PI and Ny T (2003) A redox-sensitive loop regulates plasminogen activator inhibitor type 2 (PAI-2) polymerization. Embo J 22:1753–1761. 63. Dickinson JL, Bates EJ, Ferrante A and Antalis TM (1995) Plasminogen activator inhibitor type 2 inhibits tumor necrosis factor alpha-induced apoptosis. Evidence for an alternate biological function. J Biol Chem 270:27894–27904. 64. Springhetti EM, Istomina NE, Whisstock JC, Nikitina T, Woodcock CL and Grigoryev SA (2003) Role of the M-loop and reactive center loop domains in the folding and bridging of nucleosome arrays by MENT. J Biol Chem 278:43384–43393. 65. Kaiserman D, Knaggs S, Scarff KL, Gillard A, Mirza G, Cadman M, McKeone R, Denny P, Cooley J, Benarafa C, Remold-O’Donnell E, Ragoussis J and Bird PI (2002) Comparison of human chromosome 6p25 with mouse chromosome 13 reveals a greatly expanded ov-serpin gene repertoire in the mouse. Genomics 79:349–362. 66. Mallon AM, Wilming L, Weekes J, Gilbert JG, Ashurst J, Peyrefitte S, Matthews L, Cadman M, McKeone R, Sellick CA, Arkell R, Botcherby MR, Strivens MA, Campbell RD, Gregory S, Denny P, Hancock JM, Rogers J and Brown SD (2004) Organization and evolution of a gene-rich region of the mouse genome: A 12.7-Mb region deleted in the Del(13)Svea36H mouse. Genome Res 14:1888–1901. 67. Pratt CW and Church FC (1993) General features of the heparin-binding serpins antithrombin, heparin cofactor II and protein C inhibitor. Blood Coagul Fibrinolysis 4:479–490.

ch01

Jan. 10, 2007

8:57

SPI-B429



29

68. Jin L, Abrahams JP, Skinner R, Petitou M, Pike RN and Carrell RW (1997) The anticoagulant activation of antithrombin by heparin. Proc Natl Acad Sci USA 94:14683–14688. 69. Geiben-Lynn R, Brown N, Walker BD and Luster AD (2002) Purification of a modified form of bovine antithrombin III as an HIV-1 CD8+ T-cell antiviral factor. J Biol Chem 277:42352–42357. 70. Oelschlager C, Romisch J, Staubitz A, Stauss H, Leithauser B, Tillmanns H and Holschermann H (2002) Antithrombin III inhibits nuclear factor kappaB activation in human monocytes and vascular endothelial cells. Blood 99:4015–4020. 71. O’Reilly MS, Pirie-Shepherd S, Lane WS and Folkman J (1999) Antiangiogenic activity of the cleaved conformation of the serpin antithrombin. Science 285:1926–1928. 72. Backovic M and Gettins PG (2002) Insight into residues critical for antithrombin function from analysis of an expanded database of sequences that includes frog, turtle, and ostrich antithrombins. J Proteome Res 1: 367–373. 73. Gils A and Declerck PJ (2004) The structural basis for the pathophysiological relevance of PAI-I in cardiovascular diseases and the development of potential PAI-I inhibitors. Thromb Haemost 91:425–437. 74. Baker JB, Low DA, Simmer RL and Cunningham DD (1980) Protease-nexin: A cellular component that links thrombin and plasminogen activator and mediates their binding to cells. Cell 21:37–45. 75. Festoff BW, Rao JS and Hantai D (1991) Plasminogen activators and inhibitors in the neuromuscular system: III. The serpin protease nexin I is synthesized by muscle and localized at neuromuscular synapses. J Cell Physiol 147:76–86. 76. Steele FR, Chader GJ, Johnson LV and Tombran-Tink J (1993) Pigment epithelium-derived factor: Neurotrophic activity and identification as a member of the serine protease inhibitor gene family. Proc Natl Acad Sci USA 90:1526–1530. 77. Coughlin PB (2005) Antiplasmin: The forgotten serpin? Febs J 272: 4852–4857. 78. Davis AE 3rd, Cai S and Liu D (2004) The biological role of the C1 inhibitor in regulation of vascular permeability and modulation of inflammation. Adv Immunol 82:331–363. 79. Dafforn TR, Della M and Miller AD (2001) The molecular interactions of heat shock protein 47 (Hsp47) and their implications for collagen biosynthesis. J Biol Chem 276:49310–49319.

ch01

Jan. 10, 2007

8:57

SPI-B429

30


J. A. Irving et al.

80. Razzaque MS and Taguchi T (1999) The possible role of colligin/HSP47, a collagen-binding protein, in the pathogenesis of human and experimental fibrotic diseases. Histol Histopathol 14:1199–1212. 81. Yepes M and Lawrence DA (2004) Neuroserpin: A selective inhibitor of tissue-type plasminogen activator in the central nervous system. Thromb Haemost 91:457–464. 82. Davis RL, Shrimpton AE, Holohan PD, Bradshaw C, Feiglin D, Collins GH, Sonderegger P, Kinter J, Becker LM, Lacbawan F, Krasnewich D, Muenke M, Lawrence DA, Yerby MS, Shaw CM, Gooptu B, Elliott PR, Finch JT, Carrell RW and Lomas DA (1999) Familial dementia caused by polymerization of mutant neuroserpin. Nature 401:376–379. 83. Whisstock J, Skinner R and LeskAM (1998)An atlas of serpin conformations. Trends Biochem Sci 23:63–67. 84. Na YR and Im H (2005) The length of the reactive center loop modulates the latency transition of plasminogen activator inhibitor-1. Protein Sci 14:55–63. 85. Gooptu B, Hazes B, Chang WS, Dafforn TR, Carrell RW, Read RJ and Lomas DA (2000) Inactive conformation of the serpin alpha(1)-antichymotrypsin indicates two-stage insertion of the reactive loop: Implications for inhibitory function and conformational disease. Proc Natl Acad Sci USA 97:67–72. 86. Whisstock JC, Skinner R, Carrell RW and Lesk AM (2000) Conformational changes in serpins: I. The native and cleaved conformations of alpha(1)antitrypsin. J Mol Biol 296:685–699. 87. Huntington JA, Read RJ and Carrell RW (2000) Structure of a serpin-protease complex shows inhibition by deformation. Nature 407:923–926. 88. Buck MJ and Atchley WR (2005) Networks of coevolving sites in structural and functional domains of serpin proteins. Mol Biol Evol 22:1627–1634. 89. Elliott PR, Abrahams JP and Lomas DA (1998) Wild-type alpha 1-antitrypsin is in the canonical inhibitory conformation. J Mol Biol 275:419–425. 90. Horvath AJ, Irving JA, Rossjohn J, Law RH, Bottomley SP, Quinsey NS, Pike RN, Coughlin PB and Whisstock JC (2005) The murine orthologue of human antichymotrypsin: A structural paradigm for clade A3 serpins. J Biol Chem 280:43168–43178. 91. Harrop SJ, Jankova L, Coles M, Jardine D, Whittaker JS, Gould AR, Meister A, King GC, Mabbutt BC and Curmi PM (1999) The crystal structure of plasminogen activator inhibitor 2 at 2.0 A resolution: Implications for serpin function. Structure 7:43–54. 92. Fulton KF, Buckle AM, Cabrita LD, Irving JA, Butcher RE, Smith I, Reeve S, Lesk AM, Bottomley SP, Rossjohn J and Whisstock JC (2005) The high resolution crystal structure of a native thermostable serpin reveals the complex

ch01

Jan. 10, 2007

8:57

SPI-B429



93. 94. 95.

96. 97.

98.

99.

100.

101.

102. 103. 104.

31

mechanism underpinning the stressed to relaxed transition. J Biol Chem 280:8435–8442. Hopkins PC, Carrell RW and Stone SR (1993) Effects of mutations in the hinge region of serpins. Biochemistry 32:7650–7657. Stein PE, Tewkesbury DA and Carrell RW (1989) Ovalbumin and angiotensinogen lack serpin S-R conformational change. Biochem J 262:103–107. Yamasaki M, Arii Y, Mikami B and Hirose M (2002) Loop-inserted and thermostabilized structure of P1-P1 cleaved ovalbumin mutant R339T. J Mol Biol 315:113–120. Gettins PG (2000) Keeping the serpin machine running smoothly. Genome Res 10:1833–1835. Ray CA, Black RA, Kronheim SR, Greenstreet TA, Sleath PR, Salvesen GS and Pickup DJ (1992) Viral inhibition of inflammation: Cowpox virus encodes an inhibitor of the interleukin-1 beta converting enzyme. Cell 69: 597–604. Schick C, Pemberton PA, Shi GP, Kamachi Y, Cataltepe S, Bartuski AJ, Gornstein ER, Bromme D, Chapman HA and A SG (1998) Cross-class inhibition of the cysteine proteases cathepsins K, L, and S by the serpin squamous cell carcinoma antigen 1: A kinetic analysis. Biochemistry 37: 5258–5266. Al-Khunaizi M, Luke CJ, Askew YS, Pak SC, Askew DJ, Cataltepe S, Miller D, Mills DR, Tsu C, Bromme D, Irving JA, Whisstock JC and Silverman GA (2002) The serpin SQN-5 is a dual mechanistic-class inhibitor of serine and cysteine proteinases. Biochemistry 41:3189–3199. Koonin EV, Fedorova ND, Jackson JD, Jacobs AR, Krylov, DM, Makarova KS, Mazumder R, Mekhedov SL, Nikolskaya AN, Rao BS, Rogozin IB, Smirnov S, Sorokin AV, Sverdlov AV, Vasudevan S, Wolf YI, Yin JJ and Natale DA (2004) A comprehensive evolutionary classification of proteins encoded in complete eukaryotic genomes. Genome Biol 5:R7. Dahl SW, Rasmussen SK and Hejgaard J (1996) Heterologous expression of three plant serpins with distinct inhibitory specificities. J Biol Chem 271:25083–25088. Wright HT (1993) Introns and higher-order structure in the evolution of serpins. J Molecular Evol 36:136–143. Whisstock JC, Irving JA, Bottomley SP, Pike RN and Lesk AM (1999) Serpins in the Caenorhabditis elegans genome. Proteins 36:31–41. Renatus M, Zhou Q, Stennicke HR, Snipas SJ, Turk D, Bankston LA, Liddington RC and Salvesen GS (2000) Crystal structure of the apoptotic suppressor CrmA in its cleaved form. Structure 8:789–797.

ch01

Jan. 10, 2007

8:57

SPI-B429

32


J. A. Irving et al.

105. Elliott PR, Pei XY, Dafforn TR and Lomas DA (2000) Topography of a 2.0 A structure of alpha1-antitrypsin reveals targets for rational drug design to prevent conformational disease. Protein Sci 9:1274–1281. 106. Lukacs CM, Rubin H and Christianson DW (1998) Engineering an anionbinding cavity in antichymotrypsin modulates the “spring-loaded” serpinprotease interaction. Biochemistry 37:3297–3304. 107. Huntington JA, Kjellberg M and Stenflo J (2003) Crystal structure of protein C inhibitor provides insights into hormone binding and heparin activation. Structure 11:205–215. 108. Baumann U, Bode W, Huber R, Travis J and Potempa J (1992) Crystal structure of cleaved equine leucocyte elastase inhibitor determined at 1.95 A resolution. J Mol Biol 226:1207–1218. 109. Law RH, Irving JA, BuckleAM, Ruzyla K, Buzza M, Bashtannyk-Puhalovich TA, Beddoe TC, Nguyen K, Worrall DM, Bottomley SP, Bird PI, Rossjohn J and Whisstock JC (2005) The high resolution crystal structure of the human tumor suppressor maspin reveals a novel conformational switch in the G-helix. J Biol Chem 280:22356–22364. 110. Skinner R, Abrahams JP, Whisstock JC, Lesk AM, Carrell, RW and Wardell MR (1997) The 2.6 A structure of antithrombin indicates a conformational change at the heparin binding site. J Mol Biol 266:601–609. 111. Stout TJ, Graham H, Buckley DI and Matthews DJ (2000) Structures of active and latent PAI-1: A possible stabilizing role for chloride ions. Biochemistry 39:8460–8469. 112. Aertgeerts K, De Bondt HL, De Ranter C and Declerck PJ (1995) Crystallization and X-ray diffraction data of the cleaved form of plasminogen activator inhibitor-1. Proteins 23:118–121. 113. Simonovic M, Gettins PG and Volz K (2001) Crystal structure of human PEDF, a potent anti-angiogenic and neurite growth-promoting factor. Proc Natl Acad Sci USA 98:11131–11135. 114. Briand C, Kozlov SV, Sonderegger P and Grutter MG (2001) Crystal structure of neuroserpin: A neuronal serpin involved in a conformational disease. FEBS Lett 505:18–22. 115. Li J, Wang Z, Canagarajah B, Jiang H, Kanost M and Goldsmith EJ (1999) The structure of active serpin 1K from Manduca sexta. Structure Fold Des 7:103–109. 116. Irving JA, Cabrita LD, Rossjohn J, Pike RN, Bottomley SP and Whisstock JC (2003) The 1.5 A crystal structure of a prokaryote serpin: Controlling conformational change in a heated environment. Structure 11:387–397.

ch01

Jan. 10, 2007

8:57

SPI-B429



33

117. Berman HM, Westbrook J, Feng Z, Gilliland G, Bhat TN, Weissig H, Shindyalov IN and Bourne PE (2000) The Protein data bank. Nucleic Acids Res 28:235–242. 118. Bairoch A, Apweiler R, Wu CH, Barker WC, Boeckmann B, Ferro S, Gasteiger E, Huang H, Lopez R, Magrane M, Martin MJ, Natale DA, O’Donovan C, Redaschi N and Yeh LS (2005) The universal Protein resource (UniProt). Nucleic Acids Res 33:D154–D159. 119. Kraulis PJ (1991) MOLSCRIPT: A program to produce both detailed and schematic plots of Protein structures. J App Cryst 24:946–950. 120. Irving JA, Askew DJ and Whisstock JC (2004) Computational analysis of evolution and conservation in a protein superfamily. Methods 32:73–92. 121. Merritt EA and Bacon DJ (1997) Raster3D: Photorealistic Molecular Graphics. Meth Enzymol 277:505–524. 122. Holm L and Park J (2000) DaliLite workbench for protein structure comparison. Bioinformatics 16:566–567. 123. Lesk AM and Fordham WD (1996) Conservation and variability in the structures of serine proteinases of the chymotrypsin family. J Mol Biol 258:501–537.

ch01

Jan. 10, 2007 8:57 SPI-B429



ch02

Jan. 10, 2007

8:57

SPI-B429


2 Serpin Conformations Mary C. Pearce, Robert N. Pike, Arthur M. Lesk and Stephen P. Bottomley

1. Introduction Serpins, found in many different species and performing a wide range of roles, provide a valuable and exciting opportunity to understand the relationship between protein structure, function, and associated diseases. This is largely due to the abundance of high-resolution crystal structures available in serpins in a range of conformational states (see Table 1). The native serpin fold has been studied in detail and has revealed many of the molecular mechanisms involved in controlling metastability and proteinase inhibition, as well as the folding and misfolding pathways of the family.1,2 Crystallography trials of several disease-causing serpin variants have been initiated; however, few have yet resulted in high-quality data for interpretation. Classification of serpin conformations is complicated by the fact that many members of the family display minor variations in structure, yet do not justify formation of a new category. Examples of this can be seen when comparing the various serpin native states that have been crystallographically determined, where slight differences in the location or size of secondary structural elements have been reported (see Table 1). Some of these differences will be elaborated upon subsequently in this chapter. Therefore, we first define what features can be used to categorize the various serpin states. The current pool of serpin conformations that exist are native, latent, cleaved, delta (δ), and polymer (Table 2). In addition to this, the structures 35

ch02

Jan. 10, 2007 8:57

List of all crystallized serpins in various states. Conformation

Resolution (Å)

α1 -Antitrypsin α1 -Antitrypsin α1 -Antitrypsin α1 -Antitrypsin α1 -Antitrypsin α1 -Antitrypsin α1 -Antitrypsin α1 -Antitrypsin α1 -Antitrypsin α1 -Antitrypsin α1 -Antitrypsin α1 -Antitrypsin

Human Human Human Human Human Human Human Human Human Human Human Human

1kct 1psi 1qlp 1hp7 1d5s 7api 8api 9api 1ezx 1qmb 1iz2 1jvq

Native Native Native Native Polymer Cleaved Cleaved Cleaved Cleaved, Complex Cleaved, Polymer Latent Polymer

3.46 2.90 2.00 2.10 3.00 3.00 3.00 3.00 2.60 2.60 2.20 2.60

α1 -Antitrypsin precursor α1 -Antitrypsin precursor Antithrombin III (β) Antithrombin III (α) Antithrombin III Antithrombin III

Human Human Human Human Human Human

1oo8 1oph 1e04 1e04 1azx 2ant

Native Native Native/Latent Native Native/Latent Native/Latent

2.65 2.30 2.60 2.60 2.90 2.90

Ligand

Mutation

F51L

M358R E264V E264V E264V Trypsin

P14-P8 RL peptide, tetrapeptide residues 26–418 Trypsin-S195A

Pentasaccharide Heparin

Ref

80 23 78 81 17 22 22 22 53 16 40 82

83 83 36 36 44 36 (Continued)


PDB entry

M. C. Pearce et al.

Species of origin

SPI-B429

Molecule

36

Table 1

ch02

Jan. 10, 2007 8:57

Molecule

(Continued)

Conformation

Resolution (Å)

Ligand

Antithrombin III Antithrombin III Antithrombin III

Human Human Human

1dzg 1dzh 1r1l

Native/Latent Native/Latent Native/Cleaved

2.80 2.80 2.70

Antithrombin III Antithrombin III Antithrombin III

Human Human Human

1ath 1e03 1lk6

Native Native/Latent Polymer

3.20 2.90 2.90

Antithrombin III Antithrombin III Antithrombin III Plasminogen Activator Inhibitor – 1 Plasminogen Activator Inhibitor – 1 Plasminogen Activator Inhibitor – 1 Plasminogen Activator Inhibitor – 1 Plasminogen Activator Inhibitor – 1

cow Human Human

1att 1sr5 1tb6

Cleaved Complex Complex

3.20 3.10 2.50

Human

1b3k

Native

2.99

72

Human

1dvm

Native

2.10

63

Human

1db2

Native

2.70

Somatomedin

18

Human

1oc0

Native

2.28

Somatomedin

77

Human

1dvn

Latent

2.10

P14-fluorescein P14-P9 RL peptide, tetrapeptide Pentasaccharide P14-P9 RL peptide, tripeptide Anhydrothrombin, heparin Thrombin, heparin

Mutation

Ref

N135Q, S380C N135Q, S380C

84 84

85 44

49 51 50

63 37

(Continued)


PDB entry

Serpin Conformations

Species of origin

SPI-B429

Table 1

ch02

Jan. 10, 2007 8:57

Conformation

Resolution (Å)

Human

1c5g

Latent

2.60

86

Human

9pai

Cleaved

1.60

87

Human

1by7

Native

2.00

Human Cowpox virus Cowpox virus Cowpox virus Human Human Human Human Human Human Human Human Mouse

1jrr 1c8o 1m93 1f0c 1jmj 1jmo 1as4 3caa 4caa 1qmn 1lq8 1m6q 1m37

Native Cleaved Cleaved Cleaved Native Complex Cleaved Cleaved Cleaved δ Cleaved Native Native?

1.60 2.90 1.65 2.25 2.35 2.20 2.10 2.40 2.90 2.27 2.40 model model

Ligand

peptide

Mutation

66–98

88

66–98

89 90 90 91 37 37 92 93 93 14 94 95

Res. 56–300 Res. 1–305 Thrombin-S195A

Ref

13+N-terminal A349R A347R T345R

(Continued)


PDB entry

M. C. Pearce et al.

Plasminogen Activator Inhibitor – 1 Plasminogen Activator Inhibitor – 1 Plasminogen Activator Inhibitor – 2 Plasminogen Activator Inhibitor – 2 Crma Crma Crma Heparin cofactor II Heparin cofactor II α1 -Antichymotrypsin α1 -Antichymotrypsin α1 -Antichymotrypsin α1 -Antichymotrypsin Protein C inhibitor C1 esterase inhibitor Placental thrombin inhibitor

Species of origin

(Continued)

SPI-B429

Molecule

38

Table 1

ch02

Jan. 10, 2007 8:57 SPI-B429

(Continued)

Species of Origin

PDB entry

Conformation

Resolution (Å)

Ovalbumin s-Ovalbumin Ovalbumin Neuroserpin PEDF Maspin Maspin Thermopin Thermopin Serpin K Serpin K Serpin a3n

Chicken Chicken Chicken Mouse Human Human Human T. fusca T. fusca M. sexta M. sexta Mouse

1ova 1uhg 1jti 1jjo 1imv 1xqg 1xqj 1mtp 1sng 1sek 1k90 1yxa

Uncleaved Uncleaved Loop-inserted mutant Cleaved Native Native Native Cleaved Native Active Complex Native

1.95 1.90 2.30 3.06 2.85 3.10 3.10 1.50 1.80 2.10 2.30 2.10

Ligand

Mutation

R339T

Trypsin

Ref

11 96 97 98 99 100 100 6 38 101 52 5


Molecule

39


Table 1

ch02

Jan. 10, 2007

8:57

SPI-B429


40

M. C. Pearce et al. Table 2 Structural and functional features of the various serpin folds.

Feature Monomeric Intact Inhibitory Relative stability Exposed RCL

Native

Latent

Cleaved

Delta

Polymer

Yes Yes Yes + Yes

Yes Yes No +++ No

Yes No No +++++ No

Yes Yes No ++ Partial

No Either No ++++ Either

Schematic representation of serpin conformations

The schematics illustrate the different serpin folds. Specific structural elements that differ between states are highlighted: A β-sheet (red), reactive center loop (RCL; green), s1C (orange), and the F-helix (magenta).

of several serpins in complex with other molecules ranging from short polypeptides to proteinases have been solved (Table 1). Each serpin conformation is not only defined by specific structural features, but also according to properties such as function and stability. Table 2 describes the properties that we have chosen to use in order to aid in the classification of the many serpin structures that are formed. The native serpin state generally represents the active conformation of the protein. In this conformation, the serpin molecule, is intact, monomeric, and in most cases, active as a proteinase inhibitor. The native state is also the least stable intact conformation, as demonstrated by a range of biochemical and biophysical techniques, which demonstrate that the average melting temperature for a human serpin is 58◦ C.3–10 Structurally, the best feature that can be used to define the native state is partial or complete exposure of the reactive center loop (RCL). Not all serpins meet each of these criteria. These serpins can be referred to as adopting a pseudo-native conformation, such as that adopted by the chicken serpin, ovalbumin.11 Latent serpin molecules share many structural features with the native protein; however, there are some marked differences in the properties of the protein that can be used to classify this conformation (Table 2). Similar to

ch02

Jan. 10, 2007

8:57

SPI-B429



41

native serpins, the latent conformation is monomeric and intact, but it is not active as a proteinase inhibitor. This is due to the complete insertion of the RCL into the A β-sheet, which in turn removes a strand from the adjacent C β-sheet. The latent conformation demonstrates greater stability than the native state, and is the most stable state an intact serpin can typically adopt, displaying melting temperatures between 85 and 90◦ C (Ref. 9 and unpublished observations, Pearce and Bottomley). The definition of the latent state indicates that the native conformation can be reformed through denaturation and subsequent renaturation of the protein. To date, most latent serpins studied conform to each of these properties, yet not all latent serpins refold back to the native state following denaturation (unpublished observations, Pearce and Bottomley). By definition, the cleaved serpin conformation must include a proteolyzed site (Table 2). However, to avoid confusion, the cleaved state described here applies only to those proteins where the cleaved bond is located within the RCL. Cleavage at this location leads to complete insertion of the RCL into the A β-sheet, as in the latent conformation, but strand 1C from the C β-sheet remains in place. Cleaved serpins are monomeric and highly stable, representing the most stable conformation available for this family, with a melting temperature typically >110◦ C.12,13 As the RCL is completely inserted, the cleaved state is not active as a proteinase inhibitor. Only one protein has been characterized in the δ-conformation,14 yet it is hypothesized that most serpins adopt this state during times of conformational change such as folding or misfolding. The δ-conformation is monomeric and intact, and is not active as a proteinase inhibitor and demonstrates partial insertion of the RCL. The remaining space within the A β-sheet is occupied by a β-strand consisting of the partially unwound F-helix and the loop linking this structure to s3A. The δ serpins are more stable than the native state, but not as stable as either the latent or the cleaved states.14 Serpin polymers are implicated in several disease states associated with loss of inhibitory activity and excessive deposition in tissues leading to organ damage.15 Although the structures of two serpin polymers have been determined,16,17 it is believed that there are many other possible conformations that form both in vitro and in vivo. Many polymer structures have been proposed from interactions observed in the crystal packing of

ch02

Jan. 10, 2007

8:57

SPI-B429

42


M. C. Pearce et al.

various serpins.18–20 The key feature that distinguishes a serpin polymer is the presence of more than one serpin monomer interacting with another molecule through some type of β-sheet linkage. The individual serpin molecules may be in a variety of conformations, which will be discussed in detail later in this chapter.

2. Flexibility of the Serpin Native State The serpin native state can best be defined as the conformation in which the protein is active as a proteinase inhibitor. Minor differences exist between the native states of various inhibitory serpins although, all share a common inherent molecular tension that provides the energy for the dramatic conformational changes associated with proteinase inhibition. Serpins are composed of a mixture of α-helices, β-sheets, β-turns, and loops (Fig. 1). There are three central β-sheets that constitute the core of the molecule, while the helices and loops surround these structures, consolidating the tertiary structure. The B β-sheet lies in the core of the molecule, and is the region that is first formed during folding8,21 (blue; Fig. 1). Composed of six anti-parallel β-strands, this is the most stable of the three sheets and acts as a scaffold for the rest of the secondary structural elements to fold around. Each of the β-strands is numbered as s1B, s2B, and so forth (Fig. 1). Lying perpendicular to the B β-sheet is the five stranded A β-sheet (red; Fig. 1). The first crystal structure of a serpin to be determined was that of cleaved α1 -AT,22 which revealed a six stranded anti-parallel A β-sheet. The two sides of the cleavage site, between adjacent residues within the RCL, were separated by 70 Å, one side of which was located at the end of strand s4A within the A β-sheet. Additional structures determined later revealed that in the native serpin state, theA β-sheet was composed of only five strands, with strands s3A and s5A lying parallel to each other, while all other strands were anti-parallel.11,23 This finding highlighted the role of s4A as the location of the reactive center. In the native state of all serpins, the RCL (green; Fig. 1) is exposed from the body of the molecule; however, differences in the extent to which this loop is exposed have been identified for several serpins. In some cases, these changes are related to inhibitory activity, as seen later. The C β-sheet is also composed of five strands, lying in close contact with the upper region of the B β-sheet (orange; Fig. 1). It has been previously

ch02

Jan. 10, 2007

8:57

SPI-B429



43

Fig. 1. Crystal structure of α1 -AT (1qlp) demonstrates the classical serpin fold. The A β-sheet is shown in red, B β-sheet in blue, and the C β-sheet in orange. The F-helix is in purple, and the RCL is in green. All other loops and helices are off-white. Regions of high flexibility are highlighted by white circles.

proposed that the C β-sheet is also a part of the serpin core that directs correct folding and prevents misfolding; this region is often referred to as the B/C barrel.2,15 Several mutations within this region have been characterized, and in many cases, these affect either the stability or the function of the protein.15 The RCL is an important component of the serpin structure, as it is highly mobile and contains the scissile bond cleaved by the proteinase during inhibition. The conformation adopted by the RCL varies from protein to protein, according to crystallographic data, as it is largely exposed from the body of the molecule. Several serpin structures within the Protein Databank (PDB) do not have a defined RCL structure, highlighting the flexibility of this loop. In native α1 -AT, the RCL is completely exposed, located as a loop linking the A and C β-sheets through strands s5A and s1C23 (Fig. 1). In this conformation, the scissile bond is readily available for proteinase cleavage. The mobility of the RCL is central to both function and (concomitantly)

ch02

Jan. 10, 2007

8:57

SPI-B429

44


M. C. Pearce et al.

dysfunction, and can adopt a range of conformations even within the realms of native serpin structures. Proteinase inhibition by a serpin requires conformational change and several regions of the serpin fold demonstrate flexibility to accommodate this.24,25 These regions include the “hinge”, ‘breach”, “gate”, and “shutter” (Fig. 1). At either end of the RCL are the “hinge” regions. The proximal hinge, located seven residues N-terminal to the scissile bond, is comprised of residues P15 -P9 and is the portion of the RCL that undergoes the greatest conformational change during proteinase inhibition, as the amino acids here must flex in order to allow RCL insertion into the A β-sheet.26 Residues in this region tend to possess small side chains, which permit the flexibility required of the area.26 The distal hinge is located N-terminal to the cleavage site and comprises the loop linking the RCL with the C β-sheet. Several studies have highlighted the importance of mobility within the distal hinge in relation to efficient loop insertion during proteinase inhibition.27–31 The breach (Fig. 1) encompasses the initial point of insertion of the RCL into the A β-sheet.25 There are many highly conserved residues within this region, most of them being buried within the hydrophobic core and interacting with other conserved residues stabilizing the area.32 A number of mutations located within the breach have been identified, including the wellcharacterized Glu342 Lys mutation in Z α1 -AT, which can lead to misfolding and disease, highlighting the importance of maintaining structural integrity in this region.15,33 The gate is composed of a turn in the C β-sheet between strands s1C and s2C (Fig. 1). Mobility of s1C is essential to structural movements that involve the RCL, which is accommodated by the gate.27 Such a conformational change occurs in the transition to the inactive latent state, as observed for the serpins plasminogen activator inhibitor 1 (PAI-1)34 and antithrombin III (ATIII).20 One region especially sensitive to detrimental mutations is the shutter region (Fig. 1). The shutter incorporates regions of the A β-sheet that are involved in the conformational change associated with RCL insertion.24,25 Involving many highly conserved residues, the shutter must also be flexible to allow the β-sheet to open, the helix F to distort, and the RCL to be inserted.32 While there are many common elements found in the different serpin structures, crystallographic studies have highlighted the possibility that

ch02

Jan. 10, 2007

8:57

SPI-B429



45

native serpins need not all conform to a single structural model. In particular, small differences in the location of the RCL have been identified between native serpin structures. Figure 2 shows common examples of RCL conformations identified in native serpin states. The characteristic, canonical state is seen for α1 -AT (Fig. 2a), where the loop is fully expelled from the body of the molecule. In this state, the scissile bond is optimally presented for cleavage. The noninhibitory serpin ovalbumin adopts a pseudo-native state, which presents a helical RCL (Fig. 2b) that is also fully exposed.35 Several serpins have been crystallized with partially inserted RCLs, as represented here by ATIII, where insertion involves both the P14 and P15 residues36 (Fig. 2c). When partial insertion occurs, the A β-sheet adopts a slightly more open/expanded conformation in order to accommodate this. Serpins that possess a partially inserted RCL, such asATIII, heparin cofactor II (HCII), and the murineACT homologue Serpin a1n, often show some disruption to the A β-sheet.5,20,37 As seen in Fig. 2, the region of the A β-sheet closest to the RCL is expanded slightly, allowing access for the loop. Generally, serpins that show partial insertion of their RCL possess longer loops than α1 -AT, which may be required to allow this to occur without disruption of the adjoining C β-sheet. As the RCL is highly flexible, many crystallographic data sets for a number of serpins lack density for the RCL, so that its orientation cannot be determined. Thermopin (Fig. 2d) is a bacterial serpin that shares many features with other members of the family, even though the RCL cannot be seen in its native crystal structure.38 Other structural features that can differ between native states include the absence of conserved secondary structural elements and the presence of new features. Due to the metastable nature of the native fold, all inhibitory serpins possess inherent tension in order to drive inactivation of the proteinase; however, this can predispose the proteins to unwanted conformational changes that mimic inhibition. It is exciting that inhibitory serpins are found within extremophiles, organisms that live in conditions that would certainly induce these inappropriate changes in most human serpins. Two crystal structures have been solved for just such a serpin, thermopin, found in the Archaeon Thermobifida fusca.39 Thermopin has been solved in both the cleaved and native states,6,38 both of which conform to most rules regarding serpin conformation as defined earlier (Table 2). In both states, thermopin lacks the

ch02

Jan. 10, 2007

8:57

SPI-B429

46


M. C. Pearce et al.

Fig. 2. Crystal structures of various serpins in their native state. Native α1 -AT (A; 7api), psuedo-native ovalbumin (B; 1ova), native ATIII (C; 2ant), and native thermopin (D; 1sng) can be seen in this figure. The A β-sheet is shown in red, the B β-sheet in blue, and the C β-sheet in orange. The F-helix is in purple, and the RCL is in green. All other loops and helices are off-white.

ch02

Jan. 10, 2007

8:57

SPI-B429



47

G-helix, one of the highly conserved secondary elements that make up the serpin structure (Fig. 2d). In addition to this, thermopin possesses an additional C-terminal tail that interacts with the A β-sheet, as seen in yellow in Fig. 2d. It was initially proposed that this element was the key as to why this protein was both inhibitory and able to withstand extreme temperatures; however, experimental studies have shown that this was not so.6 Studies of thermopin were initiated to gain a better understanding of the basis of serpin metastability. The information gained, however, highlighted the high conservation of the native serpin state, without providing any hints as to what really regulates the stability of the molecule. Extensive studies of other serpins have demonstrated that there are many sites spread throughout the body of the molecule that control this feature, and disruption of any one of these can be detrimental to the protein.40,41 Non-inhibitory serpins, such as ovalbumin, often share high structural similarity with their inhibitory cousins, yet there are small differences that account for their varied properties. Although the RCL of ovalbumin differs from most other members of the family (Fig. 2b), the basis for its non-inhibitory nature lies in the amino acid sequence of the hinge region. The residues located within the hinge of ovalbumin prevent gross conformational change, reducing the flexibility of the overall structure. The inhibitory potential of ovalbumin is reduced through a slowing of the rate of loop conformational change associated with loop insertion. This in turn allows the complex between proteinase and serpin to dissociate before a stable covalent complex is formed.42,43 Binding of co-factors can alter the native serpin state, which in the case of ATIII is required for some of its intrinsic roles. When alone, ATIII presents a partially inserted RCL; however, upon interaction with heparin, this loop is expelled. Biochemical studies show that expulsion and the overall conformational change within the molecule increases the affinity of ATIII for proteinases, such as factor Xa. ATIII crystallized in the presence of heparin reveals that the binding site for the polysaccharide is in a pocket alongside the D-helix, the A-helix, and the N-terminus.44 Upon binding, the D-helix is elongated by two turns (cyan, Fig. 3), while a new I-helix is formed perpendicular to this (yellow, Fig. 3). These structural changes are translated through the rest of the molecule, resulting in expulsion of the RCL.

ch02

Jan. 10, 2007

8:57

SPI-B429

48


M. C. Pearce et al.

Fig. 3. Crystal structure of native ATIII alone (A; 2ant) and bound to heparin pentasaccharide (B; 1e03). The A β-sheet is shown in red, B β-sheet in blue, and the C β-sheet in orange. The F-helix is in purple, and the RCL is in green. The D-helix with extension is shown in cyan, with the I-helix in yellow. Heparin is shown in brown. All other loops and helices are off-white.

3. Serpin Structures During Proteinase Inhibition The metastability of the native state is central to the inhibitory activity of serpins. A dramatic conformational change is initiated upon interaction with a proteinase, which results in adoption of the hyper-stable cleaved state (Table 1).12 Serpins cleaved by their target proteinase are designed to form irreversible complexes, but some complexes do break down over time, releasing the cleaved serpin and an active enzyme45,46 (Fig. 4). However, these complexes are often taken up by cellular receptors prior to degradation. Interaction with other proteinases can lead to cleavage at other locations within the RCL, which can lead to adoption of a very similar structure to that released from the complex; but the extent to which the RCL is inserted into the A β-sheet differs.47

ch02

Jan. 10, 2007

8:57

SPI-B429



49

Fig. 4. Crystal structure of cleaved α1 -AT (7api). The A β-sheet is shown in red, B β-sheet in blue, and the C β-sheet in orange. The F-helix is purple, while the RCL is in green. All other loops and helices are off-white.

When the RCL of most serpins is cleaved, a large conformational change is initiated, resulting in the incorporation of the RCL into the A β-sheet (Fig. 4). The majority of the serpin molecule, such as the B and C β-sheets, and many of the helices, still resembles that of the native state. The new s4A forms hydrogen bonds with the neighboring strands, s3A and s5A, which in turn stabilizes the A β-sheet, thus conferring greater stability to the protein. Cleaved structures of many serpins have been determined (see Table 1) with few apparent differences between proteins.6,22,48,49 This is confirmed by a comparison of all the cleaved serpin structures available, as reflected by R.M.S. deviations ranging between 1.01 and 1.63. Those differences that do appear are most likely attributable to sequence-specific differences in the structural elements of each serpin. The mechanism of proteinase inhibition by serpins has been studied intensively for many years, resulting in a proposed mechanism that involves a number of conformational intermediates, although such numbers are

ch02

Jan. 10, 2007

8:57

SPI-B429

50


M. C. Pearce et al.

unclear and under debate. Two of these conformationally distinct complexes have been characterized by crystallography; the encounter complex or Michaelis complex, and the covalent acyl complex. The Michaelis complex represents the initial structure adopted when the proteinase and inhibitor first interact. Three Michaelis complex structures have been solved, with two independent studies characterizing a single union, that of antithrombin with thrombin that is stabilized by heparin50,51 (Fig. 5). Two other Michaelis complexes include that formed between serpin 1B from Manduca sexta and trypsin52 (Fig. 5), as well as the heparin cofactor II (HCII) and thrombin union37 (Fig. 5). Three of the four crystals were solved by modifying the proteinase, which prevented proteolysis of the scissile bond, essentially freezing the inhibitor and enzyme in this form. The RCL remains intact in all of the Michaelis complexes solved with no insertion apparent, indicating that this conformational change only occurs once cleavage is initiated. The conformation of the RCL appears to be dependent on the serpin/proteinase pair; however, the solved structures suggest that the RCL forms tight interactions with the proteinase through sites other than the P1 residue. The location of the proteinase in each complex appears to differ, depending on the enzyme/inhibitor pair, as seen in Fig. 5. The HCII forms an intimate partnership with thrombin in this complex, as the proteinase is in close contact with the bulk of the molecule, while trypsin remains distant from its inhibitor (Fig. 5). The intimate relationship between HCII and thrombin is facilitated by interactions between a long tail that extends from the body of the serpin (yellow, Fig. 5) and the proteinase. A less intimate relationship is demonstrated by the other Michaelis complexes structurally characterized. In the serpin 1B/trypsin and antithrombin/thrombin structures, the proteinase is apparently held away from the serpin. This is facilitated in the latter structure through a tight association of both proteinase and inhibitor with heparin, which forms a bridge between the two molecules. All three Michaelis complexes imply that the serpin is in a native conformation at this point of inhibition; however, proteins that possess a partially inserted RCL may have undergone a conformational change in order to expel this.37 The covalent complex highlights the location of the proteinase following complete inhibition (Fig. 5). In the complex structure formed between α1 -AT and trypsin,53 the proteinase is translocated to the opposite pole

ch02

Jan. 10, 2007

8:57

SPI-B429



51

Fig. 5. Crystal structures of complexes formed during proteinase inhibition. Michaelis complexes of heparin cofactor II with thrombin (A; 1jmo); ATIII, heparin and thrombin (B; 1tb6); and Serpin 1B with trypsin (C; 1k9o). Covalent complex of α1 -AT with trypsin (D; 1ezx). A β-sheet shown in red, B β-sheet in blue, and the C β-sheet in orange. The F-helix is in purple, and the RCL is in green. In each figure the proteinase is shown in cyan. The extra tail of HCII is seen in yellow (B). All other loops and helices are off-white.

ch02

Jan. 10, 2007

8:57

SPI-B429

52


M. C. Pearce et al.

of the molecule from that which is seen in the Michaelis complexes, supporting several studies that proposed this mechanism of inhibition.53–57 Trypsin has lost 37% of its native structure in this state, bound tightly to the P1 residue and therefore held crushed against the lower portion of the A β-sheet. α1 -AT is in the cleaved conformation (Fig. 4) that is released when there is breakdown of the complex and therefore cannot inactivate any other proteinases.

4. Serpin Conformations Associated with Misfolding and Disease Serpin-associated diseases occur due to either decreased levels of active material, deposition of aggregated material in tissues, or a combination of both.58 An extensive listing of the many serpin mutations and the diseases they are associated with was published several years ago by Stein and Carrell,15 which highlighted the importance of this protein family in many disease states. Many mutations affect the functional properties of the serpin, while others lead to adoption of other inactive and more stable states such as the latent and polymeric structures (Table 1).15 In each case, the resultant levels of active protein are reduced, which can lead to many varied disorders. In particular, deposition of aggregated serpins can lead to organ damage such as brain injury as seen in familiar encephalopathy with neuroserpin inclusion bodies (FENIBs) and liver disease as observed in α1 -AT deficiency.59,60 Adoption of all disease-related conformations is thought to occur via a similar mechanism, mediated by formation of a partially unfolded structure, thought to resemble the δ-ACT structure solved in 200014 (Fig. 6). In this conformation, the s4A position is occupied, mimicking that which is seen in the cleaved and latent states, confering a greater stability to the molecule than when it is in the native state.14 Residues P14 to P12 are inserted into the upper portion of the A β-sheet, while the lower portion is filled by residues 167–170 from the F-helix (purple, Fig. 6). The partial insertion into the A β-sheet is thought to represent the first conformational change that must occur before the more stable, loop-inserted states can be adopted. The latent serpin state has been identified in vivo for a small number of serpins, namely ATIII, ACT, and PAI-1, and has been linked to diseases caused by reduced levels of active inhibitor.61–63 Other serpins can adopt

ch02

Jan. 10, 2007

8:57

SPI-B429



53

Fig. 6. Crystal structure of Leu55 Pro ACT in the δ-conformation (1qmn). The A β-sheet is shown in red, B β-sheet in blue, and the C β-sheet in orange. The F-helix is in purple and the RCL is in green. All other loops and helices are off-white.

the latent state in vitro, such as α1 -AT and neuroserpin.9,64 This conformation is more stable than the native conformation, and shares some common features with the cleaved serpin structure, as the RCL is inserted into the A β-sheet. Insertion is accommodated through conformational changes within the C β-sheet (Fig. 7). Serpin polymers have been isolated from the livers of patients suffering from α1 -AT deficiency, and can be readily formed in vitro.7,33,65,66 Destabilization of the native state through mutation is thought to be the main means by which polymers are formed in vivo, leading to a wide range of diseases, including liver cirrhosis, emphysema, and brain injury.59,67,68 The exact structure of the serpin polymer is thought to differ between proteins, and perhaps even between the conditions under which they are formed. A number of models have been proposed, several from interactions seen within crystal structures, while one polymer structure has been determined crystallographically (Fig. 8). The common basis for the serpin

ch02

Jan. 10, 2007

8:57

SPI-B429

54


M. C. Pearce et al.

Fig. 7. Crystal structure of latent ATIII (1azx). The A β-sheet is shown in red, B β-sheet in blue, and the C β-sheet in orange. The F-helix is in purple and the RCL is in green. All other loops and helices are off-white.

polymer structure is the linkage mediated by β-sheet interactions; however, the β-sheets that are involved may differ. In most cases, the subunits of the polymer chain adopt the same conformation. The cleaved polymer model is the only one to have been verified crystallographically.16,17 In this polymer, the RCL of each monomer must be cleaved, which in turn becomes incorporated in the A β-sheet. As cleavage in this model occurs from C-terminal to the scissile bond, a space is created at the s4A position. This position is occupied in the polymer by the short C-terminal portion of another cleaved subunit (Fig. 8a). It is unknown whether this type of polymer can be formed in vivo. The loopA-sheet model was the first polymer structure proposed, which utilizes the promiscuity of both the RCL and the A β-sheet.33,69 In this model, the β-sheet linkage is mediated by the A β-sheet, which accepts the RCL of another molecule in the s4A position (Fig. 8b). Biochemical and biophysical evidence suggests that partial disruption of the C β-sheet is also required for this type of interaction.70,71 Electron micrographs of polymers,

ch02

Jan. 10, 2007

8:57

SPI-B429



55

Fig. 8. Structures and models of proposed serpin polymers. (A) Cleaved polymer structure (1b3k); (B) Loop A sheet dimer model; (C) loop C sheet dimer model; (D) s7A polymer dimer (1qmb); (E) disulfide linked dimer — cis conformation; (F) disulfide linked dimer — trans conformation. The A β-sheet is shown in red, the C β-sheet is shown in orange, and the RCL is shown in green. The cysteine residue is in yellow, where relevant.

ch02

Jan. 10, 2007

8:57

SPI-B429

56


M. C. Pearce et al.

extracted from the livers of patients with α1 -AT deficiency, appear as long chains, resembling beads on a string and are thought to represent loop A sheet polymers. Another popular polymer model, termed the loop C-sheet model, was proposed from the interactions seen in the first crystal structure of native and latent ATIII.20 These polymers are thought to adopt shorter chains, as they become capped by the presence of a latent molecule without an exposed RCL. Loop C sheet polymerization is thought to be propagated by a portion of the RCL interacting with the vacant s1C position within the C β-sheet (Fig. 8c). Strand 1C is thought to be removed either by adoption of the latent conformation, or through partial insertion, similar to that seen in the δ-conformation (Fig. 6). Some biochemical studies suggest that the Mmalton and Siiyama variants of α1 -AT adopt this type of polymer, as they appear to be shorter than those proposed to form via the loop A-sheet model. Crystal packing of the native PAI-1 structure revealed a new potential polymer model, which also involves interactions mediated by the A β-sheet.72 In this structure, the RCL of one PAI-1 molecule adopts a β-strand conformation and forms contacts with s6A of the A β-sheet in an anti-parallel manner to become an artificial s7A (Fig. 8d). Biochemical analysis of PAI-1 polymers indicates that this type of polymerization is highly unstable, and therefore may not represent a structure capable of persisting in vivo.73 The most recent polymer model proposed for α1 -AT suggests polymers that are composed of disulfide-bonded α1 -AT dimers.19 The dimers can be adopted through disulfide linkage between the free cysteine residue within each α1 -AT monomer, which may occur in either a cis (Fig. 8e) or trans conformation (Fig. 8f). The β-sheet linkage is then thought to be propagated through interaction of the face of the A β-sheet with the same structure of another molecule.

5. Stabilization of Serpin Structure Through Interaction with other Molecules Binding of cofactors and small molecules has been shown to stabilize many serpins against inappropriate conformational change, some of which have been put forward as potential therapeutic agents. Some crystal structures

ch02

Jan. 10, 2007

8:57

SPI-B429



57

Fig. 9. Various crystal structures of serpins stabilized by binding of another molecule. (A) Binary complex formed between ATIII and an RCL peptide (1r1l); (B) ATIII bound to a short peptide with glycerol bound in a pocket within the shutter (1lk6), and (C) PAI-1 with the somatomedin B domain of vitronectin bound (1db2). A β-sheet shown in red, B β-sheet in blue, and the C β-sheet in orange. The F-helix is in purple, and the RCL is in green. All other loops and helices are off-white.

of serpins with such molecules have been solved, revealing some of the molecular secrets to these types of stabilization. One such structure is that of ATIII with the RCL peptide bound in the A β-sheet (Fig. 9a). This peptide occupies the s4A position of the A β-sheet, mimicking loop insertion. By stabilizing this sheet, the peptide blocks polymerization via all mechanisms mediated by this β-sheet, as well as preventing proteinase inhibition. Smaller molecules, including glycerol and a range of naturally occurring osmolytes, have also been shown to stabilize serpins.74–76 The structure of an ATIII variant demonstrated that a single glycerol molecule can bind to a pocket behind the A β-sheet, maintaining a hydrogen bonding network that is sufficient to slow polymerization and promote refolding from the denatured state (Fig. 9b). The majority of the serpin molecules in this glycerol-bound conformation remains unchanged from that seen in the native state. In vivo binding partners can also offer stability to serpins, an example of which is seen in the pairing of PAI-1 and vitronectin. PAI-1 readily

ch02

Jan. 10, 2007

8:57

SPI-B429


58

M. C. Pearce et al.

converts from its native state to the more stable latent conformation, which is thought to aid in protecting the body from over-activity of the inhibitor. Conversion back to the active, native state can be induced through interaction with vitronectin, which is also required for efficient function. Studies have shown that it is the somatomedin B (SMB) domain of vitronectin that binds PAI-1 (Fig. 9c), so a crystal structure of this domain with the serpin was solved.77 The SMB domain interacts with PAI-1 in a location similar to that which is occupied by heparin on ATIII. It suggests that this region of the serpin structure is important in mediating or regulating conformational change. Design of other molecules to regulate unfavorable conformational changes may find success by focusing on this location, which has indeed been proposed as a potential target for drug design in the past.78,79

6. Conclusions Serpins are highly mobile proteins that are capable of gross conformational change, which can be associated with both function and dysfunction. Several regions of the serpin structure are implicated in these changes; however, the two regions that perhaps contribute the most to the differences seen are the RCL and the A β-sheet.

References 1. Whisstock JC, Pike RN, Jin L, Skinner R, Pei XY, Carrell RW and Lesk AM (2000) Conformational changes in serpins: II. The mechanism of activation of antithrombin by heparin. J Mol Biol 301:1287–1305. 2. Whisstock JC, Skinner R, Carrell RW and Lesk AM (2000) Conformational changes in serpins: I. The native and cleaved conformations of alpha(1)antitrypsin. J Mol Biol 295:651–665. 3. Lawrence DA, Olson ST, Palaniappan S and Ginsburg D (1994) Serpin reactive center loop mobility is required for inhibitor function but not for enzyme recognition. J Biol Chem 269:27657–27662. 4. Lawrence DA, Olson ST, Palaniappan S and Ginsburg D (1994) Engineering plasminogen activator inhibitor 1 mutants with increased functional stability. Biochemistry 33:3643–3648. 5. Horvath AJ, Irving JA, Rossjohn J, Law RH, Bottomley SP, Quinsey NS, Pike RN, Coughlin PB and Whisstock JC (2005) The murine orthologue

ch02

Jan. 10, 2007

8:57

SPI-B429



6.

7.

8.

9.

10.

11. 12.

13.

14.

15. 16.

17.

59

of human antichymotrypsin: A structural paradigm for clade A3 serpins. J Biol Chem. Irving JA, Cabrita LD, Rossjohn J, Pike RN, Bottomley SP and Whisstock JC (2003) The 1.5 A crystal structure of a prokaryote serpin: controlling conformational change in a heated environment. Structure (Camb) 11:387–397. Dafforn TR, Mahadeva R, Elliott PR, Sivasothy P and Lomas DA (1999) A kinetic mechanism for the polymerization of alpha1-antitrypsin. J Biol Chem 274:9548–9555. Pearce MC, Cabrita LD, Rubin H, Gore MG and Bottomley SP (2004) Identification of residual structure within denatured antichymotrypsin: implications for serpin folding and misfolding. Biochem Biophys Res Commun 324:729–735. Onda M, Belorgey D, Sharp LK and Lomas DA (2005) Latent S49P neuroserpin forms polymers in the dementia familial encephalopathy with neuroserpin inclusion bodies. J Biol Chem 280:13735–13741. Cabrita LD, Whisstock JC and Bottomley SP (2002) Probing the role of the F-helix in serpin stability through a single tryptophan substitution. Biochemistry 41:4575–4581. Stein PE, Leslie AG, Finch JT and Carrell RW (1991) Crystal structure of uncleaved ovalbumin at 1.95 A resolution. J Mol Biol 221:941–959. Bruch M, Weiss V and Engel J (1988) Plasma serine proteinase inhibitors (serpins) exhibit major conformational changes and a large increase in conformational stability upon cleavage at their reactive sites. J Biol Chem 263:16626–16630. Herve M and Ghelis C (1990) Conformational changes in intact and papainmodified alpha 1-proteinase inhibitor induced by guanidinium chloride. Eur J Biochem 191:653–658. Gooptu B, Hazes B, Chang WS, Dafforn TR, Carrell RW, Read RJ and Lomas DA (2000) Inactive conformation of the serpin alpha(1)antichymotrypsin indicates two-stage insertion of the reactive loop: implications for inhibitory function and conformational disease. Proc Natl Acad Sci USA 97:67–72. Stein PE and Carrell RW (1995) What do dysfunctional serpins tell us about molecular mobility and disease? Nat Struct Biol 2:96–113. Huntington JA, Pannu NS, Hazes B, Read RJ, Lomas DA and Carrell RW (1999) A 2.6 A structure of a serpin polymer and implications for conformational disease. J Mol Biol 293:449–455. Dunstone MA, Dai W, Whisstock JC, Rossjohn J, Pike RN, Feil SC, Le Bonniec BF, Parker MW and Bottomley SP (2000) Cleaved antitrypsin polymers at atomic resolution. Protein Sci 9:417–420.

ch02

Jan. 10, 2007

8:57

SPI-B429

60


M. C. Pearce et al.

18. Nar H, Bauer M, Stassen JM, Lang D, Gils A and Declerck PJ (2000) Plasminogen activator inhibitor 1. Structure of the native serpin, comparison to its other conformers and implications for serpin inactivation. J Mol Biol 297:683–695. 19. Marszal E, Danino D and Shrake A (2003) A novel mode of polymerization of alpha1-proteinase inhibitor. J Biol Chem 278:19611–19618. 20. Carrell RW, Stein PE, Fermi G and Wardell MR (1994) Biological implications of a 3 A structure of dimeric antithrombin. Structure 2:257–270. 21. Tew DJ and Bottomley SP (2001) Probing the equilibrium denaturation of the serpin alpha(1)-antitrypsin with single tryptophan mutants: Evidence for structure in the urea unfolded state. J Mol Biol 313:1161–1169. 22. Loebermann H, Tokuoka R, Deisenhofer J and Huber R (1984) Human alpha 1-proteinase inhibitor. Crystal structure analysis of two crystal modifications, molecular model and preliminary analysis of the implications for function. J Mol Biol 177:531–557. 23. Elliott PR, Lomas DA, Carrell RW andAbrahams JP (1996) Inhibitory conformation of the reactive loop of alpha 1-antitrypsin. Nat Struct Biol 3:676–681. 24. Stein PE and Carrell RW (1995) What do dysfunctional serpins tell us about molecular mobility and disease? Nat Struct Biol 2:96–113. 25. Whisstock JC, Skinner R, Carrell RW and Lesk AM (2000) Conformational changes in serpins: I. The native and cleaved conformations of alpha(1)antitrypsin. J Mol Biol 296:685–699. 26. Hopkins PC, Carrell RW and Stone SR (1993) Effects of mutations in the hinge region of serpins. Biochemistry 32:7650–7657. 27. Bottomley SP, Lawrenson ID, Tew D, Dai W, Whisstock JC and Pike RN (2001) The role of strand 1 of the C beta-sheet in the structure and function of alpha(1)-antitrypsin. Protein Sci 10:2518–2524. 28. Wang Q and Shaltiel S (2003) Distal hinge of plasminogen activator inhibitor-1 involves its latency transition and specificities toward serine proteases. BMC Biochem 4:5. 29. He S, Yang JC, Tsang S, Sim RB and Whaley K (1997) Role of the distal hinge region of C1-inhibitor in the regulation of C1s activity. FEBS Lett 412:506–510. 30. Bijnens AP, Gils A, Stassen JM, Komissarov AA, Knockaert I, Brouwers E, Shore JD and Declerck PJ (2001) The distal hinge of the reactive site loop and its proximity: a target to modulate plasminogen activator inhibitor-1 activity. J Biol Chem 276:44912–44918. 31. De Taeye B, Compernolle G, Dewilde M, Biesemans W and Declerck PJ (2003) Immobilization of the distal hinge in the labile serpin plasminogen

ch02

Jan. 10, 2007

8:57

SPI-B429



32.

33. 34.

35.

36.

37.

38.

39. 40. 41. 42.

43.

44.

61

activator inhibitor 1: Identification of a transition state with distinct conformational and functional properties. J Biol Chem 278:23899–23905. Irving JA, Pike RN, Lesk AM and Whisstock JC (2000) Phylogeny of the serpin superfamily: Implications of patterns of amino acid conservation for structure and function. Genome Res 10:1845–1864. Lomas DA, Evans DL, Finch JT and Carrell RW (1992) The mechanism of Z alpha 1-antitrypsin accumulation in the liver. Nature 357:605–607. Mottonen J, Strand A, Symersky J, Sweet RM, Danley DE, Geoghegan KF, Gerard RD and Goldsmith EJ (1992) Structural basis of latency in plasminogen activator inhibitor-1. Nature 355:270–273. Stein PE, Leslie AG, Finch JT, Turnell WG, McLaughlin PJ and Carrell RW (1990) Crystal structure of ovalbumin as a model for the reactive centre of serpins. Nature 347:99–102. Skinner R, Abrahams JP, Whisstock JC, Lesk AM, Carrell RW and Wardell MR (1997) The 2.6 A structure of antithrombin indicates a conformational change at the heparin binding site. J Mol Biol 266:601–609. Baglin TP, Carrell RW, Church FC, Esmon CT and Huntington JA (2002) Crystal structures of native and thrombin-complexed heparin cofactor II reveal a multistep allosteric mechanism. Proc Natl Acad Sci USA 99:11079–11084. Fulton KF, Buckle AM, Cabrita LD, Irving JA, Butcher RE, Smith I, Reeve S, Lesk AM, Bottomley SP, Rossjohn J and Whisstock JC (2005) The high resolution crystal structure of a native thermostable serpin reveals the complex mechanism underpinning the stressed to relaxed transition. J Biol Chem 280:8435–8442. Irving JA, Steenbakkers PJ, Lesk AM, Op den Camp HJ, Pike RN and Whisstock JC (2002) Serpins in prokaryotes. Mol Biol Evol 19:1881–1890. Im H, Woo MS, Hwang KY and Yu MH (2002) Interactions causing the kinetic trap in serpin protein folding. J Biol Chem 277:46347–46354. Im H and Yu MH (2000) Role of Lys335 in the metastability and function of inhibitory serpins. Protein Sci 9:934–941. Huntington JA, Fan B, Karlsson KE, Deinum J, Lawrence DA and Gettins PG (1997) Serpin conformational change in ovalbumin. Enhanced reactive center loop insertion through hinge region mutations. Biochemistry 36:5432–5440. Chaillan-Huntington CE, Gettins PG, Huntington JA and Patston PA (1997) The P6-P2 region of serpins is critical for proteinase inhibition and complex stability. Biochemistry 36:9562–9570. Jin L, Abrahams JP, Skinner R, Petitou M, Pike RN and Carrell RW (1997) The anticoagulant activation of antithrombin by heparin. Proc Natl Acad Sci USA 94:14683–14688.

ch02

Jan. 10, 2007

8:57

SPI-B429

62


M. C. Pearce et al.

45. Plotnick MI, Schechter NM, Wang ZM, Liu X and Rubin H (1997) Role of the P6-P3 region of the serpin reactive loop in the formation and breakdown of the inhibitory complex. Biochemistry 36:14601–14608. 46. Plotnick MI, Samakur M, Wang ZM, Liu X, Rubin H, Schechter NM and Selwood T (2002) Heterogeneity in serpin-protease complexes as demonstrated by differences in the mechanism of complex breakdown. Biochemistry 41:334–342. 47. Mast AE, Enghild JJ and Salvesen G (1992) Conformation of the reactive site loop of alpha 1-proteinase inhibitor probed by limited proteolysis. Biochemistry 31:2720–2728. 48. Baumann U, Huber R, Bode W, Grosse D, Lesjak M and Laurell CB (1991) Crystal structure of cleaved human α1 antichymotrypsin at 2.7 Å resolution and its comparison with other serpins. J Mol Biol 218:595–606. 49. Mourey L, Samama JP, Delarue M, Petitou M, Choay J and Moras D (1993) Crystal structure of cleaved bovine antithrombin III at 3.2 A resolution. J Mol Biol 232:223–241. 50. Li W, Johnson DJ, Esmon CT and Huntington JA (2004) Structure of the antithrombin-thrombin-heparin ternary complex reveals the antithrombotic mechanism of heparin. Nat Struct Mol Biol 11:857–862. 51. Dementiev A, Petitou M, Herbert JM and Gettins PG (2004) The ternary complex of antithrombin-anhydrothrombin-heparin reveals the basis of inhibitor specificity. Nat Struct Mol Biol 11:863–867. 52. Ye S, Cech AL, Belmares R, Bergstrom RC, Tong Y, Corey DR, Kanost MR and Goldsmith EJ (2001) The structure of a Michaelis serpin-protease complex. Nat Struct Biol 8:979–983. 53. Huntington JA, Read RJ and Carrell RW (2000) Structure of a serpin-protease complex shows inhibition by deformation. Nature 407:923–926. 54. Stratikos E and Gettins PG (1997) Major proteinase movement upon stable serpin-proteinase complex formation. Proc Natl Acad Sci USA 94:453–458. 55. Stratikos E and Gettins PG (1998) Mapping the serpin-proteinase complex using single cysteine variants of alpha1-proteinase inhibitor Pittsburgh. J Biol Chem 273:15582–15589. 56. Stratikos E and Gettins PG (1999) Formation of the covalent serpinproteinase complex involves translocation of the proteinase by more than 70 A and full insertion of the reactive center loop into beta-sheet A. Proc Natl Acad Sci USA 96:4808–4813. 57. Mellet P, Boudier C, Mely Y and Bieth JG (1998) Stopped flow fluorescence energy transfer measurement of the rate constants describing the

ch02

Jan. 10, 2007

8:57

SPI-B429



58. 59.

60.

61.

62. 63.

64.

65.

66.

67.

68.

69.

63

reversible formation and the irreversible rearrangement of the elastasealpha1-proteinase inhibitor complex. J Biol Chem 273:9119–9123. Carrell RW and Lomas DA (1997) Conformational disease. Lancet 350:134–138. Davis RL, Shrimpton AE, Holohan PD, Bradshaw C, Feiglin D, Collins GH, Sonderegger P, Kinter J, Becker LM, Lacbawan F, Krasnewich D, Muenke M, Lawrence DA, Yerby MS, Shaw CM, Gooptu B, Elliott PR, Finch JT, Carrell RW and Lomas DA (1999) Familial dementia caused by polymerization of mutant neuroserpin. Nature 401:376–379. Rodriguez-Soriano J, Fidalgo I, Camarero C, Vallo A and Oliveros R (1978) Juvenile cirrhosis and membranous glomerulonephritis in a child with alpha1antitrypsin deficiency PiSZ. Acta Paediatr Scand 67:793–796. Beauchamp NJ, Pike RN, Daly M, Butler L, Makris M, Dafforn TR, Zhou A, Fitton HL, Preston FE, Peake IR and Carrell RW (1998) Antithrombins Wibble and Wobble (T85M/K): Archetypal conformational diseases with in vivo latent-transition, thrombosis and heparin activation. Blood 92:2696–2706. Chang W-SW and Lomas DA (1998) Latent α1-antichymotrypsin. J Biol Chem 273:3695–3701. Stout TJ, Graham H, Buckley DI and Matthews DJ (2000) Structures of active and latent PAI-1: A possible stabilizing role for chloride ions. Biochemistry 39:8460–8469. Lomas DA, Elliott PR, Chang WS, Wardell MR and Carrell RW (1995) Preparation and characterization of latent alpha 1-antitrypsin. J Biol Chem 270:5282–5288. Lomas DA, Finch JT, Seyama K, Nukiwa T and Carrell RW (1993) Alpha 1-antitrypsin Siiyama (Ser53–>Phe). Further evidence for intracellular loopsheet polymerization. J Biol Chem 268:15333–15335. James EL and Bottomley SP (1998) The mechanism of alpha 1-antitrypsin polymerization probed by fluorescence spectroscopy. Arch Biochem Biophys 356:296–300. Crowther DC, Belorgey D, Miranda E, Kinghorn KJ, Sharp LK and Lomas DA (2004) Practical genetics: Alpha-1-antitrypsin deficiency and the serpinopathies. Eur J Hum Genet 12:167–172. Lomas DA, Belorgey D, Mallya M, Onda M, Kinghorn KJ, Sharp LK, Phillips RL, Page R, Crowther DC and Miranda E (2004) Polymerisation underlies alpha1-antitrypsin deficiency, dementia and other serpinopathies. Front Biosci 9:2873–2891. Schulze AJ, Baumann U, Knof S, Jaeger E, Huber R and Laurell CB (1990) Structural transition of alpha 1-antitrypsin by a peptide sequentially similar to beta-strand s4A. Eur J Biochem 194:51–56.

ch02

Jan. 10, 2007

8:57

SPI-B429

64


M. C. Pearce et al.

70. Bottomley SP, Hopkins PC and Whisstock JC (1998) Alpha 1-antitrypsin polymerisation can occur by both loop A and C sheet mechanisms. Biochem Biophys Res Commun 251:1–5. 71. Chang WS, Whisstock J, Hopkins PC, Lesk AM, Carrell RW and Wardell MR (1997) Importance of the release of strand 1C to the polymerization mechanism of inhibitory serpins. Protein Sci 6:89–98. 72. Sharp AM, Stein PE, Pannu NS, Carrell RW, Berkenpas MB, Ginsburg D, Lawrence DA and Read RJ (1999) The active conformation of plasminogen activator inhibitor 1, a target for drugs to control fibrinolysis and cell adhesion. Struct Fold Des 7:111–118. 73. Zhou A, Faint R, Charlton P, Dafforn TR, Carrell RW and Lomas DA (2001) Polymerization of plasminogen activator inhibitor-1. J Biol Chem 276:9115–9122. 74. Devlin GL, Parfrey H, Tew DJ, Lomas DA and Bottomley SP (2001) Prevention of polymerization of M and Z alpha1-Antitrypsin (alpha1-AT) with trimethylamine N-oxide. Implications for the treatment of alpha1-at deficiency. Am J Respir Cell Mol Biol 24:727–732. 75. Chow MK, Devlin GL and Bottomley SP (2001) Osmolytes as modulators of conformational changes in serpins. Biol Chem 382:1593–1599. 76. Zhou A, Stein PE, Huntington JA and Carrell RW (2003) Serpin polymerization is prevented by a hydrogen bond network that is centered on his-334 and stabilized by glycerol. J Biol Chem 278:15116–15122. 77. Zhou A, Huntington JA, Pannu NS, Carrell RW and Read RJ (2003) How vitronectin binds PAI-1 to modulate fibrinolysis and cell migration. Nat Struct Biol 10:541–544. 78. Elliott PR, Pei XY, Dafforn TR and Lomas DA (2000) Topography of a 2.0 A structure of alpha1-antitrypsin reveals targets for rational drug design to prevent conformational disease. Protein Sci 9:1274–1281. 79. Parfrey H, Mahadeva R, Ravenhill NA, Zhou A, Dafforn TR, Foreman RC and Lomas DA (2003) Targeting a surface cavity of alpha 1-antitrypsin to prevent conformational disease. J Biol Chem 278:33060– 33066. 80. Song HK, Lee KN, Kwon KS, Yu MH and Suh SW (1995) Crystal structure of an uncleaved alpha 1-antitrypsin reveals the conformation of its inhibitory reactive loop. FEBS Lett 377:150–154. 81. Kim S, Woo J, Seo EJ, Yu M and Ryu S (2001) A 2.1 Å resolution structure of an uncleaved alpha(1)-antitrypsin shows variability of the reactive center and other loops. J Mol Biol 306:109–119.

ch02

Jan. 10, 2007

8:57

SPI-B429



65

82. Zhou A, Stein PE, Huntington JA, Sivasothy P, Lomas DA and Carrell RW (2004) How small peptides block and reverse serpin polymerisation. J Mol Biol 342:931–941. 83. Dementiev A, Simonovic M, Volz K and Gettins PG (2003) Canonical inhibitor-like interactions explain reactivity of alpha1-proteinase inhibitor Pittsburgh and antithrombin with proteinases. J Biol Chem 278:37881– 37887. 84. Huntington JA, McCoy A, Belzar KJ, Pei XY, Gettins PG and Carrell RW (2000) The conformational activation of antithrombin. A 2.85-A structure of a fluorescein derivative reveals an electrostatic link between the hinge and heparin binding regions. J Biol Chem 275:15377–15383. 85. Schreuder HA, de Boer B, Dijkema R, Mulders J, Theunissen HJ, Grootenhuis PD and Hol WG (1994) The intact and cleaved human antithrombin III complex as a model for serpin-proteinase interactions. Nat Struct Biol 1:48–54. 86. Tucker HM, Mottonen J, Goldsmith EJ and Gerard RD (1995) Engineering of plasminogen activator inhibitor-1 to reduce the rate of latency transition. Nat Struct Biol 2:442–445. 87. Aertgeerts K, De Bondt HL, De Ranter C and Declerck PJ (1995) Crystallization and X-ray diffraction data of the cleaved form of plasminogen activator inhibitor-1. Proteins 23:118–121. 88. Harrop SJ, Jankova L, Coles M, Jardine D, Whittaker JS, Gould AR, Meister A, King GC, Mabbutt BC and Curmi PM (1999) The crystal structure of plasminogen activator inhibitor 2 at 2.0 A resolution: Implications for serpin function. Struct Fold Des 7:43–54. 89. Jankova L, Harrop SJ, Saunders DN, Andrews JL, Bertram KC, Gould AR, Baker MS and Curmi PM (2001) Crystal structure of the complex of plasminogen activator inhibitor 2 with a peptide mimicking the reactive center loop. J Biol Chem 276:43374–43382. 90. Simonovic M, Gettins PGW and Volz K (2000) Crystal structure of viral serpin crmA provides insights into its mechanism of cysteine proteinase inhibition. Protein Sci 9:1423–1427. 91. Renatus M, Zhou Q, Stennicke HR, Snipas SJ, Turk D, Bankston LA, Liddington RC and Salvesen GS (2000) Crystal structure of the apoptotic suppressor CrmA in its cleaved form. Struct Fold Des 8:789–797. 92. Lukacs CM, Rubin H and Christianson DW (1998) Engineering an anion-binding cavity in antichymotrypsin modulates the “spring-loaded” serpin-protease interaction. Biochemistry 37:3297–3304.

ch02

Jan. 10, 2007

8:57

SPI-B429

66


M. C. Pearce et al.

93. Lukacs CM, Zhong JQ, Plotnick MI, Rubin H, Cooperman BS and Christianson DW (1996) Arginine substitutions in the hinge region of antichymotrypsin affect serpin beta-sheet rearrangement. Nat Struct Biol 3:888–893. 94. Huntington JA, Kjellberg M and Stenflo J (2003) Crystal structure of protein C inhibitor provides insights into hormone binding and heparin activation. Structure (Camb) 11:205–215. 95. Bos IG, Hack CE and Abrahams JP (2002) Structural and functional aspects of C1-inhibitor. Immunobiology 205:518–533. 96. Yamasaki M, Takahashi N and Hirose M (2003) Crystal structure of Sovalbumin as a non-loop-inserted thermostabilized serpin form. J Biol Chem 278:35524–35530. 97. Yamasaki M, Arii Y, Mikami B and Hirose M (2002) Loop-inserted and thermostabilized structure of P1-P1 cleaved ovalbumin mutant R339T. J Mol Biol 315:113–120. 98. Briand C, Kozlov SV, Sonderegger P and Grutter MG (2001) Crystal structure of neuroserpin: A neuronal serpin involved in a conformational disease. FEBS Lett 505:18–22. 99. Simonovic M, Gettins PG and Volz K (2001) Crystal structure of human PEDF, a potent anti-angiogenic and neurite growth-promoting factor. Proc Natl Acad Sci USA 98:11131–11135. 100. Al-Ayyoubi M, Gettins PG and Volz K (2004) Crystal structure of human maspin, a serpin with antitumor properties: Reactive center loop of maspin is exposed but constrained. J Biol Chem 279:55540–55544. 101. Li J, Wang Z, Canagarajah B, Jiang H, Kanost M and Goldsmith EJ (1999) The structure of active serpin 1K from Manduca sexta. Struct Fold Des 7:103–109.

ch02

Jan. 10, 2007

10:36 SPI-B429


3 Mechanisms of Serpin Inhibition Peter G. W. Gettins

1. Introduction Although members of the widely distributed families of small protein proteinase inhibitors, exemplified by the Kazal, Kunitz, , and Bowman– Birk inhibitors, possess folds that are distinct to their respective families, they share common features that are critical to their mechanism of inhibition.1 All have an exposed loop that contains the target site of interaction with target proteinase. This loop is believed to be held relatively rigidly by the body of the inhibitor in an extended canonical conformation, and to undergo little or no conformational change in interacting optimally with the target proteinase. All form complexes with proteinase that are non-covalent and therefore reversible, with affinities determined principally by the interactions of residues within P6-P6 with corresponding sub-sites on the proteinase. Dissociation of these complexes can occur either back to the native inhibitor or to a P1-P1 cleaved form. Since the free energy of the cleaved form is often similar to that of the native protein, the cleaved form can also bind to proteinase, with reformation of the scissile bond.2 In contrast to these families of protein proteinase inhibitors, members of the serpin family employ a more complex mechanism, in which the ability of the serpin fold to undergo a dramatic rearrangement plays a key role.3 Serpins are single-use, suicide substrate inhibitors that follow a branched pathway which is outlined in Fig. 1. As with the other families of protein proteinase inhibitors, serpins possess an exposed reactive center loop (RCL) 67

ch03

Jan. 10, 2007

10:36 SPI-B429


68

P. G. W. Gettins E + I* k3 k1 E+I

EI

k2

[EI†]

k5

k-1 k4

E-I*

Fig. 1. Serpin branched pathway mechanism. Minimal reaction scheme for interaction of serpins (I) with proteinases (E), showing only well-defined essential intermediates, reactants, and products. The initial interaction is a reversible non-covalent one to form the Michaelis complex E.I with recognition of the P1 residue by the proteinase. This progresses through a tetrahedral intermediate (not shown) to an acyl intermediate (E–I†) that are normal species in the substrate pathway of serine proteinases. Once the RCL has been cleaved through formation of the acyl intermediate, insertion of the RCL into β-sheet A can occur. Two outcomes are possible for this intermediate. One involves successful completion of the substrate cleavage pathway (upper branch) with release of proteinase and cleaved serpin (I*) with rate constant k3 and the other is successful translocation of the proteinase to the distal end of the serpin and kinetic trapping of the acyl intermediate (E–I*) through distortion of the proteinase active site (overall rate of k4 ). E–I* can be hydrolyzed slowly (k5 ) to the same cleaved serpin I* and active enzyme as produced directly by the substrate pathway.

that contains residues necessary for recognition by proteinase. Thus, the initial interaction with proteinase results in the formation of a non-covalent Michaelis-like complex (E.I) that has many similarities to the end-point complex of canonical inhibitors with their target proteinases. However, unlike the canonical inhibitors, cleavage of the scissile bond and the formation of the acyl intermediate of the serine proteinase substrate pathway (E-I†) removes the constraint that keeps the RCL exposed. Instead, the much greater stability of the cleaved form that has the RCL inserted into β-sheet A promotes a rapid insertion of the cleaved loop into the sheet, with concomitant translocation of the proteinase to the distal end of the serpin (E-I*). The length of the RCL appears to be carefully tailored to match the length of β-sheet A such that full loop insertion will result in distortion of the proteinase active site. It is this distortion of the proteinase that kinetically traps the acyl intermediate and slows down its hydrolysis to release the active proteinase and the highly stable cleaved, loop-inserted serpin (I*). Since distortion of the proteinase is the means of slowing and ultimately abolishing the catalyzed hydrolysis of the acyl intermediate, this

ch03

Jan. 10, 2007

10:36 SPI-B429


Mechanisms of Serpin Inhibition

69

intermediate has a finite opportunity for deacylation, between the time that it is first formed within the exposed RCL and the time when the RCL has inserted into β-sheet A and compressed the proteinase sufficiently to greatly reduce its catalytic efficiency. This catalyzed deacylation, to give the same products of active proteinase and cleaved, loop-inserted serpin as from the decay of the kinetically trapped complex, is therefore an alternative fate for the acyl intermediate, one that is represented by the upper branch in Fig. 1.4,5

2. Detailed Mechanism of Inhibition of Serine Proteinases by Serpins 2.1. Step 1. recognition of the RCL The initial event in proteinase inhibition by serpins is the formation of the non-covalent Michaelis-like complex. Unlike the complexes formed by members of the families of small reversible protein proteinase inhibitors, complexes formed by serpins usually lead irreversibly to the kinetically trapped covalent intermediate E-I*. Their rate of formation typically reflects the specificity of interaction between the proteinase active site and residues within the RCL, as well as any exosite interactions. While it is not meaningful to speak of a Ki for a serpin-proteinase interaction, both the second-order rate constant for inhibition and the Kd for the complex between an active site-inactivated proteinase (either an anhydroproteinase or an S195A variant) can be used as measures of the degree of complementarity between the proteinase and the specificity determinants in the serpin. In addition, specific information on the contact interactions is now available for six serpin-proteinase pairs through NMR or x-ray crystallographic studies on non-covalent Michaelis complexes. Structures of non-covalent complexes High-resolution x-ray structures have now been determined for S195A trypsin in complex with the serpins α1 -proteinase inhibitor (α1 PI, SERPINA1) Pittsburgh variant6 and Manduca sexta 1B,7 for S195A thrombin or anhydrothrombin in complex with heparin cofactor II (SERPIND1)8 or with the binary antithrombin–heparin complex,9,10 and for antithrombin (SERPINC 1) with S195A factor Xa II. In addition, NMR

ch03

Jan. 10, 2007

10:36 SPI-B429


70

P. G. W. Gettins Table 1

X-ray and NMR structures of serpin-proteinase Michaelis complexes.

Serpin α1 PI α1 PI-Pittsburgh α1 PI-Pittsburgh Manduca sexta serpin 1K Heparin cofactor II Antithrombin–heparin Antithrombin–heparin Antithrombin–heparin

Proteinase Porcine pancreatic anhydroelastase S195A bovine trypsin S195A bovine trypsin S195A rat trypsin S195A thrombin Anhydrothrombin S195A thrombin S195A factor Xa

Technique

Reference

NMR

12

NMR X-ray crystallography X-ray crystallography X-ray crystallography X-ray crystallography X-ray crystallography X-ray crystallography

11 6 7 8 9 10 11

studies have provided information on the effects of complex formation on the conformation of the proteinase or serpin for complexes of α1 PI Pittsburgh with S195A trypsin12 and of wild-type α1 PI with porcine pancreatic elastase13 (Table 1). These various complexes allow comparisons to be made between complexes formed between a common proteinase (trypsin or thrombin) and different serpins, as well as for complexes of a common serpin with different proteinases. The first striking, though not surprising, feature of the complexes is that they all involve docking of the proteinase active site with the P1 residue of the serpin RCL, as well as contacts between subsites on the proteinase and additional RCL residues on both P and P sides of the scissile bond (Fig. 2). In the case of α1 PI Pittsburgh with S195A trypsin, the contacts are limited to residues from P4 to P4 , whereas in the antithrombin-anhydrothrombinheparin ternary complex, there are contacts from P6 to P6 , as well as additional exosite interactions involving residues in the proteinase gamma loop and residues in the body of the serpin. Somewhat more surprising was the revelation that the RCLs in each of the x-ray structures adopt a canonical-like conformation in and around the scissile bond in complex with proteinase (Fig. 3). The serpin RCLs thus strongly resemble the RCLs of the smaller canonical inhibitors in both conformation and nature of the contacts with the proteinase. Although there are also x-ray structures of the free serpin for each of the four different complexes, those of antithrombin and heparin cofactor II

ch03

Jan. 10, 2007

10:36 SPI-B429



71

Fig. 2. Michaelis complexes of four serpin-proteinase pairs. X-ray structures of the non-covalent complexes of (A and E) Manduca sexta serpin 1K with S195A rat trypsin (1K9O),7 (B and F) α1 PI-Pittsburgh with S195A bovine trypsin (pdb 1OPH),6 (C and G) antithrombin in ternary complex with anhydrothrombin and a bridging synthetic heparin (1SR5),9 and (D and H) heparin cofactor II with S195A thrombin (1JMO).8 Views E and F are derived from structures A—D by rotation of 90◦ about a vertical axis. For each complex the proteinase is shown in green, and the body of the serpin in cyan. The RCLs are shown in purple, with the P1 residue in red space-filling representation. Strands s3A and s5A are shown in red for emphasis, and helix F is shown in orange. The structurally conserved C-terminal helix of the proteinase is shown in black to help establish the relative orientations of the proteinase in the different complexes. The conformations of the bodies of the serpin and the proteinase are largely the same as for the free proteins, with no insertion of the RCL into β-sheet A occurring as a result of complex formation. Although the portions of the RCL in contact with the proteinase active site and immediately adjacent to this are in canonical conformation, the connecting stalks of the RCL are flexible enough to allow wide movement of the proteinase with respect to the serpin, influenced by more distant protein–protein interactions (notably in the heparin cofactor II thrombin complex) or by packing effects in the crystal.

ch03

Jan. 10, 2007

10:36 SPI-B429

72


P. G. W. Gettins

Fig. 3. Canonical conformations of serpin RCLs. (A) Overlay of RCL residues P5–P5 from α1 PI-Pittsburgh alone (brown) (pdb 1OOH) and in non-covalent complex with S195A trypsin (cyan) (pdb 1OPH) showing nearly identical canonical backbone conformations, and thus indicating that complex formation requires no change in RCL conformation. (B) Overlay of the P5–P5 regions of the RCLs of α1 PI-Pittsburgh (cyan) and heparin cofactor II (purple) each in non-covalent complex with proteinase (S195A trypsin for α1 PI-Pittsburgh and S195A thrombin for heparin cofactor II), showing similar canonical conformations for each and even similar side chain orientations for the P1 side chains.

may not accurately represent the solution conformations for the RCLs, since the x-ray structures in each case involve dimerization of the serpin through RCL contacts.8,14 However, in the case of Manduca sexta 1K and α1 PI Pittsburgh, the x-ray structures are of monomeric serpin and are most

ch03

Jan. 10, 2007

10:36 SPI-B429



73

likely to be closer to the solution conformations. For these serpins, there is no change in RCL conformation in the immediate vicinity of the scissile bond upon complex formation, although in the complex of Manduca sexta 1K with rat S195A trypsin there are contacts between the distal portion of the RCL and the proteinase that cause an alteration in the RCL conformation around P10.7 A second striking feature of all of the x-ray structures of Michaelis complexes is the absence of any insertion of the RCL into β-sheet A. For a long time there had been controversy over the nature of the RCL interaction with target proteinase in the Michaelis complex, with suggestions that partial loop insertion in the vicinity of the hinge region of the RCL was necessary to “stretch” and constrain the loop into a canonical conformation and hence promote complex formation analogous to the complexes of small canonical inhibitors with their targets.15 It is clear from the x-ray structures that the RCLs adopt a canonical conformation without any need for such loop insertion. NMR of 15 N-alanine labeled serpin has also shown (in solution) that there is no partial insertion of the RCL, both for a complex for which there is an equivalent x-ray structure (α1 PI Pittsburgh with S195A trypsin)12 and for the related complex of α1 PI with anhydro-elastase.13 The x-ray crystallography and NMR of complexes have also shed light on the changes that occur within the proteinase upon forming Michaelis complexes with serpins. In all of the complexes examined, the proteinase undergoes no changes within the proteinase fold or at most minor alterations in surface loops that directly contact the serpin. This lack of structural change is manifested in very small rmsd between complexed and free proteinase and in a lack of chemical shift perturbation for backbone amide resonances in NMR spectra of free and complexed proteinase. The catalytic residues also remain unperturbed in the x-ray structures, consistent with their lack of involvement in this initial stage of complex formation. This behavior again parallels that of complex formation between small canonical proteinase inhibitors and their target proteinases.1 Although only a limited number of serpin-proteinase Michaelis complexes have so far been examined by direct structural means, i.e x-ray crystallography or NMR spectroscopy, common features have been found that might be expected to be widely, and perhaps universally, present in other serpin-proteinase Michaelis complexes. These are (i) the full exposure of

ch03

Jan. 10, 2007

10:36 SPI-B429

74


P. G. W. Gettins

the RCL with a canonical conformation flanking the scissile bond and (ii) complex formation through rigid body docking of the proteinase with the RCL, with engagement of the active site with the P1 residue and of primed and non-primed subsites on either side of the scissile bond. The principal study that maintains an alternative mechanism for Michaelis complex formation is for the complex between α1 -antichymotrypsin and chymotrypsin, for which it is proposed that partial loop insertion and movement of the proteinase occur.16,17 While this may be correct, it should be noted that the proposals are largely based on deconvolution of the time-dependence of fluorescence changes during complex formation into a number of distinct intermediate states and inference of the nature of each of these states. An unambiguous resolution of this remaining controversy will require direct x-ray or NMR structural studies on this specific complex. Determinants of specificity Although serpins function as inhibitors by kinetically trapping the acyl intermediate, E-I†, the specificity for target proteinases is manifested in the steps leading up to formation of the acyl intermediate, thus reflecting the interactions between serpin and proteinase in the Michaelis complex (KM ), as well as considerations of the rate of formation of the acyl intermediate (kcat ). It was already clear from correlations between the known target proteinases of certain serpins and the P1 residue of the serpin RCL that the primary determinant of specificity is the P1 residue. With a few exceptions, Table 2 shows that the P1 residue is as expected based on the known P1 preferences of the target proteinase. As well-known examples, antithrombin has P1 arginine and is the principal inhibitor of the argininespecific blood coagulation proteinases thrombin, factor IXa and factor Xa, while the viral serpin crmA has a P1 aspartate and inhibits aspartate specific proteinases of the caspase family, as well as granzyme B. Mutation of the P1 residue, whether occurring spontaneously in vivo or carried out in vitro by mutagenesis has produced variant serpins with inhibitory specificities that reinforce the idea of the pre-eminence of the P1 residue in determining specificity. Thus, the natural P1 methionine-to-arginine mutation of α1 PI resulted in a change in target specificity from neutrophil elastase to thrombin, arising from a 30 000-fold reduction in the rate of inhibition of elastase

ch03

Jan. 10, 2007

10:36 SPI-B429



75

Table 2 P1 residues and cognate proteinase specificity.

Common name

Serpin Systematic name

P1 residue

Target proteinases

Comments

References

α1 PI

SERPINA1

Methionine

Neutrophil elastase

18

α1ACT

SERPINA2

Leucine

Cathepsin G

18

Kallistatin SERPINA3

Phenylalanine Tissue kallikrein

19

PCI

SERPINA5

Arginine

20

Centerin

SERPINA9

Arginine

ZPI

SERPINA10 Tyrosine

Factor Xa

LEI

SERPINB1

Cysteine

Neutrophil elastase

23

PAI-2

SERPINB2

Arginine

uPA

24

SCCA-1

SERPINB3

Serine

Cathepsins K, L, and S

25

SCCA-2

SERPINB4

Leucine

Cathepsin G, chymase

26

PI6

SERPINB6

Arginine

Plasmin, chymotrypsin, cathepsin G

27

Megsin

SERPINB7

Lysine

PI8

SERPINB8

Arginine

Furin

PI9

SERPINB9

Glutamic acid

Granzyme B, subtilisin A, caspase 1

Bomapin

SERPINB10 Arginine

Epipin

SERPINB11 Lysine

Yukopin

SERPINB12 Arginine

Activated protein C

21 Requires protein Z cofactor

Not characterized

28

Arginine also at P4

29 (30, 31)

Thrombin

32 Not characterized

Trypsin

22

33 33 (Continued)

ch03

Jan. 10, 2007

10:36 SPI-B429


76

P. G. W. Gettins Table 2

Common name

Serpin Systematic name

P1 residue

(Continued) Target Comments proteinases

Hurpin

SERPINB13 Threonine Cathepsin L

Antithrombin

SERPINC1

Arginine

Heparin cofactor II

SERPIND1

Leucine

PAI-1 Protease nexin I

SERPINE1 SERPINE2

Arginine Arginine

uPA, tPA uPA, tPA, thrombin

α2 -Antiplasmin SERPINF2 C1 inhibitor SERPING1

Arginine Arginine

Neuroserpin

SERPINN1

Arginine

Plasmin C1r, C1s, kallikrein uPA, tPA, plasmin

MEPI

SERPINN2

Isoleucine

Thrombin, factor IXa, factor Xa Thrombin

References

34 Requires heparin cofactor

35

Requires heparin or dermatan sulfate cofactor

36

Heparinaccelerated rates of inhibition

37 38

39 40 41 Not characterized

42

and a 6000-fold increase in the rate of inhibition of thrombin. As a result, the individual afflicted with this mutation suffered fatal bleeding.43 Many recombinant serpins have been generated in which the P1 residue has been changed to examine its involvement in specificity and have mostly confirmed the critical importance of this residue for determining overall specificity. Two examples that appear to be contrary to expectations are the tissue kallikrein inhibitor kallistatin (SERPINA4) and the thrombin inhibitor heparin cofactor II. Both of these target arginine-specific proteinases, yet the former has a phenylalanine at P1, whereas the latter has leucine at this position. In both cases, mutation of the P1 residue to arginine gave a rate enhancement for target proteinase inhibition, with the larger effect for heparin cofactor II. The explanation for the use of an apparently non-optimal P1 residue may in both cases be to permit reaction

ch03

Jan. 10, 2007

10:36 SPI-B429



77

with the target proteinase, while sufficiently reducing the rate with other arginine-specific proteinases so that the serpin becomes highly selective. Thus, in the case of heparin cofactor II, use of both of thrombin’s exosites to permit bridging of the serpin’s N-terminal tail and of long-chain GAGs more than compensates for the poor recognition of the P1 leucine, while ensuring that other arginine-specific proteinases that cannot engage these bridging species react very poorly. Although there are a number of studies on the effects of mutations at other positions in the RCL, most notably P2 and P3, but also on the P side, most of these are at best semi-quantitative, with evaluation of relative rates of reaction by comparison of gels after a fixed period of incubation of serpin and proteinase, or else without taking into account the correction for SI. An exception was a study on the effect of the P2 residue in antithrombin on reactivity and specificity.44 The normal glycine was mutated to proline, which is a residue preferred by thrombin when cleaving peptide substrates. Although this mutation gave a 15-fold increase in reactivity towards thrombin in the absence of heparin, it both decreased the rate of reaction and increased the SI relative to wild-type antithrombin in the presence of bridging heparin. This suggests that the P2 glycine is preferred over proline to give sufficient conformational flexibility in the Michaelis complex, so that the optimal relative orientation of the three components can be achieved to give efficient inhibition. At the same time, the wild-type P2 glycine may reduce the reactivity with the anticoagulant proteinase, activated protein C, ensuring that coagulant and anticoagulant pathways are not simultaneously down-regulated. Overall, the studies on both the specificity and the reactivity of wildtype serpins and of recombinant or naturally occurring variants support the primary importance of the P1 residue and the contributory, even though lesser, importance of immediately flanking residues. This is a situation analogous to that found in small canonical proteinase inhibitors and reflects the fact that although serpins undergo subsequent essential conformational changes to effect inhibition, the specificity-determining step is analogous to that for the canonical inhibitors. While such conclusion appears to hold out the prospect of being able to design high target-proteinase specificity into a serpin through a combination of mutations in the RCL, it should be appreciated that other considerations

ch03

Jan. 10, 2007

10:36 SPI-B429

78


P. G. W. Gettins

must be taken into account concerning the compatibility of the mutations with subsequent steps, including facile loop insertion, as well as stability of the covalent complex. Thus, the nature of the covalent complex, with the RCL inserted into β-sheet A and residues P2–P7 buried behind helix F, requires that changes to residues in this region remain compatible with such burial, if efficient inhibition is to be retained. A good example of adverse effects of the introduction of mutations in this region to enhance reactivity and selectivity was an attempt to design a furin-specific serpin based on α1 PI.46 Substrate cleavage specificity of furin suggests that P1, P2, P4, and P6 arginines give optimal reactivity. However, while introduction of the P1 and P2 arginines into α1 PI gave the desired greatly enhanced reactivity, the P4 and P6 arginines gave little further increase in rate of reaction, but shifted the reaction predominantly to the substrate branch of the pathway to give an SI of 24, meaning that only 4% of Michaelis complexes were successfully converted into stable, loop-inserted covalent complex. Another example was the introduction of the P6-P2 residues of ovalbumin into α1 PI, with the result that the variant not only had greatly enhanced SI towards elastase and trypsin, but gave covalent complexes that were highly unstable and which regenerated active proteinase on a time scale of minutes.47 A final consideration in determining the specificity is the use by some serpins of exosite interactions with target proteinases. In the case of heparin co-factor II there is interaction of the N-terminal tail with exosite I of thrombin,8 while for antithrombin in complex with S195A factor Xa an exosite on antithrombin, composed of Tyr 253 and Glu 255, interacts with the autolysis loop of the proteinase.11,48 There are other examples for which there are no structural data, where rates of reaction under different conditions strongly imply a significant contribution from such interactions. Thus, the failure to completely convert α1 -antichymotrypsin into an α1 PI-like inhibitor of neutrophil elastase through swapping of increasing stretches of the α1 PI RCL into α1 -antichymotrypsin implied a 60-fold contribution from exosite interactions in the α1 -antichymotrypsin-elastase complex to the overall rate of reaction.49 2.2. Step 2. proteolysis of the scissile bond Notwithstanding early observations that the SDS-PAGE analysis of many serpin-proteinase incubations showed the formation of one or more bands of

ch03

Jan. 10, 2007

10:36 SPI-B429



79

lower mobility than the starting serpin, suggesting formation of a covalent linkage between the serpin and the proteinase, for a long period of time, there was concern that the explanation offered that the new species represented the covalent acyl intermediate of the serine proteinase substrate pathway was an artifact of the analytical method used. Thus, it was suggested that the trapped intermediate prior to denaturation was either a non-covalent Michaelis-like species or a covalent tetrahedral intermediate, and that denaturation pushed the reaction further to the acyl stage.50 Careful analysis of the intact complex in solution, however, demonstrated the presence of a new free amino terminus, strongly supporting cleavage of the scissile peptide bond and formation of an acyl intermediate.51 The final proof came from the x-ray structure of the covalent complex of trypsin with α1 PI, which gave clear electron density for the covalent intermediate between the P1 residue carbonyl and the Ser 195 γ-O.52 The same covalent linkage is also clearly visible in the second x-ray structure of a serpin-proteinase complex beetween the same serpin and pancreatic elastase.53 NMR and fluorescence studies on a number of other serpin-proteinase pairs also indicate that the same organization of serpin and proteinase, involving a translocated proteinase, occurs in all of the complexes, requiring that the scissile bond be cleaved in each case.13,54–57 The structural studies on the various non-covalent Michaelis complexes described above all showed that the initial interaction with proteinase resulted in no insertion of the RCL into β-sheet A, suggesting that subsequent loop insertion to form the covalent complex only occurred after cleavage of the scissile bond, thus implying that the attachment to β-sheet C on the P’ side acted as a constraint to loop insertion. A recent study on the effect of RCL length on the conversion of PAI-1 to the latent state showed the importance of such constraint in maintaining the loop-exposed conformation. Thus, increasing the RCL length promoted conversion to the loop-inserted (latent) state through the removal of the constraint, while RCL shortening retarded such loop insertion.58 Indeed, an earlier study involving insertion of an extra 30 residues into the RCL of α1 PI that permitted conversion to an extremely stable loop-inserted form, demonstrates that the much smaller free energy difference between active and latent states of normal serpins most probably arises as a difference between the gains from the insertion of the RCL into β-sheet A and the losses resulting from extraction of strand 1 from β-sheet C.59

ch03

Jan. 10, 2007

10:36 SPI-B429

80


P. G. W. Gettins

2.3. Step 3. Rapid loop insertion and proteinase translocation The importance of rapid loop insertion for efficient proteinase inhibition had been suggested by the occurrence of a number of dysfunctional serpins, in which mutations were present in the hinge region that might be expected to hinder such loop insertion. In particular, the presence of proline or charged residues at even-P numbered residues was found to decrease or abolish inhibitory properties of the serpin.60 The significance of the mutations being at even-numbered P residues is that these are the ones whose side chains become buried in the protein interior upon insertion of the RCL into β-sheet A. This apparent relationship between non-optimal residues in the hinge region and an adverse effect on inhibition suggested that the P14 residue, being the residue to initiate loop insertion, might be the most influential in decreasing inhibitory properties. This was shown to be the case for α1 PI, for which a change of P14 from threonine to arginine, greatly reduced the fraction of acyl intermediate species that was successfully trapped, with a corresponding increase in the fraction that generated cleaved serpin.61 A more extensive study on PAI-1, in which the P14 residue was changed to each of the other amino acids, showed that the effects seen were a function of the nature of the side chain with charged residues, whether positive or negative, having the largest negative effect on the efficiency of inhibition and the rate of loop insertion.62 The demonstrated importance of the P14 residue in influencing the rate of loop insertion, and consequently, the efficiency of proteinase inhibition, suggested that the reason that non-inhibitory serpins lack inhibitory properties might be because of insufficiently rapid loop insertion. This was first tested for ovalbumin, which has a P14 arginine. Mutation of this residue to serine greatly enhanced the rate of loop insertion,63 such that RCL cleavage was able to give a fully loop-inserted form analogous to inhibitory serpins.64 However, the rate of insertion appears to be still too slow for the variant to be a proteinase inhibitor within the limits of detection.65 A recent study suggested that the stabilizing interactions between helix F and the underlying β-sheet A raised the activation energy for the relative movement of these two secondary structural elements, which would be required to occur during loop insertion into the opened β-sheet.66

ch03

Jan. 10, 2007

10:36 SPI-B429



81

A consequence of the serpin pathway being a branched one leading to complex formation or cleaved serpin, rather than two separate pathways with the same outcomes, is that mutations that affect the overall rate of loop insertion (k4 ) or the rate of hydrolysis of the acyl intermediate (k3 ) are manifested in an alteration in the product distribution rather than in the overall rate of reaction of serpin and proteinase.61 Once the typically rate-limiting formation of the acyl intermediate has occurred between the proteinase and the exposed RCL, there is a competition between inhibitory and substrate pathways, with relative fluxes determined by the composite rate constants for each of these pathways (k3 and k4 ). A useful measure of the relative fluxes along each pathway, and hence of the efficiency of the serpin as a proteinase inhibitor, is the stoichiometry of inhibition (SI). This is the number of moles of serpin needed to generate one mole of kinetically trapped covalent serpin–proteinase complex (E-I*). The relationship of SI to the two rate constants is therefore SI = (k3 + k4 )/k4 . Efficient inhibitors are therefore those with SI values close to 1 (k4 >> k3 ), whereas predominantly substrate reactions have large SI values corresponding to k3 >> k4 . It should therefore be realized that if rate constants for inhibition are reported by measuring the rate of loss of proteinase activity rather than from total consumption of serpin, the second-order rate constant is only an apparent one that underestimates the true rate of formation of the acyl intermediate by a factor of 1/SI. It is therefore important if comparisons are to be made between the rates of acyl intermediate formation for different serpin-proteinase pairs or between different variants, that the apparent second-order rate constants be corrected for flux along the substrate pathway by being multiplied by SI.

2.4. Step 4. Proteinase crushing/distortion Evidence for active site distortion Prior to the elucidation of the x-ray structure of the trypsin-α1 PI covalent complex, there were several lines of evidence that indicated that the kinetic trap involved distortion of the proteinase active site. 1 H NMR spectroscopy on the low-field hydrogen resonance of His 57 showed a fundamental difference between chymotrypsin that was either free or in complex with standard inhibitors, and chymotrypsin in complex with serpins, suggesting a

ch03

Jan. 10, 2007

10:36 SPI-B429


82

P. G. W. Gettins

distortion of the proteinase catalytic triad in the serpin-proteinase complex.67 This was supported by a fluorescence study, which showed that the tryptophan side chain of a P1W variant of α1 PI had a very different environment in covalent complex with either trypsin or chymotrypsin than when in non-covalent complex. Since it was expected that the non-covalent complex would have the P1 side chain lodged in the S1 specificity pocket of the proteinase, this implied that progression to the covalent complex involved a change in this relationship.68 Examination of the kinetics and pHdependence of deacylation of the acyl intermediate in the covalent complex has indicated that the dominant species present contains a non-catalytically functional proteinase, from which deacylation is only possible by direct, uncatalyzed OH− -mediated hydrolysis.69 Finally, a study on the susceptibility of the acyl intermediate to nucleophilic attack supports the presence of some distortion even during translocation.70 Proteinase crushing The x-ray structure of the covalent trypsin–α1 PI covalent complex (see Fig. 4A, structure E) showed not only that the active site of the proteinase had been distorted, as expected from the results in the preceding section, but that a large part of the proteinase body had been “crushed”. The term “crush” was chosen to reflect the inability to trace defined electron density for ∼40% of the trypsin polypeptide, representing one of the two lobes of the proteinase fold.52 This suggested some plasticity of conformation caused by being “crushed” against the bottom of the serpin. Additional evidence for such plasticity was obtained for the same serpin-proteinase complex in solution using 2D NMR spectroscopy, which showed that a large part of the proteinase domain must be undergoing dynamic averaging to account for the large loss of dispersion in the 1 H dimension upon forming complex with the serpin.56 Additional evidence for the widespread occurrence of large-scale conformational alterations in the proteinase domain in the serpin-proteinase complex comes from increased susceptibility of loop regions to proteolysis. This has been characterized in greater detail for the PAI-1 complex with uPA,71 but has been observed for a number of other complexes, including antithrombin with thrombin or factor Xa,72 α1 -antichymotrypsin with chymotrypsin,73,74 and α1 PI with trypsin.75,76 Thermodynamic data

ch03

Jan. 10, 2007

10:36 SPI-B429



83

Fig. 4. Thermodynamics and structural changes in forming covalent serpin–proteinase complex. (a) Possible sequence of interactions and conformational changes associated with formation of the final covalent serpin–proteinase complex that also allows energy coupling of loop insertion with proteinase distortion. A represents the initial non-covalent Michaelis complex as well as acyl intermediate just at the point of formation and before any loop insertion has occurred. B represents an early intermediate during loop insertion and proteinase translocation, but before the proteinase has encountered opposition to movement through impingement on helix F. C represents an intermediate during displacement of helix F that permits continued passage of the proteinase, but also stores part of the energy of loop insertion in a “tensioned” helix F. D represents the penultimate conformation in which the RCL has nearly fully inserted, but before the proteinase has been “locked” at the bottom of the serpin and crushed (see the side views for D and E above the front view). (b) Schematic energy diagram for the sequence of changes depicted in panel (a). The letters A–E correspond to the conformations depicted in panel (a). Partial loop insertion in going from A to B involves a favorable G of G1 . The additional energy available from insertion of the additional RCL residues is G2 . If G4 represents the energy of displacement of helix F and G5 the energy for extracting the P1 side chain from the S1 proteinase pocket, then the energy of state D is given by G4 or G4 + G5 , depending on whether only helix F displacement or both displacement and extraction of P1 have occurred at this stage. If G3 is the total energy required to distort the proteinase, the energy of inhibition, Ginh is given by G4 + G5 − G3 , if there are no additional stabilizing interactions between serpin and proteinase in state E, or by this amount plus an additional amount G6 if there are such interactions.

ch03

Jan. 10, 2007

10:36 SPI-B429


84

P. G. W. Gettins

Fig. 4.

(Continued)

have also indicated a reduced stability for the proteinase in complex with serpin.77 consistent with significant structural distortion. Whereas large-scale proteinase distortion may be a consequence of the compression of proteinase and serpin needed to pull the P1 side chain from the S1 pocket and otherwise disrupt the proper alignment of the catalytic triad, it does not appear to be a requirement for the catalytic inactivation of the proteinase. The structural changes seen in the proteinase active site in the x-ray structures of the two covalent complexes would probably be sufficient on their own to effect kinetic trapping. Additional x-ray structures or NMR studies on different serpin-proteinase complexes will be necessary to determine whether the large-scale conformational changes are a common feature, and therefore of possible mechanistic significance. Importance of RCL length The mechanism of kinetic trapping of proteinases by serpins thus requires at least a distortion of the proteinase active site to compromise or abrogate its catalytic competence, brought about by an over-tight apposition of the

ch03

Jan. 10, 2007

10:36 SPI-B429



85

two proteins in the covalent complex, through insertion of the cleaved RCL into β-sheet A and the concomitant translocation of the proteinase to the distal end of the serpin. It is clear from such a mechanically based means of introducing active site distortion that the RCL plays an additional role in ensuring that the proteinase is held so close to the bottom of the serpin that distortion must occur to permit such positioning of the two proteins. This implies that the length of the RCL must be critical. If it is too short, the proteinase cannot reach the bottom of the serpin to be distorted, and if it is too long, there would be no compression of proteinase against serpin, and hence no distortion. These predictions have been borne out experimentally and account for the conservation of RCL length in serpins between the hinge point for insertion into β-sheet A and the P1 residue. Counting from the highly conserved glutamic acid residue that occurs just after the end of strand s5A, the RCL is 17 residues in length to the P1 residue, making the glutamic acid P17.78 With some exceptions this length of 17 residues to P1 is found for most serpins. In the viral serpin crmA, the RCL length to the P1 aspartate is only 16 residues. However, it was shown from an x-ray structure of the cleaved, loop-inserted form of crmA that this slightly shorter length matches a slightly shorter β-sheetA,79 so that the critical relationship is between the RCL length and the sheet length. A study in which the effects of RCL length were examined in the P1 arginine variant of α1 -proteinase inhibitor showed that lengthening or shortening the RCL by more than two residues completely removed the ability to trap the proteinase, while an increase of one or two residues gave very large increases in the rate of hydrolysis of the trapped acyl intermediate. Shortening the RCL by one or two residues, while giving a complex with slower rate of hydrolysis, caused the SI to increase.80 The same study showed similar effects for PAI-1 and PAI-2 complexes with tPA, while a separate study came to the same conclusion for complexes of α1 -antichymotrypsin with several proteinases.81 2.5. Need for energetic coupling The much greater stability of the cleaved forms of inhibitory serpins relative to the native state implies that loop-insertion, following cleavage of the P1-P1’ bond, is an energetically very favorable process. The few thermodynamic studies that have examined this support this idea.82,83 Furthermore,

ch03

Jan. 10, 2007

10:36 SPI-B429

86


P. G. W. Gettins

the expectation is that the favorable energetics arise from the conversion of an exposed RCL that makes few or no contacts with the serpin body into one that is an integral part of the major β-sheet. However, in order for the energy that derives from loop insertion to be available to distort the proteinase at the appropriate time, i.e., when the proteinase has been translocated to the distal end of the serpin, which must be after the RCL has inserted into β-sheet A, there must be a mean of temporarily storing some or all of this energy. A hypothesis that proposes both a means for such temporary energy storage and for coupling this to proteinase distortion in an overall energetically favorable way posits that movement of the proteinase over the face of β-sheet A necessitates temporary displacement of helix F, which lies in the path of the translocating proteinase.84 Furthermore, since this might be expected to be an energy requiring process, the displacement would serve to generate a higher-energy conformation for the helix in which some of the energy of loop insertion would be stored. Once the proteinase had passed this potential roadblock, helix F would be free to return to its lower energy conformation in a process that would also require the final step of distortion of the proteinase, though now with a favorable G (Fig. 4). Although this hypothesis has yet to be rigorously tested in terms of determining structures of intermediate states, there are several pieces of experimental data that are explained by it that would otherwise be hard to rationalize. One study involves a monoclonal antibody that discriminated between antithrombin in covalent complex and antithrombin in native, cleaved, or latent conformations.85 The epitope was furthermore shown to involve residues that are buried under helix F in all of the species. The helix F displacement mechanism can accommodate these findings by postulating that the antibody binds to the intermediate that precedes proteinase distortion in which helix F has been displaced and the epitope is exposed. Figure 4(b) shows that the affinity for the serpin in complex would be expected to be higher than that in either cleaved or latent states, since helix displacement would be coupled to the reversal of proteinase distortion, which on its own would be a favorable process. Another study showed that a PAI-1 variant in which helix F had been deleted was only capable of acting as a proteinase substrate. If helix F acted only as an impediment to proteinase translocation rather than an active component of energy coupling to enable proteinase distortion, the variant might have been expected to be an even better inhibitor than wild-type PAI-1.86 A mutagenesis study has

ch03

Jan. 10, 2007

10:36 SPI-B429



87

also shown that interacting residues in helix F and neighboring structural elements play important roles in determining the loop-insertion-dependent processes of proteinase inhibition and the transition to latency.87 2.6. Complex dissociation and stability Although the final covalent serpin-proteinase complex shown in Fig. 4(a), structure E, contains a distorted proteinase, with no catalytic activity, hydrolysis of the acyl intermediate still occurs at a measurable, though low rate. However, it should be noted, that this route to formation of cleaved serpin and regeneration of active proteinase is distinct from the direct substrate pathway, in that the former passes through the kinetically trapped complex and typically occurs very slowly (k5 ), with t1/2 on the order of days to weeks. An analysis of the pH-dependence of this deacylation, together with the effects of calcium binding to trypsin in the complex on the dissociation rates, led to the conclusion that deacylation occurs from two distinct conformations that are in equilibrium with one another.69 The dominant species contains a non-functional proteinase and is best equated with structure E of Fig. 4, while the minor species (in the absence of calcium) is capable of deacylation in a manner dependent on the catalytic triad, but at a rate that is much faster than the non-catalyzed rate, but still several orders of magnitude lower than the maximum rate of hydrolysis by the native proteinase. This species is best equated with structure D of Fig. 4, in which the proteinase has not been fully compromised, but may not be as active as the original native molecule. Nevertheless, the activity of the proteinase in conformation D is sufficient that most of the deacylation occurring for serpin-proteinase complexes at physiological pH does so from this minor species that retains a modest degree of catalytic competence. There is a twofold significance to this study. Firstly it provides additional support for the hypothesis of helix F displacement, in that it supports an equilibrium between a partly active conformation D and an inactive conformation E that favors state E. Secondly, it suggests that the stability of a given serpin-proteinase complex will depend largely on the position of equilibrium between the two states D and E. Factors that stabilize conformation E, such as better complementarity between serpin and proteinase contact surfaces should enhance stability, whereas factors that favor conformation D, such as calcium binding to trypsin, which helps

ch03

Jan. 10, 2007

10:36 SPI-B429

88


P. G. W. Gettins

rigidify it and hence makes it more energetically costly to distort, would promote deacylation. A separate study that examined the breakdown of α1 -antichymotrypsin-HNE pairs for wild-type and variant forms of the serpin was also consistent with an equilibrium between a partially active, D-like state and an inactive E-like state.88

3. Inhibition of Cysteine Proteinases Cysteine proteinases catalyze peptide bond hydrolysis by a mechanism closely analogous to that of serine proteinases, although with the obvious substitution of cysteine for serine as the attacking nucleophilic group. The above mechanism described for inhibition of serine proteinases by serpins could therefore be readily adapted to the inhibition of cysteine proteinases, by alteration of the trapped acyl intermediate from an oxyacyl species to a thioacyl intermediate. Consistent with this, there are well-documented examples of serpins inhibiting members of the cathepsin and caspase families of cysteine proteinases. What is lacking is an x-ray structure of a covalent cathepsin-serpin or caspase-serpin complex comparable to that of trypsin with α1 PI, which would reveal whether an analogous mechanism of acyl intermediate formation, loop-insertion, proteinase translocation, and distortion operates. However, it has been shown that the viral serpin crmA, which targets host caspases for inhibition, is capable of the same kind of facile loop insertion into β-sheet A upon RCL cleavage, as serpins that inhibit serine proteinases.79,89 It has also been shown for the reaction of crmA with serine proteinases that there is the same critical dependence of RCL length on inhibitory effectiveness, as seen with serpins that normally inhibit serine proteinases.90 FRET measurements on the relative positioning of serpin and proteinase for crmA and caspase 1 have also shown that the caspase appears to have been translocated to the distal end of the serpin.91 The principal reason for doubting that cysteine proteinases might be inhibited by the same conformation change-based mechanism as serine proteinases, is that no high molecular weight SDS-stable band has been demonstrated on PAGE for a serpin-cysteine proteinase complex that would correspond to a covalent thioacyl intermediate between serpin and proteinase. However, this could very easily be due to the ease of cleavage of the thioacyl intermediate in the denaturing buffers typically used to

ch03

Jan. 10, 2007

10:36 SPI-B429



89

run such gels. Thus, while the rates of hydrolysis of thioesters at physiological pH are similar to those for oxyesters, the rate of cleavage of the former by nitrogen-containing nucleophiles is much higher. As a result, unless care is taken to exclude such nucleophiles or use lower pH to reduce their efficacy, it is unlikely that a thioester that is present in the intact cysteine proteinase-serpin complex would survive the analytical procedure. This reasonable concern has been demonstrated for the complex of a cysteine proteinase with a different type of inhibitor. The x-ray structure of the complex between the inhibitor p35 and caspase 8 revealed the presence of an intact thioacyl linkage between the two proteins.89 Nevertheless, it was only possible to detect this covalent complex by PAGE when mild denaturation conditions and lower pH was used, and even then only a fraction of the protein that was known to be in complex before denaturation survived to appear in the higher MW band.

4. Reversible Inhibition 4.1. Reversible inhibition by otherwise irreversible serpin inhibitors While it is now clear from x-ray, fluorescence, and NMR studies that the normal mode of serpin inhibition of serine proteinases involves an effectively irreversible mechanism of loop insertion and proteinase translocation to give a kinetically trapped covalent intermediate, it is also clear that the formation of this intermediate, involving steps 2–4 described earlier, is preceded by the formation of the initial reversible Michaelis-like complex in step 1. As explained earlier, cleavage of the scissile bond to form the acyl intermediate is the initiating event for rapid and irreversible loop insertion in the case of serpins that can be shown to form SDS-stable complexes. Even if there are restraining interactions such as the exosite interaction of tPA on the P’ side of the RCL in PAI-1,93 these only serve to slow the rate of loop insertion slightly, rather than to prevent it. It follows that the reversibility is only practically possible before the serpin and proteinase have formed the acyl intermediate. If the proteinase forms either the tetrahedral or the acyl intermediates very slowly, the preceding complex may have a greatly prolonged existence and establish an apparent equilibrium with free proteinase

ch03

Jan. 10, 2007

10:36 SPI-B429

90


P. G. W. Gettins

that is reversible in practice. Given this requirement for very slow formation of the acyl intermediate, it is therefore not surprising that the best documented example of such reversible complex formation is between the zymogen form of uPA (single chain uPA or scuPA) and protein C inhibitor (PCI).94 Here, the zymogen is not completely inactive, but has ∼ 0.1% of the activity of the activated two-chain uPA. ScuPA is able to form complex with PCI that is reversible for long periods, though with very slow conversion to an SDS-stable high MW complex. That such reversible complex formation depends primarily on the properties of the proteinase rather than the specific features of the serpin was shown by the ability of the Pittsburgh variant of α1 -proteinase inhibitor (P1 Arg) to also form reversible complex with scuPA (unpublished results). It is less clear why chymotrypsin can form a reversible complex with α2 -antiplasmin (SERPINF2), with a dissociation rate constant of 5.5 × 10−5 M−1 s−1 that results in the regeneration of active chymotrypsin and of α2 -antiplasmin that is still functional, rather than being cleaved.75 It may be significant that chymotrypsin, not normally a target of α2 -antiplasmin, may recognize the residue adjacent to the normal P1 arginine as its P1, but may only be able to cleave this peptide bond slowly. 4.2. Reversible inhibition by ¡‘non-inhibitory” serpins The corollary of the need for appropriate hinge region residues and helix F-sheet A interactions to permit favorable and rapid loop insertion upon formation of the serpin-proteinase acyl intermediate is that serpins that have unfavorable hinge region residues or stabilizing helix F-sheet A interactions are not capable of acting as irreversible proteinase inhibitors. This has led to a common classification of serpins into “inhibitory” and “non-inhibitory,” based either on the empirically determined inability to form SDS-stable complexes with serine proteinases or the failure of RCL cleavage to give a species with greatly enhanced thermal stability, indicating a failure of the cleaved RCL to insert into β-sheet A. Of course, since effective proteinase inhibition depends on the relative rates of the loop insertion pathway and the substrate pathway, a serpin may give a cleaved loop-inserted form with enhanced stability, yet not be an inhibitory serpin by virtue of too slow a rate of loop insertion; thyroxine binding globulin and corticosteroid binding globulin are examples.

ch03

Jan. 10, 2007

10:36 SPI-B429



91

Whereas irreversible serpin inhibition requires appropriate structural features within the serpin (the examples above of reversible inhibition require appropriate low activity properties of the proteinase), there is at least one example of a serpin that would be expected to be “non-inhibitory,” yet is not only an inhibitor, but also a reversible one. This is the yellow fever mosquito serpin AFXa, which targets human factor Xa, and inhibits it in a tight reversible non-competitive manner, with a KI of 3 nM.96–98 The RCL of AFXa contains a pair of arginines located 3 and 4 residues N-terminal from where the normal position for a P1 would occur for serpin inhibition. In addition, the RCL is significantly shorter than normal. This suggests that AFXa may inhibit in a more canonical-like manner, with the shortened loop providing sufficient rigidification that it can form tight reversible complex with proteinase. The example of AFXa might suggest that other “non-inhibitory” serpins might also act as proteinase inhibitors under the right circumstances. One controversial example is maspin, whose RCL has been shown to be necessary and sufficient for the ability of maspin to induce cell adhesion and inhibit cell invasion.99 Despite the presence of glutamic acid at P10 and proline at P8, as well as the demonstration that RCL cleavage of maspin does not give a more stable form, there are reports of maspin being able to inhibit arginine-specific proteinases such as tPA and uPA,100 even though this would involve recognition of an arginine four residues closer to the hinge than normal. Equally, there are studies that maintain that maspin shows no ability to inhibit these proteinases.101 Another controversial example of proteinase inhibition by a non-inhibitory serpin is that of ovalbumin heated to 95◦ C to generate a form that can act as a tight reversible inhibitor of several proteinases.102 A concern here is that heating serpins even at much lower temperatures leads to polymerization. The significance of reversible inhibition by an ovalbumin species that might be polymerized is therefore questionable.

References 1. Bode W and Huber R (1992) Natural protein proteinase inhibitors and their interaction with proteinases. Eur J Biochem 204:433–451. 2. Estell DA, Wilson KA and Laskowski M (1980) Thermodynamics and kinetics of the hydrolysis of the reactive-site peptide bond in pancreatic

ch03

Jan. 10, 2007

10:36 SPI-B429

92

3. 4.

5.

6.

7.

8.

9.

10.

11.

12.

13.


P. G. W. Gettins

trypsin inhibitor (Kunitz) by Dermasterias imbricata trypsin 1. Biochemistry 19:131–137. Gettins PGW (2002) Serpin structure, mechanism and function. Chem Rev 102:4751–4804. Rubin H, Wang ZM, Nickbarg EB, McLarney S, Naidoo N, Schoenberger OL, Johnson JL and Cooperman BS (1990) Cloning expression, purification, and biological activity of recombinant native and variant human α1 -antichymotrypsin. J Biol Chem 265:1199–1207. Patston PA, Gettins P, Beechem J and Schapira M (1991) Mechanism of serpin action: Evidence that C1 inhibitor functions as a suicide substrate. Biochemistry 30:8876–8882. Dementiev A, Simonovic M, Volz K and Gettins PGW (2003) Canonical inhibitor-like interactions explain reactivity of α1 -proteinase inhibitor Pittsburgh and antithrombin with proteinases. J Biol Chem 278:37881–37887. Ye S, Cech AL, Belmares R, Bergstrom RC, TongY, Corey DR, Kanost M and Goldsmith EJ (2001) The structure of a Michaelis serpin-protease complex. Nat Struct Biol 8:979–983. Baglin TP, Carrell RW, Church FC, Esmon CT and Huntington JA (2002) Crystal structure of native and thrombin-complexed heparin cofactor II reveal a multistep allosteric mechanism. Proc Natl Acad Sci USA 99:11079–11084. Dementiev A, Petitou M, Herbert JM and Gettins PGW (2004) The ternary complex of antithrombin-anhydrothrombin-heparin reveals the basis of inhibitor specificity. Nat Struct Mol Biol 11:863–867. Li W, Johnson DJ, Esmon CT and Huntington JA (2004) Structure of the antithrombin-thrombin-heparin ternary complex reveals the antithrombotic mechanism of heparin. Nat Struct Mol Biol 11:857–862. Johnson DJ, Li W, Asams TE and Huntington JA (2006) Antithrombin-S195A factor Xa-heparin structure reveals the allosteric mechanism of antithrombin activation. EMBO J 25:2029–2037. Peterson FC, Gordon NC and Gettins PGW (2000) Formation of non-covalent serpin-proteinase complex involves no conformational change in the serpin. Use of 1 H–15 N HSQC NMR as a sensitive non-perturbing monitor of conformation. Biochemistry 39:11884–11892. Dobó J and Gettins PGW (2004) α1 -Proteinase inhibitor forms initial noncovalent and final covalent complexes with elastase analogously to other serpin-proteinase pairs, suggesting a common mechanism of inhibition. J Biol Chem 274:9264–9269.

ch03

Jan. 10, 2007

10:36 SPI-B429



93

14. Jin L, Abrahams JP, Skinner R, Petitou M, Pike RN and Carrell RW (1997) The anticoagulant activation of antithrombin by heparin. Proc Natl Acad Sci USA 94:14683–14688. 15. Hopkins PCR and Stone SR (1995) The contribution of the conserved hinge region residues of α1 -antitrypsin to its reaction with elastase. Biochemistry 34:15872–15879. 16. O’Malley KM and Cooperman BS (2001) Formation of the covalent chymotrypsin: Antichymotrypsin complex involves no large-scale movement of the enzyme. J Biol Chem 276:6631–6637. 17. LuoY, ZhouY and Cooperman BS (1999) Antichymotrypsin interaction with chymotrypsin. Intermediates on the way to inhibited complex formation. J Biol Chem 274:17733–17741. 18. Beatty K, Bieth J and Travis J (1980) Kinetics of association of serine proteinases with native and oxidized α-1-proteinase inhibitor and α-1antichymotrypsin. J Biol Chem 255:3931–3934. 19. Chen VC, Chao L and Chao J (2000) Roles of the P1, P2, and P3 residues in determining inhibitory specificity of kallistatin toward human tissue kallikrein. J Biol Chem 275:38457–38466. 20. Stump DC, Thienpont M and Collen D (1986) Purification and characterization of a novel inhibitor of urokinase from human urine. Quantitation and preliminary characterization in plasma. J Biol Chem 261:12759–12766. 21. Frazer JK, Jackson DG, Gaillard JP, Lutter M, Liu YJ, Banchereau J, Capra JD and Pascual V (2000) Identification of centerin: A novel human germinal center B cell-restricted serpin. Eur J Immunol 30:3039–3048. 22. Han X, Fiehler R and Broze GJ Jr (2000) Characterization of the protein Z-dependent protease inhibitor. Blood 96:3049–3055. 23. Remold-O’Donnell E, Chin J and Alberts M (1992) Sequence and molecular characterization of human monocyte/neutrophil elastase inhibitor. Proc Natl Acad Sci USA 89:5635–5639. 24. Ye RD, Wun T and Sadler JE (1987) cDNA cloning and expression in Escherichia coli of a plasminogen activator inhibitor from human placenta. J Biol Chem 262:3718–3725. 25. Schick C, Pemberton PA, Shi GP, Kamachi Y, Cataltepe S, Bartuski AJ, Gornstein ER, Brömme D, Chapman HA and Silverman GA (1998) Crossclass inhibition of the cysteine proteinases cathepsins K, L, and S by the serpin squamous cell carcinoma antigen 1: A kinetic analysis. Biochemistry 37:5258–5266.

ch03

Jan. 10, 2007

10:36 SPI-B429

94


P. G. W. Gettins

26. Schick C, Kamachi Y, Bartuski AJ, Çataltepe S, Schechter NM, Pemberton PA and Silverman GA (1997) Squamous cell carcinoma antigen 2 is a novel serpin that inhibits the chymotrypsin-like proteinases cathepsin G and mast cell chymase. J Biol Chem 272:1849–1855. 27. Coughlin P, Sun JR, Cerruti L, Salem HH and Bird P (1993) Cloning and molecular characterization of a human intracellular serine proteinase inhibitor. Proc Natl Acad Sci USA 90:9417–9421. 28. Miyata T, Nangaku M, Suzuki D, Inagi R, Uragami K, Sakai H, Okubo K and Kurokawa K (1998) A mesangium-predominant gene, megsin, is a new serpin upregulated in IgA nephropathy. J Clin Invest 102:828–836. 29. Dahlen JR, Jean F, Thomas G, Foster DC and Kisiel W (1998) Inhibition of soluble recombinant furin by human proteinase inhibitor 8. J Biol Chem 273:1851–1854. 30. Sprecher CA, Morgenstern KA, Mathewes S, Dahlen JR, Schrader SK, Foster DC and Kisiel W (1995) Molecular cloning, expression, and partial characterization of two novel members of the ovalbumin family of serine proteinase inhibitors. J Biol Chem 270:29854–29861. 31. Dahlen JR, Foster DC and Kisiel W (1997) Human proteinase inhibitor 9 (P19) is a potent inhibitor of subtilisin A. Biochem Biophys Res Commun 238: 329–333. 32. Riewald M and Schleef RR (1995) Molecular cloning of bomapin (protease inhibitor 10), a novel human serpin that is expressed specifically in the bone marrow. J Biol Chem 270:26754–26757. 33. Askew YS, Pak SC, Luke CJ, Askew DJ, Cataltepe S, Mills DR, Kato H, Lehoczk YJ, Dewar K, Birren B and Silverman GA (2001) SERPINB12 is a novel member of the human ov-serpin family that is widely expressed and inhibits trypsin-like serine proteinases. J Biol Chem 276: 49320–49330. 34. Welss T, Sun J, Irving JA, Blum R, Smith AI, Whisstock JC, Pike RN, von Mikecz A, Ruzicka T, Bird PI and Abts HF (2003) Hurpin is a selective inhibitor of lysosomal cathepsin L and protects keratinocytes from ultravioletinduced apoptosis. Biochemistry 42:7381–7389. 35. Olson ST and Björk I (1991) Predominant contribution of surface approximation to the mechanism of heparin acceleration of the antithrombinthrombin reaction. Elucidation from salt concentration effects. J Biol Chem 266: 6353–6364. 36. Sheehan JP, Wu Q, Tollefsen DM and Sadler JE (1993) Mutagenesis of thrombin selectively modulates inhibition by serpins heparin cofactor II and antithrombin III. Interaction with the anion-binding exosite determines heparin cofactor II specificity. J Biol Chem 268:3639–3645.

ch03

Jan. 10, 2007

10:36 SPI-B429



95

37. Lawrence D, Strandberg L, Grundström T and Ny T (1989) Purification of active human plasminogen activator inhibitor-1 from Escherichia coli. Comparison with natural and recombinant forms purified from eucaryotic cells. Eur J Biochem 186:523–533. 38. Scott RW, Bergman BL, BajpaiA, Hersh RT, Rodriguez H, Jones BN, Barreda C, Watts S and Baker JB (1985) Protease nexin. Properties and a modified purification procedure. J Biol Chem 260: 7029–7034. 39. Wiman B and Collen D (1978) On the kinetics of the reaction between human antiplasmin and plasmin. Eur J Biochem 84:573–578. 40. Sim RB, Arlaud G and Colomb M (1980) Kinetics of reaction of human C1inhibitor with the human complement system proteases C1r and C1s. Biochim Biophys Acta 612:433–449. 41. Krueger SR, Ghisu GP, Cinelli P, Gschwend TP, Osterwalder T, Wolfer DP and Sonderegger P (1997) Expression of neuroserpin, an inhibitor of tissue plasminogen activator, in the developing and adult nervous system of the mouse. J Neurosci 17:8984–8996. 42. Ozaki K, Nagata M, Suzuki M, Fujiwara T, Miyoshi Y, Ishikawa O, Ohigashi H, Imaoka S, Takahashi E and NakamuraY (1998) Isolation and characterization of a novel human pancreas-specific gene, pancpin, that is down-regulated in pancreatic cancer cells. Genes Chromosomes Cancer 22:179–185. 43. Owen MC, Brennan SO, Lewis JH and Carrell RW (1983) Mutation of antitrypsin to antithrombin. α1 -Antitrypsin Pittsburgh (358 Met Arg), a fatal bleeding disorder. N Engl J Med 309:694–698. 44. Chuang Y-J, Gettins PGW and Olson ST (1999) Importance of the P2 glycine of antithrombin in target proteinase specificity, heparin activation and the efficiency of proteinase trapping as revealed by a P2 Gly Pro mutation. J Biol Chem 274:28142–18149. 45. Bianchini EP, Louvain VB, Marque P-E, Juliano MA, Juliano L and Le Bonniec BF (2002) Mapping of the catalytic groove preferences of factor Xa reveals an inadequate selectivity for its macromolecule substrates. J Biol Chem 277:20527–20534. 46. Dufour EK, Denault J-B, Bissonnette L, Hopkins PCR, Lavigne P and Leduc R (2001) The contribution of arginine residues within the P6-P1 region of α1 -antitrypsin to its reaction with furin. J Biol Chem 276:38971–38979. 47. Chaillan-Huntington CE, Gettins PGW, Huntington JA and Patston PA (1997) The P6 -P2 region of serpins is critical for proteinase inhibition and complex stability. Biochemistry 36:9562–9570. 48. Izaguirre G and Oison ST (2006) Residues Tyr253 and Glu255 in strand 3 of beta-sheet C of antithrombin are key determinants of an exosite made

ch03

Jan. 10, 2007

10:36 SPI-B429

96

49.

50.

51.

52. 53.

54.

55.

56.

57.

58. 59.


P. G. W. Gettins

accessible by heparin activation to promote rapid inhibition of factors Xa and IXa. J Biol Chem 281:13424–13432. Rubin H, Plotnick M, Wang Z, Liu X, Zhong Q, Schechter NM and Cooperman BS (1994) Conversion of α1 -antichymotrypsin into a human neutrophil elastase inhibitor: Demonstration of variants with different association rate constants, stoichiometries of inhibition, and complex stabilities. Biochemistry 33:7627–7633. Matheson NR, van Halbeek H and Travis J (1991) Evidence for a tetrahedral intermediate complex during serpin-proteinase interactions. J Biol Chem 266:13489–13491. Lawrence DA, Ginsburg D, Day DE, Berkenpas MB, Verhamme IM, Kvassman J-O and Shore JD (1995) Serpin-protease complexes are trapped as stable acyl-enzyme intermediates. J Biol Chem 270:25309–25312. Huntington JA, Read RJ and Carrell RW (2000) Structure of a serpin-protease complex shows inhibition by deformation. Nature (London) 407: 923–926. Dementiev A, Dobo J and Gettins PGW (2006) Active site distortion is sufficient for proteinase inhibtion by serpins. Structure of the covalent complex of alpha-1 proteinase inhibitor with porcine pancreatic elastase. J Biol Chem 281:3452–3457. Stratikos E and Gettins PGW (1998) Mapping the serpin-proteinase complex using single cysteine variants of α-proteinase inhibitor Pittsburgh. J Biol Chem 273:15582–15589. Stratikos E and Gettins PGW (1999) Formation of the covalent serpinproteinase complex involves translocation of the proteinase by more than 70Å and full insertion of the reactive center loop into β-sheet A. Proc Natl Acad Sci USA 96:4808–4813. Peterson FC and Gettins PGW (2001) Insight into the mechanism of serpin-proteinase inhibition from 2D [1 H–15 N] NMR studies of the 69kDa α1-proteinase inhibitor Pittsburgh-trypsin covalent complex. Biochemistry 40:6284–6292. Backovic M, Stratikos E, Lawrence DA and Gettins PGW (2002) Structural similarity of the covalent complexes formed between the serpin plasminogen activator inhibitor-1 and the arginine-specific proteinases trypsin, LMW urokinase, HMW urokinase and tPA: Use of site-specific fluorescent probes of local environment. Protein Sci 11:1182–1191. Na Y-R and Im H (2005) The length of the reactive center loop modulates the latency transition of plasminogen activator inhibitor-1. Protein Sci 14:55–63. Im H, Ahn HY andYu MH (2000) Bypassing the kinetic trap of serpin protein folding by loop extension. Protein Sci 9:1497–1502.

ch03

Jan. 10, 2007

10:36 SPI-B429



97

60. Stein PE and Carrell RW (1995) What do dysfunctional serpins tell us about molecular mobility and disease? Nat Struct Biol 2:96–113. 61. Hood DB, Huntington JA and Gettins PGW (1994) α1 -Proteinase inhibitor variant T345R. Influence of P14 residue on substrate and inhibitory pathways. Biochemistry 33:8538–8547. 62. Lawrence DA, Olson ST, Muhammad S, Day DE, Kvassman J-O, Ginsburg D and Shore JD (2000) Partitioning of serpin-proteinase reactions between stable inhibition and substrate cleavage is regulated by the rate of serpin reactive center loop insertion into β-sheet A*. J Biol Chem 275:5839–5844. 63. Huntington JA, Fan B, Karlsson KE, Deinum J, Lawrence DA and Gettins PGW (1997) Serpin conformational change in ovalbumin. Enhanced reactive center loop insertion through hinge region mutations. Biochemistry 36:5432–5440. 64. Yamasaki M, Arii Y, Mikami B and Hirose M (2002) Loop-inserted and thermostabilized structure of P1-P1’ cleaved ovalbumin mutant R339T. J Mol Biol 315:113–120. 65. Arii Y and Hirose M (2002) Probing the serpin structural-transition mechanism in ovalbumi mutant R339T by proteolytic-cleavage kinetics of the reactive centre loop. Biochem J 363:403–409. 66. Takahashi N, Terakado K, Nakamura G, Soekmadji C, Masuoka T,Yamasakio M and Hirose M (2005) Dynamic mechanism for the serpin loop insertion as revealed by quantitative kinetics. J Mol Biol 348:409–418. 67. Plotnick MI, Mayne L, Schechter NM and Rubin H (1996) Distortion of the active site of chymotrypsin complexed with a serpin. Biochemistry 35:7586– 7590. 68. Futamura A, Stratikos E, Olson ST and Gettins PGW (1998) Change in environment of the P1 side chain upon progression from the Michaelis complex to the covalent serpin-proteinase complex. Biochemistry 37:13110–13119. 69. Calugaru SV, Swanson R and Olson ST (2001) The pH dependence of serpinproteinase complex dissociation reveals a mechanism of complex stabilization involving inactive and active conformational states of the proteinase which are perturbable by calcium. J Biol Chem 276:32446–32455. 70. Shin J-S and Yu M-H (2002) Kinetic dissection of α1 -antitrypsin inhibition mechanism. J Biol Chem 277:11629–11635. 71. Egelund R, Petersen TE and Andreasen PAA (2001) serpin-induced extensive proteolytic susceptibility of urokinase-type plasminogen activator implicates distortion of the proteinase substrate binding pocket and oxyanion hole in the serpin inhibitory mechanism. Eur J Biochem 268:673–685.

ch03

Jan. 10, 2007

10:36 SPI-B429

98


P. G. W. Gettins

72. Fish WW, Orre K and Björk I (1979) Routes of thrombin action in the production of proteolytically modified, secondary forms of antithrombin-thrombin complex. Eur J Biochem 101:39–44. 73. Cooperman BS, Stavridi E, Nickbarg E, Rescorla E, Schechter NM and Rubin H (1993) Antichymotrypsin interaction with chymotrypsin — partitioning of the complex. J Biol Chem 268:23616–23625. 74. Stavridi ES, O’Malley K, Lukacs CM, Moore WT, Lambris JD, Christianson DW, Rubin H and Cooperman BS (1996) Structural change in α-chymotrypsin induced by complexation with α1 -antichymotrypsin as seen by enhanced sensitivity to proteolysis. Biochemistry 35:10608–10615. 75. Kaslik G, Patthy A, Bálint M and Gráf L (1995) Trypsin complexed with α1 -proteinase inhibitor has an increased structural flexibility. FEBS Lett 370:179–183. 76. Kaslik G, Kardos J, Szabó L, Závodszky P, Westler WM, Markley JL and Gráf L (1997) Effects of serpin binding on the target proteinase: Global stabilization, localized increased structural flexibility, and conserved hydrogen bonding at the active site. Biochemistry 36:5455–5474. 77. Hervé M and Ghélis C (1991) Conformational stability of the covalent complex between elastase and α1 -proteinase inhibitor. Arch Biochem Biophys 285:142–146. 78. Irving JA, Pike RN, Lesk AM and Whisstock JC (2000) Phylogeny of the serpin superfamily: implications of patterns of amino acid conservation for structure and function. Genome Res 10:1845–1864. 79. Simonovic M, Gettins PGW and Volz K (2000) Crystal structure of viral serpin crmA provides insights into its mechanism of cysteine proteinase inhibition. Protein Sci 9:1423–1427. 80. Zhou A, Carrell RW and Huntington JA (2001) The serpin inhibitory mechanism is critically dependent on the length of the reactive center loop. J Biol Chem 276:27541–27547. 81. Plotnick MI, Rubin H and Schechter NM (2002) The effects of reactive site location on the inhibitory properties of the serpin alpha 1-antichymotrypsin. J Biol Chem 277:29927–29935. 82. Boudier C and Bieth JG (2001) The reaction of serpins with proteinases involves important enthalpy changes. Biochemistry 40:9962–9967. 83. Boudier C, Gils A, Declerck PJ and Bieth JG (2005) The conversion of active to latent plasminogen activator inhibitor-1 is an energetically silent event. Biochem J 88:2848–2854. 84. Gettins PGW (2002) The F-helix of serpins plays an essential, active role in the proteinase inhibition mechanism. FEBS Lett 523:2–6.

ch03

Jan. 10, 2007

10:36 SPI-B429



99

85. Picard V, Marque P-E, Paolucci F, Aiach M and Le Bonniec BF (1999) Topology of the stable serpin-protease complexes revealed by an autoantibody that fails to react with the monomeric conformers of antithrombin. J Biol Chem 274:4586–4593. 86. Vleugels N, Gils A, Bijnens A-P, Knockaert I and Declerck P (2000) The importance of helix F in plasminogen activator inhibitor-1. Biochim Biophys Acta 1476:20–26. 87. Wind T, Jensen JK, Dupont DM, Kulig P andAndreasen PA (2003) Mutational analysis of plasminogen activator inhibitor-1. Interactions of α-helix F and its neighbouring structural elements regulates the activity and the rate of latency transition. Eur J Biochem 270:1680–1688. 88. Plotnick MI, Samakur MMWZ, Liu X, Rubin H, Schechter NM and Selwood T (2002) Heterogeneity in serpin-proteinase complexes as demonstrated by differences in the mechanism of complex breakdown. Biochemistry 41:334–342. 89. Renatus M, Zhou Q, Stennike HR, Snipas SJ,Turk D, Bankston LA, Liddington RC and Salvesen GS (2000) Crystal structure of the apoptotic suppressor CrmA in its cleaved form. Structure 8:789–797. 90. Tesch LD, Raghavendra MP, Bedsted-Faarvang T, Gettins PGW and Olson ST (2005) Specificity and reactive loop length requirements for CrmA inhibition of serine proteases. Protein Sci 14:533–542. 91. Raghavendra MP, Bedsted T, Zhang W, Froelich CJ, Thornberry N, Gettins PGW and Olson ST (2006) Evidence for similar extents of protease translocation in crmA complexes with granzyme B and caspase-1, submitted. 92. Xu G, Cirilli M, Huang Y, Rich RL, Myszka DG and Wu H (2001) Covalent inhibition revealed by the crystal structure of the caspase-8/p35 complex. Nature (London) 410:494–497. 93. Kvassman JO, Verhamme I and Shore JD (1998) Inhibitory mechanism of serpins: loop insertion forces acylation of plasminogen activator by plasminogen activator inhibitor-1. Biochemistry 37:15491–15502. 94. Schwartz BS and España F (1999) Two distinct urokinase-serpin interactions regulate the initiation of cell-surface-associated plasminogen activation. J Biol Chem 274:15278–15283. 95. Shieh B-H, Potempa J and Travis J (1989) The use of α2 -antiplasmin as a model for the demonstration of complex reversibility in serpins. J Biol Chem 264:13420–13423. 96. Stark KR and James AA (1996) Salivary gland anticoagulants in culicine and anopheline mosquitoes (Diptera: Culicidae). J Med Entomol 33:645–650.

ch03

Jan. 10, 2007

10:36 SPI-B429

100


P. G. W. Gettins

97. Stark KR and James AA (1998) Isolation and characterization of the gene encoding a novel factor Xa-directed anticoagulant from the yellow fever mosquito, Aedes aegypti. J Biol Chem 273:20802–20809. 98. Stark KR and James AA (1995) A factor Xa-directed anticoagulant from the salivary glands of the yellow fever mosquito Aedes aegypti. Exp Parasitol 81:321–331. 99. Ngamkitidechakul A, Warejcka DJ, Burke JM, O’Brien WJ and Twining SS (2003) Sufficiency of the reactive site loop of maspin for induction of cell-matrix adhesion and inhibition of cell invasion. Conversion of ovalbumin to a maspin-like molecule. J Biol Chem 278:31796–31806. 100. McGowen R, Biliran HJ, Sager R and Sheng S (2000) The surface of prostate carcinoma DU145 cells mediates the inhibition of urokinase-type plasminogen activator by maspin. Cancer Res 60:4771–4778. 101. Bass R, Fernandéz A-MM and Ellis V (2002) Maspin inhibits cell migration in the absence of protease inhibitory activity. J Biol Chem 277:46845–46848. 102. Mellet P, Michels B and Bieth JG (1996) Heat-induced conversion of ovalbumin into a proteinase inhibitor. J Biol Chem 271:30311–30314.

ch03

Jan. 10, 2007

8:57

SPI-B429


4 Mouse Serpins and Transgenic Studies David J. Askew, Paul Coughlin and Phillip I. Bird

1. Introduction The serpin gene family encodes a group of proteins with diverse biological functions. Best defined are those serpins that are produced in the liver and secreted into the plasma, where they regulate proteolytic cascades involved in hemostasis, complement and inflammatory response via serine proteinase inhibition. Other serpins are expressed in tissues throughout the body, as extracellular or intracellular proteins, and have inhibitory or non-inhibitory functions. Within mammals, the serpin genes are divided into nine clades, A-I, based on amino acid sequence homology,1,2 Comparison of the serpin genes of humans (35 genes) and mouse (64 genes) suggests conservation of many of these genes and their activities, as evidenced by the existence of at least 26 orthologous gene pairings (Table 1). This suggests that results obtained in mouse models may contribute to our understanding of human gene function. The purpose of this review is to briefly describe the mouse serpin genes and to discuss the current and future use of transgenic mouse models in defining serpin function, developing human disease models, and testing novel therapies against serpin-associated diseases.

101

ch04

Jan. 10, 2007

8:57

SPI-B429


102 Table 1

D. J. Askew, P. Coughlin & P. I. Bird Comparison of human and mouse serpin gene content, chromosomal location. Human

Common name

a-1-antitrypsin Antitrypsin-related protein a-1-antichymotrypsin kallikrein inhibitor, kallistatin Protein C inhibitor Corticoid-steroid binding protein AGT Centerin Protein Z-dependent prot. inhib. novel Visceral adipose tissue-derived serpin Monocyte Neutrophil Elastase Inhib. PI-6, Cytoplasmic antiprotease PI-9, Cytoplasmic antiprotease 3 Plasminogen activator inhibitor 2 Squamous cell carcinoma antigen 1 Squamous cell carcinoma antigen 2 Maspin Megsin PI-8 Cytoplasmic antiprotease 2 Bomapin Epipin

Mouse

SERPIN XX

A1 A2

Chr locus

Chr Serpin XX Common name locus (total #) of predicted orthologue, abbreviation

14q32 12F1 a1a-e5 ”

A1AT1-5, Spi1-1-5

A3 A4

” ”

”

a3a-n14

A5 A6

” ”

” ”

a5 a6

PCI CBG

A8 A9 A10

” ” ”

” ” ”

a8 a9 a10

AGT Centerin PZI

A11 A12

” ”

” ”

a11 a12

B1

6p25

B6

”

”

b6a-e3

SPI3

B9

”

”

b9a-g7

SPI6

B2

18q21

1D

b2

PAI-2

B3

”

”

b3a-d4

B4

”

B5 B7 B8

” ” ”

” ” ”

b5 b7 b8

Maspin Megsin CAP2, Spi8

B10 B11

” ”

” ”

b10 b11

Bomapin Epipin

13A3 b1a-c3

MNEI

SCCAs

(Continued)

ch04

Jan. 10, 2007

8:57

SPI-B429


Mouse Serpins and Transgenic Studies

103

Table 1 (Continued ) Human Common name

Yukopin Headpin Antithrombin, ATIII Heparin cofactor II (HCFII) Plasminogen activator inhibitor 1 Protease Nexin-1, GDN Pigment-epithelium derived factor Alpha-2-antiplasmin C1 Inhibitor Collagen binding protein Heat shock protein 47 Neuroserpin Pancpin

Mouse SERPIN Chr XX locus

Chr Serpin XX Common name locus (total #) of predicted orthologue, abbreviation

B12 B13 C1 D1

” ” 11q 22q

” ” 1H2 16A2

b12 b13 c1 d1

Yukopin Headpin ATIII HCFII

E1

7q

5G2

e1

Pai-1

E2

2q33

1C5

e2

PN-1

F1

17q

11B5 f1

PEDF

F2 G1 H1

” 11q12 11q13

” 2D 7E1

f2 g1 h1

A2AP C1I Cbp1

H2 I1 I2

” 3q26 ”

” 3E3 ”

h2 i1 i2

HSP47, Cbp2 Neuroserpin PI14

Human and mouse serpin genes, Clades A–I, chromosomal location, standard nomenclature (SERPIN XX), common names and abbreviations. Refs. 5–8.

2. The Mouse Serpins and Transgenic Studies 2.1. Comparison of serpin gene content between mouse and humans Analysis of the human genome has shown that proteases and inhibitors — the “degradome” — make up >2% of genes.3 The mouse “degradome” is larger than that of humans and inter-species differences are most apparent in genes involved in reproduction and immunity. 4 The expansion in mouse protease and inhibitor genes has occurred via gene duplication, generating families of paralogues. Likewise, there are proteases and inhibitors unique to humans that have arisen by duplication after divergence from the

ch04

Jan. 10, 2007

8:57

SPI-B429

104


D. J. Askew, P. Coughlin & P. I. Bird

human/mouse common ancestor. Thus, in any mouse model developed to elucidate human gene function, it is important to identify the orthologous mouse gene, and to consider the potential effects of homologous and paralogous genes on the model phenotype. Such effects include redundancy where a paralogue with essentially the same reactive center loop (RCL) sequence and expression pattern can substitute for a deleted gene; and compensation where a homologue with a non-identical RCL sequence but overlapping inhibitory and expression profile can provide key functions of the missing protein. Of the nine serpin clades in humans and mouse, seven show perfect concordance of gene number and synteny (Clades C–I, Table 1). This predicts orthologous relationships between the genes in these clades and suggests conservation of function. In contrast, conservation of gene number is not observed in Clades A and B. In the mouse, gene expansion has produced Serpina1a–Serpina1e (5 genes) and Serpina3a–Serpina3n (14 genes) at chr 12F1. It has also produced Serpinb1a–Serpinb1c (3 genes), Serpinb6a– Serpinb6e (5 genes) and Serpinb9a–Serpinb9g (7 genes) at chr 13A3.5–7 Finally, a small expansion is observed in a subset of Clade B genes of both humans and mouse. Human SERPINB3 and SERPINB4 and mouse Serpinb3a–Serpinb3d (4 genes) are the result of gene duplication events, which occurred after the division of human and rodent lineages.8 The functions of most of these extra mouse serpins have yet to be established, but it should be noted that many have unique RCLs, implying that they control distinct proteases. This is consistent with the fact that the mouse has ∼ 60 more serine and cysteine proteases than those found in humans, and with evidence for accelerated mutation of the RCL.9 Genes with accelerated evolution due to positive selection have been associated with immunity and reproduction,10 and many of the Clades A and B serpins seem to function in these systems.1 In summary, modeling human serpin gene function in the mouse using transgenic and targeted gene deletion techniques will probably be successful where there is a 1:1 orthologous relationship between human and mouse genes. However, analysis of serpins within the regions of gene expansion in Clades A and B is most likely to be problematic. Identification of Clades A and B human–mouse orthologues cannot be done by sequence comparisons alone. It requires additional information such as inhibitory profile,

ch04

Jan. 10, 2007

8:57

SPI-B429



105

expression patterns, and cellular distribution and the recognition that a physiological role that is played by one human serpin may be played by two or more paralagous mouse serpins. Large-scale deletions encompassing all of a paralagous gene set may be necessary to generate a phenotype in the mouse reflecting loss of the corresponding unique gene in humans. For example, mouse genes Serpina3a – Serpina3n might be deleted to model SERPINA3 gene function. Transgenic alleles of SERPINA3 or individual mouse Serpina3 genes could then be added back to examine the function of each gene individually in complementation studies.

2.2. Hemostasis and thrombosis systems in mice Hemostasis is maintained through a balance between the coagulationthrombosis and fibrinolysis systems. The serpin genes playing roles in these cascades are mostly conserved between humans and mouse (see Table 2 for comparison of human diseases and mouse phenotypes). Like humans, mouse Serpinc1 (antithrombin) deficiency is homozygous-lethal, while heterozygotes are at risk for thromboembolism.11 Serpind1 (heparin cofactor II) is required for dermatan sulfate responsiveness and normal clotting.12,13 In an example of non-conservation of function between predicted orthologues, mouse Serpina5 protein (protein C inhibitor) is not found in blood plasma, and Serpina5-deficiency unexpectedly results in male infertility.14 To generate a useful model of human protein C inhibitor, a human SERPINA5 transgene was introduced into mice, and protein C inhibitor was then evident in the plasma and tissues of the transgenic animals, in a pattern identical to its distribution in humans.15 Within the fibrinolysis system, Serpine1 (PAI-1) and Serpinf2 (α2-AP) function appear to be conserved, as indicated by transgenic studies. Small amounts of plasmin-inhibitor activity are present in the plasma of Serpinf2-deficient mice, and these animals show limited bleeding disorders.16 SERPINE1 function is not strictly confined to the regulation of plasminogen activators in the blood. Serpine1-deficient and transgenic models have been used extensively to demonstrate its multifaceted potential in modulating thrombosis, fibrosis, obesity and tumor-associated angiogenesis (see Table 2 for specific models, phenotypes, and refs.).

ch04

Jan. 10, 2007 8:57

Serpin-associated human diseases and mouse phenotypes.

HUMAN Disease

Serpin allele

Mouse represents human?

Phenotype

Reference

Liver cirrhosis, Emphysema

Serpina1: (a) Tg-Z-A1AT, (b) Tg-M-A1AT

(a, b) A1AT accumulation in hepatocytes, liver degeneration, decreased plasma A1AT, no emphysema

Yes

50–52

A5 (PCI)

Coronary thrombosis marker

Serpina5 (a) PCI-/(b) PCI-/-; Tg-huPCI (humanized)

(a) Male infertility (b) Tg-Human PCI in plasma

?

14, 15, 62–64

A8 (AGT)

Hyper tension

Serpina8: (a) AGT-/(b) Tg-huAGT;Tg-hRenin (humanized)

(a) Embryonic lethal (b) Blood pressure, [AGT] correlates directly with gene copy number

Yes

17, 18, 65

(Continued)


A1 (A1AT)


SERPINXX (common name)

Mouse

SPI-B429

106

Table 2

ch04

Jan. 10, 2007 8:57

HUMAN

(Continued)

Mouse Disease

Serpin allele


Phenotype

Reference

Tumor marker

Serpinb2 (a) Pai-2-/-, (b) Pai-1-/-;Pai-2/(c) Tg-K5-Pai-2

(a) Normal, LPS-sensitive BMD-macrophage cells (b) Normal (c) Normal skin, increased papilloma potential

Yes

25, 26, 66

B3 (SCCA1) B4 (SCCA2)

Tumor marker

Serpinb3a Serpinb3a-/-

Normal

?

DJA, GAS, unpub

B5 (Maspin)

Tumor marker

Serpinb5 (a) Maspin-/(b) Maspin-/+ (c) Tg-WAP-Maspin

(a) Embryonic lethal 5.5 dpc (b) Defective ovarian, mammary gland development, reduced progesterone

Yes

24, 67–69

107

(Continued)


B2 (PAI-2)



SPI-B429

Table 2

ch04

Jan. 10, 2007 8:57

Disease

Serpin allele


Phenotype

Reference

(c) Decreased breast tumor growth, angiogenesis, and metastasis. Disrupted mammary gland development B6 (PI6)

None detected

Serpinb6 (SPI3) Serpinb6-/-

Upregulate Serpinb1

?

29, 30

B9 (PI-9)

None detected

Serpinb9 (SPI6) Serpinb9-/-

Decreased survival of CTL; inability to clear LCMV infection

?

27

(Continued)



Mouse


HUMAN

(Continued)

SPI-B429

108

Table 2

ch04

Jan. 10, 2007 8:57

Mouse Disease

Serpin allele

Phenotype


Reference

CI (AT)

Homozygous lethal, thromboembolism, deep vein thrombosis, end stage renal disease

Serpinc1 (a) AT-/(b) AT+/(c) ATm/m (R48C heparin binding deficient)

(a) Embryonic lethal 15.5-16.5 dg (b) At risk for thromboembolism (c) Non-responsive to heparin, suffer thrombotic events in heart, liver, and ocular, placental and penile vessels

Yes

11, 70, 71

D1 (HCII)

Potential risk factor for thrombosis

Serpind1 HCII-/-

Shortened arterial clotting time, dermatan sulfate non-responsive

Yes

12, 13, 72, 73



(Continued)

(Continued) 109


HUMAN

SPI-B429

Table 2

ch04

Jan. 10, 2007 8:57

Disease

Serpin allele

Phenotype


Reference

1. Thrombosis: Coronary artery disease, Pre-eclampsia, “Postoperative fibrinolytic shutdown”

Serpine1 (a) Pai-1-/(b) Pai-1-/-; vitronectin-/(b) Tg-Pai-1

(a) Protected against thrombosis, increased thrombo-embolism (b) Same as (a) above (c) Increase thrombi in coronary artery, renal and adipose tissue

Yes

74–78

”

2. Fibrosis: Glomerulonephritis, diabetic nephropathy, adult respiratory distress syndrome

(a) Pai-1-/-

Protected against fibrosis in Pulmonary and Urethra models

Yes

79, 80

(Continued)


E1 (PAI-1)



Mouse

SPI-B429

HUMAN

(Continued)

110

Table 2

ch04

Jan. 10, 2007 8:57

Mouse Disease

Serpin allele

Phenotype


Reference

”

3. Obesity

(a) Pai-1-/(b) Pai-1-/-;ob-/-

Pai-1-/- protects against diet-induced and genetic (ob/ob) obesity

Yes

81, 82

”

4. Cancer: poor prognostic marker for breast, gastric, pulmonary adenocarcinoma, and ovarian cancers

(a) Pai-1-/(b) Tg-Pai-1 (high) (c) Tg-Pai-1 (low)

(a, b) Suppressed angiogenesis, tumor growth (c) Increased angiogenesis, tumor growth

Yes

83–85



(Continued)

(Continued) 111


HUMAN

SPI-B429

Table 2

ch04

Jan. 10, 2007 8:57

Disease

None detected

Serpin allele

Serpine2 (a) PN-1-/(b) Tg-Thy1.2- PN-1

Phenotype


(a) Male infertility (b) Shortened lifespan (a, b) See Table 3 for CNS-related phenotypes

Reference

46, 86

F1 (PEDF)

Retinal diseases associated with decreased PEDF

Serpinf1 Serpinf1-/-

Epithelial and stromal hyperplasia associated with increase vasculature in prostate, pancreas Retinas had misplaced vasculature, reduced ganglion

Yes

20, 21, 87, 88

F2 (α2-AP)

Bleeding disorders

Serpinf2 α2-AP

Increased fibrinolysis at 1 to 3 weeks, residual plasmin inhibitory activity

Yes/?

16

(Continued)


E2 (Protease Nexin-1, PN-1)

Mouse



(Continued)

SPI-B429

HUMAN

112

Table 2

ch04

Jan. 10, 2007 8:57

Mouse Disease

Serpin allele

Phenotype


Reference

G1 (C1INH)

Hereditary angiodema

Serping1 (a) Serping1-/(b) Serping1-/-; Bk2R-/-

(a) Vascular permeability (b) Normal vasculature

Yes

19

H1 (HSP47)

None detected

SerpinH1 Hsp47-/-

Embryonic lethal 11.5 dpc

Yes

89, 90

I1 (Neuro-serpin)

Familial dimentia

Serpini1 (a) Ns-/(b) Tg-Thy1.2-Ns

(See Table 3 for CNS-related phenotypes, Refs.)

Yes

I2 (Pancpin)

Pancreatic insufficiency

Serpini2 pequeno (null)

Pancreatic acinar cell apoptosis, malnutrition

Yes



(Continued)

90, 91

113


HUMAN

SPI-B429

Table 2

ch04

Jan. 10, 2007

8:57

SPI-B429

114



2.3. Regulation of blood pressure and vascular permeability Serpina8 (angiotensinogen) and Serping1 (C1-inhibitor) regulate blood pressure and vascular permeability in mice, respectively. Serpina8 function was elegantly demonstrated using Serpina8-deficient and transgenic mice to produce a range of plasma Serpina8 concentrations.17,18 Blood pressure was shown to directly correlate with gene number and Serpina8 level. Vascular permeability in Serping1-deficient animals could be rescued by deletion of the bradykinin receptor type 2 gene.19 2.4. SERPINF1 (PEDF) is a promising therapeutic agent for retinal diseases Humans have not been identified with SERPINF1 (PEDF) deficiency or mutations, but reduced levels are associated with increased angiogenesis.20 Inappropriate angiogenesis around the retina is associated with ocular disease and leads to loss of sight. Serpinf1-deficient animals display the predicted increases in angiogenesis in the eyes and other tissues.21 Importantly, mouse models of retinal disease confirm the efficacy of SERPINF1 replacement therapy via direct injection and by adenoviral gene therapy as well.22,23 2.5. Identification of intracellular Clade B serpin function in mice Clade B serpins are largely intracellular and associated with processes in development, tumorigenesis, inflammation and immunity. The lack of human Clade B mutants makes mouse models the most practical approach to uncovering their physiological roles. Human SERPINB2, B5, B7, B8, B10, B11, B12 and B13 have single orthologues in mice (Table 1) and represent genes that could be analyzed in a relatively straightforward manner. Manipulation of Serpinb5 (Maspin) has yielded robust phenotypes in deletion and transgenic studies. Serpinb5-deficiency is homozygous lethal at 5.5 dpc.24 Extensive analysis of heterozygotes, Tg animals and adenoviral delivery of Serpinb5 protein has implicated this gene in ovarian and mammary development, and in angiogenesis associated with tumor growth

ch04

Jan. 10, 2007

8:57

SPI-B429



115

and regression (see Table 2 for details, refs.). In contrast, Serpinb2-deficient mice appear normal in development and growth, but further analysis has demonstrated the critical role of this serpin (PAI-2) in the macrophage defense pathway against bacterial LPS stimulation.25,26 As discussed earlier, the expansion of the remaining Clade B genes in the mouse (B1, B3/B4, B6, and B9) is quite likely to obscure a phenotype resulting from a single gene deletion, when one or more paralogues remain intact (redundancy). Thus far, targeted deletions of the Serpinb3a, Serpinb6, and Serpinb9 genes have been generated. Despite having six highly related paralogues in the mouse (Serpinb9b–Serpinb9g), deletion of Serpinb9 (SPI6) yields a phenotype predicted for the loss of human SERPINB9 — a decrease in the number and function of cytotoxic T cells due to failure to regulate granzyme B.27,28 In contrast, deletion of a combination of Serpinb3a, the functional orthologue of SERPINB3 and SERPINB4, resulted in animals without an overt phenotype (Askew and Silverman, unpublished data). Serpinb6 (SPI3)-deficient animals also have no detectable phenotype, but upregulate Serpinb1.29 This is an example of functional compensation, where a related (non-paralagous) serpin with an overlapping inhibitory profile substitutes for the missing protein. On a technical note, it was also determined that leaving the neomycin selection gene cassette in the targeted locus results in a 2–10-fold elevated level of a GFP reporter gene positioned to analyze Serpinb6 expression,30 and leads to the formation of tumors in older animals (Scarff and Bird, unpublished data). This illustrates the importance of designing gene deletion strategies that allow removal of selection cassettes and suggests that phenotypes of animals with retained cassettes may be misleading. 2.6. Dissection of Serpini1, Serpine2, and Serpine1 activities in the brain Serine proteinases from the coagulation and fibrinolysis pathways including plasminogen, its two main activators tissue types- and urokinase plasminogen activator, and thrombin have critical roles in the brain (reviewed in refs. 31 and 32). Working in concert with these proteinases are three active serine proteinase inhibitors (reviewed in ref. 33). SERPINI1 (Neuroserpin, Ns) and SERPINE2 (Protease Nexin-1, PN-1) are both expressed

ch04

Jan. 10, 2007

8:57

SPI-B429

116



extensively in the brain as well as a few specific tissues outside the Central Nervous System (CNS). The third plasminogen activator inhibitor, SERPINE1 (PAI-1), is found throughout the body, including low levels in the CNS, is inducible by both normal neuronal excitatory signals, and it responses to numerous neuronal damage-associated signals. These three serpins are predicted to play major roles in CNS development, synapse plasticity and learning, and in the context of multiple neuronal pathologies. Genetic analysis in the mouse using serpin-deficient animals in combination with proteinase-deficient models has made it possible to dissect out specific roles of these serpins, as well as those of the proteinases (Table 3). Initial characterization of Serpini1-deficient mice found significant decreases in animal activity and elevated anxiety levels, phenotypes associated with defects in emotion and behavior, without any detectable change in tPA activity.34 Transgenic mice over-expressing Serpini1 under the control of the Thy1.2 gene promoter (Tg-Thy1.2-Ns) were used to further define the role of Serpini1 in the brain, acting in concert with tPA. In Tg-Thy1.2Ns animals, Serpini1 reduces tPA activity in the brain and protects against ischemia-induced infarction.35 The direct interaction between Serpini1 and tPA in vivo was demonstrated by the isolation of proteinase–inhibitor complexes from Tg-Thy1.2-Ns brain lysates.36,37 To fully develop our understanding of where serpins such as Serpini1 fit into the complex picture of the brain, genetic analysis of interacting target proteases is critical. Using the tPA-/- mouse model, tPA was the first gene genetically identified as a component of long-term potentiation (LTP).38 Additional experiments using tPA-/- animals confirmed a requirement for tPA in neuron migration during development, and normal brain activities such as motor learning, emotion, and stress-induced anxiety behavior.39–41 The tPA-/- animals exhibited resistance to neuronal damage, and tPA was also found to be directly involved in opening the blood-brain barrier (BBB).36,37,42 Addition of recombinant Serpini1 protein in these models reduced tPA-dependent seizure progression and BBB opening, supporting its predicted role as the major tPA inhibitory molecule in the brain.36,37 Serpine2 (protease nexin-1) activities in the brain were also identified using Serpine2-deficient (PN-1-/-) and over-expressing (Tg-Thy1.2-PN-1) mice.43–46 Serpine2 was found to directly modulate protease activity inversely associated with reduced NMDA receptor levels. Specifically,

ch04

Jan. 10, 2007

8:57

SPI-B429


Mouse Serpins and Transgenic Studies Table 3

117

Phenotypes in the CNS of serpin and proteinase transgenic models.

Serpin

Genotype

Phenotype

Serpini1 (Neuroserpin, Ns)

(a) Ns-/(b) Tg-Thy1.2-Ns

(a) Emotion defects, increased anxiety.34 (b) Reduced tPA activity and infarct size after occlusion;92 detectable Serpini1-tPA complexes in brain extracts.36,37

Serpine2 (Protease Nexin-1, PN-1)

(a) PN-1-/(b) Tg-Thy1.2- PN-1

(a) Increased epileptic activity and reduced long-term potentiation;43 elevated tPA activity, reduced NMDA receptor levels, impaired whisker-dependent sensory function.44 (b) Increased epileptic activity and long-term potentiation;43 motor behavior defects and neurogenic muscle atrophy.46

Proteinase Tissue type plasminogen activator (tPA)

(a) tPA-/(b) tPA-/-; Plg-/-

(a) Resistant to excitoxin-induced degeneration;42 reduced late Long-term potentiaation (LTP);38 slowed neuron migration during development;39 reduced mossy fiber synapse formation after seizure;93 reduced motor learning;40 loss of stress-induced anxiety behavior;41 decreased ischemia-induced blood brain barrier (BBB) opening, Plg-independent;37 (Continued)

ch04

Jan. 10, 2007

8:57

SPI-B429


118

D. J. Askew, P. Coughlin & P. I. Bird Table 3 (Continued )

Serpin

Genotype

Phenotype (b) Plg-independent delayed seizure progression in tPA-/model;36 Plg-independent tPA-induced BBB permeability.37 (Reviewed in32 )

Urokinase plasminogen activator (uPA)

Plasminogen (zymogen)

(a) uPA-/(b) uPA-/-; tPA-/-

Plg-/-

(a) Dramatically delayed recovery in sciatic nerve crush model of nerve regeneration (axon growth, reinnervation of synaptic target).94 (b) Double PA-deficient animals showed less dramatic neuron recovery defect than uPA-/- alone.94 Accelerated supine righting response, faster latency for audiogenic seizure susceptibility.95

NMDA subunit NR-1 protein levels were reduced in PN-1-/- animals, in association with elevated tPA protein levels and proteinase activity.44 PN1-/- animals also have impaired whisker-dependent sensory function, in line with the role of NMDA receptor activity in many neuronal activities, including sensory response pathways. In addition, PN-1-/- animals were used to isolate a novel serine protease inhibitor activity from PN-1-/- brain homogenates, identified as the phosphatidylethanolamine-binding protein (PEBP).45 Continual use and development of such genetic model systems should not only further define the roles of Serpini1, Serpine2, and tPA in the brain, but should aid in also the identification and the testing of specific therapeutic targets of neurological diseases.

ch04

Jan. 10, 2007

8:57

SPI-B429



119

2.7. Modeling conformational diseases in the mouse Many diseases share the underlying mechanism of aberrant protein aggregation, often exhibiting the common feature of intracellular protein deposits or inclusions (reviewed in ref. 47). This collection of conformational diseases includes the common neurodegenerative disorders: Alzheimer’s, Huntington’s, and Parkinson’s diseases.Among the conformational diseases, the serpinopathies are best understood with respect to the mechanisms of protein aggregation. This is due to a myriad of crystallographic analyses of both pathogenic and non-pathogenic serpin protein forms. It is proposed that serpin polymer formation is a form of domain swap aggregation. In the case of serpins, the RCL of one molecule is inserted into a beta sheet of another molecule. Domain exchange repetition leads to the formation of large insoluble polymers. The primary cause of the associated pathology is cell death due to polymer retention in the rough endoplasmic reticulum or its accumulation in inclusion bodies, which leads to significant destruction and failure of the host tissue or organ.A secondary source of pathology is tissue damage or system failure due to the absence of functional protein at its normal site of action. Given the advanced stage of understanding serpinopathies mechanism, the use of genetically engineered mouse models to identify aggravating factors and potential therapeutic targets should prove to be very productive. The serpin-dependent conformational diseases are associated with at least five different serpins. SERPINA1 (A1AT) is expressed at high level in the liver and secreted into the plasma under normal conditions. It primarily acts to control neutrophil elastase in the lung. SERPINA1 alleles Z-, S-, and Mmalton-AT exhibit aggregation in the rough endoplasmic reticulum of liver cells, resulting in liver degeneration (reviewed in ref. 48). A second pathology of the SERPINA1 conformational disease is emphysema, primarily due to reduced serum levels of SERPINA1 (10–15% in homozygous Z-AT individuals) and loss of control of neutrophil elastase. In addition, mutant SERPINA1 that does make it to the lung may contribute further to the disease by polymerizing there. In humans and in mice, Z-AT polymers in the lung are associated with neutrophil influx and elevated elastase activity.49 Mouse models of the SERPINA1 serpinopathy consist of transgenic expression of either the Z-AT human genes, and animals exhibit liver

ch04

Jan. 10, 2007

8:57

SPI-B429

120



degeneration.50−52 Emphysema is not observed, presumably because endogenous Serpina1 genes are still functional. Such models have been used to explore therapeutic avenues for the treatment of A1AT deficiency. For example, the mouse model Pi-Z (over-expresses human Z-AT) was used to screen for “chemical chaperones” that may ameliorate A1AT retention in the rough endoplasmic reticulum, and 4-phenylbutyric acid (PBA) was identified.53 Oral administration of PBA resulted in a 20–50% increase of serum Z-AT protein levels. In other studies, Tg-mice expressing low levels of SERPINA1 proved immune-tolerant to recombinant SERPINA1 replacement therapy, and were successfully used to demonstrate its protective effect on smoke-induced lung damage.54 The onset of liver and lung pathology associated with Z-AT suggests that both environmental and genetic factors are critical in the progression of Z-AT associated disease and in other conformational diseases. Among individuals homozygous for Z-AT, there is a 2–3% risk of liver disease fatality in children, while all adults who undergo increasing hepatic damage, 25–50% will present with cirrhosis and hepatocellular carcinoma.48,55 Mouse genetics is a proven tool in the identification of factors such as modifier genes that give rise to variability in many human diseases, and can be applied to examine genes associated with the onset of liver cirrhosis in the Z-AT serpinopathy.56−58 The Pi-Z mouse is also a good model to test the role of different environmental factors on the progression of this disease, as has been done extensively for cigarette smoke induction of lung damage.59 Additional members of the Serpin gene family implicated in conformational disease include three of these expressed in the liver, SERPING1 (C1INH), SERPINA3 (ACT), and SERPINC1 (AT). SERPINI1 (Neuroserpin) causes the dementia familial encephalopathy with Neuroserpin inclusion bodies (FENIB).60 As yet, there is no report of a neuroserpinpolymerization FENIB mouse model. Finally, in addition to mutations that directly affect serpin folding and stability, a recent study has showed that simple over-expression of serpins in transgenic animals can result in pathological aggregation.61 Transgenic rats over-expressing Megsin (SERPINB7) resulted in protein inclusion bodies in the kidney and pancreas. This simple model fits well with the current hypothesis that many factors have the potential to play an important role in the development of conformational diseases. Critical to these diseases is a

ch04

Jan. 10, 2007

8:57

SPI-B429



121

switch from a normal folding pathway to a pathway that leads to protein aggregation. How easily this switch is flipped depends on protein sequence and concentration, and the cumulative input of protein folding and transport through the secretory pathway, protein degradation, and stress placed on the protein within the environment of the host cell.47 Imaginative use of mouse models will elucidate the relative contribution of these factors to serpin conformational disease.

3. Conclusions Serpin function in the mouse overlaps extensively but not completely with that in humans. With the exception of genes in Clades A and B, the total serpin gene content in the genomes of humans and mouse is well conserved. Many human diseases linked to serpin defects in the blood and multiple tissues can be mimicked in mice. Humanized mice expressing human serpins have also proven useful for therapeutic research. Several serpin gene deletion models, however, have failed to produce a phenotype. While this could reflect gene redundancy or compensation, it is more likely that the animals have not yet been exposed to critical environmental or immune challenges, and that more work (and imagination) is required to uncover a phenotype. With the availability of mouse genomic sequence and extensive techniques to modify and analyze specific target genes, there is little to limit the development of new mouse models of human serpin pathophysiology. It should be noted, however, that many serpin pathologies arise as a result of single amino acid substitution, and that simple gene deletion models may not fully recapitulate all the features of a human disease. As yet, there has been no report of a knock-in mouse model, or of a serpin mutation underpinning a developmental or disease phenotype identified in a genome-wide mutagenesis strategy. Nevertheless, transgenic and knockout mice are providing key insights into this fascinating protein superfamily.

References 1. Silverman GA, Bird PI, Carrell RW, et al. (2001) The serpins are an expanding superfamily of structurally similar but functionally diverse proteins. Evolution,

ch04

Jan. 10, 2007

8:57

SPI-B429

122

2.

3.

4.

5. 6.

7.

8.

9.

10. 11. 12.

13.

14.



mechanism of inhibition, novel functions, and a revised nomenclature. J Biol Chem 276(36):33293–33296. Irving JA, Pike RN, Lesk AM and Whisstock JC (2000) Phylogeny of the serpin superfamily: Implications of patterns of amino acid conservation for structure and function. Genome Res 10(12):1845–1864. Puente XS, Sanchez LM, Gutierrez-Fernandez A, Velasco G and Lopez-Otin C (2005) A genomic view of the complexity of mammalian proteolytic systems. Biochem Soc Trans 33(Pt 2):331–334. Puente XS, Sanchez LM, Overall CM and Lopez-Otin C (2003) Human and mouse proteases: A comparative genomic approach. Nat Rev Genet 4(7): 544–558. Charron Y, Madani R, Nef S, et al.(2005) Expression of Serpinb6 Serpins in Germ and Somatic Cells of Mouse Gonads. Mol Reprod Develop. Forsyth S, Horvath A and Coughlin P (2003) A review and comparison of the murine alpha1-antitrypsin and alpha1-antichymotrypsin multigene clusters with the human clade A serpins. Genomics 81(3):336–345. Kaiserman D, Knaggs S, Scarff KL, et al. (2002) Comparison of human chromosome 6p25 with mouse chromosome 13 reveals a greatly expanded ov-serpin gene repertoire in the mouse. Genomics 79(3):349–362. Askew DJ,AskewYS, KatoY, et al. (2004) Comparative genomic analysis of the clade B serpin cluster at human chromosome 18q21: Amplification within the mouse squamous cell carcinoma antigen gene locus. Genomics 84(1):176–184. Barbour KW, Goodwin RL, Guillonneau F, Wang Y, Baumann H and Berger FG (2002) Functional diversification during evolution of the murine alpha(1)proteinase inhibitor family: Role of the hypervariable reactive center loop. Mol Biol Evol 19(5):718–727. Hughes AL (2002) Natural selection and the diversification of vertebrate immune effectors. Immunol Rev 190:161–168. Ishiguro K, Kojima T, Kadomatsu K, et al. (2000) Complete antithrombin deficiency in mice results in embryonic lethality. J Clin Invest 106 (7):873–878. He L, Vicente CP, Westrick RJ, Eitzman DT and Tollefsen DM (2002) Heparin cofactor II inhibits arterial thrombosis after endothelial injury. J Clin Invest 109(2):213–219. Vicente CP, He L, Pavao MS and Tollefsen DM (2004) Antithrombotic activity of dermatan sulfate in heparin cofactor II-deficient mice. Blood 104(13): 3965–3970. Uhrin P, Dewerchin M, Hilpert M, et al. (2000) Disruption of the protein C inhibitor gene results in impaired spermatogenesis and male infertility. J Clin Invest 106(12):1531–1539.

ch04

Jan. 10, 2007

8:57

SPI-B429



123

15. Hayashi T, Nishioka J, Kamada H, et al. (2004) Characterization of a novel human protein C inhibitor (PCI) gene transgenic mouse useful for studying the role of PCI in physiological and pathological conditions. J Thromb Haemost 2(6):949–961. 16. Lijnen HR, Okada K, Matsuo O, Collen D and Dewerchin M (1999) Alpha2antiplasmin gene deficiency in mice is associated with enhanced fibrinolytic potential without overt bleeding. Blood 93(7):2274–2281. 17. Kim HS, Krege JH, Kluckman KD, et al. (1995) Genetic control of blood pressure and the angiotensinogen locus. Proc Natl Acad Sci USA 92(7):2735–2739. 18. Davisson RL, Kim HS, Krege JH, Lager DJ, Smithies O and Sigmund CD (1997) Complementation of reduced survival, hypotension, and renal abnormalities in angiotensinogen-deficient mice by the human renin and human angiotensinogen genes. J Clin Invest 99(6):1258–1264. 19. Han ED, MacFarlane RC, Mulligan AN, Scafidi J and Davis AE III (2002) Increased vascular permeability in C1 inhibitor-deficient mice mediated by the bradykinin type 2 receptor.[see comment]. J Clin Invest 109(8):1057–1063. 20. Funatsu H, Yamashita H, Nakamura S, et al. (2006) Vitreous levels of pigment epithelium-derived factor and vascular endothelial growth factor are related to diabetic macular edema. Ophthalmology 113(2):294–301. 21. Doll JA, Stellmach VM, Bouck NP, et al. (2003) Pigment epithelium-derived factor regulates the vasculature and mass of the prostate and pancreas. Nat Med 9(6):774–780. 22. Duh EJ, Yang HS, Suzuma I, et al. (2002) Pigment epithelium-derived factor suppresses ischemia-induced retinal neovascularization and VEGF-induced migration and growth. Invest Ophthalmol Vis Sci 43(3):821–829. 23. Mori K, Gehlbach P, Yamamoto S, et al. (2002) AAV-mediated gene transfer of pigment epithelium-derived factor inhibits choroidal neovascularization. Invest Ophthalmol Vis Sci 43(6):1994–2000. 24. Gao F, Shi HY, Daughty C, Cella N and Zhang M (2004) Maspin plays an essential role in early embryonic development. Development 131(7):1479–1489. 25. Park JM, Greten FR, Wong A, et al. (2005) Signaling pathways and genes that inhibit pathogen-induced macrophage apoptosis–CREB and NF-kappaB as key regulators. Immunity 23(3):319–329. 26. Dougherty KM, Pearson JM,Yang AY, Westrick RJ, Baker MS and Ginsburg D (1999) The plasminogen activator inhibitor-2 gene is not re quired for normal murine development or survival. Proc Natl Acad Sci USA 96(2):686–691. 27. Zhang M, Park SM, Wang Y, et al. (2006) Serine protease inhibitor 6 protects cytotoxic T cells from self-inflicted injury by ensuring the integrity of cytotoxic granules. Immunity 24(4):451–461.

ch04

Jan. 10, 2007

8:57

SPI-B429

124



28. Bird CH SV, Sun J, Hirst CE, Novak A, Kumar S, Trapani JA and Bird PI (1998) Selective regulation of apoptosis: The cytotoxic lymphocyte serpin proteinase inhibitor 9 protects against granzyme B-mediated apoptosis without perturbing the Fascell death pathway. Mol Cell Biol 18(11):6387– 6398. 29. Scarff KL, Ung KS, Nandurkar H, Crack PJ, Bird CH and Bird PI (2004) Targeted disruption of SPI3/Serpinb6 does not result in developmental or growth defects, leukocyte dysfunction, or susceptibility to stroke. Mol Cell Biol 24(9):4075–4082. 30. Scarff KL, Ung KS, Sun J and Bird PI (2003) A retained selection cassette increases reporter gene expression without affecting tissue distribution in SPI3 knockout/GFP knock-in mice. Genesis: J Genet Develop 36(3):149–157. 31. Vivien D and Buisson A (2000) Serine protease inhibitors: Novel therapeutic targets for stroke? J Cereb Blood Flow Metab 20(5):755–764. 32. Yepes M and Lawrence DA (2004) New functions for an old enzyme: Nonhemostatic roles for tissue-type plasminogen activator in the central nervous system. Exp Biol Med (Maywood) 229(11):1097–1104. 33. Miranda E and Lomas DA (2006) Neuroserpin: A serpin to think about. Cell Mol Life Sci 34. Madani R, Kozlov S, Akhmedov A, et al. (2003) Impaired explorative behavior and neophobia in genetically modified mice lacking or overexpressing the extracellular serine protease inhibitor neuroserpin. Mol Cell Neurosci 23(3):473–494. 35. Cinelli P, Madani R, Tsuzuki N, et al. (2001) Neuroserpin, a neuroprotective factor in focal ischemic stroke. Mol Cell Neurosci 18(5):443–457. 36. Yepes M, Sandkvist M, Coleman TA, et al. (2002) Regulation of seizure spreading by neuroserpin and tissue-type plasminogen activator is plasminogenindependent. J Clin Invest 109(12):1571–1578. 37. Yepes M, Sandkvist M, Moore EG, Bugge TH, Strickl DK and Lawrence DA (2003) Tissue-type plasminogen activator induces opening of the blood-brain barrier via the LDL receptor-related protein. J Clin Invest 112(10):1533–1540. 38. Huang YY, Bach ME, Lipp HP, et al. (1996) Mice lacking the gene encoding tissue-type plasminogen activator show a selective interference with late-phase long-term potentiation in both Schaffer collateral and mossy fiber pathways. Proc Natl Acad Sci USA 93(16):8699–8704. 39. Seeds NW, Basham ME and Haffke SP (1999) Neuronal migration is retarded in mice lacking the tissue plasminogen activator gene. Proc Natl Acad Sci USA 96(24):14118–14123.

ch04

Jan. 10, 2007

8:57

SPI-B429



125

40. Seeds NW, Basham ME and Ferguson JE (2003) Absence of tissue plasminogen activator gene or activity impairs mouse cerebellar motor learning. J Neurosci 23(19):7368–7375. 41. Pawlak R, Magarinos AM, Melchor J, McEwen B and Strickland S (2003) Tissue plasminogen activator in the amygdala is critical for stress-induced anxiety-like behavior. Nat Neurosci 6 (2):168–174. 42. Tsirka SE, Gualandris A, Amaral DG and Strickland S (1995) Excitotoxininduced neuronal degeneration and seizure are mediated by tissue plasminogen activator. Nature 377(6547):340–344. 43. LuthiA,Van der Putten H, Botteri FM, et al. (1997) Endogenous serine protease inhibitor modulates epileptic activity and hippocampal long-term potentiation. J Neurosci 17(12):4688–4699. 44. Kvajo M, Albrecht H, Meins M, et al. (2004) Regulation of brain proteolytic activity is necessary for the in vivo function of NMDA receptors. J Neurosci 24(43):9734–9743. 45. Hengst U, Albrecht H, Hess D and Monard D (2001) The phosphatidylethanolamine-binding protein is the prototype of a novel family of serine protease inhibitors. J Biol Chem 276(1):535–540. 46. Meins M, Piosik P, Schaeren-Wiemers N, et al. (2001) Progressive neuronal and motor dysfunction in mice overexpressing the serine protease inhibitor protease nexin-1 in postmitotic neurons. J Neurosci 21(22): 8830–8841. 47. Carrell RW (2005) Cell toxicity and conformational disease. Trends Cell Biol 15(11):574–580. 48. Crowther DC, Belorgey D, Miranda E, Kinghorn KJ, Sharp LK and Lomas DA (2004) Practical genetics: Alpha-1-antitrypsin deficiency and the serpinopathies. European J Hum Genet 12(3):167–172. 49. Mahadeva R, Atkinson C, Li Z, et al. (2005) Polymers of Z alpha1-antitrypsin co-localize with neutrophils in emphysematous alveoli and are chemotactic in vivo. Am J Pathol 166(2):377–386. 50. Dycaico MJ, Grant SG, Felts K, et al. (1988) Neonatal hepatitis induced by alpha 1-antitrypsin: A transgenic mouse model. Science 242(4884): 1409–1412. 51. Carlson JA, Rogers BB, Sifers RN, et al. (1989) Accumulation of PiZ alpha 1-antitrypsin causes liver damage in transgenic mice. J Clin Invest 83(4): 1183–1190. 52. Ali R, Perfumo S, della Rocca C, et al. (1994) Evaluation of a transgenic mouse model for alpha-1-antitrypsin (AAT) related liver disease. Ann Hum Genet 58 (Pt 4):305–320.

ch04

Jan. 10, 2007

8:57

SPI-B429

126



53. Burrows JA, Willis LK and Perlmutter DH (2000) Chemical chaperones mediate increased secretion of mutant alpha 1-antitrypsin (alpha 1-AT) Z: A potential pharmacological strategy for prevention of liver injury and emphysema in alpha 1-AT deficiency. Proc Natl Acad Sci USA 97(4):1796–1801. 54. Churg A, Wang RD, Xie C and Wright JL (2003) Alpha-1-antitrypsin ameliorates cigarette smoke-induced emphysema in the mouse [see comment]. Am J Resp Crit Care Med 168(2):199–207. 55. WHO (1997) Alpha 1-antitrypsin deficiency: memorandum from a WHO meeting. Bull World Health Organization 75(5):397–415. 56. Dragani TA (2003) 10 years of mouse cancer modifier loci: Human relevance. Cancer Res 63(12):3011–3018. 57. Guo Z, Mori M, Fu X, et al. (2002) Amyloidosis modifier genes in the less amyloidogenic a/j mouse strain. Lab Invest 83(11):1605–1613. 58. Falvella FS, Pascale RM, Gariboldi M et al. (2002) Stearoyl-CoA desaturase 1 (Scd1) gene overexpression is associated with genetic predisposition to hepatocarcinogenesis in mice and rats. Carcinogenesis 23(11):1933–1936. 59. Wright JL and Churg A (2002) Animal models of cigarette smoke-induced COPD. Chest 122(6 Suppl):301S–306S. 60. Davis RL, Shrimpton AE, Holohan PD, et al. (1999) Familial dementia caused by polymerization of mutant neuroserpin. Nature 401(6751):376–379. 61. Inagi R, Nangaku M, Usuda N, et al. (2005) Novel serpinopathy in rat kidney and pancreas induced by overexpression of megsin. J Am Soc Nephrol 16(5):1339–1349. 62. He S, Lin YL and Liu YX (1999) Functionally inactive protein C inhibitor in seminal plasma may be associated with infertility. Mol Hum Reprod 5(6): 513–519. 63. Stenflo J, Kjellberg M, Strandberg K, Svensson PJ, Kolbel T and Lindblad B (2006) The complex between activated protein C and protein C inhibitor: A clinically useful indicator of aortic aneurysms? Blood Cells Mol Dis 36(2): 118–121. 64. Bhiladvala P, Strandberg K, Stenflo J and Holm J (2005) Early identification of acute myocardial infarction by activated protein C-protein C inhibitor complex. Thromb Res 65. Ding Y, Stec DE and Sigmund CD (2001) Genetic evidence that lethality in angiotensinogen-deficient mice is due to loss of systemic but not renal angiotensinogen. J Biol Chem 276(10):7431–7436. 66. Zhou HM, Bolon I, Nichols A, Wohlwend A and Vassalli JD (2005) Overexpression of plasminogen activator inhibitor type 2 in basal keratinocytes enhances papilloma formation in transgenic mice. Cancer Res 61(3):970–976.

ch04

Jan. 10, 2007

8:57

SPI-B429



127

67. Shi HY, Lydon JP and Zhang M (2004) Hormonal defect in maspin heterozygous mice reveals a role of progesterone in pubertal ductal development. Mol Endocrinol 18(9):2196–2207. 68. Zhang M, Magit D, Botteri F, et al. (1999) Maspin plays an important role in mammary gland development. Develop Biol 215(2):278–287. 69. Zhang M, Shi Y, Magit D, Furth PA and Sager R (2002) Reduced mammary tumor progression in WAP-TAg/WAP-maspin bitransgenic mice. Oncogene 19(52):6053–6058. 70. Dewerchin M, Herault JP, Wallays G, et al. (2003) Life-threatening thrombosis in mice with targeted Arg48-to-Cys mutation of the heparin-binding domain of antithrombin. Circ Res 93(11):1120–1126. 71. Yanada M, Kojima T, Ishiguro K, et al. (2002) Impact of antithrombin deficiency in thrombogenesis: lipopolysaccharide and stress-induced thrombus formation in heterozygous antithrombin-deficient mice. Blood 99(7): 2455–2458. 72. Bernardi F, Legnani C, Micheletti F, et al. (1996)A heparin cofactor II mutation (HCII Rimini) combined with factor V Leiden or type I protein C deficiency in two unrelated thrombophilic subjects. Thromb Haemost 76(4):505–509. 73. Tollefsen DM (2002) Heparin cofactor II deficiency. Arch Pathol Lab Med 126(11):1394–1400. 74. Erickson LA, Fici GJ, Lund JE, Boyle TP, Polites HG and Marotti KR (1990) Development of venous occlusions in mice transgenic for the plasminogen activator inhibitor-1 gene. Nature 346(6279):74–76. 75. Eren M, Painter CA, Atkinson JB, Declerck PJ and Vaughan DE (2002) Agedependent spontaneous coronary arterial thrombosis in transgenic mice that express a stable form of human plasminogen activator inhibitor-1. Circulation 106(4):491–496. 76. Carmeliet P, Stassen JM, Schoonjans L, et al. (1993) Plasminogen activator inhibitor-1 gene-deficient mice. II. Effects on hemostasis, thrombosis, and thrombolysis. J Clin Invest 92(6):2756–2760. 77. Meade TW, Ruddock V, Stirling Y, Chakrabarti R and Miller GJ (1993) Fibrinolytic activity, clotting factors, and long-term incidence of ischaemic heart disease in the Northwick Park Heart Study. Lancet 342(8879):1076– 1079. 78. Yamamoto K, Takeshita K, Shimokawa T, et al. (2002) Plasminogen activator inhibitor-1 is a major stress-regulated gene: Implications for stress-induced thrombosis in aged individuals. Proc Natl Acad Sci USA 99(2):890–895. 79. Oda T, Jung YO, Kim HS, et al. (2001) PAI-1 deficiency attenuates the fibrogenic response to ureteral obstruction. Kidney Int 60(2):587–596.

ch04

Jan. 10, 2007

8:57

SPI-B429

128



80. Eitzman DT, Westrick RJ, Xu Z, Tyson J and Ginsburg D (2000) Plasminogen activator inhibitor-1 deficiency protects against atherosclerosis progression in the mouse carotid artery. Blood 96(13):4212–4215. 81. Schafer K, Fujisawa K, Konstantinides S and Loskutoff DJ (2001) Disruption of the plasminogen activator inhibitor 1 gene reduces the adiposity and improves the metabolic profile of genetically obese and diabetic ob/ob mice. Faseb J 15(10):1840–1842. 82. Ma LJ, Mao SL, Taylor KL, et al. (2004) Prevention of obesity and insulin resistance in mice lacking plasminogen activator inhibitor 1. Diabetes 53(2): 336–346. 83. Bajou K, Maillard C, Jost M, et al. (2004) Host-derived plasminogen activator inhibitor-1 (PAI-1) concentration is critical for in vivo tumoral angiogenesis and growth. Oncogene 23(41):6986–6990. 84. McMahon GA, Petitclerc E, Stefansson S, et al. (2001) Plasminogen activator inhibitor-1 regulates tumor growth and angiogenesis. J Biol Chem 276(36):33964–33968. 85. Devy L, Blacher S, Grignet-Debrus C, et al. (2002) The pro- or antiangiogenic effect of plasminogen activator inhibitor 1 is dose dependent. Faseb J 16(2):147–154. 86. MurerV, Spetz JF, Hengst U,Altrogge LM, deAgostiniA and Monard D (2001) Male fertility defects in mice lacking the serine protease inhibitor protease nexin-1. Proc Natl Acad Sci USA 98(6):3029–3033. 87. Holekamp NM, Bouck N and Volpert O (2002) Pigment epithelium-derived factor is deficient in the vitreous of patients with choroidal neovascularization due to age-related macular degeneration. Am J Ophthalmol 134(2):220–227. 88. Bouck N (2002) PEDF: anti-angiogenic guardian of ocular function. Trends Mol Med 8(7):330–334. 89. Matsuoka Y, Kubota H, Adachi E, et al. (2004) Insufficient folding of type IV collagen and formation of abnormal basement membrane-like structure in embryoid bodies derived from Hsp47-null embryonic stem cells. Mol Biol Cell 15(10):4467–4475. 90. Nagai N, Hosokawa M, Itohara S, et al. (2000) Embryonic lethality of molecular chaperone hsp47 knockout mice is associated with defects in collagen biosynthesis. J Cell Biol 150(6):1499–1506. 91. Loftus SK, Cannons JL, IncaoA, et al. (2005)Acinar cell apoptosis in Serpini2deficient mice models pancreatic insufficiency. PLoS Genet 1(3):e38. 92. Yepes M, Sandkvist M, Wong MK, et al. (2000) Neuroserpin reduces cerebral infarct volume and protects neurons from ischemia-induced apoptosis. Blood 96(2):569–576.

ch04

Jan. 10, 2007

8:57

SPI-B429



129

93. Wu YP, Siao CJ, Lu W, et al. (2000) The tissue plasminogen activator (tPA)/plasmin extracellular proteolytic system regulates seizure-induced hippocampal mossy fiber outgrowth through a proteoglycan substrate. J Cell Biol 148(6):1295–1304. 94. Siconolfi LB, Seeds NW (2001) Mice lacking tPA, uPA, or plasminogen genes showed delayed functional recovery after sciatic nerve crush. J Neurosci 21(12):4348–4355. 95. Hoover-Plow J, Wang N and Ploplis V (1999) Growth and behavioral development in plasminogen gene-targeted mice. Growth Dev Aging 63(1-2):13–32.

ch04

Jan. 10, 2007 8:57 SPI-B429



ch05

Jan. 10, 2007

8:57

SPI-B429


5 Serpins in Prokaryotes Qingwei Zhang, Ruby Law, Ashley M. Buckle, Lisa Cabrita, Sheena McGowan, James A. Irving, Noel G. Faux, Arthur M. Lesk, Stephen P. Bottomley and James C. Whisstock

1. Introduction An extensive bioinformatic analysis of the genomic sequence data available in 2000 revealed that serpin-like sequences could only be identified in higher animals, their viral pathogens (the poxviridae), and higher plants.1 Despite the presence of serpins in higher plants and animals, no obvious ancestral serpin-like sequence could be identified in fungi or prokaryotes using sensitive informatic approaches. This distribution suggested that serpins had evolved as a specialized group of protease inhibitors associated with the development of complex proteolytic cascades, present in multicellular organisms.1 However, since 2000, the amount of genomic data available has substantially increased and classical serpin-like sequences have been detected in a number of unusual prokaryotes.2,3 In addition, serpins could be identified in several primitive eukaryotes, including Alexandrium tamarense, Chlamydomonas reinhardtii, Eimeria tenella, Entamoeba histolytica, Entamoeba invadens, Entamoeba moshkovskii, Giardia lamblia, and Toxoplasma gondii.3 Sequence alignments revealed that most prokaryote serpins contained the classical “inhibitory hinge” motif in the reactive center loop (RCL) characteristic of mammalian inhibitory serpins.4 These data suggested that prokaryote serpins, similar to their mammalian counterparts, might 131

ch05

Jan. 10, 2007

8:57

SPI-B429

132


Q. Zhang et al.

generally function to inhibit proteolytic enzymes utilizing serpin conformational change and “inhibition by distortion”.5 Interestingly, given their likely role in protease inhibition, several serpin-like sequences could be identified in the genomes of thermophilic organisms that grow in extreme (55–100◦ C) environments. This was surprising since mammalian serpins adopt a metastable conformation and are heat labile, as a result of the requirement to undergo conformational change as part of inhibitory function (see Chap. 3). Due to their metastability, mammalian serpins are also vulnerable to destabilizing mutations that cause the polymerization event underlying serpin-related human disease.6,7 The presence of serpins in thermophilic organisms, therefore, suggested that these proteins may have evolved to resist inappropriate conformational change, while still functioning to inhibit target proteases. This hypothesis was supported by a structural and functional investigation of the serpin (thermopin) from the thermophile Thermobifida fusca,8 which revealed that this molecule is able to adopt both the native and cleaved conformation and inhibit model chymotrypsin-like proteases.9,10 Thus, despite somewhat sporadic distribution, serpins are the only protease inhibitor family that can be identified in all the major branches of life,11 and it is suggested that the serpin-like fold may be to dated back to the archean era (3.8–2.5 billion years ago). In this review, the distribution, structure, and evolution of serpins from prokaryotes are discussed. We further investigate putative functions of prokaryote serpins and describe how some of these molecules may have evolved to survive and function in an extreme environment.

2. Distribution of Prokaryotic Serpins An iterative PSI-BLAST search using the sequence of α1 -antitrypsin as a probe identified 36 putative serpins from 23 different bacteria and 5 archaea (Table 1). Six organisms have more than one serpin genes: Oceanobacillus iheyensis, Gloeobacter violaceus and Symbiobacterium thermophilum possess two serpin molecules, and Clostridium thermocellum, Methanosarcina acetivorans, and Methanococcoides burtonii possess three serpin molecules. It is important to note that serpins are not broadly distributed in prokaryotes, since they are absent from all other sequenced bacterial organisms (260 finished and 762 ongoing bacterial

ch05

Jan. 10, 2007 8:57 SPI-B429

Table 1 Serpins in bacteria and archaea. Id (%)a

P1 P1

Peptidase present

Taxonomy

References

B

Alkaliphilus metalliredigenes QYMF [1]

ZP_00799881

25

EEGSEAAAATVVVMTES A.....MA

?

Firmicutes; Clostridia; Clostridiales; Clostridiaceae; Alkaliphilus

JGI-PGF

B

Anabaena variabilis ATCC 29413 [2]

YP_325165

25

EEGTEASAATSVGMVAT S.....LREPQ

1 S8B

B

Arthrobacter sp. FB24 [3]

ZP_00413322

24

EKGTVAAAVTQINGAVT SAP.....PQ

1 S1A unassigned 1 S1C unassigned 1 S8A unassigned

Cyanobacteria; Nostocaceae; Anabaena Actinobacteria; Actinomycetales; Micrococcaceae; Arthrobacter

NCBI

JGI-PGF

[1] Novel alkaliphilic microorganism that can reduce metals at pH values up to 11.0 in the presence of elevated salt levels.12 [2] Filamentous nitrogen-fixing cyanobacterium, has two Mo-dependent nitrogenises: Nif1 functions in heterocyst formation and Nif2 functions under anoxic conditions in vegetative cells.13 [3] Aerobic high GC content Gram positive bacteria, extreme chromium tolerance (http:// genome.jgi-psf.org/ draft_microbes/ art_f/art_f. home.html). (Continued)

133


NCBI accession number

Serpins in Prokaryotes

Organism

Bacteria (B) or archae (A)

ch05

Jan. 10, 2007 8:57

(Continued)


Id (%)a

B

Bacillus licheniformis ATCC 14580 [4]

AAU21907b

25

EAGTEASAATAVEIIES AP.....V

1 S1B 1 S1C 5 S8A 1 S8A unassigned

Firmicutes; Bacillaceae; Bacillus

B

Bifidobacterium longum DJO10A [5]

ZP_00121847

20

EAGAKAMSFTKVGADSV S.....APV

2 C1B unassigned 1 S1C unassigned

Actinobacteria; Bifidobacteriaceae; Bifidobacterium

NCBI

B

Bifidobacterium longum NCC2705 (See Ref. 17)

NP_695337

20

EAGAKAMSFTKVGADSV S.....APV

2 C1B unassigned 1 S1C unassigned

Actinobacteria; Actinobacteridae; Bifidobacteriaceae; Bifidobacterium

17

Taxonomy

References

15

B. licheniformis is a Gram-positive, spore-forming soil bacterium that is used in the biotechnology industry to manufacture enzymes, antibiotics, biochemicals and consumer products.14 [5] An obligate anaerobe, belonging to the Actinomycetales, a branch of the high-GC Gram-positive Bacteria. B. longum, a harmless bacterium in human, is considered to play an important role in maintaining a healthy gastrointestinal (GIT) tract by preventing diarrhea, improving lactose intolerance, and participating in immunomodulation.16 (Continued)

Q. Zhang et al.

[4]

Peptidase present


P1 P1

Organism


SPI-B429

134

Table 1

ch05

Jan. 10, 2007 8:57

Organism

(Continued)


Id (%)a

P1 P1

Peptidase present

Taxonomy

References

Chlorobium limicola DSM 245 [6]

ZP_00511389

30

EKGTEAAAATAVVMQLR S.....AMPM

1 S1C unassigned 1 S8A

Chlorobi; Chlorobiales; Chlorobiaceae; Chlorobium

JGI-PGF

B

Chlorobium phaeobacteroides BS1 (See [6])

ZP_00530649

29

ETGTEASAATGVVVGLT S.....AVQA

1 C1A unassigned 1 S1C unassigned

Chlorobi; Chlorobia; Chlorobiaceae; Chlorobium

JGI-PGF

B

Clostridium thermocellum ATCC 27405 [7]

ZP_00503703

23 26

3 S1C unassigned 1 S8A unassigned

Firmicutes; Clostridia; Clostridiaceae; Clostridium

JGI-PGF

ZP_00503704 ZP_00503588

24

EKGTEASGVVVIPIAPT S.....IA EDGSTAAGSTVVRMIDG AAIG..... EKGTEASSSVVVIPVPG FG.....


B

[6] The Chlorobiaceae sp. (green sulfur bacteria) carry out anoxygenic photosynthesis in which reduced sulfur compounds (sulfide, sulfur, thiosulfate) serve as electron donors for the reduction of carbon dioxide. These organisms are strictly anaerobic and obligately phototrophic.18 [7] A thermophilic, anaerobic, cellulolytic bacterium that produces ethanol and acetic acid as major fermentation end products.19 (Continued)

135



SPI-B429

Table 1

ch05

Jan. 10, 2007 8:57

Id (%)a

B

Crocosphaera watsonii WH 8501 [8]

ZP_00515710

24

B

Dehalococcoides ethenogenes 195 [9]

YP_180970

B

Dehalococcoides sp. CBDB1 [10]

CAI82463

P1 P1

Peptidase present

Taxonomy

References

EEGVTSPPEDDLLIKKA S.....MDNTK


Cyanobacteria; Chroococcales; Crocosphaera.

JGI

27

EDGTEAAAATAVSMNLT S.....APT

1 S1C unassigned

Chloroflexi; Dehalococcoidetes; Dehalococcoides

21

28

EAGTEAAAATAVTMNLT S.....APA

Chloroflexi; Dehalococcoidetes; Dehalococcoides.

NCBI

[8] A unicellular marine cyanobacterium that fixes nitrogen, presumably at night, while growing photosynthetically during the day. It is an abundant organism and contributes significantly to global carbon and nitrogen biogeochemical cycles (http:// genome.jgi-psf.org/draft_microbes/ crowa/crowa.home.html). [9] Is able to dechlorinate tetrachloroethene to ethylene.20 [10] Oxygen sensitive strain capable of reductive dechlorination of chlorobenzenes.22 (Continued)



(Continued)

Q. Zhang et al.

Organism


SPI-B429

136

Table 1

ch05

Jan. 10, 2007 8:57

Organism

P1 P1

26

EKGTEAAAVTGVEVNTT S.....MPI

NP_924916b

24

EEGTEAAAATGVIVART S.....AVAT

NP_924792

20

ETGLTRPSAATGGERFA LP.....LRPGE

B

Desulfitobacterium ZP_00559302b hafniense DCB-2 [11]

B

Gloeobacter violaceus PCC 7421 [12]

Peptidase present

Taxonomy

References

1 S1B unassigned 3 S1C unassigned 4 S8A unassigned 2 S8B unassigned

Firmicutes; Clostridia; Peptococcaceae; Desulfitobacterium

JGI-PGF

4 C14 unassigned 1 S1C 3 S1C unassigned 2 S1X unassigned 2 S8A unassigned

Cyanobacteria; Gloeobacterales; Gloeobacter

25

(Continued)

137

[11] Gram-positive, of low G+C content, spore-forming bacteria Known for reductively dechlorinating chlorophenols.23 [12] A rod-shape unicellular cyanobacterium lacking thylakoid membranes.24


Id (%)a




(Continued)

SPI-B429

Table 1

ch05

Jan. 10, 2007 8:57

Organism


Id (%)a

P1 P1

Peptidase present

Taxonomy

References

AAV34162b

24

ELGTEAAAATAVGMVPV S.....

1 S1C unassigned

Proteobacteria; Deltaproteobacteria; Cystobacterineae; Myxococcus

B

Nostoc punctiforme PCC 73102 [14]

ZP_00107412

25

EEGTEAAAATSVGIVAT S.....LRDEP

6 S1C unassigned 1 S1X unassigned 3 S8A unassigned 1 S8B

Cyanobacteria; Nostocales; Nostocaceae; Nostoc

NCBI

B

Nostoc sp. PCC 7120 [15]

BAB72735

27

EEGTEASAATSVGMVAT S.....LREPQ

4 C14 unassigned 4 S1C unassigned 4 S8A unassigned 1 S8B

Cyanobacteria; Nostocales; Nostocaceae; Nostoc

29

[13] Gram-negative soil bacterium that exhibits a communal lifestyle during vegetative growth and multicellular development.26 [14] Filamentous nitrogen-fixing cyanobacterium, is able to form a symbiotic relationship with a variety of terrestrial plants, the only known example of endocytobiosis between a fungus and cyanobacterium.27 [15] Cyanobacteria that possess an oxygenic photosynthetic system very similar to that found in higher plants.28 (Continued)


Myxococcus xanthus [13]

Q. Zhang et al.

B

SPI-B429


(Continued)

138

Table 1

ch05

Jan. 10, 2007 8:57

(Continued)

Id (%)a

P1 P1

Oceanobacillus iheyensis HTE831 [16]

NP_694094b

28

NP_694093

25

VRGTEAAGATSVEMKEE S.....AVVSP EEGTEAAGATSVEMKTT S.....IEIG

B

Parachlamydia sp. UWE25 [17]

CAF23443b

27

B

Symbiobacterium thermophilum IAM 14863 [18]

YP_076854

27

YP_076853

26

B

Peptidase present

Taxonomy

References

2 S1C 1 S1C unassigned 5 S8A 3 S8A unassigned

Firmicutes; Bacillaceae; Oceanobacillus

30

EEGSDSKAWGVNMEEPE D.....S

1 S1C unassigned

Chlamydiae; Parachlamydiaceae; Parachlamydia

31

EEGTEAAAATGVAVTVT S.....APAG EEGTEAAAATVVGITAG A.....APPP

1 S1C unassigned 1 S8A 1 S8A unassigned

Actinobacteria; Symbiobacterium

32

139

[16] Extremely halotolerant and facultatively alkaliphilic bacterium, isolated from deep-sea sediment collected at a depth of 1050 m. This is the first spore-forming Gram-positive bacterium isolated from the bathypelagic zone.30 [17] Obligate highly specialized intracellular bacteria that live within the eukaryotic cell. This organism has unique adaptations, for example, the reduction of bacterial metabolism and the exploitation of host metabolites.31 [18] Gram-positive themophile (optimum growth temperature 60◦ C) that produces thermostable tryptophanase. Can grow only with the coexistence of an associating thermophilic Bacillus spp.32 (Continued)




Organism


SPI-B429

Table 1

ch05

Jan. 10, 2007 8:57

(Continued) P1 P1


Id (%)a

B

Thermoanaerobacter tengcongensis MB4 [19]

NP_623161b

23

ELGTKAGAVTSVDITAA G.....IPV


Firmicutes; Clostridia; Thermoanaerobacter

34

B

Thermobifida fusca YX [20]

YP_289989

23

ERGAEGAAATAAMMLLA G.....AMPPR

3 S1C unassigned 1 S1E unassigned 1 S1X unassigned

Actinobacteria; Actinomycetales; Streptosporangineae; Nocardiopsaceae; Thermobifida

US DOE Joint Genome Institute

A

Methanococcoides burtonii DSM 6242 [21]

ZP_00563267

26

1 S8A

27

Euryarchaeota; Methanomicrobia; Methanosarcinaceae; Methanococcoides

JGI-PGF

ZP_00561746 ZP_00562519

25

EEGTEATAATAIEAVDS CLP.....LPG EEGTEAAAATAIEATDS AP.....MPG EQGTEAAAATLVSFFMG YS.....A

Peptidase present

Taxonomy

References

Q. Zhang et al.

[19] An obligately anaerobic, rod-shaped, gram-negative, saccharolytic bacterium. Lives at 75◦ C in hot spring water.33 [20] Thermophilic soil bacterium, optimal growth temperature is 55◦ C, plays an important role in degradation of plant detritus.8 [21] A flagellated, motile methanogen isolated from permanently cold (1–2◦ C), methanesaturated waters from the bottom of Ace Lake, Antarctica.35 (Continued)


Organism


SPI-B429

140

Table 1

ch05

Jan. 10, 2007 8:57

A

Methanosarcina acetivorans C2A [22]

Methanosarcina mazei Go1 [23]

P1 P1


Id (%)a

NP_617161b

27

EKGTEAAAATGAIILEE E.....IE

NP_618277b

25

NP_617514c

23

EEGTEAAAATMEEMAMG VS.....ISWDA EEVTEAAASTTVFIGSV SS.....SKS

NP_634699b

27

EEGTEAAAATGVGMTVG MDFS.....WDS

Peptidase present

Taxonomy

References

5 C1A unassigned 1 S1X unassigned

Euryarchaeota; Methanosarcinales; Methanosarcinaceae; Methanosarcina

36

3 C1A unassigned 1 S1X unassigned 2 S8A unassigned

Euryarchaeota; Methanomicrobia; Methanosarcinaceae; Methanosarcina

37


A

Organism

(Continued)

[22] An acetate-utilizing methanogen Methanosarcineae.36 [23] Great ecological importance since they are the only organisms fermenting acetate, methylamines and methanol to methane, carbon dioxide and ammonia.37 (Continued)

141



SPI-B429

Table 1

ch05

Jan. 10, 2007 8:57

(Continued)

Id (%)a

A

Pyrobaculum aerophilum str IM2 [24]

NP_558344

23

ENGVVAAAATAVVFKPV CAK.....G

A

Thermococcus kodakaraensis KOD1 [25]

BAD85971

31

ENGTEAAAATAVTLTMN AP.....MQEK

Peptidase present

Taxonomy

References

1 S1C unassigned 1 S8A 1 S8A unassigned

Crenarchaeota; Thermoprotei; Thermoproteaceae; Pyrobaculum

38

1 C1A unassigned 1 S8A unassigned

Euryarchaeota; Thermococci Thermococcaceae; Thermococcus

40

[24] First isolated from a boiling marine water hole, this is a hyperthermophilic (Tmax 104◦ C, Topt 100◦ C) and metabolically versatile member of the crenarchaea.38 [25] A hyperthermophilic archaeon isolated from a solfatara (102◦ C, pH 5.8) on the shore of Kodakara Island Kagoshima.39 a b c

Percent identity when aligned with α1 -antitrypsin. N-terminuses contain signal peptide predicted by program GLOBPLOT.41 In this sequence, there is a stop codon after 341Y.

Q. Zhang et al.



P1 P1

Organism


SPI-B429

142

Table 1

ch05

Jan. 10, 2007

8:57

SPI-B429



143

genome projects, and 24 finished and 38 ongoing archaeal genome projects) (as on November 18, 2005). To date, no putative serpins could be identified in fungi; however, given the broad and sporadic distribution of serpins, it is suggested that additional sequencing data may reveal the presence of such molecules in more specialized fungi.

3. Evolutionary Relationships Prokaryote serpins contain the majority of the conserved motifs characteristic of mammalian inhibitory serpins.1 In particular, all but one (NP_924792 from G. violaceus) of the serpins discussed here contain the conserved pattern of small residues (Ala, Ser, or Gly) from P9 to P12 within the RCL (the “hinge region,” Ref. 4), as well as many of the highly conserved residues in two other functionally important regions, the breach and the shutter (Fig. 1). Together, these data suggest that prokaryote serpins are able to undergo conformational change and are likely to be able to inhibit proteases. Sequence alignments (Fig. 1) reveal that in general, the prokaryotic serpins are distantly related to mammalian serpins such as antitrypsin (∼ 20– 31% identity to α1 -antitrypsin, see Table 1). Phylogenetic analysis of large, diverse protein families is often hampered by a paucity of sequence data and precise patterns of orthology are difficult to elucidate. Indeed, to date, previous phylogenetic analyses of prokaryote serpins demonstrated that these molecules do not cluster with any of the higher eukaryote clades identified,2,3 and no support for horizontal gene transfer could be found. In this study, a bootstrapped neighbor-joining tree was built revealing that the 36 prokaryotic serpins identified, fell into nine well-supported clade groups and eight orphans (Fig. 2). Seven archaeal serpins from M. acetivorans, M. mazei, and M. burtonii form a well-supported clade with 100% bootstrap support. The other two archaeal serpins from T. kodakaraensis and Pyrobaculum aerophilum cannot be placed within this or any other clade with significant support, and are therefore orphans (Fig. 2). Generally, where more than one bacterial serpins are present in a single organism, these molecules are closely related, suggesting a recent duplication event, for example, the two serpins from O. iheyensis share a common ancestor as do the two serpins from S. thermophilum. Three

ch05

Jan. 10, 2007 8:57 SPI-B429

144


Q. Zhang et al. Fig. 1. Alignment of prokaryotic serpins. Residues are colored according to type: polar uncharged (green), basic (blue), and non-polar (yellow). Those residues that are identical to residues conserved in >70% of superfamily1 appear in white-on-black. Elements of consensus secondary structure are shown at the top. The P1 P1 residues are boxed in black lines. Regions of the alignment >60 residues N-terminal to hA are not shown. To identify bacterial serpins, PSI-Blast42 searches were performed till convergence on GENPEPT (Novermber 23, 2005) using the sequence of human α1 -antitrypsin as a probe. Default parameters were used with an inclusion threshold of 0.005.

ch05

Jan. 10, 2007

8:57

SPI-B429

145



(Continued) Fig. 1.

ch05

Jan. 10, 2007

8:57

SPI-B429

146


Q. Zhang et al.

(Continued) Fig. 1.

ch05

Jan. 10, 2007

8:57

SPI-B429

147



(Continued) Fig. 1.

ch05

Jan. 10, 2007

8:57

SPI-B429

148


Q. Zhang et al.

(Continued) Fig. 1.

ch05

Jan. 10, 2007

8:57

SPI-B429

149



(Continued) Fig. 1.

ch05

Jan. 10, 2007

8:57

SPI-B429

150


Q. Zhang et al.

Fig. 2. The bootstrap neighbor-joining consensus tree illustrating the evolutionary relationships between the bacterial serpins. Yellow filled ovals indicate the major clades discussed in this study. The P1 P1 is shown in blue next to each serpin. A structure-based alignment was generated of α1-antitrypsin (pdb accession 1qlp), human antithrombin (1azx), Manduca sexta serpin (1sek), and native thermopin (1sng) using program “MUSTANG” (Konagurthu et al., in press by Proteins: Strucutre, Function, and Bioinformatics). The 36 prokaryotic serpins were aligned to this “seed” with reference to secondary structure-specific gap penalties using CLUSTALW.43 The sequence alignment was used to derive bootstrapped neighborjoining trees (500 replicates, seed =64,238) and clusters of evolutionarily related sequences were obtained using MEGA 3.1.44

ch05

Jan. 10, 2007

8:57

SPI-B429



151

cyanobacteria, Nostoc punctiforme, Nostoc sp., and Anabaena variabilis form a well-supported clade as do four serpins from two Fermicutes (Alkaliphilus metalliredigenes and Clostridium thermocellum). Three other such clades are apparent: Clostridia (Desulfitobacterium hafniense and Thermoanaerobacter tengcongensis), Chlorobium (Chlorobium limicola and C. phaeobacteroides), Dehalococcoides (D. ethenogenes 195 and D. sp. CBDB1), and Bifidobacterium (Bifidobacterium longum spp. DJO10A and NCC2705). These associations reflect the close evolutionary origin of these prokaryotic species. Eight bacterial serpins (from Myxococcus xanthus, Parachlamydia, T. fusca, Crocosphaera watsonii, Bacillus licheniformis, and Arthrobacter) failed to group into any well-supported clades. Interestingly, these include the two proteins from same organism, G. violaceus, suggesting a different ancestor for these molecules.

4. Structural Features of Prokaryote Serpins Serpins are single domain proteins that generally contain three β-sheets (A, B, and C) and 9 α-helices (termed A to I). However, structural studies on viral serpins reveal that the serpin fold can be substantially modified — the x-ray crystal structure of the caspase inhibitor crmA revealed that the D-helix was entirely deleted and that part of the E-helix and A-helix was substantially shortened45 [Fig. 3(A)]. Sequence analysis and structural studies reveal that prokaryote serpins have also undergone such modifications — the X-ray crystal structure of thermopin revealed that the G-helix is deleted [Figs. 3(B) and 3(C)] and analysis of sequence alignments suggests strongly that the serpin from P. aerophilum lacks the D-helix.2,9,10 Interestingly, 31 out of 36 prokaryotic serpins contain an additional region, 19–222 residues in length, N-terminal to the serpin domain (Fig. 1). Serpins from closely related species share similarities in their N-termini; however, these regions do not share detectable similarity with any other protein. Many mammalian serpins contain N-terminal extensions that are fundamental to their function, for example, antithrombin contains a disulfide linked N-terminus that forms part of the heparin binding site.46

ch05

Jan. 10, 2007

8:57

SPI-B429

152


Q. Zhang et al.

The role of the N-terminal region in prokaryote serpins remains to be understood. However, it is suggested that this region may be important for mediating protein-protein or protein-ligand interactions. Interestingly, 13 out of 36 prokaryotic serpins also contain a short C-terminal tail that is between 1 and 4 residues longer than the analogous region in α1 -antitrypsin: one serpin (serpin3-ZP_00503588 from C. thermocellum) possesses an extended C-terminus-12 residues in length. Certain mammalian serpins also contain an extensive C-terminal region — most notably the C-terminus of α2 -antiplasmin has been demonstrated to mediate the initial interaction with plasmin.47 In contrast, a recent study on the serpin thermopin from T. fusca reveals that this region appears to be important for efficient folding of this molecule (see later).

Fig. 3. Structure of cleaved crmA.45 Elements of secondary structure are labelled. The A β-sheet is in red; B β-sheet in green; C β-sheet in yellow, and α-helices are in cyan; the RCL is in magenta. The location of hD in other serpins is shown. 3(B) and (C) structure of native (B) and cleaved (C) thermopin. Coloring is as for 3(a). The location of hG in other serpins is shown. The C-terminal region is shown in dark blue.

ch05

Jan. 10, 2007

8:57

SPI-B429



Fig. 3.

(Continued)

153

ch05

Jan. 10, 2007

8:57

SPI-B429

154


Q. Zhang et al.

5. Serpins in Thermophilic Organisms The presence of serpins in thermophilic organisms is interesting, since all inhibitory mammalian serpins characterized to date polymerize and are inactivated at elevated temperatures. Therefore, it is suggested that serpins from thermopiles must have evolved the ability to fold and undergo conformational change at elevated temperatures, while resisting heat-induced polymerization. Previous studies on mammalian serpins have revealed that mutations alter the conformational mobility often resulting in impaired inhibitory activity.4,48 Thus, the study of serpins from thermophiles may shed some light on the underlying molecular basis for metastability, conformational change, and polymerization. To date, six thermophilic prokaryotes have been shown to contain serpin homologues: T. fusca (55◦ C), S. thermophilum (60◦ C), C. thermocellum (60◦ C), T. tengcongensis (75◦ C), P. aerophilum (Tmax 104◦ C and Topt 100◦ C), and T. kodakaraensis (102◦ C). The most extensively characterized serpin of these thermophiles is thermopin from T. fusca. Studies on thermopin reveal that this molecule possesses dramatically enhanced thermal stability, in comparison to α1 -antitrypsin and is also capable of inhibiting chymotrypsin-like proteases, albeit relatively poorly. Structural and biophysical studies revealed that thermopin can adopt both a native metastable state as well as an RCL cleaved form9,10 [Figs. 3(B) and C)]. The x-ray crystal structure of thermopin revealed that this molecule contained a unique C-terminal tail, which interacts with the highly conserved residues at the top of strands s5A and s6A.10 In particular, the C-terminal residues interact with the residue equivalent to Glu 342 in α1 -antitrypsin, a crucial conserved position important for serpin stability that is mutated in the polymerogenic Z-variant of antitrypsin. Therefore, it was initially hypothesized that the additional interactions conferred by the C-terminus on the body of the serpin may be important for stability. However, mutagenesis and biophysical studies demonstrate that this is not the case — a variant of thermopin lacking the C-terminus (thermopinC ) has an identical thermal melt to wild-type material. Interestingly, the thermopinC variant unfolds via a two-state rather than a three-state pathway, supporting a role for the C-terminus in stabilizing an intermediate ensemble. It is suggested that the C-terminus of thermopin is important for

ch05

Jan. 10, 2007

8:57

SPI-B429



155

efficient folding at elevated temperatures. The x-ray crystal structure of thermopin also revealed several unusual features of this molecule in comparison with their mesophilic counterparts. In particular, thermopin contains > 1.5 times more salt bridges and hydrogen bonds, lacks the G-helix and is considerably more compact [Fig. 3(B)]. Furthermore, the absolute stabilities of the native and cleaved conformations of thermopin are both increased, providing resistance to heat induced polymerization, while still allowing the rapid conformational change essential for inhibition of proteases.9,10

6. Specificity of Prokaryotic Serpins and Possible Protease Targets Using the MEROPS resource,49 we investigated the distribution of proteases in serpin-containing organisms (Table 1). Serpins have been demonstrated to be capable of inhibiting A1-pepsin, C1-papain, C14-caspase, C25-gingipain, S1-trypsin-like, and S8-subtilisin proteinase families.50,51 The MEROPS database49 revealed that serpin-bearing prokaryotes contained four different “serpin vulnerable” proteinase families: trypsin-like, subtilisin-like, papain-like, and caspase-like sequences. Pepsin and gingipain-like sequences were not found in any of the serpincontaining species. Broadly across the data set, 18 out of the 22 serpin-bearing bacteria and 3 out of 5 serpin-bearing archae contained trypsin-like sequences in their genomes (Table 1). Most of the trypsin-like serine proteases generally closely resemble the protease Do/HtrA-like molecules. These proteases are involved in heat shock response and function as chaperones at low temperature before becoming active proteases at higher temperature. It is interesting to speculate that one possible function of prokaryote serpins could be to protect against inappropriate activity of heat shock activated protease Do/HtrA-like molecules. Subtilisin-like proteases were also widely distributed with representative sequences found in 14 of the 22 serpin containing bacteria. The majority of these molecules belong to S8A subfamily. Papain-like sequences can be identified in three archae (M. acetivorans, T. kodakaraensis, and M. mazei) and three bacteria (B. longum, C. phaeobacteroides, and B. longum). All papain-like sequences from

ch05

Jan. 10, 2007

8:57

SPI-B429

156


Q. Zhang et al.

archae belong to C1A subfamily. The C. phaeobacteroides contains a C1A subfamily-like sequence, while the other two bacterial papains belong to C1B subfamily. Finally, caspase-like sequences have only been found in G. violaceus and Nostoc sp. The primary specificity of serpins is generally governed by P1 and P1 residues; with rare exceptions, this region is located 17 amino acids from the top of the A β-sheet (Fig. 1). Strikingly, 12 of the bacterial serpins possess identical predicted P1 and P1 residues (T/S) (Table 1 and Fig. 2), suggesting that these serpins may target a similar enzyme, the identity of which remains obscure. In addition, no single protease class is common to organisms containing a serpin with a P1 -P1 sequence of T/S — for example, not all organisms that contain a serpin with a T/S P1 -P1 also contain S8-like proteases. Serpins within several well-supported bacterial serpin-clades also contain identical, or similar P1 -P1 sequences, suggesting that these proteins may perform orthologous functions (Fig. 2). None of archae serpins share common predicted P1 -P1 residues, suggesting that these serpins target different enzymes in each organism. In prokaryotes containing more than one serpin (i.e., where a clear gene duplication event has occurred), the P1 -P1 residues are generally physiochemically different, suggesting that these serpins have undergone functional diversification and evolved to target distinct proteases. To date, only two serpins (the granzyme B inhibitor PI-952 and the caspase inhibitor crmA53 ) have been shown to utilize an acidic residue at the P1 position. Interestingly, throughout the prokaryote data set, three prokaryotic serpins are predicted to possess a P1 Glu: serpin-NP_617161 from M. acetivorans ( P1 -P1 E/E), serpin from Parachlamydia sp. (P1 -P1 E/D), and serpin-NP_694094 from O. iheyensis (P1 -P1 E/S) (Table 1). Rarely, serpin-containing prokaryotes contain only a single member of serine protease family that has previously been shown to be able to be inhibited by serpins. Specifically, M. burtonii contains a single S8A-like peptidase, A. variabilis has a single S8B-like peptidase; D. ethenogenes contains an S1C-like peptidase as does Parachlamydia sp. and M. xanthus. It is not possible to unequivocally match serpin to proteinase target based on the P1 /P1 sequence alone or to identify a direct correlation between any of the current serine proteinase family classifications and serpins in the

ch05

Jan. 10, 2007

8:57

SPI-B429



157

prokaryotic genomes. The precise function of bacterial serpins in the control of proteolysis therefore remains to be understood. We suggest three possibilities that are apparent: these molecules could target a “classical” serpin-vulnerable protease, they could target a novel protease not yet identified or listed in MEROPS, or they could target an environmental protease from another organism.

7. Conclusions The discovery of serpins in bacteria and archae has important implications for the evolution of the serpin superfamily, which was previously believed to be restricted only to higher eukaryotic organisms and their viral pathogens. Recent data reveals that serpins are present in a diverse array of prokaryotes; however, the role of these molecules remains elusive. It is even difficult to ascertain whether bacterial serpins are predominantly intracellular or extracellular, since the unusual nature of the serpin-containing prokaryotes precludes accurate sorting prediction by informatic engines such as PSORT.54 Interestingly, a proteomic study revealed that the serpin from T. tengcongensis localizes to the cytoplasm,55 and thus we suggest that it is likely that at least some of the bacterial serpins perform a role in controlling endogenous intracellular proteases. Finally, it is possible that certain prokaryote serpins, and in particular the putative non-inhibitory serpin2-NP_924792 from G. violaceus is not involved in the control of proteolytic enzymes. Further experimental investigation on prokaryote serpins together with additional genomic data will no doubt shed light on their function and may also suggest novel serpin functions in higher organisms.

Acknowledgments Qingwei Zhang is a recipient of a Monash Graduate Scholarship. Lisa Cabrita and James Irving are National Health and Medical Research Council of Australia CJ Martin Fellows. Steve Bottomley is an NHMRC RD Wright Fellow and Monash University Logan Fellow. James Whisstock is a NHMRC Senior Research Fellow and Monash Unviersity Logan Fellow. We thank the NHMRC and the ARC for support.

ch05

Jan. 10, 2007

8:57

SPI-B429


158

Q. Zhang et al.

References 1. Irving JA, Pike RN, Lesk AM and Whisstock JC (2000) Phylogeny of the serpin superfamily: Implications of patterns of amino acid conservation for structure and function. Genome Res 10(12):1845–1864. 2. Irving JA, Steenbakkers PJ, Lesk AM, Op den Camp HJ, Pike RN and Whisstock JC (2002) Serpins in prokaryotes. Mol Biol Evol 19(11): 1881–1890. 3. Roberts TH, Hejgaard J, Saunders NF, Cavicchioli R and Curmi PM (2004) Serpins in unicellular Eukarya, Archaea, and Bacteria: Sequence analysis and evolution. J Mol Evol 59(4):437–447. 4. Hopkins PC, Carrell RW and Stone SR (1993) Effects of mutations in the hinge region of serpins. Biochemistry 32(30):7650–7657. 5. Huntington JA, Read RJ and Carrell RW. (2000) Structure of a serpin-protease complex shows inhibition by deformation. Nature 407(6806):923–926. 6. Lomas DA and Carrell RW (2002) Serpinopathies and the conformational dementias. Nat Rev Genet 3(10):759–768. 7. Carrell RW (2005) Cell toxicity and conformational disease. Trends Cell Biol 15(11):574–580. 8. Crawford DL (1975) Cultural, morphological, and physiological characteristics of Thermomonospora fusca (strain 190Th). Can J Microbiol 21(11):1842–1848. 9. Fulton KF, Buckle AM, Cabrita LD, Irving JA, Butcher RE, Smith I, et al. (2005) The high resolution crystal structure of a native thermostable serpin reveals the complex mechanism underpinning the stressed to relaxed transition. J Biol Chem 280(9):8435–8442. 10. Irving JA, Cabrita LD, Rossjohn J, Pike RN, Bottomley SP and Whisstock JC (2003) The 1.5 A crystal structure of a prokaryote serpin: Controlling conformational change in a heated environment. Struct (Camb) 11(4):387–397. 11. Barrett AJ and Rawlings ND (2001) Evolutionary lines of cysteine peptidases. Biol Chem 382(5):727–733. 12. Ye Q, Roh Y, Carroll SL, Blair B, Zhou J, Zhang CL, et al. (2004) Alkaline anaerobic respiration: Isolation and characterization of a novel alkaliphilic and metal-reducing bacterium. Appl Environ Microbiol 70(9):5595–5602. 13. Thiel T and Pratte B (2001) Effect on heterocyst differentiation of nitrogen fixation in vegetative cells of the cyanobacterium Anabaena variabilis ATCC 29413. J Bacteriol 183(1):280–286. 14. Haavik HI (1975) Bacitracin production by the neotype; bacillus licheniformis ATCC 14580. Acta Pathol Microbiol Scand Suppl 83(6):534–540.

ch05

Jan. 10, 2007

8:57

SPI-B429



159

15. Rey MW, Ramaiya P, Nelson BA, Brody-Karpin SD, Zaretsky EJ, Tang M, et al. (2004) Complete genome sequence of the industrial bacterium Bacillus licheniformis and comparisons with closely related Bacillus species. Genome Biol 5(10):R77. 16. Klijn A, Mercenier A and Arigoni F (2005) Lessons from the genomes of bifidobacteria. FEMS Microbiol Rev 29(3):491–509. 17. Schell MA, Karmirantzou M, Snel B,Vilanova D, Berger B, Pessi G, et al. (2002) The genome sequence of Bifidobacterium longum reflects its adaptation to the human gastrointestinal tract. Proc Natl Acad Sci USA 99(22):14422–14427. 18. Verte F, Kostanjevecki V, De Smet L, Meyer TE, Cusanovich MA and Van Beeumen JJ (2002) Identification of a thiosulfate utilization gene cluster from the green phototrophic bacterium Chlorobium limicola. Biochemistry 41(9):2932–2945. 19. Herrero AA and Gomez RF. (1980) Development of ethanol tolerance in Clostridium thermocellum: Effect of growth temperature. Appl Environ Microbiol 40(3):571–577. 20. Damborsky J (1994) Tetrachloroethene-dehalogenating bacteria. Folia Microbiol (Praha) 44(3):247–262. 21. Seshadri R, Adrian L, Fouts DE, Eisen JA, Phillippy AM, Methe BA, et al. (2005) Genome sequence of the PCE-dechlorinating bacterium Dehalococcoides ethenogenes. Science 307(5706):105–108. 22. Adrian L, Szewzyk U, Wecke J and Gorisch H (2000) Bacterial dehalorespiration with chlorinated benzenes. Nature 408(6812):580–583. 23. Niggemyer A, Spring S, Stackebrandt E and Rosenzweig RF (2001) Isolation and characterization of a novel As(V)-reducing bacterium: Implications for arsenic mobilization and the genus Desulfitobacterium. Appl Environ Microbiol 67(12):5568–5580. 24. Mimuro M, Ookubo T, Takahashi D, Sakawa T, Akimoto S, Yamazaki I, et al. (2002) Unique fluorescence properties of a cyanobacterium Gloeobacter violaceus PCC 7421: Reasons for absence of the long-wavelength PSI Chl a fluorescence at −196 degrees C. Plant Cell Physiol 43(6):587–594. 25. Nakamura Y, Kaneko T, Sato S, Mimuro M, Miyashita H, Tsuchiya T, et al. Complete genome structure of Gloeobacter violaceus PCC 7421, a cyanobacterium that lacks thylakoids (supplement). DNA Res 10(4):181–201. 26. Dworkin M (1996) Recent advances in the social and developmental biology of the myxobacteria. Microbiol Rev 60(1):70–102. 27. Gehrig H, Schussler A and Kluge M (1996) Geosiphon pyriforme, a fungus forming endocytobiosis with Nostoc (cyanobacteria), is an ancestral member of the Glomales: Evidence by SSU rRNA analysis. J Mol Evol 43(1):71–81.

ch05

Jan. 10, 2007

8:57

SPI-B429

160


Q. Zhang et al.

28. Fleming H and Haselkorn R (1974) The program of protein synthesis during heterocyst differentiation in nitrogen-fixing blue-green algae. Cell 3(2): 169–170. 29. Kaneko T, Nakamura Y, Wolk CP, Kuritz T, Sasamoto S, Watanabe A, et al. (2001) Complete genomic sequence of the filamentous nitrogen-fixing cyanobacterium Anabaena sp. strain PCC 7120. DNA Res 8(5):205–213; 227–253. 30. Takami H, TakakiY and Uchiyama I (2002) Genome sequence of Oceanobacillus iheyensis isolated from the Iheya Ridge and its unexpected adaptive capabilities to extreme environments. Nucleic Acids Res 30(18):3927–3935. 31. Horn M, Collingro A, Schmitz-Esser S, Beier CL, Purkhold U, Fartmann B, et al. (2004) Illuminating the evolutionary history of chlamydiae. Science 304(5671):728–730. 32. Ueda K, Yamashita A, Ishikawa J, Shimada M, Watsuji TO, Morimura K, et al. (2004) Genome sequence of Symbiobacterium thermophilum, an uncultivable bacterium that depends on microbial commensalism. Nucleic Acids Res 32(16):4937–4944. 33. XueY, XuY, LiuY, MaY and Zhou P (2001) Thermoanaerobacter tengcongensis sp. nov., a novel anaerobic, saccharolytic, thermophilic bacterium isolated from a hot spring in Tengcong, China. Int J Syst Evol Microbiol 51(Pt 4): 1335–1341. 34. Bao Q, Tian Y, Li W, Xu Z, Xuan Z, Hu S, et al. (2002) A complete sequence of the T. tengcongensis genome. Genome Res 12(5):689–700. 35. Franzmann PD, Hopfl P, Weiss N and Tindall BJ (1991) Psychrotrophic, lactic acid-producing bacteria from anoxic waters in Ace Lake, Antarctica; Carnobacterium funditum sp. nov. and Carnobacterium alterfunditum sp. nov. Arch Microbiol 156(4):255–262. 36. Galagan JE, Nusbaum C, Roy A, Endrizzi MG, Macdonald P, FitzHugh W, et al. (2002) The genome of M. acetivorans reveals extensive metabolic and physiological diversity. Genome Res 12(4):532–542. 37. Deppenmeier U, Johann A, Hartsch T, Merkl R, Schmitz RA, MartinezArias R, et al. (2002) The genome of Methanosarcina mazei: Evidence for lateral gene transfer between bacteria and archaea. J Mol Microbiol Biotechnol 4(4):453–461. 38. Fitz-Gibbon ST, Ladner H, Kim UJ, Stetter KO, Simon MI and Miller JH (2002) Genome sequence of the hyperthermophilic crenarchaeon Pyrobaculum aerophilum. Proc Natl Acad Sci USA 99(2):984–989. 39. Morikawa M, Izawa Y, Rashid N, Hoaki T and Imanaka T (1994) Purification and characterization of a thermostable thiol protease from a newly

ch05

Jan. 10, 2007

8:57

SPI-B429



40.

41.

42.

43. 44.

45.

46.

47.

48. 49. 50.

51.

161

isolated hyperthermophilic Pyrococcus sp. Appl Environ Microbiol 60(12): 4559–4566. Fukui T, Atomi H, Kanai T, Matsumi R, Fujiwara S and Imanaka T (2005) Complete genome sequence of the hyperthermophilic archaeon Thermococcus kodakaraensis KOD1 and comparison with Pyrococcus genomes. Genome Res 15(3):352–363. Linding R, Russell RB, Neduva V and Gibson TJ (2003) GlobPlot: Exploring protein sequences for globularity and disorder. Nucleic Acids Res 31(13): 3701–3708. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, et al. (1997) Gapped BLAST and PSI-BLAST: A new generation of protein database search programs. Nucleic Acids Res 25(17):3389–3402. Higgins DG, Thompson JD and Gibson TJ (1996) Using CLUSTAL for multiple sequence alignments. Methods Enzymol 266:383–402. Kumar S, Tamura K and Nei M (2004) MEGA3: Integrated software for molecular evolutionary genetics analysis and sequence alignment. Brief Bioinform 5(2):150–163. Simonovic M, Gettins PGW and Volz K (2000) Crystal structure of viral serpin crmA provides insights into its mechanism of cysteine proteinase inhibition. Protein Sci 9(8):1423–1427. Fitton HL, Skinner R, Dafforn TR, Jin L and Pike RN (1998) The N-terminal segment of antithrombin acts as a steric gate for the binding of heparin. Protein Sci 7(3):782–788. Zhai P, Wakeham N, Loy JA and Zhang XC (2003) Functional roles of streptokinase C-terminal flexible peptide in active site formation and substrate recognition in plasminogen activation. Biochemistry 42(1):114–120. Stein PE and Carrell RW (1995) What do dysfunctional serpins tell us about molecular mobility and disease? Nat Struct Biol 2(2):96–113. Rawlings ND, O’Brien E and Barrett AJ (2002) MEROPS: The protease database. Nucleic Acids Res 30(1):343–346. Komiyama T, Ray CA, Pickup DJ, Howard AD, Thornberry NA, Peterson EP, et al. (1994) Inhibition of interleukin-1 beta converting enzyme by the cowpox virus serpin CrmA. An example of cross-class inhibition. J Biol Chem 269(30):19331–19337. Schick C, Pemberton PA, Shi GP, Kamachi Y, Cataltepe S, Bartuski AJ, et al. (1998) Cross-class inhibition of the cysteine proteinases cathepsins K, L, and S by the serpin squamous cell carcinoma antigen 1: A kinetic analysis. Biochemistry 37(15):5258–5266.

ch05

Jan. 10, 2007

8:57

SPI-B429

162


Q. Zhang et al.

52. Sun J, Whisstock JC, Harriott P, Walker B, Novak A, Thompson PE, et al. (2001) Importance of the P4 residue in human granzyme B inhibitors and substrates revealed by scanning mutagenesis of the proteinase inhibitor 9 reactive center loop. J Biol Chem 276(18):15177–151784. 53. Ray CA, Black RA, Kronheim SR, Greenstreet TA, Sleath PR, Salvesen GS, et al. (1992) Viral inhibition of inflammation: Cowpox virus encodes an inhibitor of the interleukin-1 beta converting enzyme. Cell 69(4):597–604. 54. Gardy JL, Laird MR, Chen F, Rey S, Walsh CJ, Ester M, et al. (2005) PSORTb v.2.0: Expanded prediction of bacterial protein subcellular localization and insights gained from comparative proteome analysis. Bioinformatics 21(5):617–623. 55. Wang J, Xue Y, Feng X, Li X, Wang H, Li W, et al. (2004) An analysis of the proteomic profile for Thermoanaerobacter tengcongensis under optimal culture conditions. Proteomics 4(1):136–150.

ch05

Jan. 10, 2007

8:57

SPI-B429


6 New Lessons from Poxvirus Serpins Peter C. Turner, Priscilla F. McAuliffe, Amy L. MacNeill, and Richard W. Moyer

1. Introduction Poxviruses are a large, diverse and successful group of DNA viruses that replicate in the cytoplasm,1 and encode an array of immune modulators. The ability of poxviruses to cause disease depends on subversion of the hostimmune response by proteins that interfere with inflammation and apoptosis, and disrupt signaling by cytokines and chemokines.2,3 Despite the recent discovery of serpins in prokaryotes and archaebacteria (see Chap. 5), poxviruses remain the only virus group to express serpins that function as proteinase inhibitors. Murine gammaherpesvirus 68 contains ORF M1 with sequence homology to the poxvirus serpin SPI-14 but the sequence indicates that the M1 protein is unlikely to function as a proteinase inhibitor. Presumably, the poxvirus serpins were acquired from the host cell at some time during evolution, perhaps from a cDNA copy of an mRNA transcript, since poxviruses do not enter the nucleus and their genomes contain no introns. However, databases of vertebrate sequence data do not currently contain host serpins that are closely related to poxvirus serpins, either because the databases are incomplete or because the host and poxvirus serpins have diverged substantially since the time of acquisition by the virus. The Poxviridae family is divided into two subfamilies, the Chordopoxvirinae of vertebrates, and the Entomopoxvirinae of invertebrates. Not all poxvirus genera encode serpins. At present, the poxvirus genera 163

ch06

Jan. 10, 2007

8:57

SPI-B429

164


P. C. Turner et al.

encoding serpins include six of the eight Chordopoxvirus genera, namely, the orthopoxviruses, leporipoxviruses, avipoxviruses, capripoxviruses, suipoxvirus, and yatapoxviruses. Evidence for some of these serpins is inferred purely from sequence data, and in many cases, the effects of knockout mutations on viruses or the properties of purified proteins have not been determined. The molluscipoxvirus genus, which includes the human virus molluscum contagiosum, and the parapoxviruses, including orf virus, do not encode serpins; none have been found in the complete genomic sequences of molluscum contagiosum virus or orf virus. Similarly, there are no serpins in the entomopoxvirus (EPV) subfamily, comprising poxviruses of insect hosts. Two insect poxvirus genomes, those of Amsacta moorei EPV and Melanoplus sanguinipes EPV, have been completely sequenced but are found to lack serpins. The poxvirus serpins that have been functionally characterized are listed in Table 1. Orthopoxviruses encode three serpins, SPI-1, SPI-2/crmA, and SPI-3, which function as proteinase inhibitors. The leporipoxvirus myxoma virus encodes two serpins, SERP1 and SERP2, plus the non-inhibitory serpin SERP3.5 Rather than grouping poxvirus serpins by the virus from which they originate, we have found a classification based on the P1 residue within the reactive center loop (RCL) to be more meaningful. Using this system, the poxvirus serpins fall into three main groups, with a P1 residue of Phe, Asp, or Arg. The first category with P1 = Phe is only found in orthopoxviruses, the genus that includes the type species vaccinia virus (VV); rabbitpox virus (RPV); cowpox virus (CPV); ectromelia virus (mousepox); variola virus, the causative agent of smallpox; monkeypox; and camelpox. When representatives of all known functional poxvirus serpins are grouped in a guide tree (Fig. 1), the P1 = Phe serpins form a well-defined cluster indicated by the blue color in Fig. 1. Several serpins have aspartic acid at the P1 position, but the tree in Fig. 1 suggests that some P1 Asp serpins are quite distinct from others and probably derived from different ancestral host serpins. The orthopoxvirus P1 = Asp serpins form a well-defined cluster of orthologous genes which we have labeled as SPI-2/crmA. The P1 = Asp serpins from other genera form a much more heterogeneous group that includes myxoma virus SERP2, and genes from capripoxviruses (goatpox, lumpy skin disease, and sheeppox viruses), the suipoxvirus genus (swinepox virus), and yatapoxviruses (Yaba monkey tumor virus and Yaba-like disease virus),

ch06

Jan. 10, 2007 8:57 SPI-B429

serpin

P1 residue

N-linked glycosylation

Location

Orthopoxvirus

SPI-1

Phe (F)

No

Cytoplasmic

Orthopoxvirus

SPI-2/crmA

Asp (D)

No

Cytoplasmic

Leporipoxvirus

SERP2

Asp (D)

No

Cytoplasmic

Orthopoxvirus

SPI-3

Arg (R)

Yes

Cell surface, EEV

Leporipoxvirus

SERP1

Arg (R)

Yes

Secreted

Biochemical function/comments Forms complex with cathepsin G; determines host range Inhibits caspases and gzB; blocks apoptosis and inflammation Inhibits caspases; blocks apoptosis and inflammation Inhibits plasmin, uPA, tPA; blocks cell-cell fusion Inhibits plasmin, uPA, tPA; required for virulence

165


Virus group

Poxvirus serpins.

New Lessons from Poxvirus Serpins

Table 1

ch06

Jan. 10, 2007

8:57

SPI-B429

166


P. C. Turner et al.

in addition to avipoxviruses (canarypox and fowlpox viruses). The third group, the P1 = Arg serpins, again fall into distinct groups, and are clearly not all orthologues, but instead appear to have been acquired from distinct hosts during evolution. The orthopoxvirus P1 = Arg serpins are a tight cluster of closely related SPI-3 orthologues. The remaining P1 Arg serpins derived from other poxvirus genera include myxoma virus SERP1, and a yatapoxvirus serpin. The avipoxviruses include a number of serpins whose P1 residues cannot be easily identified based on alignments with other poxvirus serpins, and these serpins were therefore not included in Fig. 1. An alignment of key serpins is shown in Fig. 2. The following sections

Fig. 1. Phylogenetic grouping of poxvirus serpins. Protein sequences of poxvirus serpins were aligned using AlignX, a component of Vector NTI Suite 9 (Invitrogen). The guide tree was then built using the Neighbor Joining method of Saitou and Nei (1987). The boxes indicate clusters of serpins with P1 = Phe, P1 = Asp, and P1 = Arg. For clarity only one strain of each virus was chosen, for example, vaccinia WR serpins were included, and the many other vaccinia strains for which serpin sequences are available were excluded. Names of orthopoxviruses are italicized. The number after each virus name is the designation of the open reading frame in the genome.

ch06

Jan. 10, 2007 8:57 SPI-B429

167



Fig. 2. Alignment of representative poxvirus serpins and human α1 -proteinase inhibitor (antitrypsin). The portion of the serpins encompassing the RCL is shown. Black letters on white show non-similar residues. Red lettering on a yellow background indicates complete conservation of a residue for the aligned serpins; black on cyan shows a consensus residue occurring at greater than 50% at a given position; and black on green indicates consensus residue(s) derived from a block of similar residues at a given position. Note that residues flanking the RCL are highly conserved but the sequences within the RCL are more variable. The position of the P14 Thr residue is marked. The P1 residue for each serpin is underlined, and the position indicated above by either of the two arrows. For the P1 = Asp serpins, the P1 residue is displaced one position to the left (smaller arrow).

ch06

Jan. 10, 2007

8:57

SPI-B429

168


P. C. Turner et al.

describe the biochemical properties of individual poxvirus serpins, and the roles of serpins in poxvirus infections. Before discussing the properties of individual serpins, we will consider animal models that are appropriate for evaluating the virulence of mutant poxviruses.

2. Selection of Animal Models To date, none of the poxvirus serpins studied are generally required for growth of the virus in cell culture. However, the in vitro biochemical activities discussed below, along with the evolutionary conservation of serpins, suggest that the poxviral serpin proteins do have a function in vivo.6 The comparison of poxvirus natural infections with the results of infecting laboratory animals with knock-out or genetically altered viruses is not always straightforward.6 Frequently, interpretation of the in vivo result of infections with laboratory-generated recombinant viruses is difficult, because during a poxvirus infection, there is functional redundancy in the expressed viral proteins that can disrupt or modulate multiple pathways in a susceptible host.2,7 Selecting a suitable laboratory animal as a host and choosing an appropriate inoculation route that resembles the process of natural infection are also very important.6 For the Orthopoxviruses, including vaccinia and CPV, which encode SPI-1, SPI-2, and SPI-3, mouse models have most commonly been employed, because rodents are the natural reservoir of some of these poxvirus strains, including CPV.8 Furthermore, the use of laboratory mice is advantageous because they are easier to handle and the cost is less than larger animals, and many well-developed immune markers as well as transgenic and knock-out strains are readily available for mice.6 However, other animals, particularly the rabbit, may provide a model of disease more similar to human smallpox, including transmission from animal to animal (Rice, Adams, and Moyer, unpublished results). Inoculation routes evaluated in the laboratory setting include aerosol, intranasal, intratracheal, intracerebral, intraperitoneal, and intradermal infection. Intradermal models include inoculation of both the footpad9 and the ear pinna.10 Intradermal inoculation is thought to correlate best with the natural route of infection for most Orthopoxviruses. One notable exception to this is the smallpox infection in humans, which is transmitted by the

ch06

Jan. 10, 2007

8:57

SPI-B429



169

respiratory route,11 and can be modeled by aerosol, intranasal or intratracheal administration of Orthopoxviruses to laboratory animals.12 The New Zealand White (NZW) laboratory rabbit is most commonly used to evaluate the Leporipoxvirus myxoma,13 which encodes the serpins SERP1, SERP2 and SERP3. Although rabbits are the natural host of Myxoma virus, rabbits have the disadvantage of being more costly than mice, with less well-developed genetic analyses, and fewer reagents and immune markers specific for rabbits.6 In the European rabbit Oryctolagus cuniculus, myxoma causes myxomatosis, characterized by large myxomatous lesions, fever, severe respiratory distress, and death. It was thought that death results from profound immune compromise that leads to overwhelming Gram-negative infection.3,13 However, specific-pathogenfree rabbits infected with Myxoma virus have a typical course of disease and die within 14 days of infection, but bacterial sepsis is not observed (MacNeill, Turner, and Moyer, unpublished data). Lesion formation, fever and severe respiratory distress occur, and cellular damage due to viral infection appears to be the primary cause of death. This interpretation has also been reached with variola virus infection of humans (smallpox). Original descriptions of fatal cases of smallpox attributed death to bacterial sepsis or immune complex deposition, but a retrospective review of smallpox cases concluded that death was due to the cytopathic effects of variola virus.14 The effects of serpin gene deletions and alterations have been measurable both at the primary inoculation site and at subsequent spread of myxoma virus infection to secondary organs.5,15−19 Compared with the European rabbit in which myxoma was deliberately introduced for population control, the normal host for myxoma is the American brush rabbit, Sylvilagus, which develops only a localized benign fibroma and mild disease.20 Myxoma naturally infects Sylvilagus rabbits through the skin via biting insect vectors, and in the laboratory, intradermal injections are the most commonly used route of inoculation.6 As alluded above, the rabbit is also gaining increased attention as a model for orthopoxvirus (variola virus) infection of humans. Rabbits are extremely sensitive to infection by vaccinia or rabbitpox virus (RPV), and succumb to pulmonary disease typical of smallpox. In addition, virus can be transmitted from infected index rabbits to sentinel animals.

ch06

Jan. 10, 2007

8:57

SPI-B429

170


P. C. Turner et al.

3. The Properties of Individual Poxvirus Serpins The following sections deal with the function of specific poxvirus serpins, as deduced from studies of infected tissue culture cells, from biochemical work with purified proteins, and from infected animals. 3.1. SPI-1 with P1 = Phe The orthopoxvirus SPI-1 gene was found to play a role in host range. Both spontaneous and engineered deletion mutants of Rabbitpox virus (RPV) lacking the SPI-1 gene were unable to grow on pig kidney PK-15 epithelial cells or human lung carcinoma A549 cells,21 although growth of RPV SPI-1 deletion mutants on permissive cell lines such as CV-1 (African green monkey kidney fibroblasts) was completely normal. Some strains of vaccinia virus (including modified vaccinia virus Ankara) lack the SPI-1 gene. The introduction of the SPI-1 gene into modified vaccinia virus Ankara (MVA) was associated with the gain of the ability to propagate in A549 cells.22 The phenotype of RPV SPI-1 mutants on non-permissive A549 cells was initially suggestive of apoptosis, as nuclear condensation was observed.23 However, later studies failed to demonstrate that caspases were induced on the infection of non-permissive cells with RPV SPI-1 mutants, and no cleavage of poly(ADP-ribose)polymerase (PARP) or lamin A was observed.24 A SPI-1 deletion mutant of vaccinia virus strain WR (VVSPI-1) was unable to plaque on A549 cells or on primary human keratinocytes.25 Chromatin condensation was seen in some A549 cells infected with VVSPI-1, but other apoptotic correlates such as caspase-3 activation, PARP cleavage, and DNA fragmentation were not observed.25 In these respects, the RPV SPI-1 deletion mutant and VVSPI-1 resembled one another. The lack of association between SPI-1 and apoptosis is consistent with the identification of Phe as the P1 residue (Fig. 2). To date, all poxvirus serpins with anti-apoptotic activity have P1= Asp. In other aspects, deletion of SPI-1 from vaccinia virus (VV) and from RPV had different effects. Infection of A549 cells with VVSPI-1 resulted in low levels of intermediate and late mRNAs, and reduced levels of late proteins and virus particles,25 ; features that were not seen in corresponding RPVSPI-1 infections.23

ch06

Jan. 10, 2007

8:57

SPI-B429



171

Radiolabeled RPV SPI-1 protein synthesized by coupled in vitro transcription/translation formed an SDS-stable complex with cathepsin G, a serine protease with chymotrypsin-like activity.24 Cathepsin G is a component of neutrophils, monocytes and mast cells, and is involved in neutrophil chemotaxis.26 The observation of SPI-1/cathepsin G complexes strongly suggests that SPI-1 inhibits cathepsin G, and enzymatic studies with purified SPI-1 protein confirm this (Turner and Moyer, unpublished data). Kinetic constants for SPI-1 against cathepsin G have not been determined, and the significance of cathepsin G inhibition during natural infections has not been established. The fact that site-directed mutations in the RPV SPI1 RCL confer a restricted host range phenotype during infection in cell culture24 suggests that proteinase inhibition by SPI-1 is important for the full host range of RPV. On the other hand, mutation of the P1 and P1 residues of SPI-1 in VV strain WR destabilized the SPI-1 protein, thus preventing any conclusions to be drawn on how SPI-1 functions in host range.25 Recent genetic experiments designed to generate mutations that suppress the restricted host range phenotype of RPV SPI-1 mutants have been successful, and generated both intragenic and extragenic suppressors. The fact that extragenic suppressors map to the DNA polymerase E9L gene and to D5R, an NTPase involved in DNA replication, was unexpected.27 Interactions between SPI-1 and E9L or D5R do not appear to be allelespecific, indicating that direct protein-protein interactions between SPI-1 and E9L or D5R are unlikely to occur. The mechanism by which mutations in E9L and D5R suppress the host range phenotype of SPI-1 mutants remains to be understood. 3.2. The role of P1=Phe serpins in infected animal models After intranasal administration to BALB/c mice of cowpox virus or rabbitpox virus deletion mutants of SPI-1, there were no differences in clinical disease manifestation when compared with wild-type(wt) viruses.28 Similar results were obtained with vaccinia virus, in which intranasal inoculation of BALB/c mice with an SPI-1 deletion mutant caused disease that was indistinguishable from that of wt virus and also that of a revertant virus in which the deleted SPI-1 gene had been restored.29 However, another isolate

ch06

Jan. 10, 2007

8:57

SPI-B429

172


P. C. Turner et al.

of VV deleted for SPI-1 was attenuated after intranasal inoculation of the 8- to 9-week-old female BALB/c × C57BL/6 mice.30 Despite the varying effect of SPI-1 knock-outs in these systems, the conservation of the SPI-1 gene in all orthopoxviruses suggests that this serpin has a role in natural infections. 3.3. SPI-2/crmA and SERP2 with P1 = Asp We will consider the orthopoxvirus SPI-2 and the leporipoxvirus SERP2 proteins together, as the properties of these serpins are quite similar, despite the fact that they are only 33% identical in overall sequence. The SPI-2 gene in CPV is known as crmA (cytokine response modifier A), and was the first poxvirus serpin to be discovered.31 Initially, crmA was identified by the phenotype conferred by the deletion of this gene. wt CPV causes red, hemorrhagic lesions (pocks) on the chorioallantoic membranes of embryonated chicken eggs, but deletion mutants lacking crmA produce white, inflammatory pocks.32 Purified crmA protein was later found33 to be an excellent inhibitor of human caspase-1 (ICE, interleukin-1β converting enzyme), which processes the precursors of interleukin-1β (IL-1β) and IL-18 to the mature pro-inflammatory forms by proteolytic cleavage. CrmA theoretically acts to control inflammation by this mechanism. Subsequently, crmA was found to have potent antiapoptosis activity by virtue of inhibition of caspase-8, an initiator caspase that regulates the extrinsic pathway.34,35 The biochemical properties of crmA as a caspase inhibitor are discussed in Chap. 13. Inhibition of caspases by crmA is an example of cross-class inhibition, where a serine protease inhibitor has evolved to inhibit proteases that use cysteine as the nucleophile in catalysis. CrmA is unlike the baculovirus caspase inhibitor p35 in that crmA has no activity against the executioner (terminal) caspases, namely caspase-3, -6, and -7.34,35 CrmA inhibits the serine proteinase granzyme B, a component of cytotoxic lymphocytes that has Asp-ase activity and can activate executioner caspases.36 In contrast, p35 is not active against granzyme B.37 Deletion of crmA from CPV does not have a profound effect on the growth of virus in most tissue culture cell lines. However, CPVcrmA does trigger apoptosis on the infection of LLC-PK1 pig kidney epithelial cells.38

ch06

Jan. 10, 2007

8:57

SPI-B429



173

Under these conditions, the yield of infectious virus was not reduced, presumably because apoptosis occurred sufficiently late after the completion of virus replication. The myxoma virus SERP2 protein has broadly similar function to crmA, with activity against caspase-139 and granzyme B.40 The inhibition constants suggested that SERP2 was a less effective inhibitor of human caspase-1 and human granzyme B than crmA.40 Myxoma virus (MYX) deleted for SERP2 induces apoptosis on the infection of RK-13 rabbit kidney epithelial cells, in contrast with wtMYX.41 This indicates that SERP2 does have anti-apoptotic activity in rabbit cells. A role for viral genes that are non-essential for growth is frequently revealed by examining behavior of the virus in animal models. We will consider the effect of mutations or substitutions affecting P1 = Asp serpins in three models: embryonated chicken eggs, mice, and rabbits.

3.4. The role of P1=Asp serpins in the chicken chorioallantoic membrane A CPV-crmA-D303A mutant engineered to change the P1 residue of crmA from Asp to Ala behaved essentially as a CPV crmA deletion mutant (CPVcrmA) when inoculated onto the chicken chorioallantoic membrane (CAM). White pocks were produced, the virus yield was reduced relative to wild-type infection, and infected CAMs exhibited increased apoptosis.18 All of the biological properties of crmA therefore appear to be dependent on the proteinase inhibition conferred by the P1 Asp residue. A recent biochemical study found that a crmA P1 Arg mutant was unable to inhibit caspase-1 or granzyme B,42 consistent with the pivotal role of the Asp normally found at this position. Despite the apparently similar inhibitory profiles of crmA and SERP2, SERP2 was unable to substitute fully for crmA in the CPV-infected CAM model. CPV with SERP2 under the control of the crmA promoter in place of the natural crmA gene produced white inflammatory pocks.18 SERP2 clearly could not inhibit inflammation under these circumstances, unlike crmA. On the other hand, apoptosis in CAMs was inhibited following infection with the CPV recombinant expressing SERP2 instead of crmA.18 The

ch06

Jan. 10, 2007

8:57

SPI-B429

174


P. C. Turner et al.

ability of SERP2 to substitute for crmA in controlling apoptosis but not in controlling inflammation may reflect the fact that crmA is an excellent caspase-1 inhibitor, but SERP2 is less effective against caspase-1. In the chicken CAM, the reduction in virus yield following the deletion of crmA from CPV was originally interpreted as reflecting the infiltration of inflammatory cells into the pock. However, the CPV recombinant with SERP2 in place of crmA had a restored virus titer, with evidence of inflammation but no indicators of apoptosis.18 Therefore, control of apoptosis rather than inflammation appears to be important for virus yield in this system.

3.5. The role of P1=Asp serpins in infected mice Conflicting results have been reported in mouse models after intranasal inoculation with vaccinia virus lacking the SPI-2 gene (the orthologue of the CPV crmA gene). Deletion of SPI-2 from VV did not cause attenuation of the virus during intranasal inoculation of 5 to 6-week-old female BALB/c mice.29,42 However, 8 to 9-week-old female BALB/c × C57BL/6 mice infected with VV deleted for SPI-2 lost significantly less weight than animals infected with wtVV.30 Despite the ability of SPI-2 to inhibit the interleukin-1β converting enzyme (ICE), deletion of this gene did not prevent fever, a systemic effect mediated by IL-1β, in infected mice.43 Instead, fever appeared to be controlled by the vaccinia IL-1β receptor, which when deleted, did increase the febrile response.43 Deletion of the viral IL-1β receptor and SPI-2 together had no additional effect above that of deleting the viral IL-1β receptor alone.43 Mice infected intranasally with a cowpox mutant containing a deletion of the crmA gene had attenuated disease, compared with wild-type infection.28 Mice infected with the crmA deletion mutant of cowpox virus showed reduced inflammation and decreased alveolar edema and hemorrhage.28 A similar analysis of RPV deleted for SPI-2, compared with wtRPV, was flawed in that the deletion mutant contained an additional attenuating mutation elsewhere in the genome.28 Differences in pulmonary pathology are seen with wild-type cowpox and rabbitpox infection.28 Mice infected with wild-type cowpox develop epithelial cell hyperplasia along the bronchi and bronchioles, with a mixed inflammatory cell infiltrate,

ch06

Jan. 10, 2007

8:57

SPI-B429



175

along with severe hemorrhage at these levels and also involving the alveoli, whereas wild-type rabbitpox infection, which also induced mixed inflammatory cell infiltrates, causes a remarkable absence of pulmonary hemorrhage, and instead produces multifocal areas of coagulation necrosis of epithelial cells of the bronchioles.28 Intratracheal and intradermal (ear pinnae) models of CPV infection have also shown attenuation after the deletion of SPI-2/crmA (CPVcrmA). Intratracheal inoculation of 104 plaque forming units (pfu) of CPV caused severe respiratory distress necessitating euthanasia in > 60% of infected mice, whereas 104 pfu of CPVcrmA caused death in only 20% of infected mice. After intradermal inoculation of the ear pinna, C57BL/6 mice infected with 105 pfu CPVcrmA had less severe inflammation and necrosis at the inoculation site, compared with mice infected with wtCPV.At Day 16 post inoculation, high virus titers were isolated from ear pinnae of mice infected with wtCPV, but no virus could be isolated from the ears of mice infected with CPVcrmA (MacNeill, Moldawer and Moyer, unpublished data). Intradermal inoculation of VV into the ear pinnae of BALB/c mice also revealed differences in pathogenesis caused by deletion of SPI-2/crmA. However, in this case, VVSPI-2 caused increased swelling and inflammation compared with wtVV.44 The divergent effect of serpin deletion from different poxvirus species in the same genus is intriguing. However, it is important to be aware of both differences in the animal models and the viruses themselves used in these studies. The genome of CPV is considerably larger than that in either VV or RPV, allowing for the possibility of genes that are functionally redundant in CPV. BALB/c mice have a weak cell-mediated immune response and are deficient in the complement component C3, compared with that in C57BL/6 mice.45,46 Since cell-mediated immunity and the complement cascade are important during the immune response to poxvirus infection, the effects of serpin deletion during poxvirus infection of BALB/c mice may be different from the effects in C57BL/6 mice, simply because C57BL/6 mice mount a more effective immune response to the virus. 3.6. The role of P1=Asp serpins in infected rabbits Improved survival concomitant with highly attenuated disease also developed in European rabbits after the deletion of SERP2 when compared with

ch06

Jan. 10, 2007

8:57

SPI-B429

176


P. C. Turner et al.

the uniformly lethal infectious myxomatosis characteristic of wild-type myxoma virus infection.16 Histological evaluation suggests a more profound inflammatory response at the vascular level in rabbits treated with the mutant myxoma from which SERP2 was deleted, as opposed to the suppression of inflammation and the prevention of apoptosis seen after wild-type infection.16 Experiments involving substitution of the P1 amino acid of SERP2 were designed to ask if SERP2 functioned solely as a proteinase inhibitor during infection with MYX or if additional non-inhibitory protein interactions were involved. The P1 Asp at amino acid 294 of SERP2 was changed to Ala to generate a SERP2 D294A mutant that did not inhibit granzyme B or caspase-1, -8, or -10. MYX expressing SERP2 D294A in place of SERP2 caused signs of disease and lesions that were identical to MYX deleted for SERP2,41 suggesting that SERP2 functions solely as a proteinase inhibitor during intradermal infection of NZW rabbits. A second P1 site mutation was made in SERP2 to alter the inhibitory specificity of SERP2. The Asp at the P1 position was replaced with glutamate to produce SERP2 D294E. This mutation was based on human serine proteinase inhibitor-9 (PI-9), which has glutamate at the P1 site and inhibits granzyme B, but it is not an effective inhibitor of caspase-1, -8 or -10.47 We expected SERP2 D294E to have a similar inhibitory profile to PI-9. However, although SERP2-D294E was not able to inhibit caspase-8 or -10, SERP2 D294E retained ability to inhibit caspase-1. Surprisingly, the D294E substitution resulted in decreased inhibition of human granzyme B. MYX expressing SERP2 D294E in place of SERP2 was clearly unable to prevent caspase-mediated apoptosis in rabbit kidney epithelial cells, consistent with the defect in caspase-8 and -10 inhibition. However, MYX SERP2 D294E was fully virulent in rabbits,41 suggesting that the control of apoptosis is not required for virulence, and implying that the control of inflammation via caspase-1 is more important in the immunocompetent animal model. Caspases-8 and -10 are apparently not the primary targets of SERP2 during infection of NZW rabbits. A MYXSERP2::crmA recombinant with crmA expressed from the SERP2 promoter in place of SERP2 was tested in tissue culture and in the rabbit model. The MYXSERP2::crmA was able to control apoptosis in infected rabbit kidney RK-13 cells, indicating that crmA

ch06

Jan. 10, 2007

8:57

SPI-B429



177

was able to inhibit rabbit caspases involved in apoptosis. However, the behavior of MYXSERP2::crmA in rabbits infected intradermally did not completely resemble either wtMYX or the MYXSERP2 mutant. MYXSERP2::crmA did not give myxomatous primary lesions in infected animals; instead, it gave flat lesions somewhat similar to those produced by the MYXSERP2 mutant. However, this virus caused death in 7 of the 10 infected rabbits, and was clearly not as attenuated as MYXSERP2.41 The results show that lethality can still occur in the absence of a myxomatous primary lesion, and that crmA was only partially effective in substituting for SERP2. There are at least two reasons that poxvirus serpins with similar proteinase inhibitory profiles would fail to fully substitute for each other during infection of animals. One possibility involves non-inhibitory protein interactions with the serpin outside of the RCL. Poxvirus serpins that appear to have evolved independently of each other (like crmA and SERP2, Fig. 1) are quite different at the amino acid level. Interactions that may occur with the backbone of one serpin are unlikely to occur with the backbone of the substituted serpin. The fact that the SERP2 D294A mutant was as attenuated as MYXSERP2 argues against a non-inhibitory role of SERP2 in pathogenesis. Another possibility is that the poxvirus serpins have evolved to be extremely effective inhibitors of proteinases in the animal species that the virus infects naturally. This would allow Leporipoxvirus serpins to inhibit rabbit proteinases more effectively than Orthopoxvirus serpins and Orthopoxvirus serpins to inhibit mouse or human proteinases more effectively than Leporipoxvirus serpins. SERP2 may be a better inhibitor of rabbit proinflammatory proteinases than crmA. Experimental testing of this hypothesis requires purified rabbit caspases, which are not yet available. 3.7. SPI-3 and SERP1 with P1 = Arg The myxoma virus SERP1 and cowpox virus SPI-3 proteins both have P1 = Arg, but they are different in sequence terms with only 24% identity. Presumably they have distinct origins in evolution.Although the two serpins have very similar activity profiles in vitro, their biological functions in natural virus infections appear to be dissimilar.

ch06

Jan. 10, 2007

8:57

SPI-B429

178


P. C. Turner et al.

Purified SERP1 protein is an effective inhibitor of the serine proteases plasmin, uPA, tPA, and thrombin in vitro.26 Kinetic studies indicate that the SERP1 association rate constant, ka , varies from 3.4 ± 0.7×105 M−1 s−1 for thrombin to 7.0 ± 0.5×104 M−1 s−1 for uPA.48 SERP1 is naturally secreted following myxoma virus infection of cells, and is modified by N-linked glycosylation.15 However, changing the glycosylation and sialylation state of SERP1 had no effect on proteinase inhibition in vitro.49 The SERP1 gene of MYX is located in each inverted terminal repetition of the virus genome and is therefore present in two copies. SERP1 is not required for the growth of MYX in tissue culture cells, and plays no obvious role in the virus lifecycle under these circumstances. The CPV SPI-3 protein is active against plasmin, uPA, and tPA in vitro,50 and the kinetic values were similar to those for SERP1 against the same enzymes. The only significant difference between SPI-3 and SERP1 occurs with thrombin, where SPI-3 gives little or no inhibition50 but SERP1 is a reasonably good inhibitor.48 When wtCPV infects cells in tissue culture, the cells show evidence of cytopathic effect, but do not fuse with one another. However, inactivation of SPI-3 results in a syncytial phenotype, with fusion between infected cells.51–53 Surprisingly, mutations in the SPI-3 RCL that prevented protease inhibition50 had no effect on inhibition of cell-cell fusion.54 SPI-3 therefore has two distinct activities: as a negative regulator of cell-cell fusion, and as a proteinase inhibitor. SPI-3 is present both on the surface of cells infected with wtCPV, and in extracellular enveloped virus (EEV).55 The apparent membrane localization of SPI-3 is unique for poxvirus serpins, and was initially puzzling as SPI-3 lacks a predicted transmembrane domain. Based purely on the sequence of SPI-3, one would predict that it would be secreted like SERP1, as both proteins have a signal sequence. A CPV mutant with SERP1 in place of SPI-3 still caused fusion of infected cells, and SERP1 is therefore unable to block cell-cell fusion.19 As expected, SERP1 was secreted from cells infected from this CPV recombinant. A reciprocal MYX recombinant with SPI-3 in place of SERP1 gave secretion of SPI-3 on the infection of cells.19 The fact that SPI-3 was associated with cells and EEV in a wtCPV infection, but secreted from cells infected with the MYX/SPI-3 recombinant, suggested that SPI-3 might be associated with a protein present in wtCPV-infected cells, but absent

ch06

Jan. 10, 2007

8:57

SPI-B429



179

from MYX-infected cells. This protein has now been identified as the CPV hemagglutinin (HA), a membrane protein that was initially found to mediate adhesion of chicken erythrocytes to cells infected with orthopoxviruses. HA is not present in the leporipoxvirus MYX. The orthopoxvirus HA is known to inhibit cell-cell fusion,56 and like SPI-3, it is present on the surface of infected cells and on EEV. SPI-3 and HA co-localize on the surface of CPV-infected cells.55 HA was found to localize normally to the cell membrane in the presence or absence of SPI-3.55 However, SPI-3 only localizes to the surface if HA is present; in the absence of HA, some SPI-3 is secreted, and some remains inside the cell.55 Recent work shows that SPI-3 and HA can be coimmunoprecipited from the infected cells, using antibodies against either SPI-3 or against HA.57 The association between SPI-3 and HA can be demonstrated in uninfected cells that have been transfected with plasmids expressing SPI-3 and HA. The SPI-3/HA interaction does not require any additional viral proteins, and is unaffected by mutations in the SPI-3 RCL. A deletion mutant of SPI-3 lacking residues 23–144 was still able to bind to HA, indicating that the binding domain in SPI-3 for HA resided in residues 145–373. The deletion mutant lacks helices A, B, C, D, and E, two strands of β-sheet A, and one strand of β-sheet B. The SPI-3/HA interaction clearly does not require helix D, which is involved in cofactor binding to serpins in many instances, for example, heparin binding to antithrombin.58 The mechanism by which SPI-3 and HA together regulate fusion between the infected cells is currently under investigation. 3.8. The role of P1-Arg serpins in animal models The orthopoxvirus SPI-3 gene appears to have little, if any, effect on disease in animal models. A vaccinia virus mutant deleted for SPI-3 had equivalent immunogenicity and virulence, compared with the wild-type virus in intranasally inoculated mice.51 Likewise, disease manifestations in mice infected intranasally with cowpox SPI-3 deletion mutants or the parent wild-type virus were indistinguishable.28 The conservation of SPI-3 in orthopoxviruses may not reflect any direct role of SPI-3 in pathogenesis, but rather some role that affects cell-cell or virus-cell fusion. One possibility is that SPI-3 may reduce superinfection of previously infected cells by

ch06

Jan. 10, 2007

8:57

SPI-B429

180


P. C. Turner et al.

released virions. At present, there is no data to support this notion, but it remains an attractive hypothesis. Inactivation of SERP1 in MYX altered the course of myxomatosis in infected rabbits, the rapidly fatal disease characterized by the downregulation of host immune function.2,3 Mutation of both copies of the SERP-1 gene from myxoma results in the attenuation of disease and improved survival in rabbits after intradermal infection,59 due to the development of an inflammatory response that allows resolution of the infection.15 Likewise, in the closely related malignant rabbit fibroma virus, deletion of SERP-1 causes attenuated disease, with an effective immune response, recovery and protection from subsequent wild-type viral infection.59 The MYX recombinant with SPI-3 in place of SERP1 was tested for virulence in rabbits and found to be as attenuated as a MYX SERP1 deletion mutant.19 SPI-3 could not substitute for SERP1 in the rabbit/MYX model. SPI-3 was secreted from cells infected with the MYX recombinant as SERP1 would be naturally and the inability of SPI-3 to function in MYX virulence was therefore not due to incorrect localization. Similarly, experiments in which SERP1 was substituted for SPI-3 in CPV, indicated that SERP1 could not substitute for SPI-3 in the context of virus infection. Clearly, the natural functions of CPV SPI-3 and MYX SERP1 differ, despite the similar behavior of the two serpins in vitro.

4. The Use of Virus Serpins as Therapeutic Agents to Control Human Disease The known protease inhibitory profile of SERP148 and the ability to control the inflammatory response15 suggest that this serpin could be useful in the control of undesirable inflammation or immune modulation associated with certain clinical syndromes. SERP1 targets proteinases associated with the thrombolytic system, reinforcing the idea that SERP1 might have efficacy against chronic inflammatory disorders. SERP1 inhibits plasmin, as well as urokinase-type and tissue-type plasminogen activators, both of which activate plasmin, which itself, in turn, activates several matrixmetalloproteinases that are associated with vascular disorders and commonly present in vascular lesions, namely collagenase, stromelysin and

ch06

Jan. 10, 2007

8:57

SPI-B429



181

gelatinase.60 It is possible that decreased matrix-metalloproteinase activation by the presence of SERP1 could decrease the development of vessel narrowing and thickening.60 SERP1 protein has been evaluated as an anti-inflammatory treatment in many disease models with this purpose in mind, including arthritis, atherosclerosis restenosis and transplant vasculopathy.60−66 Inflammation is a known factor in many of these diseases. For example, vascular injury is thought to cause inflammation with a transmigration of monocytes and T-cells from the blood to the intimal layer of the vessel wall, which then leads to migration and proliferation of smooth muscle cells, and subsequent thickening and narrowing of the vessel lumen.67,68 It has been hypothesized that reducing the immune cell infiltrate might eliminate or ameliorate much of the disease-associated pathology.

5. Clinical Evaluation of SERP1 Administered Locally One of the major complications of angioplasty is restenosis, the development of plaque and intimal hyperplasia at the primary site of balloon injury and at adjacent areas upstream and downstream. In initial experiments to evaluate the clinical potential of SERP1, rabbits undergoing balloonmediated aortic injury were treated with partially purified SERP1 isolated from the supernatant of cells infected with a recombinant vaccinia virus over-expressing SERP1 or with aVV control vector.64 Restenosis was found to be markedly reduced in rabbits treated with the SERP1 protein in a single dose at the time of angioplasty, but not in animals receiving vaccinia control samples.64 Subsequent experiments which utilized highly purified SERP1 indicated that major inhibition of plaque development and intimal hyperplasia occurred at very low levels of the serpin. In control experiments, a mutated SERP1 with alanine at the P1 active site of the RCL, which lacks the ability to bind thrombolytic serine proteinases, was evaluated in several models63,64,69 and found to be inactive, suggesting that the effect of SERP1 in reducing chronic inflammation and intimal hyperplasia is linked to the ability of SERP1 to inhibit serine proteases. Studies in many different animal models have since confirmed and extended the efficacy of SERP1 delivered locally. For example, local,

ch06

Jan. 10, 2007

8:57

SPI-B429

182


P. C. Turner et al.

intra-arterial delivery of SERP1 after balloon-mediated aortic injury in rabbits and rats as well as in aortic allografts demonstrated reduced plaque formation.64,69 Furthermore, local, intra-articular administration of SERP1, 2 and 4 weeks after induction of antigen-induced arthritis, a form of chronic inflammation resembling rheumatoid arthritis, reduced cellular hyperplasia in the joint synovial lining and preserved the articular cartilage.65 Treatment efficacy has been demonstrated at picogram to nanogram doses.63,64

6. Clinical Evaluation of SERP1 Administered Systemically Systemic delivery of SERP1 also demonstrated efficacy in multiple animal studies, albeit at slightly higher doses.63,64,69 For instance, the effect of intravenous, systemic SERP1 treatment after arterial balloon angioplasty on the reduction of intimal hyperplasia has been evaluated in a wide range of species, including rat, rabbit, rooster, and microswine.63 Also, rats undergoing heterotopic abdominal heart transplantation were treated with 10 daily intravenous injections of SERP1, in conjunction with cyclosporine to reduce the risk of transplant rejection. A two-fold reduction in intimal thickening in the coronary arteries of the transplanted hearts was found at 90 days of observation, compared with rats receiving only cyclosporine.70 The effect of SERP1 on reducing intimal hyperplasia in the coronary vessels increased as dose increased. Additionally, in both Lewis to Sprague– Dawley and August–Copenhagen–Irish to Lewis rat aortic allograft transplant models, there was a reduction in intimal hyperplasia at a single SERP1 infusion dose of 1 and 0.1 pg/g, respectively.63 SERP1 has also shown efficacy when delivered using a subcutaneous osmotic pump. In apolipoprotein E knock-out mice, in which atherosclerotic plaques were accelerated by placing perivascular collars around the carotid artery, SERP1 was administered for four weeks using pumps that release 2 mg/kg/day.60 Mice treated with SERP1 early, at 1 week after collar placement, had three-fold reduction in early plaque size as well as a doubled luminal area, compared with PBS-treated mice.60 When SERP1 treatment was initiated five weeks after collar placement, there was also a lesser but still significant improvement in luminal size.60

ch06

Jan. 10, 2007

8:57

SPI-B429



183

7. Cellular and Vascular Changes Following SERP1 Treatment SERP1 treatment has been associated with long-term reduction in macrophage infiltrates. In rabbits, rats, roosters and microswine subjected to balloon angioplasty and SERP1 treatment, plaques showed decreased macrophage influx at 12–24 h, and persisting for four weeks.63,64 In some models, there was also a reduction in non-specific T and B lymphocytes, as well as natural killer cell invasion, as early as 12 h after treatment, that was still apparent at 28 days.63,66 Furthermore, in multiple models of transplant vasculopathy in rats, including heterotopic cardiac transplants and aortic allografts, rats treated with SERP1 also had decreased macrophage invasion at 12–48 h post-transplantation.66,70 In addition, immunohistochemical analysis of the plaques that developed in apolipoprotein E-deficient mice undergoing carotid collar placement revealed a significant reduction in macrophages in the SERP1 treated mice, with no evidence of apoptosis, suggesting that SERP1 impaired monocyte recruitment rather than contributing to their destruction.60 The reduction in immune cell entry at the vascular level did not increase infectious complications in any disease model evaluated. The reduction of inflammatory cell influx after treatment with SERP1 may have led to decreased chemotaxis that resulted in lower migration of smooth muscle cells from the arterial media layer.66 Multiple studies showed that after treatment with SERP1, there was a reduction in smooth muscle cells in the intimal or adventitial cellular layers of the treated vessels visualized at 28 days.63,66 In aortic allografts in rats, decreased intimal smooth muscle cells after treatment with SERP1 resulted in reduced local connective tissue deposition, and in apolipoprotein E-deficient mice, the increased smooth muscle cellularity of the media resulted in increased collagen content of the plaques in mice treated with SERP1, which was inferred to improve plaque stability.60,66

8. The use of Gene Therapy to Provide Sustained Delivery of SERP1 Frequently, a single dose of a therapeutic agent is insufficient to have a lasting clinical effect. On the other hand, systemic, intravenous infusion is

ch06

Jan. 10, 2007

8:57

SPI-B429

184


P. C. Turner et al.

costly and expensive to maintain. To create a sustained and non-intravenous mode of systemic delivery for SERP1, we evaluated the efficacy of SERP1 gene delivery using a recombinant adeno-associated viral vector, administered intramuscularly at a dose of 1011 particles. Utilizing carotid ligation, a murine model of intimal hyperplasia, we found equivalent reduction in intimal area after treatment with this AAV serotype 1 vector and treatment with intravenous SERP1 (McAuliffe, Moldawer, and Moyer, unpublished data).

9. Immune Response to SERP1 Rabbits treated with radiolabeled SERP1 did not have a sudden decline in plasma clearance curves after SERP1 treatment over an evaluation period lasting six days, suggesting that there was no immune response generated against SERP1 in the rabbits.71 Little, if any, immune response was also noted after a variety of single dose administration of SERP1 by various routes. However, the administered doses were quite low. The carotid ligation model and, in particular, the use of AAV to deliver SERP1 provides a much more rigid test of the immune response to this foreign protein, as there is constant infusion and the systemic doses tend to be relatively high. However, in mice treated with SERP1 after carotid ligation, we found the presence of low titer anti-SERP1 IgG immunoglobulins in 15% of mice treated with 15 µg of SERP1 and 5% of mice treated with 1.5 µg of SERP1. Continued infusion of SERP1 by the use of recombinant AAV vectors was as efficacious at reducing intimal area after carotid ligation, as the treatment of animals with 15 µg of purified SERP1 protein. However, the mice treated with AAV-delivered SERP1 developed less antibody against SERP1 (5%) than those injected with purified protein.

10. Effect on Thrombotic/Thrombolytic Systems Following Administration of Serp1 No thrombotic or thrombolytic side effects were seen after SERP1 treatment in any animal model evaluated. Despite extensive knowledge of the in vitro inhibitory profile of SERP1,26,48 the precise biological target of SERP1 has not been defined in vivo.64 In fact, SERP1 has been proposed to act as a specific mimic and up-regulator of PAI-1, binding to and blocking

ch06

Jan. 10, 2007

8:57

SPI-B429



185

the pro-inflammatory complex between uPA and its receptor uPAR.69 Evidence consistent with this hypothesis is the observation that after the SERP1 treatment of rabbits or rats undergoing balloon angioplasty, tissue plasminogen activator levels were found to be reduced, while plasminogen activator inhibitor-1 expression was elevated using RT-PCR or Western blot analysis.63,69 Also, urokinase plasminogen activator was elevated69 and matrix metalloproteinase activity was reduced.63 These changes occurred early at 4–12 h, and resolved by the tenth day. An intact RCL was required to detect these alterations, as they were not observed after treatment with the mutated SERP1 in which alanine replaced the arginine at the P1 residue.69 To further evaluate the effect of SERP1 on the thrombolytic enzymes, plaque growth was evaluated in mice with no functional plasminogen activator inhibitor-1.72,73 These mice underwent aortic isograft, a model of chronic rejection with a marked vascular inflammatory response, followed by the treatment with SERP1.69 Without SERP1 treatment, plasminogen-activator inhibitor-1 deficient mice normally develop robust intimal hyperplasia.72,73 After SERP1 treatment, there was a reduction in intimal hyperplasia and plaque growth, suggesting that the presence of plasminogen activator inhibitor-1 was not required for SERP1 function.69 This was corroborated in a wire-injury model in plasminogen activator inhibitor1 deficient mice, which also showed a reduction in plaque size, compared with those treated with saline.69 In contrast, in mice deficient in urokinase plasminogen activator receptor, SERP1 did not inhibit plaque growth after aortic allograft transplantation. This implies that the presence of native urokinase plasminogen activator receptors is required for SERP1 to reduce intimal hyperplasia.69 Taken together, this work suggests that SERP1 inhibits intimal hyperplasia by interacting with and blocking the down-stream activity of the urokinase plasminogen activator and receptor complex, perhaps by mimicking plasminogen activator inhibitor-1, but without a direct requirement for the latter compound.69

11. Conclusions This review of the poxvirus serpins has several recurring themes. Firstly, the serpins are not required for virus growth, but rather function during the

ch06

Jan. 10, 2007

8:57

SPI-B429

186


P. C. Turner et al.

infection of natural hosts, as suggested by their biology and biochemistry. Genes whose purpose is to neutralize host responses to infection are commonly “non-essential” for growth in tissue culture. Frequent deletion of such genes greatly attenuates the virus during the infection of animal hosts. In poxviruses, perhaps a third of the total genome encodes genes which belong to this category. It is most likely that such genes are derived from the hosts which the virus infects. However, since poxviruses develop in the cytoplasm and do not contain introns, the host genes were probably not introduced into the virus by direct capture from DNA within the nucleus of host cells. Rather, the genes were captured in the cytoplasm following reverse transcription of mRNAs by retroviruses or retrovirus-like particles. Once a host gene had been acquired by a poxvirus, the gene could rapidly mutate with the modification of function to specifically fit the needs of the virus. This functional honing can occur within a relatively short period of time since the virus has a generation time of hours. A theme which has emerged through the study of viral genes involved in pathogenesis is that major immunological/host defense pathways are targeted, frequently at critical points where other signal pathways intersect. Therefore, the serpins and other immune modulator genes serve as guides to which host defense pathways must be neutralized for the virus to be successful. Scientists studying such genes have taken advantage of this virus strategy to delineate novel pathways modified by the virus, often in very unusual ways. Finally, perhaps the most intriguing possibility is the use of proteins that neutralize the host’s immune responses as therapeutic agents in the treatment of clinical disorders. Realization of this possibility has triggered an intensive study of such genes to determine their function. In this regard, SERP1 provides an exciting paradigm as the protein is a powerful inhibitor of inflammation and epithelial hyperplasia in a variety of clinical settings. SERP1 treatment has been so successful that phase two clinical trials are about to begin. How and under what conditions to administer such proteins requires careful study. We have shown for the first time that AAV-vectored gene therapy is a promising way to deliver such viral proteins in a therapeutically efficacious manner. In summary, there is little doubt that the viral serpins and other nonessential gene products will serve as model compounds for present and

ch06

Jan. 10, 2007

8:57

SPI-B429



187

future therapeutics. This area of research is rapidly becoming one of the most exciting areas of clinical, translational research.

References 1. Moss B (2001) Poxviridae: The viruses and their replication, in Fields BN, Knipe DM and Howley PM (eds.) Virology, Raven Press: NewYork, pp. 2849– 2883. 2. Johnston JB and McFadden G (2003) Poxvirus immunomodulatory strategies: Current perspectives. J Virol 77:6093–6100. 3. McFadden G, Graham K, Ellison K, Barry M, Macen J, Schreiber M, Mossman K, Nash P, Lalani A and Everett H (1995) Interruption of cytokine networks by poxviruses: Lessons from myxoma virus. J Leukoc Biol 57: 731–738. 4. Clambey ET, Virgin HW and Speck SH (2000) Disruption of the murine gammaherpesvirus 68 M1 open reading frame leads to enhanced reactivation from latency. J Virol 74:1973–1984. 5. Guerin JL, Gelfi J, Camus C, Delverdier M, Whisstock JC, Amardeihl MF, Py R, Bertagnoli S and Messud-Petit F (2001) Characterization and functional analysis of Serp3:A novel myxoma virus-encoded serpin involved in virulence. J Gen Virol 82:1407–1417. 6. Turner PC and Moyer RW (2002) Poxvirus immune modulators: Functional insights from animal models. Virus Res 88:35–53. 7. Johnston JB and McFadden G (2004) Technical knockout: Understanding poxvirus pathogenesis by selectively deleting viral immunomodulatory genes. Cell Microbiol 6:695–705. 8. Bennett M, Crouch AJ, Begon M, Duffy B, Feore S, Gaskell RM, Kelly DF, McCracken CM, Vicary L and Baxby D (1997) Cowpox in British voles and mice. J Comp Pathol 116:35–44. 9. Miller CG, Justus DE, Jayaraman S and Kotwal GJ (1995) Severe and prolonged inflammatory response to localized cowpox virus infection in footpads of C5-deficient mice: Investigation of the role of host complement in poxvirus pathogenesis. Cell Immunol 162:326–332. 10. Tscharke DC and Smith GL (1999) A model for vaccinia virus pathogenesis and immunity based on intradermal injection of mouse ear pinnae. J Gen Virol 80:2751–2755. 11. Shchelkunov SN, Marennikova SS and Moyer RW (2005) Orthopoxviruses pathogenic for humans. Plenum: US.

ch06

Jan. 10, 2007

8:57

SPI-B429

188


P. C. Turner et al.

12. Martinez MJ, Bray MP and Huggins JW (2000) A mouse model of aerosoltransmitted orthopoxviral disease: Morphology of experimental aerosoltransmitted orthopoxviral disease in a cowpox virus-BALB/c mouse system. Arch Pathol Lab Med 124:362–377. 13. Nash P, Barrett J, Cao JX, Hota-Mitchell S, Lalani AS, Everett H, Xu XM, Robichaud J, Hnatiuk S, Ainslie C, Seet BT and McFadden G (1999) Immunomodulation by viruses: The myxoma virus story. Immunol Rev 168:103–120. 14. Martin DB (2002) The cause of death in smallpox: An examination of the pathology record. Mil Med 167:546–551. 15. Macen JL, Upton C, Nation N and McFadden G (1993) SERP1, a serine proteinase inhibitor encoded by myxoma virus, is a secreted glycoprotein that interferes with inflammation. Virology 195:348–363. 16. Messud-Petit F, Gelfi J, Delverdier M, Amardeilh MF, Py R, Sutter G and Bertagnoli S (1998) Serp2, an inhibitor of the interleukin-1β-converting enzyme, is critical in the pathobiology of Myxoma virus. J Virol 72:7830–7839. 17. Nash P, Lucas A and McFadden G (1997) SERP-1, a poxvirus-encoded serpin, is expressed as a secreted glycoprotein that inhibits the inflammatory response to myxoma virus infection. Adv Exp Med Biol 425:195–205. 18. Nathaniel R, MacNeill AL, Wang YX, Turner PC and Moyer RW (2004) Cowpox virus CrmA, Myxoma virus SERP2 and baculovirus P35 are not functionally interchangeable caspase inhibitors in poxvirus infections. J Gen Virol 85:1267–1278. 19. Wang YX, Turner PC, Ness TL, Moon KB, Schoeb TR and Moyer RW (2000) The cowpox virus SPI-3 and Myxoma virus SERP1 serpins are not functionally interchangeable despite their similar proteinase inhibition profiles in vitro. Virology 272:281–292. 20. Fenner F and Woodroofe GM (1953) The pathogenesis of infectious myxomatosis; the mechanism of infection and the immunological response in the European rabbit (Oryctolagus cuniculus). Br J Exp Pathol 34:400–411. 21. Ali AN, Turner PC, Brooks MA and Moyer RW (1994) The SPI-1 gene of rabbitpox virus determines host range and is required for hemorrhagic pock formation. Virology 202:305–314. 22. Wyatt LS, Carroll MW, Czerny CP, Merchlinsky M, Sisler JR and Moss B (1998) Marker rescue of the host range restriction defects of modified vaccinia virus Ankara. Virology 251:334–342. 23. Brooks MA, Ali AN, Turner PC and Moyer RW (1995) A rabbitpox virus serpin gene controls host range by inhibiting apoptosis in restrictive cells. J Virol 69:7688–7698.

ch06

Jan. 10, 2007

8:57

SPI-B429



189

24. Moon KB, Turner PC and Moyer RW (1999) SPI-1 dependent host range of Rabbitpox virus and complex formation with cathepsin G is associated with serpin motifs. J Virol 73:8999–9010. 25. Shisler JL, Isaacs SN and Moss B (1999) Vaccinia virus serpin-1 deletion mutant exhibits a host range defect characterized by low levels of intermediate and late mRNAs. Virology 262:298–311. 26. Lomas DA, Evans DL, Upton C, McFadden G and Carrell RW (1993) Inhibition of plasmin, urokinase, tissue plasminogen activator, and C1S by a myxoma virus serine proteinase inhibitor. J Biol Chem 268:516–521. 27. Luttge BG and Moyer RW (2005) Suppressors of a host range mutation in the rabbitpox virus serpin SPI-1 map to proteins essential for viral DNA replication. J Virol 79:9168–9179. 28. Thompson JP, Turner PC, Ali AN, Crenshaw BC and Moyer RW (1983) The effects of serpin gene mutations on the distinctive pathobiology of cowpox and rabbitpox virus following intranasal inoculation of Balb/c mice. Virology 197:328–338. 29. Kettle S, Blake NW, Law KM and Smith GL (1995) Vaccinia virus serpins B13R (SPI-2) and B22R (SPI-1) encode Mr 38.5 and 40K, intracellular polypeptides that do not affect virus virulence in a murine intranasal model. Virology 206:136–147. 30. Legrand FA, Verardi PH, Jones LA, Chan KS, Peng Y and Yilma TD (2004) Induction of potent humoral and cell-mediated immune responses by attenuated vaccinia virus vectors with deleted serpin genes. J Virol 78:2770–2779. 31. Pickup DJ, Ink BS, Hu W Ray CA and Joklik WK (1986) Hemorrhage in lesions caused by cowpox virus is induced by a viral protein that is related to plasma protein inhibitors of serine proteases. Proc Natl Acad Sci USA 83:7698–7702. 32. Pickup DJ, Ink BS, Parsons BL, Hu W and Joklik WK (1984) Spontaneous deletions and duplications of sequences in the genome of cowpox virus. Proc Natl Acad Sci USA 81:6817–6821. 33. Ray CA, Black RA, Kronheim SR, Greenstreet TA, Sleath, PR, Salvesen GS and Pickup DJ (1992) Viral inhibition of inflammation: Cowpox virus encodes an inhibitor of the interleukin-1 beta converting enzyme. Cell 69:597–604. 34. Garcia-Calvo M, Peterson EP, Leiting B, Ruel R, Nicholson DW and Thornberry NA (1998) Inhibition of human caspases by peptide-based and macromolecular inhibitors. J Biol Chem 273:32608–32613. 35. Zhou Q, Snipas S, Orth K, Muzio M, Dixit VM and Salvesen GS (1997) Target protease specificity of the viral serpin CrmA. Analysis of five caspases. J Biol Chem 272:7797–7800.

ch06

Jan. 10, 2007

8:57

SPI-B429

190


P. C. Turner et al.

36. Quan LT, Caputo A, Bleackley RC, Pickup DJ and Salvesen GS (1995) Granzyme B is inhibited by the cowpox virus serpin cytokine response modifier A. J Biol Chem 270:10377–10379. 37. Bump NJ, Hackett M, Hugunin M, Seshagiri S, Brady K, Chen P, Ferenz C, Franklin S, Ghayur T, Li P, Licari P, Mankovitch J, Shi L, Greenberg AH, Miller LK and Wong WW (1995) Inhibition of ICE family proteases by baculovirus antiapoptotic protein p35. Science 269:1885–1888. 38. Ray CA and Pickup DJ (1996) The mode of death of pig kidney cells infected with cowpox virus is governed by the expression of the crmA gene. Virology 217:384–391. 39. Petit F, Bertagnoli S, Gelfi J, Fassy F, Boucraut-Baralon C and Milon A (1996) Characterization of a myxoma virus-encoded serpin-like protein with activity against interleukin-1 β-converting enzyme. J Virol 70:5860–5866. 40. Turner PC, Sancho MC, Thoennes SR, CaputoA, Bleackley RC and Moyer RW (1999) Myxoma virus serp2 is a weak inhibitor of granzyme B and interleukin1beta-converting enzyme in vitro and unlike CrmA cannot block apoptosis in cowpox virus-infected cells. J Virol 73:6394–6404. 41. MacNeill AL, Turner PC and Moyer RW (2006) Mutation of the Myxoma virus SERP2 P1-site to prevent proteinase inhibition causes apoptosis in cultured RK-13 cells and attenuates disease in rabbits, but mutation to alter specificity causes apoptosis without reducing virulence. Virology 356:12–22. 42. Tesch LD, Raghavendra MP, Bedsted-Faarvang T, Gettins PG and Olson ST (2005) Specificity and reactive loop length requirements for crmA inhibition of serine proteases. Protein Sci 14:533–542. 43. Kettle S, Alcami A, Khanna A, Ehret R, Jassoy C and Smith GL (1997) Vaccinia virus serpin B13R (SPI-2) inhibits interleukin-1beta-converting enzyme and protects virus-infected cells from TNF- and Fas-mediated apoptosis, but does not prevent IL-1beta-induced fever. J Gen Virol 78 (Pt 3):677–685. 44. Tscharke DC, Reading PC and Smith GL (2002) Dermal infection with vaccinia virus reveals roles for virus proteins not seen using other inoculation routes. J Gen Virol 83:1977–1986. 45. Dieli F, Lio D, Sireci G and Salerno A (1988) Genetic control of C3 production by the S region of the mouse MHC. J Immunogenet 15:339–343. 46. Janeway CA, Travers P, Walport M and Capra JD (1999) Immunobiology. Garland Publishing: New York 47. Bird CH, Sutton VR, Sun J, Hirst CE, Novak A, Kumar S, Trapani JA and Bird PI (1998) Selective regulation of apoptosis: The cytotoxic lymphocyte serpin proteinase inhibitor 9 protects against granzyme B-mediated apoptosis without perturbing the Fas cell death pathway. Mol Cell Biol 18:6387–6398.

ch06

Jan. 10, 2007

8:57

SPI-B429



191

48. Nash P, Whitty A, Handwerker J, Macen J and McFadden G (1998) Inhibitory specificity of the anti-inflammatory myxoma virus serpin, SERP-1. J Biol Chem 273:20982–20991. 49. Nash P, Barry M, Seet BT, Veugelers K, Hota S, Heger J, Hodgkinson C, Graham K, Jackson RJ and McFadden G (2000) Post-translational modification of the myxoma-virus anti-inflammatory serpin SERP-1 by a virally encoded sialyltransferase. Biochem J 347:375–382. 50. Turner PC, Baquero MT, Yuan S, Thoennes SR and Moyer RW (2000) The cowpox virus serpin SPI-3 complexes with and inhibits urokinase-type and tissue-type plasminogen activators and plasmin. Virology 272:267–280. 51. Law KM and Smith GL (1992) A vaccinia serine protease inhibitor which prevents virus-induced cell fusion. J Gen Virol 73:549–557. 52. Turner PC and Moyer RW (1992) An orthopoxvirus serpinlike gene controls the ability of infected cells to fuse. J Virol 66:2076–2085. 53. Zhou J, Sun XY, Fernando GJ and Frazer IH (1992) The vaccinia virus K2L gene encodes a serine protease inhibitor which inhibits cell-cell fusion. Virology 189:678–686. 54. Turner PC and Moyer RW (1995) Orthopoxvirus fusion inhibitor glycoprotein SPI-3 (open reading frame K2L) contains motifs characteristic of serine proteinase inhibitors that are not required for control of cell fusion. J Virol 69:5978–5987. 55. Brum LM, Turner PC, Devick H, Baquero MT and Moyer RW (2003) Plasma membrane localization and fusion inhibitory activity of the cowpox virus serpin SPI-3 require a functional signal sequence and the virus encoded hemagglutinin. Virology 306:289–302. 56. Ichihashi Y and Dales S (1971) Biogenesis of poxviruses: Interrelationship between hemagglutinin production and polykaryocytosis. Virology 46:533–543. 57. Turner PC and Moyer RW (2006) The cowpox virus fusion regulator proteins SPI-3 and hemagglutinin interact in infected and uninfected cells. Virology 347:88–99. 58. Jin L, Abrahams JP, Skinner R, Petitou M, Pike RN and Carrell RW (1997) The anticoagulant activation of antithrombin by heparin. Proc nat acad sci USA 94:14683–14688. 59. Upton C, Macen JL, Wishart DS and McFadden G (1990) Myxoma virus and malignant rabbit fibroma virus encode a serpin-like protein important for virus virulence. Virology 179:618–631. 60. Bot I, der Thusen JH, Donners MM, Lucas A, Fekkes ML de Jager SC, Kuiper J, Daemen MJ, Van Berkel TJ, Heeneman S and Biessen EA (2003)

ch06

Jan. 10, 2007

8:57

SPI-B429

192

61.

62.

63.

64.

65.

66.

67. 68. 69.

70.

71.


P. C. Turner et al.

Serine protease inhibitor Serp-1 strongly impairs atherosclerotic lesion formation and induces a stable plaque phenotype in ApoE-/-mice. Circ Res 93: 464–471. Liu L, Dai E, Miller L, Seet B, Lalani A, Macauley C, Li X, Virgin HW, Bunce C, Turner P, Moyer R, McFadden G and Lucas A (2004) Viral chemokinebinding proteins inhibit inflammatory responses and aortic allograft transplant vasculopathy in rat models. Transplantation 77:1652–1660. Liu LY, Viswanathan K, Dai EB, Turner PC, Togonu-Bickersteth J, Pang B, Li Y, Moyer RW, McFadden G and Lucas A (2004) Analysis of anti-apoptotic intracellular viral serpins as anti-atherosclerotic agents. J Am College Cardiol 43:54A. Lucas A, Dai E, Liu L, Guan H, Nash P, McFadden G and Miller L (2000) Transplant vasculopathy: Viral anti-inflammatory serpin regulation of atherogenesis. J Heart Lung Transplant 19:1029–1038. Lucas A, Liu L.-Y, Macen J, Nash P, Dai E, Stewart M, Graham K, Etches W, Boshkov L, Nation PN, Humen D, Hobman ML and McFadden G (1996)Virusencoded serine proteinase-inhibitor SERP-1 inhibits atherosclerotic plaque development after balloon angioplasty. Circulation 94:2890–2900. Maksymowych WP, Nation N, Nash P, Macen J, Lucas A, McFadden G and Russell AS (1996) Amelioration of antigen induced arthritis in rabbits treated with a secreted viral serine proteinase inhibitor. J Rheumatol 23:878–882. Miller LW, Dai E, Nash P, Liu L, Icton C, Klironomos D, Fan L, Nation PN, Zhong R, McFadden G and Lucas A (2000) Inhibition of transplant vasculopathy in a rat aortic allograft model after infusion of anti-inflammatory viral serpin. Circualtion 101:1598–1605. Haber E (1996) Can a viral serine proteinase inhibitor prevent postangioplasty restenosis? Circulation 94:2694–2695. Ross R (1995) Cell biology of atherosclerosis. Annu Rev Physiol 57:791–804. Dai E, Guan H, Liu L, Little S, McFadden G, Vaziri S, Cao H, Ivanova IA, Bocksch L and Lucas A (2003) Serp-1, a viral anti-inflammatory serpin, regulates cellular serine proteinase and serpin responses to vascular injury. J Biol Chem 278:18563–18572. Hausen B, Boeke K, Berry GJ and Morris RE (2001) Viral serine proteinase inhibitor (SERP-1) effectively decreases the incidence of graft vasculopathy in heterotopic heart allografts. Transplantation 72:364–368. Hatton MW, Ross B, Southward SM and Lucas A (2000) Metabolism and distribution of the virus-encoded serine proteinase inhibitor SERP-1 in healthy rabbits. Metabolism 49:1449–1452.

ch06

Jan. 10, 2007

8:57

SPI-B429



193

72. Carmeliet P, Stassen JM, Schoonjans L, Ream B, van den Oord JJ, De Mol M, Mulligan RC and Collen D (1993) Plasminogen activator inhibitor-1 genedeficient mice. II. Effects on hemostasis, thrombosis and thrombolysis. J Clin Invest 92:2756–2760. 73. Carmeliet P, Kieckens L, Schoonjans L, Ream B, van Nuffelen A, Prendergast G, Cole M, Bronson R, Collen D and Mulligan RC (1993) Plasminogen activator inhibitor-1 gene-deficient mice. I. Generation by homologous recombination and characterization. J Clin Invest 92:2746–2755.

ch06

Jan. 10, 2007 8:57 SPI-B429



ch07

Jan. 10, 2007

8:57

SPI-B429


7 The Intracellular Serpins of Caenorhabditis elegans Stephen C. Pak, Yuko. S. Askew and Cliff J. Luke

1. Introduction Attempts to identify the function of intracellular serpins in humans and other mammalian systems have been in vain, partly because of the lack of an association between naturally occurring mutations/variations and disease phenotypes. Moreover, targeted gene inactivation approaches in mouse models have not been very informative and are complicated by the expanded repertoire of the serpin loci, leading to functional redundancy/overlap. Simpler organisms have less complex genomes, are easier to manipulate genetically, and provide a more tractable system for studying gene function. The nematode, C. elegans, is a simple experimental organism whose cellular and biological processes are well conserved throughout evolution. This chapter describes how this nematode has been used to study the biological function of serpins. C. elegans is a free-living nematode and the simplest, multicellular organism known to harbor a serpin gene family (Fig. 1). The nematodes are found in the soil of temperate regions and grow to a maximum length of 1.5 mm. Among their many advantages are a sequenced genome (97 megabases), ease of laboratory cultivation, a short life cycle (3 days), and high fecundity. Hermaphrodites are self-fertilizing and capable of generating in excess of 300 progeny (more if fertilized by a male) within 3 to 4 days. This means that an individual carrying a recessive mutation will 195

ch07

Jan. 10, 2007

8:57

SPI-B429

196


S. C. Pak, Y. S. Askew & C. J. Luke

Fig. 1. Caenorhabditis elegans (A) Nomarski photomicrograph of an adult hermaphrodite. The head is to the right and the tail is to the left.Abbreviations: emb, embryo; sp, spermatheca. (B) Life-cyle. The number above the diagram represents the time of (h) postfertilization at 25◦ C. Developmental stage is indicated inside the box.

produce homozygous progeny without the need to set up specific crosses. Moreover, if required, hermaphrodites can be mated with males to produce progeny with desired genotypes. The ability to rapidly produce numerous genotypes and phenotypes makes C. elegans particularly suited for genetic studies. Anatomically, the worm is a simple tubular organism that consists largely of a gonad and a gastrointestinal tract surrounded by an outer tube consisting of neurons, muscles, hypodermis, and cuticle (Fig. 1). A wild-type hermaphrodite is made up of exactly 959 somatic cells, the number and position of which remain invariant between animals. In addition, the worm is transparent throughout all stages of development. This makes it possible to follow and record the origin and fate of each individual cell created from fertilization to a mature adult using differential interference contrast (DIC) microscopy. A complete knowledge of cell lineage and wiring of all (∼350) neurons is available, providing investigators with a powerful tool to study how genes influence cell fate. In addition, the availability of well-established genetic tools has enabled the study of various biological processes relevant to human health (e.g. genetic

ch07

Jan. 10, 2007

8:57

SPI-B429


The Intracellular Serpins of Caenorhabditis elegans

197

components controlling programmed cell death, cancer, and neurodegenerative diseases). C. elegans development is divided into two stages: embryonic and postembryonic. Embryogenesis begins at fertilization and ends 14 h (at 25◦ C) later when the worm hatches from its egg. Postembryonic development entails growth through four larval stages punctuated by molts at 29, 38, 47, and 59 h postfertilization (Fig. 1).1 When conditions become unfavorable for survival, e.g. in the absence of food and high population density, worms can enter an alternative developmental stage, the dauer, at the second molt. Dauers are well equipped for surviving adverse conditions and can remain viable in this dormant state for several months. When conditions become suitable for growth, dauers can resume development at the L4 stage and proceed to become normal adults.

2. Serpins in C. elegans The C. elegans genome harbors a total of nine serpin-like genes.2–4 Characteristics of the C. elegans serpins are presented in Fig. 2. All of the serpins identified are located on chromosome V. Many of the serpin genes are closely linked and share significant sequence similarity. srp-5/srp-6 and srp-9/srp-10 share 89% and 90% identity, respectively, at the cDNA level. The high degree of sequence homology and conservation of gene structure suggests that these genes occurred through a duplication of an ancestral precursor, after C. elegans diverged from other nematodes. srp-1, srp-2, srp-3, srp-6, and srp-7 transcripts encode full-length serpins with well-conserved structural elements, such as a hinge region rich in alanines, characteristic of functional inhibitory-type serpins. Reminiscent of the Manduca sexta serpin gene-1,5 srp-7 encodes three chimeric serpins with distinct peptidase inhibitory activity resulting from alternate utilization of 3 RSL coding exons. Of the cDNAs isolated thus far, srp-5, srp-9, and srp-10 transcripts contain multiple premature translational stop codons suggesting that they encode non-functional pseudogenes. Interestingly, srp-8 encodes a full-length protein that lacks a noticeable RSL motif, suggesting that it is unlikely to function as a peptidase inhibitor. However, a role in a non-proteolytic process has not been assessed.

ch07

Jan. 10, 2007

8:57

SPI-B429

198



Fig. 2. Schematic representation of the nine C. elegans serpin genes on chromosome V. The genomic positions of C. elegans serpins are marked with both map units and nucleotide sequence positions. The transcriptional orientations of the genes are indicated with black arrows. In accordance with genetic nomenclature guidelines provided by the Caenorhabditis Genetics Center, the genes are named srp-1 to srp-10. Note, srp-4 cDNA was initially isolated from Yuji Kohara’s EST database and thought to encode a novel serpin, but, subsequent analysis indicated that it was identical to srp-3 except for a nucleotide substitution in the RSL. Since this difference is most likely to be a sequencing error or a cloning artifact, srp-4 was omitted from this figure. Genes marked with an asterisk contain premature termination codons (i.e., the RSL sequence was deduced by reading through the stop codon). Exons are indicated as open boxes, and the numbers under each indicate exon size (bp). The position and size of the introns are indicated between the exons by lines. The letters within the exons of srp-7 represent alternately spliced RSL-containing exons. This figure was modified from Ref. 4 and reproduced here with permission from Frontiers in Bioscience.

ch07

Jan. 10, 2007

8:57

SPI-B429



199

3. Evolution of C. elegans Serpins Delineating the phylogenetic associations between members of a protein family can provide insight into the evolution of domain structure and function relationships. While unable to determine precise mechanistic and phenotypic knowledge, evolutionary relationships may shed light on the broad functional roles, as highly conserved genes will tend to have similar biologic functions and are more likely to be involved in fundamental processes than that of non-conserved genes. When serpins from Caenorhabditis species are compared, the nematode serpins fall into three subclades (Fig. 3).4 The first subclade contains the srp-1-like serpins. The second subclade contains srp-2 and srp-3. The remaining serpins, srp-5, -6, -7, -8, -9, and -10, belong to the third subclade. Interestingly, members belonging to the first and second subclade (i.e. srp-1, -2, and -3) have clear orthologues in C. briggsae (CBG06351A; CBG06352B) and C. remanei (Cre128.5.3; Cre128.5.2). In addition, both C. briggsae (CBG06351A; CBG06352B) and C. remanei (Cre128.5.3; Cre128.5.2) have an expanded repertoire of srp-3-like genes that are not found in C. elegans (Fig. 3), suggesting that these genes arose from a srp-3 precursor after C. briggsae and C. remanei diverged from C. elegans. Of the six serpins belonging to the third subclade, only srp-7 appears to have clear orthologues in C. remanei and C. briggsae. While the precise roles of srp-1, -2, -3, and -7 are currently unknown, their conservation throughout the other Caenorhabditis species suggests that these serpins are essential and play important roles in fundamental biological processes such as development. It is interesting to note that all of the predicted non-functional serpins lack orthologous genes in the other Caenorhabditis species. Of the predicted functional serpins, srp-6 does not have an orthologue in the other species. The lack of orthologues may indicate that srp-6 performs a function specific to C. elegans, or that other serpin genes in other species are utilized to replace the biological role of srp-6.

4. C. elegans Serpin Gene Structure In mammals and insects, many serpins possess N-terminal hydrophobic signal peptides and function extracellularly. The nematode serpins do

ch07

Jan. 10, 2007

8:57

SPI-B429

200



Fig. 3. Phylogenetic analysis of the Clade L serpins. The nematode serpins were aligned with 276 other serpin sequences and a bootstrapped neighbor-joining tree was created using ClustalX v1.83. For clarity, the tree branch that contained the Clade L serpins was isolated using NJplot. The bootstrap values are shown at the branch splits. The length of the branches indicates the number of base-pair substitutions per 100 amino acids (scale bar = 0.05 substitutions/100 residues). Three subgroups were identified; srp-1-like family (blue); srp-2and srp-3-like family (green); srp-7-like family (red). This figure was reproduced with permission from Ref. 4.

ch07

Jan. 10, 2007

8:57

SPI-B429



201

not possess cleavable hydrophobic signal sequences, suggesting that they reside intracellularly and function cell autonomously. When compared with members of other clades, the C. elegans serpins show high degree of similarity with members of the Clade B, intracellular serpin family. Apart from lacking N-terminal signal sequences, the nematode serpins also have other features that are strongly conserved in clade B serpins such as the lack of N- and C-terminal extensions and the presence of an Ala in the P10 position. Thus, a study of C. elegans serpins may provide much needed insight into the functions of intracellular serpins and expand our current knowledge of serpins in higher organisms.

5. The Biochemistry of SRP-2 The best-characterized C. elegans serpin to date is SRP-2.2 Studies using recombinant protein expressed in E. coli demonstrated that SRP-2 is a dual cross-class inhibitor of serine and cysteine peptidases. The reactive center (P2–P2 ) of SRP-2, Leu-Glu-Met-Met, is rare (Fig. 4). To date, of all (>700) serpins known, only SERPINB9 is known to have a P1 Glu. Serpins with acidic (Glu or Asp) residues at their P1 position are involved in regulating peptidases that are important in the execution of apoptosis. For example, SERPINB9 is a known inhibitor of granzyme B, serine peptidase that is produced by cytotoxic lymphocytes to destroy virus infected and malignant cells by inducing apoptosis.6 In vitro peptidase inhibition studies have shown that SRP-2, like SERPINB9, is capable of efficiently inactivating human granzyme B with second-order rate constants in the physiologically relevant range (>104 M−1 s−1 ). The presence of a granzyme B-like peptidase inhibitor is noteworthy. Infectious organisms have developed mechanisms for evading the host-defense system. For example, the viral serpin, crmA (P1-Asp), inhibits proteolytic activation of interleukin-β and protects virus-infected cells from Fas- and TNF-induced apoptosis.7 Although C. elegans is a non-parasitic nematode that is unlikely to find its way into the human blood stream, it can nevertheless serve as a model for parasitic nematodes such as Brugia malayi (associated with filariasis) and Onchocerca vovulus (causative agent of river blindness), thus providing crucial insights into the biochemical pathways open to therapeutic intervention. A search of the C. elegans genome identified a candidate serine peptidase, F31D4.6,

ch07

Jan. 10, 2007

8:57

SPI-B429

202



Fig. 4. Amino acid alignment of the nematode serpin RSLs. The RSL regions of the nematode serpins, containing the putative reactive center (P4–P4 ) were aligned using ClustalX v1.83. The colors indicate polar (green), nonpolar/hydrophobic (yellow), acidic (red), and basic (blue) residues. The putative scissile bond is marked by an arrowhead.

with an Arg residue in position 189 (based on chymotrypsin numbering) of the active site similar to granzyme B. Although multiple residues in the active site pocket of a peptidase affect substrate specificity, the residue at position 189 provides insight into the peptidase specificity. For example, the human granzyme B harbors a positively charged residue (Arg) at the active site to facilitate interaction with negatively charged residues such as Glu and Asp. Further work is required to determine whether F31D4.6 is granzyme B-like and whether it is an in vivo target of SRP-2. The presence of a Leu residue in the P2 position suggested that SRP-2 might also be capable of neutralizing members of the cysteine peptidase family. Indeed, kinetic studies have shown that SRP-2 can exhibit crossclass inhibitory activity against cathepsins (cat) K, L, S, and V with second order rate constants in the physiological range.2 Serpins with dual crossclass activity have also been detected in mammals (e.g., the mouse serpin, Serpinb3b).8 These results suggest that dual cross-class activity may have originated around the divergence of nematodes and that it may be more common than originally appreciated to be.

6. The Biology of SRP-2 A survey of SRP-2 temporal and spatial expression profile using promoter::GFP fusions indicates that SRP-2 is expressed in all major stages of

ch07

Jan. 10, 2007

8:57

SPI-B429



203

Fig. 5. SRP-2 expression pattern and phenotype. (A,C) Normaski and (B,D) fluorescent photomicrograph of an adult hermaphrodite expressing a srp-2::gfp transcriptional fusion construct. Note strong expression of srp-2 in the intestine, anterior and posterior (phasmid) neurons. (E) A transgenic L1 larvae overexpressing srp-2. The animal is trapped in its old cuticle (arrowheads) and developmentally arrested. Abbreviations: int, intestine; phx, pharynx; nr, nerve ring; phm, phasmid neurons.

the male and hermaphrodite development. In particular, strong GFP expression is seen in numerous hypodermal and intestinal cells of the embryo. During the postembryonic development, strong expression is evident in the hypodermal cells prior to the larval molts. In adult hermaphrodites, expression is also evident in the intestine, hypoderm, seam cells, fibrous organelles, neurons, and sensory support cells (Fig. 5). The strong expression of SRP-2 during embryonic and postembryonic phases suggests that

ch07

Jan. 10, 2007

8:57

SPI-B429

204



SRP-2 may play a role in regulating peptidases that are critical for normal C. elegans development. Despite the low number (five) of functional serpins in C. elegans, an srp-2 null mutant failed to display an overt phenotype. This is not surprising as numerous knockouts in the fruit fly and mouse models have also failed to reveal an overt phenotype. The lack of a phenotype suggests that SRP-2 does not play a critical role in a fundamental biological process, or that other serpins with similar inhibitory activity are capable of compensating for the absence of srp-2. Support for the latter hypothesis comes from the observation that mice deficient in Serpinb6 demonstrated compensatory upregulation of functionally related serpins.9 Moreover, a number of other C. elegans serpins may have overlapping functional activity with SRP-2. Based on the presence of hydrophobic residues in the P2 positions, SRP-1, SRP-6, and SRP-7 may also neutralize papain-like cysteine peptidases and have similar peptidase targets as SRP-2 (Fig. 4). Thus, to circumvent problems associated with functional overlap, multiple serpins with similar function (and expression pattern) may need to be inactivated to generate a phenotype. Alternatively, the unique feature of the serpinpeptidase interaction may be exploited to understand serpin gene function. Since serpins are capable of irreversibly inactivating their target peptidases, it may be possible to overcome the issue of redundancy/functional overlap and generate a phenotype by overexpression of a given serpin. Indeed, transgenic animals engineered to overexpress SRP-2 displayed gross developmental abnormalities (Fig. 5). The most severely affected animals were lethal at the early larval stages and many were trapped in their old cuticles, suggesting a defect in the molting pathway. The molting process in the worm requires the activity of peptidases to digest cuticular and other proteins for shedding of the old cuticle. Thus, a surplus of serpin activity may explain why the molting pathway would be disrupted in animals overexpressing SRP-2. Recently, cathepsin-like cysteine peptidases have been implicated in the progression of development in C. elegans. RNA-mediated interference studies aimed at debilitating the function of CPL-1, a cathepsin-L-like peptidase, resulting in larval arrest, and slow growth of affected animals.10 In addition, a null mutant of cpz-1, a C. elegans cathepsin-Z-like peptidase, displayed gross developmental and molting abnormalities.11 The relationships between CPL-1, CPZ-1,

ch07

Jan. 10, 2007

8:57

SPI-B429



205

and SRP-2 are striking. Firstly, CPL-1 and CPZ-1 are members of the papain-like cysteine peptidases and a possible target of SRP-2. Secondly, the two peptidases are expressed in similar cell types, hypoderm and cuticle during worm development. Thirdly, the SRP-2 overexpression phenotype is remarkably similar to the hypomorphic phenotype of the peptidase cpl-1 and the null phenotype of cpz-1. Although, it is unknown at this point whether CPL-1 or CPZ-1 are indeed in vivo target of SRP-2, these observations suggest that SRP-2 may function by neutralizing peptidase activity required for normal C. elegans development.

7. Conclusion The study of serpins in nematodes is only at its embryonic stages. Nonetheless, the observation that SRP-2 overexpression in wild-type animals leads to significant defects in postembryonic development underscores the significance of serpins in regulating biological processes and demonstrates the usefulness of C. elegans as a model organism for the study of serpin biology. Biochemical characterization of the other serpins and the generation of more null mutants and overexpressors should help define the biological importance of the intracellular serpins, not only in C. elegans but also in higher organisms.

References 1. Wood WB, Hecht R, Carr S, Vanderslice R, Wolf N and Hirsh D (1980) Parental effects and phenotypic characterization of mutations that affect early development in Caenorhabditis elegans. Dev Biol 74(2):446–469. 2. Pak SC, Kumar V, Tsu C, et al. (2004) Srp-2 is a cross-class inhibitor that participates in post-embryonic development of the nematode Caenorhabditis elegans: initial characterization of the Clade L serpins. J Biol Chem 279: 15448–15459. 3. Pak SC, Tsu C, Luke CJ, AskewYS and Silverman GA (2006) The Caenorhabditis elegans muscle specific serpin, SRP-3 neutralizes chymotrypsin-like serine peptidases. Biochemistry 45(14):4474–4480. 4. Luke CJ, Pak SC, Askew DJ, Askew YS, Smith JE and Silverman GA (2006) Selective conservation of the RSL-encoding, proteinase inhibitorytype, clade I serpins in Caenorhabditis species. Front Biosci 11:581–594.

ch07

Jan. 10, 2007

8:57

SPI-B429

206



5. Jiang H, Wang Y, Huang Y, et al. (1996) Organization of serpin gene-1 from Manduca sexta. Evolution of a family of alternate exons encoding the reactive site loop. J Biol Chem 271(45):28017–28023. 6. Sun J, Bird CH, Sutton V, et al. (1996) A cytosolic granzyme B inhibitor related to the viral apoptotic regulator cytokine response modifier A is present in cytotoxic lymphocytes. J Biol Chem 271(44):27802–27809. 7. Kettle S,AlcamiA, KhannaA, Ehret R, Jassoy C and Smith GL (1997) Vaccinia virus serpin B13R (SPI-2) inhibits interleukin-1beta-converting enzyme and protects virus-infected cells from TNF- and Fas-mediated apoptosis, but does not prevent IL-1beta-induced fever. J Gen Virol 78(Pt 3):677–685. 8. Askew DJ, Askew YS, Kato Y, et al. (2004) The amplified mouse squamous cell carcinoma antigen gene locus contains a serpin (serpinb3b) that inhibits both papain-like cysteine and trypsin-like serine proteinases. Genomics 84(1): 166–175. 9. Scarff KL, Ung KS, Nandurkar H, Crack PJ, Bird CH and Bird PI (2004) Targeted disruption of SPI3/Serpinb6 does not result in developmental or growth defects, leukocyte dysfunction, or susceptibility to stroke. Mol Cell Biol 24(9):4075–4082. 10. Hashmi S, Britton C, Liu J, Guiliano DB, Oksov Y and Lustigman S (2002) Cathepsin L is essential for embryogenesis and development of Caenorhabditis elegans. J Biol Chem 277(5):3477–3486. 11. Hashmi S, Zhang J, Oksov Y and Lustigman S (2004) The Caenorhabditis elegans cathepsin Z-like cysteine protease, Ce-CPZ-1, has a multifunctional role during the worms’ development. J Biol Chem 279(7):6035–6045.

ch07

Jan. 10, 2007

8:57

SPI-B429


8 Drosophila Serpins: Regulatory Cascades in Innate Immunity and Morphogenesis David Gubb, Andrew Robertson, Tim Dafforn, Laurent Troxler and Jean-Marc Reichhart

1. Introduction The serpins form a divergent group of proteins, many of which control the activity of proteolytic cascades. Serpins were first recognized as a discrete family of protease inhibitors in humans,1 where they regulate many biological processes including the blood-coagulation, fibrinolysis, complementactivation and inflammatory pathways.2,3 The serpins share a common core structure and have been found in all types of organisms, although rarely in prokaryotes. In mammals, serpins are found intracellularly as well as circulating in the blood plasma, often at very high concentrations. In addition, a related group of proteins has been identified, which retain the serpin fold while lacking inhibitory activity. This group of “non-inhibitory serpins” includes proteins involved in chromosome condensation,4 tumour suppression,5 hormone transport,3 molecular chaperones,6,7 and storage proteins.8 In many cases, it remains unclear what the functional requirement for a serpin fold is in the activities of these non-inhibitory molecules. The serpin/protease inhibitory mechanism involves cleavage of the reactive center loop (RCL), which extends outwards from the “hinge region” of the serpin core and presents itself as bait for the target protease (see Chap. 2). Following cleavage, the serpin and protease form a covalently linked complex, with the cleaved RCL inserted within β-sheet A of the 207

ch08

Jan. 10, 2007

8:57

SPI-B429

208


D. Gubb et al.

serpin. During this process, the protease core is distorted, hydrophobic residues are exposed and the complex is targeted for destruction.9 The turnover of both protease and serpin is rapid, so that reduction in serpin activity can give explosive activation of downstream proteolytic cascades. However, the metabolic cost of this mechanism is high and it is subjected to two further constraints. Firstly, serpins have an inherent tendency to form inactive polymers, with the RCL of one serpin inserted within β-sheet A of the next (Chap. 2). Secondly, the target specificity of serpins depends almost entirely on the amino acid sequence of the exposed RCL, in particular, the P1 and P1 residues. Both of these constraints can be modified by mutations and the resultant “serpinopathies” form a major group of human genetic diseases. The most common serpin-polymer disease in Northern European Caucasian populations is associated with Z-variant α1 -antitrypsin (Z-ATT) polymers in the liver.10,11 Homozygous individuals are susceptible to cirrhosis, particularly during neonatal life. Surviving adults develop emphysema, as the reduced ATT activity is insufficient to protect lung tissue from proteolysis by elastases activated during the inflammatory response. On the other hand, the serpinopathy associated with Pittsburgh-ATT results from altered target protease-specificity. In this variant, a single amino-acid transition (Met → Arg) within the RCL converts ATT from an elastase inhibitor to a thrombin inhibitor, giving a hemorrhagic pathology.12 These two examples illustrate how the gain of function mutations can lead to disease. In general, the broad specificity of serpin/protease interactions means that in the absence of one serpin proteases remain inhibited by the activity of a serpin with a similar inhibitory profile. Due to this genetic redundancy, null mutations of serpins rarely give a mutant phenotype. The broad target specificity also makes it difficult to assign individual serpins to their target protease(s) reliably on the basis of in vitro measurements. Despite this observation, there are several serpin-activated pathways which respond much more specifically under physiological conditions than biochemical assays would suggest. Arthropod serpins have received much less attention than their vertebrate homologues. The family includes active inhibitors found at high concentrations in extracellular hemolymph (the insect blood fluid).13 In

ch08

Jan. 10, 2007

8:57

SPI-B429


Regulatory Cascades in Innate Immunity and Morphogenesis

209

addition, presumptive intracellular inhibitors and non-inhibitory serpins have been identified by insect genome sequencing projects. In particular, the Drosophila genome encodes 29 serpin genes14 and around 211 chymotrypsin-fold serine proteases15 (compared with the human compliment of 37 serpins and ∼100 chymotrypsin-fold serine proteases). The serpin transcripts tend to be grouped within the fly genome, with 17 of the 29 organized into six gene clusters, while the remaining 12 are distributed singly. The properties of these serpins are summarized in Table 1; the nomenclature follows the FlyBase rules which reflect chromosomal location (e.g. to chromosome band 42D) rather than the serpin clade classifications. It is notable that the serpins within a gene cluster tend not to be closely homologous (Fig. 1). The serpin gene clusters in 77B and 88E presumably originate from recent duplication events, but the 28D and 43A clusters are less related and the serpins of the 42D cluster share very little homology with each other. The wide variance within gene clusters implies that the clade classifications reflect functional constrains more strongly than the long-term evolutionary relationships within the serpin family. The similar excess of proteases to serpins in Drosophila and humans suggests broadly comparable levels of genetic redundancy in serpin/protease interactions. Consistent with this view, very few mutant phenotypes have been linked to serpin genes in Drosophila; again, the lack of function of individual serpins is presumably being complemented by family members with similar inhibitory profiles.

2. Physiological Functions of Drosophila Serpins In recent years, the identification of serpins involved in the innate immune response and morphogenesis has begun to give a more detailed picture of these biological pathways. In two cases, Spn43Ac and Spn27A, null mutations do give a visible mutant phenotype, allowing their physiological functions to be identified.16–19,23,24 In a further case, Spn42Da, overexpression gives the same phenotype as that identified with the lack of function of its predicted target protease28 (see below). In general, however, the lack of mutant phenotypes associated with particular serpin transcripts has left the serpin biology field in Drosophila with a “black-box” of unassigned

ch08

Jan. 10, 2007

8:57

SPI-B429


210

D. Gubb et al. Table 1 Summary of Drosophila melanogaster serpin transcripts.

Name Spn27A Spn28D Spn28F Spn38F Spn42Da Spn42Da Spn42Da Spn42Da Spn42Da Spn42Dd Spn42De Spn43Aa Spn43Ac Spn77Ba Spn88Ea Spn88Eb

Synonym Chromosomal CG Swiss-Prot Size location number accession (AA)

P1/P1

Signal peptide

Reference

27A 28D6 28F5 38F1 42D6 42D6 42D6 42D6 42D6 42D6 42D6 43A1.2 43A1.2 77B4 88E3 88E3

11331 7219 8137 9334 9453-B 9453-C 9453-E/F 9453-D 9453-I 9456 9460 12172 1857 6680 18525 6687

Q9V3N1 Q9VLU4 Q9VLQ7 Q9U1I6 Q917G6 Q8TOM5 Q8MM49 Q8MPN8 Q8MPN7 Q9U1I8 Q8SZF4 Q8MQZ7 Q9U5W7 Q81GD7 Q9U1I4 Q9VFC1

447 536 375 372 424 418 411 411 406 372 404 390 476 450 427 426

K/F S/G Y/S K/S R/A S/M S/M S/L V/A R/A E/S M/S L/S K/A S/A S/S

ATG/NGN 16–19 CCE/SEL RFA/DDF 20 SCR/FTD 20 HTA/DVT 21, 22 HTA/DVT HTA/DVT HTA/DVT HTA/DVT VAC/QTS 20 MAN/TLN ISA/PEG 23 FAQ/ELI 23–26, 57 SAD/PQN ALA/GLC 20 1AA/LDK

Spn28B Spn28Da Spn42Da Spn4-A Spn42Da Sp4-G Spn42Da Spn4-H/K Spn42Da Spn4-J Spn42Dd Spn42Dc Spn55B sp6

28B3 28D2 42D6 42D6 42D6 42D6 42D6 42D6 55B7

6717 31902 9453-A 9453-G 9453-H/K 9453-J 9454 49455 10913

Q9VLZ8 Q8IPH2 Q9I7G5 Q8MPN6 Q8MM39 Q8MPN5 Q9V990 Q9V989 Q9V3L3

397 416 392 379 379 374 351 403 374

K/K L/S R/A S/L S/M V/A K/G M/M R/M

Cell autonomous

Spn43Ab Spn43Ad Spn53F Spn76A Acp76A Spn77Bc Spn85F Spn100A

43A1.2 43A1.2 53F12 75F5 77B4 85F5 100A2

1865 1859 10956 3801 6289 12807 1342

Q9U5W8 Q8SYY7 Q9V7Y9 Q9VVW1 Q9VPH9 Q9VH46 Q9VA48

393 NonSQS/KTV 23 407 inhibitory IAL/TLP 362 RCL LGI/SRY 15 386 IQQ/NVS 20, 22, 27 416 RAH/ICA 555 DAH/YIQ 649 AAI/GQG

28D2 31A2 47C7 74F9 77B4

33121 4804 7722 32203 6663

Q9VLV3 Q9VL44 Q9V5S3 Q81QT8 Q9VWB4

309 NonCell 382 inhibitory autonomous 382 RCL 456 362

Spn28Db Spn31A Spn47C Spn75F Spn77Bb

sp2 sp3 Spn4-B Spn4-C Spn4-E/F Spn4-D Snp4-I sp1

necrotic sp5

20

The table is divided into four sections: extracellular inhibitors (with a secretion signal peptide and inhibitory RCL and hinge region), intracellular inhibitors (lacking signal peptide), extracellular serpinfold proteins (with non-inhibitory RCL and hinge region), and intracellular serpin-fold proteins (with non-inhibitory RCL + hinge region and lacking signal peptide). The Drosophila serpin nomenclature reflects chromosomal location to band (e.g.: Spn42D, with adjacent transcripts in a cluster labeled a, b, c, etc. Alternatively spliced transcripts are designated A, B, C, etc. and the different Swiss-Prot Accessions of the individual cDNA sequences are given).

ch08

Jan. 10, 2007

8:57

SPI-B429



211

Fig. 1. Evolutionary tree of Drosophila serpins. Sequences were aligned against the α1 -antitrypsin scaffold using ClustalW. Subsequent assembly into a majority consensus minimum evolution bootstrap tree was made using the MEGA3 software (Kumar). While some genes within the Spn28D, Spn77B, and Spn88E gene clusters are closely related, the Spn28Dc and Spn77Ba transcripts are more diverged. Within the Spn42D, Spn43A clusters most of the transcripts have diverged to a similar degree as non-clustered transcripts: so that little evidence remains of the origin of these clusters by duplication of adjacent transcripts.

ch08

Jan. 10, 2007

8:57

SPI-B429

212


D. Gubb et al.

physiological functions. While there has been no systematic study of serpin genes, in several cases such as Spn43Aa, Spn43Ab and Spn43Ad, complete genetic deletion of their transcripts gives no mutant phenotype. As in mammals, the lack of individual serpin transcripts mostly causes little or no physiological response.

2.1. Serpins and the innate immune response The innate immune responses of flies and mammals share many features in common. In Drosophila, immune-challenge by fungi, or Gram-positive bacteria, induces the synthesis of antibiotic peptides in the fat body, the insect homologue of the liver. In both cases, signaling is through a Toll receptor and the intracellular components of these signaling pathways proceed through homologous proteins to activate the nuclear transcription factor NF-κB (or its Drosophila homologue, Dif ).29 The extracellular activation pathways, however, differ significantly in insects and mammals. In mammals, Tolllike-receptors (TLRs) recognize pathenogenic determinants directly; TLR3 binds viral dsRNA, TLR-4 binds lipopolysaccaride, and TLR-5 binds flagellin (Akira, 2003). In flies, by contrast, the Toll receptor is activated by binding the Spaetzle ligand, a cystein-knot cytokine (Fig. 2). The active ligand is cleaved from its inactive precursor,31 via a proteolytic cascade resembling that of the mammalian acute-phase response. The serpin that controls this response, Necrotic (Nec, synonym: Spn43Ac) was identified as one of a cluster of serpin transcripts mapping to the chromosome band 43A.23 Necrotic resembles the human acute phase serpins antichymotrypsin and antithrombin, and carries the same P1–P1 scissile bond as antichymotrypsin (Leu–Ser). In addition, Nec has a long, glutamine-rich, Nterminal extension which is unique in the serpin family. Expression of the Nec-N peptide (lacking the N-terminal extension) in E. coli gives a correctly folded serpin with broad specificity for elastase, thrombin, and chymotrypsin-like proteases.32 Despite this broad specificity, the necrotic phenotype corresponds to the recessive lack of function of the nec transcript.25 Given this genetic result, it would be expected that the Nec inhibitor might have a high affinity for a single target protease. Furthermore, this expectation is strengthened by the identification of the downstream

ch08

Jan. 10, 2007

8:57

SPI-B429



213

Fig. 2. Serpin-controlled proteolytic cascades in the Toll and PPO signaling pathways. Both the embryonic dorso-ventral polarity pathway and the adult innate immune response are activated by binding of the cleaved cytokine, Spaetzle, to the Toll receptor. The two proteolytic cascades that activate Toll are controlled by different serpins, however, and involve different serine proteases. In the innate immune response, Spn43Ac inactivates a cascade, which includes Persephone and Spaezle processing enzyme (Spe).30 During embryonic morpogenesis, Spn27 inactivates the Gastulation defective, Snake and Easter proteases, the last of which, Easter, cleaves the Spaetzle pro-cytokine to its active form.

serine protease, Persephone (Psh), in a suppressor screen to rescue the lethality of nec mutations.33 In psh; nec double-mutants, all aspects of the necrotic phenotype are suppressed. In the whole organism, therefore neither Nec, nor Psh, is redundant. Taken together, these results, exemplify a central question in serpin biology: how can inhibitors with broad target specificity under in vitro conditions control discrete physiological responses? An important component of the antimicrobial response in insects involves melanization of the exposed surfaces of pathogens, which is mediated via the phenoloxidase (PO) pathway. The activation of this pathway

ch08

Jan. 10, 2007

8:57

SPI-B429

214


D. Gubb et al.

is controlled by Spn27A.16,18 As with nec, deletion of the Spn27A transcript gives a recessive phenotype: both larvae and adults show spontaneous melanization of internal tissue associated with high levels of expression of PO in the hemolymph. This phenotype differs from that of nec mutants, which display cellular necrosis and melanotic patches restricted to the barrier epithelium of the body wall. It is not clear how the Toll and PO signaling pathways are connected in the adult immune response. Furthermore, in addition to its role in the immune response, Spn27A has an embryonic function. After fertilization in Drosophila, the formation of the dorso-ventral axis is mediated by a cascade of three serine proteases: Gastrulation defective, Snake and Easter.34 In this process, Spn27A inhibits Easter, the last protease in the cascade, which cleaves the Toll ligand, Spaetzle to activate signaling. In eggs from Spn27A mutant females, the uniform activation of these proteases gives ventralized embryos. It is clear, therefore, that Spn27A targets (at least) two different serine proteases: Easter in the embryo and an unidentified PO-activating protease in the adult.

2.2. The Spn42D gene cluster A compact stretch of 12.4 kb DNA in chromosomal region 42D contains five adjacent serpin transcription units (Spn42Da, Spn42Db, Spn42Dc, Spn42Dd and Spn42De). This gene cluster must represent a series of duplication-divergence events, but only Spn42Da and Spn42Db encode closely related serpins, the remainder being widely diverged (Fig. 1). The Spn42D gene cluster encodes far more than five serpin activities; however, as the Spn42Da transcription unit (synonym Sp4) has alternatively spliced 5 and 3 exons.21 This complex genetic structure allows for 11 different protein isoforms, with four alternative RCL baits and N-terminal sequences, with or without a signal peptide. The isoform encoded by one splice-variant, Spn42DaA, carries a furin cleavage site (RRKR/) and an endoplasmic reticulum retention signal (HDEL). Three groups have shown that Spn42DaA is a potent inhibitor of mammalian furin, a pro-protein convertase (PC) required for protein maturation along the secretory pathway.28,35,36 No lack of function phenotype has been described, but overexpression would be expected to mimic the loss of function of the target protease and cause

ch08

Jan. 10, 2007

8:57

SPI-B429



215

abnormal maturation of several neuropeptides.28 This prediction has been supported by the observation that the phenotype of Spn42Da overexpression mimics that of hypomorphic mutations in the Drosophila PC2 homologue, amontillado.37,38 Intriguingly, the Spn42Da furin inhibitor is made both with (Spn42DaB) and without (Spn42DaA), an N-terminal secretion peptide. The other three RCL forms encode the putative P1-P1 sites S/M, S/L, and V/A, indicating that they target different ranges of proteases.

2.3. The role of the serpin core and alternatively spliced serpin transcripts The major function of the core of an inhibitory serpin is to deliver a potential energy store held in its metastable, stressed conformation (S). Following RCL cleavage, the serpin adopts a relaxed conformation (R), distorts the active site of the protease and traps it in a covalent serpin/protease complex. This is quite different from the mechanism of Kunitz-type inhibitors, in which a lock and key fit between inhibitor and protease gives very high specificity. The serpin core is essential for inhibitory function, but the functional constraints of the S → R transition on the serpin fold leave little potential for the core structure to contribute to target specificity. As discussed above, this conclusion is supported by analysis of the mutations underlying human serpinopathies and the genetic redundancy in serpin/protease interactions. The finding of chimeric serpins, with different RCL loops attached to the same serpin core, confirms this conclusion. Chimeric serpins were first described in Bombyx mori,39 although the underlying genetic mechanism was subsequently identified in Manduca sexta. This species has a serpin gene with 12 alternative forms of the exon encoding the RCL.40,41 Alternatively-spliced RCL exons have since been identified in Drosophila melanogaster, Caenorhabditis elegans,21 and Anopheles gambiae.42 Chimeric serpins have yet to be described in mammals, but it will be surprising if no examples are eventually found. These alternatively-spliced transcripts represent an elegant mechanism for generating multiple inhibitory activities. More critically, the existence of RCL

ch08

Jan. 10, 2007

8:57

SPI-B429

216


D. Gubb et al.

chimeras demonstrates that the same core serpin can drive inhibitory baits targeted to the different families of protease in a non-pathological context.

2.4. Non-inhibitory serpins: Spn28Db, Spn31A, Spn43Ab, Spn43Ad, Spn47C, Spn53F, Spn75F, Spn76A, Spn77Bb, Spn77Bc, Spn85F, and Spn100A The 12 putative serpin transcripts in the Drosophila genome with hinge regions and RCLs that do not correspond with inhibitory serpins are listed in Table 1. Little is known about the functions of these proteins. The best studied of the non-inhibitory serpins, Spn76A, is a male accessory gland protein (synonym: Acp76A) that is transferred to the female genital tract during mating.27 Spn76A has been postulated to have a role in regulating proteolysis in seminal fluid,20 but lacks the alanine-rich hinge region typical of an inhibitory serpin. However, Spn28F does have a hinge region typical of an inhibitory serpin and it is also a component of seminal fluid,22 so Spn28F may well have the functions attributed to Spn76A. The Spn43Ab gene encodes a strongly basic peptide with a predicted iso-electric point of 10.0. This might suggest a nucleic acid binding function, similar to that of the vertebrate chromatin inactivating serpin MENT.4 Homozygous deletion of the Spn43Ab gene, however, does not modify the phenotype of the wm4 allele, a sensitive assay of heterochromatinization in Drosophila.23 Thus, while nucleic acid binding remains a possibility, there is no indication of a role for Spn43Ab in chromosome condensation. Assigning functions to the putative non-inhibitory serpins in Drosophila is a task that has only just begun.

3. Structure and Function of the Nec serpin, Spn43Ac The most comprehensive biochemical characterization of a Drosophila serpin to date is that of Nec.32,43 Nec is a broad-range inhibitor, with particular affinity for elastase- and chymotrypsin-like proteases (giving complete protease inhibition of Cathepsin G and Neutrophil Elastase at

ch08

Jan. 10, 2007

8:57

SPI-B429



217

a serpin to enzyme ratio of 1.1 and 1.2, respectively). These proteases cleave non-polar substrates, making the P1 leucine of Nec a good match. Conversely, Nec shows virtually no inhibition of proteases that preferentially cut positively charged substrates such as thrombin and trypsin, which retain more than 50% activity at serpin to enzyme ratios exceeding 100:1. 3.1. In silico prediction of the putative target protease of Nec The specificity of chymotrypsin-fold serine proteases is mainly dependent on the structure of the S1 substrate-binding pocket.44,45 This pocket is easily identifiable through sequence alignments and structural modeling. Of the 211 predicted serine proteases in the Drosophila genome, approximately 30 have structurally similar substrate-binding pockets to chymotrypsin or elastase. A further 90 proteases show structural similarity to thrombin and trypsin and are therefore unlikely targets for Nec. Thrombin and trypsin-type proteases carry a negatively charged substrate binding domain with a conserved aspartate residue at the mouth of the specificity pocket. Analysis of the Psh sequence predicts a substrate binding pocket unlike neutrophil elastase, but very similar to that of thrombin and trypsin (Fig. 3). Therefore, although Psh is clearly a downstream protease in the Nec → Toll pathway, the high affinity of Nec for elastases, together with its lack of inhibition of thrombin and trypsin, suggests that Psh is unlikely to be the direct target of Nec. In principle, the specificity of the Toll-signaling pathway might be modified through the Necrotic N-terminal extension. Although N-terminal extensions are an unusual feature of serpin inhibitors, several are known to modify target-specificity, including heparin cofactor II, α1 -antitrypsin, angiotensin and C1-protease inhibitor.46–49 In this context, it is striking that the Nec N-terminal peptide is cleaved following immunological challenge.43 This cleavage is Psh-dependent and does not occur in psh mutant flies, although it remains uncertain whether the N-terminal cleavage of Nec is directly mediated by Psh. When tested against a large panel of mammalian serine proteases, the specificity of the N-terminally deleted serpin, Nec-N, remains unchanged when compared with the specificity

ch08

Jan. 10, 2007

8:57

SPI-B429

218


D. Gubb et al.

Fig. 3. Structural alignment of Psh specicity-pocket. Alignment of the S1 residues of Psh (yellow) with trypsin (left, green) and neutrophil elastase (right, purple) indicates that the specificity pocket of Psh resembles that of trypsin more closely than that of neutrophil elastase.

of the full-length Nec inhibitor, with the exception of Nec-N that has 13-fold increase in specificity for porcine pancreatic elastase. This result suggests a feedback suppression mechanism to downregulate the Toll pathway. Such a mechanism is clearly required since the constitutive activation of the Toll pathway that occurs in necrotic mutant flies is lethal within a few days. However, these experiments provide no evidence that the Nterminal cleavage of Necrotic gives the reduction in specificity that would allow its target protease to escape inhibition; at least not under in vitro conditions. An additional feature of the Nec serpin was identified by protein modeling using the antithrombin structure as a template. This domain consists of three Lys residues in the D-helix and single Lys and Arg residues in the A-helix (Fig. 4) and is homologous to the heparin-binding domain of antithrombin50 and heparin co-factor II. Heparan-sulfate-binding by these domains activates and localizes the serpin to the internal surface of the blood vessels, in addition to stabilizing protease-serpin docking. We have found that substituting the three charged Lys residues of the D-helix with Glu residues inactivates the Nec protein, both in vivo and in vitro (unpublished results). Biochemical analysis, however, indicates that neither the original Nec protein, nor the substituted NecK175Q,K177Q,K178Q protein binds

ch08

Jan. 10, 2007

8:57

SPI-B429



219

Fig. 4. Alignment of Nec with the sugar-binding groove of antithrombin. Structural alignment of antithrombin (red) and Nec (blue) identifies a homologous domain carrying basic residues at the C-terminal domain of the D-helix. In antithrombin, this domain forms a heparin sulfate-binding groove. The homologous structure in the Nec protein does not bind heparin in vitro, but is required for inhibitory activity in vivo. The β-sheet A stands of Nec are shown as gray ribbons to aid in orientation.

heparin. Although we have identified an essential domain within the Nec core serpin, the function of this domain differs from that in the antithrombin serpin family.

4. Drosophila Melanogaster as a Model of Human Serpinopathies In recent years, fly models have been developed for a range of human protein conformational diseases, in particular, the neurodegenerative, Machado–Joseph’s, Huntington’s, Parkinson’, and Alzheimer’s51–56 diseases (see also http://superfly.ucsd.edu/homophilia). The success of these models suggests that many of the physiological mechanisms of dealing with misfolded peptides are conserved between flies and mammals.

ch08

Jan. 10, 2007

8:57

SPI-B429


220

D. Gubb et al.

4.1. Homologous mutations in human serpins and Necrotic Of the 12 sequenced nec mutations, 5 correspond to premature stop codons, confirming that the phenotype results from the lack of gene function.23,26 The remaining seven mutations include five that carry homologous amino acid substitutions to those associated with human serpinopathies, one results from a Gly → Ser transition in the conserved P15 position of the RCL (nec16 ) and one results from a two amino acid deletion (nec1 ), which apparently destabilizes β-sheet A (Fig. 5).26,57 Of particular interest are the temperatute-sensitive serpins, Nec9 and 20 Nec , which carry the homologous Glu → Lys substitution to Z α1 antitrypsin. In humans, Z-ATT forms stable polymers under physiological conditions and the polymerization rate is temperature-dependent in vitro, being accelerated at 41◦ C.10 This correlates with the role of episodic fevers in the pathogenesis of neonatal Z-ATT liver disease.58 Similarly,

nec 10

nec 16 nec 9 & 20

nec 7 & 22

Siiyama nec 1

Fig. 5. Location of nec mutations on the serpin molecule. All the amino acid substitutions affect residues that are conserved within the serpin super-family. The nec9 and nec20 carry homologous Glu → Lys substitutions in the hinge-region as human Z-variant ATT; while nec7 and nec22 carry a Gly → Ser substitution in the homologous position as a polymerogenic variant of antithrombin.

ch08

Jan. 10, 2007

8:57

SPI-B429



221

the Gly386 → Arg substitution in antithrombin (homologous to nec7 and nec22 ) gives temperature-dependent polymerization and is associated with thrombosis triggered by fevers.26 The temperature-dependent polymerization of serpins in vivo can be studied more readily in insect than in mammalian models, simply by culturing insects at different temperatures. In fact, nec9 flies show a temperature-dependent phenotype that reduces longevity of adult flies cultured at 29◦ C. Moreover, engineering NecS408F , the equivalent substitution to that which is responsible for the Siiyama variant α1 -antitrypsin serpinopathy, gives a dominant temperature sensitive nec phenotype.57 These results underline the importance in controlling body temperature in the clinical management of serpinopathies, which as come from anecdotal observations.26 It is striking that the nec alleles that result from amino-acid substitutions are all in conserved residues and, with the exception of nec16 , correspond to human variants associated with known serpinopathies. In each case, the nec phenotype results from the lack of function, equivalent to the emphysema pathology of Z-ATT, rather than the liver pathology of Z-ATT. In this context, the nec16 substitution of a conserved amino-acid in the RCL gives a null mutation, which would not cause an associated pathology in most human serpins. Therefore, with the polymerogenic nec alleles as with the null mutations, it is the reduced Nec activity in the hemolymph that leads to constitutive activation of the Toll pathway. One explanation for the absence of a gain of function nec phenotype might be that the insect fat body is less sensitive to serpin polymer toxicity than the mammalian liver. In general, the neurodegenerative disease models indicate that the fly is not refractory to toxic peptide accumulation. A more intriguing possibility is that the Nec core serpin does not form long-chain polymers in vivo. Polymerogenic nec alleles do form high molecular weight material and, like Z-ATT, they fail to show a S → R transition on Transverse Urea Gradient Gels, but a clear Nec polymer ladder has not been observed.57 Like Hsp47,7 it may be that Nec forms closed loop oligomers. This suggests the possibility of engineering chimeric human serpins with the RCL of one serpin attached to a core of another that does not form long chain polymers (see Refs. 59 and 60). Such a chimeric protein would have biological activity as an inhibitor, without contributing to the gain of function polymer pathology.

ch08

Jan. 10, 2007

8:57

SPI-B429


222

D. Gubb et al.

5. Conclusions The serpins form a group of proteins that are highly diverged at the amino acid sequence level, while retaining a conserved fold within which a few critical residues are conserved strongly. A primary function of the serpin fold is to store potential energy within a metastable configuration. The release of this potential energy drives the formation of an irreversible serpin/protease complex, in which the denatured proteinase moiety is targeted for destruction. Additional uses of the serpin-fold within non-inhibitory molecules may also involve the S → R transition, or exploit the tendency of this fold structure to generate homopolymers. In general, the study of serpins in arthropods, particularly Drosophila has only scratched the surface of the complexity of biological processes mediated by these molecules. The contribution that a model organism can bring to the field, in conjunction with the sophisticated biochemical tools that have been developed in mammalian systems, is the ease of manipulation of genetic functions. This genetic technology will facilitate the tracing of serpin-controlled signaling pathways and the cross-talk between related pathways. More importantly, we have the tools in hand to knock out multiple genetic functions and to approach an understanding of the specificity of serpin-regulated responses under physiological conditions.To date, the results have shown a remarkable conservation between flies and humans, particularly within the innate immune response. The recovery of a series of amino-acid substitutions in the Necrotic serpin that are homologous to substitutions underlying a range of human serpinopathies, implies that critical physiological mechanisms are also conserved. Within living organism, the mechanisms for localizing inhibitory serpins in the extracellular environment, trafficking across membranes and partitioning to sub-cellular compartments are critical in maintaining the serpin to target proteinase balance. From this perspective, it is not surprising that the serpinopathies form a major group of human genetic diseases. Rather, it is the mechanisms that permit most individuals to remain free from serpinopathies that are incompletely understood. The fly model will provide us with an insight into how the control of such processes contributes to the modulation of serpin-regulated pathways. Although the details may differ, these mechanisms are likely to be highly conserved within multicellular organisms.

ch08

Jan. 10, 2007

8:57

SPI-B429



223

References 1. Carrell RW and Travis J (1986) Alpha 1-antitrypsin and the serpins: Variation and countervariation. TIBS 10:20–24. 2. Boswell DR and Carrell RW (1988) Genetic engineering and the serpins. Bioessays 8:83–87. 3. Potempa J, Korzus E and Travis J (1994) The serpin superfamily of proteinase inhibitors: structure, function, and regulation. J Biol Chem 269:15957–15960. 4. Grigoryev SA, Bednar J and Woodcock CL (1999) MENT, a heterochromatin protein that mediates higher order chromatin folding, is a new serpin family member. J Biol Chem 274:5626–5636. 5. Liu T, Pemberton PA and Robertson AD (1999) Three-state unfolding and self-association of maspin, a tumor-suppressing serpin. J Biol Chem 274: 29628–29632. 6. Clarke EP, Cates GA, Ball EH and Sanwal BD (1991) A collagen-binding protein in the endoplasmic reticulum of myoblasts exhibits relationship with serine protease inhibitors. J Biol Chem 266:17230–17235. 7. Dafforn TR, Della M and Miller AD (2001) The molecular interactions of heat shock protein 47 (Hsp47) and their implications for collagen biosynthesis. J Biol Chem 276:49310–49319. 8. Huntington JA and Stein PE (2001) Structure and properties of ovalbumin. J Chromatogr B Biomed Sci Appl 756:189–198. 9. Huntington JA, Read RJ and Carrell RW (2000) Structure of a serpin-protease complex shows inhibition by deformation. Nature 407:923–926. 10. Lomas DA, Evans DL, Finch JT and Carrell RW (1992) The mechanism of Z alpha 1-antitrypsin accumulation in the liver. Nature 357:605–607. 11. Lomas DA and Mahadeva R (2002) Alpha 1-antitrypsin polymerization and the serpinopathies: Pathobiology and prospects for therapy. J Clin Invest 110:1585–1590. 12. Owen MC, Brennan SO, Lewis JH and Carrell RW (1983) Mutation of antitrypsin to antithrombin. Alpha 1-antitrypsin Pittsburgh (358 Met leads to Arg), a fatal bleeding disorder. N Engl J Med 309:694–698. 13. Kanost MR (1999) Serine proteinase inhibitors in arthropod immunity. Dev Comp Immunol 23:291–301. 14. Reichhart JM (2005) Tip of another iceberg: Drosophila serpins. Trends Cell Biol 15. Ross J, Jiang H, Kanost MR and Wang Y (2003) Serine proteases and their homologs in the Drosophila melanogaster genome: An initial analysis of sequence conservation and phylogenetic relationships. Gene 304:117–131.

ch08

Jan. 10, 2007

8:57

SPI-B429

224


D. Gubb et al.

16. Ligoxygakis P, Pelte N, Ji C, Leclerc V, Duvic B, Belvin M, Jiang H, Hoffmann JA and Reichhart JM (2002) A serpin mutant links Toll activation to melanization in the host defence of Drosophila. Embo J 21:6330–6337. 17. Ligoxygakis P, Roth S and Reichhart JM (2003) A serpin regulates dorsalventral axis formation in the Drosophila embryo. Curr Biol 13:2097–2102. 18. De Gregorio E, Han SJ, Lee WJ, Baek MJ, Osaki T, Kawabata S, Lee BL, Iwanaga S, Lemaitre B and Brey PT (2002) An immune-responsive serpin regulates the melanization cascade in Drosophila. Dev Cell 3:581–592. 19. Hashimoto C, Kim DR, Weiss LA, Miller JW and Morisato D (2003) Spatial regulation of developmental signaling by a serpin. Dev Cell 5:945–950. 20. Wolfner MF, Harada HA, Bertram MJ, Stelick TJ, Kraus KW, Kalb JM, Lung YO, Neubaum DM, Park M and Tram U (1997) New genes for male accessory gland proteins in Drosophila melanogaster. Insect Biochem Mol Biol 27:825–834. 21. Kruger O, Ladewig J, Koster K and Ragg H (2002) Widespread occurrence of serpin genes with multiple reactive centre-containing exon cassettes in insects and nematodes. Gene 293:97–105. 22. Han J, Zhang H, Min G, Kemler D and Hashimoto CA (2000) Novel Drosophila serpin that inhibits serine proteases. FEBS Lett 468:194–198. 23. Green C, Levashina E, McKimmie C, Dafforn T, Reichhart JM and Gubb D (2000) The necrotic gene in Drosophila corresponds to one of a cluster of three serpin transcripts mapping at 43A1.2. Genetics 156:1117–1127. 24. Levashina EA, Langley E, Green C, Gubb D, Ashburner M, Hoffmann JA and Reichhart JM (1999) Constitutive activation of toll-mediated antifungal defense in serpin-deficient Drosophila. Science 285:1917–1919. 25. Heitzler P, Coulson D, Saenz-Robles MT, Ashburner M, Roote J, Simpson P and Gubb D (1993) Genetic and cytogenetic analysis of the 43A-E region containing the segment polarity gene costa and the cellular polarity genes prickle and spiny-legs in Drosophila melanogaster. Genetics 135:105–115. 26. Carrell R and Corral J (2004) What can Drosophila tell us about serpins, thrombosis and dementia? Bioessays 26:1–5. 27. Coleman S, Drahn B, Petersen G, Stolorov J and Kraus KA (1995) Drosophila male accessory gland protein that is a member of the serpin superfamily of proteinase inhibitors is transferred to females during mating. Insect Biochem Mol Biol 25:203–207. 28. Osterwalder T, Kuhnen A, Leiserson WM, Kim YS and Keshishian H (2004) Drosophila serpin 4 functions as a neuroserpin-like inhibitor of subtilisin-like proprotein convertases. J Neurosci 24:5482–5491.

ch08

Jan. 10, 2007

8:57

SPI-B429



225

29. Lemaitre B, Reichhart JM and Hoffmann JA (1997) Drosophila host defense: Differential induction of antimicrobial peptide genes after infection by various classes of microorganisms. Proc Natl Acad Sci USA 94:14614–14619. 30. Jang IH, Chosa N, Kim SH, Nam HJ, Lemaitre B, Ochiai M, Kambris Z, Brun S, Hashimoto C, Ashida M, Brey PT and Lee WJ (2006) A spatzle-processing enzyme required for toll signaling activation in Drosophila innate immunity. Dev Cell 10:45–55. 31. DeLotto Y and DeLotto R (1998) Proteolytic processing of the Drosophila Spatzle protein by easter generates a dimeric NGF-like molecule with ventralising activity. Mech Dev 72:141–148. 32. Robertson AS, Belorgey D, Lilley KS, Lomas DA, Gubb D and Dafforn TR (2003) Characterization of the necrotic protein that regulates the Toll-mediated immune response in Drosophila. J Biol Chem 278:6175–6180. 33. Ligoxygakis P, Pelte N, Hoffmann JA and Reichhart JM (2002) Activation of Drosophila Toll during fungal infection by a blood serine protease. Science 297:114–116. 34. LeMosy EK, Hong CC and Hashimoto C (1999) Signal transduction by a protease cascade. Trends Cell Biol 9:102–107. 35. Oley M, Letzel MC and Ragg H (2004) Inhibition of furin by serpin Spn4A from Drosophila melanogaster. FEBS Lett 577:165–169. 36. Richer MJ, Keays CA, Waterhouse J, Minhas J, Hashimoto C and Jean F (2004) The Spn4 gene of Drosophila encodes a potent furin-directed secretory pathway serpin. Proc Natl Acad Sci USA 101:10560–10565. 37. Siekhaus DE and Fuller RS (1999) A role for amontillado, the Drosophila homolog of the neuropeptide precursor processing protease PC2, in triggering hatching behavior. J Neurosci 19:6942–6954. 38. Rayburn LY, Gooding HC, Choksi SP, Maloney D, Kidd AR, III, Siekhaus DE and Bender M (2003)Amontillado, the Drosophila homolog of the prohormone processing protease PC2, is required during embryogenesis and early larval development. Genetics 163:227–237. 39. Sasaki T (1991) Patchwork-structure serpins from silkworm (Bombyx mori) larval hemolymph. Eur J Biochem 202:255–261. 40. Jiang H, Wang Y, Huang Y, Mulnix AB, Kadel J, Cole K and Kanost MR (1996) Organization of serpin gene-1 from Manduca sexta. Evolution of a family of alternate exons encoding the reactive site loop. J Biol Chem 271: 28017–28023. 41. Jiang H, Wang Y and Kanost MR (1994) Mutually exclusive exon use and reactive center diversity in insect serpins. J Biol Chem 269:55–58.

ch08

Jan. 10, 2007

8:57

SPI-B429

226


D. Gubb et al.

42. Danielli A, Kafatos FC and Loukeris TG (2003) Cloning and characterization of four Anopheles gambiae serpin isoforms, differentially induced in the midgut by Plasmodium berghei invasion. J Biol Chem 278:4184–4193. 43. Pelte N, Robertson AS, Zou Z, Belorgey D, Dafforn TR, Jiang H, Lomas D, Reichhart JM and Gubb D (2006) Immune challenge induces N-terminal cleavage of the Drosophila serpin Necrotic. Insect Biochem Mol Biol 36:37–46. 44. Greer J (1990) Comparative modeling methods: Application to the family of the mammalian serine proteases. Proteins 7:317–334. 45. Perona JJ and Craik CS (1995) Structural basis of substrate specificity in the serine proteases. Protein Sci 4:337–360. 46. Van Deerlin VM and Tollefsen DM (1991) The N-terminal acidic domain of heparin cofactor II mediates the inhibition of alpha-thrombin in the presence of glycosaminoglycans. J Biol Chem 266:20223–20231. 47. Fitton HL, Skinner R, Dafforn TR, Jin L and Pike RN (1998) The N-terminal segment of antithrombin acts as a steric gate for the binding of heparin. Protein Sci 7:782–788. 48. Costa-Neto CM, Miyakawa AA, Oliveira L, Hjorth SA, Schwartz TW and PaivaAC (2000) Mutational analysis of the interaction of the N- and C-terminal ends of angiotensin II with the rat AT(1A) receptor. Br J Pharmacol 130:1263– 1268. 49. Schedin-Weiss S, Desai UR, Bock SC, Olson ST and Bjork I (2004) Roles of N-terminal region residues Lys11, Arg13, and Arg24 of antithrombin in heparin recognition and in promotion and stabilization of the heparin-induced conformational change. Biochemistry 43:675–683. 50. Huntington JA, Olson ST, Fan B and Gettins PG (1996) Mechanism of heparin activation of antithrombin. Evidence for reactive center loop preinsertion with expulsion upon heparin binding. Biochemistry 35:8495–8503. 51. Crowther DC, Kinghorn KJ, Miranda E, Page R, Curry JA, Duthie FA, Gubb DC and Lomas DA (2005) Intraneuronal Abeta, non-amyloid aggregates and neurodegeneration in a Drosophila model of Alzheimer’s disease. Neuroscience 132:123–135. 52. Kazemi-Esfarjani P and Benzer S (2000) Genetic suppression of polyglutamine toxicity in Drosophila. Science 287:1837–1840. 53. Marsh JL, Walker H, Theisen H, Zhu YZ, Fielder T, Purcell J and Thompson LM (2000) Expanded polyglutamine peptides alone are intrinsically cytotoxic and cause neurodegeneration in Drosophila. Hum Mol Genet 9:13–25. 54. Warrick JM, Chan HY, Gray-Board GL, Chai Y, Paulson HL and Bonini NM (1999) Suppression of polyglutamine-mediated neurodegeneration in Drosophila by the molecular chaperone HSP70. Nat Genet 23:425–428.

ch08

Jan. 10, 2007

8:57

SPI-B429



227

55. Warrick JM, Paulson HL, Gray-Board GL, Bui QT, Fischbeck KH, Pittman RN and Bonini NM (1998) Expanded polyglutamine protein forms nuclear inclusions and causes neural degeneration in Drosophila. Cell 93:939–949. 56. Crowther DC, Kinghorn KJ, Page R and Lomas DA (2004) Therapeutic targets from a Drosophila model of Alzheimer’s disease. Curr Opin Pharmacol 4:513–516. 57. Green C, Brown G, Dafforn TR, Reichhart JM, Morley T, Lomas DA and Gubb D (2003) Drosophila necrotic mutations mirror disease-associated variants of human serpins. Development 130:1473–1478. 58. Sveger T (1978) Alpha 1-antitrypsin deficiency in early childhood. Pediatrics 62:22–25. 59. Hopkins PC, Crowther DC, Carrell RW and Stone SR (1995) Development of a novel recombinant serpin with potential antithrombotic properties. J Biol Chem 270:11866–11871. 60. Djie MZ, Stone SR and Le Bonniec BF (1997) Intrinsic specificity of the reactive site loop of alpha 1-antitrypsin, alpha 1-antichymotrypsin, antithrombin III, and protease nexin I. J Biol Chem 272:16268–16273. 61. Okuyama E, Tachida H and Yamazaki T (1997) Molecular analysis of the intergenic region of the duplicated Amy genes of Drosophila melanogaster and Drosophila teissieri. J Mol Evol 45:32–42.

ch08

Jan. 10, 2007 8:57 SPI-B429



ch09

Jan. 10, 2007

8:57

SPI-B429


9 Serpins in a Lepidopteran Insect, Manduca sexta Michael R. Kanost

1. Introduction Insect plasma contains fairly high concentrations of serine protease inhibitors from several different sequence families, including serpins.1–3 The first insect serpins to be well characterized biochemically were from the silkworm, Bombyx mori. Sasaki and his group purified inhibitors of trypsin and chymotrypsin from Bombyx plasma and demonstrated in the late 1980s that these ∼45 kDa proteins were cleaved near their carboxyl-termini and formed SDS-stable complexes with proteases during inhibition reactions, 4–7 and they speculated that these proteins might be related to human serpins. Similar inhibitors were isolated from another lepidopteran insect, Manduca sexta,8,9 and amino acid sequences obtained by Edman degradation and deduced from cDNA cloning, indicated that the Manduca and Bombyx proteins were indeed serpins.6,8,10,11 Insect serpin sequences lack sufficient similarity to vertebrate serpins to permit assignment of orthologous serpins between insects and vertebrates.12 These lepidopteran serpins have been assigned to serpin clade K (insect), based on their alignment with sequences available in 2000.12,13 However, several insect serpins were orphans that did not fall in clade K,13 and considering the diversity of the Insecta, it is quite likely that when more insect serpin sequences are known, definition of additional clades may be needed. The large size of caterpillars such as Manduca, compared with most other insect species, facilitates 229

ch09

Jan. 10, 2007

8:57

SPI-B429

230


M. R. Kanost

collection of sufficient volumes of hemolymph (blood) for purification of plasma serpins, and this advantage continues to benefit the biochemical study of serpin function in insects.14–17

2. Manduca Serpin-1: 12 Serpins from 1 Gene When multiple cDNA clones for the first Manduca serpin8 were sequenced, it became apparent that they included variants, which differed only in the region encoding the carboxyl-terminal 35–40 residues of the protein. This region, which includes the reactive center loop (RCL) and two β strands in sheet B, is encoded by 12 alternative versions of the ninth exon in the serpin-1 gene18,19 (Fig. 1). Duplication and sequence divergence in the alternative exons during evolution, resulted in a gene that encodes serpins with 12 different reactive site sequences (Table 1), each with a unique inhibitory capability.19,22 Recombinant serpins expressed from each of the 12 cDNAs inhibit mammalian proteases consistent with predictions based on the serpin RCL sequences. For example, serpin-1A, with Arg at the P1 position inhibits trypsin and plasmin, whereas serpin-1B, with Ala at P1 inhibits elastase, and serpin-1K, with Tyr at P1 inhibits chymotrypsin and cathepsin G.22 Serpin-1 is constitutively expressed in fat body, a tissue with functions analogous to mammalian liver and adipose tissue, and in granular hemocytes. Its expression is negatively regulated by the steroid hormone 20-hydroxyecdysone, which causes cessation of serpin-1 expression as it triggers molting.23 From the relative abundance of different cDNA isoforms in fat body and hemocyte cDNA libraries, it appears that serpin-1F is preferentially expressed in fat body, whereas the 12 isoforms are approximately represented equally in hemocytes.18 Regulation of the mutually exclusive alternative exon splicing that must occur in serpin-1 expression is a fascinating topic that has not yet been investigated. Serpin genes with alternative exons encoding the RCL have been identified in other insects. These include the apparent orthologous genes for serpin-1 in silkworms10 and another lepidopteran, Mamestra configurata24 (each with three alternative exons), a serpin gene in the cat flea, Ctenocephalides felis (with 14 alternative exons), serpin-4 in Drosophila melanogaster25 (four alternative exons), and serpin-10 from the mosquito,

ch09

Jan. 10, 2007

8:57

SPI-B429


Serpins in a Lepidopteran Insect, Manduca sexta

231

A Exon

C

B

Fig. 1. Alternative exons in the Manduca serpin-1 gene encode variant RCL sequences. (A) Organization of the serpin-1 gene. Exons 1–8 encode the 5 -untranslated region, secretion signal peptide, and amino acid residues 1–336 or the mature protein. Each mRNA contains one of the 12 alternative forms of exon 9, which encodes the carboxyl-terminal ∼40 residues of the protein followed by a translation stop codon and then exon 10, which contains the 3 untranslated sequence.19 The entire gene is 20.5 kb long. (B) Structure of serpin-1K (1SEK) determined by X-ray crystallography.20 The sequence encoded by exon 9K is in red, including the P1 Tyr residue, whose sidechain is shown. (C) Structure of a Michaelis complex formed between Manduca serpin-1B (Ala353 Lys mutant) and rat trypsin (Ser105Ala mutant) (1K9O).21 The sequence encoded by exon 9B is in red, with the P1 Lys residue in green. The trypsin structure is shown in purple.

Anopheles gambiae26 (four alternative exons). In all of these genes, an intron is present at the same position as the intron at the 3 end of exon 8 in Manduca serpin-1, suggesting that an intron at this position existed in an ancient serpin gene in the lineage that gave rise to at least several insect

ch09

Jan. 10, 2007

8:57

SPI-B429


232

M. R. Kanost Table 1

Manduca serpin Serpin-1

Serpins from hemolymph of Manduca sexta.

AA residues

Tissue source

Secreted

Bacteriainduced

P4–P4

376–381

Fat body hemocytes

Yes

no

12 variants

No

Yes

FITRQARL IVPASLIL FIIESYSS RGIRPRPS RVVKKKFR IAVVDSID IVGITSIQ FITYVESI IVALSLEF LTDRCCSD ITTYSFHF GIAYLSAV IRSISFMA

Yes Yes

Yes Yes

IQNKFGED FSNRIGII

Yes

Yes

FADRISTP

Yes

Yes

FGFRSSRP

Yes

?

LEDRILGP

1A 1B 1C 1D 1E 1F 1G 1H 1I 1J 1K 1L Serpin-2

381

Serpin-3 Serpin-4

435 391

Serpin-5

379

Serpin-6

395

Serpin-7

384

Granular hemocytes Fat body Fat body hemocytes Fat body hemocytes Fat body hemocytes Fat body

orders. Alternative exons encoding RCL variants have also been identified in serpin genes of the nematode Caenorhabditis elegans.25 It thus appears that some insects and other invertebrate species have evolved this mechanism of duplication of a reactive site loop cassette to produce serpins with diverse activities. Structures for two Manduca serpin-1 variants have been solved by x-ray crystallography by the group of Elizabeth Goldsmith. Serpin-1K, a chymotrypsin inhibitor, has a typical serpin fold, and its RCL is in an extended conformation from residues P3 to P3 , ready to bind to a target protease active site.20 Serpin-1B has also been used for structural studies.

ch09

Jan. 10, 2007

8:57

SPI-B429



233

We mutated the P1 residue of serpin-1B from Ala to Lys, converting its inhibitory specificity from elastase to trypsin.27 The interaction of this serpin-1B (A353K) with rat trypsin was investigated by x-ray crystallography and by the biochemical study of additional serpin mutants designed based on examination of the crystal structure.21 Serpin-1B (A353K) complexed with trypsin S195A (in which the active site serine is changed to alanine) shares a nearly identical protein framework with serpin-1K but differs in the conformation of the RCL. The serpin-1B (A353K) RCL is shifted in comparison to serpin-1B, and the region from P3 to P13 forms extensive interactions with the surface of trypsin. Site-directed mutations in these interacting residues of serpin-1B (A353K), distant from the reactive site, reduced inhibitory activity. The Manduca serpin-rat trypsin complex is one of a few available structures of such an encounter complex between a serpin and a protease, and it is useful as a model for understanding the mechanism of mammalian serpin inhibitory activity.13,28,29

3. Manduca Serpins 2–7 Six additional Manduca serpins have been identified at this time (Table 1). As there is no Manduca genome project, these probably represent only a partial list of serpin genes in this species. For comparison, the genomes of Drosophila and Anopheles contain 28 and 14 serpin genes, respectively. Manduca serpin-2 lacks a secretion signal peptide and is expressed in cytoplasm of granular hemocytes.30 Serpin-2 synthesis is induced by bacterial challenge. It weakly inhibits human neutrophil cathepsin G, but its natural target protease is unknown. Perhaps, it functions to protect hemocytes from the proteases they produce31,32 as proposed for the human intracellular clade B serpins related to ovalbumin.33 Manduca serpins 3–6 are plasma proteins, which are expressed constitutively by fat body at lower levels than serpin-1, but their mRNAs and plasma concentration increase significantly in response to microbial infection, and thus they behave as acute phase proteins with potential roles in innate immune responses. Serpins 4–6 are also expressed by hemocytes. Serpin-3 contains an amino-terminal extension of ∼50 amino acid residues that is not present in other Manduca serpins.34 It is most similar and probably orthologous to a serpin from another moth, Hyphantria cunea35 and

ch09

Jan. 10, 2007

8:57

SPI-B429

234


M. R. Kanost

Drosophila serpin 27A,36,37 which also have amino-terminal extensions, and to Anopheles gambiae serpin-2 (accession XP_308845). Manduca serpins 4 and 517 are most similar to each other (36% identity) among available sequences and may represent a somewhat recent gene duplication within Lepidoptera, as Bombyx orthologs of serpin-4 (AAS68505) and serpin-5 (AAS68506) have been identified. Serpin-638,39 is most similar to Drosophila serpin-5 (AAF55138) and Anopheles serpin-9 (EAA08448). No alternative exons similar to those in serpin-1 are present in serpins 2–6. A cDNA for serpin-7 has recently been cloned, but the corresponding protein has not yet been characterized (Ochieng, Zhuang, Kanost, unpublished results). It is most similar (31% identity) to serpin-4 and is predicted to have an Arg residue at the P1 position of its reactive center.

4. Physiological Functions of Manduca Serpins The insect genomes sequenced so far contain very large numbers of serine protease genes, ∼200 in Drosophila40,41 and ∼300 in Anopheles.42 Some of them are certainly digestive enzymes expressed in the gut, but the majority may have other functions in insect biology.41 Nondigestive serine proteases in arthropods are known to function in cascade pathways, which regulate aspects of embryonic development, blood coagulation and innate immune responses.43 We are interested in identifying serine proteases that are present in hemolymph and which may be regulated by the plasma serpins in Manduca. At least 25 serine proteases are present in Manduca hemolymph,32 but their substrates and physiological roles are known for only a few. There are several pathways in arthropod hemolymph in which the presence of microorganisms triggers activation of a serine protease cascade (Fig. 2). A response to infection common to insects and other arthropods is the proteolytic activation of a prophenoloxidase (proPO) zymogen. Phenoloxidase then oxidizes diphenols related to dopamine to the corresponding quinones, which undergo further nonenzymatic reactions that result in synthesis and deposition of melanin on the surface of the invading pathogen. Three hemolymph proteases in Manduca are capable of precisely cleaving proPO to produce active phenoloxidase. They are

ch09

Jan. 10, 2007

8:57

SPI-B429



235

Fig. 2. Proteins in plasma or on the surface of hemocytes bind to microbial surface molecules such as lipopolysaccharide or peptidoglycan, which stimulates several types of cellular and humoral innate immune responses that involve serine protease cascades. The final protease in the pathway cleaves a substrate protein to produce an active form that aggregates to form fibers (hemolymph coagulation in horseshoe crabs),44 activates a phenolxodase zymogen,15,45 or releases a protein fragment that acts as a cytokine to cause hemocyte adhesion (plasmatocyte spreading peptide)46 or to stimulate synthesis of immune effector proteins (spätzle).47

called prophenoloxidase activating proteases (PAP)-1, -2, and -3.48–50 The Manduca PAPs also require as a cofactor the presence of a serine protease homolog (SPH), which lacks catalytic activity because the position normally occupied by the active site serine is changed to glycine.50–52 Both the PAPs and the SPHs have a structure in which one or two small amino-terminal regulatory domains called clip domains are followed by a carboxyl-terminal protease domain with sequence similarity to the chymotrypsin (S1) family.53 In addition to the three PAPs and four SPHs, 10 other clip-domain proteases are present in Manduca hemolymph.32 It is quite likely that some of them take part in the proPO activation pathway, upstream from the PAPs. Recombinant Manduca serpins are proving useful for exploring the functions of the plasma proteases. The PAPs cleave proPO at Arg-51 within the sequence Asn-ArgPhe-Gly. The similarity of this sequence to the reactive site of serpin-3, Asn-Lys-Phe-Gly from P2 to P2 , led to a prediction that serpin-3 might function as an inhibitor of the PAPs. This hypothesis was demonstrated to be correct. Addition of recombinant serpin-3 to plasma decreased proPO activation, and serpin-3 to inhibited purified PAPs with association rate

ch09

Jan. 10, 2007

8:57

SPI-B429


236

M. R. Kanost

constants of ∼7 × 105 M−1 s−1 . Serpin-3 forms stable complexes with PAP-1 and PAP-3, and these can be detected in plasma activated by exposure to bacteria.34 These data along with the knowledge that mutations in Drosophila serpin 27A, an ortholog of Manduca serpin-3, result in spontaneous melanization in the absence of infection,36,37 are consistent with a role for serpin-3 as a regulator of PAP, the last protease in the prophenoloxidase activation pathway. Serpin-1J and serpin 6 can also inhibit PAP-3.39,50 Multiple plasma serpins that can inhibit each individual PAP provides redundance in function that may be beneficial in protecting insects from unregulated phenoloxidase activation. Drosophila 27A also functions in embryonic development54,55 to regulate a protease pathway, including clip domain proteases that determine dorsal-ventral patterning.43 It may be feasible to test by RNA interference methods whether Manduca serpin-3 also regulates this pathway in lepidopteran embryos. Recombinant serpin-4 and serpin-5 added to plasma block the proPO activation pathway, but they are poor inhibitors of PAPs,17 suggesting that they inhibit one or more proteases upstream from the PAPs in the pathway. The identity of such proteases was determined by immunoaffinity purification of the protease complexes with serpin-4 and serpin-5 that formed in bacteria-activated plasma.16 Serpin-4 inhibits hemolymph clip domain proteases 1, 6, and 21. Serpin-5 also inhibits proteases 1 and 6, suggesting that these proteases may function in the proPO activation pathway. Hemolymph protease-6 is an orthologue of a Drosophila protease known as Persephone, which functions in a pathway that leads to the activation of the Toll ligand spätzle.56 This presents the possibility that Manduca hemolymph protease-6 may also have a role in the activation of the Toll pathway, in response to infection, and that this process could be regulated by serpin-4 and serpin-5.

5. Conclusion Study of insects and other arthropods has helped to demonstrate that serine protease cascade pathways are ancient mechanisms for triggering extracellular responses to wounding and infection, and that serpins regulate such pathways in a wide variety of organisms. The best understood

ch09

Jan. 10, 2007

8:57

SPI-B429



237

protease cascade of invertebrates is the hemolymph clotting pathway in horseshoe crabs,44 which is used as the basis for sensitive assays to detect bacterial endotoxins (the Limulus amoebocyte lysate test). This pathway, which is regulated by three different serpins that inhibit several proteolytic steps,1 serves as an excellent example that invertebrate systems can be useful as models for human biochemistry, for understanding evolution of complex proteolytic pathways, and as a source of molecules with potential biomedical value. Investigation of insect serpins can be expected to expand rapidly with the availability of the genome sequences of several insect species: Drosophila melanogaster and other Drosophila species, Anopheles gambiae (malaria mosquito), Apis mellifera (honeybee), Tribolium castaneum (red flour beetle), and Bombyx mori (silkworm). Genomic information will make the identification and study of a much larger number of insect serpins possible, leading to a better understanding of serpin evolution and physiological function in insects. Detailed investigation of serpins in the Manduca model system has revealed unexpected complexity of serine protease biochemistry in insect hemolymph, and has provided information at a molecular level that has proven useful for understanding serpin structure and inhibitory mechanism. Future insect serpin research can be expected to yield additional surprises as fundamental knowledge in serpin biochemistry, and perhaps ways to manipulate insect serpin biology for disrupting transmission of diseases by blood-feeding insects, or to manage populations of insect pests.

Acknowledgment Research from the author’s laboratory was supported by NIH grant GM41247.

References 1. Kanost MR (1999) Serine proteinase inhibitors in arthropod immunity. Dev Comp Immunol 23(4–5):291–301. 2. Polanowski A, Wilusz T, Blum MS, Escoubas P, Schmidt JO and Travis J (1992) Serine proteinase inhibitor profiles in the hemolymph of a wide range of insect species. Comp Biochem Physiol B 102(4):757–760.

ch09

Jan. 10, 2007

8:57

SPI-B429

238


M. R. Kanost

3. Polanowski A and Wilusz T (1996) Serine proteinase inhibitors from insect hemolymph. Acta Biochim Pol 43(3):445–453. 4. Sasaki T, Kobayashi K and Ozeki T (1987) Interaction of silkworm larval hemolymph antitrypsin and bovine trypsin. J Biochem (Tokyo) 102(2):433–441. 5. Sasaki T, Kohara A, Shimidzu T and Kobayashi, K (1990) Single site proteolysis in silkworm antitrypsin causes structural changes in behavior against denaturing reagents. Agric Biol Chem 54(1):139–145. 6. Sasaki T, Kohara A, Takagi H and Shimidzu T (1990) Limited proteolysis of silkworm antitrypsin by several proteinases. Agric Biol Chem 54(1):131–137. 7. Sasaki T (1985) The reactive site of silkworm hemolymph antichymotrypsin is located at the COOH-terminal region of the molecule. Biochem Biophys Res Commun 132 (1):320–326. 8. Kanost MR, Prasad SV and Wells MA (1989) Primary structure of a member of the serpin superfamily of proteinase inhibitors from an insect, Manduca sexta. J Biol Chem 264(2):965–972. 9. Kanost MR (1990) Isolation and characterization of four serine proteinase inhibitors (serpins) from hemolymph of Manduca sexta. Insect Biochem 20:141–147. 10. Sasaki T (1991) Patchwork-structure serpins from silkworm (Bombyx mori) larval hemolymph. Eur J Biochem 202(2):255–261. 11. Narumi H, Hishida T, Sasaki T, Feng DF and Doolittle RF (1993) Molecular cloning of silkworm (Bombyx mori antichymotrypsin. A new member of the serpin superfamily of proteins from insects. Eur J Biochem 214(1):181–187. 12. Irving JA, Pike RN, Lesk AM and Whisstock JC (2000) Phylogeny of the serpin superfamily: Implications of patterns of amino acid conservation for structure and function. Genome Res 10(12):1845–1864. 13. Silverman GA, Bird PI, Carrell RWet al. (2001) The serpins are an expanding superfamily of structurally similar but functionally diverse proteins. Evolution, mechanism of inhibition, novel functions, and a revised nomenclature. J Biol Chem 276(36):33293–33296. 14. Kanost MR and Jiang H (1997) Serpins from an insect, Manduca sexta. Adv Exp Med Biol 425:155–161. 15. Kanost MR, Jiang H and Yu (2004) XQ Innate immune responses of a lepidopteran insect, Manduca sexta. Immunol Rev 198:97–105. 16. Tong Y, Jiang H and Kanost MR (2005) Identification of plasma proteases inhibited by Manduca sexta serpin-4 and serpin-5 and their association with components of the prophenol oxidase activation pathway. J Biol Chem 280(15):14932–14942.

ch09

Jan. 10, 2007

8:57

SPI-B429



239

17. Tong Y and Kanost MR (2005) Manduca sexta serpin-4 and serpin-5 inhibit the prophenol oxidase activation pathway: cDNA cloning, protein expression, and characterization. J Biol Chem 280(15):14923–14931. 18. Jiang H, Wang Y and Kanost MR (1994) Mutually exclusive exon use and reactive center diversity in insect serpins. J Biol Chem 269(1):55–58. 19. Jiang H, Wang Y, Huang Y et al. (1996) Organization of serpin gene-1 from Manduca sexta. Evolution of a family of alternate exons encoding the reactive site loop. J Biol Chem 271(45):28017–28023. 20. Li J, Wang Z, Canagarajah B, Jiang H, Kanost M and Goldsmith EJ (1999) The structure of active serpin 1K from Manduca sexta. Struct Fold Des 7(1): 103–109. 21. Ye S, Cech AL, Belmares R et al. (2001) The structure of a Michaelis serpinprotease complex. Nat Struct Biol 8(11):979–983. 22. Jiang H and Kanost MR (1997) Characterization and functional analysis of 12 naturally occurring reactive site variants of serpin-1 from Manduca sexta. J Biol Chem 272(2):1082–1087. 23. Kanost MR, Prasad SV, Huang Y and Willott E (1995) Regulation of serpin gene-1 in Manduca sexta. Insect Biochem Mol Biol 25(2):285–291. 24. Chamankhah M, Braun L, Visal-Shah S et al. (2003) Mamestra configurata serpin-1 homologues: Cloning, localization and developmental regulation. Insect Biochem Mol Biol 33(3):355–369. 25. Kruger O, Ladewig J, Koster K and Ragg H (2002) Widespread occurrence of serpin genes with multiple reactive centre-containing exon cassettes in insects and nematodes. Gene 293(1–2):97–105. 26. Danielli A, Kafatos FC and Loukeris TG (2003) Cloning and characterization of four Anopheles gambiae serpin isoforms, differentially induced in the midgut by Plasmodium berghei invasion. J Biol Chem 278(6):4184–4193. 27. Jiang H, Mulnix AB and Kanost MR (1995) Expression and characterization of recombinant Manduca sexta serpin-1B and site-directed mutants that change its inhibitory selectivity. Insect Biochem Mol Biol 25(10):1093–1100. 28. Ye S and Goldsmith EJ (2001) Serpins and other covalent protease inhibitors. Curr Opin Struct Biol 11(6):740–745. 29. Gettins PG (2002) Serpin structure, mechanism, and function. Chem Rev 102(12):4751–4804. 30. Gan H, Wang Y, Jiang H, Mita K and Kanost MR (2001) A bacteria-induced, intracellular serpin in granular hemocytes of Manduca sexta. Insect Biochem Mol Biol 31(9):887–898. 31. Jiang H, Wang Y and Kanost MR (1999) Four serine proteinases expressed in Manduca sexta haemocytes. Insect Mol Biol 8(1):39–53.

ch09

Jan. 10, 2007

8:57

SPI-B429

240


M. R. Kanost

32. Jiang H, Wang Y, Gu Y et al. (2005) Molecular identification of a bevy of serine proteinases in Manduca sexta hemolymph. Insect Biochem Mol Biol 35(8):931-943. 33. Silverman GA, Whisstock JC, Askew DJ et al. (2004) Human clade B serpins (ov-serpins) belong to a cohort of evolutionarily dispersed intracellular proteinase inhibitor clades that protect cells from promiscuous proteolysis. Cell Mol Life Sci 61(3):301–325. 34. Zhu Y, Wang Y, Gorman MJ, Jiang H and Kanost MR (2003) Manduca sexta serpin-3 regulates prophenoloxidase activation in response to infection by inhibiting prophenoloxidase-activating proteinases. J Biol Chem 278(47):46556–46564. 35. Park DS, Shin SW, Hong SD and Park HY (2000) Immunological detection of serpin in the fall webworm, Hyphantria cunea and its inhibitory activity on the prophenoloxidase system. Mol Cells 10(2):186–192. 36. De Gregorio E, Han SJ, Lee WJ et al. (2002) An immune-responsive Serpin regulates the melanization cascade in Drosophila. Dev Cell 3(4):581–592. 37. Ligoxygakis P, Pelte N, Ji C et al. (2002) A serpin mutant links Toll activation to melanization in the host defence of Drosophila. Embo J 21(23): 6330–6337. 38. Wang Y and Jiang H (2004) Purification and characterization of Manduca sexta serpin-6: A serine proteinase inhibitor that selectively inhibits prophenoloxidase-activating proteinase-3. Insect Biochem Mol Biol 34(4):387–395. 39. Zou Z and Jiang H (2005) Manduca sexta serpin-6 regulates immune serine proteinases PAP-3 and HP8. cDNA cloning, protein expression, inhibition kinetics, and function elucidation. J Biol Chem 280(14):14341–14348. 40. Adams MD, Celniker SE, Holt RA et al. (2000) The genome sequence of Drosophila melanogaster. Science 287(5461):2185–2195. 41. Ross J, Jiang H, Kanost MR and Wang Y (2003) Serine proteases and their homologs in the Drosophila melanogaster genome: An initial analysis of sequence conservation and phylogenetic relationships. Gene 304:117–131. 42. Christophides GK, Zdobnov E, Barillas-Mury C et al. (2002) Immunity-related genes and gene families in Anopheles gambiae. Science 298(5591):159–165. 43. Kanost M and Clarke T (2005) Proteases, in L Gilbert, K Iatrou and S Gill (eds.), Comprehensive Molecular Insect Science. Elsevier, New York, pp. 247–266. 44. Iwanaga S (2002) The molecular basis of innate immunity in the horseshoe crab. Curr Opin Immunol 14(1):87–95.

ch09

Jan. 10, 2007

8:57

SPI-B429



241

45. Cerenius L and Soderhall K (2004) The prophenoloxidase-activating system in invertebrates. Immunol Rev 198:116–126. 46. Yu X-Q Prakash O and Kanost MR (2001) Insect ENF family peptides with diverse biological activities display similar core structures in solution. Medicinal Chem Res 10:493–501. 47. Ferrandon D, Imler JL and Hoffmann JA (2004) Sensing infection in Drosophila: Toll and beyond. Semin Immunol 16(1):43–53. 48. Jiang H, Wang Y and Kanost MR (1998) Pro-phenol oxidase activating proteinase from an insect, Manduca sexta: A bacteria-inducible protein similar to Drosophila easter. Proc Natl Acad Sci USA 95(21):12220–12225. 49. Jiang H, Wang Y, Yu XQ and Kanost MR (2003) Prophenoloxidase-activating proteinase-2 from hemolymph of Manduca sexta. A bacteria-inducible serine proteinase containing two clip domains. J Biol Chem 278(6):3552–3561. 50. Jiang H, Wang Y, Yu XQ, Zhu Y and Kanost M (2003) Prophenoloxidaseactivating proteinase-3 (PAP-3) from Manduca sexta hemolymph: A clipdomain serine proteinase regulated by serpin-1J and serine proteinase homologs. Insect Biochem Mol Biol 33(10):1049–1060. 51. Yu XQ, Jiang H, Wang Y and Kanost MR (2003) Nonproteolytic serine proteinase homologs are involved in prophenoloxidase activation in the tobacco hornworm, Manduca sexta. Insect Biochem Mol Biol 33(2):197–208. 52. Gupta S, Wang Y and Jiang H (2005) Manduca sexta prophenoloxidase (proPO) activation requires proPO-activating proteinase (PAP) and serine proteinase homologs (SPHs) simultaneously. Insect Biochem Mol Biol 35:241–248. 53. Jiang H and Kanost MR (2000) The clip-domain family of serine proteinases in arthropods. Insect Biochem Mol Biol 30(2):95–105. 54. Hashimoto C, Kim DR, Weiss LA, Miller JW and Morisato D (2003) Spatial regulation of developmental signaling by a serpin. Dev Cell 5(6):945–950. 55. Ligoxygakis P, Roth S and Reichhart JM (2003) A serpin regulates dorsalventral axis formation in the Drosophila embryo. Curr Biol 13(23):2097–2102. 56. Ligoxygakis P, Pelte N, Hoffmann JA and Reichhart JM (2002) Activation of Drosophila toll during fungal infection by a blood serine protease. Science 297(5578):114–116.

ch09

Jan. 10, 2007 8:57 SPI-B429



ch10

Jan. 10, 2007

8:57

SPI-B429


10 MENT, a Chromatin-Associated Serpin from Avian Blood Cells: Structure and Function Sergei A. Grigoryev, Yaroslava A. Bulynko, Robert N. Pike and James C. Whisstock

1. Introduction Myeloid and erythroid nuclear termination (MENT) stage-specific protein is a developmentally regulated nuclear protein that accumulates in mature chicken blood cells, binds to repressed chromatin and promotes its condensation. MENT belongs to Clade B of the serpin superfamily and its gene is located within a cluster of Clade B serpins in the chicken genome. MENT is a specific inhibitor of cathepsin L, a lysomal protease that can also enter the cell nucleus. The DNA- and chromatin-binding properties of MENT are determined by its unique DNA-binding domain, the M-loop, that is located between the C- and D-helixes of the protein. Recent studies established MENT as the first chromatin-condensing protein, where the particular protein architecture and amino acids involved in chromatin condensation were identified. Remarkably, MENT uses its reactive center loop (RCL) both for protease inhibition and for chromatin condensation, and these two biochemical functions may be intimately related in chromatin regulating mechanisms. The function of MENT is thus of particular interest to two research communities, linking the fields of chromatin structure and protease biology.

243

ch10

Jan. 10, 2007

8:57

SPI-B429

244


S. A. Grigoryev et al.

2. MENT Expression and Accumulation in Heterochromatin of Chicken Blood Cells MENT was first identified as a developmentally regulated nuclear protein selectively associated with repressed and condensed chromatin in nucleated chicken erythrocytes.1 Subsequent studies showed that MENT is highly expressed at the terminal stage of avian hemopoiesis in all major blood cell lines2,3 , when most of the previously active genes become inactivated.4,5 In mature chicken erythrocytes, MENT is localized in relatively small number of nuclear foci.2 These foci are co-localized with dimethylated histone H3(K9), a general marker of repressed genes, but not with trimethylated histone H3(K9), a marker of constitutive heterochromatin.6 This binding specificity towards histone H3 dimethylation supports the proposed role of MENT in the formation of developmentally regulated facultative heterochromatin rather than constitutive heterochromatin. Erythrocytes express about 1 molecule of MENT per 40 nucleosomes, i.e. less than other avian blood cells. MENT is expressed to a very high level in white blood cells: lymphocytes express 1 MENT per 5 nucleosomes, and granulocytes express it at a level equivalent to 2 protein molecules per nucleosome, making it the most abundant nonhistone chromatin protein in vertebrates. In granulocytes, MENT is associated with large electron-dense heterochromatic areas.3 Thus, the expression and location of native MENT in chicken erythrocytes and granulocytes is in excellent agreement with total chromatin condensation in these cells.

3. MENT cDNA Sequence, Gene, and Chromosomal Location MENT cDNA was first cloned from chicken bone marrow. The complete cDNA clone (GenBank accession number AF053401) contains a 1230 bp ORF, which encodes a 410 amino acid polypeptide.7 An identical cDNA clone was also isolated from chicken lymphocyte cDNA library (Acc.No: BG625247). According to the year 2004 assembly of the chicken genome available through the NCBI on-line database, ment-1 gene is located on the Gallus gallus chromosome 2 (gene ID 395715) and is surrounded by other

ch10

Jan. 10, 2007

8:57

SPI-B429


MENT, a Chromatin-Associated Serpin from Avian Blood Cells

245

Fig. 1. Clade B serpin gene cluster on Gallus gallus chromosome 2. (A) Physical mapping of serpin Clade B genes (black boxes) cluster on chromosome 2. Chromosome coordinates represent physical distance (in Kb) from the centromere. The MNEI-, MG664421-and MENT-similar genes are separated by 1 Kb from one another. The ovalbumin-relatedY indicated by asterisks were previously identified as ovalbumin-related Y- (*) and X- (**) genes by O’Malley and co-workers.8 (B) Rooted phylogenetic tree of chicken Clade B serpins. Genes are indicated by their gene ID numbers and their similarity to human serpins (with the exception of MGC64421, which is frog serpin).(C) Exon–intron structure of ment-1 gene (adapted from the NCBI databank). Numbers at the gene borders indicate the chromosome coordinates.

Clade B serpins (Fig. 1A). Such gene arrangement is characteristic of this serpin group and is consistent with rapid expansion of serpin Clade B during vertebrate phylogeny and close evolutionary relationship between neighboring serpin genes. The chromosomal localization as well as sequence comparison allows one to predict the evolutionary relationship between the chicken clade B serpins (Fig. 1C). Interestingly, a MENT-like gene (gene ID 428483) located next to MENT on chromosome 2 (Fig. 1B), shares 76% identity with MENT by nucleotide sequence, but lacks the signature chromatin-binding domain of MENT, the M-loop (see below), suggesting that MENT rapidly evolved to acquire chromatin condensing function.

ch10

Jan. 10, 2007

8:57

SPI-B429

246



Sequence analysis of ment-1 shows that the gene spans 6.38 Kb of DNA and is split into eight exons ranging from 58 to 1193 bp in size (Fig. 1B). The intron size varies between 320 and 1255 bp. The first short exon (58 bp) and the first intron (1255 bp) are located within the 5 -untranslated region. The DNA sequence of the putative upstream transcriptional regulatory region, including the transcription start site (between nucleotides 67423965 and 67423365 at Gallus gallus chromosome 2, gene ID NC_006089), contains consensus binding site for hemapoietic regulators consistent with MENT expression in mature myeloid, erythroid, and lymphoid cells. These include PU.1, Oct-2, GATA-1, and most prominently, C/EBPalpha, a master switch for myeloid differentiation.9 The actual occupancy of these sites by transcription factors remains to be determined.

4. MENT Primary Structure and its Relation to Other Proteins Multiple sequence alignments of MENT amino acid sequence and their comparison with known serpin structures revealed a significant sequence similarity to the serpin family (Fig. 2). The closest sequence similarities of MENT belong to serpin Clade B.Although among Clade B serpins, there are a number of proteins with nuclear or nucleocytoplasmic localization,10–12 as far as we know, MENT is the only serpin with a confirmed nuclear function. Like other serpins, MENT possesses a RCL domain through which interaction with a cognate proteinase occurs. The presence of a conserved inhibitory hinge motif7 in its RCL is consistent with protease inhibition activity of MENT (see later). However, two structural features distinguish MENT from other serpins. These include: (i) the M-loop (61–91), a region rich in positively charged residues that contains an AT-Hook DNA binding motif and a nuclear localization signal, and (ii) a relatively large number of positively charged residues in MENT (pI 9.4) compared with those in most serpins. It is suggested that the high positive charge of the protein is required for neutralizing the negative charge of DNA upon chromatin binding and condensation. The presence of an AT-hook motif in the M-loop of MENT is entirely consistent with the localization of MENT to AT-rich centromeric

ch10

Jan. 10, 2007

8:57

SPI-B429



247

Fig. 2. Amino acid sequence of MENT and its comparison with other serpins. (A) Alignment of the amino acid sequence of MENT deduced from its cDNA with those of the related serpins: human bomapin (Acc.No: P48595), horse leukocyte elastase inhibitor (hlei,Acc.No: P05619), and chicken ovalbumin (oval, Acc.No: P01012). Consensus key: , single, fully conserved residue; :, conservation of strong groups; ., conservation of weak groups; - no consensus. Triangle marks the reactive site P1 residue (371) in MENT. The M-loop domain (61–91) and the “hinge” motif (357–364) are underlined. The AT-hook motif (73–80) is marked by a line on top of the sequence.

heterochromatin.13 The AT-hook is found in many DNA-binding proteins,14 and often mediates binding of a protein to AT-rich DNA15 or chromatin.14,16,17 AT-hook also mediates the nuclear relocation of matrix-associated regions (MARs) or scaffold-associated regions (SARs)

ch10

Jan. 10, 2007

8:57

SPI-B429

248



on DNA.16,17 The M-loop also contains a region of homology to nuclear lamins. In the latter proteins, this region is known to bind core histones H2A and H2B.18 The nuclear lamins line the inner nuclear membrane and regulate the size, shape and assembly of the nuclear envelope. They are also involved in the regulation of higher-order chromatin organization and DNA replication.19 MENT has been shown to be active in detaching chromatin from the nuclear matrix,1 and its competition for histone H2A binding with nuclear lamins is a possible mechanism for this action.

5. 3D Structural Modeling of MENT Protein The structure of a number of serpins including MENT have been determined,20–22 these data reveal a common 3D fold comprising 3 β-sheets (termed A, B and C) and 8-9 α-helices (termed A-I). In native serpins the RCL is located in an exposed conformation, whereas in the cleaved form this region forms an extra strand in the center of the A β-sheet. A molecular model of MENT shown on Fig. 3A is based on serpin homology7 and confirmed by the X-ray crystal structure of native wild-type MENT22 (PDB: 2H4R). The protein contains all of the basic elements of secondary structure present in most serpins, however, it was not possible to determine the structure of variable regions such as the RCL (352–359) and M-loop (61– 91); the position of these regions is indicated by dashed lines in Fig. 3A. Structural analysis of MENT reveals several large clusters of positively charged residues (Fig. 3B), in particular in the region around the D α-helix, which are important for interaction with DNA.22 Several patches of negative charge can also be seen and it is suggested that these regions may coordinate basic regions of nucleosomal proteins, such as histones or indeed other MENT molecules. In addition, the X-ray crystal structure of several forms of MENT have been determined, these data confirm the basic features of the model.22

6. Cathepsin L Inhibition The potential protease cleavage P1-P1 amino acids in MENT RCL (T-T) resembled those in another serpin inhibiting papain-like cysteine proteases, SCCA1.23 Indeed, native MENT isolated from chicken erythrocytes as well

ch10

Jan. 10, 2007

8:57

SPI-B429



249

Fig. 3. MENT 3D structure and esential functional elements. (A) MENT 3D model is based on on serpin homology7 and confirmed by MENT X-ray crystal structure.22 The RCL and the M-loop are indicated by the dashed line. (B) Surface electrostatic potential modeling of MENT using GRASP shows the regions that have positive electrostatic potential in blue, negative electrostatic potential in red, and neutral regions in white. The molecular surface is viewed from the same side as in (A) on the left and from the opposite side on the right. Positions of the M-loop (M) and the RCL domains are indicated.

as recombinant material are effective inhibitors of cathepsin L and V.13 At pH 5.5, MENT inhibited human cathepsins L and V with kass values of 1.4 × 106 M−1 s−1 and 4.0 × 105 M−1 s−1 , respectively (Fig. 4). The stoichiometry of inhibition was 1.0 with both proteases.13 This interaction was highly specific as MENT did not inhibit closely related cathepsins B and S13 , as well as trypsin, chymotrypsin, elastase, thrombin, granzyme B, or granzyme 3.3 The mutations in RCL (but not in M-loop) completely abolished the inhibitory function of MENT, demonstrating that the RCL is essential for MENT functioning as a cysteine protease inhibitor.

ch10

Jan. 10, 2007

8:57

SPI-B429

250



Fig. 4. MENT inhibits cathepsins K, L, and V. (a) Determination of the MENTWT -toproteinase ratio (SI) required to fully inhibit human cathepsins L and V (open square and filled triangle, left panel). The plot of kobs against inhibitor concentration used to derive the second-order rate constant (kass ) is shown at the right, including that for cathepsin K (open diamond).

7. Chromatin Condensation by MENT MENT association with highly condensed chromatin suggested that it might directly promote chromatin condensation. The ability of native MENT isolated from chicken blood cells to condense chromatin has been demonstrated in several in vitro systems. Firstly, MENT was reconstituted with nuclei of immature erythrocytes with very little MENT and an incompletely condensed chromatin. In this system, MENT reconstituted at protein/chromatin ratio equal to that found in native erythrocytes, i.e., 1 molecule per 40 nucleosomes was shown to bring about complete condensation of nuclear chromatin as observed by electron microscopy of ultrathin nuclear sections.1 Such condensation, however, has not been observed with undifferentiated erythroblasts that had a much lower level of linker histone H5, suggesting that other proteins, most probably histone H5, was needed to cooperate with MENT in chromatin condensation. Later experiments (Ref. 24, also see below) confirmed this conclusion. Usually, nuclear chromatin is decondensed at low ionic strength and in the absence of multivalent cations. However, the isolated nuclei of chicken granulocytes, enriched with MENT, were remarkably resistant for chromatin decondensation at low salt. Removing MENT alleviated this resistance, and reassociation of MENT-depleted nuclei with purified

ch10

Jan. 10, 2007

8:57

SPI-B429



251

MENT at a ratio of 1 molecule per nucleosome was sufficient to protect the nuclei from decondensation.3 Thus, the two sets of experiments with isolated and MENT-reconstituted nuclei from avian erythrocytes and granulocytes showed that the protein was sufficient to bring about the highly condensed state of nuclear chromatin at its natural protein/DNA ratio. To examine if purified MENT was sufficient to cause chromatin condensation in a fully defined biochemical system, native MENT was reconstituted with hexanucleosomes assembled in vitro on a strong nucleosome positioning sequence from the 5S rDNA of L. variegatus.25,26 The 5S DNA allows to reconstitute nucleosome arrays with predetermined number of nucleosomes from purified components. When increasing amounts of MENT were added to reconstituted hexanucleosomes, the resulting complexes were gradually retarded in agarose gels indicating a progressive binding of MENT. At a threshold ratio of 2 molecules per nucleosome, MENT caused the particles to self-associate in the presence of 2 mM MgCl2 . Precipitation was reversed when Mg2+ was chelated by EDTA. Electron microscopy of the MENT-reconstituted hexanucleosomes and native chromatin has shown that when 1 molecule of MENT per nucleosome was added, extended linkers were no longer observed, and the chromatin was seen as electron-dense particles of heterogeneous shape and size.3 Thus, upon its binding to nucleosome arrays reconstituted from biochemically defined purified components, MENT was sufficient to cause their condensation.

8. Mechanism of Chromatin Condensation by MENT: Separate Roles of the M-loop and the RCL Domains Protein structural analysis of MENT (Figs. 2 and 3) has revealed the conserved serpin RCL and the unique M-loop domains as potentially most important for its function. The particular role(s) of these domains in chromatin condensation was examined in experiments with wild-type and mutant MENT reconstituted with either naked DNA or the biochemically defined 5S DNA nucleosome arrays.24

ch10

Jan. 10, 2007

8:57

SPI-B429

252



8.1. The M-loop mediates cooperative DNA binding and nucleosome array folding Experiments with MENT binding to naked DNA showed that while MENTwt and all mutants had similar Kd values of approximately 2 nM, the M-loop deletion (but not RCL mutations) caused a striking decrease in DNA binding, indicating that the M-loop is responsible for cooperative binding of MENT to DNA. Under electron microscope, MENT-DNA complexes showed large polymeric structures associated with closely apposed DNA duplexes.24 Similar structures have been previously observed with complexes between globular domains of linker histones and DNA, and are known as “tramline structures”.27 The RCL mutants did not affect the formation of DNA tramlines, while with the M-loop deletion, only irregular DNA-protein complexes were observed but not tramlines. Furthermore, electron microscopy of MENT reconstituted on 12-mer positioned 5S DNA oligonucleosome arrays showed that the M-loop deletion (Fig. 5, panels 7 and 8) but not the RCL swapping, impaired the tight folding of nucleosome linkers. Thus, the M-loop but not the RCL is important for tight folding of either naked DNA or the nucleosome arrays presumably by forming parallel runs of DNA linkers between the nucleosomes.24

8.2. The RCL domain and linker histone promote self-association of nucleosome arrays in vitro In its native setting, MENT is associated with chromatin that also contains linker histones.3 Reconstitution of the 12-mer nucleosome arrays with both linker histone H5 and MENTwt resulted in the extensive self-association of the nucleosome arrays. In contrast, similar particles reassociated with an RCL mutant (MENTov) did not self-associate. These results showed for the first time that chromatin is compacted by a non-histone protein factor through both folding of linker DNA and bridging of nucleosome arrays. Since MENT binding to naked DNA and nucleosome folding both did not require an active RCL, it appeared that MENT oligomerization was not essential for intra-array chromatin folding. Furthermore, electron microscopy revealed a stronger folding of nucleosome “core arrays” by a mutant with inactive RCL, MENTov , than by wild-type MENT (Fig. 5A).

ch10

Jan. 10, 2007

8:57

SPI-B429



253

This suggests that the folding and the bridging activities of MENT are not only separable but may compete with each other, such that MENTfolds more efficiently when its bridging activity is inactivated.24 8.3. Chromatin-dependent MENT oligomerization requires RCL In complexes with naked DNA and chromatin, wild-type MENT formed dimers (predominantly) and higher oligomers revealed by EDC crosslinking. These oligomers were not formed with MENT RCL mutant.24 Since other serpins are known to oligomerize via the formation of RCL–Asheet linkages,28,29 we suggested that such linkages are responsible for making protein bridges connecting chromatin fibers in condensed chromatin (Fig. 5).

Fig. 5. Chromatin condensation by MENT. Panels 1–9: transmission electron microscopy (rotary shadowing technique) showing 12-mer nucleosome arrays (intact on panels 1–3) and reconstituted with MENTwt (4–6), MENTMloop− (7,8), and MENTwt plus histone H5 (9). Adapted from.24 Panel 10 shows a model for higher-order folding of nucleosome arrays in heterochromatin. MENT first binds to nucleosome linker DNA through its M-loop and folds it into a stem-like structure with closely apposed DNA duplexes. MENT binding is stabilized by histone H3 N-terminal domains methylated at lysine 9 and linker histones. MENT then dimerizes through its RCL and forms bridges connecting separate chromatin fibers. Histone acetylation prevents histone methylation and protects active chromatin from condensation.

ch10

Jan. 10, 2007

8:57

SPI-B429

254



With most serpins, the formation of loop A-sheet polymers is irreversible and is associated with the loss of inhibitory activity against target proteases.30 However, MENT isolated from terminally differentiated avian erythocytes is in the native, inhibitory conformation, suggesting that in the case of MENT the oligomerization process is reversible.13 Furthermore, MENTMLoop− , which is inhibitory active, multimerizes in vitro in the absence of any ligand24 , whereas significant loop A-sheet polymerization would be expected to result in the loss of inhibitory activity, which was not observed. In summary, current data suggest that the RCL mediates reversible higher order multimers of MENT that are distinct from typical loop A-sheet polymers. One crystal structure of the active serpin, PAI-1, revealed edge strand linkages between the RCL of one molecule and the Hβ-sheet of the next, forming an infinite chain within the crystalline lattice.31 In this case, the loop-sheet interaction is reversible, since active material was obtained from dissolved crystals. X-ray crystal structure analysis of MENT revealed a similar loop-sheet interaction in the crystals suggesting that MENT may also undergo reversible polymerization on chromatin.22

9. Regulation of MENT Localization and Action Inside the Nuclei 9.1. Histone modifications In terminally differentiated cells with globally condensed chromatin, some genes such as globin genes in nucleated chicken erythrocytes remain transcriptionally active. These genes contain a high level of multiple histone acetylation and a low level of histone methylation (at histone H3 lysine 9), and are protected from chromatin condensation by special boundary elements.32 Mapping the topography of MENT across the active chicken β-globin locus by chromatin immunoprecipitation (ChIP) with anti-MENT antibodies revealed two sharp transitions of the protein level coinciding with the β-globin boundary elements. The MENT distribution profile was opposite to that of acetylated core histones, but correlated with that of histone H3 dimethylated at lysine 9 (H3me2 K9). MENT was also co-localized with H3me2 K9 by in situ immunofluorescence experiments.6 Thus, like other heterochromatin proteins such as HP1 and polycomb, MENT was physically associated with chromatin marked by histone H3 methylation.

ch10

Jan. 10, 2007

8:57

SPI-B429



255

Ectopic MENT expression in cells that normally do not express it (NIH/3T3) caused large-scale and specific remodeling of chromatin marked by histone H3 methylation. Mutational analysis of MENT and experiments with deacetylase inhibitors revealed the essential role of the RCL domain and an inhibitory affect of histone hyperacetylation on MENT-induced chromatin remodeling. In vitro, reconstitution with dimethylated but not acetylated N-terminal domain of histone H3 specifically promoted chromatin self-association by MENT and its oligomerization in MENT-DNA complexes. These results suggest that histone H3 modification at lysine 9 directly regulates chromatin condensation by promoting MENT oligomerization in a fashion that requires the RCL domain of MENT, and is spatially constrained from active genes by gene boundary elements and histone hyperacetylation.6 Apparently, the chromosomal loci that become inactivated during terminal differentiation first acquire histone H3(K9) methylation and then become selectively condensed when developmentally regulated serpins accumulate in the nucleus. 9.2. Cathepsin L In vitro MENT inhibits Cathepsin L, an enzyme that has been long known as a lysosomal protease. Remarkably, the inhibition of cathepsin L in vivo rearranges MENT inside the nucleus, promoting its association with constitutive heterochromatin and suggesting that cathepsin L or a similar protease directly interacts with MENT inside the nucleus.13 Indeed, using MENT ectopically expressed in NIH/3T3 cells and highly specific in vivo reactive center labeling of cathepsin L with 125 I-FYAC, it was found that nuclear MENT protects cathepsin L from labeling in vivo and that overexpression of cathepsin L relocates MENT to euchromatin in a striking contrast to the effect of the protease inhibitors. These results suggest that cathepsin L interacts with MENT inside the cell and may at least temporarily enter the nucleus.33 Nuclear translocation of cathepsin L at the G1/S phase transition has been clearly demonstrated by Goulet et al.34 Furthermore, it has been recently shown that targeting of cathepsin L gene leads to a dramatic rearrangement of heterochromatin proteins and histone H3 (K9) methylation in mouse knockout cells.35 It thus appears that MENT may be capable of modifying chromatin structure via two molecular reactions: first by destabilizing the preexisting constitutive heterochromatin structures via

ch10

Jan. 10, 2007

8:57

SPI-B429


256


cathepsin L inhibition and then by forming polymers that bridge distinct chromatin fibers together and cause their condensation. MENT is found in greatest abundance in the nuclei of avian granulocytes. The complete repression of chromatin may become especially important during myeloid differentiation, where strong oxidative stress often causes DNA damage and chromosomal translocations, leading to oncogenic transformation and leukemia. The entry of cathepsin L in the nuclei of proliferating cells and its strong effect on MENT, which is not expressed during proliferation, suggested an interesting possibility that the serpin-protease interaction may be engaged in preventing cells from resuming proliferation, after being committed to terminal differentiation. Indeed, if a cell expressing chromatin condensing factor starts a proliferation cycle, then a spurt of cathepsin L into the nucleus at the G1/S transition would relocate MENT to euchromatin and induce its binding to active genes, thus repressing gene activity and stopping cell proliferation. Although MENT is not yet found in humans, several Clade B serpins, closely related to MENT, are known to enter the nuclei of mammalian cells.12,36 Future studies should reveal their relationship within human chromatin and gene regulatory mechanisms.

References 1. Grigoryev SA, Solovieva VO, Spirin KS and Krasheninnikov IA (1992) A Novel nonhistone protein (MENT) promotes nuclear collapse at the terminal stage of avian erythropoiesis. Exp Cell Res 198:268–275. 2. Grigoryev SA and Woodcock CL (1993) Stage-specific expression and localization of MENT, a nuclear protein associated with chromatin condensation in terminally differentiating avian erythroid cells. Exp Cell Res 206:335–343. 3. Grigoryev SA and Woodcock CL (1998) Chromatin structure in granulocytes. A link between tight compaction and accumulation of a heterochromatinassociated protein (MENT). J Biol Chem 273:3082–3089. 4. Tobin AJ, Selvig SE and Lasky L (1978) RNA synthesis in avian erythroid cells. Dev Biol 67:11–22. 5. Beaulieu AD, Paquin R, Rathanaswami P and McColl SR (1992) Nuclear signaling in human neutrophils. Stimulation of RNA synthesis is a response to a limited number of proinflammatory agonists. J Biol Chem 267:426–432.

ch10

Jan. 10, 2007

8:57

SPI-B429



257

6. Istomina NE, Shushanov SS, Springhetti EM, Karpov VL, Krasheninnikov IA, Stevens K, Zaret KS, Singh PB and Grigoryev SA (2003) Insulation of the Chicken b-globin Chromosomal Domain from a Chromatin-Condensing Protein, MENT. Mol Cell Biol 23:6455–6468. 7. Grigoryev SA, Bednar J and Woodcock CL (1999) MENT a heterochromatin protein that mediates higher order chromatin folding, is a new serpin family member. J Biol Chem 274:5626–5636. 8. Lawson GM, Tsai MJ and O’Malley BW (1980) Deoxyribonuclease I sensitivity of the nontranscribed sequences flanking the 5’and 3’ends of the ovomucoid gene and the ovalbumin and its related X and Y genes in hen oviduct nuclei Biochemistry 19:4403–4441. 9. Friedman AD (2002) Transcriptional regulation of granulocyte and monocyte development. Oncogene 21:3377–3390. 10. Chuang TL and Schleef RR (1999) Identification of a nuclear targeting domain in the insertion between helices C and D in protease inhibitor-10. J Biol Chem 274:11194–11198. 11. Irving JA, Pike RN, Lesk AM and Whisstock JC (2000) Phylogeny of the serpin superfamily: Implications of patterns of amino acid conservation for structure and function. Genome Res 10:1845–1864. 12. Bird CH, Blink EJ, Hirst CE, Buzza MS, Steele PM, Sun J, Jans DA and Bird PI (2001) Nucleocytoplasmic distribution of the ovalbumin serpin pi-9 requires a nonconventional nuclear import pathway and the export factor crm1. Mol Cell Biol 21:5396–5407. 13. Irving JA, Shushanov SS, Pike RN, Popova EY, Bromme D, Coetzer TH, Bottomley SP, Boulynko IA, Grigoryev SA and Whisstock JC (2002) Inhibitory activity of a heterochromatin-associated serpin (MENT) against papain-like cysteine proteinases affects chromatin structure and blocks cell proliferation. J Biol Chem 277:13192–13201. 14. Aravind L and Landsman D (1998) AT-hook motifs identified in a wide variety of DNA-binding proteins. NAR 26:4413–4421. 15. Reeves R and Nissen MS (1990) The AT-DNA-binding domain of mammalian high mobility group I chromosomal proteins. A novel peptide motif for recognizing DNA structure. J Biol Chem 265:8573–8582. 16. Strick R and Laemmli UK (1995) SARs are cis DNA elements of chromosome dynamics: Synthesis of a SAR repressor protein. Cell 83:1137–1148. 17. Girard F, Bello B, Laemmli UK and Gehring WJ (1998) In vivo analysis of scaffold-associated regions in Drosophila: A synthetic high-affinity SAR binding protein suppresses position effect variegation. Embo J 17: 2079–2085.

ch10

Jan. 10, 2007

8:57

SPI-B429

258



18. Goldberg M et al. (1999) The tail domain binding of lamin Dmo binds histones H2A and H2B. Proc Natl Acad Sci 96(6):2852–2857. 19. Hutchison CJ (2002) Lamins: Building blocks or regulators of gene expression? Nat Rev Mol Cell Biol 3:848–858. 20. Whisstock J, Skinner R and Lesk AM (1998) An atlas of serpin conformations. Trends Biochem Sci 23:63–67. 21. Silverman GA, Bird PI, Carrell RW, Church FC, Coughlin PB, Gettins PG, Irving JA, Lomas DA, Luke CJ, Moyer RW, Pemberton PA, RemoldO’Donnell E, Salvesen GS, Travis J and Whisstock JC (2001) The serpins are an expanding superfamily of structurally similar but functionally diverse proteins. Evolution, mechanism of inhibition, novel functions, and a revised nomenclature. J Biol Chem 276:33293–33296. 22. McGowan S, Buckle AM, Irving JA, Ong P-C, Bashtannyk-Puhalovich TA, Kan W-T, Henderson KN, Bulynko YA, Popova EY, Smith AI, Bottomley SP, Rossjohn J, Grigoryev SA, Pike RN and Whisstock JC. X-ray crystal structure of MENT: Evidence for functional loop-sheet polymers in chromatin condensation. EMBO J 25:3144–3155. 23. Schick C, Bromme D, Bartuski AJ, Uemura Y, Schechter NM, and Silverman GA (1998) The reactive site loop of the serpin SCCA1 is essential for cysteine proteinase inhibition. Proc Natl Acad Sci USA 95:13465–13470. 24. Springhetti EM, Istomina NE, Whisstock JC, Nikitina TV, Woodcock CL and Grigoryev SA (2003) Role of the M-loop and reactive center loop domains in the folding and bridging of nucleosome arrays by MENT. J Biol Chem 278:43384–43393. 25. Simpson RT, Thoma F and Brubaker JM (1985) Chromatin reconstituted from tandemly repeated cloned DNA fragments and core histones: A model system for study of higher order structure. Cell 42:799–808. 26. Meersseman G, Pennings S, and Bradbury EM (1991) Chromatosome positioning on assembled long chromatin. Linker histones affect nucleosome placement on 5 S rDNA. J Mol Biol 220:89–100. 27. Thomas JO, Rees C and Finch JT (1992) Cooperative binding of the globular domains of histones H1 and H5 to DNA. Nucleic Acids Res 20: 187–194. 28. Dunstone MA, Dai W, Whisstock JC, Rossjohn J, Pike RN, Feil SC, Le Bonniec BF, Parker MW and Bottomley SP (2000) Cleaved antitrypsin polymers at atomic resolution. Protein Sci 9:417–420. 29. Huntington JA, Pannu NS, Hazes B, Read RJ, Lomas DA and Carrell RW (1999) A 2.6 A structure of a serpin polymer and implications for conformational disease. J Mol Biol 293:449–455.

ch10

Jan. 10, 2007

8:57

SPI-B429



259

30. Lomas DA (2000) Loop-sheet polymerization: The mechanism of alpha1antitrypsin deficiency. Respir Med 94(Suppl C):S3–S6. 31. Sharp AM, Stein PE, Pannu NS, Carrell RW, Berkenpas MB, Ginsburg D, Lawrence DA and Read RJ (1999) The active conformation of plasminogen activator inhibitor 1, a target for drugs to control fibrinolysis and cell adhesion. Struct Fold Des 7:111–118. 32. Litt MD, Simpson M, Gaszner M, Allis CD and Felsenfeld G (2001) Correlation between histone lysine methylation and developmental changes at the chicken beta-globin locus. Science 293:2453–2455. 33. Bulynko YA and Grigoryev SA (2003) Interaction with cathepsin L modifies chromatin-associated serpin, MENT, in the cell nucleus: Is cathepsin L a nuclear protease? Mol Biol Cell 14:2283. 34. Goulet B, Baruch A, Moon NS, Poirier M, Sansregret LL, Erickson A, Bogyo M and Nepveu A (2004) A cathepsin L isoform that is devoid of a signal peptide localizes to the nucleus in S phase and processes the CDP/Cux transcription factor. Mol Cell 14:207–219. 35. Bulynko YA, Hsing LC, Mason RW, Tremethick D and Grigoryev SA (2006) Cathepsin L stabilizes histone modification landscape on Y chromosome and pericentromeric heterochromatin. Mol Cell Biol 26:4172–4184. 36. Popova EY, Claxton DF, Lukasova E, Bird PI and Grigoryev SA (2006) Epigenetic Heterochromatin markers distinguish terminally differentiated leukocytes from incompletely differentiated Leukemia cells in human blood. Exp Hematol 34:453–462.

ch10

Jan. 10, 2007 8:57 SPI-B429



ch11

Jan. 10, 2007

8:57

SPI-B429


11 Uterine Serpins Peter J. Hansen, Saban Tekin and Maria B. Padua

1. Introduction The mammalian uterus creates and maintains an environment in which the developing conceptus can undergo placentation and differentiation, meet its nutritional and respiratory requirements, be protected from maternal immune responses, and be buffered from major disturbances in maternal physiology. This environment is organized under the direction of maternal and conceptus signaling molecules, and is dependent on secretions of the uterine endometrium and conceptus, as well as blood transudate. Serpins in this progestational milieu include plasminogen activator inhibitor-1 (PAI-1), secreted from endometrial, stromal, and epithelial cells1,2 and trophoblast,2,3 protease nexin-1, produced by endometrial stroma,4 α1-antitrypsin, derived from serum transudate,5 and maspin, present in trophoblast.6 These proteins may function during pregnancy to limit remodeling of the endometrium, participate in local hemostasis, and regulate placental invasiveness. The endometrial epithelium of at least some mammals also secrete a serpin, often in abundant quantity, whose expression appears endometriumspecific and which exerts roles in pregnancy, distinct from those associated with proteinase inhibition. The so-called uterine serpins (US) have been identified as products of the endometrium in sheep,7 goats,8 cattle,9 and pigs.9,10 Evolutionary analysis suggests the possibility that the proteins could be found in other species also.11 The sheep uterine serpin is the 261

ch11

Jan. 10, 2007

8:57

SPI-B429

262


Peter J. Hansen, Saban Tekin and Maria B. Padua

predominant protein in the uterine fluid during pregnancy and has been postulated to protect the conceptus from destruction by the maternal immune system through the inhibition of lymphocyte function.12 The two porcine uterine serpins form complexes with the iron-containing protein, uteroferrin,13 and may be involved in the metabolism of that protein. The function of the uterine serpins in the goat and cow has not been investigated. The purpose of this chapter is to summarize the existing knowledge about this novel group of serpins.

2. Uterine Serpin Genes The uterine serpins have been classified either as a separate clade of the serpin superfamily or as a highly diverged group of the α1-antitrypsin Clade A.11,14 While only identified so far in artiodactyls, the estimated divergence of uterine serpins from other serpins predates divergence of mammals,11 and it is possible that functional uterine serpin genes have been retained in other mammals as well. Alignment of the inferred amino acid sequences of the uterine serpins are displayed in Fig. 1. The gene sequence of uterine serpin is well conserved. For example, the predicted amino acid sequence of the CaUS precursor showed 96%, 82%, 55%, and 56% identity to OvUS, BoUS, PoUS1, and PoUS2, respectively.8 There are two start codons located in each of the uterine serpin genes, resulting in proteins that differ in length by five amino acids. The signal peptide extends from amino acids 1 to 25; there is a mutation at the site of cleavage of the signal sequence in the goat gene, where a single nucleotide substitution (G→C) changed the cysteine in the sheep, bovine and porcine genes to a serine. Nonetheless, it is most likely that the signal sequence is cleaved at a similar position for all four species, since the requirements for a small and neutral amino acid at position –115 is maintained in the goat. There are two N-glycosylation sites in all species, except the cow in which only one site has been conserved. Both the sheep and goat uterine serpins have a nine amino acid insert in the Helix I region that is not found in bovine or porcine uterine serpins. All species have a conserved KVP at P4-P2, but there are other differences between species at the putative reactive center. The amino acid at the putative P1-P1 site (the scissile bond for antiproteinase activity) is a

ch11

Jan. 10, 2007

8:57

SPI-B429


Uterine Serpins

263

Fig. 1. Alignment of predicted uterine serpin sequences. Shaded bases are those that are identical to the goat sequence. Accession numbers are as follows: CaUS, TPA: BK005554; OvUS, M21027; BoUS, L22095; PoUS 1, M30315; and PoUS2: NM213845. The alignment is reproduced with permission from Molecular Reproduction and Development.8

valine for CaUS, BoUS, PoUS1, and PoUS2 versus an alanine for OvUS. In the bovine, there is a 39-amino acid insertion containing two repeats of KAKEVPAVVKVPM and a similar KEVPVVVKVP located between the P1 and P1 . The P2 site is not well conserved between species, being threonine for sheep and cow, alanine for goat, and lysine for the two pig genes. The hinge region of all five of the uterine serpins (P17-P9) is distinct from the consensus pattern for inhibitory sequences.16 Rather than the consensus glycine at P15, uterine serpins have threonine (OvUS and CaUS), valine (BoUS), or lysine (PoUS1 and PoUS2). Instead of threonine or serine at P14, uterine serpins have aspartic acid (OvUS, CaUS, and BoUS) or asparagine (PoUS1 and PoUS) and the P12-P9 region of uterine serpins are not filled with the alanines, glycines, and serines that predominate

ch11

Jan. 10, 2007

8:57

SPI-B429

264



in the consensus sequence for inhibitory serpins. All of the uterine serpins have a conserved KEVPVVVK motif upstream of the P1-P1 reactive center.

3. Physicochemical Properties Depending on the species, the uterine serpins are secreted from endometrial epithelium as proteins of Mr = 50,000–57,000. In sheep, for example, endometrial-conditioned medium cultured with radiolabeled amino acid precursors contains two major forms of US at Mr = 55,000 and 57,000 (OvUS-55 and OvUS-57).17,18 Only one OvUS gene has been identified and microheterogeneity appears not to be due to differential processing of the preprotein because only one N-terminal amino acid sequence that is present.7,18 Rather, the OvUS-57K and OvUS-55K proteins are formed from a tissue form of Mr = 53,000.18 Ovine US is approximately 5.6% carbohydrate by weight and oligosaccharide chains have been processed at least partially to complex-type forms.18 Differential glycosylation may therefore be responsible for the formation of OvUS-57 and OvUS-55. The size and lectin-binding characteristics of the oligosaccharides on OvUS-57 and OvUS-55 are similar18 and it is unlikely that the oligosaccharide on OvUS-57 is a more complex oligosaccharide than the oligosaccharide on OvUS-55. Given the two potential sites for N-glycosylation (Fig. 1), it is possible that OvUS-57 has two oligosaccharide chains, while OvUS-55 has only one chain. The OvUS also contains mannose 6-phosphate on at least a proportion of its oligosaccharides.18 In addition to the major forms of OvUS, uterine secretions and endometrium-conditioned medium contain other immunoreactive proteins of lower molecular weight.17,19–21 The most prominent of these are at Mr ∼ 47,000–49,000, and are probably proteolytic breakdown products.22 The OvUS can also form large aggregates in the uterus17 and during longterm storage at −20◦ C.23 Uterine fluid from unilaterally pregnant ewes contains large amounts of colloidal or precipitated protein in uterine fluid from unilaterally pregnant ewes and these are composed of OvUS aggregates.17 Aggregation is partly due, to the intermolecular disulfide bond formation because aggregates of precipitated OvUS can be dissolved with reducing agents.17

ch11

Jan. 10, 2007

8:57

SPI-B429


Uterine Serpins

265

The pig also contains several uterine serpins in uterine fluid with major species at Mr = 50,000, 46,000, and 39,000.13,24 There are two separate genes encoding proteins of predicted molecular weights of 46,009 and 45,123.9,10 Like the sheep protein, there is also evidence for differential glycosylation, presence of mannose 6-phosphate and proteolytic cleavage.10,24 Microheterogeneous variants in the CaUS and BoS have not been thoroughly examined, although immunoreactive proteins of lower molecular weight can be found in uterine secretions for both species.8,20,21

4. Tissue Source and Patterns of Secretion During Pregnancy Surprisingly, it is still not clear whether expression of uterine serpin genes is limited to the uterus. In that tissue, synthesis of uterine serpins is restricted to the endometrial epithelium.8,20,21,25 The pattern of synthesis and the secretion of uterine serpin during the estrous cycle and pregnancy has been characterized in most detail in sheep. There is no evidence of OvUS synthesis in the endometrium during the estrous cycle until day 16 of the cycle.26 Detectable OvUS mRNA is first detected in the endometrium on days 13 and 14 of pregnancy,25,26 and the protein can first be identified in uterine secretions on day 16 of the pregnancy.26 From day 30 of pregnancy until term, OvUS is the predominant protein present in uterine fluid.17 Steady-state amounts of OvUS mRNA in endometrium increased between day 20 and 80 of gestation, and were then reduced slightly by day 120 of gestation.25 The protein is gone by day 7 postpartum.27 In the endometrium, the epithelial lining of the endometrial glands is continuous with the epithelium covering the surface of the uterine lumen. Experiments with sheep and cattle indicate that uterine serpins are initially produced by deep to middle-depth glandular epithelium and that synthesis eventually extends to the glandular regions near the surface of the uterine lumen.21,26 Another study suggests that, the deepest endometrial glands have lower expression of OvUS25 until day 50 of pregnancy. Expression of OvUS mRNA was limited to glandular epithelium through day 120 of the pregnancy25 and OvUS protein was found in the glandular epithelium but

ch11

Jan. 10, 2007

8:57

SPI-B429

266



not the luminal epithelium on day 60 of the pregnancy.19 By day 120–140 of the pregnancy, however, immunoreactive OvUS was also detected in luminal epithelium.17,19 An exhaustive analysis of immunolocalization of uterine serpins from other species has not been done, but at least at the stages of pregnancy examined, immunoreactive product was limited to glandular epithelium in cow,20 and pig,24 and primarily to glandular epithelium in goat with a few scattered luminal epithelial cells having reaction product.8

5. Hormonal Regulation of Uterine Serpin Secretion The major regulator of gene expression for uterine serpin is progesterone. Treatment of ovariectomized animals induces endometrial secretion of uterine serpin and accumulation of uterine serpin RNA in sheep17,21,26,28 and cattle.21 Treatment with progesterone for as little as 6 days can induce OvUS synthesis,26 but prolonged treatment with progesterone is necessary to induce high levels of uterine serpin synthesis. Amounts of OvUS in uterine flushings after 10 days of progesterone treatment were small, compared with those from ovariectomized ewes treated for 30 days.21 During the estrous cycle, OvUS is not detected in the uterus until day 16.26 Perhaps, induction of uterine serpin gene expression requires differentiation events in the endometrial epithelium, and the long-term nature of induction by progesterone explains why the protein is absent during the estrous cycle and in early pregnancy. Among the differentiation events required may be a reduction in progesterone receptor concentrations in endometrial epithelium. Amounts of progesterone receptor mRNA and protein in endometrial epithelium are decreased by progesterone treatment.28 The magnitude of the effect of progesterone on uterine serpin secretion is modified by other hormones. Estradiol-17β has been shown to inhibit OvUS synthesis and reduce steady-state levels of OvUS mRNA.28 Estradiol may exert this effect by upregulating progesterone receptor expression because it increases steady-state levels of progesterone receptor mRNA in endometrium.28 The induction of OvUS by progesterone can be enhanced by treatment with intrauterine treatment, with either

ch11

Jan. 10, 2007

8:57

SPI-B429


Uterine Serpins

267

placental lactogen or growth hormone, provided that the animal is also treated with interferon-τ, a product of the peri-implantation ruminant conceptus.28

6. Proteinase Inhibitory Activity and Protein Conformation More research into the conformational properties and enzymatic activity of uterine serpins is needed before it can be determined whether uterine serpins are prototypical inhibitory serpins. To date, the results obtained are not supportive of this idea. Screening of OvUS for inhibitory activity towards trypsin, chymotrypsin, plasmin, thrombin, elastase, plasminogen activator, cathepsin D, cathepsin E, and dipeptidyl proteinase IV (DPPIV) indicates that the protein does not inhibit these proteinases.7,29,30 Ovine uterine serpin can bind the aspartic proteinase, pepsin, and weakly inhibit its activity.30 It was speculated that pepsin inhibitory activity was related to motifs in the putative reactive center loop (RCL) that were similar to pepstatin (VVVK) and the propepsinogens (KVP). However, peptides corresponding to these regions had no pepsin-inhibitory activity.22 There is evidence that the conformation of uterine serpin is distinct from the prototypical serpin. While mild denaturation of OvUS with 0.5 M guanidine HCl increased thermal stability, the increase in thermal stability was lost upon removal of the denaturant.22 This is in contrast to the situation with α1-antitrypsin and antithrombin III, where proteins remain thermostabile after the removal of denaturant.31 Also, OvUS displayed only a single, continuous unfolding pattern in response to increasing the concentrations of guanidine HCl, and observable changes in circular dichroism at 222 nm was seen at 0.2 M guanidine HCl. This result is in contrast to antithrombin III, which demonstrated a two phase denaturation curve with inflections at 0.7 and 4 M guanidine HCl.32 Results suggest that OvUS is very unstable to guanidine HCl denaturation (Fig. 2). Moreover, incubation of OvUS with 100-fold molar excess of a peptide corresponding to the putative P14 –P2 region of the RCL to induce binary complex formation had no effect on the protein’s secondary structure, and did not alter the biological activity of the protein.22

ch11

Jan. 10, 2007

8:57

SPI-B429

268



Fig. 2. Effect of guanidine HCl on physical properties of OvUS. The top panel shows effect of guanidine HCl on thermal stability. Incubation of OvUS in DPBS + 0.5 M guanidine HCl (open circles) increased the thermal stability of OvUS relative to protein in DPBS alone (closed circles). Incubation of OvUS in DPBS + 0.5 M NaCl (closed squares) had no effect on thermal stability. The bottom panel illustrates unfolding of OvUS with guanidine HCl. There was a single unfolding transition as measured by the loss in circular dichroism at 222 nm. Y is the predicted change in circular dichroism at concentration X of guanidine HCl. Data are reproduced with permission of Biochimica et Biophysica Acta.22

ch11

Jan. 10, 2007

8:57

SPI-B429


Uterine Serpins

269

7. Lymphocyte Inhibitory Activity In the sheep, uterine serpin can inhibit T-lymphocyte proliferation and other aspects of immune function. Ovine uterine serpin inhibits proliferation of cultured peripheral blood lymphocytes induced by several T-cell activators, including phytohemagglutinnin, concanavalin A, mixed lymphocyte reactions, and Candida albicans antigen.33−36 The protein also reduced antibody titer in sheep to ovalbumin, a T-cell dependent antigen, when mixed with adjuvant.35 In mice, OvUS reduced natural killer (NK) cell activity in splenocytes and blocked abortion mediated by the activation of NK cells.37 While the existence of natural killer cells in ruminants has not demonstrated unequivocally in ruminants, OvUS inhibits activity of NK-like cells in peripheral blood and endometrium of sheep.38 In contrast, OvUS does not appear to inhibit γδ-T cell proliferation.36 The resistance of these cells to OvUS may partly explain why their numbers in the luminal epithelium of the endometrium rise in mid and late pregnancy.39 Inhibition of lymphocyte function by OvUS is likely to be physiologically relevant. Concentrations of OvUS required to inhibit lymphocyte proliferation (∼40–500 µg/ml) are high (see Fig. 3), but these concentrations are well within the range of concentrations present in uterine fluid after day 30 of the pregnancy. In addition, recent data indicates that recombinant OvUS is more active than OvUS purified from uterine fluid.40 Consistent with a role for OvUS in the inhibition of uterine immune responses are the findings that treatment with progesterone reduces the rejection response to intrauterine transplants of allografts41,42 and xenografts.43 Thus, OvUS may mediate this action of progesterone.

8. Inhibition of Proliferation of Other Cells In addition to inhibiting lymphocyte proliferation, OvUS can also inhibit the growth of other cells. In particular, OvUS inhibited proliferation of PC-3 adenocarcinoma cells and D17 osteosarcoma cells.40,44 Interestingly, both of these cell lines are tumor cells, and OvUS did not inhibit the growth of MDBK cells or BEND cells, which are the two non-tumor cell lines.44 Perhaps, OvUS inhibits a signal transduction system for proliferation present in rapidly growing cells.

ch11

Jan. 10, 2007

8:57

SPI-B429

270



Fig. 3. Inhibition of lymphocyte proliferation by phytohemagglutinin (PHA) or phorbol myristate acetate (PMA) by OvUS. Lymphocytes were cultured for either 24 or 72 h in the presence of OvUS and then pulsed for an additional 16–18 h with 0.5 µCi 3 H-thymidine. Data reproduced with permission from the Society for Experimental Biology and Medicine.38

Ovine uterine serpin also inhibits the growth of preimplantation bovine embryos.44 It may seem counterintuitive that a progestational protein would prove inhibitory to the growth of the embryo. It must be kept in mind, however, that OvUS is not first produced until day 16 of pregnancy,26 a time when the embryo has passed the period of growth shown to be sensitive to

ch11

Jan. 10, 2007

8:57

SPI-B429


Uterine Serpins

271

OvUS. Pharmacological treatment with progesterone to hasten the rise of progesterone concentrations in circulation during early pregnancy can lead to reduced fertility in cattle,45 and one possible reason is because of the premature secretion of uterine serpin.

9. Mechanism of Action of OvUS to Block Cell Proliferation Ovine uterine serpin has been demonstrated to specifically bind to lymphocytes and MDBK cells22,46 but it is not known whether this binding represents interactions of OvUS with a specific receptor. If so, the interpretation of the dose-response curve for inhibition of cell proliferation would imply that it is a receptor with a low Kd . Binding is not due to interactions with a serpin enzyme complex receptor.46 One possibility is that a uterine serpin receptor does not exist, but that OvUS inhibits cell proliferation by weakly binding to receptors for other molecules to either inhibit activation of stimulatory receptors or to activate inhibitory receptors. Lymphocyte proliferation induced by phorbol myristate acetate can be blocked by OvUS (see Fig. 3; Ref. 36). Thus, OvUS is likely to inhibit lymphocyte proliferation by either blocking phospholipase C-induced activation of protein kinase C or some downstream event. Uterine serpin does not block the ability of mitogen to increase interleukin-2 (IL-2) mRNA, but OvUS can decrease mitogen induced expression of CD25 (the α chain of IL-2 receptor) and IL-2 induced proliferation of PBL.36 One possibility is that OvUS inhibits proliferation of cells through intracellular actions. Bovine seminal ribonuclease is an example of a protein with basic pI that inhibits lymphocyte proliferation and that which enters the lymphocyte.47,48 Peltier and Hansen (49) have postulated that OvUS could act as a competitive inhibitor of protein kinase C because it contains five protein kinase C phosphorylation sites.11 It is also possible that OvUS inhibits cell proliferation by inducing cell death. At high concentrations, the protein can be toxic.34 There is, however, no evidence for a pro-apoptotic role for the protein.44 Proteinases such as DPPIV are involved in lymphocyte activation50,51 and one could postulate that OvUS might inhibit lymphocyte function by blocking a proteinase involved in lymphocyte activation.At least for DPPIV,

ch11

Jan. 10, 2007

8:57

SPI-B429

272



however, OvUS does not inhibit DPPIV coactivation of lymphocytes or DPPIV enzymatic activity.29

10. Other Activities of Uterine Serpins Uterine serpins have been shown to cross the placenta and enter allantoic and amniotic fluids in sheep and pigs.13,52−54 This observation, as well as the fact that uterine serpins are so abundant in uterine fluid, suggests that they may function during pregnancy to transfer poorly soluble molecules across the placenta to the fetus. The serpins corticosteroid binding globulin and thyroxine binding globulin serve such a role. To date, no such poorly soluble ligand has been found for a uterine serpin. Uterine serpins do have a propensity for binding other proteins. The ovine uterine serpin binds IgA and IgM,55 the growth factor activin,54 as well as the pregnancy-associated glycoproteins secreted by the sheep trophoblast.9 The functional importance of these binding reactions is unknown. Interactions with immunoglobulins are electrostatic in nature55 and do not reflect a transport function of OvUS, because immunoglobulins do not cross the ruminant placenta.56 The pregnancy-associated glycoproteins are inactive aspartyl proteinases and perhaps the binding of OvUS to these proteins is exerted through a similar domain as the one responsible for the inhibition of pepsin. The pig uterine serpin binds to another uterine protein called uteroferrin which is an iron-containing acid phosphatase.13 In fact, the protein’s original name in the literature was uteroferrin-associated protein. One function of PoUS is to stabilize uteroferrin: association with PoUS maintains uteroferrin’s acid phosphatase activity when exposed to air.13 Uteroferrin itself also crosses the placenta and its major role seems to be that of delivering iron to the fetus.57

11. Conclusions Originally defined as inhibitors of a specific class of proteinase, it is now recognized that the serpins consist of a large number (∼500) of proteins that play diverse roles in many organ systems of the body.14,58 Some of the roles assumed by serpins during evolution reflect functional properties separate from proteinase inhibition. The uterine serpins represent a distinct

ch11

Jan. 10, 2007

8:57

SPI-B429


Uterine Serpins

273

group of serpins by virtue of their evolution,11,14 abundance (OvUS is the most copious steroid-induced protein described), and functional properties (inhibition of cell proliferation). It remains to be seen whether the uterine serpin genes have survived as functional genes in a wide range of mammals or whether functional genes are limited to the artiodactyls. Investigation is also required to determine whether expression in species in which functional uterine serpin genes exist is limited to the uterus or occurs for other tissues as well. In addition, more research is required to better understand the function of the uterine serpins and whether those functions require the prototypical serpin conformation that confers proteinase inhibitory activity to serpins.

Acknowledgements Appreciation is expressed to all the colleagues who have contributed to the development of data and concepts discussed in this paper. The authors’ research was supported by USDA National Research Initiative Competitive Grant No. 2001-35204-10797.

References 1. Chou CS, MacCalman CD and Leung PC (2003) Differential effects of gonadotropin-releasing hormone I and II on the urokinase-type plasminogen activator/plasminogen activator inhibitor system in human decidual stromal cells in vitro. J Clin Endocrinol Metab 88:3806–3815. 2. Ueyama M, Kasatori N, Urayama T, Maemura T, Yao Y, Shiraishi T, Saito S and Kubo H. (2002) Quantitative evaluation of the influence of ovarian steroids on plasminogen activators and inhibitors in human endometrial cells and trophoblasts. Thromb Res 108:235–244. 3. Chou CS, Zhu H, Shalev E, MacCalman CD and Leung PCK. The effects of gonadotropin-releasing hormone (GnRH) I and GnRH II on the urokinase type plasminogen activator/plasminogen activator inhibitor system in human extravillous cytotrophoblasts in vitro. J Clin Endocrinol Metab 87:5594–5603. 4. Kim NK, Choi MJ, Chung HM, Sohn TJ, Hwang SG, Oh D, Lee HH, Lee YH and Ko JJ (2001) Increased expression and localization of a serine protease inhibitor, protease nexin-1 (PN-1), in the ovary and uterus during implantation in rat. Thromb Res 103:135–142.

ch11

Jan. 10, 2007

8:57

SPI-B429

274



5. Bany BM and McRae AC (1992) Uterine uptake of α2 -macroglobulin and α1 -proteinase inhibitor from the blood during early implantation in the mouse. Biol Reprod 47:514–519. 6. Dokras A, Gardner LM, Kirschmann DA, Seftor EA and Hendrix MJ (2002) The tumour suppressor gene maspin is differentially regulated in cytotrophoblasts during human placental development. Placenta 23:274–280. 7. Ing NH and Roberts RM (1989) The major progesterone-induced proteins secreted into the sheep uterus are members of the serpin superfamily of protease inhibitors. J Biol Chem 264:3372–3379. 8. Tekin S, Padua MB, Newton GR and Hansen PJ (2005) Identification and cloning of caprine uterine serpin. Mol Reprod Dev 70:262–270. 9. Mathialagan N and Hansen TR (1996) Pepsin-inhibitory activity of the uterine serpins. Proc Natl Acad Sci USA 93:13653–13658. 10. Malathy PV, Imakawa K, Simmen RC and Roberts RM (1990) Molecular cloning of the uteroferrin-associated protein, a major progesterone-induced serpin secreted by the porcine uterus, and the expression of its mRNA during pregnancy. Mol Endocrinol 4:428–440. 11. Peltier MR, Raley LC, Liberles DA, Benner SA and Hansen PJ (2000) Evolutionary history of the uterine serpins. J Exp Zool 288:165–174. 12. Hansen PJ (1998) Regulation of uterine immune function by progesterone lessons from the sheep. J Reprod Immunol 40:63–79. 13. Baumbach GA, Ketcham CM, Richardson DE, Bazer FW and Roberts RM (1986) Isolation and characterization of a high molecular weight stable pink form of uteroferrin from uterine secretions and allantoic fluid of pigs. J Biol Chem 261:12869–12878. 14. Irving JA, Pike RN, Lesk AM and Whisstock JC (2000) Phylogeny of the serpin superfamily: Implications of patterns of amino acid conservation for structure and function. Genome Res 10:1845–1864. 15. von Heijne G (1985) Signal sequences. The limits of variation. J Mol Biol 84:99–105. 16. Hopkins PC, Carrell RW and Stone SR (1993) Effects of mutations in the hinge region of serpins. Biochemistry 32:7650–7657. 17. Moffatt J, Bazer FW, Hansen PJ, Chun PW and Roberts RM (1987) Purification, secretion and immunocytochemical localization of the uterine milk proteins, major progesterone-induced proteins in the uterine secretions of the sheep. Biol Reprod 36:419–430. 18. Hansen PJ, Ing NH, Moffatt RJ, Baumbach GA, Saunders PT, Bazer FW and Roberts RM (1987) Biochemical characterization and biosynthesis of the uterine milk proteins of the pregnant sheep uterus. Biol Reprod 36: 405–418.

ch11

Jan. 10, 2007

8:57

SPI-B429


Uterine Serpins

275

19. Stephenson DC, Low BG, Newton GR, Leslie MV, Hansen PJ and Bazer FW (1989) Secretion of the major progesterone-induced proteins of the sheep uterus by caruncular and intercaruncular endometrium of the pregnant ewe from days 20-140 of gestation. Domest Anim Endocrinol 6:349–362. 20. Leslie MV, Hansen PJ and Newton GR (1990) Uterine secretions of the cow contain proteins that are immunochemically related to the major progesteroneinduced proteins of the sheep uterus. Domest Anim Endocrinol 7:517–526. 21. Leslie MV and Hansen PJ (1991) Progesterone-regulated secretion of the serpin-like proteins of the ovine and bovine uterus. Steroids 56:589–597. 22. Peltier MR, Grant TR and Hansen PJ (2000) Distinct physical and structural properties of the ovine uterine serpin. Biochim Biophys Acta 1479:37–51. 23. Hansen PJ and Liu W-J (1994) Biochemical/physiological properties of endometrial serpin-like proteins. Adv Contracept Delivery Sys 10:339–353. 24. Murray MK, Malathy PV, Bazer FW and Roberts RM (1989) Structural relationship, biosynthesis, and immunocytochemical localization of uteroferrinassociated basic glycoprotein. J Biol Chem 264:4143–4150. 25. Stewart MD, Johnson GA, Burghardt RC, Schuler LA, Joyce MM, Bazer FW and Spencer TE (2000) Prolactin receptor and uterine milk protein expression in the ovine uterus. Biol Reprod 62:1779–1789. 26. Ing NH, Francis H, Mc Donnell JJ, Amann JF and Roberts RM (1989) Progesterone induction of the uterine milk proteins: Major secretory proteins of sheep endometrium. Biol Reprod 41:643–654. 27. Gray CA, Stewart MD, Johnson GA and Spencer TE (2003) Postpartum uterine involution in sheep: Histoarchitecture and changes in endometrial gene expression. Reproduction 125:185–198. 28. Spencer TE, Gray A, Johnson GA, Taylor KM, Gertler A, Gootwine E, Ott TL and Bazer FW (1999) Effects of recombinant ovine interferon tau, placental lactogen, and growth hormone on the ovine uterus. Biol Reprod 61:1409–1418. 29. Liu W-J and Hansen PJ (1995) Progesterone-induced secretion of dipeptidyl peptidase-IV (cell differentiation antigen-26) by the uterine endometrium of the ewe and cow that costimulates lymphocyte proliferation. Endocrinology 136: 779–787. 30. Mathialagan N and Hansen TR (1996) Pepsin-inhibitory activity of the uterine serpins. Proc Natl Acad Sci USA 93:13653–13658. 31. Carrell RW, Evans DL and Stein PE (1991) Mobile reactive centre of serpins and the control of thrombosis. Nature 353:576–578. 32. Villanueva GB and Allen N (1983) Demonstration of a two-domain structure of antithrombin III during its denaturation in guanidinium chloride. J Biol Chem 258:11010–11013.

ch11

Jan. 10, 2007

8:57

SPI-B429

276



33. Segerson EC, Moffatt RJ, Bazer FW and Roberts RM (1984) Suppression of phytohemagglutinin-stimulated lymphocyte blastogenesis by ovine uterine milk protein. Biol Reprod 30:1175–1186. 34. Skopets B and Hansen PJ (1993) Identification of the predominant proteins in uterine fluids of unilaterally pregnant ewes that inhibit lymphocyte proliferation. Biol Reprod 49:997–1007. 35. Skopets B, Liu W-J and Hansen PJ (1995) Effects of endometrial serpin-like proteins on immune responses in sheep. Am J Reprod Immunol 33:86–93. 36. Peltier MR, Liu W-J and Hansen PJ (2000) Regulation of lymphocyte proliferation by uterine serpin: interleukin-2 mRNA production, CD25 expression and responsiveness to interleukin-2. Proc Soc Exp Biol Med 223:75–81. 37. Liu W-J and Hansen PJ (1993) Effect of the progesterone-induced serpin-like proteins of the sheep endometrium on natural-killer cell activity in sheep and mice. Biol Reprod 49:1008–1114. 38. Tekin S and Hansen PJ (2002) Natural-killer like cells in the sheep: Functional characterization and regulation by pregnancy-associated proteins. Exp Biol Med 227: 803–811. 39. Lee CS, Meeusen E, Gogolin-Ewens K and Brandon MR (1992) Quantitative and qualitative changes in the intraepithelial lymphocyte population in the uterus of nonpregnant and pregnant sheep. Am J Reprod Immunol 28:90–96. 40. Padua MB, Tekin S and Hansen PJ (2005) Antiproliferative actions of ovine uterine serpin on PC-3 prostate cancer cells — comparison of recombinant and native forms of the protein. Am J Reprod Immunol 53:295 (abstract). 41. Hansen PJ, Bazer FW and Segerson EC (1986) Skin graft survival in the uterine lumen of ewes treated with progesterone. Am J Reprod Immunol Microbiol 12:48–54. 42. Padua MB, Tekin S, Spencer TE and Hansen PJ (2005) Actions of progesterone on uterine immunosuppression and endometrial gland development in the uterine gland knockout (UGKO) ewe. Mol Reprod Dev 71:347–357. 43. Majewski AC and Hansen PJ (2002) Progesterone inhibits rejection of xenogeneic transplants in the sheep uterus. Horm Res 58:128–135. 44. Tekin S, Padua MB and Hansen PJ (2005) Antiproliferative and embryoinhibitory actions of ovine uterine serpin. Am J Reprod Immunol 53:136–43. 45. Van Cleeff J, Drost M and Thatcher WW (1991) Effects of postinsemination progesterone supplementation on fertility and subsequent estrous responses of dairy heifers. Theriogenology 36:795–807. 46. Liu W-J, Peltier MR and Hansen PJ (1999) Binding of ovine uterine serpin to lymphocytes. Am J Reprod Immunol 41:428–432.

ch11

Jan. 10, 2007

8:57

SPI-B429


Uterine Serpins

277

47. Mancheno JM, Gasset M, Onaderra M, Gavilnes JG and DAlessio G (1994) Bovine seminal ribonuclease destabilizes negatively charged membranes. Biochem Biophys Res Commun 199:119–124. 48. Mastronicola MR, Piccoli R and DAlessio G (1995) Key extracellular and intracellular steps in the antitumor action of seminal ribonuclease. Eur J Biochem 230:242–249. 49. Peltier MR and Hansen PJ (2001) Immunoregulatory activity, biochemistry, and phylogeny of ovine uterine serpin. Am J Reprod Immunol 45:266–272. 50. Schon E, Demuth H-U, Eichmann E, Horst H-J, Korner I-J, Kopp J, Mattern T, Neubert K, Noll F, Ulmer AJ, Barth A and Ansorge S (1989) Dipeptidyl peptidase IV in human T lymphocytes: Impaired induction of interleukin 2 and gamma interferon due to specific inhibition of dipeptidyl peptidase IV. Scand J Immunol 29:127–132. 51. Tanaka T, Kameoka J, Yaron A, Schlossman ST and Morimoto C (1993) The costimulatory activity of the CD26 antigen requires dipeptidyl peptidase IV enzymatic activity. Proc Natl Acad Sci USA 90:4586–4590. 52. Moffatt RJ, Bazer FW, Roberts RM and Thatcher WW (1987) Secretory function of the ovine uterus: effects of gestation and steroid replacement therapy. J Anim Sci 65:1400–1410. 53. Newton GR, Hansen PJ, Bazer FW, Leslie MV, Stephenson DC and Low BG (1989) Presence of the major progesterone-induced proteins of the sheep endometrium in fetal fluids. Biol Reprod 40:417–424. 54. McFarlane JR, Foulds LM, O’ConnorAE, Phillips DJ, Jenkin G, Hearn MT and de Kretser DM (1999) Uterine milk protein, a novel activin-binding protein, is present in ovine allantoic fluid. Endocrinology 140:4745–4752. 55. Hansen PJ and Newton GR (1988) Binding of immunoglobulins to the major progesterone-induced proteins of the sheep uterus. Arch Biochem Biophys 260:208–217. 56. Powell JR, Barratt MEJ and Porter P (1984) The aquisition of immunity by the neonate, in Crighton DB (ed.) The Aquisition of Immunity by the Neonate. Butterworths, London, pp 265–290. 57. Buhi WC, Ducsay CA, Bazer FW and Roberts RM (1982) Iron transfer between the purple phosphatase uteroferrin and transferrin and its possible role in iron metabolism of the fetal pig. J Biol Chem 257:1712–1723. 58. Silverman GA, Bird PI, Carrell RW, Church FC, Coughlin PB, Gettins PG, Irving JA, Lomas DA, Luke CJ, Moyer RW, Pemberton PA, RemoldO’Donnell E, Salvesen GS, Travis J and Whisstock JC (2001) The serpins are an expanding superfamily of structurally similar but functionally diverse proteins. Evolution, mechanism of inhibition, novel functions, and a revised nomenclature. J Biol Chem 276:33293–33296.

ch11

Jan. 10, 2007 8:57 SPI-B429



ch12

Jan. 10, 2007

8:57

SPI-B429


12 Plant Serpins Jørn Hejgaard and Thomas H. Roberts

1. Introduction Serpins are present in only some species of Archaea, Bacteria and unicellular Eukarya, but are probably found in all species of animals and plants.1–4 Sequence databases contain serpin genes in species from the major groups in the Plant Kingdom, including the eudicots (e.g. Arabidopsis, tomato), monocots (e.g. rice, wheat), conifers, and mosses, as well as green algae (e.g. Chlamydomonas), the closest relatives of the plants. Consistent with the monophyletic nature of the Kingdom Plantae and the existence of land plants for more than 400 million years, one of the 16 serpin clades identified by Irving et al.5 in phylogenetic analyses of higher eukaryotes and viruses was composed entirely of plant serpins. Many serpin genes are known to be expressed in various vegetative and reproductive plant organs, such as leaves and seeds, but little is known about the tissue- and cell-specific localization of the corresponding proteins.6 Plant serpins appear to be intracellular; no cleavable signal peptides have been experimentally identified and serpins have not been detected in extracellular fluids.6 The characterized plant serpins, mainly from cereal grains7–11 but also from vegetative tissues of other plants,12 are all efficient inhibitors of mammalian serine proteinases in vitro, suggesting that they act as inhibitors in vivo. However, target proteinases for plant serpins have not been identified; neither have interactions with other ligands or proteins known to regulate or modulate functions of 279

ch12

Jan. 10, 2007

8:57

SPI-B429

280


J. Hejgaard and T. H. Roberts

many serpins in mammals.3 Thus, physiological functions of plant serpins are yet to be discovered, and whether these proteins are functionally diverse or share a few common functions remains unknown.

2. The Beer Connection: A Brief History of Plant Serpin Research For decades, the proteinaceous material remaining in finished beer was assumed to be a heterogenous mixture of “proteoses”; i.e. smaller peptides covalently modified and crosslinked by phenolics. Notwithstanding this assumption, early immunoelectrophoretic experiments13 clearly showed that immunochemically distinct components were present in beer. During the last part of the 1960s, the application of variants of twodimensional immunoelectrophoresis led to a more profound understanding of the complexity of proteins in blood serum. Using similar methods for beer protein characterization allowed identification of a ∼ 40 kDa major barley grain albumin that was resistant to proteolytic degradation and heat denaturation.14 This polypeptide was present in beer as the only major proteinaceous component with an essentially retained peptide backbone structure. Its concentration in finished beer varied depending on wort strength (20–170 mg/l). The protein antigen was shown to be the major protein of beer foam and haze and was thus of industrial interest. The barley protein giving rise to this major beer antigen was termed protein Z in 1976,15 the year before a structurally unrelated plasma protein was given the same name;16 no one knew then that eventually these two distinct “protein Z’s” would (in different ways) become associated with the emerging family of serpins. Further characterization of beer protein Z by comparative immunoelectrophoretic methods17 showed that this beer antigen was a mixture of at least four related products of genes on two barley chromosomes (Fig. 1). The first partial sequence of barley protein Z4 (encoded by a gene on barley chromosome 4) was obtained in 1984. Similarities with ovalbumin, both in terms of protein size and a transformation to a stable molecular variant (plakalbumin) upon specific proteolytic cleavage of a single, peptide bond, led to a comparison of sequences, and thus the identification of the first plant serpin18 years before databases and search engines became everyday

ch12

Jan. 10, 2007

8:57

SPI-B429


Plant Serpins

281

Fig. 1. Comparison of serpins in wheat–barley addition lines with those found in beer by tandem crossed immunoelectrophoresis. Extracts of grains of wheat lines containing homologous pairs of single barley chromosomes (1–7) in addition to the complete set of wheat chromosomes were compared with a beer protein concentrate (X), using antibodies against freeze-dried beer.17 (A) Immunochemical identity between beer antigen 1a and a gene product of barley chromosome 4, now 4H (BSZ4a). (B) Partial immunochemical identity between beer antigen 1b and several (2–4) gene products of barley chromosome 7, now 1H (molecular forms of BSZ7).

tools. The first plant serpin gene was characterized in 199019 and the first plant serpin with confirmed inhibitory properties was isolated from wheat grain in 1994.20 Since then, more than 20 serpins from monocot and eudicot seeds have been characterized with respect to primary structure and inhibitory kinetics, after cloning or chromatographic isolation from tissue extracts.7–11,21 The first serpin from a vegetative tissue was isolated from pumpkin phloem, cloned and characterized in 2000.12 Immunomicroscopy, expression studies and a growing number of serpin EST sequences from tissue-specific libraries have confirmed that serpins are widespread in plant tissues,6 fuelling interest in discovering their physiological functions. To prevent confusion in the literature on serpins and blood coagulation due to the designation “plant protein Z,” we changed the plant serpin nomenclature in later publications, but retained the Z-signature in the names for individual plant serpins. For the cereal serpins, we used BSZ4

ch12

Jan. 10, 2007

8:57

SPI-B429

282



for barley (Z-type) serpin 4, WSZ1a for wheat serpin 1a, etc., with numbers designating subfamilies (irrespective of chromosomal origin). With detection of an increasing number of plant serpins, a slightly modified nomenclature was used; e.g. MdZ2a for Malus domestica (apple) serpin 2a, in accordance with the two-letter species abbreviations used for plant gene numbering (Table 1). Plant serpins have gained some interest in human medicine. The proteins of wheat flour are allergens for more than 60% of bakers with workplace-induced respiratory symptoms. As abundant soluble grain proteins, the serpins have been identified among the predominant IgE-binding allergens in several studies.22 Although beer allergy is rare, serpin-derived allergens have been identified as the most common causative agent in several cases of contact urticaria and IgE-mediated anaphylaxis, following ingestion of barley or wheat malt beer.23,24

3. Serpin Genes in Plant Genomes: Chlamydomonas, Arabidopsis, and Rice The three fully sequenced genomes of organisms belonging to the Viridiplantae (green plants according to the NCBI Taxonomy browser) are those of Chlamydomonas reinhardtii (a green alga), Arabidopsis thaliana (thale cress) and Oryza sativa (rice). In all three organisms, serpin genes are found only in the nuclear genomes (Table 1). Serpins have not yet been characterized from these species, but are the subjects of current research. Chlamydomonas reinhardtii is the simplest model organism of the green plants. Members of the genus are water- and soil-dwelling green alga able to grow both in the dark and in the light, and are found in a great variety of habitats. The genome of C. reinhardtii has an estimated size of ∼ 125 Mbp on 16 chromosomes. The C. reinhardtii genome (http://genome.jgi-psf.org.chlre3/chlre3.info.html) contains only a single serpin gene (Table 1). Analysis of the putative amino acid sequence and comparisons to those of serpins from other unicellular eukaryotes and prokaryotes suggested that the RCL length of serpins from the unicellular eukaryotes was one residue shorter than that found in most serpins,4 and thus the 390-residue C. reinhardtii serpin is an inhibitor with P2-P1

ch12

Jan. 10, 2007

8:57

SPI-B429


Plant Serpins Table 1

Serpin genes in model plant genomes.a

Chlamydomonas reinhardtii C_1140006

283

P1

P1

DEEGTVAAAAT-AVMMLR

CALPM----PTPEF

p

Arabidopsis thaliana At1g47710 NEEGTEAAAASAGVIKLR At1g64030* DEEGAEAAAATADGDCGC At2g14540* DEEGTEAAAATTVVVVTG At2g25240 DEEGTEAAAVSVGVVSCT At2g26390 DEEGTEAAAVSVAIMMPQ At3g45220 DEEGTEAAAVSVASMTKD

GLLMEE---DEIDF SLDFVEPP-KKIDF SCLWEPK--KKIDF SFR------RNPDF CLM------RNPDF ML-------LMGDF

p

Oryza sativa cv. Nipponbare Os01g16200 DELGTVAAASTAVVMMQK Os01g56010 NEEGTEAAAATAVVMTLG Os03g41410 NEEGTEAAAATAAVITLR Os03g41438 DEEGTEAAAASAAVVSFR Os11g11760* NEEGTVAAAATMTRMLPS Os11g12410* NEEGTEAAAATAVLMEGA Os11g12420 NEEGTEVAAATVVIMKGR Os11g12460 NEEGTEAAAATAVCLTFA Os11g12520* NEEGTEAAASAINMVCGM Os11g13530 NEEGTEAAASTACTIRLL Os11g13540 NEEGTEAAAATACTMKFL

GSSL-----PPVDF CAAPSAPVH-VVDF SAP------IAEDF SAP------VTVDF GVPP-----PPVDF ARYAPPPP-PREDF ARRPSPAP-APVDF SAAPSSRRPARVDF SMTPEPRP-VPVDF SMS------YPEDF CLTLT----SPVDF

n

p p n

n

n

a RCL sequences of putative inhibitory serpins from fully sequenced Viridiplantae genomes (Sept 2006): The unicellular green alga Chlamydomonas reinhardtii (∼125 Mbp genome on 16 chromosomes), the eudicot model plant Arabidopsis thaliana (thale cress; ∼125 Mbp on five chromosomes) and the monocot cereal Oryza sativa (rice; ∼420 Mbp genome on 12 chromosomes; serpins from the subspecies japonica cultivar Nipponbare are shown). Serpin genes are found only in the nuclear genomes in all three organisms. The table lists all genes coding for full-length serpins (∼400 residues) with conserved structural motifs predicted to be functional proteinase inhibitors. An asterisk indicates that the sequence contains unusual insertions/deletions/extensions (discussed in text; some may be due to misannotations). The two-letter organism initials are followed by the chromosome number and the annotated gene (g) number. In the Chlamydomonas sequence a gap has been introduced at P7 to fit with the predicted 16-residue RCL.4 Expression has been confirmed at the RNA (n) or protein (p) level for only some of the putative genes as indicated to the right. The expression of a single serpin with a predicted P2-P1 Leu-Arg (LR serpin) in each species is highlighted.

Leu-Arg-Cys (Table 1). EST sequences confirming almost the entire sequence showed that the serpin is expressed. Recent proteomic studies using whole-cell extracts of C. reinhardtii confirmed expression at the protein level (Dr. R.D. Willows, personal communication).

ch12

Jan. 10, 2007

8:57

SPI-B429

284



The small ∼ 125-Mbp genome of the model eudicot A. thaliana, a member of the mustard (Brassicaceae) family, contains ∼ 25,500 genes on five chromosomes.25 Searches in The Arabidopsis Information Resource TAIR database (http://www.arabidopsis.org) followed by examination of the amino acid sequences of the putative full-length gene products, suggested that six genes code for inhibitory serpins (Table 1). Two of the six putative serpins (genes At1g64030 and At14540) appear to lack the surface loop linking helixI1 to strand5A found in serpins generally, but homology modeling suggests that this deletion of up to 24 residues may not interfere with the mechanism of irreversible inhibition. All of the serpin genes contain a single intron, as found for barley BSZ4.19 A substantial number of serpin pseudogenes in Arabidopsis (there are ∼ 25 genes and pseudogenes in total) is consistent with the genome’s history of duplication (duplicated regions encompass 67.9 Mbp or 60% of the Arabidopsis genome). It is also consistent with the finding that the proportion of proteins belonging to families of more than five members is substantially higher in Arabidopsis (37.4%) than in fully sequenced organisms of a similar genome size, namely Drosophila (12.1%) or C. elegans (24.0%).25 The majority of serpin pseudogenes are at positions immediately adjacent to the “normal” serpin genes, suggesting gene duplication followed by loss-of-function mutations. The EST sequences confirm expression of At1g47710 and At2g14540 and results of A. thaliana gene expression studies, involving partial sequences of four of the serpin genes and based on Affymetrix® gene chips and on cDNA microarrays (www.arabidopsis.org), provided clear confirmation of the expression of only At1g47710 (in many tissues, At2g14540 and At3g45220 were not covered).26 Serpin expression at the protein level has been confirmed for At1g47710 in an Arabidopsis proteome study of siliques.27 Rice (Oryza sativa) serves as a model genome for the second main group of flowering plants, the monocots, as Arabidopsis is the model for the eudicots (monocots and eudicots comprise 97% of the ∼ 300,000 flowering plant species). Rice represents 30% of current world cereal production, providing staple food for more than half the world’s population. The rice genome was selected to be sequenced as a priority also because it has the smallest genome of all the monocot cereal crops. The euchromatic part of the rice genome, ∼ 430 Mb found on 12 chromosomes, is only one-fifth

ch12

Jan. 10, 2007

8:57

SPI-B429


Plant Serpins

285

of the size of the maize genome, and one-fortieth of that of wheat. However, comparisons between cereal genomes reveal large blocks of homologous genes whose order is relatively conserved. This synteny makes rice a good entry point for characterizing the genes of other cereals and their possible associations with agronomic traits. Approximately one-half of the rice genome corresponds to the A. thaliana gene complement, whereas the other half of the predicted rice genes have no obvious homology to genes in A. thaliana. Complete genomes have been sequenced for two subspecies of rice, ssp. japonica and ssp. indica. Searching for serpin genes in the public rice genome databases, primarily The TIGR Rice Genome Annotation Database (http:www.tigr.org) and Gramene (http://www.gramene.org), which provides links to all rice genome sequencing efforts, suggests that ∼10 genes coding for functional serpins are present in rice. Tentatively, a maximum of 11 genes are proposed to code for inhibitory serpins, on the basis of similarity with other plant serpin sequences, including highly conserved structural motifs such as the hinge and hydrophobic C-terminal sequence (Table 1). In addition, the rice genome contains ∼10 longer pseudogenes, mainly occurring in clusters with the functional genes as found for Arabidopsis. Information about protein expression in rice is limited and evidence from EST/cDNA databases and cDNA microarray studies confirmed expression of only five of the putative rice serpin genes (Table 1). A comprehensive rice proteome exploration28 confirmed that two of these genes were expressed at the protein level, Os01g56010 in the roots and Os03g41410 in the seed. The recently constructed Japanese rice proteome database (http://gene64.dna.affrc.go.jp/RPD), which has amino acid sequences of more than 5,000 proteins identified in one or more of 23 reference maps, based on 2D electrophoresis of proteins from different tissues, does not identify any serpin spots. Comprehensive EST libraries exist for wheat and barley, allowing assembly of several full-length sequences of serpins with putative inhibitory properties (unpublished results), in addition to the previously characterized grain serpins from these cereals.7,8 Based on these observations, we predict that at least seven serpins are expressed from the haploid genomes of wheat and barley. These results are in good agreement with the number of

ch12

Jan. 10, 2007

8:57

SPI-B429

286



putative serpin genes in Arabidopsis and in rice, suggesting that at least five but probably less than 10 distinct serpin genes may be expected to be expressed in any flowering plant.

4. The QQ Serpins: Glutamine-rich Reactive Centers of Cereal Grain Serpins Glutamine (Gln) is an extremely rare residue at the reactive center of serpins. To our knowledge, the only serpins with P1 Gln and a confirmed inhibitory activity at this residue are found in plants. In humans, Gln is found only in the sequence position corresponding to P1 in a non-inhibitory serpin, thyroxine-binding globulin (TBG), and only at P1 in megsin.3 In some plant serpins, sequences of two, three and even four glutamine residues, including P1, have been found at the reactive center (Table 2). With one exception, these reactive centers have until now been found only in the related cereals of the Triticeae tribe wheat, barley and rye. Several serpins of wheat grain were found to contain the reactive center bonds P1-P1 Gln-Gln or X-Gln,8 whereas reactive centers with P2-P1 Gln-Gln-Ser or X-Gln-Ser were present in rye.9 Irrespective of the identity of P1 (Gln or Ser), these serpins were found to be efficient inhibitors at P1 Gln of chymotrypsin and cathepsin G but not of serine proteinases with trypsinor elastase-like specificity. For example, two wheat serpins with P1-P1 Gln-Gln were found to inhibit chymotrypsin (ka ∼ 104 –105 M−1 s−1 and SI 1.3–1.5) as efficiently as cereal inhibitors with P1-P1 Leu-Ser, a reactive center found in many mammalian serpins.8 More recently, serpins from wheat and barley with three consecutive Gln residues at the putative reactive centers P2-P1 or P3-P1 purified from wheat and barley, respectively, were characterized, but only the barley serpin (BSZ3) was confirmed to be inhibitory10 (Table 2). These two serpins have not been characterized kinetically, but are structurally most similar to barley serpin BSZ7 with P1 Arg (∼ 80% amino acid identity). Thus, serpins with Gln at the reactive center are found in three distinct serpin subfamilies in wheat. Database BLAST searches have revealed the presence of one additional putative plant serpin with more than one Gln residue in the reactive center.10 Overlapping cDNAs that could be assembled to cover the entire sequence of a serpin

ch12

Jan. 10, 2007

8:57

SPI-B429


Plant Serpins

287

Table 2 Plant serpins with Gln residues at the reactive center bond P1 –P1 .a Gln (Q)

Position

Serpinsb

Inhibition

Q Q Q QQ QQ QQQ QQQ QQQQ

P2 P1 P1 P1–P1 P2–P1 P2–P1 P3–P1 P3–P1

OSZb RSZc1 WSZ1c+2b/c WSZ1a/b RSZb/c2/d/e WSZ3 BSZ3 GaZ1

P1-specific CT+CatG P1-specific CT+CatG CT+CatG Not CT CT Not tested

a

The number of glutamine residues (Q) in the sequence, and the position of each in the RCL is indicated. The serpins referred to are described in Refs. 8–11 or in the text. Serpins with P1 Gln tested were inhibitors of chymotrypsin (CT) and cathepsin G (Cat G). b B, barley; Ga, Gossypium arboreum (cotton); O, oat; R, rye; W, wheat.

with putative P3-P1 Gln-Gln-Gln-Gln were identified in cotton (Gossypium arboreum) fiber EST libraries (GaZ1 in Table 2).

5. The LR Serpins: A Reactive Center Conserved in Plants In addition to the obvious need for the conservation of residues close to P1 to retain inhibitory specificity, there is no strict conservation of residues in the P7-P4 region of inhibitory serpins in general. Only the general requirement for hydrophobic residues in the even-numbered positions of the RCL is heeded in most cases. The RCL sequences of serpins present in high concentrations in seeds from closely related cereals may illustrate this point for plant serpins (Fig. 2). The differences close to P1 may reflect evolutionary adaptation for defense against a range of pests or diseases as discussed in Section 7, but atypical acidic or basic residues can also be found in the P7-P3 region. In contrast to the serpins abundantly present in plant seeds, serpins expressed at lower levels in vegetative tissues (and possibly also in seed tissues) are dominated by molecular forms with the conserved P2-P1

ch12

Jan. 10, 2007

8:57

SPI-B429

288


J. Hejgaard and T. H. Roberts A BSZ4 BSZ7 WSZ1b OSZ

↓ P8 TVAMGVAM S TGDVIVER S TAIKMVPQ Q TATTIMQT S

B AM/AP SMS/GP PS Physc Chlamy

SAGVIKLR SAATVVLR TAATVLLR TAATITLR TA-VMMLR

G S S S C

Fig. 2. Reactive center sequences (P8-P1 ) in serpins from cereal seeds (A) and in LR serpins from plant vegetative tissues (B). (A) RCL sequences of barley serpins BSZ4 and BSZ7, wheat serpin WSZ1b and oat serpin OSZc, illustrating the diversity found in RCL sequences of seed serpins from related cereals. (B) The conserved RCLs of LR serpins expressed in vegetative tissues of most plants studied: AM/AP, Arabidopsis and mustard (cabbage family)/apple and peach (rose family). SMS/GP, sugar cane, maize and sorghum (monocots)/grape and Poncirus (rue family). PS, pine and spruce (conifers); Physc, Physcomitrella (a moss); Chlamy, Chlamydomonas reinhardtii (a green alga); presumably with a 16-residue RCL4 (gap introduced).

sequence Leu-Arg-Ser/Gly. BLAST searches in plant EST databases (May 2006) with query sequences corresponding to 60-residue C-terminal regions of selected plant serpins (containing the highly conserved serpin signature motifs on both sides of the RCL sequence) identified expressed serpin genes or pseudogenes in 48 plant genera: 34 eudicots, 11 monocots, two conifers, and a moss. Except for the plants examined in greater detail as discussed above (Arabidopsis and the cereals) serpins in these genera were represented by only a few available sequences, in most cases, one or two. The ESTs were derived from various libraries, mainly from vegetative or mixed tissues. Unexpectedly, 44 of the 48 genera (>90%) expressed a serpin sequence with the P2-P1 Leu-Arg and either Ser, Gly or Ala (in two cases) at P1 , indicating that a serpin with this main determinant of specificity conserved is present in most, if not all, plants. However, a high degree of conservation was also observed in the P7-P3 region of these 44 serpins. Eighteen of the eudicot genera had a conserved P3 Lys, while a hydrophobic P3 was found in the remaining 26 sequences. Residues in the remaining positions were also conserved (P7-P4 A/S-A/G-V/T-V/I, with only conservative single deviations from this pattern). Generally, genera from the same subfamily or

ch12

Jan. 10, 2007

8:57

SPI-B429


Plant Serpins

289

tribe had identical P8-P1 sequences, as illustrated in Fig. 2. For example, the serpins from the rose and cabbage families (two genera from each, including the Arabidopsis At1g47710 gene product, Table 1) had identical sequences. Similarly, the P8-P1 sequence found in serpins from the three related monocots, maize, sorghum and sugar cane, was identical with that of serpins from the eudicot rue family (two genera). Notably, a single gene coding for a serpin with these properties is present both in the Arabidopsis and rice genomes (Table 1, in bold). The protein encoded by the single serpin gene identified in the C. reinhardtii genome (Table 1, in bold) might also fit into the pattern. With a 16-residue RCL the C. reinhardtii serpin will have the conserved P2-P1 Leu-Arg (Fig. 2). Expression of these three genes has been confirmed at the protein level. The in vitro inhibitory properties of one of the serpins with P2-P1 Leu-Arg-Ser, BSZx from barley, have been characterized in detail.21 Recombinant BSZx was found to be an efficient inhibitor of many serine proteinases of mammalian (non-physiological) origin with trypsin- or chymotrypsin-like specificity at the overlapping reactive center bonds P1-P1 Arg-Ser or P2-P1 Leu-Arg, respectively.

6. Vascular Serpins: The Phloem Conduit of Higher Plants The most abundant serpins in mammals are the extracellular plasma serpins involved in blood coagulation, fibrinolysis and inflammation.3 These functions are not relevant in plants, and plant serpins appear to be intracellular. However, plants contain an extended cell-based vasculature and serpins have been localized to the specialized sieve elements of this system. The conducting tissue of flowering plants is composed of two major systems (Fig. 3): the xylem, which transports water and minerals (salt ions) unidirectionally from the root system to the rest of the plant, and consists of dead cells; and the phloem, which distributes the products of photosynthesis (sugars) and other solutes throughout the plant, and consists of living sieve elements supported by companion cells, which load and unload photosynthate. In addition to non-reducing sugars, it has been shown that proteins, mRNA, small RNA molecules, and hormones are transported long-distance in the phloem, providing inter-organ communication that helps to control plant development and provide defense signaling.29 The sieve elements are

ch12

Jan. 10, 2007

8:57

SPI-B429

290



Fig. 3. Serpin in pumpkin phloem. (A) Immunolocalization of serpin in a cross-section of a pumpkin stem vascular bundle. A polyclonal anti-barley serpin was used and immunogold labeling was visualized in the light microscope by silver enhancement. Arrows indicate detection of serpin in sieve elements only. (B) Enlarged section of A showing detection in cell with sieve plate containing pores (left arrow). Courtesy of Mette la Cour Petersen and Alexander Schulz.

elongated cells stacked in vertical rows and interconnected by sieve plates to form the sieve tubes responsible for the translocation in the phloem. The functional sieve elements have relatively few cytoplasmic organelles, and are connected via numerous plasmodesmata with one or more companion cells. These cells are enriched in cytoplasmic organelles and support the sieve elements with metabolic functions (such as synthesis of proteins and ATP) that were lost or reduced during their specialization for long-distance transport. Mechanisms based on phloem proteins and polysaccharide synthesis may seal off wounded sieve elements; this process is very different from, but with a function similar to, that of blood coagulation. Numerous proteins, including many low-molecular-weight proteinase inhibitors, have been

ch12

Jan. 10, 2007

8:57

SPI-B429


Plant Serpins

291

identified in the phloem sap of many plants.30 Serpins have been detected immunochemically in the phloem of pumpkin12 and barley,6 but the precise cellular localization was not determined. Expression of the pumpkin serpin (CmPS-1) was developmentally regulated, as the level of CmPS-1 in phloem sap increased from >11 nM in 10-day-old seedlings to ∼ 5 µM in 42-day-old plants.12 In contrast, levels of other phloem proteins were found to be more or less invariant with plant age. CmPS-1 was cloned and found to be an efficient serine proteinase inhibitor with elastase-like specificity (P2-P1 Gly-Val-Ser). The study also indicated the presence of a second immunochemically distinct serpin in the pumpkin phloem. In more recent work, a cucumber phloem serpin was cloned and found to belong to the LR serpins with P2-P1 Leu-Arg-Gly.31 Developmental regulation of the expression of this serpin was similar to that found for pumpkin CmPS-1. In the same study, long-distance transport of the cucurbit phloem serpins was confirmed in pumpkin/cucumber heterografting experiments. Immunolocalization experiments detected serpins only in the sieve elements (Fig. 3), whereas other long-distance transported proteins, including a trypsin inhibitor,32 were detected in both sieve elements and companion cells.

7. Target Proteinases: Endogenous or Exogenous? Generally, serpins are inhibitors of serine proteinases with trypsin-like structure and some may inhibit subtilisin-like serine proteinases. However, in several cases, serpin inhibition of cysteine proteinases, with quite different folds but similar mechanisms, has been documented. These proteinases include cathepsins and related papain-like enzymes from plants, as well as caspases.3 Plants contain large numbers of proteinases of different mechanistic classes with key roles in “general housekeeping” (removal of nonfunctional proteins and mobilization of storage proteins to recover amino acids for de novo synthesis), as well as in specific defense responses triggered by pests or pathogens. According to the MEROPS database (http://merops.sanger.ac.uk), the Arabidopsis genome contains 673 known

ch12

Jan. 10, 2007

8:57

SPI-B429

292



or putative peptidase genes (Sept. 2006), mechanistically classified in serine, threonine, aspartic, cysteine, and metalloproteinases.26,33 A number of these proteases are restricted to mitochondria or chloroplasts.26,33 There are very few serine proteinases of the chymotrypsin family (family S1 in clan PA) in the Arabidopsis genome. The DegP2 and SppA1 peptidases are found in chloroplasts and are thus unlikely to be serpin targets. Of the 462 putative peptidases in the rice genome only the DegP2 peptidase is listed as belonging to clan PA. Most of the plant serine proteinases belong to the subtilisin family (Clan SB). Although classified with animal and microbial enzymes shown to be inhibited by some non-plant serpins, plant subtilases have distinct structural features, and inhibition of these proteinases by serpins has not been demonstrated. We investigated the possible inhibition of two plant subtilases: cucumisin from melon mesocarp and hordolisin from germinated barley. Both enzymes have a broad substrate specificity similar to that of bacterial subtilisins, but none of a group of wheat grain serpins with distinct reactive centers was capable of inhibiting these enzymes.8 Although some mammalian serpins can inhibit plant cysteine proteinases, we found no inhibition of papain or chymopapain by the wheat serpins. In barley and wheat, homologous cysteine proteinases are of major importance for the degradation of prolamin storage proteins during germination. The specificity of the barley malt cysteine proteinase EP-B is determined primarily by a large hydrophobic residue at P2 in the substrate, and Gln is among the readily accepted residues at P1.34 Among the cleavage sites for EP-B in the highly repetitive prolamin sequences are PQ↓Q and LQ↓Q; thus, the reactive centers of some of the wheat serpins (Table 2) were expected to be attractive baits for EPB, but none of the wheat serpins did inhibit EP-B at pH 5.0.8 There is evidence for cysteine proteinases with caspase-like structural properties in plants (metacaspases), as well as for inhibition of plant cell death by specific inhibitors that mimic the caspase substrate recognition site; however, caspases specifically involved in programmed cell death remain to be identified.33 We conclude that endogenous target enzymes for plant serpins remain to be identified. Investigations in this direction have been limited so far to tests with a few available proteinases.

ch12

Jan. 10, 2007

8:57

SPI-B429


Plant Serpins

293

8. Functional Hypothesis 1: Defense Against Pest and Disease Organisms Apparently very few serine proteinases of the chymotrypsin family are present in plants and the serpins seem to be intracellular. Consequently, a search for the functions of plant serpins falls outside the areas most central to human serpin research. Wheat serpins did not inhibit the plant subtilases or cysteine proteinases tested.8,10 Neither were selected bacterial or fungal serine proteinases, with either trypsin or subtilisin folds, inhibited by the cereal serpins.7,21 However, almost all plant serpins characterized to date have been inhibitors of animal serine proteinases. Plant seeds contain relatively high amounts of numerous low-molecular-weight protein inhibitors of digestive hydrolytic enzymes, including serine and cysteine proteinases and amylases.35 With very few exceptions, these proteins are inactive against endogenous enzymes, and many studies have supported the generally accepted view that most seed inhibitors are accumulated to defend the developing, resting, or germinating seed against plant pests or pathogens. The storage tissues of specific plant seeds are the preferred feed of many insects or their larvae. Many seeds, including wheat grain, contain specific inhibitors of insect α-amylases that do not affect plant or mammalian amylases. Further evidence for the participation of inhibitors of hydrolytic enzymes in defense against insect attacks comes from in vitro feeding experiments with isolated inhibitors, or with seed lines with different levels of inhibitor expression, as well as from studies using transgenic plants or plants with manipulated production of endogenous inhibitors.35,36 The major seed serpins are most likely to be an integral part of a broad spectrum defense of the seed against attacks on the storage protein reserves, which are essential for growth of the seedling before it can become autotrophic. Many insects have digestive serine proteinases of the trypsin family active in a neutral or alkaline gut, providing conditions under which exogenous serpins could be inhibitory. It is possible that evolutionary adaption of the proteolytic digestion system of some insects to feed efficiently on cereal storage proteins has occurred. In the three related cereals, wheat, rye and barley, Gln-and Prorich prolamin storage proteins constitute approximately half of the total grain protein. These prolamins are rich in octa- or hepta-peptide repeats

ch12

Jan. 10, 2007

8:57

SPI-B429

294



containing Gln-Gln sequences.35 For example, prolamins with repetitive sequences based on the octapeptide consensus motif PQQPFPQQ are abundant, and often repeats with a sequence of three Gln residues are found. Reactive centers of cereal grain serpins rich in Gln residues (Table 2) are clearly reminiscent of the Gln-rich repeats of the prolamins.8,9 Several studies have shown hypervariability of the residues of serpin RCLs.37,38 For example, the presence of multiple α1 -proteinase inhibitor variants with distinct inhibitory specificity in some animals was suggested to be the result of accelerated evolution of RCL sequences to match the specificity of digestive proteinases secreted by infectious parasites or predators.39 Similarly, the Gln-rich RCLs of serpins present in wheat, rye and barley grain, but not found in other plant seeds, may reflect an evolutionary adaption to defend the plant against digestive proteinases of specific seed-eating insects, adapted to the ingestion of prolamin storage proteins as their main amino acid source. The prolamins of rice are different from those of the Triticeae cereals and lack the typical Gln- and Pro-rich repeats, although single Glnrich octapeptides are found in their structures.35 Although very little is known about seed expression of rice serpins, the rice genome does not contain genes for serpins with Gln-rich reactive centers (Table 2), similar to those found in the Triticeae cereals (Table 2). In the cereal oat, the dominant storage proteins are globulins without repetitive structures.35 Together the oat grain serpins, with either Thr, Arg, or Tyr at P1, represent a broader inhibitory specificity against digestive serine proteinases than those found in other cereals; pancreas elastase, trypsin, and chymotrypsin were efficiently inhibited by the oat grain serpin complement.11 Barley grain contains major serpins with P2-P1 Asp-Arg-Ser or Glu-Arg-Ser, and similar serpins were found in apple seeds, the first serpins of eudicot seeds studied.40 These serpins could be the result of adaptation to inhibition of serine proteinases from pests or pathogens with a trypsin-like specificity and a preference for an acidic residue at P2. Although plant serpins are inactivated at the acidic stomach pH of mammals, the serpins may participate in protection of whole seeds against serine proteinases active in parts of the digestive system of other seed-eating animals (including birds), contributing to seed dispersal. Low-mass enzyme inhibitors, expressed constitutively or after induction, in vegetative tissues of plants undoubtedly have similar defense

ch12

Jan. 10, 2007

8:57

SPI-B429


Plant Serpins

295

roles in these tissues, although they are present in much lower concentrations than in seeds.36 Although serpins are expressed in leaves, roots and other vegetative tissues, direct evidence for their involvement in plant defense is very limited. Generally, genes related to stress and defense are responsive to insect or mechanical damage of vegetative tissues. Many proteinase inhibitors, but not serpins, have been identified among the proteins that were induced or upregulated in a range of mostly eudicot plants.41 The majority of the 45 proteins identified thus far in cucurbit phloem exudate appear to be involved in stress and defense reactions.30 A clear correlation was found between the developmentally increased concentration of serpin in pumpkin phloem and the survival of pierce-sucking aphids feeding on phloem sap.12 However, a direct effect on aphid survival in feeding experiments with the cloned and purified serpin was not observed. A single set of studies on transgenic plants may support the defense hypothesis for vegetative tissues: natural and engineered variants of a serpin from the hemolymph of the insect Manduca sexta, expressed in the leaves of alfalfa, cotton and tobacco, were in some experiments shown to limit plant damage after challenging the herbivorous insects.42–44

9. Functional Hypothesis 2: Endogenous Proteinase Regulation It appears quite likely that serpin genes related to pest or disease resistance would have evolved rapidly. The presence of serpins that together represent a distinct inhibitory repertoire in the seeds of a specific plant species (or closely related group) supports roles in defense against a complement of more-or-less species-specific pests or pathogens (as discussed in the previous section). In contrast, the apparent presence of a single serpin with the highly conserved reactive center P2-P1 Leu-Arg-Ser/Gly (LR serpins) in various vegetative tissues of most (if not all) plants suggests an endogenous (“housekeeping”) role of this serpin. In particular, it seems more likely that the single serpin found in the unicellular alga C. reinhardtii, apparently an LR serpin, is involved in the regulation of specific proteolytic events in a conserved fundamental process than in broad-spectrum defense. Thus, in spite of the differences in proteinase complement between plants and animals, the LR serpins appear to be the probable candidates for a role

ch12

Jan. 10, 2007

8:57

SPI-B429


296


in regulatory inhibition of one or more specific proteinases, such as most animal serpins. It is also probable that a plant serpin may directly share functions with one or more intracellular serpins in animals; for example, a role related to the regulation of programmed cell death (plant cells do not undergo apoptosis as such). In higher plants, the LR serpins might participate in the protection of surrounding tissues from proteolysis that occurs during development, such as in the formation of tracheal elements or in the hypersensitive response. The latter function would be analogous to the function of mammalian serpins that protect adjacent tissue against proteinases released during inflammation. The presence and long distance transport of serpins that may be related to the LR serpins in the phloem sieve elements may indicate roles in the functional protection of this circulatory system, while suggestions of a direct involvement in inter-organ communication would be speculative.

10. Conclusions Identification of endogenous or exogenous target proteinases leading to the discovery of physiological functions of plant serpins expressed in distinct tissues or cells will be a major challenge in future plant serpin research. The first hypothesis presented in this chapter that distinct serpins are essential components of a broad-spectrum plant defense system, is not easy to support rigorously in experiments. However, results of many different approaches, including heterologous expression of genes in plant tissues, have supported defensive roles of other proteinase inhibitors (problems with interpretation of such studies have been discussed.35,36 ) Ideally, the evaluation of defense benefits should be based on challenge experiments comparing plants/tissues expressing a single endogenous serpin gene to different levels, keeping all other defense traits and their signaling intact.36 Only recently was it concluded from experiments along this line that a tobacco leaf trypsin inhibitor was effective in defense against herbivorous insects native to the plant.36 We are currently focusing on the second hypothesis that a cellular function of the single serpin in C. reinhardtii is conserved and possibly expanded in higher plants. As a first step, the LR serpins from C. reinhardtii and A. thaliana (Table 1) are being produced heterologously to confirm their inhibitory properties, in particular the proposed 16-residue RCL of the

ch12

Jan. 10, 2007

8:57

SPI-B429


Plant Serpins

297

C. reinhardtii serpin.4 For both organisms, we have started to examine gene expression and the effects of gene knockouts on growth, morphology and the proteome. We will attempt to identify serpin-proteinase complexes, and will study serpin localization at the tissue and cellular level.

Note The first inhibition of a plant proteinase by a plant serpin has now been demonstrated.45 A yeast two-hybrid screen for Arabidopsis thaliana proteins that interact with the cysteine proteinase metacaspase 9 from this species (AtMC9) yielded the LR serpin encoded by Atlg47710 (Table 1; named AtSerpin1 by the authors). Recombinant AtSerpin1 was shown to inhibit AtMC9 in vitro, forming SDS-stable 1:1 complexes. Metacaspases are most probably not responsible for caspase-like activities in plants, but may act upstream of Asp-specific proteinases.

References 1. Irving JA, Steenbakkers PJM, Lesk AM et al. (2002) Serpins in prokaryotes. Mol Biol Evol 19:1881–1890. 2. Silverman GA, Bird PI, Carrell RW et al. (2001) The serpins are an expanding superfamily of structurally similar but functionally diverse proteins — Evolution, mechanism of inhibition, novel functions, and a revised nomenclature. J Biol Chem 276:33293–33296. 3. Gettins PGW (2002) Serpin structure, mechanism, and function. Chem Rev 102:4751–4803. 4. Roberts TH, Hejgaard J, Saunders NFW, Cavicchioli R and Curmi PMG (2004) Serpins in unicellular Eukarya, Archaea, and Bacteria: Sequence analysis and evolution. J Mol Evol 59:437–447. 5. Irving JA, Pike RN, Lesk AM and Whisstock JC (2000) Phylogeny of the serpin superfamily. Implications of amino acid conservation for structure and function. Genome Res 10:1845–1864. 6. Roberts TH, Marttila S, Rasmussen SK and Hejgaard J (2003) Differential gene expression for suicide-substrate serine proteinase inhibitors (serpins) in vegetative and grain tissues of barley. J Exp Bot 54:2251–2263. 7. Dahl SW, Rasmussen SK and Hejgaard J (1996) Heterologous expression of three plant serpins with distinct inhibitory specificities. J Biol Chem 271:25083–25088.

ch12

Jan. 10, 2007

8:57

SPI-B429

298



8. Ostergaard H, Rasmussen SK, Roberts TH and Hejgaard J (2000) Inhibitory serpins from wheat grain with reactive centers resembling glutamine-rich repeats of prolamin storage proteins — Cloning and characterization of five major molecular forms. J Biol Chem 275:33272–33279. 9. Hejgaard J (2001) Inhibitory serpins from rye grain with glutamine as P1 and P2 residues in the reactive center. FEBS Lett 488:149–153. 10. Hejgaard J (2005) Inhibitory plant serpins with a sequence of three glutamine residues in the reactive center. Biol Chem 386:1319–1323. 11. Hejgaard J and Hauge S (2002) Serpins of oat (Avena sativa) grain with distinct reactive centres and inhibitory specificity. Physiol Plant 116:155–163. 12. Yoo BC, Aoki K, Xiang Y et al. (2000) Characterization of Cucurbita maxima phloem serpin-1 (CmPS-1) — a developmentally regulated elastase inhibitor. J Biol Chem 275:35122–35128. 13. Grabar P (1957) Etudes Immunochimiques sur la Biere. In Proceedings of the 6th Congress of the European Brewing Convention, pp. 147–154. 14. Hejgaard J and Sorensen SB (1975) Characterization of a protein-rich beer fraction by two-dimensional immunoelectrophoretic techniques. Comptes Rendus des Travaux du Laboratoire Carlsberg 40:187–203. 15. Hejgaard J (1976) Free and protein-bound beta-amylases of barley grain: Characterization by two-dimensional immunoelectrophoresis. Physiol Plant 38:293–299. 16. Prouse CV and Esnouf MP (1977) The isolation of a new warfarin-sensitive protein from bovine plasma. Biochem Soc Trans 5:255–256. 17. Hejgaard J (1984) Gene products of barley chromosomes 4 and 7 are precursors of the major antigenic beer protein. J Inst Brew 90:85–87. 18. Hejgaard J, Rasmussen SK, Brandt A and Svendsen I (1985) Sequence homology between barley endosperm protein Z and protease inhibitors of the alpha-1-antitrypsin family. FEBS Lett 180:89–94. 19. Brandt A, Svendsen I and Hejgaard J (1990) A plant serpin gene. Structure, organization and expression of the gene encoding barley protein Z4. Eur J Biochem 194:499–505. 20. Rosenkrands I, Hejgaard J, Rasmussen SK and Bjorn SE (1994) Serpins from wheat grain. FEBS Lett 343:75–80. 21. Dahl SW, Rasmussen SK, Petersen LC and Hejgaard J (1996) Inhibition of coagulation factors by recombinant barley serpin BSZx. FEBS Lett 394:165–168. 22. Sander I, Flagge A, Merget R et al. (2001) Identification of wheat flour allergens by means of 2-dimensional immunoblotting. J Allergy Clin Immunol 107:907–913.

ch12

Jan. 10, 2007

8:57

SPI-B429


Plant Serpins

299

23. Herzinger T, Kick G, Ludolph-Hauser D and Przybilla B (2004) Anaphylaxis to wheat beer. Ann Allergy Asthma Immunol 92:673–675. 24. Garcia-Casado G, Crespo JF, Rodriguez J and Salcedo G (2001) Isolation and characterization of barley lipid transfer protein and protein Z as beer allergens. J Allergy Clin Immunol 108:647–649. 25. The Arabidopsis Genome Initiative (2000) Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408:796–815. 26. Zimmermann P, Hirsch-Hoffmann M, Hennig L and Gruissem W (2004) Genevestigator: Arabidopsis microarray database and analysis toolbox. Plant Physiol 136:2621–2632. 27. Giavalisco P, Nordhoff E, Kreitler T et al. (2005) Proteome analysis of Arabidopsis thaliana by two-dimensional gel electrophoresis and matrixassisted laser desorption/ionisation-time of flight mass spectrometry. Proteomics 5:1902–1913. 28. KollerA, Washburn MP, Lange BM et al. (2002) Proteomic survey of metabolic pathways in rice. Proc Natl Acad Sci USA 99:11969–11974. 29. Lough TJ and Lucas WJ (2006) Integrative plant biology: role of phloem long-distance macromolecular trafficking. Annu Rev Plant Biol 57:203–232. 30. Walz C, Giavalisco P, Schad M et al. (2001) Proteomics of curcurbit phloem exudate reveals a network of defence proteins. Phytochemistry 65:1795–1804. 31. la Cour Petersen M, Hejgaard J, Thompson GA and Schulz A (2005) Cucurbit phloem serpins are graft-transmissible and appear to be resistant to turnover in the sieve element-companion cell complex. J Exp Bot 56:3111–3120. 32. Dannenhoffer JM, Suhr RC and Thompson GA (2001) Phloem-specific expression of the pumpkin fruit trypsin inhibitor. Planta 212:155–162. 33. Schaller A (2004) A cut above the rest: the regulatory function of plant proteases. Planta 220:183–197. 34. Davy A, Svendsen I, Sorensen SO et al. (1998) Substrate specificity of barley cysteine endoproteases EP-A and EP-B. Plant Physiol 117:255–261. 35. Shewry PR, Casey SR (eds.) (1999) Seed Proteins. Kluwer Academic: Dordrecht. 36. Zavala JA, Patankar AG, Gase K, Hui DQ and Baldwin IT (2004) Manipulation of endogenous trypsin proteinase inhibitor production in Nicotiana attenuata demonstrates their function as antiherbivore defenses. Plant Physiol 134:1181–1190. 37. Hill RE and Hastie ND (1987) Accelerated evolution in the reactive centre regions of serine protease inhibitors. Nature 326:96–99. 38. Ohta T (1994) On hypervariability at the reactive center of proteolytic enzymes and their inhibitors. J Mol Evol 39:614–619.

ch12

Jan. 10, 2007

8:57

SPI-B429

300



39. Goodwin RL, Baumann H and Berger FG (1996) Patterns of divergence during evolution of alpha 1-proteinase inhibitors in mammals. Mol Biol Evol 13:346–358. 40. Hejgaard J, Laing WA, Marttila S, Gleave AP and Roberts TH (2005) Serpins in fruit and vegetative tissues of apple (Malus domestica): Expression of four serpins with distinct reactive centres and characterization of a major inhibitory seed form, MdZ1b. Funct Plant Biol 32:517–527. 41. Ryan CA (1990) Proteinase inhibitors in plants: Genes for improving defenses against insects and pathogens. Annu Rev Phytopathol 28:425–449. 42. Thomas JC, Adams DG, Keppenne VD et al. (1995) Protease inhibitors of Manduca sexta expressed in transgenic cotton. Plant Cell Rep 14:758–762. 43. Thomas JC, Adams DG, Keppenne VD et al. (1995) Manduca sexta encoded protease inhibitors expressed in Nicotiana tabacum provide protection against insects. Plant Physiol Biochem 33:611–614. 44. Thomas JC, Wasmann CC, Echt C et al. (1994) Introduction and expression of an insect proteinase inhibitor in alfalfa (Medicago sativa L.). Plant Cell Rep 14:31–36. 45. Vercammen D, Belenghi B, van de Cotte B, Beunens T, Gavigan J-A, De Rycke R, Brackenier A, Inzé D, Harris JL and Van Breusegem F (2006) Serpin1 of Arabidopsis thaliana is a suicide inhibitor for metacaspase 9. J Mol Biol (in press).

ch12

Jan. 10, 2007

8:57

SPI-B429


13 Serpins, Apoptosis and Other Aspects of Cell Death Fiona L. Scott

1. Introduction to Programmed Cell Death and Apoptosis Maintenance of tissue and organism size in metazoans is achieved through a balance between cell proliferation and cell elimination. The concerted removal of cells in the developing organism is exquisitely timed and is usually achieved by the activation of a cellular suicide program. The term “apoptosis” from Greek, meaning “the dropping of leaves from a tree,” was coined in 1972 to describe the classic morphological features seen during this orchestrated dismantling of the cell and packaging into fragments for efficient removal by neighboring phagocytes.1 Apoptosis is evolutionarily conserved from worms to flies to mammals and many of the genes involved have been identified.2–6 Crucial control of apoptosis must be maintained in order to prevent pathologies that arise from excessive or insufficient cell death.7 In the absence of cell replenishment from proliferation, ischemic injury, AIDS, and neurodegenerative disorders are associated with too much cell death.8,9 A deficit in cell death is involved in the pathogenesis of autoimmune diseases, and most obviously, in cancer.10–13 Components of the apoptotic cellular machinery have become attractive targets for drug design and have far reaching potential in the treatment of many chronic and acute diseases with a cell death component.14 301

ch13

Jan. 10, 2007

8:57

SPI-B429


302

F. Scott

2. The Caspase Family At the center of the apoptotic process is the C14 family of the clan CD cysteine proteases, i.e. the caspases.15 Although the CD clan has other families that exist in all kingdoms of life (except viruses), caspases are found only in metazoans with the earliest members in nematodes. The 1990s saw an explosion of knowledge within the field of cell death research. It all began with the identification of the human protease responsible for the proteolytic maturation of interleukin-1β (IL-1β), interleukin-1β converting enzyme (ICE/caspase-1). At the same time, extensive work on the nematode Caenorhabditis elegans identified the CED3 gene product as being essential for cellular commitment to apoptosis and having 29% identity with ICE.16 These exciting initial findings provoked many researchers to search for other cysteine proteases with specificity for aspartic acid residues involved in cell death. It is now recognized that there are 11 human caspases, of which seven are agreed to be involved in apoptosis. These apoptotic caspases can be further divided into those responsible for the initiation of apoptosis (caspases-2, -8, -9, and -10) or the execution of apoptosis (caspases-3, -6, and -7). The others are either involved in inflammation as cytokine activators (caspases-1, -4, and -5), or they played a role in keratinocyte differentiation (caspase-14). Even though these enzymes are involved in different cellular processes, they all have strict requirement for aspartic acid in the P1 position of both their physiological substrates and within small man-made peptidic substrates. The initiator caspases are the first to be activated in response to insult, transmitting a death-inducing signal into proteolytic activity. They are characterized by the presence of large amino-terminal adaptor domains such as caspase-recruitment domain (CARD) or death effecter domain (DED) that facilitate their recruitment to large multi-component complexes upon a death trigger. Activation of the initiator zymogens at these oligomeric complexes is mediated by dimerization, and subsequent proteolytic processing stabilizes the active enzyme.17–19 There are two main pathways to initiate caspase activation. The “intrinsic pathway” activates caspase-9 in response to stress, DNA damage and developmental cues (Fig. 1). This pathway can be thought of as a sensor for intracellular damage and is initiated at the apoptosome — a large protein complex in the cytoplasm. The apoptosome comprises apoptotic protease activating factor 1(Apaf1), dATP, and

ch13

Jan. 10, 2007

8:57

SPI-B429


Serpins, Apoptosis and other Aspects of Cell Death

303

Fig. 1. Pathways to apoptosis. The extrinsic pathway is activated by ligation of death receptors at the cell surface. Monomeric caspase-8 is recruited to the death inducing signaling complex (DISC) by adaptor proteins and is activated by dimerization. The intrinsic pathway is activated in response to cell stress. Cytochrome C is released from the mitochondrial intermembrane space and is a cofactor for apoptosome formation. Monomeric caspase-9 is activated by dimerization at the apoptosome. Once the initiators are activated, caspase-8 and-9 proteolytically activate the executioner caspase-3, -6, and -7, which then cleave cellular death substrates, resulting in apoptosis. The cytotoxic lymphocyte pathway is activated after delivery of granzymes to the target cell. Granzymes may directly activate executioner caspases, cleave bid to engage the intrinsic pathway, or directly cleave death substrates. It is less clear how the lysosomal pathway is initiated but may involve death receptors, lipid mediators, or photo damage. Lysosomal cathepsins in the cytoplasm may cleave bid to engage the intrinsic pathway or may cleave other death substrates. Protease inhibitors that mediate cell death, including intracellular serpins proposed to play a role, are indicated.

cytochrome C. Assembly of the apoptosome only occurs after cytochrome C is released from the mitochondria in response to cellular stress. Caspase-9 is then recruited to this death-signaling complex by virtue of a homotypic interaction between its amino-terminal (CARD) and the CARD of Apaf1.

ch13

Jan. 10, 2007

8:57

SPI-B429

304


F. Scott

The “extrinsic pathway” activates caspase-8 and is triggered by ligation of some members of the tumor necrosis factor (TNF) transmembrane receptor family, a sensor for extracellular signals (Fig. 1). Although activated in response to different cues, caspases-8 and -9 converge to activate the executioner caspases, caspase-3, -6, and –7. The executioners have a short prodomain and exist as preformed, inactive dimeric zymogens in resting cells. Using a mechanism reminiscent of classical protease activation, the executioner caspase zymogens are activated by interchain proteolysis. Executioner caspases are usually activated by the initiator caspases, resulting in the amplification of the apoptotic signal in a cascade-like manner. In addition, some other immune cell derived proteases (discussed in detail below) can also perform this function. Most of the limited proteolysis required to produce the classic apoptotic morphology changes are due to the actions of executioner caspases. Reference 20 shows a current list of known death substrates.

3. Serpins as Caspase Inhibitors This two-step (at minimum) cascade of caspase activation not only amplifies the death signal, but also allows for a point of regulation before commitment to death. Caspase regulators have evolved in both mammals and viruses to mediate cell death. Inhibitor of apoptosis proteins (IAPs) are the only endogenous mammalian caspase inhibitors known to date and target caspase-3, -7, and -9. In contrast, P35 from baculovirus is a very potent inhibitor of all known caspases. Orthopox viruses contain many genes that are involved in protecting the virus from the hosts’ anti-viral defense mechanisms, including apoptosis inhibitors.21 The best studied of these viruses, the cowpox virus, has three serpin genes: cytokine response modifier A (CrmA/SPI-2), SPI-1 (B24R) and SPI-3 (K2L). The rising interest in CrmA began when Gagliardini observed that microinjection of CrmA into chicken neurons provides protection from death induced by growth factor withdrawal.22 This sparked a frenzy of publications that examined the protective effect CrmA had on researcher’s favorite apoptosis model system. This original observation placed CrmA as a modulator of the cell stress, intrinsic death pathway. Although pivotal at the time, many groups have subsequently demonstrated

ch13

Jan. 10, 2007

8:57

SPI-B429



305

that, as a rule, CrmA does not modulate the intrinsic cell death pathway in many cell types killed with many different intrinsic apoptotic stimuli from kinase inhibitors to DNA damaging agents.23–28 Bcl2 family members keep this pathway in check. Importantly, CrmA can block apoptosis induced by all extrinsic pathway stimuli examined [soluble Fas ligand, anti-Fas agonistic antibody, TNF, and TNF-related apoptosis inducing ligand (TRAIL)] in fibroblasts, epithelial cells, lymphocytes and neurons.29–34 It can also protect epithelial and endothelial cell death induced by detachment from the extracellular matrix, also known as anoikis,35 a process thought to involve some components of the extrinsic apoptosis machinery (see Ref. 36 for review). CrmA orthologue (SPI-2) from ectromelia virus and vaccinia virus share 92% identity with cowpox CrmA. Transfection with ectromelia SPI-2 protects cells from TNF-induced apoptosis,37 while transfection of vaccinia SPI-2 or infection with vaccinia virus expressing SPI-2 protects from Fas and TNF-induced cell death.38,39 In addition, infection with cowpox or rabbitpox virus deleted in SPI-2 no longer protects cells from Fas-mediated cell death.40 So how does CrmA inhibit apoptosis? The anti-apoptotic function of ectopically expressed CrmA was shown to be completely dependent on a functional reactive center loop (RCL), confirming that it inhibits a protease.41 In addition, the unusual Asp residue at the P1 position is also essential.42 With that, CrmA has evolved the perfect bait with which to hook aspartic acid-specific proteases of the apoptotic machinery. Of the cell death related caspases, CrmA directly inactivates caspase-8 most efficiently with a kass = 9.1 × 104 M−1 s−1 and Ki =< 0.3 − 0.95 nM.43,44 The potency of this interaction and the placement of caspase-8 as the initiator of the extrinsic pathway explain why CrmA is able to block all caspase-8 dependent apoptosis pathways. In addition, CrmA inhibits caspase-9, the initiator caspase of the intrinsic pathway, most potently with a Ki ≤ 2 nM.44,45 This inhibition constant is unlikely to allow for a physiological interaction between caspase-9 and CrmA in virally infected cells, and is reflected by the finding that CrmA expression does not block most caspase-9 dependent cell death stimuli. We assume that the original studies placing CrmA as an inhibitor of serum withdrawal induced cell death probably resulted from massive over-expression of CrmA or microinjection of quantities greater than ever seen during cowpox viral infection.

ch13

Jan. 10, 2007

8:57

SPI-B429

306


F. Scott

Cowpox viruses that lack CrmA are unable to suppress the inflammatory response in a chicken embryo infection model, although they are still lethal.46 This anti-inflammatory function is primarily through the inhibition of inflammatory caspase-1 (kass = 1.7 × 107 M−1 s−1 and Ki ≤ 0.01 nM) and to a lesser extent, caspase-4 (Ki = 1.1 nM) and caspase-5 (Ki < 0.1 nM; Refs. 43, 44, 47). Rabbitpox and vaccinia SPI-2 also inhibits caspase1.39,48 These inflammatory caspases do not play a role in apoptosis or cell death. Instead they are processing enzymes that activate cytokines. Caspase-1 is the protease that converts pro-IL-1β and pro-interleukin-18 to their mature, secreted forms.46,49,50 Consequently, virally infected host cells expressing CrmA have depressed caspase-mediated cytokine processing, suppressing the inflammatory response and thereby facilitating viral replication and further infection. It seems logical for virus to acquire as many properties as possible to promote host cell survival during infection. The anti-inflammatory function of CrmA may initially be the most important as it prevents the recruitment of immune cells to the site of infection. CrmA inhibition of caspase-8 dependent Fas and TNF-induced apoptosis is also crucial for the etiology of virus infection in that it prevents the infected host cell from being killed by immune cells, before virus can replicate. Until the biochemical characterization of CrmA, serpins were considered to be strictly serine protease inhibitors. CrmA inhibition of cysteine protease family members was the first demonstration of cross-class inhibition, where a serpin has adapted to inhibit proteases that utilize either cysteine or serine residues as the catalytic nucleophile. This characteristic has now been demonstrated for a number of related human intracellular Clade B serpins (discussed below). Although it is a strong inhibitor of caspase-1 and -8, CrmA is a very poor inhibitor of executioner caspase-3 (Ki = 500–1600 nM), despite their structural and functional similarity.43,44 This poor inhibition is not only due to suboptimal specificity of the RCL sequence of CrmA for caspase-3 — the P4 –P1 of the CrmA RCL is Leu-Val-Ala-Asp, while caspase-3 prefers Asp-Glu-Val-Asp. Molecular modeling studies predict major structural clashes between the specificity determining 381 loop of caspase-3 with surface residues of CrmA, preventing the RCL from gaining access to the

ch13

Jan. 10, 2007

8:57

SPI-B429



307

substrate binding site of caspase-3.51 No such clashes were evident when caspase-1 or -8 were modeled with CrmA. In agreement, in an in vitro yeast survival assay, ectromelia SPI-2 inhibits caspase-1 and -8 but not caspase-3 or -6.37 The only other serpin suggested to be a caspase inhibitor is SERPINB9 (PI9/cytoplasmic antiproteinase 3/CAP3). SERPINB9 is the only human serpin with an acidic residue at the P1 position of its RCL.52,53 It is also found in both the cytoplasm and nucleus of many cell types, placing it in the appropriate subcellular compartment to be a caspase regulator.53,54 CrmA and SERPINB9 also have similar RCL sequences — P4 –P1 residues are Leu-Val-Ala-Asp-Cys and Val-Val-Ala-Glu-Cys, respectively. A notable difference is that SERPINB9 contains a glutamic acid at P1 instead of an aspartic acid, the preferred P1 residue for caspases. The second-order rate constant for SERPINB9 with caspase-1, -4, and -8 (< 10 × 103 M−1 s−1 ) is greater than 10,000-fold weaker than CrmA inhibition of caspase-1.55 SERPINB9 and its mouse orthologue are also ineffectual against Fas mediated apoptosis.56,57 In contrast, a Glu to Asp mutant at P1 had improved inhibitor potency against most caspases and was as effective as CrmA at protecting cells from Fas-induced cell death. This supports the general view that SERPINB9 does not regulate apoptosis through the inhibition of caspases. Despite its low potency for caspase-1, there is an inverse correlation between SERPINB9 expression and mature IL-1β in atherosclerotic plaques, implying a possible role for SERPINB9 in inflammation. There is much stronger evidence from numerous studies that SERPINB9 plays an important role in the regulation of granzyme B (GraB) in cytotoxic lymphocytes, antigen presenting cells, and at sites of immune privilege (discussed below).

4. Serpins as Cell Death Regulators A number of studies have addressed putative roles for serpins in cell death and tumorigenesis. Here, we will focus on the involvement of serpins in the extrinsic and intrinsic apoptotic pathways, lysosomal cell death pathway, and cytoprotection in cytotoxic lymphocyte and granulocyte populations.

ch13

Jan. 10, 2007

8:57

SPI-B429

308


F. Scott

Where relevant, the interplay between cell death and cancer will be discussed; however, for a detailed review of serpins in malignancy, we refer readers to Chaps. 14 and 15 of this book.

5. Regulation of Extrinsic Apoptosis Pathway The extrinsic apoptotic pathway is triggered by the engagement of death receptors of the TNF-transmembrane receptor family. Many of the Clade B serpins have been reported to provide a cytoprotective function when upregulated or ectopically expressed in cells triggered to die with TNF-α. Intracellular SERPINB2 (plasminogen activator inhibitor 2, PAI2) protects cells from TNF-α induced cell death, while antisense-mediated downregulation of SERPINB2 sensitized cells.58,59 Both the P1Arg residue of the RCL and the unusually long C-D interhelical loop are required for this protection.59,60 The loop between helices C and D is a common feature of the intracellular Clade B serpins coded by genes clustered at the chromosome 18q21.3 locus. It varies in both length and amino acid composition between serpins.61 It is longest in SERPINB2 at 33 residues and can be a substrate for transglutaminase. This is implicated in the covalent crosslinking of SERPINB2 to large complex structures in trophoblast membranes and its incorporation into the cornified envelopes of differentiated keratinocytes.62–64 No cross-linking of SERPINB2 to the cell membrane of apoptotic cells was observed, despite reports of increased transglutaminase activity during apoptosis. Another study has shown that the C-D loop of SERPINB2 reversibly binds some annexins, in a calcium independent manner, and other unidentified proteins.65 An interesting observation is that SERPINB2 can bind retinoblastoma protein (Rb) via its C-D loop, preventing Rb degradation by an unknown protease.66 Finally, the C-D loop acts as a redox sensor, controlling SERPINB2 polymerization. In the reducing environment of the cytosol where it is predominantly localized, SERPINB2 is in a functional monomeric conformation. When forced into the secretory pathway, SERPINB2 is a non-functional polymer.67 It is unclear how any of these C-D loop properties convey an anti-apoptotic phenotype and warrants further study. Although TNF-α also induced a moderate accumulation of urokinasetype plasminogen activator (uPA) protein in cells, a SDS-stable complex

ch13

Jan. 10, 2007

8:57

SPI-B429



309

was not detected between SERPINB2 and uPA. In addition, a role for uPA, an extracellular protease that activates the plasminogen-plasmin fibrinolytic cascade in TNF-α induced apoptosis was excluded by the use of neutralizing antibodies.59 Unfortunately, no protease-SERPINB2 complex has been detected, although the reduction of SERPINB2 protein with prolonged TNF-α may be due to proteasome-mediated removal of SERPINB2protease complexes.58 SERPINB3 (squamous-cell carcinoma antigen-1, SCCA-1) also provides some resistance to TNF-α induced cell death in squamous carcinoma cell lines, through either endogenous expression or by retroviral transduction of the SERPINB3 cDNA.68 Cells transfected with SERPINB4 (squamous-cell carcinoma antigen-2, SCCA2) were protected from TNFα induced apoptosis, whilst a RCL mutant that cannot inhibit cathepsin G (CatG) in vitro (P3 –P2 : SERPINB4 = Val-Glu-Leu-Ser-Ser; RCL mutant = Val-Gln-Gly-Ala-Ser) was ineffective.69 SERPINB4 does not inhibit caspase-3 or -7 directly yet both SERPINB3 and SERPINB4 expressing cells produce less executioner enzymatic activity in response to TNF-α. This suggests that the protease, or proteases, regulated by these inhibitors act upstream of executioner caspase activation.69 Finally, ectopically expressed SERPINB10 (bomapin/PI10) protects cells from TNF-α in a dose-dependent manner.70 Although SERPINB10 is restricted to bone marrow and hemopoietic cells of the monocytic lineage, a portion of ectopically expressed SERPINB10 is part of a high-molecular weight complex in the cytosolic fraction of Hela cells. This complex is assumed to contain a protease of greater than 100 kDa in molecular weight and is enriched when Hela cells are treated with TNF-α in the presence of a proteasome inhibitor.

6. Regulation of Intrinsic Apoptosis Pathway The potential role of SERPINB2 in the regulation of the intrinsic apoptotic pathway may depend on the cell stress that initiates the response. SERPINB2 does not protect Hela cells from UV or ionizing radiation.59 It also does not protect gastric cancer cells from etoposide, but does marginally protect these cells from Helicobacter pylori induced apoptosis.71 The mechanism for this is unclear; however, it also correlates with a decrease in the migration of H. pylori infected cells, suggesting the

ch13

Jan. 10, 2007

8:57

SPI-B429

310


F. Scott

involvement of the cell surface uPA-plasmin system. It is unknown whether the intracellular pool of SERPINB2 or whether an extracellular pool, a consequence of facultative secretion, is responsible for this phenomenon.72 In a chemical model for papilloma-skin carcinogenesis, transgenic mice expressing SERPINB2 in basal keratinocytes were highly susceptible to papilloma lesions, compared with wild-type mice. Upon cessation of treatment, papillomas in wild-type mice regressed and were highly apoptotic, whilst papillomas in SERPINB2 transgenic mice did not regress, were not apoptotic and some even became carcinomas.73 SERPINB2 is normally located in the suprabasal layers of the epidermis. It is not understood how the deregulation of SERPINB2 expression in the epidermis blocks apoptosis in this disease model; however, uPA is not involved. SERPINB3 and SERPINB4 have a clearer capacity for intrinsic cell death pathway regulation. A chemotherapeutic topoisomerase I inhibitor (SN-38) induces apoptosis in a SERPINB3-negative human head and neck carcinoma cell line. When transduced with SERPINB3, they were less susceptible to SN-38 induced cell death.68 In agreement with this finding, antisense-mediated down-regulation of SERPINB3 in a cervical squamous carcinoma cell line increases its sensitivity to etoposide.68 In addition, both SERPINB3 and SERPINB4 protect cells from gamma irradiation, possibly through the suppression of the mitogen-activated protein kinase pathway.74 Finally, when SERPINB13 (hurpin/headpin/PI13) is ectopically expressed under a heterologous promoter in keratinocytes, it protects them from UV induced apoptosis, as measured by executioner caspase activity.75 It is currently unknown what proteases these serpins might regulate in the extrinsic or intrinsic cell death pathways. Based on the RCL sequence, none are expected to target caspases. In fact, SERPINB4 is the only one with an acidic amino acid within its RCL (P2 Glu) and does not inhibit executioner caspases.69 There is mounting evidence that cysteine proteases of the calpain and lysosomal cathepsin families (see later) may play a role in cell death.76–78 Considering the broad inhibitory profile of these intracellular serpins, it is unlikely that they target the same protease or group of proteases. SERPINB2 and SERPINB10 have a P1Arg residue, yet inhibiting different tryptic-like proteases. SERPINB3 and SERPINB4, although 92% identical throughout the entire protein sequence, vary dramatically in their

ch13

Jan. 10, 2007

8:57

SPI-B429



311

RCL. SERPINB3 and SERPINB13 inhibit lysosomal cysteine proteases while SERPINB4 inhibits the chymotryptic-like proteases, including chymase and CatG.79,80 CatG is normally confined to polymorphonuclear leukocytes, yet an immunoreactive species was detected by CatG antibody in TNF-α treated epithelial adenosarcoma cells.69 It has also been reported that CatG relocalizes from azurophilic granules to the cytoplasm and nucleus of apoptotic promyelocytic leukemia cells where it can cleave Brm.81 Brm forms a repressor complex with Rb and histone deacetylase, regulating chromatin conformation and subsequent gene transcription. CatG can also directly activate caspase-7 in vitro.82 It is unknown whether the expression of CatG, and other spatially restricted proteases, can be induced to participate in apoptosis in cells and tissues where they are not normally found. It could also be argued that Clade B serpins have adapted other functions aside from protease inhibition for apoptosis regulation. This would not be supported by the observation that the RCL is required in many of these cases. Alternatively, the RCL of these intracellular serpins may have adopted another function in addition to being bait for proteases. The finding that chemical inhibitors of chymotrypsin-like proteases (TPCK) and trypsin-like proteases (TLCK) have differential effects on both extrinsic and intrinsic death pathways, depending on the cell type, does not help to clarify the issue. Identification of the binding partners regulated by these serpins, be they proteases or other cellular components, will provide a better understanding of the protease and inhibitor networks that operate in cell survival and apoptosis.

7. Lysosomal Death Pathway It is well known that lysosomes contain a myriad of degradative enzymes, including proteases, and are one of the central garbage dumps of the cell for the disposal of exhausted or damaged proteins, carbohydrates, and lipids. Of the human cathepsins (clan CA, family C1, subfamily C1A), the most ubiquitously expressed and most abundant in lysosomes are cathepsin B, cathepsin H and cathepsin L along with cathepsin D (clan AA, family A1). Although the pH optimum for these enzymes is in the acidic range, they are capable of substrate hydrolysis in the neutral pH of the cytoplasm

ch13

Jan. 10, 2007

8:57

SPI-B429

312


F. Scott

and because they are stored as active enzymes, it is crucial that they be sequestered from the rest of the cell for the duration of their lifespan. It has been accepted for some time that lysosomal proteases contribute to necrotic and autophagic cell death. More recently, it has become evident that lysosomal proteases can specifically activate apoptotic pathways (Fig. 1). Cell death induced by lysosomal proteases leads to necrosis or apoptosis and depends on the degree of lysosomal permeabilization and hence protease leakage into the cytoplasm. Selective permeabilization of the lysosomal membrane results in apoptosis, while massive organelle rupture leads to necrosis.83–85 For either pathway of cell death to ensue, the amount of protease released must outweigh the amount of regulators in the cytoplasm, including cystatins and serpins, which provide a protective barrier against adventitious proteolysis. Activation of death receptors, lipid mediators, and photo damage have been suggested to trigger the lysosomal death pathway (see Refs. 86 and 87 for review). Only two of the human Clade B serpins have been demonstrated to inhibit lysosomal cysteine proteases — SERPINB3 and SERPINB13. SERPINB3 efficiently inhibits cathepsin K, L, S, and V,79 while SERPINB13 primarily targets cathepsin L.75 SERPINB3 can modulate apoptosis induced by TNF-α, chemotherapeutic drugs, and irradiation, while SERPINB13 inhibits UV irradiation induced cell death. It is tempting to speculate that their pro-survival function is due to neutralization of lysosomal proteases that might gain access to the cytoplasm in response to these stimuli. However, SERPINB3 and SERPINB13 are apparently restricted to epithelial cells, while the lysosomal death pathway can be induced in many other cell types. There is a formal possibility that other, as yet unidentified, and more broadly distributed intracellular serpins may regulate these enzymes in the cytoplasm of other cells. Murine serpinb3a (SQN-5), an evolutionarily close relative of human SERPINB3 and SERPINB4, can also inhibit the lysosomal proteases cathepsin K, S, L, and V, even though a pro-survival role in the lysosomal death pathway for this serpin is yet to be demonstrated.88 To date, a serpin that can inhibit cathepsin D has not been identified. There may also be a role for the murine Clade A serpin, serpina3g (serpin 2A/SPI2A/Spi2-1) in regulating the lysosomal death pathway (Fig. 1). In response to TNF-α, cathepsin B leakage into the cytoplasm has been observed, inducing mitochondria-dependent cell death.89–91 Serpina3g can

ch13

Jan. 10, 2007

8:57

SPI-B429



313

inhibit lysosomal proteases cathepsin B, L, V, K, and H in vitro. The second-order rate constant for cathepsin B inhibition was estimated to be >106 M−1 s−1 . Cathepsin B activity in the cytoplasm, in response to TNF-α, may be regulated by serpina3g, the expression of which is induced by TNF-α in a nuclear factor-κB (NF-κB) dependent manner.92 Ectopic expression of serpina3g can also prevent TNF-α induced caspase-independent apoptosis of fibroblasts through neutralization of cathepsin B activity in the cytoplasm.92,93 The localization of serpina3g in the cytoplasm and nucleus places it in a convenient location for the neutralization of proteases that may escape from the lysosomes.92,94 A caveat to this hypothesis is that a human counterpart of the serpina3g gene does not appear to exist, although a functional homolog cannot be excluded.

8. SERPINB5 is a Pro-apoptotic Serpin SERPINB5 (maspin/PI5) was first identified as a differentially expressed serpin in normal breast epithelial and myoepithelial cells. SERPINB5 was downregulated or absent from most mammary carcinoma cell lines and advanced mammary carcinoma tumors samples from patients.95 Paradoxically, SERPINB5 is elevated in some ovarian carcinoma and pancreatic cancers.96,97 Regardless, it is considered a class II suppressor gene and its properties include the inhibition of cell migration and invasion, the inhibition of tumor growth and metastasis and also the suppression of angiogenesis (see Ref. 98 for review). SERPINB5 may also play a role in apoptosis. Targeting SERPINB5 to the developing mammary gland inhibited alveolar development during pregnancy through the induction of mammary cell apoptosis.99 In agreement with this in vivo observation, overexpression of SERPINB5 in breast and prostate cancer cells was not pro-apoptotic per se, but sensitizes cells to other apoptotic stimuli.100–102 In contrast, when targeted to endothelial cells, SERPINB5 can actively induce apoptosis.103 Although it can be released from cells, where it functions to suppress cell invasion and motility, it is intracellular maspin that sensitizes cells to apoptosis and the RCL is required.100,103 The requirement of the RCL is very interesting because SERPINB5 is not considered a typical inhibitory serpin. The consensus residues within the hinge region of the RCL

ch13

Jan. 10, 2007

8:57

SPI-B429

314


F. Scott

(P14Thr, P12 Ala, P10 Ala, and P8Thr) required for serpins to undergo the conformational change that traps a protease in a structurally distorted, inhibited state, are absent in SERPINB5.104 At this point, SERPINB5 is unable to inhibit any proteases it has been tested with. If it is capable of protease inhibition, it is unlikely to be through the typical pseudo-substrate serpin mechanism. The requirement for the RCL in apoptosis regulation invokes a novel function for this region of the serpin. In the absence of inhibitory potential, the RCL may act as a substrate for proteolysis, acting as a sensor for protease activity in the cell. This would be analogous to the function of Bid, a pro-apoptotic Bcl2 family member. Bid can be cleaved in the cytosol by lysosomal proteases, caspases and granzyme B. Once cleaved, it translocates to the mitochondria, invoking the intrinsic apoptotic pathway (Fig. 1). SERPINB5 translocation to mitochondria in response to serum starvation, with subsequent cytochrome C release and loss of mitochondrial membrane potential has been reported to occur in an RCL dependent manner.101 It is unclear whether the cleavage of SERPINB5 is required for mitochondrial translocation, whether it binds to another protein (e.g. Bax or truncated Bid) and is “piggy backed” into the mitochondria, or whether it is translocated on its own. Although the mechanism of SERPINB5 induced apoptosis is unclear, it may involve deregulation of pro- and anti-apoptotic Bcl2-family members.102,103

9. Intracellular Serpins Protect Immune Cells from their Own Cytotoxic Proteases Cells in both the acquired immune system (cytotoxic T lymphocytes (CTLs)) and innate immune system (monocytes, macrophages, neutrophils, mast cells, natural killer cells (NKs), and lymphokine activated killer cells (LAKs)) utilize a range of cytotoxic defense mechanisms in their fight against bacterial, viral and parasitic infections, as well as tumor development and growth in the case of CTLs and NK cells. One of these defense strategies comprises serine proteases of the chymotrypsin family (clan PA, family S1, subfamily S1A). The immune cells can use serine proteases to kill either a pathogen, a virally infected cell, or a tumor cell, in a number of ways (Fig. 2). Firstly, in the case of neutrophils and mononuclear phagocytes, they can fuse their protease-containing azurophilic granules with a

ch13

Jan. 10, 2007

8:57

SPI-B429



315

Fig. 2. Model for the anti-apoptotic function of some intracellular serpins in immune cells. A common theme is that exposure of the cytoplasm to proteases will result in cell death, and expression of intracellular serpins that regulate those proteases provides protection. The route of access to the cytoplasm by proteases normally confined to membrane bound vesicles is unclear but may include: (1) miscompartmentalization during packaging into granules; (2) rupture of granules during cell stress; (3) leakage of granule contents during fusion with phagocytic vesicles; (4) internalization of extracellular protease after degranulation. (A) CTLs and NK cells store active granzymes in their cytotoxic granules, for release into the immune synapse and destruction of the target cell. The presence of serpins in the cytoplasm of the cytotoxic cell protects it from its own degradative enzymes. SERPINB9 inhibits GraB while mouse Serpinb9b inhibits GraM. A human counterpart to Serpinb9b, and serpins that regulate other granzymes (GraA, GraH, and GraK) are yet to be identified. (B) Granulocytes and monocytes contain active elastase, proteinase 3, and catG in their azurophilic granules. Their granule contents are either delivered to the phagosome, containing ingested pathogens or are secreted outside the cell. SERPINB1 and SERPINB6 together neutralize azurophilic granule proteases inside the cell while SERPINA1 and SERPINA3 inhibit extracellular protease. (C) Mast cell granules contain CatG, chymase and heparin-associated tetrameric tryptase. SERPINB6 inhibits CatG and can bind monomeric tryptase, while SERPINB9 can bind an unidentified protease. The physiological relevance of monomeric tryptase is unknown.

ch13

Jan. 10, 2007

8:57

SPI-B429

316


F. Scott

phagosome that encases recently phagocytosed bacteria, fungus or residual apoptotic bodies of neighboring cells.105 The combined activities of these neutral proteases and other mediators (e.g. reactive oxygen intermediates) in the newly generated phagolysosome destroy the ingested pathogen or cell corpse. Secondly, they can secrete their granule contents in a regulated manner out of the cell to kill pathogens in the extracellular milieu. For instance, the secretion of elastase by neutrophils occurs at sights of inflammation. As a side effect in the lungs, elastase can have a devastating effect on tissue architecture if its activity goes unregulated by the extracellular SERPINA1 (α1-antitrypsin/α1-antiproteinase). This imbalance between enzyme and inhibitor underpins emphysema that develops in patients containing mutations in the SERPINA1 gene (see Chapter 20). Finally, as is the case for NK cells and CTLs, after ligation of a virally infected target cell or tumor cell, they direct the release of their cytotoxic granules, which contain serine proteases (granzymes) and membrane permeabilizing molecules (perforin), directly into the immune synapse between the target and killer cell. It was originally thought that perforin created pores in the target cells, allowing the granzymes to enter the target cell by diffusion.106,107 Recent investigations have led to the revision of this theory, though the mechanism of entry into the target cell is still not clearly understood. One group has observed that GraB and perforin are delivered to target cells as a macromolecular complex with granule derived proteoglycans (serglycin), in the absence of pore formation.108 In contrast, the 280 kDa cation-independent mannose6-phosphate receptor was identified is a cell surface receptor for GraB.109 Finally, others have shown that free GraB is internalized by both a receptordependent and a receptor-independent (fluid phase) mechanism.110,111 Each leukocyte type expresses a different, but sometimes overlapping repertoire of hydrolases. Although the range of enzymes expressed varies, regardless of cell type, a common feature is that they are stored in membrane-bound intracellular vesicles or granules as active enzymes. This poses a problem for the cell, a question of how to retain an arsenal of weaponry with which to attack invaders without compromising one’s own barracks? These cells must function in an environment, high in free radicals that are deleterious to lipid membranes. The opportunity for the release of cytotoxic proteases into the cytoplasm of the killer cell itself may in fact be reasonably high. Immune cells have developed defense mechanisms to

ch13

Jan. 10, 2007

8:57

SPI-B429



317

protect them from their own toxic proteases: intracellular serpins that eliminate the cytotoxic activity of proteases that may leak out of membrane bound intracellular organelles (see Refs. 112 and 113 for review).

10. Cytotoxic Lymphocytes and Natural Killer Cells The human genome contains five granzymes (GraA, GraB, GraH, GraK, and GraM), while in the mouse genome, the number is expanded to 11 (GraA-G, GraK, and GraM-O). Each hemopoietic cytotoxic cell-type contains a subset of these serine proteases (clan PA, family S1, subfamily S1A). In CTLs, the most abundant are GraA and GraB, the best studied of the granzymes. Researchers have historically focused on GraB because it was initially found to be absolutely required for rapid DNA degradation associated with CTL-induced cell death.114 It is clear that GraB can directly initiate apoptosis through the proteolytic activation of executioner caspases.115–117 In addition, GraB can directly cleave a number of protein substrates in the cell that are a part of the apoptotic pathway, such as Bid (Fig. 1) and polyADP ribose polymerase, as well as other intracellular substrates (see Ref. 111 for review). Like caspases, GraB cleaves substrates at P1Asp. Although the function and specificity of some of the other granzymes has remained elusive, there is mounting evidence that they too have cytotoxic properties that are important for tumor surveillance and control of viral infection in vivo. GraA can induce cell death in the absence of caspase activation.118,119 GraA and GraK are tryptase-like proteases that cleave substrates after basic amino acids. GraK can also kill cells in a perforindependent manner in vitro, although the mechanism is unknown.120 GraC and GraM can kill cells in a caspase-independent manner in vitro.121,122 The substrate specificity of GraC (the closest murine ortholog of human GraH) is unknown, while GraM, an NK cell specific granzyme, prefers residues with long hydrophobic side chains in P1 . Finally, some evidence suggests that murine GraA-D and GraF-G are required for efficient tumor clearance in a murine tumor model, while other studies indicate that none of these granzymes are required.123,124 It seems that murine CTLs lacking GraA-D and GraF-G are able to lyse target cells as efficiently as wild-type cells, although DNA cleavage, a hallmark of apoptosis, is not observed.125

ch13

Jan. 10, 2007

8:57

SPI-B429

318


F. Scott

How do CTLs and NK cells protect themselves from their own cytotoxic granzymes? In humans, an intracellular serpin, SERPINB9, appears to have evolved this function, at least for neutralizing GraB.112 SERPINB9 is the only human serpin with an acidic residue (Glu340) at the P1 position of its RCL making it a logical candidate, considering the GraB substrate preference for aspartic acid, followed by glutamic acid.52,53 In addition, SERPINB9 is expressed in the nucleus and cytosol of the killer cell, where it can localize to the cytoplasmic face of the cytotoxic granules, the ideal location to neutralize any GraB activity that escapes from the granules.126 This hypothesis has been confirmed in NK cells. Within one minute of NK cell activation with anti-CD2 antibody, GraB leaks from the granules to the cytoplasm where it is inactivated by SERPINB9.127 Further activation leads to apoptosis, presumably because the amount of GraB that leaks into the cytoplasm overcomes the amount of SERPINB9. GraB leakage from the granules to the cytosol after in vitro T-cell activation was also seen by microscopy.128 Protection of CTLs against their own cytotoxic proteases may be of critical importance for the development of long-term immunity. After exposure to antigen, there is clonal expansion and differentiation of naive T-cells into CTLs, followed by a contraction of the CTL population, where 90–95% die by apoptosis (see Ref. 129 for review). Only a small number survived to become memory cells (CD44 high ). There is some evidence to suggest that mouse Serpinb9 (SPI6) contributes to the development of long-lived memory T-cells. Mouse Serpinb9 mRNA levels are elevated in effector CTLs after antigen challenge, and although they decrease over time, antigenrestricted memory cells retain Serpinb9 expression at levels five-fold higher than naive cells before the antigen challenge. This coincides with elevated active GraB in the granules of memory versus naive T-cells. Significantly, more virus specific memory cells persist after infection in transgenic mice overexpressing Serpinb9 in the T-cell lineage. By increasing the survival potential of CTLs through protection from their own GraB, SERPINB9 may promote memory cell development.57 Another predominantly intracellular mouse serpin, Serpina3g, is also slightly elevated in mouse memory T-cells. The authors have suggested that Serpina3g contributes to memory cell maintenance through cathepsin B inhibition, although a role for this lysosomal cysteine protease in the contraction phase is unclear.130

ch13

Jan. 10, 2007

8:57

SPI-B429



319

Serpina3g is an NF-κB responsive gene, and coincidently, NF-κB is required for memory T-cell formation.92 Again, a caveat to this hypothesis is that there is no human counterpart of the Serpina3g gene, although a functional homologue cannot be excluded. In murine NK cells and CTLs, expression of the serpinb9 is accompanied by Serpinb9b (SPI-CI/R86).131,132 Serpinb9b can associate with GraM, forming an SDS-stable complex. Although the second-order rate constant for this interaction is yet to be determined, ectopic expression of serpinb9b conveys resistance to apoptosis induced by delivery of purified GraM and perforin to the target cell. We await further study of this serpin and identification of its human counterpart. SERPINB9 is also found in the cytoplasm of antigen-presenting cells, as well as mesothelial and endothelial cells, and may provide protection to these “bystander” cells during an immune response.126,133–135 This hypothesis is supported by experimental evidence that cells expressing SERPINB9, or its murine orthologue, Serpinb9, resist apoptosis induced by delivery of purified GraB and perforin, as well as by cytotoxic lymphocytes.53,56,136 At first glance, it seems counter-intuitive for an endogenous inhibitor to evolve a less efficiently recognized P1 residue, when compared with the substrate preference of the enzyme. However, the mechanism of serpins as pseudo-substrate inhibitors may account for this. The serpin must bind the enzyme and trap it by conformational change, in an irreversible complex prior to the deacylation step. To prevent the serpin acting as a substrate and being rapidly cleaved and inactivated before it can undergo the conformational change, it may be beneficial to have an RCL with slightly lower binding affinity. This is reflected in the second-order rate constant of 1.7 × 106 M−1 s−1 between SERPINB9 and GraB.53 In contrast, CrmA has a P1Asp residue and is less efficient at inhibiting GraB, with a second-order rate constant of 2.9 × 105 M−1 s−1 .137

11. Granulocytes and Mast Cells The membrane-bound azurophilic granules of neutrophils and mononuclear phagocytes contain the neutral serine proteases CatG, elastase and proteinase 3 (see Ref. 138 for review). They are closely related both at the

ch13

Jan. 10, 2007

8:57

SPI-B429

320


F. Scott

protein and genetic level and have similar structures, although their substrate specificities differ. CatG has an unusual dual specificity, preferring substrates with P1 Lys, Phe, Tyr and to a lesser extent, Arg, while elastase and proteinase 3 have broader specificity, with some preference for small hydrophobic residues (elastinolytic activity). The physiological function of these serine proteases is still not completely resolved. However, it is clear from both in vitro and in vivo studies that inside the phagolysosome of granulocytes, these serine proteases contribute to the degradation of ingested bacterial and fungal pathogens during inflammation.139 Among other cytotoxic mediators including histamine, mast cells also contain CatG, in conjunction with tryptase-like and chymase-like proteases that are released into the tissue upon mast cell activation. Aside from the release of granulocyte proteases into the phagosome to destroy phagocytosed material, these cytotoxins may also be secreted into the extracellular milieu. Many functions have been ascribed to one or all of these proteases, including the processing of interleukin-8, chemotaxis, platelet activation, and degradation of extracellular matrix components, including fibronectin, elastin, proteoglycans, and collagen types III and IV (refer to Ref. 140). The extracellular activity of these proteases is regulated by the SERPINA1 and SERPINA3 (α1 -antichymotrypsin). In the cytoplasm of the granulocytes themselves are SERPINB1 (monocyte neutrophil elastase inhibitor (MNEI), horse leukocyte elastase inhibitor (LEI), and PI-2) and SERPINB6 (PI-6/cytoplasmic antiproteinase/placental thrombin inhibitor), which, can neutralize all three of these proteases when collaborated. SERPINB1 was one of the first intracellular serpins discovered. The protein was originally isolated from monocytes in 1985 as a potent inhibitor of both neutrophil and pancreatic elastase.141 Besides elastase, SERPINB1 is also able to inhibit proteinase 3 and CatG from monocytes and neutrophils with second-order rate constants of > 106 M−1 s−1 ,142,143 while SERPINB6 is a very potent CatG inhibitor with a kass > 107 M−1 s−1 .144 Taken together with their expression in both the cytoplasm and nucleus of monocytes and neutrophils, the inhibitory activity of SERPINB1 and SERPINB6 against all the azurophilic granule proteases provides a buffer zone against leakage of these cytotoxic proteases into the cytoplasm of the granulocyte (Fig. 2).

ch13

Jan. 10, 2007

8:57

SPI-B429



321

The hypothesis of Clade B serpin-induced cytoprotection in immune cells may also extend to mast cells, although studies addressing this are only beginning to appear in the literature. In mast cells, SERPINB6 can form a complex with monomeric tryptase, although this may be a postlysis phenomenon.145 SERPINB1 is a potent chymase inhibitor (kass > 105 M−1 s−1 ), even though the expression of SERPINB1 in mast cells is yet to be reported.143 SERPINB9 also forms a complex with an unidentified protein in mast cell extracts, but this may yet be a post-lysis event.146 In contrast to lysosomal proteases and granzymes, there is less evidence to suggest that the serine proteases of granulocytes and mast cells escape from membrane-bound vesicles into the cytoplasm, before, during, or after phagocytosis or granule secretion. However, it can be predicted that if it were to occur, the effect on cell viability would be catastrophic. There is some evidence that CatG can move from the azurophilic granules to the cytoplasm and nucleus of apoptotic promyelocytic leukemia cells.81 In addition, CatG can directly activate caspase-7 by proteolysis in vitro.82 Although there is a formal possibility that CatG observed in the cytoplasm is a consequence of, and not a cause of apoptosis, introduction of any serine protease into the cytoplasm seems to kill the cell by apoptosis.147 An alternative route to the cytoplasm may be internalization from the extracellular environment at the sites of inflammation, which may in fact be the case for proteinase-3.148 This model of SERPINB1 and SERPINB6 function as housekeeping proteins, providing protection from intracellular proteolysis that may also be extended to other tissues. In prostate tissue, SERPINB1 may inhibit prostate-specific antigen (PSA), and in the pancreas, it may inhibit pancreatic elastase.143 In prostate cancer cells, SERPINB6 forms a complex with kallikrein2, although this seems to be in a post-lysis manner.149,150 At this stage, it is unclear whether PSA, kallikrein2, and pancreatic elastase can also gain access to the cytoplasm of the cells where they are produced. In addition, SERPINB1 and SERPINB6 are also found in many tissues that do not produce proteases they have been reported to regulate,113 so other functions for SERPINB1 and SERPINB6 are probably yet to be revealed. More recently, it has been suggested that SERPINB1 may acquire DNase activity, becoming L-DNase II (leukocyte elastase inhibitor-derived DNase II), upon cleavage with either elastase or 24 kDa-apoptotic

ch13

Jan. 10, 2007

8:57

SPI-B429


322

F. Scott

protease, or acid treatment.151–153 It seems that L-DNase II is activated during lens cell differentiation, a process postulated to be a modified form of apoptosis.154,155 Curiously, enforced expression of porcine SERPINB1 protects BHK cells from etoposide induced apoptosis, yet sensitizes the same cells to apoptosis induced by acidification with hexamethylene amiloride.156 This would be a novel function for a serpin and the mechanism of endonuclease activity has not been demonstrated. We await the development of a SERPINB1 null mouse to verify the endonuclease activity ascribed to cleaved SERPINB1 is physiologically significant.

12. Conclusions A common element of all aspects of cell death discussed herein is the pivotal role of proteases. The decision to die in response to apoptotic or necrotic stimuli is governed in part by a finely tuned balance between proteases and their inhibitors. Serpins contribute to this balance in both the extracellular and intracellular compartments. For instance, SERPINE2 (protease nexin 1) regulates neuronal apoptosis in response to the activation of the cell surface thrombin receptor (protease activated receptor-1, PAR-1) by thrombin (see Ref. 157 for review). This is an example of a serpin that regulates apoptotic signaling from the outside of the cell. From within the cell, a subset of serpins can regulate cytotoxic proteases employed by immune cells in their attack on invading pathogens. This can provide the immune cell itself with a buffer zone that protects cytosolic and nuclear components from aberrant proteolytic activity, potentially resulting in sustained lifespan of the immune cell, and hence a more efficient and prolonged immune response.112 This hypothesis is also supported by the presence of serpins in the cytosol of nonimmune cells at sites of immune privilege. Finally, some viral serpins can directly regulate the apoptotic response through the inhibition of caspases, the effector molecules of apoptosis. It is less clear how intracellular serpins that do not regulate immune cell proteases or caspases contribute to apoptosis regulation. This is an area of increasing interest due to recent evidence that lysosomal cathepsins and calpains may have apoptotic roles in some pathological situations. With mounting evidence that some intracellular serpins with non-overlapping

ch13

Jan. 10, 2007

8:57

SPI-B429



323

inhibitory profiles can mediate apoptosis in response to different stimuli, there is a real need for further studies into the mechanisms employed. Information about the individual cellular compartments of intracellular serpins are found, and the identification of the types of proteases they can inhibit will provide insight into the proteases they actually inhibit in vivo, and subsequently, the physiological pathways they regulate. Since the completion of the human genome project, it is unlikely that there are many new human proteases to be discovered, so the re-examination of well-characterized proteases, previously considered unlikely serpin targets might provide some interesting new information. It is also entirely possible that some serpins have evolved new functions aside from direct protease inhibition, as it appears to be the case for proapoptotic SERPINB5. Some serpins may have adapted to bind other cellular components, including proteins, lipids, DNA, or nucleotides, in an RCLdependent manner. Alternatively, they may bind these components via as yet unidentified exosites on the body of the serpin. This may occur in a manner analogous to heparin binding to the circulatory serpin, SERPINC1 (antithrombin III). Heparin binding substantially increases its inhibitory capacity for the blood coagulation proteases, thrombin and factor Xa (see Ref. 158 for review). Then again, exosite binding may induce other as yet unidentified non-inhibitory functions that may or may not require the RCL of the serpin. The propensity of some intracellular serpins to selfassociate and polymerize, or aggregate under physiological conditions may also be important.159,160 Polymerization may be a means of sequestering serpin activity or it may have a more pro-active function that is unknown at this stage. The mechanisms intracellular serpins employ for regulating apoptosis, in either a positive or negative manner, is still largely a mystery. Research into these areas should provide interesting new information about serpin biology within the cell.

References 1. Kerr JF, Wyllie AH and Currie AR (1972) Apoptosis: A basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 26:239–257.

ch13

Jan. 10, 2007

8:57

SPI-B429

324


F. Scott

2. Horvitz HR (1999) Genetic control of programmed cell death in the nematode Caenorhabditis elegans. Cancer Res 59:1701s–1706s. 3. Jacobson MD, Weil M and Raff MC (1997) Programmed cell death in animal development. Cell 88:347–354. 4. Tittel JN and Steller H (2000) A comparison of programmed cell death between species. Genome Biol 1:REVIEWS0003. 5. Hay BA, Huh JR and Guo M (2004) The genetics of cell death: Approaches, insights and opportunities in Drosophila. Nat Rev Genet 5: 911–922. 6. Abrams JM (1995) An emerging blueprint for apoptosis in Drosophila. Trends Cell Biol 9:435–440. 7. Thompson CB (1995) Apoptosis in the pathogenesis and treatment of disease. Science 267:1456–1462. 8. Mattson MP (2000) Apoptosis in neurodegenerative disorders. Nat Rev Mol Cell Biol 1:120–129. 9. Yuan J and Yankner BA (2000) Apoptosis in the nervous system. Nature 407:802–809. 10. Vaux DL and Flavell RA (2000) Apoptosis genes and autoimmunity. Curr Opin Immunol 12:719–724. 11. Hanahan D and Weinberg RA (2000) The hallmarks of cancer. Cell 100: 57–70. 12. Lowe SW, Cepero E and Evan G (2004) Intrinsic tumour suppression. Nature 432:307–315. 13. Green DR and Evan GI (2002) A matter of life and death. Cancer Cell 1:19–30. 14. Alam JJ (2003) Apoptosis: Target for novel drugs. Trends Biotechnol 21: 479–483. 15. Rawlings ND and Barrett AJ (1993) Evolutionary families of peptidases. Biochem J 290( Pt 1):205–218. 16. Yuan J, Shaham S, Ledoux S, Ellis HM and Horvitz HM (1993) The C. elegans cell death gene ced-3 encodes a protein similar to mammalian interleukin1β-converting enzyme. Cell 75:641–652. 17. Boatright KM, Renatus M, Scott FL, Sperandio S, Shin H, Pedersen IM, Ricci JE, Edris WA, Sutherlin DP, Green DR and Salvesen GS (2003) A unified model for apical caspase activation. Mol Cell 11:529–541. 18. Renatus M, Stennicke HR, Scott FL, Liddington RC and Salvesen GS (2001) Dimer formation drives the activation of the cell death protease caspase 9. Proc Natl Acad Sci USA 98:14250–14255. 19. Donepudi M, Mac Sweeney A, Briand C and Grutter MG (2003) Insights into the regulatory mechanism for caspase-8 activation. Mol Cell 11:543–549.

ch13

Jan. 10, 2007

8:57

SPI-B429



325

20. Fischer U, Janicke RU and Schulze-Osthoff K (2003) Many cuts to ruin: A comprehensive update of caspase substrates. Cell Death Differ 10:76–100. 21. Pickup DJ (1994) Poxviral modifiers of cytokine responses to infection. Infect Agents Dis 3:116–127. 22. Gagliardini V, Fernandez V-A, Lee RK, Drexler HCA, Rotello RJ, Fishman MC and Yuan J (1994) Prevention of vertebrate neuronal death by the crmA gene. Science 263:826–828. 23. Strasser A, Harris AW, Huang DC, Krammer PH and Cory S (1995) Bcl-2 and Fas/APO-1 regulate distinct pathways to lymphocyte apoptosis. Embo J 14:6136–6147. 24. Memon SA, Moreno MB, Petrak D and Zacharchuk CM (1995) Bcl-2 blocks glucocorticoid- but not Fas- or activation-induced apoptosis in a T cell hybridoma. J Immunol 155:4644–4652. 25. Moreno MB, Memon SA and Zacharchuk CM (1996) Apoptosis signaling pathways in normal T cells: Differential activity of Bcl-2 and IL-1betaconverting enzyme family protease inhibitors on glucocorticoid- and Fas-mediated cytotoxicity. J Immunol 157:3845–3849. 26. Chinnaiyan AM, Orth K, O’Rourke K, Duan H, Poirier GG and Dixit VM (1996) Molecular ordering of the cell death pathway. Bcl-2 and Bcl-xL function upstream of the CED-3-like apoptotic proteases. J Biol Chem 271: 4573–4576. 27. Datta R, Kojima H, Banach D, Bump NJ, Talanian RV, Alnemri ES, Weichselbaum RR, Wong WW and Kufe DW (1997) Activation of a CrmAinsensitive, p35-sensitive pathway in ionizing radiation-induced apoptosis. J Biol Chem 272:1965–1969. 28. Datta R, Banach D, Kojima H, Talanian RV, Alnemri ES, Wong WW and Kufe DW (1996) Activation of the CPP32 protease in apoptosis induced by 1-beta-D-arabinofuranosylcytosine and other DNA-damaging agents. Blood 88:1936–1943. 29. Los M, Van de Craen M, Penning LC, Schenk H, Westendorp M, Baeuerle PA, Droge W, Krammer PH, Fiers W and Schulze-Osthoff K (1995) Requirement of an ICE/CED-3 protease for Fas/APO-1 mediated apoptosis. Nature (London) 375:81–83. 30. Enari M, Hug H and Nagata S (1995) Involvement of an ICE-like protease in Fas-mediated apoptosis. Nature 375:78–81. 31. Tewari M and Dixit VM (1995) Fas- and tumor necrosis factor-induced apoptosis is inhibited by the poxvirus crmA gene product. J Biol Chem 270: 3255–3260.

ch13

Jan. 10, 2007

8:57

SPI-B429

326


F. Scott

32. Miura M, Friedlander RM and Yuan J (1995) Tumor necrosis factor-induced apoptosis is mediated by a CrmA-sensitive cell death pathway. Proc Natl Acad Sci USA 92:8318–8322. 33. Talley AK, Dewhurst S, Perry SW, Dollard SC, Gummuluru S, Fine SM, New D, Epstein LG, Gendelman HE and Gelbard HA (1995) Tumor necrosis factor alpha-induced apoptosis in human neuronal cells: Protection by the antioxidant N-acetylcysteine and the genes bcl-2 and crmA. Mol Cell Biol 15:2359–2366. 34. Marsters SA, Pitti RM, Donahue CJ, Ruppert S, Bauer KD and Ashkenazi A (1996) Activation of apoptosis by Apo-2 ligand is independent of FADD but blocked by CrmA. Curr Biol 6:750–752. 35. Boudreau N, Sympson CJ, Werb Z and Bissell MJ (1995) Suppression of ICE and apoptosis in mammary epithelial cells by extracellular matrix. Science 267:891–893. 36. Frisch SM and Screaton RA (2001) Anoikis mechanisms. Curr Opin Cell Biol 13:555–562. 37. Turner SJ, Silke J, Kenshole B and Ruby J (2000) Characterization of the ectromelia virus serpin, SPI-2. J Gen Virol 81:2425–2430. 38. Dobbelstein M and Shenk T (1996) Protection against apoptosis by the vaccinia virus SPI-2 (B13R) gene product. J Virol 70:6479–6485. 39. Kettle S, Alcami A, Khanna A, Ehret R, Jassoy C and Smith GL (1997) Vaccinia virus serpin B13R (SPI-2) inhibits interleukin-1beta-converting enzyme and protects virus-infected cells from TNF- and Fas-mediated apoptosis, but does not prevent IL-1beta-induced fever. J Gen Virol 78( Pt 3):677–685. 40. Macen JL, Garner RS, Musy PY, Brooks MA, Turner PC, Moyer RW, McFadden G and Bleackley RC (1996) Differential inhibition of the Fas- and granule-mediated cytolysis pathways by the orthopoxvirus cytokine response modifier A/SPI-2 and SPI-1 protein. Proc Natl Acad Sci USA 93:9108–9113. 41. Tewari M, Quan LT, O’Rourke K, Desnoyers S, Zeng Z, Beidler DR, Poirier GG, Salvesen GS and Dixit VM (1995) Yama/CPP32 beta, a mammalian homolog of CED-3, is a CrmA-inhibitable protease that cleaves the death substrate poly(ADP-ribose) polymerase. Cell 81:801–809. 42. Tesch LD, Raghavendra MP, Bedsted-Faarvang T, Gettins PG and Olson ST (2005) Specificity and reactive loop length requirements for crmA inhibition of serine proteases. Protein Sci 14:533–542. 43. Zhou Q, Snipas S, Orth K, Muzio M, Dixit VM and Salvesen GS (1997) Target protease specificity of the viral serpin CrmA. Analysis of five caspases. J Biol Chem 272:7797–7800.

ch13

Jan. 10, 2007

8:57

SPI-B429



327

44. Garcia-Calvo M, Peterson EP, Leiting B, Ruel R, Nicholson DW and Thornberry NA (1998) Inhibition of human caspases by peptide-based and macromolecular inhibitors. J Biol Chem 273:32608–32613. 45. Ryan CA, Stennicke HR, Nava VE, Burch JB, Hardwick JM and Salvesen GS (2002) Inhibitor specificity of recombinant and endogenous caspase-9. Biochem J 366:595–601. 46. Ray CA, Black RA, Kronheim SR, Greenstreet TA, Sleath PR, Salvesen GS and Pickup DJ (1992) Viral inhibition of inflammation: Cowpox virus encodes an inhibitor of the interleukin-1 beta converting enzyme. Cell 69:597–604. 47. Komiyama T, Ray CA, Pickup DJ, Howard AD, Thornberry NA, Peterson EP and Salvesen G (1994) Inhibition of interleukin-1 beta converting enzyme by the cowpox virus serpin crmA. An example of cross-class inhibition. J Biol Chem 269:19331–19337. 48. Macen J, Takahashi A, Moon KB, Nathaniel R, Turner PC and Moyer RW (1998) Activation of caspases in pig kidney cells infected with wild-type and CrmA/SPI-2 mutants of cowpox and rabbitpox viruses. J Virol 72:3524–3533. 49. Ghayur T, Banerjee S, Hugunin M, Butler D, Herzog L, Carter A, Quintal L, Sekut L, Talanian R, Paskind M, Wong W, Kamen R, Tracey D and Allen H (1997) Caspase-1 processes IFN-gamma-inducing factor and regulates LPSinduced IFN-gamma production. Nature 386:619–623. 50. Gu Y, Kuida K, Tsutsui H, Ku G, Hsiao K, Fleming MA, Hayashi N, Higashino K, Okamura H, Nakanishi K, Kurimoto M, Tanimoto T, Flavell RA, Sato V, Harding MW, Livingston DJ and Su MS (1997) Activation of interferon-gamma inducing factor mediated by interleukin-1beta converting enzyme. Science 275:206–209. 51. Renatus M, Zhou Q, Stennicke HR, Snipas SJ, Turk D, Bankston LA, Liddington RC and Salvesen GS (2000) Crystal structure of the apoptotic suppressor CrmA in its cleaved form. Struct Fold Des 8:789–797. 52. Sprecher CA, Morgenstern KA, Mathews S, Dahlen JR, Schrader SK, Foster DC and Kisiel W (1995) Molecular cloning, expression and partial characterization of two novel members of the ovalbumin family of serine proteinase inhibitors. J Biol Chem 270:29854–29861. 53. Sun J, Bird C, Sutton V, McDonald L, Coughlin PB, De Jong, TA, Trapani JA and Bird P (1996) A cytosolic granzyme B inhibitor related to the viral apoptotic regulator cytokine response modifier A is present in cytotoxc lymphocytes. J Biol Chem 271:27802–27809. 54. Bird CH, Blink EJ, Hirst CE, Buzza MS, Steele PM, Sun J, Jans DA and Bird PI (2001) Nucleocytoplasmic distribution of the ovalbumin serpin

ch13

Jan. 10, 2007

8:57

SPI-B429

328

55.

56.

57.

58.

59.

60.

61. 62.

63.

64.

65.


F. Scott

PI-9 requires a nonconventional nuclear import pathway and the export factor Crm1. Mol Cell Biol 21:5396–5407. Annand RR, Dahlen JR, Sprecher CA, De Dreu P, Foster DC, Mankovich JA, Talanian RV, Kisiel W and Giegel DA (1999) Caspase-1 (interleukin-1betaconverting enzyme) is inhibited by the human serpin analogue proteinase inhibitor 9. Biochem J 342 Pt 3:655–665. Bird CH, Sutton V, Sun J, Hirst C, Novak A, Kumar S, Trapani JA and Bird PI (1998) Selective regulation of apoptosis: The cytotoxic lymphocyte serpin proteinase inhibitor 9 protects against granzyme B-mediated apoptosis without perturbing the Fas cell death pathway. Mol Cell Biol 18:6387–6398. Phillips T, Opferman JT, Shah R, Liu N, Froelich CJ and Ashton-Rickardt PG (2004) A role for the granzyme B inhibitor serine protease inhibitor 6 in CD8+ memory cell homeostasis. J Immunol 173:3801–3809. Kumar S and Baglioni C (1991) Protection from tumor necrosis factormediated cytolysis by overexpression of plasminogen activator inhibitor type 2. J Biol Chem 266:20960–20964. Dickinson JL, Bates EJ, Ferrante A and Antalis TM (1995) Plasminogen activator inhibitor-type 2 inhibits TNFα-induced apoptosis: Evidence for an alternative biological function. J Biol Chem 270:27894–27904. Dickinson JL, Norris BJ, Jensen PH and Antalis TM (1998) The C-D interhelical domain of the serpin plasminogen activator inhibitor type-2 is required for protection from TNFα induced apoptosis. Cell Death Diff 5:163–171. Remold-O’Donnell E (1993) The ovalbumin family of serpin proteins. FEBS Lett 315:105–108. Jensen PH, Lorand L, Ebbesen P and Gliemann J (1993) Type-2 plasminogenactivator inhibitor is a substrate for trophoblast transglutaminase and factor XIIIa. Transglutaminase-catalyzed cross-linking to cellular and extracellular structures. Eur J Biochem 214:141–146. Jensen PH, Schuler E, Woodrow G, Richardson M, Goss N, Hojrup P, Petersen TE and Rasmussen LK (1994) A unique interhelical insertion in plasminogen activator inhibitor 2 contains three glutamines, Gln83, Gln84, Gln86, essential for transglutaminase-mediated cross-linking. J Biol Chem 269:15394– 15398. Jensen PJ, Wu Q, Janowitz P, AndoY and Schechter NM (1995) Plasminogen activator inhibitor type 2:An intracellular keratinocyte differentiation product that is incorporated into the cornified envelope. Experimental Cell Research 217:65–71. Jensen PH, Jensen TG, Laug WE, Hager H, Gliemann J and Pepinsky B (1996) The exon 3 encoded sequence of the intracellular proteinase inhibitor

ch13

Jan. 10, 2007

8:57

SPI-B429



66.

67.

68.

69.

70.

71.

72.

73.

74.

75.

329

plasminogen activator inhibitor 2 is a protein binding domain. J Biol Chem 271:26892–26899. Darnell GA, Antalis TM, Johnstone RW, Stringer BW, Ogbourne SM, Harrich D and Suhrbier A (2003) Inhibition of retinoblastoma protein degradation by interaction with the serpin plasminogen activator inhibitor 2 via a novel consensus motif. Mol Cell Biol 23:6520–6532. Wilczynska M, Lobov S, Ohlsson PI and Ny T (2003) A redox-sensitive loop regulates plasminogen activator inhibitor type 2 (PAI-2) polymerization. Embo J 22:1753–1761. SuminamiY, Nagashima S, Vujanovic NL, Hirabayashi K, Kato H and Whiteside TL (2000) Inhibition of apoptosis in human tumour cells by the tumourassociated serpin, SCC antigen-1. British J Cancer, 82:981–989. McGettrick AF, Barnes RC and Worrall DM (2001) SCCA2 inhibits TNFmediated apoptosis in transfected HeLa cells. The reactive centre loop sequence is essential for this function and TNF-induced cathepsin G is a candidate target. Eur J Biochem 268:5868–5875. Schleef RR and Chuang TL (2000) Protease inhibitor 10 inhibits tumor necrosis factor alpha -induced cell death. Evidence for the formation of intracellular high M(r) protease inhibitor 10-containing complexes. J Biol Chem 275:26385–26389. VarroA, Noble PJ, Pritchard DM, Kennedy S, Hart CA, Dimaline R and Dockray GJ (2004) Helicobacter pylori induces plasminogen activator inhibitor 2 in gastric epithelial cells through nuclear factor-kappaB and RhoA: Implications for invasion and apoptosis. Cancer Res 64:1695–1702. Belin D, Wohlwend A, Schleuning W-D, Kruithof EKO and Vassalli J-D (1989) Facultative polypeptide translocation allows a single mRNA to encode the secreted and cytosolic forms of plasminogen activator inhibitor 2. EMBO J 8:3287–3294. Zhou HM, Bolon I, Nichols A, Wohlwend A and Vassalli JD (2001) Overexpression of plasminogen activator inhibitor type 2 in basal keratinocytes enhances papilloma formation in transgenic mice. Cancer Res 61: 970–976. Murakami A, Suminami Y, Hirakawa H, Nawata S, Numa F and Kato H (2001) Squamous cell carcinoma antigen suppresses radiation-induced cell death. Br J Cancer 84:851–858. Welss T, Sun J, Irving JA, Blum R, Smith AI, Whisstock JC, Pike RN, von Mikecz A, Ruzicka T, Bird PI and Abts HF (2003) Hurpin is a selective inhibitor of lysosomal cathepsin L and protects keratinocytes from ultravioletinduced apoptosis. Biochemistry 42:7381–7389.

ch13

Jan. 10, 2007

8:57

SPI-B429

330


F. Scott

76. Gil-Parrado S, Fernandez-Montalvan A, Assfalg-Machleidt I, Popp O, Bestvater F, Holloschi A, Knoch TA, Auerswald EA, Welsh K, Reed JC, Fritz H, Fuentes-Prior P, Spiess E, Salvesen GS and Machleidt W (2002) Ionomycin-activated calpain triggers apoptosis. A probable role for Bcl-2 family members. J Biol Chem 277:27217–27226. 77. Johnson DE (2000) Noncaspase proteases in apoptosis. Leukemia 14:1695– 1703. 78. Jaattela M (2002) Programmed cell death: Many ways for cells to die decently. Ann Med 34:480–488. 79. Schick C, Pemberton PA, Shi G-P, Kamachi Y, Cataltepe S, Bartuski AJ, Gornstein ER, Bromme D, Chapman HA and Silverman GA (1998) Crossclass inhibition of the cysteine proteinases cathepsins K, L, and S by the serpin squamous cell carcinoma antigen 1: A kinetic analysis. Biochemistry 37:5258–5266. 80. Schick C, KamachiY, BartuskiAJ, Cataltepe S, Schechter NM, Pemberton PA and Silverman GA (1997) Squamous cell carcinoma antigen 2 is a novel serpin that inhibits the chymotrypsin-like proteinases cathepsin G and mast cell chymase. J Biol Chem 272:1849–1855. 81. Biggs JR, Yang J, Gullberg U, Muchardt C, Yaniv M and Kraft AS (2001) The human brm protein is cleaved during apoptosis: The role of cathepsin G. Proc Natl Acad Sci USA 98:3814–3819. 82. Zhou Q and Salvesen GS (1997) Activation of pro-caspase-7 by serine proteinases includes a non-canonical specificity. Biochem J 324:361–364. 83. Li W, Yuan X, Nordgren G, Dalen H, Dubowchik GM, Firestone RA and Brunk UT (2000) Induction of cell death by the lysosomotropic detergent MSDH. FEBS Lett 470:35–39. 84. Bursch W (2001) The autophagosomal-lysosomal compartment in programmed cell death. Cell Death Diff 8:569–581. 85. Turk B, Stoka V, Rozman-Pungercar J, Cirman T, Droga-Mazovec G, Oresic K and Turk V (2002) Apoptotic pathways: Involvement of lysosomal proteases. Biol Chem 383:1035–1044. 86. Guicciardi ME, Leist M and Gores GJ (2004) Lysosomes in cell death. Oncogene 23:2881–2890. 87. Lockshin RA and Zakeri Z (2002) Caspase-independent cell deaths. Curr Opin Cell Biol 14:727–733. 88. Al-Khunaizi M, Luke CJ, Askew YS, Pak SC, Askew DJ, Cataltepe S, Miller D, Mills DR, Tsu C, Bromme D, Irving JA, Whisstock JC and Silverman GA (2002) The serpin SQN-5 is a dual mechanistic-class inhibitor of serine and cysteine proteinases. Biochemistry 41:3189–3199.

ch13

Jan. 10, 2007

8:57

SPI-B429



331

89. Guicciardi ME, Deussing J, Miyoshi H, Bronk SF, Svingen PA, Peters C, Kaufmann SH and Gores GJ (2000) Cathepsin B contributes to TNFa– mediated hepatocyte apoptosis by promoting mitochondrial release of cytochrome c. J Clin Invest 106:1127–1137. 90. Foghsgaard L, Wissing D, Mauch D, Lademann U, Bastholm L, Boes M, Elling F, Leist M and Jaattela M (2001) Cathepsin B acts as a dominant execution protease in tumor cell apoptosis induced by tumor necrosis factor. J Cell Biol 153:999–1010. 91. Werneburg NW, Guicciardi ME, Bronk SF and Gores GJ (2002) Tumor necrosis factor-alpha-associated lysosomal permeabilization is cathepsin B dependent. Am J Physiol Gastrointest Liver Physiol 283:G947–G956. 92. Liu N, Raja SM, Zazzeroni F, Metkar SS, Shah R, Zhang M, WangY, Bromme D, Russin WA, Lee JC, Peter ME, Froelich CJ, Franzoso G and AshtonRickardt PG (2003) NF-kappaB protects from the lysosomal pathway of cell death. Embo J 22:5313–5322. 93. Liu N, Wang Y and Ashton-Rickardt PG (2004) Serine protease inhibitor 2A inhibits caspase-independent cell death. FEBS Lett 569:49–53. 94. Morris EC, Dafforn TR, Forsyth SL, Missen MA, Horvath AJ, Hampson L, Hampson IN, Currie G, Carrell RW and Coughlin PB (2003) Murine serpin 2A is a redox-sensitive intracellular protein. Biochem J 371:165–173. 95. Zou Z, Anisowicz A, Hendrix MJC, Thor A, Neveu M, Sheng S, Rafidi K, Seftor E and Sager R (1994) Maspin, a serpin with tumor-suppressing activity in human mammary epithelial cells. Science 263:526–529. 96. Sood AK, Fletcher MS, Gruman LM, Coffin JE, Jabbari S, Khalkhali-Ellis Z, Arbour N, Seftor EA and Hendrix MJ (2002) The paradoxical expression of maspin in ovarian carcinoma. Clin Cancer Res 8: 2924–2932. 97. Maass N, Hojo T, Ueding M, Luttges J, Kloppel G, Jonat W and Nagasaki K (2001) Expression of the tumor suppressor gene Maspin in human pancreatic cancers. Clin Cancer Res 7:812–817. 98. Schaefer JS and Zhang M (2003) Role of maspin in tumor metastasis and angiogenesis. Curr Mol Med 3:653–658. 99. Zhang M, Magit D, Botteri F, Shi YH, He K, Li M, Furth P and Sager R (1999) Maspin plays an important role in mammary gland development. Develop Biol 215:278–287. 100. Jiang N, Meng Y, Zhang S, Mensah-Osman E and Sheng S (2002) Maspin sensitizes breast carcinoma cells to induced apoptosis. Oncogene 21:4089– 4098. 101. Latha K, Zhang W, Cella N, Shi HY and Zhang M (2005) Maspin mediates increased tumor cell apoptosis upon induction of the mitochondrial permeability transition. Mol Cell Biol 25:1737–1748.

ch13

Jan. 10, 2007

8:57

SPI-B429

332


F. Scott

102. Liu J, Yin S, Reddy N, Spencer C and Sheng S (2004) Bax mediates the apoptosis-sensitizing effect of maspin. Cancer Res 64:1703–1711. 103. Li Z, Shi HY and Zhang M (2005) Targeted expression of maspin in tumor vasculatures induces endothelial cell apoptosis. Oncogene 24:2008–2019. 104. Pemberton PA, Wong DT, Gibson HL, Kiefer MC, Fitzpatrick PA, Sager R and Barr PJ (1995) The tumor suppressor maspin does not undergo the stressed to relaxed transition or inhibit trypsin-like proteases. J Biol Chem 270:15832–15837. 105. Baggiolini M, Bretz U, Dewald B and Feigenson ME (1987) The polymorphonuclear leukocyte. Agents Actions 8:3–10. 106. Podack ER (1992) Perforin: Structure, function, and regulation. Curr Top Microbiol Immunol 178:175–184. 107. Yagita H, Nakata M, Kawasaki A, Shinkai Y and Okumura K (1992) Role of perforin in lymphocyte-mediated cytolysis. Adv Immunol 51:215–242. 108. Metkar SS, Wang B, Aguilar-Santelises M, Raja SM, Uhlin-Hansen L, Podack E, Trapani JA and Froelich CJ (2002) Cytotoxic cell granule-mediated apoptosis: Perforin delivers granzyme B-serglycin complexes into target cells without plasma membrane pore formation. Immunity 16:417–428. 109. Motyka B, Korbutt G, Pinkoski MJ, Heibein JA, Caputo A, Hobman M, Barry M, Shostak I, Sawchuk T, Holmes CF, Gauldie J and Bleackley RC (2000) Mannose 6-phosphate/insulin-like growth factor II receptor is a death receptor for granzyme B during cytotoxic T cell-induced apoptosis. Cell 103:491–500. 110. Russell JH and Ley TJ (2002) Lymphocyte-mediated cytotoxicity. Annu Rev Immunol 20:323–370. 111. Trapani JA and Sutton VR (2003) Granzyme B: Pro-apoptotic, antiviral and antitumor functions. Curr Opin Immunol 15:533–543. 112. Bird PI (1999) Regulation of pro-apoptotic leucocyte granule serine proteinases by intracellular serpins. Immunol Cell Biol 77:47–57. 113. Silverman GA, Whisstock JC, Askew DJ, Pak SC, Luke CJ, Cataltepe S, Irving JA and Bird PI (2004) Human clade B serpins (ov-serpins) belong to a cohort of evolutionarily dispersed intracellular proteinase inhibitor clades that protect cells from promiscuous proteolysis. Cell Mol Life Sci 61: 301–325. 114. Heusel JW, Wesselschmidt RL, Shresta S, Russell JH and Ley TJ (1994) Cytotoxic lymphocytes require granzyme B for the rapid induction of DNA fragmentation and apoptosis in allogeneic target cells. Cell 76:977–987. 115. Martin SJ, Amarante-Mendes GP, Shi L, Chuang TH, Casiano CA, O’Brien GA, Fitzgerald P, Tan EM, Bokoch GM, Greenberg AH and Green DR (1996) The cytotoxic cell protease granzyme B initiates apoptosis in a cell-free

ch13

Jan. 10, 2007

8:57

SPI-B429



116.

117.

118.

119.

120.

121. 122.

123.

124.

125.

126.

333

system by proteolytic processing and activation of the ICE/CED-3 family protease, CPP32, via a novel two-step mechanism. Embo J 15:2407–2416. Darmon AJ, Nicholson DW and Bleackley RC (1995) Activation of the apoptotic protease CPP32 by cytotoxic T-cell-derived granzyme B. Nature 377:446–448. Fernandes-Alnemri T, Armstrong R, Krebs J, Srinivasula SM, Wang L, Bullrich F, Fritz L, Trapani JA, Croce CM, Tomaselli KJ, Litwack G and Alnemri ES (1996) In vitro activation of CPP32 and Mch3 by Mch4, a novel human apoptotic cysteine protease containing two FADD-like domains. Proc Natl Acad Sci USA 93:7464–7469. Beresford PJ, Xia Z, Greenberg AH and Lieberman J (1999) Granzyme A loading induces rapid cytolysis and a novel form of DNA damage independently of caspase activation. Immunity 10:585–594. Shresta S, Graubert TA, Thomas DA, Raptis SZ and Ley TJ (1999) Granzyme A initiates an alternative pathway for granule-mediated apoptosis. Immunity 10:595–605. MacDonald G, Shi L, Vande Velde C, Lieberman J and Greenberg AH (1999) Mitochondria-dependent and -independent regulation of Granzyme B-induced apoptosis. J Exp Med 189:131–144. Johnson H, Scorrano L, Korsmeyer SJ and Ley TJ (2003) Cell death induced by granzyme C. Blood 101:3093–3101. Kelly JM, Waterhouse NJ, Cretney E, Browne KA, Ellis S, Trapani JA and Smyth MJ (2004) Granzyme M mediates a novel form of perforin-dependent cell death. J Biol Chem 279:22236–22242. Revell PA, Grossman WJ, Thomas DA, Cao X, Behl R, Ratner JA, Lu ZH and Ley TJ (2005) Granzyme B and the downstream granzymes C and/or F are important for cytotoxic lymphocyte functions. J Immunol 174:2124–2131. Davis JE, Smyth MJ and Trapani JA (2001) Granzyme A and B-deficient killer lymphocytes are defective in eliciting DNA fragmentation but retain potent in vivo anti-tumor capacity. Eur J Immunol 31:39–47. Simon MM, Hausmann M, Tran T, Ebnet K, Tschopp J, ThaHla R and Mullbacher A (1997) In vitro- and ex vivo-derived cytolytic leukocytes from granzyme A x B double knockout mice are defective in granule-mediated apoptosis but not lysis of target cells. J Exp Med 186:1781–1786. Hirst CE, Buzza MS, Bird CH, Warren HS, Cameron PU, Zhang M, AshtonRickardt PG and Bird PI (2003) The intracellular granzyme B inhibitor, proteinase inhibitor 9, is up-regulated during accessory cell maturation and effector cell degranulation, and its overexpression enhances CTL potency. J Immunol 170:805–815.

ch13

Jan. 10, 2007

8:57

SPI-B429

334


F. Scott

127. Ida H, Nakashima T, Kedersha NL, Yamasaki S, Huang M, Izumi Y, Miyashita T, Origuchi T, Kawakami A, Migita K, Bird PI, Anderson P and Eguchi K (2003) Granzyme B leakage-induced cell death: A new type of activation-induced natural killer cell death. Eur J Immunol 33:3284–3292. 128. Bidere N, Briet M, Durrbach A, Dumont C, Feldmann J, Charpentier B, de Saint-Basile G and Senik A (2002) Selective inhibition of dipeptidyl peptidase I, not caspases, prevents the partial processing of procaspase-3 in CD3activated human CD8(+) T lymphocytes. J Biol Chem 277:32339–32347. 129. Sprent J and Surh CD (2002) T cell memory. Annu Rev Immunol 20:551–579. 130. Liu N, Phillips T, Zhang M, Wang Y, Opferman JT, Shah R and AshtonRickardt PG (2004) Serine protease inhibitor 2A is a protective factor for memory T cell development. Nat Immunol 5:919–926. 131. Bots M, Kolfschoten IG, Bres SA, Rademaker MT, de Roo GM, Kruse M, Franken KL, Hahne M, Froelich CJ, Melief CJ, Offringa R and Medema JP (2005) SPI-CI and SPI-6 cooperate in the protection from effector cellmediated cytotoxicity. Blood 105:1153–1161. 132. Sun J, Ooms L, Bird CH, Sutton VR, Trapani JA and Bird PI (1997) A new family of 10 murine ovalbumin serpins includes two homologs of proteinase inhibitor 8 and two homologs of the granzyme B inhibitor (proteinase inhibitor 9). J Biol Chem 272:15434–15441. 133. Bladergroen BA, Strik MC, Bovenschen N, van Berkum O, Scheffer GL, Meijer CJ, Hack CE and Kummer JA (2001) The granzyme B inhibitor, protease inhibitor 9, is mainly expressed by dendritic cells and at immuneprivileged sites. J Immunol 166:3218–3225. 134. Buzza MS, Hirst CE, Bird CH, Hosking P, McKendrick J and Bird PI (2001) The granzyme B inhibitor, PI-9, is present in endothelial and mesothelial cells, suggesting that it protects bystander cells during immune responses. Cell Immunol 210:21–29. 135. Medema JP, Schuurhuis DH, Rea D, van Tongeren J, de Jong J, Bres SA, Laban S, Toes RE, Toebes M, Schumacher TN, Bladergroen BA, Ossendorp F, Kummer JA, Melief CJ and Offringa R (2001) Expression of the serpin serine protease inhibitor 6 protects dendritic cells from cytotoxic T lymphocyteinduced apoptosis: Differential modulation by T helper type 1 and type 2 cells. J Exp Med 194:657–667. 136. Medema JP, de Jong J, Peltenburg LT, Verdegaal EM, Gorter A, Bres SA, Franken KL, Hahne M, Albar JP, Melief CJ and Offringa R (2001) Blockade of the granzyme B/perforin pathway through overexpression of the serine protease inhibitor PI-9/SPI-6 constitutes a mechanism for immune escape by tumors. Proc Natl Acad Sci USA 98:11515–11520.

ch13

Jan. 10, 2007

8:57

SPI-B429



335

137. Quan LT, Caputo A, Bleackley RC, Pickup DJ and Salvesen GS (1995) Granzyme B is inhibited by the cowpox virus serpin cytokine response modifier A. J Biol Chem 270:10377–10379. 138. Borregaard N and Cowland JB (1997) Granules of the human neutrophilic polymorphonuclear leukocyte. Blood 89:3503–3521. 139. Belaaouaj A, McCarthy R, Baumann M, Gao Z, Ley TJ, Abraham, SN and Shapiro SD (1998) Mice lacking neutrophil elastase reveal impaired host defense against gram negative bacterial sepsis. Nat Med 4:615–618. 140. (2004) Handbook of Proteolytic Enzymes, 2nd edn, Vol. 1–2. Elsevier. 141. Remold-O’Donnell E (1985) A fast-acting elastase inhibitor in human monocytes. J Exp Med 162:2142–2155. 142. Sugimori T, Cooley J, Hoidal JR and Remold-O’Donnell E (1995) Inhibitory properties of recombinant monocyte/neutrophil elastase inhibitor. Am J Resp Cell Mol Biol 13:314–322. 143. Cooley J, Takayama TK, Shapiro SD, Schechter NM and RemoldO’Donnell E (2001) The serpin MNEI inhibits elastase-like and chymotrypsin-like serine proteases through efficient reactions at two active sites. Biochemistry 40:15762–15770. 144. Scott FL, Hirst CE, Sun J, Bird CH, Bottomley SP and Bird PI (1999) The intracellular serpin proteinase inhibitor 6 is expressed in monocytes and granulocytes and is a potent inhibitor of the azurophilic granule protease, cathepsin G. Blood 93:2089–2097. 145. Strik MC, WolbinkA, Wouters D, Bladergroen BA,VerlaanAR, van Houdt IS, Hijlkema S, Hack CE and Kummer JA (2004) Intracellular serpin SERPINB6 (PI6) is abundantly expressed by human mast cells and forms complexes with beta-tryptase monomers. Blood 103:2710–2717. 146. Bladergroen BA, Strik MC, Wolbink AM, Wouters D, Broekhuizen R, Kummer JA and Hack CE (2005) The granzyme B inhibitor proteinase inhibitor 9 (PI9) is expressed by human mast cells. Eur J Immunol 35: 1175–1183. 147. Williams MS and Henkart PA (1994) Apoptotic cell death induced by intracellular proteolysis. J Immunol 153:4247–4255. 148. Yang JJ, Preston GA, Pendergraft WF, Segelmark M, Heeringa P, Hogan SL, Jennette JC and Falk RJ (2001) Internalization of proteinase 3 is concomitant with endothelial cell apoptosis and internalization of myeloperoxidase with generation of intracellular oxidants. Am J Pathol 158:581–592. 149. Mikolajczyk SD, Millar LS, Marker KM, Rittenhouse HG, Wolfert RL, Marks LS, Charlesworth MC and Tindall DJ (1999) Identification of a novel complex between human kallikrein 2 and protease inhibitor-6 in prostate cancer tissue. Cancer Res 59:3927–3930.

ch13

Jan. 10, 2007

8:57

SPI-B429

336


F. Scott

150. Saedi MS, Zhu Z, Marker K, Liu RS, Carpenter PM, Rittenhouse H and Mikolajczyk SD (2001) Human kallikrein 2 (hK2), but not prostate-specific antigen (PSA), rapidly complexes with protease inhibitor 6 (PI-6) released from prostate carcinoma cells. Int J Cancer 94:558–563. 151. Belmokhtar CA, Torriglia A, Counis MF, Courtois Y, Jacquemin-Sablon A and Segal-Bendirdjian E (2000) Nuclear translocation of a leukocyte elastase Inhibitor/Elastase complex during staurosporine-induced apoptosis: Role in the generation of nuclear L-DNase II activity. Exp Cell Res 254:99–109. 152. Torriglia A, Perani P, Brossas JY, Chaudun E, Treton J, CourtoisY and Counis MF (1998) L-DNase II, a molecule that links proteases and endonucleases in apoptosis, derives from the ubiquitous serpin leukocyte elastase inhibitor. Mol Cell Biol 18:3612–3619. 153. Altairac S, Wright SC, Courtois Y and Torriglia A (2003) L-DNase II activation by the 24 kDa apoptotic protease (AP24) in TNFalpha-induced apoptosis. Cell Death Differ 10:1109–1111. 154. Torriglia A, Chaudun E, Chany-Fournier F, Jeanny JC, CourtoisY and Counis MF (1995) Involvement of DNase II in nuclear degeneration during lens cell differentiation. J Biol Chem 270:28579–28585. 155. Counis MF, Chaudun E, Arruti C, Oliver L, Sanwal M, Courtois Y and Torriglia A (1998) Analysis of nuclear degradation during lens cell differentiation. Cell Death Differ 5:251–261. 156. Altairac S, Zeggai S, Perani P, Courtois Y and Torriglia A (2003) Apoptosis induced by Na+/H+ antiport inhibition activates the LEI/L-DNase II pathway. Cell Death Differ 10:548–557. 157. Festoff BW, Smirnova IV, Ma J and Citron BA (1996) Thrombin, its receptor and protease nexin I, its potent serpin, in the nervous system. Semin Thromb Hemost 22:267–271. 158. Huntington JA (2003) Mechanisms of glycosaminoglycan activation of the serpins in hemostasis. J Thromb Haemost 1:1535–1549. 159. Mikus P and Ny T (1996) Intracellular polymerization of the serpin plasminogen activator inhibitor type 2. J Biol Chem 271:10048–10053. 160. Benning LN, Whisstock JC, Sun J, Bird PI and Bottomley SP (2004) The human serpin proteinase inhibitor-9 self-associates at physiological temperatures. Protein Sci 13:1859–1864.

ch13

Jan. 10, 2007

8:57

SPI-B429


14 Serpins in Malignancy: Tumor Cell Invasion, Motility and Angiogenesis Philip A. Pemberton, Christian Schem, Nicolai Maass and Ming Zhang

1. Introduction Serpins and their respective proteases or non-inhibitory targets possess well-documented functions in tumor malignancy. The growth of primary tumors is dependent on the proliferation status of tumor cells and the availability of growth factors provided by the surrounding vasculature in a tightly regulated, complex process known as angiogenesis. Tumor metastasis is another complex process involving the escape of tumor cells from the local environment and the migration to a remote location to create a secondary metastatic lesion that is capable of sustained growth. Since many serpins are involved in both tumor growth and metastasis, we choose to focus in this chapter on the unique serpin maspin and the role that it plays in these processes. We first describe the structural and biochemical basis for the maspin function. We then discuss the biological function of maspin in tumor progression and in human breast cancers. Finally, we discuss the involvement of other serpins in malignancy.

2. Structure and Biological Activities of Maspin Since Zou et al.1 described the existence of the tumor suppressor maspin (mammary serpin) in 1994, research efforts have largely focused on the 337

ch14

Jan. 10, 2007

8:57

SPI-B429

338


P. A. Pemberton et al.

mechanism of the action of the protein and its utility as a prognostic indicator for other types of cancer. Maspin affects tumor cell motility, invasion, growth, apoptosis, and angiogenesis.2–6 Several potential targets or molecular effectors of maspin functions have been described, including types I and III collagen, the integrins, and the plasminogen activators tPA and uPA.7–9 In addition, post-translational modifications of the protein have been described, which may play a role in regulating one or more of its biological activities.11 Very recently, low and high resolution threedimensional structures of human maspin have been resolved, providing insight into the structural elements involved in regulating the observed biologic responses.12 2.1. Biological activities of maspin: Location and function Maspin is a Clade B serpin, a unique group of the serpin superfamily that possesses uncleaved secretion signals, yet found both inside and outside cells.13 A few other notable members of this group are ovalbumin (OVAL), plasminogen activator inhibitor 2 (PAI-2), and the squamous-cell carcinoma antigens 1 and 2 (SCCA 1, SCCA 2). Maspin has been localized to the extracellular matrix, cellular surfaces, cytoplasm, and also found in the nucleus of a number of different cell types.14–16 Its expression is differentially regulated by hormonal and epigenetic processes, depending on the tissue type. For example, in animal models of prostate cancer, androgen ablation increases the expression of maspin-whereas in mammary adenocarcinomas, its expression is down-regulated by hypermethylation.17,18 How does maspin distribute among these compartments and is this compartmentalization important in determining the behavioral properties of tumors? Studies on the related serpins OVAL and PAI-2 have demonstrated that the uncleaved N-terminus of the protein functions as a facultative secretion signal, the hydrophobicity of which determines the ratio of secretion versus retention in the cytosol.19 It is yet unclear how maspin gets into the nucleus, as it does not possess any classical nuclear targeting signals. Maspin either secreted from the mammary epithelial cells or tumor tranfectants, or purified recombinant maspin was shown to block the ability of tumor cells to invade through artificial basement membrane

ch14

Jan. 10, 2007

8:57

SPI-B429


Serpins in Malignancy: Tumor Cell Invasion, Motility and Angiogenesis 339

matrices.20,21 However, the amount and localization of maspin expressed by transfected tumor cells or normal cells have never been accurately determined in many of these studies; thus, it is hard to draw any conclusions about where the protein is having its effect. However, in vitro and in vivo studies with recombinant maspin protein(s) have clearly shown that exogenously added maspin affects tumor cell migration, motility, growth and angiogenesis,2–4,6 indicating that signaling events must be capable of mediating many of these processes, as there is no evidence to date that maspin is taken up by cells. A number of mechanisms have been described to explain how exogenous maspin may work: exogenous maspin may (1) disrupt VEGF and bFGF growth factor signaling events required for the proliferation and migration of endothelial cells during angiogenesis, and consequently, affect tumor growth indirectly6 ; (2) coordinately regulate both invasion (adhesion) and motility by signaling through the Rho GTPase pathway to down-regulate Rac activity and inhibit cell motility, while promoting cell adhesion via PI3K/ERK pathways 11 ; and (3) increase the expression of the fibronectin receptor and consequently increase adhesion to fibronectin matrices.8 The function(s) of cytoplasmic or nuclear maspin are less well understood, but evidence is emerging that cytoplasmic maspin may translocate to the mitochondria where it may be involved in the opening of the permeability transition pore, the loss of transmembrane potential, and the initiation of a mitochondrial apoptosis pathway potentially involving Bax.22,23 This is in direct contrast to the apoptotic inhibitory properties of most other serpins.24 The elevated expression of maspin in the nucleus has been associated with microsatellite instability and high tumor grade in colorectal cancer.25 2.2. Functional domains of maspin: The reactive site loop The ability of maspin to inhibit tumor cell invasion is a function of the RSL.26 However, conflicting data exist as to the exact role of the RSL in inhibiting tumor cell invasion. There are concerns whether maspin directly inhibits the activity of tPA or uPA.9,10,27 The RSL sequence of maspin has been grafted onto OVAL and shown to be sufficient to allow OVAL to

ch14

Jan. 10, 2007

8:57

SPI-B429

340



inhibit tumor cell invasion.26 These same authors proceeded to show that the sequence of the RSL alone is sufficient to induce inhibition of tumor cell invasion.26 The RSL of most serpins possesses a protease inhibitory function supported by highly conserved structural properties. Most serpin conformations consist of three β-pleated sheets (A, B, and C), and eight α-helices.24 The recent three-dimensional structure(s) of maspin confirmed the presence of characteristic features of the serpin fold in maspin.12 The reactive site loop (RSL) is a stretch of approximately 17 amino acids tethered between β-sheets A and C, which exists in a variety of conformations ranging from a compact α-helix as seen in the non-inhibitory serpin OVAL to the canonical form observed in most inhibitory serpins.28,29 All inhibitory serpins have a partial opening “breach” between strands 3 and 5 of β-sheet A, into which the RSL may be partially inserted in the native state. Protease inhibition is thought to arise from the ability of the RSL to fully insert between strands 3 and 5 of β-sheet A, dragging the target protease with it and conformationally “crushing” it against the body of the molecule.30 The insertion of the RSL also occurs, following catalytic cleavage within the RSL by non-target proteases and accompanied by an increase in conformational stability. Serpins that have no known inhibitory activity (OVAL, angiotensinogen, and pigment epithelial derived factor [PEDF]) do not undergo this increase in stability and this transitional change from a stressed (S) form to a more relaxed (R) form following proteolysis has been considered a hallmark of inhibitory potential.31–33 2.3. Is maspin a protease inhibitor? The structure of the RSL in maspin is unique. Table 1 shows an alignment of RSL sequences present in selected Clade B family members and RSL sequences present in other serpins known to inhibit tPA, uPA, or both proteases. The following structural observations do not support the notion that maspin possesses direct protease inhibitory activity: (a) maspin lacks the conservation of sequence present in all other serpins within the region of the proximal hinge and, (b) alignment of the RSL of maspin with the RSL sequences in other serpins is difficult, as the sequence leading into the first strand of the C-sheet is significantly shorter than that in most other serpins.

ch14

Jan. 10, 2007 8:57

“Proxim Hinge”

P1

SERPIN B1

G

T

E

A

A

A

A

T

A

G

I

A

T

F

C

M

L

M

P

E

elastase

SCCA-1

SERPIN B3

G

A

E

A

A

A

A

T

A

V

V

G

F

G

S

S

P

A

S

T

cat L, S, K

Megsin

SERPIN B7

G

T

E

A

T

A

A

T

G

S

N

I

V

E

K

Q

L

P

Q

S

PAI-1

SERPIN E1

G

T

V

A

S

S

S

T

A

V

I

V

S

A

R

M

A

P

E

E

uPA, tPA

PAI-2

SERPIN B2

G

T

E

A

A

A

G

T

G

G

V

M

T

G

R

T

G

H

G

G

uPA

Maspin

SERPIN B5

G

G

D

S

I

E

V

P

G

A —

— —

—

R

I

L

Q

H

K


Target P rotease(s)

Putative RSL (P15-P5′)

Gene symbol

SPI-B429

Serpin


Table 1 Alignment of selected clade B and other tPA/uPA inhibitory RSL sequences.

ch14

Jan. 10, 2007

8:57

SPI-B429

342



In the recent solved structure(s) the “breach” between the third and fifth strands of β-pleated sheet A is closed and the RSL is expelled fully from the A β-sheet, potentially denying access to the RSL, following interaction with any potential target protease. In the low-resolution structure solved by Al-Ayyoubi et al.12 the placement of the RSL was different from inhibitory serpins in that it was fully expelled from the A β-sheet, but bent toward the amino end of the second strand of β-pleated sheet C (s2C) and held in place by a number of stabilizing bonding interactions with surface amino acid side chains of β-sheet C.12 This imparted a constrained conformation on the RSL, one that is accessible to proteases, in particular, trypsin-like proteases, and is consistent with previous observations. In the higher resolution structures solved by Law et al.34 the RSL was also fully expelled from the A β-sheet, but made minimal contacts with the core of the molecule, allowing a high degree of flexibility of the RSL. The differences in RSL placements and surface contacts may be attributable to the different crystallization conditions and reflect the different conformations that serpins can adopt. Finally, maspin does not undergo the S to R transition,35 yet does associate in a manner indicative of polymerization. The recent low-resolution structure solved by Al-Ayyoubi showed that such polymerization may occur by a unique mechanism involving dimers of tetrameric maspin formed by intermolecular contacts predominantly from the RSLs.12 The relevance of this is presently unknown. The RSL is also implicated in the induction of apoptosis, although how this happens is also presently unknown.22 Curiously, a sequence almost identical to that of the RSL is found in the microtubule associated protein 1B (MAP1B).36 It is tempting to speculate that the polymerization phenomenon observed for maspin may also apply to MAP1B. 2.4. Mapping functional domains/residues of maspin Table 2 summarizes the functions described thus far and where they map on the modeled structure of maspin. The secretion of the protein appears to be dependent on an uncleaved facultative secretion signal encoded by the first 50 amino acids. Inhibition of tumor cell migration is dependent on the

ch14

Jan. 10, 2007

8:57

SPI-B429


Serpins in Malignancy: Tumor Cell Invasion, Motility and Angiogenesis 343 Table 2 Domain function in maspin. Domain(s)/residues N-terminal 1–50 residues C20, C34, C183, C205, C214, C287, C323, C373 Y84, Y93, Y112, Y116, Y356 RSL

? Y84 – Y 112

Function(s) Facultative secretion signal ? One or more phosphorylated Inhibition of tumor cell invasion Apoptosis induction (?) Polymerization (G331-E335) Anti-angiogenic activity Collagen binding (I, III)

RSL, whereas inhibition of endothelial cell migration and angiogenesis is not.6,26 Polymerization is dependent on the presence of the RSL, but not on its integrity.12,35 The protein possesses eight free cysteines — one or more of which are highly reactive and likely to possess some functional roles.37 More recently, a role in signal transduction has been described for maspin.11 Maspin is reportedly phosphorylated on tyrosine residues in normal mammary epithelial cells and when transfected into maspin-deficient tumor cell lines. Furthermore, recombinant maspin is reportedly phosphorylated by the kinase domain from the epidermal growth factor receptor in vitro, eventhough, in the high-resolution structure of maspin, none of the five tyrosine residues are sufficiently exposed to allow phosphorylation by a kinase. Subtle conformational shifts may expose Tyr 84, 112 or 166; thus the possibility of tyrosine phosphorylation as a means of regulating maspin activity is not entirely excluded.95 Maspin is also reported to bind to types I and III collagens and the binding domain suggested to lie between residues Y84 and Y112.7 However, the low-resolution structure did not reveal any well-defined acidic collagen-binding motif as exists in the related collagenbinding serpin (PEDF) (pigment epithelium derived factor).38 However, the high-resolution crystal structure revealed that maspin is capable of undergoing a novel major conformational change in and around the G α-helix, switching between an open and closed conformation. This rearrangement exposes two negatively charged glutamate residues, which contribute to a

ch14

Jan. 10, 2007

8:57

SPI-B429


344

P. A. Pemberton et al. Table 3 Anti-angiogenic serpins and their functions.

Serpin

Active form (s)

AAT

Native, C-terminal truncated

↓ Angiogenesis

ANGIO

ANG I ANG II ANG II RSL cleaved ANGIO

↓ Angiogenesis ↑ Angiogenesis ↓ Angiogenesis ↓ Angiogenesis

Latent, pre-latent, RSL-cleaved

↓ Angiogenesis

AT III

Function

Kallistatin Native

↓ Angiogenesis

Maspin

Native; C-terminal truncated

↓ Angiogenesis

PAI - 1

Native

↓ Angiogenesis

PEDF

Native, RSL-cleaved ↓ Angiogenesis

Co-factor(s) Effector(s)

AT2 Receptor AT1 Receptor

Uncouples the focal adhesion kinase (fak); ↓ Perlecan Heparin

↓ VEGF binding and phosphorylation of Akt Disrupts VEGF/bFGF signaling events

Vitronectin Heparin

Blocks hypoxia induced decline in TSP-1; Receptor binding?

newly formed elongated, negatively charged patch. The comparable region in PEDF has been implicated in collagen binding. Finally, very little is currently known about how intracellular maspin mediates induction of apoptosis or translocates to the nucleus, but the availability of these three-dimensional structures will allow us to start addressing these questions at the molecular level.

ch14

Jan. 10, 2007

8:57

SPI-B429



3. Role of Maspin in Tumor Progression and Angiogenesis Maspin is a complex and versatile protein with a multitude of effects on cells and tissues at various stages of development. Beginning with embryo development, maspin is required for the appropriate cell-matrix interactions to occur so that a healthy and viable embryo develops. Maspin plays an important role in other tissue developments. In this section, we will only discuss maspin’s role in tumorigenesis and angiogenesis. Emerging evidence indicates that extracellular and intracellular maspin plays different functions in cell adhesion and signal transduction, as well as in tumor cell apoptosis. Such diverse functions demand further investigation into the mode of maspin action. Understanding the molecular mechanism of maspin action may provide invaluable information for therapeutic intervention of certain cancers.

3.1. Role of maspin in breast cancer progression Together with same colleague, we have utilized several mouse models to study maspin’s function in tumor progression in the past few years. In the first study, MDA-MB-435 mammary carcinoma cells were transfected with maspin cDNA and implanted into nude mice. Tumor initiation and growth were recorded; mice were examined post-mortem to confirm the presence of metastasis. In agreement with the tumor suppressor designation, the transfectants that expressed maspin exhibited reduced growth rates and the incidence of metastasis.1,39 In the second study, TM40D cells and maspin TM40D transfectants were implanted into the No. 4 mammary GLANDS of BALB/c mice. In concurrence with the MDA-MB-435 nude mouse study, the maspin TM40D transfectants again exhibited reduced growth rates and incidence metastasis.40 Finally, we employed a maspin transgenic mouse model (WAP-maspin) to examine the effect of overexpression of maspin on tumor suppression. The WAP-maspin transgenic mice had targeted overexpression of maspin in mammary epithelial cells by a mammary specific promoter-WAP.41 We then crossed the WAP-maspin transgenic mice with a strain of oncogenic WAP-SV 40 T antigen (TAg) mice. WAP-TAg transgenic mice develop mammary tumors with high frequency by targeting the

ch14

Jan. 10, 2007

8:57

SPI-B429

346



inactivation of both the p53 and the pRb related family of proteins.42,43 We compared the tumor free time period and the rate of tumor growth of the WAP-TAg and the bitransgenic mice.44 The tumor free period in mice overexpressing maspin significantly increased from 36.3 to 49.3 days. The tumor growth rate was also significantly decreased in maspin overexpression mice. Such inhibition resulted from the effect of maspin overexpression on microvessel density and apoptosis. We also examined the effect of maspin overexpression on lung tumor metastasis. The bitransgenic mice had a reduced rate of metastasis, compared with that of the WAP-TAg single transgenic mice. These studies demonstrate that maspin indeed inhibits primary tumor growth and metastasis. Maspin’s tumor inhibitory effect prompted us to develop a cancer therapy, utilizing maspin gene as targeting molecule. Cancer gene therapy requires both a good animal model and an effective delivery system.45 We established a syngeneic PyV-mT tumor transplantation model for gene delivery study. The PyV-mT cells were initially isolated from the mammary tumors in MMTV-PyV-mT transgenic mice.46 The PyV-mT cells were transplanted to syngeneic FVB female mammary fat pad bilaterally. Tumors developed 100% in transplanted sites and were treated with maspin DNA:liposome complexes or with control plasmid DNA through tail vein injection. The maspin group received the DNA:liposome treatment for an average of 32.2±1.8 days, whilst the control group received vector DNA:liposome for 33.7±1.8 days (P 0.05, no significant difference). Primary mammary tumor growth was monitored by caliper measurement during the treatment and at the endpoint. Our data showed that both the tumor size and the overall tumor growth rate were significantly decreased for maspin treated tumors, when compared with that in the controls.47 In addition, maspin treatment also exhibited reduced tumor metastasis as compared with control vector treated samples.47 These data demonstrate that maspin DNA:liposome treatment offers an effective therapy for breast cancer. 3.2. Role of maspin in angiogenesis Tumor growth and metastasis require neovascular formation, a process also termed as angiogenesis.48,49 Angiogenesis provides tumors with nutrients

ch14

Jan. 10, 2007

8:57

SPI-B429



and aids in the removal of metabolic wastes.50,51 Most solid tumors cannot grow beyond a few millimeters without neovascular formation.52,53 Several non-inhibitory serpin members have been discovered that possess antiangiogenic properties. For example, PEDF, a serpin with known function of cell differentiation, is also a very potent anti-angiogenic factor.54 A cleaved product of antithrombin III was shown to directly inhibit angiogenesis.55 Our study has shown that maspin acts as an angiogenesis inhibitor as well.6 In an initial study, recombinant maspin was tested in a variety of angiogenesis assays. Maspin blocked endothelial cell migration induced by VEGF and bFGF in a dose-dependent manner with an ED50 of 0.2–0.3 µM. In vivo, purified maspin effectively inhibited neovascularization. Rat corneas were surgically implanted with non-inflammatory slow release pellets containing maspin with bFGF and examined six or seven days later for the ingrowth of vessels. Deletion and mutation analysis demonstrated that maspin’s antiangiogenesis property is not dependent on the RSL region. Recombinant maspin mutants in the RSL region retained the ability to inhibit endothelial cell migration and mitogenesis in vitro. These proteins also retained the ability to inhibit neovascularization in rat corneas in vivo. To determine if the ability of maspin to inhibit angiogenesis plays a role in its well documented anti-tumor activity, an athymic mouse xenograft model was utilized.6 LNCaP prostate tumor cells were implanted subcutaneously on the bidorsal back of nude mice and tumor growth, and neovascularization were monitored following systemic treatment with exogenous maspin. We found that maspin-treated tumors contained significantly fewer vessels as determined by CD31 immunostaining than GST treated controls. To determine whether maspin effects on the tumor-induced vasculature were maintained during a more prolonged treatment, the above experiment was replicated with tumors harvested after seven to eight weeks. Thirty-two tumor sites were treated with maspin and 37 with GST. When examined on week eight, the growth of 53% of the maspin-treated tumors had been completely inhibited. The remaining 15 maspin-treated tumors were reduced in size by an average of 3.43-fold, compared with GST control treated tumors. The effect of maspin was reversible. To examine if the reduced size of maspin-treated tumors coincided with reduced neovascularization, 20 representative tumors from either maspin-treated (10 sites) or GST-treated tumors (10 sites) were used to quantify the density

ch14

Jan. 10, 2007

8:57

SPI-B429

348



of microvessels after immunostaining with CD31 antibody. The density of vessels in maspin-treated tumors was reduced 2.6-fold in average to that in control tumors and this difference was highly significant. We also compared the treated and control tumors of similar size. A reduction of vessel density was also observed in the maspin-treated samples. These data confirm that maspin is an effective inhibitor of tumor angiogenesis and therefore could be developed into a potent anti-angiogenesis/anti-cancer therapy. One possible reason that maspin treatment is effective may be due to increased apoptosis in targeted endothelial cells. To prove this hypothesis, we constructed an adenovirus maspin (Ad-Mp) to target maspin to both proliferating neovessels and non-proliferating mature vessels. In a recent report, we demonstrated that maspin could indeed cause vascular disruption in tumors that were treated with Ad-Mp for a short time period. This disruption led to the leakage of red blood cells as analyzed by histology and by a quantitative assay for hemoglobin. However, maspin acted selectively on neovascular endothelial cells and not on mature vessels. We have compared the vasculature of tissues from other organs with the vasculature of mammary tumors after the mice were treated by Ad-Mp. No maspin-mediated vascular disruption or leakage occurred in other tissues such as lung and kidney. An important finding from this study is that maspin directly induces apoptosis of proliferating endothelial cells. We used human HUVECs to test the sensitivity of endothelial cells to maspin treatment. Infection of HUVECs with Ad-Mp caused the cells to undergo apoptosis even when they were growing in high serum medium. Additionally, we were unable to overexpress maspin in HUVECs, suggesting that these endothelial cells are extremely sensitive to maspin dosages. The ability of maspin to induce apoptosis is not limited to the endothelial cells; breast tumor cells overexpressing maspin can sensitize the tumor cells to various stress conditions, including serum starvation and other chemicals. Subcellular fractionation shows that a fraction of intracellular maspin translocates to the mitochondria during apoptosis, which is linked to the opening of the permeability transition pore, causing the loss of trans-membrane potential and initiating an apoptotic process. Our data indicate that maspin inhibits tumor progression through the mitochondrial apoptosis pathway. These findings will be useful for maspin-based therapeutic interventions against breast cancer.

ch14

Jan. 10, 2007

8:57

SPI-B429



Another study by Cher et al.56 has provided further support regarding maspin’s effect on angiogenesis in cancer progression. They showed that maspin-expressing transfectant cells derived from prostate cancer cell line DU145 were inhibited in in vitro extracellular matrix and collagen degradation assays. They injected the maspin-transfected DU145 cells into human fetal bone fragments, which were previously implanted in immunodeficient mice. Their studies showed that maspin expression decreased tumor growth, reduced osteolysis, and decreased angiogenesis.56

4. Clinical Relevance of Maspin in Human Breast and Pancreatic Cancer Specimens Since the association of maspin down-regulation with breast tumor progression was first investigated in 1994,1 several clinical studies from different laboratories have reported the relevance of maspin in human breast and pancreatic cancers. Our group57 studied 45 invasive breast cancers with a short follow-up of 38 months, using nested RT-PCR. Maspin expression was reported in 64% of the cases, but this expression was not associated with any other prognostic factors such as tumor stage, histological type, ER, c-erbB-2 or patient’s age. Of-five-patients who developed recurrence within two years, five did not show maspin expression, but 70% of the 37 patients who did not develop recurrence had maspin expression. We concluded that the lack of maspin expression in breast cancer seems to be associated with a shorter disease-free survival. Additionally, we found nuclear maspin staining in the majority of the tumors and cell lines examined.58 The other group evaluated maspin expression in 49 human samples by IHC.59 50% of the cases were reported to express maspin. However, the description of the staining was unclear (“maspin staining was seemingly positive in 20 cases, all of which exhibited maspin expression in only myoepithelial cells”). The expression of maspin appeared to be associated with low levels of p53 and c-erbB-2 and higher microvessel density,60 but these associations were not statistically significant. The level of expression of maspin in nuclear versus cytoplasmic compartments of breast cancer and the implications of that expression therefore still remained unclear.

ch14

Jan. 10, 2007

8:57

SPI-B429

350



The presence of maspin in both nuclear and cytoplasmic compartments has been shown by Pemberton et al.14 with a specific maspin antibody. In this study, cell fractionation and Western blotting showed that although maspin is predominantly present in the cytoplasm of tumor cells, it is also detectable in the nuclear fraction. Immunohistochemistry depicts nuclear maspin in normal and malignant tissues from several different organs.14 Similar illustrations can be seen in IBC in Ref. 61 and in prostate cancer in Ref. 62. Theses results clearly demonstrate that maspin is indeed present in the nuclear compartment and antibodies to maspin are specific. However, to date, the function and significance of this pattern of expression remained unknown. Mohsin et al.63 addressed this in a preliminary fashion by separately recording and analyzing maspin expression in the cytoplasm and the nuclei for breast cancer cells. This study used the same antibody as the previous IHC study of human IBC. The staining in normal breast tissue (present in areas adjacent to the tumor) showed both cytoplasmic and nuclear staining in myoepithelial cells, but predominantly nuclear staining in the luminal epithelial cells. This finding is consistent with previous reports. There was differential expression of maspin in the two cell compartments in tumor cells. By statistical analyses, these two patterns had different implications. Since previous studies using IHC, including our own, have not separated nuclear form cytoplasmic maspin, it is difficult to compare them with the results from the Allred group63 accordingly. For example, positive cases account for 35% of the cytoplasmic maspin in Allreds group, which is substantially less than the 50% maspin staining reported by Hojo et al.59 In addition, our group reported a weak inverse correlation between maspin expression and lymph node status and PR, but no relationship with ER.61 Mohsin et al., however, found statistically significant relationships between maspin (both cytoplasmic staining and nuclear staining) and both hormone receptors. These signigicant results are most likely related to the large sample size in the Mohsin et al. study. The association of cytoplasmic maspin with several markers of poor clinical outcome appears counterintuitive to what is so far known of the biological function of maspin.20,59,61,62 However, two other reports, one in oral squamouscell carcinoma and the other in prostate andenocarcinoma, have shown that the loss of maspin expression is associated with advanced stage, primarily lymph node metastasis.63,64

ch14

Jan. 10, 2007

8:57

SPI-B429



In contrast to the findings in breast cancer, the maspin gene is not expressed in normal human pancreas, but its expression is acquired during human pancreatic carcinogenesis. In other normal human cells and their malignant counterparts, maspin expression is controlled through the epigenetic state of its promoter. In studies presented herein, Fitzgerald et al. used bisulfite genomic sequencing and chromatin immunoprecipitation studies to show that maspin-negative pancreas cells have a methylated maspin promoter, and that the associated H3 and H4 histones are hypoacetylated.65 In contrast to normal pancreas, four of six human pancreatic carcinoma cell lines investigated displayed activation of maspin expression. This activation of maspin expression in pancreatic carcinoma cells was linked to demethylated promoters and hyperacetylation of the associated H3 and H4 histones. In addition, 5-aza-2 -deoxycytidine treatments activated maspin expression in the two maspin-negative pancreatic carcinoma cell lines, suggesting a causal role for cytosine methylation in the maintenance of a transcriptionally silent maspin gene. Thus, human pancreatic carcinoma cells acquire maspin expression through epigenetic derepression of the maspin locus, and in doing so, appear to co-opt a normal cellular mechanism for the regulation of this gene. Sato et al. demonstrated significant differences in the methylation pattern of maspin between pancreatic and breast cancer cell lines, suggesting cancer-type specificity for some hypomethylation patterns.66 Their results confirm that DNA hypomethylation is a frequent epigenetic event in pancreatic cancer, and suggest that gene expression profiling may help in identifying potential targets affected by this epigenetic alteration.66 Similarly, our group demonstrated in pancreatic tissue, using the same antibody as that from Mohsin et al. (clone G167-70), that no expression of maspin was found in normal tissue or normal pancreatic cells lines, whilst five of nine cancer cell lines and 23 of 24 pancreatic tumors expressed maspin.15 The staining intensity was generally strong with minimal variation, except that cells with a broad clear cytoplasm showed a faint positivity. The only case that did not stain positively was a rare type of clear-cell ductal adenocarcinoma. Intraductal non-clear-cell areas of this case, however, showed faint nuclear and cytoplasmic staining. Some cells also exhibited nuclear staining that was always accompanied by a strong cytoplasmic staining. Intraductal extensions of the carcinomas and lesions

ch14

Jan. 10, 2007

8:57

SPI-B429

352



of PanIN grade 3 also stained positive, but with lower intensity. In contrast, ductal hyperplasia without dysplasia and low-grade dysplasia such as mucinous cell hypertrophy or papillary hyperplasia stained negative. In addition, foci of sqamous intraductal metaplasia showed cytoplasmic staining. There was no correlation between the staining intensity and the histological grade or stage of the tumors. Pemberton et al. did not detect maspin mRNA expression by Northern blot analysis, but did detect very weak maspin-like protein expression in the glandular epithelia of the pancreas by immunostaining, using a polyclonal antibody. The discrepancy may be attributable to different characteristics of the anti-maspin antibodies used, such as reaction with distinct epitopes of different specificity as mentioned above or the nature of the tissue employed for the immunohistochemical analysis. A monoclonal anti-maspin antibody was used in our study. Finally, low expression was observed in low-grade precancerous lesions, and maspin expression seems to increase with increasing malignancy from normal pancreas tissue via precancerous lesions to invasive carcinomas. Recently, Lim et al. confirmed our findings that maspin may be a marker in the progression of pancreatic cancers.67 Carcinoma of the pancreas is the fourth highest cause of cancer-related death and shows the highest mortality rate of all cancers in most Western countries.68 Several tumor-associated antigens such as CEA, CA 125, and CA 19-9 are used to monitor pancreatic cancer patients.69 However, they are not tumor-specific and are commonly expressed in normal and benign conditions.70,71 The fact that maspin was detected in pancreatic cancer but was not expressed in normal pancreas tissues suggests that maspin could serve as a useful marker for primary human pancreatic cancer. Based on these preliminary studies, large numbers of separate human tumor samples in which long-term clinical follow-up is available has to be conducted to evaluate if nuclear versus cytoplasmic maspin expression is indeed a useful prognosis of predictive factor in breast cancer. The function of maspin as a tumor suppressor gene involved in tumor invasion, metastasis and angiogenesis may not be limited to breast and prostate cancer. Its relationship to the pancreas carcinoma unfolds another angle to the study of its function(s) in cancers.

ch14

Jan. 10, 2007

8:57

SPI-B429



5. Other Serpins in Malignancy Thus for, this paper has focused on the tumor suppressing properties of maspin. However, many other serpins can also regulate tumor cell invasion by directly inhibiting the extra-cellular matrix degrading enzymes. A good example of this is the ability of plasminogen activator inhibitor I to coordinately regulate the activity of surface bound urokinase.72 More recently, a growing number of other serpins or biologically active peptides derived from them have been shown to block tumor cell growth by inhibiting angiogenesis (Table 3). Very recently, the archetypal serpin α1 -antitrypsin was shown to inhibit angiogenesis by inducing apoptosis and inhibiting chemotaxis of endothelial cells.73 Like maspin, this activity was not dependent on the C-terminal RSL, indicating that the structural feature(s) responsible for this activity must be localized elsewhere within the molecule. Similar anti-angiogenic activity has been reported for native forms of other serpins, in particular, PAI-1, PEDF, kallistatin, and angiotensinogen.54,74–76 Inhibition by PAI-1 is dose-dependent in animal models of tumor growth: low concentrations stimulate angiogenesis, while high concentrations inhibit angiogenesis and tumor growth. Inhibition occurs by two overlapping pathways; the first is dependent on the inhibitory activity of PAI-1, while the second is independent and acts through binding to vitronectin.77 PEDF is responsible for the avascularity of the ocular compartments, and like maspin, its anti-angiogenic activity appears independent of protease inhibition. PEDF has also no known protease inhibitory function, but it may rather function to block a hypoxiainduced decline in thrombospondin-1.54,78 Kallistatin was first identified as a specific inhibitor of tissue kallikrein, which is localized in vascular smooth muscle and endothelial cells, and plays a role in neointimal hyperplasia by inhibiting angiogenesis and tumor growth.75 Its anti-angiogenic activity is dependent on the presence of a heparin binding domain at residues K312/K313, which interferes with VEGF binding to endothelial cells and suppresses VEGF-induced signaling through the protein kinase Akt.75 Angiotensinogen (ANGIO) is a serpin for which no known protease inhibitory function has been identified. It is the parent molecule from which angiotensin II (ANG II), a key regulator of blood pressure and body fluid homeostasis, is derived and angiotensin I — to which no known function

ch14

Jan. 10, 2007

8:57

SPI-B429


354


has been ascribed. ANGIO and all of its derivatives reportedly affect angiogenesis. ANGIO, ANG I, and the RSL cleaved form of ANGIO all inhibit angiogenesis.78 ANG II differentially regulates angiogenesis through type I (AT1) and type II (AT2) receptors.79 The AT2 receptor is a member of the G-protein-coupled receptor superfamily (GPCRs), which undergoes homoor hetero-oligomerization to induce cell-signaling. The AT2 receptor displays constitutively active homo-oligomerization and is capable of inducing apoptosis without changes in receptor conformation.80 Other, non-native forms and fragments of serpins also display antior pro-angiogenic activity. Antithrombin III is the major inhibitor of thrombin activity in vivo, but its anti-angiogenic activity resides in neo-epitope(s) present in pre-latent, latent, and cleaved forms of the protein.55,81,82 Current data suggests that latent antithrombin disrupts cellmatrix interactions by uncoupling focal adhesion kinase (FAK) and downregulating the expression of perlecan — a proangiogenic heparin sulfate proteoglycan.81,83 Finally, a number of derivatives of the non-inhibitory serpin angiotensinogen that have different effects have been described above. It is clear that these bioactive fragments signal through a group of well-characterized GPCRs to display anti-angiogenic effects similar to those described for other members of the serpin superfamily or peptide fragments derived from them. It remains to be determined if signaling through GPCRs is a common pathway by which serpins act in effecting their angiogenic responses, or if a variety of divergent pathways are involved.

6. Conclusions Serpins comprise a large family of molecules that assume a variety of roles in normal physiology and in the development in cancer.24,84–86 Many serpins are involved in cell migration and adhesion; others are actively involved in regulating proteolysis and apoptosis. Some family members do not inhibit any serine proteases. Interestingly, some inhibitory serpins have evolved additional regulatory functions. For example, plasminogen activator inhibitor 1 (PAI-1) not only specifically inhibits tPA and uPA,87,88 but also regulates cell adhesion which is independent of its protease

ch14

Jan. 10, 2007

8:57

SPI-B429



inhibitor function, by blocking integrin αvβ3 binding to vitronectin.89,90 uPA, uPAR, PAI-1, and PAI-2 play important roles in cancer progression.91 This not only implies that serpins play diverse roles as a class, but also that a single serpin molecule may possess multiple functions.92–95 It is anticipated that additional studies will clearly prove their importance in cancer diagnosis as tumor markers, and that new therapies will be developed using serpins, or structure-based (conformational or peptide) design strategies modeled on serpins to design small molecule(s) that mimic specific serpin domain functions.

References 1. Zou Z, Anisowicz A, Hendrix MJ, et al. (1994) Maspin, a serpin with tumorsuppressing activity in human mammary epithelial cells [see comments]. Science 263(5146):526–529. 2. Sheng S, Pemberton PA and Sager R (1994) Production, purification, and characterization of recombinant maspin proteins. J Biol Chem 269(49):30988– 30993. 3. Sheng S, Carey J, Seftor EA, Dias L, Hendrix MJ and Sager R (1996) Maspin acts at the cell membrane to inhibit invasion and motility of mammary and prostatic cancer cells. Proc Natl Acad Sci USA 93(21):11669–11674. 4. Shao ZM, Nguyen M, Alpaugh ML, O’Connell JT and Barsky SH (1998) The human myoepithelial cell exerts antiproliferative effects on breast carcinoma cells characterized by p21WAF1/CIP1 induction, G2/M arrest, and apoptosis. Exp Cell Res 241(2):394–403. 5. Jiang N, MengY, Zhang S, Mensah-Osman E and Sheng S (2002) Maspin sensitizes breast carcinoma cells to induced apoptosis. Oncogene 21(26):4089– 4098. 6. Zhang M, Volpert O, Shi YH and Bouck N (2000) Maspin is an angiogenesis inhibitor. Nat Med 6(2):196–199. 7. Blacque OE and Worrall DM (2002) Evidence for a direct interaction between the tumour suppressor serpin maspin, and types I and III collagen. J Biol Chem. 8. Seftor RE, Seftor EA, Sheng S, Pemberton PA, Sager R and Hendrix MJ (1998) Maspin suppresses the invasive phenotype of human breast carcinoma. Cancer Res 58(24):5681–5685. 9. Sheng S, Truong B, Fredrickson D, Wu R, Pardee AB and Sager R (1998) Tissue-type plasminogen activator is a target of the tumor suppressor gene maspin. Proc Natl Acad Sci USA 95(2):499–504.

ch14

Jan. 10, 2007

8:57

SPI-B429

356



10. McGowen R, Biliran H Jr, Sager R and Sheng S (2000) The surface of prostate carcinoma DU145 cells mediates the inhibition of urokinase-type plasminogen activator by maspin. Cancer Res 60(17):4771–4778. 11. Odero-Marah VA, Khalkhali-Ellis Z, Chunthapong J, et al. (2003) Maspin regulates different signaling pathways for motility and adhesion in aggressive breast cancer cells. Cancer Biol Ther 2(4):398–403. 12. Al-Ayyoubi M, Gettins PG and Volz K (2004) Crystal structure of human maspin, a serpin with antitumor properties: Reactive center loop of maspin is exposed but constrained. J Biol Chem 279(53):55540–55544. 13. Remold-O’Donnell E (1993) The ovalbumin family of serpin proteins. FEBS Lett 315(2):105–108. 14. Pemberton PA, Tipton AR and Pavloff N, et al. (1997) Maspin is an intracellular serpin that partitions into secretory vesicles and is present at the cell surface. J Histochem Cytochem 45(12):1697–1706. 15. Maass N, Hojo T and Ueding M et al. (2001) Expression of the tumor suppressor gene Maspin in human pancreatic cancers. Clin Cancer Res 7(4):812–817. 16. Smith SL, Watson SG, Ratschiller D, Gugger M, Betticher DC and Heighway J. (2003) Maspin — the most commonly-expressed gene of the 18q21.3 serpin cluster in lung cancer — is strongly expressed in preneoplastic bronchial lesions. Oncogene 22(54):8677–8687. 17. Zou Z, Zhang W, Young D, et al. (2002) Maspin expression profile in human prostate cancer (CaP) and in vitro induction of maspin expression by androgen ablation. Clin Cancer Res 8(5):1172–1177. 18. Domann FE, Rice JC, Hendrix MJ and Futscher BW (2000) Epigenetic silencing of maspin gene expression in human breast cancers. Int J Cancer 85(6):805–810. 19. Belin D, Wohlwend A, Schleuning WD, Kruithof EK and Vassalli JD (1989) Facultative polypeptide translocation allows a single mRNA to encode the secreted and cytosolic forms of plasminogen activators inhibitor 2. Embo J 8(11):3287–3294. 20. Sager R, Sheng S, Pemberton P and Hendrix MJ (1996) Maspin: A tumor suppressing serpin. Curr Top Microbiol Immunol 213(Pt 1):51–64. 21. Zhang M, Sheng S, Maass N and Sager R (1997) mMaspin: The mouse homolog of a human tumor suppressor gene inhibits mammary tumor invasion and motility. Mol Med 3(1):49–59. 22. Latha K, Zhang W, Cella N, Shi HY and Zhang M (2005) Maspin mediates increased tumor cell apoptosis upon induction of the mitochondrial permeability transition. Mol Cell Biol 25(5):1737–1748.

ch14

Jan. 10, 2007

8:57

SPI-B429



23. Liu J, Yin S, Reddy N, Spencer C and Sheng S (2004) Bax mediates the apoptosis-sensitizing effect of maspin. Cancer Res 64(5):1703–1711. 24. Silverman GA, Bird PI, Carrell RW et al. (2001) The serpins are an expanding superfamily of structurally similar but functionally diverse proteins. Evolution, mechanism of inhibition, novel functions, and a revised nomenclature. J Biol Chem 276(36):33293–33296. 25. Bettstetter M, Woenckhaus M, Wild PJ et al. (2005) Elevated nuclear maspin expression is associated with microsatellite instability and high tumour grade in colorectal cancer. J Pathol. 26. Ngamkitidechakul C, Warejcka DJ, Burke JM, O’Brien WJ and Twining SS (2003) Sufficiency of the reactive site loop of maspin for induction of cellmatrix adhesion and inhibition of cell invasion: Conversion of ovalbumin to a maspin-like molecule. J Biol Chem. 27. Bass R, Fernandez AM and Ellis V. Maspin inhibits cell migration in the absence of protease inhibitory activity. J Biol Chem 277(49):46845–46848. 28. Stein PE, Leslie AG, Finch JT, Turnell WG, McLaughlin PJ and Carrell RW (1990) Crystal structure of ovalbumin as a model for the reactive centre of serpins. Nature 347(6288):99–102. 29. Elliott PR, Abrahams JP and Lomas DA (1998) Wild-type alpha 1-antitrypsin is in the canonical inhibitory conformation. J Mol Biol 275(3):419–425. 30. Huntington JA, Read RJ and Carrell RW (2000) Structure of a serpin-protease complex shows inhibition by deformation. Nature 407(6806):923–926. 31. Stein PE, Tewkesbury DA and Carrell RW (1989) Ovalbumin and angiotensinogen lack serpin S-R conformational change. Biochem J 262(1): 103–107. 32. Stratikos E, Alberdi E, Gettins PG and Becerra SP (1996) Recombinant human pigment epithelium-derived factor (PEDF): Characterization of PEDF overexpressed and secreted by eukaryotic cells. Protein Sci 5(12):2575–2582. 33. Gettins P, Patston PA and Schapira M (1993) The role of conformational change in serpin structure and function. Bioessays 15(7):461–467. 34. Law RH, Irving JA, Buckle AM et al. (2005) The high resolution crystal structure of the human tumour suppressor maspin reveals a novel conformational switch in the G-helix. J Biol Chem. 35. Pemberton PA, Wong DT, Gibson HL et al. (1995) The tumor suppressor maspin does not undergo the stressed to relaxed transition or inhibit trypsinlike serine proteases. Evidence that maspin is not a protease inhibitory serpin. J Biol Chem 270(26):15832–15837. 36. Pemberton PA (1991) Maspin: Functional insights from a structural perspective, in Hendrix MJ, (ed.) Landes Bioscience p, 20–29.

ch14

Jan. 10, 2007

8:57

SPI-B429

358



37. Fitzpatrick PA, Wong DT, Barr PJ and Pemberton PA (1996) Functional implications of the modeled structure of maspin. Protein Eng 9(7):585–589. 38. Simonovic M, Gettins PG and Volz K (2001) Crystal structure of human PEDF, a potent anti-angiogenic and neurite growth-promoting factor. Proc Natl Acad Sci USA 98(20):11131–11135. 39. Sager R, Sheng S, Pemberton P and Hendrix MJ (1997) Maspin. A tumor suppressing serpin. Adv Exp Med Biol 425:77–88. 40. Shi HY, Zhang W, Liang R et al. (2001) Blocking tumor growth, invasion, and metastasis by maspin in a syngeneic breast cancer model. Cancer Res 61(18):6945–6951. 41. Pittius CW, Hennighausen L, Lee E et al. (1988) A milk protein gene promoter directs the expression of human tissue plasminogen activator cDNA to the mammary gland in transgenic mice. Proc Natl Acad Sci USA 85(16): 5874–5878. 42. Tzeng YJ, Guhl E, Graessmann M and Graessmann A (1993) Breast cancer formation in transgenic animals induced by the whey acidic protein SV40 T antigen (WAP-SV-T) hybrid gene. Oncogene 8(7):1965–1971. 43. Li M, Lewis B, Capuco AV, Laucirica R and Furth PA (2000) WAP-TAg transgenic mice and the study of dysregulated cell survival, proliferation, and mutation during breast carcinogenesis. Oncogene 19(8):1010–1019. 44. Zhang M, Shi Y, Magit D, Furth PA and Sager R (2000) Reduced mammary tumor progression in WAP-TAg/WAP-maspin bitransgenic mice. Oncogene 19(52):6053–6058. 45. McCormick F (2001) Cancer gene therapy: Fringe or cutting edge? Nat Rev 1(1):130–141. 46. Webster MA, Martin-Soudant N, Nepveu A, Cardiff RD and Muller WJ (1998) The induction of uterine leiomyomas and mammary tumors in transgenic mice expressing polyomavirus (PyV) large T (LT) antigen is associated with the ability of PyV LT antigen to form specific complexes with retinoblastoma and CUTL1 family members. Oncogene 16(15):1963–1972. 47. Shi H, Liang R, Templeton N and Zhang M (2002) Inhibition of breast tumor progression by systemic delivery of maspin in a syngeneic tumor model. Mol. Therapy. 48. Folkman J (1986) How is blood vessel growth regulated in normal and neoplastic tissue? G.H.A. Clowes memorial Award lecture. Cancer Res 46(2): 467–473. 49. Folkman J (1989) Successful treatment of an angiogenic disease [editorial]. N Engl J Med 320(18):1211–1212. 50. Folkman J (1995) Tumor angiogenesis. The Mol Basis Cancer 206–232.

ch14

Jan. 10, 2007

8:57

SPI-B429



51. Fidler IJ and Ellis LM (1994) The implications of angiogenesis for the biology and therapy of cancer metastasis [comment]. Cell 79(2):185–188. 52. Folkman J (1995) Angiogenesis in cancer, vascular, rheumatoid and other disease. Nat Med 1(1):27–31. 53. Boudreau N and Myers C (2003) Breast cancer-induced angiogenesis: Multiple mechanisms and the role of the microenvironment. Breast Cancer Res 5(3):140–146. 54. Dawson DW, Volpert OV, Gillis P et al. (1999) Pigment epithelium-derived factor: A potent inhibitor of angiogenesis. Science 285(5425):245–248. 55. O’Reilly MS, Pirie-Shepherd S, Lane WS and Folkman J (1999) Antiangiogenic activity of the cleaved conformation of the serpin antithrombin. Science 285(5435):1926–1928. 56. Cher ML, Biliran HR Jr, Bhagat S et al. (2003) Maspin expression inhibits osteolysis, tumor growth, and angiogenesis in a model of prostate cancer bone metastasis. Proc Natl Acad Sci USA 100(13):7847–7852. 57. Maass N, Hojo T, Rosel F, Ikeda T, Jonat W and Nagasaki K (2001) Down regulation of the tumor suppressor gene maspin in breast carcinoma is associated with a higher risk of distant metastasis. Clin Biochem 34(4):303–307. 58. Maass N, Rolver L, Ziebart M, Nagasaki K and Rudolph P (2002) Maspin locates to the nucleus in certain cell types. J Pathol 197(2):274–275. 59. Hojo T, AkiyamaY, Nagasaki K et al. (2001) Association of maspin expression with the malignancy grade and tumor vascularization in breast cancer tissues. Cancer Lett 171(1):103–110. 60. Zou Z, Gao C, Nagaich AK et al. (2000) p53 regulates the expression of the tumor suppressor gene maspin. J Biol Chem 275(9):6051–6054. 61. Maass N, Teffner M, Rosel F et al. (2001) Decline in the expression of the serine proteinase inhibitor maspin is associated with tumour progression in ductal carcinomas of the breast. J Pathol 195(3):321–326. 62. Mohsin SK, Zhang M, Clark GM and CraigAllred D (2003) Maspin expression in invasive breast cancer: Association with other prognostic factors. J Pathol 199(4):432–435. 63. Machtens S, Serth J, Bokemeyer C et al. (2001) Expression of the p53 and maspin protein in primary prostate cancer: Correlation with clinical features. Int J Cancer 95(5):337–342. 64. Xia W, Lau YK, Hu MC et al. (2000) High tumoral maspin expression is associated with improved survival of patients with oral squamous cell carcinoma. Oncogene 19(20):2398–2403. 65. Fitzgerald M, Oshiro M, Holtan N et al. (2003) Human pancreatic carcinoma cells activate maspin expression through loss of epigenetic control. Neoplasia 5(5):427–436.

ch14

Jan. 10, 2007

8:57

SPI-B429

360



66. Sato N, Fukushima N, Matsubayashi H and Goggins M (2004) Identification of maspin and S100P as novel hypomethylation targets in pancreatic cancer using global gene expression profiling. Oncogene 23(8):1531–1538. 67. Lim YJ, Lee JK, Jang WY et al. (2004) Prognostic significance of maspin in pancreatic ductal adenocarcinoma. Korean J Int Med 19(1):15–18. 68. Parker SL, Tong T, Bolden S and Wingo PA (1996) Cancer statistics, CA Cancer J Clin 46(1):5–27. 69. Sakamoto K, HagaY,Yoshimura R, Egami H,YokoyamaY and Akagi M (1987) Comparative effectiveness of the tumour diagnostics, CA 19-9, CA 125 and carcinoembryonic antigen in patients with diseases of the digestive system. Gut 28(3):323–329. 70. Safi F, Roscher R, Bittner R, Schenkluhn B, Dopfer HP and Beger HG (1987) High sensitivity and specificity of CA 19-9 for pancreatic carcinoma in comparison to chronic pancreatitis. Serological and immunohistochemical findings. Pancreas 2(4):398–403. 71. Maestranzi S, Przemioslo R, Mitchell H and Sherwood RA (1998) The effect of benign and malignant liver disease on the tumour markers CA19-9 and CEA. Ann Clin Biochem 35(Pt 1):99–103. 72. Behrendt N, List K, Andreasen PA and Dano K (2003) The pro-urokinase plasminogen-activation system in the presence of serpin-type inhibitors and the urokinase receptor: Rescue of activity through reciprocal pro-enzyme activation. Biochem J 371(Pt 2):277–287. 73. Huang H, Campbell SC, Nelius T et al. (2004) Alpha1-antitrypsin inhibits angiogenesis and tumor growth. Int J Cancer 112(6):1042–1048. 74. Stefansson S, McMahon GA, Petitclerc E and Lawrence DA (2003) Plasminogen activator inhibitor-1 in tumor growth, angiogenesis and vascular remodeling. Curr Pharm Des 9(19):1545–1564. 75. Miao RQ, Chen V, Chao L and Chao J (2003) Structural elements of kallistatin required for inhibition of angiogenesis. Am J Physiol Cell Physiol 284(6):C1604–c1613. 76. Celerier J, Cruz A, Lamande N, Gasc JM and Corvol P (2002) Angiotensinogen and its cleaved derivatives inhibit angiogenesis. Hypertension 39(2):224–228. 77. Stefansson S, Lawrence DA and Argraves WS (1996) Plasminogen activator inhibitor-1 and vitronectin promote the cellular clearance of thrombin by low density lipoprotein receptor-related proteins 1 and 2. J Biol Chem 271(14):8215–8220. 78. Aparicio S, Sawant S, Lara N, Barnstable CJ and Tombran-Tink J (2005) Expression of angiogenesis factors in human umbilical vein endothelial cells and their regulation by PEDF. Biochem Biophys Res Commun 326(2):387–394.

ch14

Jan. 10, 2007

8:57

SPI-B429



79. Walther T, Menrad A, Orzechowski HD, Siemeister G, Paul M and Schirner M (2003) Differential regulation of in vivo angiogenesis by angiotensin II receptors. Faseb J 17(14):2061–2067. 80. Miura SI, Karnik SS and Saku K (2005) Constitutive active homo-oligomeric angiotensin ll type 2 receptor induces cell signaling. J Biol Chem. 81. Larsson H, Sjoblom T and Dixelius J, et al. (2000) Antiangiogenic effects of latent antithrombin through perturbed cell-matrix interactions and apoptosis of endothelial cells. Cancer Res 60(23):6723–6729. 82. Larsson H, Akerud P, Nordling K, Raub-Segall E, Claesson-Welsh L and Bjork I. (2001) A novel anti-angiogenic form of antithrombin with retained proteinase binding ability and heparin affinity. J Biol Chem 276(15):11996– 12002. 83. Zhang W, Chuang YJ, Swanson R et al. (2004) Antiangiogenic antithrombin down-regulates the expression of the proangiogenic heparan sulfate proteoglycan, perlecan, in endothelial cells. Blood 103(4):1185–1191. 84. Potempa J, Korzus E and Travis J (1994) The serpin superfamily of proteinase inhibitors: structure, function and regulation. J Biol Chem 269(23):15957– 15960. 85. Worrall DM, Blacque OE and Barnes RC (1999) The expanding superfamily of serpins: searching for the real targets. Biochem Soc Trans 27(4):746–750. 86. Whisstock JC, Irving JA, Bottomley SP, Pike RN and Lesk AM (1999) Serpins in the Caenorhabditis elegans genome. Proteins 36(1):31–41. 87. Brunner G and Preissner KT (1994) Pericellular enzymatic hydrolysis: Implications for the regulation of cell proliferation in the vessel wall and the bone marrow. Blood Coagul Fibrinolysis 5(4):625–639. 88. Blasi F (1993) Urokinase and urokinase receptor:A paracrine/autocrine system regulating cell migration and invasiveness. Bioessays 15(2):105–111. 89. Deng G, Curriden SA, Wang S, Rosenberg S and Loskutoff DJ (1996) Is plasminogen activator inhibitor-1 the molecular switch that governs urokinase receptor-mediated cell adhesion and release? J Cell Biol 134(6):1563–1571. 90. Stefansson S and Lawrence DA (1996) The serpin PAI-1 inhibits cell migration by blocking integrin alpha V beta 3 binding to vitronectin [see comments]. Nature 383(6599):441–443. 91. Ossowski L and Aguirre-Ghiso JA (2000) Urokinase receptor and integrin partnership: Coordination of signaling for cell adhesion, migration and growth. Curr Opin Cell Biol 12(5):613–620. 92. Nishibori M, Chikai T, Kawabata M, Ohta J, Ubuka T and Saeki K (1995) Purification of a novel serpin-like protein from bovine brain. Neurosci Res 24(1):47–52.

ch14

Jan. 10, 2007

8:57

SPI-B429

362



93. Becerra SP (1997) Structure-function studies on PEDF. A noninhibitory serpin with neurotrophic activity. Adv Exp Med Biol 425:223–237. 94. Alberdi E, Aymerich MS and Becerra SP (1999) Binding of pigment epithelium-derived factor (PEDF) to retinoblastoma cells and cerebellar granule neurons. Evidence for a PEDF receptor. J Biol Chem 274(44):31605– 31612. 95. Bajou K, Noel A, Gerard RD et al. (1998) Absence of host plasminogen activator inhibitor 1 prevents cancer invasion and vascularization. Nat Med 4(8):923–928.

ch14

Jan. 10, 2007

8:57

SPI-B429


15 Serpins in Malignancy: Prognostic Indicators Hiroshi Kato, Yuko S. Askew, Shugo Nawata, Akihiro Murakami, Yoshinori Suminami and Gary Clayman

1. Introduction Identification of prognosis in malignant diseases is important for discriminating a high-risk group of patients for whom the adjuvant therapy and intense monitoring should be conducted. Generally, prognosis is determined by clinical features such as the extension of tumor, resistance to therapeutics and host defense response, which are the outcomes of complicated biological events, including uncontrolled proliferation, inhibition of apoptosis, neoangiogenesis, and invasion/metastasis. Recently, much attention has been focused on the biological roles of the serine protease inhibitors (serpins) in malignant behaviors of tumor cells.1,2 There are an increasing number of reports indicating that serpins are involved in apoptosis or extension of the tumor, and hence may become prognostic indicators in clinical practice. Typical examples for this are plasminogen activator inhibitors3,4 and squamous-cell carcinoma antigen (SCCA).5 The main purpose of this paper is to review the prognostic value of serpins in cancer, but we will have to first begin with a brief note on how serpins may be involved in the malignant behavior of cancer cells.

363

ch15

Jan. 10, 2007

8:57

SPI-B429

364


H. Kato et al.

2. Serpins and Malignant Behavior Generally, somatic cells in multicellular organisms live at the right place for the right period of time and drive the apoptotic process once they begin to divide. Cancer cells escape from these self-control systems, inhibiting apoptosis and promoting invasion and metastasis. Anti-tumor therapeutics block cell proliferation and promote apoptosis, but some tumor cells develop the ability to resist treatment.6 Thus, the potentials of anti-apoptosis, angiogenesis and cell migration are the most crucial biological events to influence prognosis of cancer. Apoptosis is a complicated mechanism with multistep pathways; serpins are involved in the apoptotic process (Fig. 1). Plasminogen activator inhibitor type 2 (PAI-2, SERPINB2), remains mainly intracellular7 and inhibits apoptosis.8 In fact, Dickinson et al. reported that PAI-2 protects HeLa cells from TNF-α induced apoptosis.9 SCCA is a group

Fig. 1. Serpins involved in apoptosis. SERPINB3 and SERPINB4 (SCCA) interfere with caspases 1 and 9 in apoptosis pathway and inhibit cathepsin G activity to block the infiltration of NK cells or cytotoxic T-lymphocytes into tumor tissues. SERPINB5 (maspin) activates caspase 3 activity. SERPINB9 (PI-9) inhibits granzyme B and protects cells from apoptosis induced by their own granzyme B. SERPINB2 (PAI-2), SERPINB10 (PI-10) and SERPINB13 (PI-13) also inhibit apoptosis, although the detailed mechanisms are unknown. SERPING2 inhibits α2 -antiplasmin. Roles of these serpins in prognosis, however, may become reverse depending on their tissue localizations, such as in host cells or in tumor cells.

ch15

Jan. 10, 2007

8:57

SPI-B429


Serpins in Malignancy: Prognostic Indicators

365

of proteins, divided into two groups, SCCA1 (SERPINB3) and SCCA2 (SERPINB4). SCCA1 inhibits serine proteases (chymotrypsin) and cysteine protease (cathepsin L, K, S, and papain) whereas SCCA2 inhibits serine-proteases (cathepsin G, mast cell chymase).5 Both types of SCCAs reduce the activities of caspases-3 and -9 via down-regulation of the mitogen-activated protein kinase system10 and interfere with apoptosis induced by radiation, TNF-α, ultraviolet rays, and natural killer (NK) cells.11 Treatments with antisense oligonucleotides of SCCA1 or SCCA2 apparently reduced the growth of squamous-cell carcinoma cells inoculated into nude mice.12 SCCA also inhibits cathepsin G and blocks the infiltration of NK cells into tumor tissues.12,13 Maspin (SERPINB5) activates caspase-3 or -8 to enhance apoptosis14 and can be a prognostic indicator.15 Proteinase inhibitor 9 (PI-9, SERPINB9) is generally present in cytotoxic T-lymphocytes (CTLs) and NK cells and inhibit their own granzyme B or caspase 1 to protect them from apoptosis.16–19 PI-9 also inhibits granzyme B in tumor cells inhibiting apoptosis induced by the immune cells or therapeutics.20 Other serpins such as PI-10 (SERPINB10)21 and PI-13 (headpin, hurpin, and SERPINB13)22 are reported to inhibit apoptosis, but clinical significance of these serpins as prognostic indicators remains to be clarified. Angiogenesis is essential for the growth of solid tumors23 and plasminogen activator inhibitor type 1 (PAI-1, SERPINE1) stimulates expression and the release of vascular endothelial growth factor (VEGF), the most important angiogenesis factor.24 Maspin25 and SERPINF1 (pigment epithelium-derived factor, PEDF)26,27 inhibit angiogenesis. The plasminogen activation and inhibition system appears to play the most important role in invasion and metastasis3,28 (Fig. 2). Plasminogen is converted to plasmin by two types of plasminogen activators, the urokinasetype (uPA) and the tissue-type (tPA). tPA generates plasmin for thrombolysis, while uPA generates plasmin to degrade extracellular matrix (ECM) proteins such as vitronectin (VN) and also enhances the binding of uPAR to vitronectin and β2-integrin.3,29 Thus, uPA has been regarded as the main factor involved in cancer extension. uPA is inhibited by two inhibitors, PAI-1 and PAI-2. Plasmin is inhibited by SERPING2 (α2-antiplasmin).30 Besides the inhibition of uPA, PAI-1 binds to vitronectin and interferes with the binding of uPAR to VN or integrins,3,31–34 affecting cell adhesion and

ch15

Jan. 10, 2007

8:57

SPI-B429

366


H. Kato et al.

Fig. 2. Serpins involved in cell migration. SERPINA3 (PAI-3), SERPINB6 (PI-6) and SERPINE1 (PAI-1) inhibit the effect of uPA to activate plasmin. PAI-1 inhibits the bindings of uPAR to extracellular matrix and promotes invasion or metastasis of tumor cells. PAI-1 also promotes internalization and degradation of uPA. SERPING2 (a2-antiplasmin) inhibits plasmin.

motility. However, this is one of the most complicated and paradoxical phenomena, as Palmieri et al. noted that intracellularly transfected PAI-1 stimulated adhesion of uPAR to VN, whereas exogenously added PAI-1 inhibited adhesion to VN and increased cell motility.30 The role of PAI-1 in tumor invasion may reside in the location of this protein. PAI-1 enhances the internalization of uPA/uPAR complex to degrade uPA inside the cell.35 Maspin interferes with the invasion process and can be a prognostic indicator.36 There is no clear evidence that PAI-2 is involved in invasion or metastasis of cancer.3 The role of serpins in the malignant behavior of tumor cells are summarized in Fig. 3.

3. Prognostic Value of Serpins Accumulating data demonstrates that many serpins provide prognostic information in a spectrum of malignancies,1 including SERPINB2 (PAI-2),

ch15

Jan. 10, 2007

8:57

SPI-B429



367

Fig. 3. Serpins involved in malignant behavior of the tumor. These serpins affect on tumor cells or host cells to promote (→) or inhibit ( ) apoptosis, angiogenesis and invasion/metastasis, and eventually influence the prognosis of diseases. Prognostic relevance of some serpins (PAI-2 and maspin) is reverse according to their tissue localization.

SERPINB3 (SCCA1), SERPINB4 (SCCA2), SERPINB5 (maspin), SERPINB9 (PI-9), SERPINB13 (PI-13), SERPINE-1 (PAI-1), and SERPINF1 (PEDF) (Table 1). 3.1. SERPINB2 (PAI-2) PAI-2 inhibits apoptosis and therefore overexpression of PAI-2 relates to poor prognosis.42,43 High PAI-2 and low uPA levels correlated with a poor prognosis in colorectal cancer.42 Osmak et al. demonstrated that PAI-2 levels were increased in patients with myometrial invasion of more than 50% and lymph vascular invasion in endometrial carcinoma.43 Several investigators have found a reverse relationship between high expression of PAI-2 and favorable prognosis.37,38,40,41 Nagayama et al. demonstrated that PAI-2 levels in lung cancer tissue were significantly lower in cases with lymph node involvement than in those cases without lymph node involvement.40 PAI-2 expression was also lower in tumors with peritoneal metastasis in pancreatic cancer.41 Foekens et al. reported that PAI-2 alone was not a prognostic factor in breast cancer, but in those patients with high uPA, PAI-2 was independently associated with overall survival.38 Bouchet et al. found that lower PAI-2 levels in breast cancer tissue are associated with lymph node metastasis and shorter survival time.39 Interestingly, Shiomi et al. noted in esophageal squamous-cell carcinoma

ch15

Jan. 10, 2007

8:57

SPI-B429


368

H. Kato et al. Table 1 Serpins of prognostic value in cancer.

Serpin

SERPINB2a (PAI-2)

Function

Inhibit apoptosis

Prognostic value

References

Favorable

Unfavorable

Breast,37–39 lung,40 Colorectum,42 pancreas41 endometrium,43 esophagus44 Cervix of uterus45–49 head and neck,50–52 skin,53 larynx,54 esophagus,55,56 lung55,57 bladder,58 anus,59 liver60 Breast,71–73 prostate,74

SERPINB3 and Inhibit apoptosis B4 (SCCA1, SCCA2)

SERPINB5a (maspin)

Promote apoptosis Inhibit invasion Inhibit angiogenesis

SERPINB9 (PI-9)

Inhibit apoptosis

SERPINE1 (PAI-1)

Promote invasion

SERPINF1 (PEDF)

Promote apoptosis Inhibit angiogenesis

a Roles

Breast,61–63 mouth,64 ovary78 65,66 gastric, prostate,67 pancreas,15,68 bladder,69 ovary70,a Non-Hodgkin’s lymphoma20,75 Breast,77–80 stomach,81

endometrium,82 ovary83–85 Liver27,76

of these serpins in the prognostic value depend on their tissue localization.

that high expression of PAI-2 in the fibroblasts was correlated with a good prognosis, whereas PAI-2 expression in cancer cells was associated with poor prognosis.44 The role of PAI-2 in tumor progression may depend on its tissue localization. 3.2. SERPINB3 and SERPINB4 (SCCA1 and SCCA2) Since the first report on cervical cancer by Kato et al.,45 the prognostic significance of SCCA has been demonstrated in various malignancies, including cervical cancer,46–49 head and neck cancer,50,51 skin cancer,53 laryngeal cancer,54 esophageal cancer,55,56 lung cancer,55,57 transitional

ch15

Jan. 10, 2007

8:57

SPI-B429



369

cell carcinoma of the bladder,58 and cancer of the anal canal.59 Generally, increased serum SCCA levels reflect progression of disease and poor prognosis in squamous-cell carcinoma of various origins,86 and even in early stages of diseases, elevation of pretreatment serum SCCA levels indicate a relative risk of recurrence.48 Also, elevation of serum SCCA levels indicate the presence of squamous-cell carcinoma components in the tumor that has a higher risk of recurrent disease.58 Recently, Stenman et al.52 reported that SCCA2/SCCA1 mRNA ratio in the primary tumor tissue predicted recurrence in head and neck cancer. Since it has been indicated that SCCA2 is increased in malignant cells,5 determinations of SCCA2 may provide more promising information for evaluating the prognosis of squamous-cell carcinoma. SCCA is also expressed in hepatic cancer tissues, though the clinical significance of SCCA in hepatic cancer remains to be clarified.60 Interestingly, SCCA is produced by lymphocytes or macrophages surrounding tumor tissues, which may be of some clinical value in host defense response against cancer.87 3.3. SERPINB5 (Maspin) Maspin overexpression is associated with favorable prognosis with decreased tumor extension and long survival time in breast carcinoma,61–63 oral squamous-cell carcinoma,64 gastric cancer,65,66 prostate cancer,67 pancreatic cancer,68 and bladder cancer.69 Maass et al. reported that maspin expression is decreased stepwise in the sequence of ductal carcinoma in situ to invasive cancer to lymph node metastasis.15 There are several reports which show that high maspin levels are associated with a poor prognosis.74 Negative or decreased maspin expression appeared to be associated with a short disease-free survival in breast cancer.71 Bieche et al. measured maspin mRNA in breast cancer, using a reverse transcription polymerase chain reaction (RT-PCR) assay, and demonstrated that overexpression of maspin gene in tumor tissue related significantly to shorter relapse-free survival after surgery.72 Umekita et al. demonstrated that maspin expression was increased in invasive mammary ductal carcinoma.73 These authors suggested that these discrepancies may reside in the mutation of the maspin gene, causing loss of normal function which cannot be discriminated by the ordinary antibody. Recently,

ch15

Jan. 10, 2007

8:57

SPI-B429


370

H. Kato et al.

Mohsin et al. demonstrated in breast cancer tissues88 that positive staining of maspin in the nucleus was related to the positive expression of estrogen and progesterone receptors and was also significantly associated with good prognosis, whereas cytoplasmic staining of maspin was related to the negative expression of estrogen receptors and progesterone receptors and was associated with poor prognosis. Sood et al. also noted that nuclear expression of maspin is associated with increased survival, whereas cytoplasmic localization was associated with the poor outcome in ovarian carcinoma.70 The localization of maspin in two different compartments of the cell may suggest different biological and clinical implications. Maspin is one of the p53-tageted genes, and a mutation of p53 gene is often associated with a decrease in maspin expression.89 Ito et al. found an inverse correlation between maspin expression and advanced stage, and deep invasion and abnormal p53 accumulation65 (See Chap. 14 for additional information regarding maspin’s role in the malignant phenotype). 3.4. SERPINB9 (PI-9) Bladergroen et al. demonstrated that PI-9 is expressed not only in T-cell non-Hodgkin’s lymphoma (NHL), but also in B-cell NHL and Hodgkin’s lymphomas.20 Berge et al. found a positive correlation of high PI-9 with a poor prognosis in anaplastic large-cell lymphomas.75 Since PI-9 exerts its protective effects on apoptosis, an increase in PI-9 expression in tumor cells may also induce the resistance against host immune cells or therapeutics. 3.5. SERPINE1 (PAI-1) Increase in PAI-1 expression is associated with poor prognosis in a variety of malignancies. High levels of uPA/PAI-1 complex were associated with poor overall survival and recurrence-free survival in lymph node-negative invasive breast carcinoma.77–79 Harbeck et al. also noted that high levels of uPA/PAI-1 in primary tumor tissue related to shorter disease-free survival in breast cancer and provided clinically relevant information to perform adjuvant chemotherapy with better response to adjuvant therapy.80 Ganesh et al. found a positive relationship between PAI-1 levels and poor overall survival in gastric cancer.81 The elevation of PAI-1 levels in tumor tissue

ch15

Jan. 10, 2007

8:57

SPI-B429



371

appears to correlate with unfavorable prognosis in gynecologic cancer, including endometrial carcinoma82 and ovarian cancer83–85 Since both uPA and PAI-1 often change in a similar fashion to indicate prognosis, PAI-1 may exert its effects in the ways different from inhibiting uPA functions.90 3.6. SERPINF1 (PEDF) PEDF was first discovered by Tombran-Tink et al. as a neurotrophic serpin.91 PEDF is an inhibitor of neovascularization and it also downregulates the expression of matrix metalloproteinase (MMP), and enhances apoptosis and prevents the growth of cells.26 Uehara et al. demonstrated that the positive expression of PEDF in tumor tissue was inversely associated with liver metastasis in pancreatic cancer.76 Since PEDF is particularly increased in the liver,26 the decreased production of PEDF in chronic hepatic disease may enhance angiogenesis and tumor progression in neighboring hapatocellular carcinoma.27 Serum concentrations of PEDF were decreased in proportion to the progression of premalignant liver disease.27 These authors also demonstrated that the survival of patients was significantly longer in PEDF positive cases. Palmieri et al. found the age-related loss of PEDF expression in human endometrial stromal fibroblasts and suggested that this might contribute to the age-related increase in cancer incidence.92 Guan et al. suggested the possibility of PEDF for the treatment of glioma.25 3.7. Other SERPINS with prognostic potential There are several reports suggesting the clinical value of other serpins. SERPINA1 stimulated cell proliferation in breast cancer and liver cancer93 and increased in papillary thyroid carcinoma.94 SERPINA3 (PAI-3) and SERPINB6 (PI-6) showed essentially the same function as PAI-1 and interfered within cell adhesion.33,95 SERPINB1 (MNEI) inhibits neutrophil granule proteases such as neutrophil elastase, protinase-3 and cathepsin G96 and may play a role in promoting apoptosis. SERPINB10 (PI-10) and SERPINB13 (PI-13) protect the cell from apoptosis21 and Spling et al. have reported down-regulation of PI-13 in malignant tissues, when compared with their normal counterparts in oral cavity squamous-cell carcinoma.97

ch15

Jan. 10, 2007

8:57

SPI-B429


372

H. Kato et al.

Further studies will reveal the prognostic relevance of these serpins in clinical practice.

Conclusions Undoubtedly, serpins play important roles in the biologic behavior of a variety of malignancies. A number of reports have clearly demonstrated the prognostic relevance of serpins. However, while reviewing the currently available data regarding the prognostic significance of serpins in human malignancies, it is apparent that both proteinases and their inhibitors often rise or fall together in relation to tumor extension. This may simply reflect the tumor mass to produce both types of proteins. Yet, it is also quite likely that since every proteinase and proteinase inhibitor has at least several functions and since they behave variably according to the milieu, serpins may not bind to a single target proteinase. Biologically, cancer cells tend to be resistant and develop new mechanisms for adapting to varied environments in order to survive. It will be necessary to evaluate the multiple parameters or pathways of serpins to precisely define their roles in malignant potential and to develop serpins as prognostic indicators. Serpins are most likely to become promising tools not only for understanding tumor biology, but also for identifying new targets to be attacked by therapeutics.

Acknowledgment Work in the laboratory of the author was supported in part by Grant-inAid for Scientific Research (12218224) from the Ministry of Education, Science and Culture, Japan.

References 1. Silverman GA, Whisstock JC, Askew DJ, Pak SC, Luke CJ, Cataltepe S, Irving JA and Bird PI (2004) Human clade B serpins (ov-serpins) belong to a cohort of evolutionarily dispersed intracellular proteinase inhibitor clades that protect cells from promiscuous proteolysis. Cell Mol Life Sci 61:301–325. 2. Silverman GA, Bird PI, Carrelll RW, Church FC, Coughlin PB, Gettins GW, Irving JA, Lomas DA, Luke CJ, Moyer RW, Pemberton PA,

ch15

Jan. 10, 2007

8:57

SPI-B429



3.

4.

5. 6. 7.

8.

9.

10.

11.

12.

13.

373

Remold-O’Donnell E, Salvesen GS, Travis J and Whisstock JC (2001) The serpins are an expanding superfamily of structurally similar but functionally diverse proteins. J Biol Chem 36:33293–33296. Andreasen PA, Kjoller L, Ceistensen L and Duffy MJ (1997) The urokinasetype plasminogen activator system in cancer metastasis: A review. Int J Cancer 72:1–22. Duffy MJ and Duggan C (2004) The urokinase plasminogen activator system: A rich source of tumour markers for the individualized management of patients with cancer. Clin Bchem 37:541–548. Kato H (1996) Expression and function of squamous cell carcinoma antigen. Anticancer Res 16:2149–2154. Shah MA and Schwartz GK (2001) Cell cycle-mediated drug resistance. An emerging concept in cancer therapy. Clin Cancer Res 7:2168–2181. Kumar S and Baglioni C (1991) Protection from tumor necrosis-factor mediated cytolysis by overexpression of plasminogen activator inhibitor types 2. J Biol Chem 266:20960–20964. Zhou H-M, Bolon I, Nichols A, Wohlsend A and Vassalli J-D (2001) Overexpression of plasminogen activator inhibitor type 2 in basal keratinocytes enhances papilloma formation in transgenic mice. Cancer Res 61: 970–976. Dickinson JL, Bates EJ, Ferrante A and Antalis TM (1995) Plasminogen activator inhibitor type 2 inhibits tumor necrosis factor alpha-induced apoptosis. Evidence for an alternate biological function. J Biol Chem 270: 27894–27904. Murakami A, SuminamiY, Hirakawa H, Nawata S, Numa F and Kato H (2001) Squamous cell carcinoma antigen suppresses radiation-induced cell death. Br J Cancer 94:851–858. Suminami Y, Nagashima S, Vujanovic MI, Hirabayashi K, Kato H and Whiteside TI (2000) Inhibition of apoptosis in human tumour cells by the tumour-associated serpin, SCC antigen. Br J Cancer 82:981–989. Suminami Y, Nagashima S, Murakami A, Nawata S, Gondo T, Hirakawa H, Numa F, Silverman GA and Kato H (2001) Suppression of a squamous cell carcinoma (SCC)-related serpin, SCC antigen, inhibits tumor growth with increased intratumor infiltration of natural killer cells. Cancer Res 61: 1776–1780. McGettrich S, Barnes RC and Worrall DM (2001) SCCA2 inhibits TNFmediated apoptosis in transfected HeLa cells. The reactive centre loop sequence is essential for this function and TNF-induced cathepsin G is a candidate target. Eur J Biochem 268:5868–5875.

ch15

Jan. 10, 2007

8:57

SPI-B429

374


H. Kato et al.

14. Jiang J, Meng Y, Zhang S, Mensah-Osman E and Sheng S (2002) Maspin sensitizes breast carcinoma cells to induced apoptosis. Oncogene 21: 4089–4098. 15. Maass N, Teffner M, Rosel F, Pawaresch R, Jonat W, Nagasaki K and Rdolph P (2002) Decline in the expression of the serine proteinase inhibitor maspin is associated with tumour progression in ductal carcinomas of the breast. J Pathol 195:321–326. 16. Bird CH, Sutton VR, Sun J, Hirst CE, Novak A, Kumar S, Trapani JA and Bird PI (1998) Selective regulation of apoptosis: The cytotoxic lymphocyte serpin proteinase inhibitor 9 protects against granzyme B-mediated apoptosis without perturbing the Fas cell death pathway. Mol Cell Biol 18: 6387–6398. 17. Sun J, Bird CH, Sutton V, McDonald L, Coughlin PB, de Jong TA, Trapani JA and Bird PI (1996) A cytosolic granzyme B inhibitor related to the viral apoptotic regulator cytokine response modifier A is present in cytotoxic lymphocytes. J Biol Chem 271:27802–27809. 18. Buzza MS, Hirst CE, Bird CH, Hosing P, McKendrick J and Bird PI (2001) The granzyme B inhibitor, PI-9, is present in endothelial and mesothelial cells, suggesting that it protects bystander cells during immune responses. Cell Immunol 25:21–29. 19. Annand RR, Dahlen JR, Sprecher CA, de Dreu P, Foster DC, Mankovich JA, Talanian RV, Kisiel W and Giegel DA (1999) Caspase-1 (interleukin1beta-converting enzyme) is inhibited by the human serpin analogue proteinase inhibitor 9. Biochem J 342:655–665. 20. Bladergroen BA, Meijer CJ, ten Berge RI, Hack CE, Muris JJF, Dukers DF, Chott A, Kazama Y, Oudejans JJ, van Berkum O and Kummer JA (2002) Expression of the granzyme B inhibitor, protease inhibitor 9, by tumor cells in patients with non-Hodgkin and Hodgkin lymphoma: A novel protective mechanism for tumor cells to circumvent the immune system? Blood 99: 233–237. 21. Schleef RR and Chuang RL (2000) Protease inhibitor 10 inhibits tumor necrosis factor alpha-induced cell death. Evidence for the formation of intracellular high M(r) protease inhibitor 10-containing complexes. J Biol Chem 275:26385–26389. 22. Welss T, Sun J, Irving JA, Blum R, Smith AI, Whisstock JC, Pike RN, von MikeczA, Ruzicka T, Bird PI andAbts HF (2003) Hurpin is a selective inhibitor of lysosomal cathepsin L and protects keratinocytes from ultraviolet-induced apoptosis. Biochemistry 42:7381–7389. 23. Folkman J (1985) Tumor angiogenesis. Adv Cancer Res 43:175–203.

ch15

Jan. 10, 2007

8:57

SPI-B429



375

24. Hjortland GO, Lillehammer T, Somme S, Wang J, Halvorsen T, Juell S, Hirschberg H, Fodstad O and Engebraaten O (2004) Plasminogen activator inhibitor-1 increases the expression of VEGF in human glioma cells. Exp Cell Res 294:130–139. 25. Zhang M, Volpart O, Shi YH and Bouck N (2000) Maspin is an angiogenesis inhibitor. Nat Med 6:196–199. 26. Guan M, Pang CP, Yam HF, Cheung KF, Liu WW and Lu Y (2004) Inhibition of glioma invasion by overexpression of pigment epithelium-derived factor. Cancer Gene Ther 11:325–332. 27. Matsumoto K, Ishikawa H, Nishimura D, Hamasaki K, Nakao K and Eguchi K (2004) Antiangiogenic property of pigment epithelium-derived factor in hepatocellular carcinoma. Hepatology 40:252–259. 28. Dano K, Andreasen PA, Grondahl-Hansen K, Kristensen P, Nielsen LS and Skriver L (1985) Plasminogen activators, tissue degradation and cancer. Adv Cancer Res 44:139–266. 29. Wei Y, Lukashev M, Simon DI, Bodary SC, Rosenberg S, Doyle MV and Chapman HA (1994) Regulation of integrin functions by the urokinase receptor. Science 273:1551–1555. 30. Palmieri D, Lee JW, Juliano RL and Church FC (2002) Plasminogen activator inhibitor-1 and -2 increase cell adhesion and motility of MDA-MB-435 breast cancer cells. J Biol Chem 277:40950–40957. 31. Deng G, Curriden SA, Wang S, Rosenberg S and Loskutoff DJ (1996) Is plasminogen activator inhibitor-1 the molecular switch that governs urokinase receptor-mediated cell adhesion and release? J Cell Biol 134:1563–1571. 32. Waltz DA, Natkin LR, Fujita RM, WeiY and Chapman HA (1997) Plasmin and plasminogen activator inhibitor type 1 promote cellular motility by regulating the interaction between the urokinase receptor and vitronectin. J Clin Invest 100:58–67. 33. Stefansson S and Lawrence DA (1996) The serpin PAI-1 inhibits cell migration by blocking integrin alpha V beta 3 binding to vitronectin. Nature 383: 441–443. 34. Zhou A, Huntington JA, Pannu NS, Carrell RW and Read RJ (2001) How vitronectin binds PAI-1 to modulate fibrinolysis and cell migration. Nat Struct Biol 10:541–544. 35. Cubbelis MV, Wun TC and Blasi F (1990) Receptor-mediated internalization and degradation of urokinase is caused by its specific inhibitor PAI-1. EMBO J 9:1079–1085. 36. Sheng S, Carey J, Seftor EA, Dias L, Hendrix MJ and Sager R (1996) Maspin acts at the cell membrane to inhibit invasion and motility of mammary and prostatic cancer cells. Proc Natl Acad Sci USA 93:11669–11674.

ch15

Jan. 10, 2007

8:57

SPI-B429

376


H. Kato et al.

37. Duggan C, Kennedy S, Kramer MD, Barnes C, Elvin P, McDermott E, O’Higgins N and Duffy MJ (1997) Plasminogen activator inhibitor type 2 in breast cancer. B J Cancer 75:622–627. 38. Foekens JA, Buessecker F, Peters HA, Krainick U, van Putten WL, Loo MP, Klijn JG and Kramer MD (1995) Plasminogen activator inhibitor-2: prognostic relevance in 1012 patients with primary breast cancer. Cancer Res 55: 1423–1427. 39. Bouchet C, Spyratos F, Martin PM, Hacene K, Gentile A and Oglobine J (1994) Prognostic value of urokinase-type plasminogen activator (uPA) and plasminogen activator inhibitors PAI-1 and PAI-2 in breast carcinoma. Br J Cancer 69:398–404. 40. Nagayama M, Sato A, Hayakawa H, Urano T, Takada Y and Takada A (1994) Plasminogen activators and their inhibitors in non-small cell lung cancer. Low content of type 2 plasminogen activator inhibitor associated with tumor dissemination. Cancer 73:1398–1405. 41. Takeuchi Y, Nakao A, Harada A, Nonami T, Fukatsu T and Takagi H (1993) Expression of plasminogen activators and their inhibitors in human pancreatic carcinoma: immunohistochemical study. Am J Gastroenterol 88: 1928–1933. 42. Ganesh S, Sier CFM, Griffioen G, Vloedgraven HJ, de Boer A, Welvaart H, van de Velde CJ, van Krieken JH, Verheijen JH, Lamers CBHW and Verspaget HW (1994) Prognostic relevance of plasminogen activators and their inhibitors in colorectal cancer. Cancer Res 54:4065–4071. 43. Osmak M, Babic D, Abramic M, Milicic D, Vrhovec I and Skrk J (2001) Plasminogen activator inhibitor type 2: Potential prognostic factor for endometrial carcinomas. Neoplasma 48:462–467. 44. Shiomi H, Eguchi Y, Tani T, Kodama M and Hattori T (2000) Cellular distribution and clinical value of urokinase-type plasminogen activator, its receptor, and plasminogen activator inhibitor-2 in esophageal squamous cell carcinoma. Am J Pathol 156:567–575. 45. Kato H, Morioka H, Aramaki S, Tamai K and Torigoe T (1983) Prognostic significance of the tumor antigen TA-4 in squamous cell carcinoma of the uterine cervix. Am J Obstet Gynecol 145:350–354. 46. Scambia G, Benedetti PP, Foti F, Amoroso M, Salerno G, Ferrandina G, Battaglia F, Greggi S, De Gaetano A, Puglia G and Mancuso S (1994) Squamous cell carcinoma antigen: Prognostic significance and role in the monitoring of neoadjuvant chemotherapy response in cervical cancer. J Clin Oncol 12:2309–2316.

ch15

Jan. 10, 2007

8:57

SPI-B429



377

47. Ngan HYS, Chan SY, Wong LC, Choy DTK and Ma HK (1990) Serum squamous cell carcinoma antigen in the monitoring of radiotherapy treatment response in carcinoma of the cervix. Gynecol Oncol 37:260–263. 48. Duk JM, Groenier KH, de Bruijn HWA, Hollema H, Ten Hoor KA, van der Zee AGJ and Aalders JG (1996) Pretreatment serum SCC antigen: A new prognostic factor in early stage cervical carcinoma. J Clin Oncol 14:111–118. 49. Avall-Lundqvist EH, Sjovall K, Nilsson BR and Eneroth PHE (1992) Prognostic significance of pretreatment serum levels of squamous cell carcinoma antigen and CA 125 in cervical carcinoma. Eur J Cancer 28:1695–1702. 50. Synderman CH, D’Amico F, Wagner R and Eibling DE (1995) A reappraisal of the squamous cell carcinoma antigen as a tumor marker in head and neck cancer. Arch Otolaryngol Head Neck Surg 121:1294–1297. 51. Molina R, Torres MD, Moragas M, Perez-Villa J, Jilella X, Jo J, Farrus B, Gimenez N, Traserra J and Ballesta AM (1996) Prognostic significance of SCC antigen in the serum of patients with head and neck cancer. Tumor Biol 17:81–89. 52. Stenman J, Hedstrom J, Grenman R, Leivo I, Finne P, Palotie A and Oepana A (2001) Relative levels of SCCA2 and SCCA1 mRNA in primary tumors predicts recurrent disease in squamous cell cancer of the head and neck. Int J Cancer 95: 39–43. 53. Yagi H, Danno K and Maruguchi T (1987) Significance of squamous cell carcinoma (SCC)-related antigen in cutaneous SCC: a preliminary report. Arch Dermatol 123:902–906. 54. Lara PC and Cuyas JM (1995) The role of squamous cell carcinoma antigen in the management of laryngeal and hypopharyngeal cancer. Cancer 76:758–764. 55. Mino N, Iio A and Hamamoto K (1988) Availability of tumor-antigen 4 as a marker of squamous cell carcinoma of the lung and other organs. Cancer 62:730–734. 56. Matsuda H, Mori M, Tsujitani S, Ohno S, Kuwano H and Sugimachi K (1990) Immunohistochemical evaluation of squamous cell carcinoma antigen and S-100 protein-positive cells in human malignant esophageal tissues. Cancer 65:2261–2265. 57. Body JJ, Sculier JP, Raymakers N, Paesmans M, Ravez P, Libert P, Richez M, Dabouis G, Lacroix H, Bureau G, Thiriauz J, Lecomte J, Brohee D, Mommen P, Fruling J and Klastersky J (1990) Evaluation of squamous cell carcinoma antigen as a new marker for lung cancer. Cancer 65:1552–1556. 58. Tsukamoto T, Kummamoto Y, Ohmura K, Miyao N, Nanbu A, Takagi Y and Itoh N (1992) Squamous cell carcinoma associated antigen in uroepithelial carcinoma. Urology 40:477–484.

ch15

Jan. 10, 2007

8:57

SPI-B429

378


H. Kato et al.

59. Petrelli NJ, Shaw N, Bhargava A, Herrera B, Stulc JP, Sischy J and Mittelman A (1988) Squamous cell carcinoma antigen as a marker for squamous cell carcinoma of the anal canal. J Clin Oncol 6:782–785. 60. Pontisso P, Calabrese F, Benvegnu L, Lise M, Belluco C, Ruvoletto MC, De Falco S, Marino M, Valente M, Nitti D, Gatta A and Fassina G (2004) Overexpression of squamous cell carcinoma antigen variants in hapatocellular carcinoma. Br J Cancer 90:833–837. 61. Zhang M (2004) Multiple functions of maspin in tumor progression and mouse development. Front Biosci 9:2218–2226. 62. Shi HY, Zhang W, Liang R, Abraham S, Kittrell FS, Medina D and Ahang M (2001) Blocking tumor growth, invasion, and metastasis by maspin in a syngeneic breast cancer model. Cancer Res 61:6945–6951. 63. Zou Z, Anisowicz A, Hendrix MJ, Thor A, Neveu M, Sheng S, Rafidi K, Seftor E and Sager R (1994) Maspin, a serpin with tumor-suppressing activity in human mammary epithelial cells. Science 263:526–529. 64. Xia W, Lau YK, Hu MC, Li L, Johnston DA, Sheng S, El-Naggar A and Hung MC (2000) High tumoral maspin expression is associated with improved survival of patients with oral squamous cell carcinoma. Oncogene 19: 2398–2403. 65. Ito R, Nakayama H, Yoshida K, Oda N and Yasui W (2004) Loss of maspin expression is associated with development and progression of gastric carcinoma with p53 abnormality. Oncol Rep 12:985–990. 66. Wang M-C, Yang Y-M, Li X-H, Dong F and Li Y (2004) Maspin expression and its clinicopathological significance in tumorigenesis and progress cancer. World J Gastroenterol 10:634–637. 67. Zou Z, Zhang W, Young D, Gleave MG, Rennie P, Connell T, Connelly R, Moul J, Srivastava S and Sesterhenn I (2002) Maspin expression profile in human prostate cancer (CaP) and in vitro induction of maspin expression by androgen ablation. Clin Cancer Res 8:1172–1177. 68. Maass N, Hojo T, Ueding M, Luttges J, Kloppel G, Jonat W, Nagasaki K (2001) Expression of the tumor suppressor gene maspin in human pancreatic cancers. Clin Cancer Res 7:812–817. 69. Friedrich MG, Toma MI, Petri S, Cheng JC, Hammerer P, Erbersdobler A and Huland H (2004) Expression of maspin in non-muscle invasive bladder carcinoma: Correlation with tumor angiogenesis and prognosis. Eur J Urol 45:737–743. 70. Sood AK, Fletcher MS, Gruman LM, Coffin JE, Jabbari S, Khaikhali-Ellis Z, Arbour N, Seftor EA and Hendrix MJC (2002) The paradoxical expression of maspin in ovarian cancer. Clin Cancer Res 8:2924–2932.

ch15

Jan. 10, 2007

8:57

SPI-B429



379

71. Maass N, Hojo T, Rosel F, Ikeda T, Jonat W and Nagasaki K (2001) Down regulation of the tumor suppressor gene maspin in breast carcinoma is associated with a higher risk of distant metastasis. Clin Biochem 34:303–307. 72. Bieche I, Girault I, Sabourin J-C, Toziu S, Driouch K,Vidaud M and Lidereau R (2003) Prognostic value of maspin mRNA expression in ER alpha-positive postmenopausal breast carcinomas. Br J Cancer 88:863–870. 73. Umekita Y, Ohi Y, Sagara Y and Yoshida H (2002) Expression of maspin predicts poor prognosis in breast-cancer patients. Int J Cancer 100:452–455. 74. Machtens S (2001) Expression of the p53 and maspin protein in primary prostate cancer: Correlation with clinical features. Int J Cancer 95:337–342. 75. ten Berge RL, Meijer CJLM, Dukers DF, Kummer JA, Bladergroen BA, Vos W, Hack CE, Ossenkoppele GJ and Oudejans JJ (2002) Expression levels of apoptosis-related proteins predict clinical outcome in anaplastic large cell lymphoma. Blood 99:4540–4546. 76. Uehara H, Miyamoto M, Kato K, Ebihara Y, Kaneko H, Hashimoto H, Murakami Y, Hase R, Takahashi R, Mega S, Shichinohe T, Kawarada Y, Itoh T, Okushiba S, Kondo S and Tatoh H (2004) Expression of pigment epitheliumderived factor decreases liver metastasis and correlates with favorable prognosis for patients with ductal pancreatic adenocarcinoma. Cancer Res 64: 3533–3537. 77. Manders P, Tjan-Heijnen VC, Span PN, Grebenchtchikov M, GeurtsMoespot A, van Tienoven DT, Beex LV and Sweep FC (2004) Complex of urokinase-type plasminogen activator with its type 1 inhibitor predicts poor outcome in 576 patients with lymph node negative breast carcinoma. Cancer 101:486–494. 78. Foekens JA, Schmitt M, van Putten JWL, Peter HA, Kramer MD, Janicke F and Klijn JMG (1994) Plasminogen activator inhibitor-1 and breast cancer prognosis. J Clin Oncol 12:1648–1658. 79. Duffy MJ (2004) The urokinase plasminogen activator system: Role in malignancy. Curr Pharm Des 10:39–49. 80. Harbeck N, Kates RE, Look MP, Meijer-Van Gelder ME, Klijn JG, Kruger A, Kiechle M, Janicke F, Schmitt M and Foekens JA (2002) Enhanced benefit from adjuvant chemotherapy in breast cancer patients classified high-risk according to urokinase-type plasminogen activator (uPA) and plasminogen activator inhibitor type 1 (n = 3424). Cancer Res 62:4617–4622. 81. Ganesh S, Sier CF, Heerding MM, van Krieken JH, Griffioen G, Welvaart K, van de Velde CJ, Verheijen JH, Lamers CB and Verspaget HW (1996) Prognostic value of the plasminogen activation system in patients with gastric carcinoma. Cancer 77:1035–1043.

ch15

Jan. 10, 2007

8:57

SPI-B429

380


H. Kato et al.

82. Tecimer C, Doering DL, Goldsmith LJ, Meyer JS, Abdulhay G and Wittliff JL (2001) Clinical relevance of urokinase-type plasminogen activator, its receptor, and its inhibitor type 1 in endometrial cancer. Gynecol Oncol 80:48–55. 83. Tecimer C, Doering DL, Goldsmith LJ, Meyer JS, Abdulhay G and Wittliff JL (2000) Clinical relevance of urokinase-type plasminogen activator, its receptor, and its inhibitor type 1 in ovarian cancer. Int J Gynecol Cancer 10:372–381. 84. van der Burg ME, Henzen-Logmans SC, Berns EM, van Putten WI, Klijn JG and Foekens JA (1996) Expression of urokinase-type plasminogen activator (uPA) and its inhibitor PAI-1 in benign, borderline, malignant primary and metastatic ovarian tumors. Int J Cancer 69:475–479. 85. Chambers SK, Ivins CM and Carcangiu ML (1998) Plasminogen activator inhibitor-1 is an independent poor prognostic factor for survival in advancer stage epithelial ovarian cancer patients. Int J Cancer 79:449–454. 86. Kato H (1992) Sqaumous cell carcinoma antigen, in Sell S (ed.), Serological Cancer Markers, Humana Press, Totowa, NJ, pp. 437–451. 87. Yasumatsu R, Nakashima T, Azuma K, Hirakawa N, Kuratomi Y, Yomita K, Cataltepe S, Silverman GA, Clayman GL and Komiyama S (2001) SCCA1 expression in T-lymphocytes peripheral to cancer cells is associated with the elevation of serum SCC antigen in squamous cell carcinoma of the tongue. Cancer Lett 167:205–213. 88. Mohsin SK, Zhang M, Clark GM and CraigAllred D (2003) Maspin expression in invasive breast cancer: Association with other prognostic factors. J Pathol 199:432–435. 89. Zou Z, Gao C, Nagaich A, Connell T, Saito S, Moul JW, Seth P, Appella E and Srivastava S (2000) p53 regulates the expression of the tumor suppressor gene maspin. J Biol Chem 275:6051–6054. 90. Duffy MJ (2002) Urokinase plasminogen activator and its inhibitor, PAI-1, as prognostic markers in breast cancer. From pilot to level 1 evidence studies. Clin Chem 48:1194–1197. 91. Tombran-Tink J, Chader GG and Johnson LV (1991) PEDF, a pigment epithelium-derived factor with potent neuronal differentiative activity. Exp Eye Res 53:411–414. 92. Palmieri D, Watson JM and Rinehart CA (1999) Age-related expression of PEDF/ECP-1 in human endometrial stromal fibroblasts: Implications for interactive senescence. Exp Cell Res 247:142–147. 93. Congote LF and Temmel N (2004) The C-terminal 26-residue peptide of serpin A1 stimulates proliferation of breast and liver cancer cells: Role of protein kinase C and CD47. FEBS Lett 576:343–347.

ch15

Jan. 10, 2007

8:57

SPI-B429



381

94. Hawthorn L, Stein L, Varma R, Wiseman S, Loree T and Tan D (2004) TIMP1 and SERPIN-A overexpression and TFF3 and CRABP1 underexpression as biomarker for papillary thyroid carcinoma. Head Neck 26:1069–1083. 95. Espana F, Estelles A, Fernandez RJ, Gilabert J, Sanchez-Cuenca J and Griffin JH (1993) Evidence of the regulation of urokinase and tissue type plasminogen activators by the serpin, protein C inhibitor, in semen and blood plasma. Thromb Haemost 70:989–994. 96. Cooley J, Takayama TK, Shapiro SD, Schechter MN and Remold-O’Donnell E (2001) The serpin MNEI inhibits elastase-like and chymotrypsin-like serine proteases through efficient reactions at two active sites. Biochemistry 40:15762–15770. 97. Spring P, Nakashima T, Frederick M, Henderson Y and Clayman G (1999) Identification and cDNA cloning of headpin, a novel differentially expressed serpin that maps to chromosome 18q. Biochem Biophys Res Commun 264: 299–304.

ch15

Jan. 10, 2007 8:57 SPI-B429



ch16

Jan. 10, 2007

8:57

SPI-B429


16 SERPINB7 (Megsin) Nephropathy Toshio Miyata

1. Introduction Mesangial cells play a central role in maintaining both the structure and function of the glomerulus. To elucidate the pathogenesis of renal disease, we performed molecular biological analysis of genes expressed in cultured human mesangial cells.1,2 As a result, we cloned a new human mesangiumpredominant gene, megsin, which is a new member of the serpin clade B family.3 The amino acid sequence in the reactive loop site of megsin exhibits the characteristic features of functional serpins.3 In vitro assays utilizing recombinant megsin confirmed that megsin serves as a functional serpin.4 Northern blot and reverse-transcribed polymerase chain reaction analyses of tissues and cells demonstrated that megsin was expressed predominantly in human mesangial cells.3 These findings were confirmed by in situ hybridization and by immunohistochemistry, using megsin-specific antibodies.5,6 In IgA nephropathy and diabetic nephropathy, megsin mRNA expression in glomeruli was up-regulated.5 A similar up-regulation of megsin was observed in the experimental anti-Thy1 nephritis model of rats.7 Thus, the increased expression of megsin was associated with renal disorders with mesangial expansion and proliferation. Interestingly, a recent study showed an association between megsin polymorphisms and increased susceptibility to IgA nephropathy.8

383

ch16

Jan. 10, 2007

8:57

SPI-B429

384


T. Miyata

2. Mouse Models of Megsin Nephropathy To further understand the role of megsin in mesangial function, we overexpressed the human megsin cDNA in the mouse.4 Two lines of megsin transgenic mice were obtained. Both lines developed progressive mesangial matrix expansion, an increase in the number of mesangial cells, and augmented immune complex deposition. The transgenic model was characterized by the expression of megsin in all tissues, due to the ubiquitous promoter driving the transgene. Although immunohistochemical studies revealed the presence of megsin in many tissues as well as in the nonmesangial areas of the kidney, the pathogenic effects of megsin overexpression were restricted to the glomeruli. The mechanism of glomerular abnormalities remains unknown. We speculate that megsin overexpression leads to mesangial dysfunction, impairs the disposal of immune complexes, and increases mesangial matrix by tipping the balance toward lower matrix degradation. In contrast, histological abnormalities were not evident in megsin knock-out mice (unpublished observations). To further investigate the role of megsin in vivo, we employed the anti-glomerular basement membrane (GBM) nephritis model in megsin transgenic mice.4 The animal model of anti-GBM nephritis shares major pathogenic features with the human proliferative glomerulonephritis. The development of anti-GBM nephritis in megsin transgenic mice revealed a persistent mesangial matrix expansion, supporting the possible role of megsin in regulating matrix deposition.

3. Differential Pathology in Megsin Transgenic Rats Megsin transgenic rats were generated utilizing the same ubiquitous promoter used to create the mouse over-expressers. Rats homozygous for the megsin transgene showed impaired growth, poor weight gain, and death by 10 weeks of age. Western blots revealed that megsin expression was differentially expressed: high in heart, kidney and pancreas, and moderate in lung. Homozygotes had clear evidence of renal and pancreatic dysfunction, biochemically, as shown by the nephrotic syndrome (proteinuria, hypoproteinemia, and elevated cholesterol levels) with progressive deterioration of the renal function (elevated creatinine and BUN) and diabetes

ch16

Jan. 10, 2007

8:57

SPI-B429


SERPINB7 (Megsin) Nephropathy

385

(hyperglycemia with insulin deficiency), respectively. When compared with size-matched wild-type rats, transgenic animals showed a markedly lower creatinine clearance. By histology, numerous large PAS-positive droplets were observed within the cytoplasm of glomerular epithelial cells, proximal and distal tubular cells, and collecting ducts (Fig. 1). Compared with homozygous animals, heterozygotes showed the same distribution pattern of PAS-positive droplets, but the size and number of the droplets were significantly less. In the pancreas of homozygous animals, PAS-positive droplets were found in the exocrine and Langerhans islet cells; however, the majority of Langerhans islet cells had disappeared. On electron microscopy, huge electrondense inclusion bodies were detected in the dilated rough endoplasmic reticulum (ER) of glomerular epithelial (Fig. 2) and ascinar cells. In heterozygotes, the Langerhans islet cells were better preserved.

Fig. 1. PAS staining of a megsin transgenic rat (homozygote) at the age of 8 weeks. Numerous, large, PAS-positive droplets were observed within the cytoplasm of glomerular epithelial cells, distal tubular cells, and collecting ducts (200× magnification).

ch16

Jan. 10, 2007

8:57

SPI-B429

386


T. Miyata

Fig. 2. Electron microscopy of the glomerulus of a megsin transgenic rat (homozygote) at the age of 8 weeks. Note the presence of electron-dense inclusion bodies in rough ER (right, 10,000× magnification; left: 150,000× magnification).

4. Does Megsin Overexpression Lead to a Serpinopathy? Why is the renal pathology different between mice and rats overexpressing megsin? Why is diabetes associated only with megsin transgenic rats, while it is absent in the murine system? What are the molecular properties for PAS-positive droplets and electron-dense inclusion bodies within the ER of transgenic rats over-expressing megsin? In certain α1 -antitrypsin (α1 -AT) deficiencies (see Chaps. 20 and 21), the mutated α1 -AT molecule is retained within the ER as PAS-positive and diastase-resistant (i.e. non-glycogen) droplets. ER retention appears to be secondary to conformational changes leading to serpin polymerization, which in turn leads to hepatocellular injury and cirrhosis. This conformational disorder has been referred to as a serpinopathy.9–11 To date,

ch16

Jan. 10, 2007

8:57

SPI-B429



387

two human serpinopathies have been documented: liver cirrhosis of α1 -AT deficiency and dementia of familial encephalopathy involving mutations of neuroserpin.12,13 We hypothesize that megsin overexpression in transgenic rats might lead to a pathologic condition that mimics a serpinopathy.14 The intracellular PAS-positive inclusions were diastase-resistant, and stained positive with anti-human megsin antibody. Also, the electron dense inclusion bodies were detected in the dilated rough ER by immuno-gold electron microscopy (Fig. 3). MegsinT334R has a mutation at the hinge region.14 This mutated megsin does not undergo conformational transition and it lacks serine peptidase inhibitory activity in vitro. This mutation also hinders megsin polymerization. Transgenic rats expressing megsinT334R under the same ubiquitous promoter showed expressions in the kidney and pancreas that

Fig. 3. Immunoelectron microscopic detection of human megsin in podocytes of a megsin transgenic rat (homozygot) at the age of 8 weeks (50,000× magnification).

ch16

Jan. 10, 2007

8:57

SPI-B429

388


T. Miyata

was comparable to the wild-type megsin transgenic heterozygotes. However, PAS-positive droplets were not detectable in mutant megsin animals. In contrast, they were observed in all wild-type megsin heterozygotes at the same age. Mutant megsin heterozygotes exhibited no abnormalities in biochemical analyses, whereas 13% of wild-type megsin heterozygotes suffered from renal or pancreatic dysfunction. Thus, the preservation of megsin conformational mobility appears to be required for the development of disease manifestations. The phenotypic difference between megsin transgenic mice and rats might reflect different pathogenic mechanisms, depending on the degree of gene expressions. In mice, overexpression of megsin may enhance the inactivation of its target serine protease. Histological abnormalities develop late in life at 40 weeks of age. However, megsin expression in rats was an order of magnitude higher and cellular toxicity occurred within 10 weeks of age. We speculate that megsin transgenic rats fail to manifest mesangial expansion because serpin polymerization leads to early lethality.

5. Megsin Overexpression and Endoplasmic Reticulum Stress Why is megsin overexpression toxic only for the kidney and pancreas, while megsin is also overexpressed in other organs? What are the mechanism(s) underlying the cytotoxicity and tissue damage associated with the megsin overexpresssion? The ER is an intracellular compartment that plays a critical role in the processing, folding and transporting of newly synthesized proteins. All cells regulate the capacity of their ER to process synthesized proteins and adapt to an imbalance between protein load and folding capacity, recently referred to as the ER stress response.15 The ER stress response is triggered by various stimuli and pathophysiological conditions.16 One such stress response is the unfolded protein response (UPR).17 UPR involves transient attenuation of new protein synthesis, degradation of misfolded proteins, expression of a variety of ER stress proteins such as oxygen-regulated protein (ORP) 150.18 and glucose-regulated proteins (GRPs). Under normal conditions, these ER stress inducible proteins serve as protein chaperones, complex with defective proteins, and target them for degradation. During

ch16

Jan. 10, 2007

8:57

SPI-B429



389

stress, UPR may limit accumulation of abnormal proteins within the ER, allowing cells to tolerate the ER stress. When the ER stress overcomes the UPR, the cell can undergo apoptosis by activating caspase 12 and CHOP. By immunohistochemistry, the ER stress inducible chaperones, ORP150 and GRP 78, were markedly up-regulated in the glomeruli of megsin transgenic rats, when compared with those of wild-type animals. By double immunostaining, OPR150-expressed podocytes showed the accumulation of megsin inclusion bodies. In contrast, these ER stress inducible chaperones were not detected in megsinT334R transgenic rats. Interestingly, megsin expression was abundant in the heart, but no PAS-positive droplets, no electron dense-inclusions, and no ER stress inducible chaperones were detected in the heart of homozygotes. By immunohistochemistry, ER stress in the podocytes of megsin transgenic rats was associated with cellular injury, as shown by the positive staining for desmin. A TUNEL assay detected increased apoptosis in renal tubular cells and acinar cells in the pancreas of transgenic homozygotes.

6. Entoplasmic Reticulum Stress and Pancreatic Disease Disappearance of β-cells in megsin transgenic rats might be mediated by the mechanism analogous to that highlighted in diabetic Akita mouse, a counterpart of MODY in human type II diabetes.19,20 In this model, the accumulation of polymers of abnormal insulin within the ER eventually results in β-cell apoptosis. The dilated ER with electron-dense inclusions and immature insulin granules are characteristics on electron microscopy. Indeed, in vitro treatment of β-cells with tunicamycin, a reagent that blocks N-glycosylation and causes retention of unfolded proteins in the ER, results in augmented expressions of ER stress chaperons and CHOP, and leads to the apoptosis.

7. ER Stress and Pancreatic Disease In contrast, a pathophysiological role for ER stress in kidney disease is not understood. Treatment with tunicamycin, a calcium ionophore (depletes

ch16

Jan. 10, 2007

8:57

SPI-B429


390

T. Miyata

ER calcium stores), and SNAP (NO donor), but not hypoxia and hyperglycemia, up-regulates expression of ER stress inducible chaperones in podocytes.21 Recently, Cybulsky et al. demonstrated that complement activation increases ER stress in podocytes as well.22 Thus, induction of ER stress may play a role in podocyte injury, not only in the present megsin model but also in other renal disorders.

Conclusions The phenotypic difference between transgenic mice and rats overexpressing megsin yield two different forms of pathologic injury. In the mouse model, megsin overexpression results in mesangial matrix expansion/proliferation and glomerular injury. In the rat model, megsin overexpression leads to the nephrotic syndrome and diabetes. The disorder in transgenic mice appears to be due to a perturbation of peptidase-inhibitor balance, leading to excess extracellular matrix accumulation and mesangial cell activation. In the rat model, megsin overexpression appears to result in the accumulation of megsin that exceeds the capacity of the ER-stress response and triggers podocyte and islet cell injury/death.

References 1. Yasuda Y, Miyata T, Nangaku M, et al. (1998) Funcional quantiative analysis of the genome in cultured human mesangial cells. Kidney Int 53:154–158. 2. Wada T, Miyata T, Inagi R, et al. (2001) Cloning and characterization of a novel subunit of protein serine/threonine phosphatase 4 from mesangial cells. J Am Soc Nephrol 12:2601–2608. 3. Miyata T, Nangaku M, Suzuki D, et al. (1998) A mesangium-predominant gene, megsin, is a new serpin up-regulated in IgA nephropathy. J Clin Invest 102:828–836. 4. Miyata T, Inagi R, Nangaku M, et al. (2002) Overexpression of the serpin megsin induces progressive mesangial cell proliferation and expansion. J Clin Invest 109:585–593. 5. Suzuki D, Miyata T, Nangaku M, et al. (1999) Expression of megsin mRNA, a novel mesangium-predominant gene, in the renal tissues of various glomerular diseases. J Am Soc Nephrol 10:2606–2613.

ch16

Jan. 10, 2007

8:57

SPI-B429



391

6. Inagi R, Miyata T, Suzuki D, et al. (2001) Specific tissue distribution of megsin, a novel serpin, in the glomerulus and its up-regulation in IgA nephropathy. Biochem Biophys Res Commun 286:1098–106. 7. Nangaku M, Miyata T, Suzuki D, et al. (2001) Cloning of rodent megsin revealed its up-regulation on mesangioproliferative nephritis. Kidney Int 60: 641–652. 8. Li Y, Du Y, Li CX, et al. (2004) Family-based association study showing that immunoglobulin A nephropathy is associated with the polymorphisms 2093C and 2180T in the 3 untranslated region of the megsin gene. J Am Soc Nephrol 15:1739–1743. 9. Lomas DA and Carrell RW (2002) Serpinopathies and the conformational dementias. Nat Genetics Rev 3:759–767. 10. Lomas DA and Mahadeva R (2002) Alpha1 -antitrypsin polymerization and the serpinopathies: pathobiology and prospects for therapy. J Clin Invest 110: 1585–1590. 11. Perlmutter DH (2002) Liver injury in alpha1-antitrypsin deficiency: An aggregated protein induces mitochondrial injury. J Clin Invest 110:1579–1583. 12. Davis RL, Shrimpton AE, Holohan PD, et al. (1999) Familial dementia caused by polymerization of mutant neuroserpin. Nature 401:376–379. 13. Carrell RW and Lomas DA (1997) Conformational disease. Lancet 350: 134–138. 14. Inagi R, Miyata T, Nangaku M, et al. (2005) Novel serpinopathy in rat kidney and pancreas induced by overexpression of megsin. J Am Soc Npehrol 16:1339–1349. 15. Pahl HL (1999) Signal transduction from the endoplasmic reticulum to the cell nucleus. Physiol Rev 79:683–701. 16. Aridor M and Balch WE (1999) Integration of endoplasmic reticulum signaling in health and disease. Nat Med 5:745–751. 17. Imai Y, Soda M, Inoue H, et al. (2001) An unfolded putative transmembrane polypeptide, which can lead to endoplasmic reticulum stress, is a substrate of Parkin. Cell 105:891–902. 18. Kitao Y, Ozawa K, Miyazaki M, et al. (2001) Expression of the endoplasmic reticulum molecular chaperone (ORP150) rescues hippocampal neurons from glutamate toxicity. J Clin Invest 108:1439–1450. 19. Ron D (2002) Proteotoxicity in the endoplasmic reticulum: Lessons from the Akita diavetic mouse. J Clin Invest 109:443–445. 20. Oyadomari S, Koizumi A, Takeda K, et al. (2002) Targeted disruption of the Chop gene delays endoplasmic reticulum stress-mediated diabetes. J Clin Invest 109:525–532.

ch16

Jan. 10, 2007

8:57

SPI-B429

392


T. Miyata

21. Inagi R, Pnogi H, Nangaku M, et al. (2005) Involvement of endoplasmic reticulum (ER) stress in podocyte injury associated with a novel transgenic model of serpinoapthy. Kidney Int 68:2639–2650. 22. CybulskyAV, Takano T, Papillon J, et al. (2002) Complement C5b-9 membrane attack complex increases expression of endoplasmic reticulum stress proteins in glomerular epithelial cells. J Biol Chem 277:41342–41351.

ch16

Jan. 10, 2007

8:57

SPI-B429


17 Serpin-Protein and Receptor Interactions Toni M. Antalis and Dudley K. Strickland

1. Introduction Serpins are associated with a range of biological activities that are regulated by interaction with non-protease ligands. Cellular responses to serpins or serpin-protease complexes are found to affect cytoskeletal rearrangements, cell adhesion and cell migration, cell growth and apoptosis. The serpin molecular structure and associated conformational changes are not only finely tuned for protease inhibition, but may also mediate biological activities through additional mechanisms. Serpins bind to signaling receptors, and there is emerging evidence for uncomplexed serpin interactions with non-receptor proteins that may contribute to a diverse range of biological activities. Serpin-binding proteins may include such diverse molecules as transcription factors and chromatin, as well as extracellular matrix components. In this review, we will consider the roles of signaling receptors of the LDL receptor family in the binding of serpins and serpin complexes, and will focus on LRP as a major signaling receptor. We will then consider other less well-characterized protein interactions associated with uncomplexed serpins.

2. The LDL Receptor Family The low density lipoprotein LDL receptor family consists of several related scavenger receptors that not only function as important cargo transporters, 393

ch17

Jan. 10, 2007

8:57

SPI-B429

394


T. M. Antalis & D. K. Strickland

but also participate in modulating signaling responses that inform the cell of changes in its environment (for review, see Refs. 1 and 2). This receptor family consists of six structurally related members, which include the LDL receptor, the very low density lipoprotein (VLDL) receptor, apolipoprotein E (apoE) receptor-2, LDL receptor-related protein 1 (LRP or LRP-1), LRP1b, and LRP2 (gp330/megalin). These six receptors are structurally related and contain several functional regions. First, they contain clusters of complement-like repeats of varying lengths that are responsible for recognizing ligands.3 Each cluster of complement-like repeats are preceded by an epidermal growth factor (EGF)-precursor homology region, which contains EGF-type repeats and a β-propeller domain that participates in pH-mediated ligand dissociation from the receptor, which occurs in the endosomes.4 Some of the larger members of this receptor family (LRP-1, LRP1b, and LRP-2) contain multiple clusters of complement-type repeats and additional EGF-type repeats adjacent to the plasma membrane. The next important regions are the transmembrane and cytoplasmic domains. Typically, the cytoplasmic domain of this receptor family contains NPxY, YxxL, or LL repeats that are important for interacting with adaptor proteins of the endocytic machinery and signaling pathways. Some receptors such as LRP-1 contain all of these motifs within its cytoplasmic domain. The LDL receptor was the first family member to be characterized and its function seems to be restricted to lipoprotein metabolism,5 although recent work suggests a role in factor VIII catabolism.6 In contrast, lipoprotein metabolism does not appear to be the exclusive function of the other characterized LDL receptor family members such as LRP-1. It is now apparent that cargo transport by the members of the LRP-1 and other LDL receptor family is closely associated with the regulation of cellular physiology and cellular signaling events. In the case of LRP-1, this occurs via its association scaffolding and signaling adaptor proteins mediated by its intracellular domain7 in a phosphorylation-dependent manner,8 and by its ability to function as a co-receptor partnering with other cell surface integral membrane proteins.9–11 In addition to the six closely related LDL receptor family members, other more distantly related members, such as LRP-3, LRP-4, LRP-5, LRP-6, LRP-9 and LR-11, share some but not all of the structural features found in the other family members (for review see Ref. 12).

ch17

Jan. 10, 2007

8:57

SPI-B429


Serpin-Protein and Receptor Interactions

395

3. Serpin–Receptor Interactions Shifman and Pizzo13 were the first to evaluate receptor-mediated serpin clearance. Their studies of the clearance of 125 I-labeled antithrombin III (ATIII) and 125 I-labeled ATIII-thrombin complexes from the circulation in mice revealed that ATIII-thrombin complexes were removed from the plasma much more rapidly than native forms of ATIII. Removal occurred via hepatocytes located in the liver, and importantly, was competed by excess unlabeled ATIII-thrombin complex. These pioneering experiments provided the first evidence for a receptor-mediated process. Subsequent work revealed that α1 -antitrypsin(α1 -PI)-trypsin complexes,14 ATIII-thrombin complexes,14 heparin cofactor II–thrombin complexes15 and α1 -antichymotrypsin:chymotrypsin complex16 could all compete for one another, suggesting the existence of a common receptor specific for serpin protease complexes. In further evaluating the specificity of this receptor system for various forms of serpins, Mast et al.16 observed that while serpin:enzyme complexes were rapidly cleared, the native and cleaved forms of the serpin were not cleared rapidly, revealing that the receptor is specific for serpin-enzyme complexes. Fuchs et al.14 demonstrated that another protease inhibitor, α2 macroglobulin (α2 M) was also cleared by a liver receptor, after forming a complex with a protease. Subsequently, two groups isolated the receptor responsible for the clearance of α2 M-protease complexes, and demonstrated that this receptor was identical to LRP-1, which is abundant in the liver.17–20 Since the in vivo clearance of serpin:enzyme complexes from the circulation was not blocked by α2 M–protease complexes,14 the initial thought was that the α2 M-protease receptor might be distinct from the serpin complex receptor. However, shortly after its identification as an α2 M-protease complex receptor, studies showed that LRP could bind urokinase (uPA):plasminogen activator inhibitor type 1 (PAI-1),21 and tissue plasminogen activator (tPA):PAI-1.22 complexes with high affinity and mediate their internalization. These early studies raised the possibility that LRP may also function as the liver clearance receptor for other serpin-enzyme complexes. This was confirmed when Kounnas et al.23 provided compelling evidence that the hepatic receptor responsible for the removal of serpin-enzyme complexes from the circulation is most likely to

ch17

Jan. 10, 2007

8:57

SPI-B429

396



be LRP-1. This study showed that LRP specifically binds serpin-enzyme complexes (ATIII–thrombin, Heparin cofactor II-thrombin, α1 -antitrypsin (α1 -PI):trypsin) with KD values ranging from 80 to 120 nM. In agreement with the earlier clearance studies, LRP failed to recognize the native or cleaved serpin. The binding to LRP was inhibited by the 39 kDa receptor associated protein (RAP), a known antagonist for LRP and other LDLreceptor family members. Mouse fibroblasts in which the LRP gene was genetically deleted, failed to mediate the cellular internalization and degradation of these complexes, whereas control fibroblasts from wild-type mice did internalize and degrade 125 I-labeled serpin-enzyme complexes. Furthermore, in HepG2 cells, anti-LRP IgG was shown to prevent the uptake and degradation of serpin-enzyme complexes. Finally, the rapid clearance of 125 I-labeled antithrombin-III complexes from the circulation was inhibited by RAP. These results provide compelling evidence that LRP is the hepatic clearance receptor for the serpin-enzyme complexes. The inability of α2 Mprotease complexes to inhibit the binding of serpin-enzyme complexes is explained by the finding that α2 M-protease complexes bind to a region of LRP that is distinct from where serpin-enzyme complexes bind.24 Subsequent work has shown that LRP also binds a number of serpin-enzyme complexes (Table 1).

Table 1 Serpins and serpin:enzyme complexes recognized by LRP. uPA/PAI-1 tPA/PAI-1 Thrombin/PAI-1 Thrombin/anti-thrombin III Thrombin/heparin cofactor II Thrombin/protease nexin-1 Neuroserpin Neuroserpin/tPA Elastase/α1 -anti-trypsin C1s/C1q inhibitor Protease/protein C inhibitor Factor XIa/protease nexin-1

ch17

Jan. 10, 2007

8:57

SPI-B429



397

4. Serpin Structure Recognized by Signaling Receptors The recognition of various serpin:enzyme complexes by a single hepatic clearance receptor suggests the existence of a common epitope on serpins that is recognized by this receptor. Perlmutter and colleagues25 identified a C-terminal peptide derived from α1-PI that appeared to compete with the binding of various serpin:enzyme complexes to HepG2 cells, which raised the possibility that this region constitutes the receptor binding epitope. However, the three-dimensional structure revealed that this peptide is largely buried in the cleaved form of α1-PI, and since the structure of the complexed serpin appears quite similar to the cleaved form, it seems likely that this region does not serve as a receptor recognition epitope. In support of this, mutation of this corresponding region in Heparin cofactor II had little impact on receptor-mediated clearance of protease complexes.80 For most serpins, there is very little information regarding the structural requirements for serpin binding to LRP. Basic residues are known to be important for the interaction of ligands with LRP, as it has been shown that Lys1370 and Lys1374 are essential for the binding of α2M with LRP (Nielsen et al., 1996,81 12909), while Lys256 and Lys270 contribute to the interaction of domain 3 of RAP with LRP.27 It was not surprising then when site-directed mutagenesis studies identified basic residues in PAI-1 important for the interaction of the uPA:PAI-1 complex with LRP.28 Thus, a mutant PAI-1 molecule in which Lys82 and Arg120 were converted to Ala had 10% of binding activity to LRP when complexed to uPA. In addition, a PAI-1 molecule in whichArg78 and Lys124 were converted toAla displayed 5% binding activity when compared with its wild-type counterpart. Arg78 and Lys82 localize to α-helix D of PAI-1, while Arg120 and Lys124 are located within the β-strand 1A in the PAI-1 structure. Steffanson et al.29 identified a high affinity cryptic binding site for LRP on PAI-1, generated upon complex formation with a protease, which is lost upon mutation of Arg76 to Glu76, resulting in loss of binding to LRP. Together, these studies suggest that the basic residues on α-helix D of PAI-1 might contribute to the binding interaction. In the case of one serpin, protease nexin 1 (PN-1), several studies have generated significant insight into the region that is involved with

ch17

Jan. 10, 2007

8:57

SPI-B429

398



LRP recognition.30–32 This serpin is different from other serpin:enzyme complexes in that the PN1:thrombin complex initially binds to cell surface heparin sulphate molecules and is then transferred to LRP for subsequent internalization.30 Thus, a mutant of PN1 (Lys7Glu) that renders heparin binding non-functional, showed decreased binding to cells, and was catabolized 5-10-fold less efficiently than the wild-type one.30 To identify the region on PN-1 that is recognized by LRP, overlapping synthetic peptides representing most of the sequence of mature PN-1 were synthesized. These peptides were then screened for their ability to block the LRP-mediated degradation of 125 I-thrombin:PN-1 complexes.32 The result of these studies identified a peptide (PHDNIVISPHGI), which corresponds to Pro47 to Ile58 in PN-1 and was shown to competitively inhibit the LRP-dependent endocytosis of thrombin:PN1 complexes. An antibody prepared against this synthetic peptide inhibited degradation of the PN1:thrombin complex by 70%, but had no effect on the binding of the complex to the cell surfaces.31 Furthermore, point mutations within the corresponding region of PN-1 (His48Ala and Asp49Ala) reduced the catabolism rate to 15% of wild-type.31 This region of PN-1 exists as a loop structure that is bounded by sheet-6B on the amino-terminal side and by helix-B on the carboxyl-terminal side. Since a three-dimensional structure for PN-1 is not yet available, it is not possible to determine if this site is masked in uncomplexed PN-1.

5. Functional Consequences of Receptor Binding The serpin PAI-1 seems to be able to modulate various cellular processes and some of these functions are related to its ability to bind LDL receptor family members. The physiological response occurs when PAI-1 complexes to its target protease (either tPA or uPA) bound to the cell surface and bridges this complex to LDL receptor family members. In the first example, Webb et al.33 found that the addition of uPA to MCF-7 breast cancer cells which expresses the VLDL receptor results in a transient activation of extracellular-signal-regulated kinase (ERK), by a mechanism that requires uPA-binding to its cellular receptor, the urokinase receptor (uPAR). PAI-1 does not independently activate ERK; however, in the presence of PAI-1, the

ch17

Jan. 10, 2007

8:57

SPI-B429



399

duration of ERK activation by uPA is significantly prolonged.33 This effect was blocked by uPAR-specific antibodies or the VLDL receptor antagonist RAP, implicating both uPAR and the VLDL receptor in this process. Furthermore, a mutant form of PAI-1, which fails to bind the VLDL receptor, does not sustain uPA-induced ERK activation. These results suggest that cross-bridging of uPAR with the VLDL receptor by uPA-PAI-1 complex regulates uPAR-initiated cell-signaling, which in turn might have important effects on the response of the cell. Although the mechanism by which PAI-1 sustains uPA-induced ERK activation is not known, it is possible that the VLDL receptor might function to regenerate unligated uPAR by facilitating its recycling back to the cell surface. As the binding of uPA to uPAR causes ERK activation that is transient, uPAR recycling may be necessary to obtain an integrated response that is sufficient to maintain ERK in a continuous state of activation. In support of this pathway, uPA-PAI-1 complexes induce sustained ERK phosphorylation in MCF-7 cells only when uPA-PAI-1 complexes are present in the medium during the course of the incubation.33 PAI-1 can also modulate integrin function by facilitating LRP-mediated internalization of integrins under certain conditions. Thus, Czekay et al.34 found that addition of exogenous PAI-1 and uPA to HT-1080 fibrosarcoma cells resulted in their detachment. This occurred when PAI-1 bound covalently to uPA present in a uPA/uPAR/integrin complex on the cell surface. Formation of the covalent uPA:PAI-1 complex led to internalization of the attached integrin by an LRP-mediated process, which was detected by the accumulation of αv β3 and αv β5 integrins in early endosomal fractions. This accumulation was completely blocked by anti-LRP IgG or RAP, revealing that LRP is required for mediating the uPA/PAI-1 induced internalization of integrins.

6. Binding Interactions of Uncomplexed Serpins Serpins may be linked to biological signaling responses through protease dependent as well as independent mechanisms. The structural and functional relationships associated with protease inhibition have been well studied. Domains of established functional significance include the RCL, the

ch17

Jan. 10, 2007

8:57

SPI-B429

400



hinge region, protease-binding “exo-sites” and accessory ligand-binding domains tacked onto the core serpin structure. The RCL is the region of the serpin that contains the “bait” residue and that which interacts directly with the protease active site. The RCL region is thought to control much but not all of the inhibitory specificity of serpins. Another area related to the RCL is the Hinge region, which encompasses the portion of the RCL that is most proximal to the amino-terminus of the serpin (centered around the P14 residue). This is the area where the RCL makes the transition from strand 5 of β-sheet A to the exposed RCL in an active serpin and is the first portion of the RCL to insert into β-sheet A upon protease inhibition.35–37 The protease exo-sites that influence inhibitory activity and non-protease ligand-binding domains may be present anywhere on the entire molecule, since ligand binding and protein-protein interactions can occur on any face of the serpin. The RCL regions of serpins may have non-inhibitory activities. Maspin is an example of a non-inhibitory serpin38–40 that functions to suppress tumor growth and angiogenesis, and plays a fundamental role in early embryonic development (see Chap. 14). Maspin’s activities appear to occur through non-protease ligand interactions and the RCL, as studies using synthetic maspin RCL peptides as well as maspin/ovalbumin chimeras, reveal that this region is involved in promoting cell adhesion.41,42 The intracellular role of maspin is at present unknown, but maspin has been suggested to play a role in apoptosis pathways.43 and has recently been identified as an interferon regulatory factor 6 (IRF6) binding protein.44 The crystal structure of maspin suggests that conformational switching at the G-helix may mediate protein/protein or ligand interactions.45 Homotypic serpin-serpin interactions may occur as a consequence of the insertion between the RCL of one serpin and the β-sheet of another serpin. This results in the formation of loop sheet dimers and higher order serpin oligomers (see Chaps. 2 and 20). Loop-sheet polymers formed as a consequence of point mutations in serpin sequences lead to the retention of these serpin mutants in the endoplasmic reticulum of cells, and give rise to clinical conditions that result from either protein overload and cell death or plasma deficiency.46–48 These serpinopathies result in clinical syndromes as diverse as cirrhosis, thrombosis, angio-edema, emphysema and dementia. The serpin PAI-2 may also spontaneously form loop-sheet

ch17

Jan. 10, 2007

8:57

SPI-B429



401

polymers, depending on the redox-status of the environment.49 The role of serpin polymerization in normal physiological processes is not known.

7. Non-protease Ligand Binding Domains A serpin domain frequently implicated in non-protease ligand-binding associated with biological responses is the extended polypeptide loop between β-helices C and D (CD loop) of the Clade B serpins. The Clade B or ov-serpins50,51 mostly reside intracellularly with cytoplasmic or nucleocytoplasmic distributions,52 as a result of the absence of a classical secretory signal peptide. Many ov-serpins are reported to have intracellular activities; for example, PAI-2, CrmA and PI9 in apoptosis inhibition, MENT in chromatin condensation, and maspin and headpin in tumor suppression.52–55 PAI-2 is a Clade B serpin with an extended CD-loop in comparison with that found in other ov-serpins.51 This domain is required for its cell survival function,52,56 for interaction of PAI-2 with the retinoblastoma protein (Rb)57 and is a target for transglutamination 26,58 and annexin binding.59 The CD loop of PAI-2 encodes a novel binding motif, called the PENF homology motif, which is shared by many cellular and viral Rb-binding proteins, and which interacts with the C-pocket of the Rb protein.57 Binding of PAI-2 protects the Rb protein from degradation and causes an increase in Rb-mediated activities, potentially explaining why PAI-2 expression is often able to confer a series of Rb-related phenotypes such as resistance to apoptosis,56,60,61 regulation of gene transcription,62–64 promotion of differentiation,65,66 and tumor suppression.60,61,66–69 The CD loop domains of Bomapin and MENT carry nuclear localization signals that presumably interact with nuclear importins.70–73 The MENT CD loop also contains a lamin-like chromatin binding domain, and an A-T hook DNA binding motif. MENT is concentrated in heterochromatin and appears to utilize the positive charges near the CD loop to bind tightly to nucleosomes.73 The mechanism of MENT-induced heterochromatin formation appears to involve two independent events (see Chap. 10). Ordered binding of MENT to linker DNA via its M-loop domain promotes the folding of chromatin, whereas bridging of chromatin fibers is facilitated by MENT oligomerization mediated by the RCL.71

ch17

Jan. 10, 2007

8:57

SPI-B429


402


Other non-protease ligand-binding domains of serpins have been reported to impact biological and physiological responses. For example, β1-ACT is implicated in the pathogenesis of Alzheimer’s disease (AD) through the interaction with the Alzheimer’s amyloid peptide Aβ1-42 74–76 (see Chap. 25). Pigment epithelium-derived factor (PEDF) is a survival factor for various types of neurons and inhibits angiogenesis.77 The region of human PEDF essential for motor neuron protection was identified within a 44 aa N-terminal region (positions 78–121) of the human PEDF molecule that lacks the homologous serpin-reactive region.78 Evidence of a specific 80 kDa PEDF binding protein, which is present in neurons that exhibit neuronal differentiating activity has been reported, although the molecular identity is presently unknown.79

Conclusions Serpins influence diverse range of biological processes through their abilities to inhibit specific proteases, as well as through other mechanisms. The dramatic conformational changes associated with serpins and the diversity of non-protease ligands identified for many of the serpins, suggest serpin functions beyond the regulation of proteolytic reactions. Future studies aimed at identifying additional binding partners and determining how interactions with non-protease ligands influence biological processes in vivo should assist in our understanding of the diverse roles of the serpin family in cellular physiology.

References 1. Herz J and Strickland DK (2001) LRP: A multifunctional scavenger and signaling receptor. J Clin Invest 108:779–784. 2. Strickland DK, Gonias SL and Argraves WS (2002) Diverse roles for the LDL receptor family. Trends Endocrinol Metab 13:66–74. 3. Neels JG, van Den Berg BM, Lookene A, Olivecrona G, Pannekoek H and van Zonneveld AJ (1999) The second and fourth cluster of class A cysteine-rich repeats of the low density lipoprotein receptor-related protein share ligandbinding properties. J Biol Chem 274:31305–31311. 4. Mikhailenko I, Considine W, Argraves KM, Loukinov D, Hyman BT and Strickland DK (1999) Functional domains of the very low density lipoprotein

ch17

Jan. 10, 2007

8:57

SPI-B429



5. 6.

7.

8.

9.

10.

11.

12. 13. 14.

15.

403

receptor: molecular analysis of ligand binding and acid-dependent ligand dissociation mechanisms. J Cell Sci 112 (Pt 19):3269–3281. Brown MS and Goldstein JL (1986) A receptor-mediated pathway for cholesterol homeostasis. Science 232:34–47. Bovenschen N Mertens K, Hu L, Havekes LM and van Vlijmen BJ (2005) LDL receptor cooperates with LDL receptor-related protein in regulating plasma levels of coagulation factor VIII in vivo. Blood 106:906–912. Gotthardt M, Trommsdorff M, Nevitt MF, Shelton J, Richardson JA, Stockinger W, Nimpf J and Herz J (2000) Interactions of the low density lipoprotein receptor gene family with cytosolic adaptor and scaffold proteins suggest diverse biological functions in cellular communication and signal transduction. J Biol Chem 275:25616–25624. Ranganathan S, Liu CX, Migliorini MM, Von Arnim CA, Peltan ID, Mikhailenko I, Hyman BT and Strickland DK (2004) Serine and threonine phosphorylation of the low density lipoprotein receptor-related protein by protein kinase Calpha regulates endocytosis and association with adaptor molecules. J Biol Chem 279:40536–40544. Boucher P, Liu P, Gotthardt M, Hiesberger T, Anderson RG and Herz J (2002) Platelet-derived growth factor mediates tyrosine phosphorylation of the cytoplasmic domain of the low density lipoprotein receptor-related protein in caveolae. J Biol Chem 277:15507–15513. Boucher P, Gotthardt M, Li WP, Anderson RG and Herz J (2003) LRP: Role in vascular wall integrity and protection from atherosclerosis. Science 300: 329–332. Loukinova E, Ranganathan S, Kuznetsov S, Gorlatova N, Migliorini MM, Loukinov D, Ulery PG, Mikhailenko I, Lawrence DA and Strickland DK (2002) Platelet-derived growth factor (PDGF)-induced tyrosine phosphorylation of the low density lipoprotein receptor-related protein (LRP). Evidence for integrated co-receptor function betwenn LRP and the PDGF. J Biol Chem 277:15499–15506. Makarova A, Williams SE and Strickland DK (2004) Proteases and lipoprotein receptors in Alzheimer’s disease. Cell Biochem Biophys 41:139–178. Shifman MA and Pizzo SV (1982) The in vivo metabolism of antithrombin III and antithrombin III complexes. J Biol Chem 257:3243–3248. Fuchs HE, Shifman MA and Pizzo SV (1982) In vivo catabolism of alpha 1-proteinase inhibitor-trypsin, antithrombin III-thrombin and alpha 2macroglobulin-methylamine. Biochim Biophys Acta 716:151–157. Pratt CW, Church FC and Pizzo SV (1988) In vivo catabolism of heparin cofactor II and its complex with thrombin: Evidence for a common

ch17

Jan. 10, 2007

8:57

SPI-B429

404

16.

17.

18.

19.

20.

21.

22.

23.

24.



receptor-mediated clearance pathway for three serine proteinase inhibitors. Arch Biochem Biophys 262:111–117. Mast AE, Enghild JJ, Pizzo SV and Salvesen G (1991) Analysis of the plasma elimination kinetics and conformational stabilities of native, proteinasecomplexed, and reactive site cleaved serpins: Comparison of alpha 1proteinase inhibitor, alpha 1-antichymotrypsin, antithrombin III, alpha 2antiplasmin, angiotensinogen, and ovalbumin. Biochemistry 30:1723–1730. Ashcom JD, Tiller SE, Dickerson K, Cravens JL, Argraves WS and Strickland DK (1990) The human alpha 2-macroglobulin receptor: Identification of a 420-kD cell surface glycoprotein specific for the activated conformation of alpha 2-macroglobulin. J Cell Biol 110:1041–1048. Strickland DK, Ashcom JD, Williams S, Burgess WH, Migliorini M and Argraves WS (1990) Sequence identity between the alpha 2-macroglobulin receptor and low density lipoprotein receptor-related protein suggests that this molecule is a multifunctional receptor. J Biol Chem 265:17401–17404. Moestrup SK and Gliemann J (1989) Purification of the rat hepatic alpha 2-macroglobulin receptor as an approximately 440-kDa single chain protein. J Biol Chem 264:15574–15577. Kristensen T, Moestrup SK, Gliemann J, Bendtsen L, Sand O and Sottrup-Jensen L (1990) Evidence that the newly cloned low-densitylipoprotein receptor related protein (LRP) is the alpha 2-macroglobulin receptor. FEBS Lett 276:151–155. Nykjaer A, Petersen CM, Moller B, Jensen PH, Moestrup SK, Holtet TL, Etzerodt M, Thogersen HC, Munch M and Andreasen PA (1992) Purified alpha 2-macroglobulin receptor/LDL receptor-related protein binds urokinase.plasminogen activator inhibitor type-1 complex. Evidence that the alpha 2-macroglobulin receptor mediates cellular degradation of urokinase receptorbound complexes. J Biol Chem 267:14543–14546. Orth K, Madison EL, Gething MJ, Sambrook JF and Herz J (1992) Complexes of tissue-type plasminogen activator and its serpin inhibitor plasminogenactivator inhibitor type 1 are internalized by means of the low density lipoprotein receptor-related protein/alpha 2-macroglobulin receptor. Proc Natl Acad Sci USA 89:7422–7426. Kounnas MZ, Church FC, Argraves WS and Strickland DK (1996) Cellular internalization and degradation of antithrombin III-thrombin, heparin cofactor II-thrombin, and alpha 1-antitrypsin-trypsin complexes is mediated by the low density lipoprotein receptor-related protein. J Biol Chem 271:6523–6529. Mikhailenko I, Battey FD, Migliorini M, Ruiz JF, Argraves K, Moayeri M and Strickland DK (2001) Recognition of alpha 2-macroglobulin by the low

ch17

Jan. 10, 2007

8:57

SPI-B429



25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

405

density lipoprotein receptor-related protein requires the cooperation of two ligand binding cluster regions. J Biol Chem 276:39484–39491. Joslin G, Fallon RJ, Bullock J, Adams SP and Perlmutter DH (1991) The SEC receptor recognizes a pentapeptide neodomain of alpha 1-antitrypsin-protease complexes. J Biol Chem 266:11282–11288. Jensen PH, Schuler E, Woodrow G, Richardson M, Goss N, Hojrup P, Petersen TE and Rasmussen LK (1994) A unique interhelical insertion in plasminogen activator inhibitor-2 contains three glutamines, Gln83, Gln84, Gln86, essential for transglutaminase-mediated cross-linking. J Biol Chem 269:15394–15398. Migliorini MM, Behre EH, Brew S, Ingham KC and Strickland DK (2003) Allosteric modulation of ligand binding to low density lipoprotein receptorrelated protein by the receptor-associated protein requires critical lysine residues within its carboxyl-terminal domain. J Biol Chem 278:17986–17992. Rodenburg KW, Kjoller L, Petersen HH and Andreasen PA (1998) Binding of urokinase-type plasminogen activator-plasminogen activator inhibitor-1 complex to the endocytosis receptors alpha2-macroglobulin receptor/lowdensity lipoprotein receptor-related protein and very-low-density lipoprotein receptor involves basic residues in the inhibitor. Biochem J 329 (Pt 1):55–63. Stefansson S, Muhammad S, Cheng XF, Battey FD, Strickland DK and Lawrence DA (1998) Plasminogen activator inhibitor-1 contains a cryptic high affinity binding site for the low density lipoprotein receptor-related protein. J Biol Chem 273:6358–6366. Knauer MF, Kridel SJ, Hawley SB and Knauer DJ (1997) The efficient catabolism of thrombin-protease nexin 1 complexes is a synergistic mechanism that requires both the LDL receptor-related protein and cell surface heparins. J Biol Chem 272:29039–29045. Knauer MF, Crisp RJ, Kridel SJ and Knauer DJ (1999) Analysis of a structural determinant in thrombin-protease nexin 1 complexes that mediates clearance by the low density lipoprotein receptor-related protein. J Biol Chem 274: 275–281. Knauer MF, Hawley SB and Knauer DJ (1997) Identification of a binding site in protease nexin I (PN1) required for the receptor mediated internalization of PN1-thrombin complexes. J Biol Chem 272:12261–12264. Webb DJ, Thomas KS and Gonias SL (2001) Plasminogen activator inhibitor 1 functions as a urokinase response modifier at the level of cell signaling and thereby promotes MCF-7 cell growth. J Cell Biol 152:741–752. Czekay RP, Aertgeerts K, Curriden SA and Loskutoff DJ (2003) Plasminogen activator inhibitor-1 detaches cells from extracellular matrices by inactivating integrins. J Cell Biol 160:781–791.

ch17

Jan. 10, 2007

8:57

SPI-B429

406



35. Schulze AJ, Huber R, Degryse E, Speck D and Bischoff R (1991) Inhibitory activity and conformational transition of alpha 1-proteinase inhibitor variants. Eur J Biochem 202:1147–1155. 36. Lawrence DA, Olson ST, Muhammad S, Day DE, Kvassman JO, Ginsburg D and Shore JD (2000) Partitioning of serpin-proteinase reactions between stable inhibition and substrate cleavage is regulated by the rate of serpin reactive center loop insertion into beta-sheet A. J Biol Chem 275:5839–5844. 37. Gils A, Vleugels N, Tobback K, Knockaert I and Declerck PJ (1998) Characterization of plasminogen activator inhibitor 1 mutants containing the P13 to P10 region of ovalbumin or antithrombin III: evidence that the P13 residue contributes significantly to the active to substrate transition. Biochim Biophys Acta 1387:291–297. 38. Pemberton PA, Wong DT, Gibson HL, Kiefer MC, Fitzpatrick PA, Sager R and Barr PJ (1995) The tumor suppressor maspin does not undergo the stressed to relaxed transition or inhibit trypsin-like serine proteases. Evidence that maspin is not a protease inhibitory serpin. J Biol Chem 270:15832–15837. 39. Hopkins PC and Whisstock J (1994) Function of maspin. Science 265:1893–1894. 40. Hopkins PC and Stone SR (1995) The contribution of the conserved hinge region residues of alpha1-antitrypsin to its reaction with elastase. Biochemistry 34:15872–15879. 41. Ngamkitidechakul C, Warejcka DJ, Burke JM, O’Brien WJ and Twining SS (2003) Sufficiency of the reactive site loop of maspin for induction of cellmatrix adhesion and inhibition of cell invasion. Conversion of ovalbumin to a maspin-like molecule. J Biol Chem 278:31796–31806. 42. Blacque OE and Worrall DM (2003) Evidence for a direct interaction between the tumor suppressor serpin, maspin, and types I and III collagen. J Biol Chem 277:10783–10788. 43. Bailey CM, Khalkhali-Ellis Z, Kondo S, Margaryan N, Seftor RE, Wheaton WW, Amir S, Pins MR, Schutte BC and Hendrix MJ (2005) Maspin binds directly to interferon regulatory factor 6: Identification of a novel serpin partnership. J Biol Chem 280:34210–34217. 44. Jiang N, MengY, Zhang S, Mensah-Osman E and Sheng S (2002) Maspin sensitizes breast carcinoma cells to induced apoptosis. Oncogene 21:4089–4098. 45. Law RH, Irving JA, Buckle AM, Ruzyla K, Buzza M, Bashtannyk-Puhalovich TA, Beddoe TC, Nguyen K, Worrall DM, Bottomley SP, Bird PI, Rossjohn J and Whisstock JC (2005) The high resolution crystal structure of the human tumor suppressor maspin reveals a novel conformational switch in the G-helix. J Biol Chem 280:22356–22364.

ch17

Jan. 10, 2007

8:57

SPI-B429



407

46. Lomas DA and Carrell RW (2002) Serpinopathies and the conformational dementias. Nat Rev Genet 3:759–768. 47. Carrell RW and Lomas DA (2002) Alpha1-antitrypsin deficiency–a model for conformational diseases. N Engl J Med 346:45–53. 48. Lomas DA, Belorgey D, Mallya M, Miranda E, Kinghorn KJ, Sharp LK, Phillips RL, Page R, RobertsonAS and Crowther DC (2005) Molecular mousetraps and the serpinopathies. Biochem Soc Trans 33: 321–330. 49. Lobov S, Wilczynska M, Bergstrom F, Johansson LB and Ny T (2004) Structural bases of the redox-dependent conformational switch in the serpin PAI-2. J Mol Biol 344:1359–1368. 50. Benarafa C and Remold-O’donnell E (2005) The ovalbumin serpins revisited: Perspective from the chicken genome of clade B serpin evolution in vertebrates. Proc Natl Acad Sci USA 102:11367–11372. 51. Remold-O’donnell E (1993) The ovalbumin family of serpin proteins. FEBS Lett 315:105–108. 52. Bird CH, Blink EJ, Hirst CE, Buzza MS, Steele PM, Sun J, Jans DA and Bird PI (2001) Nucleocytoplasmic distribution of the ovalbumin serpin PI-9 requires a nonconventional nuclear import pathway and the export factor Crm1. Mol Cell Biol 21:5396–5407. 53. Dickinson JL, Bates EJ, Ferrante A and Antalis TM (1995) Plasminogen activator inhibitor type 2 inhibits tumor necrosis factor alpha-induced apoptosis. Evidence for an alternate biological function. J Biol Chem 270:27894–27904. 54. Silverman GA, Whisstock JC, Askew DJ, Pak SC, Luke CJ, Cataltepe S, Irving JA and Bird PI (2004) Human clade B serpins (ov-serpins) belong to a cohort of evolutionarily dispersed intracellular proteinase inhibitor clades that protect cells from promiscuous proteolysis. Cell Mol Life Sci 61:301–325. 55. Silverman GA, Bird PI, Carrell RW, Church FC, Coughlin PB, Gettins PG, Irving JA, Lomas DA, Luke CJ, Moyer RW, Pemberton PA, RemoldO’donnell E, Salvesen GS, Travis J and Whisstock JC (2001) The serpins are an expanding superfamily of structurally similar but functionally diverse proteins. Evolution, mechanism of inhibition, novel functions, and a revised nomenclature. J Biol Chem 276:33293–33296. 56. Dickinson JL, Norris BJ, Jensen PH and Antalis TM (1998) The C-D interhelical domain of the serpin plasminogen activator inhibitor-type 2 is required for protection from TNF-alpha induced apoptosis. Cell Death Differ 5:163–171. 57. Darnell GA, Antalis TM, Johnstone RW, Stringer BW, Ogbourne SM, Harrich D and Suhrbier A (2003) Inhibition of retinoblastoma protein degradation by interaction with the serpin plasminogen activator inhibitor 2 via a novel consensus motif. Mol Cell Biol 23:6520–6532.

ch17

Jan. 10, 2007

8:57

SPI-B429

408



58. Jensen PH, Lorand L, Ebbesen P and Gliemann J (1993) Type-2 plasminogenactivator inhibitor is a substrate for trophoblast transglutaminase and factor XIIIa. Transglutaminase-catalyzed cross-linking to cellular and extracellular structures. Eur J Biochem 214:141–146. 59. Jensen PH, Jensen TG, Laug WE, Hager H, Gliemann J and Pepinsky B (1996) The exon 3 encoded sequence of the intracellular serine proteinase inhibitor plasminogen activator inhibitor 2 is a protein binding domain. J Biol Chem 271:26892–26899. 60. Harbour JW and Dean DC (2000) The Rb/E2F pathway: Expanding roles and emerging paradigms. Genes Dev 14:2393–2409. 61. Zhou HM, Bolon I, Nichols A, Wohlwend A and Vassalli JD (2001) Overexpression of plasminogen activator inhibitor type 2 in basal keratinocytes enhances papilloma formation in transgenic mice. Cancer Res 61:970–976. 62. Antalis TM, La Linn M, Donnan K, Mateo L, Gardner J, Dickinson JL, Buttigieg K and Suhrbier A (1998) The serine proteinase inhibitor (serpin) plasminogen activation inhibitor type 2 protects against viral cytopathic effects by constitutive interferon alpha/beta priming. J Exp Med 187:1799–1811. 63. Morris EJ and Dyson NJ (2001)Retinoblastoma protein partners. Adv Cancer Res 82:1–54. 64. Shafren DR, Gardner J, Mann VH, Antalis TM and Suhrbier A (1999) Picornavirus receptor down-regulation by plasminogen activator inhibitor type 2. J Virol 73:7193–7198. 65. Jensen PJ, Wu Q, Janowitz P, Ando Y and Schechter NM (1995) Plasminogen activator inhibitor type 2: An intracellular keratinocyte differentiation product that is incorporated into the cornified envelope. Exp Cell Res 217:65–71. 66. Lipinski MM and Jacks T (1999) The retinoblastoma gene family in differentiation and development. Oncogene 18:7873–7882. 67. Foekens JA, Buessecker F, Peters HA, Krainick U, van Putten WL, Look MP, Klijn JG and Kramer MD (1995) Plasminogen activator inhibitor-2: prognostic relevance in 1012 patients with primary breast cancer. Cancer Res 55:1423–1427. 68. Kruithof EK, Baker MS and Bunn CL (1995) Biological and clinical aspects of plasminogen activator inhibitor type 2. Blood 86:4007–4024. 69. Nakamura M, Konno H, Tanaka T, Maruo Y, Nishino N, Aoki K, Baba S, Sakaguchi S, Takada Y and Takada A (1992) Possible role of plasminogen activator inhibitor 2 in the prevention of the metastasis of gastric cancer tissues. Thromb Res 65:709–719.

ch17

Jan. 10, 2007

8:57

SPI-B429



409

70. Chuang TL and Schleef RR (1999) Identification of a nuclear targeting domain in the insertion between helices C and D in protease inhibitor-10. J Biol Chem 274:11194–11198. 71. Springhetti EM, Istomina NE, Whisstock JC, Nikitina T, Woodcock CL and Grigoryev SA (2003) Role of the M-loop and reactive center loop domains in the folding and bridging of nucleosome arrays by MENT. J Biol Chem 278:43384–43393. 72. Irving JA, Shushanov SS, Pike RN, Popova EY, Bromme D, Coetzer TH, Bottomley SP, Boulynko IA, Grigoryev SA and Whisstock JC (2002) Inhibitory activity of a heterochromatin-associated serpin (MENT) against papain-like cysteine proteinases affects chromatin structure and blocks cell proliferation. J Biol Chem 277:13192–13201. 73. Grigoryev SA, Bednar J and Woodcock CL (1999) MENT a heterochromatin protein that mediates higher order chromatin folding, is a new serpin family member. J Biol Chem 274:5626–5636. 74. Mucke L, Yu GQ, McConlogue L, Rockenstein EM, Abraham CR and Masliah E (2000) Astroglial expression of human alpha(1)-antichymotrypsin enhances Alzheimer-like pathology in amyloid protein precursor transgenic mice. Am J Pathol 157:2003–2010. 75. Nilsson LN, Bales KR, DiCarlo G, Gordon MN, Morgan D, Paul SM and Potter H (2001) Alpha-1-antichymotrypsin promotes beta-sheet amyloid plaque deposition in a transgenic mouse model of Alzheimer’s disease. J Neurosci 21:1444–1451. 76. Nilsson LN, Arendash GW, Leighty RE, Costa DA, Low MA, Garcia MF, Cracciolo JR, Rojiani A, Wu X, Bales KR, Paul SM and Potter H (2004) Cognitive impairment in PDAPP mice depends on ApoE and ACT-catalyzed amyloid formation. Neurobiol Aging 25:1153–1167. 77. Dawson DW, Volpert OV, Gillis P, Crawford SE, Xu H, Benedict W and Bouck NP (1999) Pigment epithelium-derived factor: A potent inhibitor of angiogenesis. Science 285:245–248. 78. Bilak MM, Becerra SP, Vincent AM, Moss BH, Aymerich MS and Kuncl RW (2002) Identification of the neuroprotective molecular region of pigment epithelium-derived factor and its binding sites on motor neurons. J Neurosci 22:9378–9386. 79. Alberdi E, Aymerich MS and Becerra SP (1999) Binding of pigment epithelium-derived factor (PEDF) to retinoblastoma cells and cerebellar granule neurons. Evidence for a PEDF receptor. J Biol Chem 274:31605–31612.

ch17

Jan. 10, 2007

8:57

SPI-B429

410



80. Maekawa H and Tollefsen DM (1996) Role of the proposed serpin-enzyme complex receptor recognition site in binding and internalization of thrombinheparin cofactor II complexes by hepatocytes. J Biol Chem 271:18604–18610. 81. Nielsen KL, Holtet TL, Etzerodt M, Moestrup SK, Gliemann J, Sottrup-Jensen L and Thogersen HC (1996) Identification of residues in α-macroglobulins important for binding to the α2 -macroglobulin receptor1 low density lipoprotein receptor-related protein. J Biol Chem 271:12909–12912.

ch17

Jan. 10, 2007

8:57

SPI-B429


18 Serpins and the Metabolic Syndrome Jun Wada

1. Introduction Metabolic syndrome is characterized by abdominal obesity and clustered with atherogenic risk factors, i.e. hypertension, dyslipidemia, hyperglycemia and hyperinsulinemic insulin resistance. Consequently, morbidity and mortality of cardiovascular disease and stroke markedly increased in the subjects with metabolic syndrome. Vigorous efforts were made to delineate the molecular link between increased adiposity and insulin resistance, and it demonstrated that a variety of substances including free fatty acids, leptin,1 tumor necrosis factor-α (TNF-α),2,3 acylation-stimulating protein (ASP),4,5 adiponectin,6,7 resistin8 and visfatin,9 are secreted by adipocytes and they modulate the insulin sensitivity. In addition to insulin sensitivity, several adipocytokines directly affect the vascular functions. For instance, the investigation using globular domain adiponectin (gAd) transgenic (Tg) mice revealed that globular adiponectin can protect against atherosclerosis in vivo6 Thus, the decreased levels of adiponectin in obese subjects may attribute to the development of atherosclerosis in patients with metabolic syndrome. In visceral adipose tissues in genetically obese rats, i.e. Otsuka Long–Evans Tokushima fatty (OLETF) rats, we have found that some serine protease inhibitors (serpins) are differentially expressed during the development of obesity revealed by DNA chip study and PCR-based cDNA subtraction methods.10,11 Although plasminogen activator inhibitor-1 (PAI-1) is a 411

ch18

Jan. 10, 2007

8:57

SPI-B429

412


J. Wada

well-characterized serpin member in the process of metabolic syndrome, the roles of other serpin genes are totally uncharacterized. During investigation of such serpin genes, we identified that Serpina12 is specifically upregulated in visceral adipose tissues of OLETF rats and it is not expressed in subdermal white adipose tissues, as well as other organs. Thus, we designated the new serpin member as visceral adipose tissue-derived serpin protease inhibitor (vaspin). This chapter describes the functional link between serpins and metabolic syndrome, especially newly identified vaspin.

2. Plasminogen Activator Inhibitor-1 and the Metabolic Syndrome Hyper-coagulable state is another feature in patients with diabetes and the metabolic syndrome. Increased platelet reactivity, augmented activity of the coagulation system, and impaired fibrinolysis are reported in these patients. PAI-1 is the most important endogenous inhibitor of tissue plasminogen activator and uro-plasminogen activator, and is a main determinant of fibrinolytic activity.12 PAI-1 also differs from other serpins, i.e. it is a trace protein in plasma, it has a short half-life in vivo, its synthesis is highly regulated, it binds to the adhesive glycoprotein vitronectin (VN) with high affinity and specificity,13 and it is involved in cell adhesion, detachment and migration. In human subjects, plasma PAI-1 levels were closely correlated with the visceral fat area, but not with subcutaneous fat area. PAI-1 mRNA was detected in both types of fat tissues in obese rats, but increased only in visceral fat during the development of obesity.14 Fibrinolytic compensation for hypercoagulation is incomplete in obese patients with type 2 diabetes, partly because of elevated PAI-1 in the blood.15 Accelerated atherothrombotic process in the metabolic syndorme is attributed not only to the metabolic abnormalities but also to a specific inflammatory state that leads to increased plasma PAI-1 levels, which is derived from abdominal adipose tissues.

3. α1-Antitrypsin and the Metabolic Syndrome As visceral and subcutaneous adipose tissues are the major sources of cytokines (adipocytokines), increased adipose tissue mass is associated

ch18

Jan. 10, 2007

8:57

SPI-B429


Serpins and the Metabolic Syndrome

413

with alteration in adipocytokine production, i.e. overexpression of TNF-α, interleukin-6 (IL-6), and PAI-1. Adipose tissue expression of inflammatory cytokines is a potential mechanism linking obesity with metabolic syndrome. The elevated levels of inflammation-sensitive plasma proteins such as fibrinogen, orosomucoid, haptoglobin, ceruloplasmin and α1-antitrypsin are also associated with insulin resistance and endothelial dysfunction, i.e. the initiation step of the atherosclerotic process. Markers of inflammation are associated with the development of diabetes, future weight gain and cardiovascular events.16,17 Reduction of adipose tissue mass through weight reduction in association with exercise reduces TNF-α, IL-6 and PAI-1, increases adiponectin, and is associated with improved insulin sensitivity and endothelial function.18–20 α1-antitrypsin is a serum protease inhibitor that is synthesized mainly in the liver and its rate of synthesis markedly increases in response to inflammation. The expression of α1-antitrypsin is transcriptionally regulated by hepatocyte nuclear factor-1 (HNF-1) and HNF-4α.21,22 The C-terminal fragment (C-36) generated during the cleavage of α1-antitrypsin by proteinases is present in atherosclerotic plaques, particularly within the fibrous cap at the base of the lipid core. Human monocyte stimulation with C-36 fibrils led to a strong activation of both peroxisome proliferator-activated receptors (PPARα and PPARγ), activator protein-1 (AP-1) and nuclear factor-κB (NF-κB).23 The C-36 peptide binds to both LDL and CD36 scavenger receptors, which involves selective up-regulation of pro-inflammatory molecules and activation of the respiratory burst in human monocytes.24 These data suggested the fact that α1-antitrypsin deficiency may protect against atherosclerosis is biologically plausible; however, α1-antitrypsin deficiency is a potential risk factor for arterial aneurysm and cerebrovascular disease.25,26 The mechanism of this connection may be due to the weakening of the arterial wall by the destruction of elastic lamellae.

4. Gene Expression of Serpins in Visceral Adipose Tissues of Otsuka Long–Evans Tokushima Fatty Rats Otsuka Long-Evans Tokushima fatty rat is an animal model of metabolic syndrome, characterized by abdominal obesity, insulin resistance,

ch18

Jan. 10, 2007

8:57

SPI-B429

414


J. Wada

hypertension and dyslipidemia.27 We isolated mRNAs visceral adipose tissues of OLETF rats and their diabetes-resistant counterparts, i.e. LongEvans Tokushima Otsuka (LETO) rats. By using a PCR-based subtraction method, we identified that vaspin mRNA was exclusively expressed in visceral adipose tissues of OLETF rats and not found in other tissues (Fig. 1A). In addition to vaspin, we found that serine protease inhibitor (Spin2c) (NM_031531) is expressed in visceral adipose tissues of OLETF rats and is barely detectable in LETO rats (Fig. 1A). Spin2c mRNA was also observed in liver; however, mRNA signals were not detected in other tissues. HomoloGene at National Center for Biotechnology Information (NCBI) indicates that serine (or cysteine) proteinase inhibitor, Clade A (α1 antiproteinase, -antitrypsin), member 3 (SERPINA3) in human, serine (or cysteine) proteinase inhibitor, Clade A, member 3N (Serpina3n) in mouse

Fig. 1. Cloning and identification of vaspin. (A) Northern blot analysis of spin2c and vaspin in various organs of obese 30-week-old OLETF and visceral adipose tissue of lean 6-week-old LETO rats. (B) Automated protein structure homology-modeling by SWISSMODEL predicted the presence of three β-sheets (blue), nine α-helices (red), and reactive center loop (yellow). Source: 2005 by the National Academy of Sciences of the USA.

ch18

Jan. 10, 2007

8:57

SPI-B429



415

and Spin2c in rat are orthologous genes (HomoloGene:40658). We further compared the gene expression in mesenteric white adipose tissues in OLETF and LETO rats, gene expression of serine (or cysteine) proteinase inhibitor, Clade H, member 1 (Serpinh1) and serine (or cysteine) proteinase inhibitor, Clade E, member 1 (Serpine1), were up-regulated in OLETF rats. Recently, Köting et al. studied the expression of human vaspin mRNA in paired samples of visceral and subcutaneous adipose tissues from 196 human subjects28. Vaspin mRNA expression was detectable in 23% of the visceral and 15% of the subcutaneous adipose samples. Visceral vaspin expression correlated with body mass index, % body fat, and 2-hour plasma glucose levels obtained after an oral glucose tolerance test. Subcutaneous vaspin mRNA expression also correlated with waist-hip-ratio, fasting plasma insulin cocentrations and the glucose infusion rate obtained during the steady state of an euglycemic-hyperinsulinemic clamp28. Secretome analysis of primary cultures of human adipose-derived stem cells revealed the presence of multiple serpins; including PAI-1, pigmented epidermal derived factor (PEDF), placental thrombin inhibitor and protease C inhibitor29. These findings suggest that the serpins are associated with the development of obesity and type 2 diabetes29.

5. Production of Recombinant Human Vaspin and its Activity The analysis of the deduced amino acid sequence of rat, mouse and human orthologues, serine (or cysteine) proteinase inhibitor, Clade A (α-1antiproteinase, -antitrypsin), member 12 (SERPINA12) revealed that vaspin has 40.5% identity with α1 -antitrypsin. Automated protein structure homology-modeling by SWISS-MODEL predicted the presence of three β-sheets, nine α-helices, and a reactive center loop, features characteristic of serpin gene family (Fig. 1B). Human vaspin consists of 395 amino acids with a mass = 45.2 kDa and theoretical pI = 9.26. Using pET expression system, ∼10% vaspin exists in soluble fraction and ∼90% in inclusion bodies in E. coli and soluble fractions were used in the investigation of proteinase inhibitory activity of vaspin. Denatured and reduced hen egg lysozyme (HEL) is relatively sensitive to the attack of proteases compared with native HEL (Fig. 2), and it was used as a substrate for assay of serpin

ch18

Jan. 10, 2007

8:57

SPI-B429

416


J. Wada

Fig. 2. Vaspin does not inhibit known serine proteases. (A) Reduced HEL is sensitive to trypsin treatment when compared with native HEL. (B) α1-antitrypsin inhibits protease activity of trypsin, but both (His)6 -tagged recombinant human vaspin (rhVaspin) and digested rhVaspin fails to inhibit tryptic activity. Source: 2005 by the National Academy of Sciences of the USA.

activity. Trypsin degraded reduced HEL and α1 -antitrypsin inhibited the protease activity of trypsin. However, human recombinant vaspin failed to inhibit protease activity of trypsin. Similarly, inhibitory activity of recombinant human vaspin was not observed for the other known proteases such as elastase, urokinase, factor Xa, collagenase and dipeptidyl peptidase.

ch18

Jan. 10, 2007

8:57

SPI-B429



417

6. Regulation of Vaspin Expression in OLETF Rats To investigate regulation of vaspin gene expression in various white adipose tissues (WATs), OLETF rats were subjected to various regimens up to 50 weeks of age, i.e. voluntary exercise (EXE), pioglitazone (TZD), and insulin (INS) therapy (Fig. 3A–C). Body weight of OLETF rats peaked at 30 weeks of age when notable hyperinsulinemia was observed. Thereafter, these control OLETF rats progressively lost weight for the next 20 weeks along with the decline in insulin secretion and the rise in HbA1c levels. In TZD and INS groups, OLETF rats progressively gained weight for the entire duration of the experiment; however, glycemia was comparable to that in the diabetes-resistant LETO rats. Vaspin mRNA was highly expressed in

Fig. 3. Vaspin mRNA expression in OLETF rats. (A) Body weight of OLETF, LETO, and OLETF rats with EXE, OLETF rats administered with pioglitazone (TZD), and insulin (INS). (B) Hemoglobin A1c levels. (C) Fasting IRI levels. Data are mean ± SEM (n = 5 per group). (D) Northern blot analysis of vaspin mRNA in various adipose tissues. β-actin is shown as an internal control. (E) Western blot of vaspin using sera of OLETF, LETO, and OLETF rats with EXE, OLETF rats administered with pioglitazone (TZD), and insulin (INS). Two major bands ∼50 and ∼45 Kd are detected. Source: 2005 by the National Academy of Sciences of the USA.

ch18

Jan. 10, 2007

8:57

SPI-B429

418


J. Wada

mesenteric and retroperitoneal WATs, and to a lesser extent in epididymal WAT in 30-week-old OLETF rats, corresponding to the peak of their body weight and insulin resistance (Fig. 3D); while it was absent both at the age of 6 and 50 weeks. Vaspin mRNA was barely detected in subdermal WATs of OLETF and LETO rats during the entire period. Voluntary exercise suppressed vaspin mRNA expression in visceral WATs, comparable to the levels of diabetes-resistant LETO rats. Long-term administration of pioglitazone notably up-regulated vaspin mRNA expression in subdermal WATs at 30 weeks of age and maintained its levels up to 50 weeks. In contrast, administration of insulin up-regulated its expression in subdermal WAT at 30 weeks and then declined at 50 weeks of age. By Western blot analyses, two major bands ∼50 and ∼45 Kd in rat sera were detected and vaspin serum levels increased in OLETF rats, compared with LETO rats (Fig. 3E). Insulin and pioglitazone treatment increased vaspin serum levels in OLETF rats, and similarly with voluntary exercise, suggesting that vaspin serum levels are determined both by production from WATs and processing or cleavage by unknown serine proteases in the serum.

7. Vaspin Sensitizes Insulin Action in High Fat High Sucrose Chow-induced Obese ICR Mice Since vaspin expression is regulated in genetically obese OLETF rats, the role of vaspin, causative or protective was assessed in the development of type 2 diabetes in diet-induced obese ICR mice. ICR mice with high fat high sucrose (HFHS) chow developed hyperglycemia, hyperinsulinemia and obesity at 18 weeks of age (Fig. 4A-C). Administration of human recombinant vaspin into 18-week-old ICR mice with HFHS chow significantly reduced glucose levels 2 h after the intraperitoneal injection of glucose. However, insulin levels were not altered by the injection of vaspin. Furthermore, a single injection of vaspin sensitized the insulin action in HFHS chow-induced obese ICR mice (p < 0.01), while vaspin did not sensitize the insulin action in 10- and 18-week-old ICR mice fed with standard chows. These data indicate that vaspin lowered blood glucose levels by improving insulin resistance in diet-induced obese mice.

ch18

Jan. 10, 2007

8:57

SPI-B429



419

Fig. 4. Vaspin improves insulin sensitivity. (A) Glucose tolerance test. 1 mg/kg rhVaspin or PBS was intraperitoneally injected to ICR mice with standard starch chow (STD) and HFHS chow 2 and 14 h before glucose administration. Data are mean ± SEM (n = 10 per group). Asterisk (∗), p < 0.01 (unpaired t-test). (B) Insulin levels in glucose tolerance test. Data are mean ± SEM (n = 10 per group). (C) Insulin tolerance test. 1 mg/kg rhVaspin or PBS was intraperitoneally injected into ICR mice with STD and HFHS chow 2 and 14 h before insulin injection. Data are mean ± SEM (n = 10 per group). (∗) p < 0.05, (∗∗) p < 0.01 (unpaired t-test). Source: 2005 by the National Academy of Sciences of the USA.

ch18

Jan. 10, 2007

8:57

SPI-B429


420

J. Wada

8. Vaspin is a New Therapeutic Target of the Metabolic Syndrome Since administration of pioglitazone significantly up-regulates vaspin mRNA levels in the subdermal adipose tissues, it is reasonable to speculate that production of vaspin may antagonize the action of unknown proteases derived from fat or other tissues that impair the insulin action. Similar paradigm is already known in relation to α1-antitrypsin and neutrophil elastase. The α1-antitrypsin is an acute-phase protein derived from liver, its concentration rises during inflammation and thereby it inhibits neutrophil elastase to prevent tissue destruction.30 Since the HFHS chowinduced expression change in genes related to glucose metabolism and adipocytokines almost completely returned to normal levels, and unsupervised principal component and hierarchical clustering analysis of Genechip data indicated that vaspin normalizes the global gene expression of WATs in ICR mice with HFHS chow-induced obesity to insulin-sensitive WAT observed in ICR mice with starch chow. This strongly suggests that vaspin may have insulin sensitizing actions on WAT.11 However, administration of vaspin in in vitro culture of cells derived from these organs did not alter the insulin-induced glucose uptake, suggesting that vaspin could influence the insulin action in the presence of its target proteases. Vascular cells are also thought to be the target of vaspin, since mesenteric fat derived PAI-1 promotes atherosclerosis.14 Additional investigations are required to understand whether vaspin accelerates or decelerates atherosclerosis in the metabolic syndrome. The available data suggest that vaspin has a modulatory role on glucose metabolism, and further identification of its potential substrate proteases may link to the development of anti-protease inhibitor therapy, which could be helpful in improving insulin sensitivity in metabolic syndrome.

Acknowledgments This work was supported by Japan Heart Foundation/Pfizer Grant for Research on Hypertension; Hyperlipidemia and Vascular Metabolism; Yamanouchi Foundation for Research on Metabolic Disorders;

ch18

Jan. 10, 2007

8:57

SPI-B429



421

Grant-in-Aid for Scientific Research (C); Ministry of Education, Science and Culture, Japan (14571025, 17590829).

References 1. Ebihara K, Ogawa Y, Masuzaki H, Shintani M, Miyanaga F, Aizawa-Abe M, Hayashi T, Hosoda K, Inoue G, Yoshimasa Y, Gavrilova O, Reitman ML and Nakao K (2001) Transgenic overexpression of leptin rescues insulin resistance and diabetes in a mouse model of lipoatrophic diabetes. Diabetes 50: 1440–1448. 2. Hotamisligil GS, Peraldi P, Budavari A, Ellis R, White MF and Spiegelman BM (1996) IRS-1-mediated inhibition of insulin receptor tyrosine kinase activity in TNF-alpha- and obesity-induced insulin resistance. Science 271:665–668. 3. Uysal KT, Wiesbrock SM, Marino MW and Hotamisligil GS (1997) Protection from obesity-induced insulin resistance in mice lacking TNF-alpha function. Nature 389:610–614. 4. Xia Z, Sniderman AD and Cianflone K (2002) Acylation-stimulating protein (ASP) deficiency induces obesity resistance and increased energy expenditure in ob/ob mice. J Biol Chem 277:45874–45879. 5. Cianflone K, Xia Z and Chen LY (2003) Critical review of acylationstimulating protein physiology in humans and rodents. Biochim Biophys Acta, 1609:127–143. 6. Yamauchi T, Kamon J, Waki H, ImaiY, Shimozawa N, Hioki K, Uchida S, ItoY, Takakuwa K, Matsui J, Takata M, Eto K, Terauchi Y, Komeda K, Tsunoda M, Murakami K, OhnishiY, Naitoh T,Yamamura K, UeyamaY, Froguel P, Kimura S, Nagai R and Kadowaki T (2003) Globular adiponectin protected ob/ob mice from diabetes and ApoE-deficient mice from atherosclerosis. J Biol Chem 278:2461–2468. 7. Maeda N, Shimomura I, Kishida K, Nishizawa H, Matsuda M, Nagaretani H, Furuyama N, Kondo H, Takahashi M, Arita Y, Komuro R, Ouchi N, Kihara S, Tochino Y, Okutomi K, Horie M, Takeda S, Aoyama T, Funahashi T and Matsuzawa Y (2002) Diet-induced insulin resistance in mice lacking adiponectin/ACRP30. Nat Med 8:731–737. 8. Steppan CM, Bailey ST, Bhat S, Brown EJ, Banerjee RR, Wright CM, Patel HR, Ahima RS and Lazar MA (2001) The hormone resistin links obesity to diabetes. Nature 409:3070–312. 9. Fukuhara A, Matsuda M, Nishizawa M, Segawa K, Tanaka M, Kishimoto K, Matsuki Y, Murakami M, Ichisaka T, Murakami H, Watanabe E, Takagi T,

ch18

Jan. 10, 2007

8:57

SPI-B429

422

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.


J. Wada

Akiyoshi M, Ohtsubo T, Kihara S, Yamashita S, Makishima M, Funahashi T, Yamanaka S, Hiramatsu R, Matsuzawa Y and Shimomura I (2005) Visfatin: A protein secreted by visceral fat that mimics the effects of insulin. Science 307:426–430. Hida K, Wada J, Zhang H, Hiragushi K, Tsuchiyama Y, Shikata K and Makino H (2000) Identification of genes specifically expressed in the accumulated visceral adipose tissue of OLETF rats. J Lipid Res 41:1615–1622. Hida K, Wada J, Eguchi J, Zhang H, Baba M, Seida A, Hashimoto I, Okada T, Yasuhara A, Nakatsuka A, Shikata K, Hourai S, Futami J, Watanabe E, Matsuki Y, Hiramatsu R, Akagi S, Makino H and Kanwar YS (2005) Visceral adipose tissue-derived serine protease inhibitor: A unique insulin-sensitizing adipocytokine in obesity. Proc Natl Acad Sci USA 102:10610–10615. Skurk T and Hauner H (2004) Obesity and impaired fibrinolysis: Role of adipose production of plasminogen activator inhibitor-1. Int J Obes Relat Metab Disord 28:1357–1364. Dellas C and Loskutoff DJ (2005) Historical analysis of PAI-1 from its discovery to its potential role in cell motility and disease. Thromb Haemost 93:631–640. Shimomura I, Funahashi T, Takahashi M, Maeda K, Kotani K, Nakamura T, Yamashita S, Miura M, Fukuda Y, Takemura K, Tokunaga, K and Matsuzawa Y (1996) Enhanced expression of PAI-1 in visceral fat: Possible contributor to vascular disease in obesity. Nat Med 2:800–803. Aso Y, Matsumoto S, Fujiwara Y, Tayama K, Inukai T and Takemura Y (2002) Impaired fibrinolytic compensation for hypercoagulability in obese patients with type 2 diabetes: Association with increased plasminogen activator inhibitor-1. Metabolism, 51:471–476. Schmidt MI, Duncan BB, Sharrett AR, Lindberg G, Savage PJ, Offenbacher S, Azambuja MI, Tracy RP and Heiss G (1999) Markers of inflammation and prediction of diabetes mellitus in adults (Atherosclerosis Risk in Communities study): a cohort study. Lancet 353:1649–1652. Engstrom G, Stavenow L, Hedblad B, Lind P, Eriksson KF, Janzon L and Lindgarde F (2003) Inflammation-sensitive plasma proteins, diabetes, and mortality and incidence of myocardial infarction and stroke: A populationbased study. Diabetes, 52:442–447. Morange PE, Alessi MC and Juhan-Vague I (2004) Relations between hemostatic variables, insulin resistance and inflammation. Hematol J 5(Suppl 3):S15–s19. Aldhahi W and Hamdy O (2003) Adipokines, inflammation, and the endothelium in diabetes. Curr Diab Rep 3:293–298.

ch18

Jan. 10, 2007

8:57

SPI-B429



423

20. Lin Y, Rajala MW, Berger JP, Moller DE, Barzilai N and Scherer PE (2001) Hyperglycemia-induced production of acute phase reactants in adipose tissue. J Biol Chem 276:42077–42083. 21. Pontoglio M (2000) Hepatocyte nuclear factor 1, a transcription factor at the crossroads of glucose homeostasis. J Am Soc Nephrol 11(Suppl 16): S140–S143. 22. Shih DQ, Dansky HM, Fleisher M, Assmann G, Fajans SS, and Stoffel M (2000) Genotype/phenotype relationships in HNF-4alpha/MODY1: Haploinsufficiency is associated with reduced apolipoprotein (AII), apolipoprotein (CIII), lipoprotein(a), and triglyceride levels. Diabetes 49:832–837. 23. Dichtl W, Moraga F, Ares MP, Crisby M, Nilsson J, Lindgren S and Janciauskiene S (2000) The carboxyl-terminal fragment of alpha1-antitrypsin is present in atherosclerotic plaques and regulates inflammatory transcription factors in primary human monocytes. Mol Cell Biol Res Commun 4:50–61. 24. Janciauskiene S, Moraga F and Lindgren S (2001) C-terminal fragment of alpha1-antitrypsin activates human monocytes to a pro-inflammatory state through interactions with the CD36 scavenger receptor and LDL receptor. Atherosclerosis 158:41–51. 25. Dahl M, Tybjaerg-Hansen A, Sillesen H, Jensen G, Steffensen R and Nordestgaard BG (2003) Blood pressure, risk of ischemic cerebrovascular and ischemic heart disease, and longevity in alpha(1)-antitrypsin deficiency: The Copenhagen City Heart Study. Circulation 107:747–752. 26. Schievink WI, Prakash UB, Piepgras DG and Mokri B (1994) Alpha 1antitrypsin deficiency in intracranial aneurysms and cervical artery dissection. Lancet 343:452–453. 27. Kawano K, Hirashima T, Mori S, Saitoh Y, Kurosumi M and Natori T (1992) Spontaneous long-term hyperglycemic rat with diabetic complications. Otsuka Long-Evans Tokushima Fatty (OLETF) strain. Diabetes 41:1422–1428. 28. Köting N, Berndt J, Kralisch S, Kovacs P, Fasshauer M, Schon MR, Stumvoll M and Blüher M (2006) Vaspin gene expression in human adipose tissue: Association with obesity and type 2 diabetes. Biochem Biophys Res Commun 339:430–436. 29. Zvonic S, Lefevre M, Kilroy G, Floyd ZE, Delany JP, Kheterpal I, Gravois A, Dow R, White A, Wu X and Gimble JM (2006) Secretome of primary cultures of human adipose-derived stem cells (ASCs): Modulation of serpins by adipogenesis. Mol Cell Protemics, in press. 30. Gettins PG (2002) Serpin structure, mechanism, and function. Chem Rev 102:4751–4804.

ch18

Jan. 10, 2007 8:57 SPI-B429



ch19

Jan. 10, 2007

8:57

SPI-B429


19 Serpin Interactions with Bacterial Peptidases Jan Potempa, Tomasz Kantyka and Daniel Nelson

1. Introduction The high-molecular weight serine protease inhibitors (serpins) are present in all domains of life, including a subset of prokaryotes. They originally emerged ∼1000 million years ago apparently in Protozoan organisms and then flourished in Metazoans.1 Thus, serpins are ancient molecules and may have arisen near the beginning of cellular life when bacterial and eukaryotic species were entangled in a primordial milieu. Some prokaryotic species lived in symbiosis with eukaryotes, others constituted prey for Protozoans/Metazoans, or conversely adopted a parasitic life-style at the expense of eukaryotic organisms. There is a little doubt that prokaryotes released proteolytic enzymes to digest available environmental proteins in order to generate a pool of free amino acids or small peptides, which served as nutrients. It is also quite likely that some proteases were used as virulence factors by bacterial pathogens. In this context, it may be expected that a subset of serpins would evolve to inhibit bacterial proteases, thus becoming part of the innate immune system. Indeed, serpins seem to be perfectly fit for this purpose since inhibitory specificity of a given serpin can be readily changed by grafting a reactive site loop from alternate exons onto a single serpin scaffold.2–6 Furthermore, the serpin inhibitory spectrum can be broadened by using distinct amino acid residues within reactive site loop (RSL) for the inhibition of proteases with different specificities.7–10 Nevertheless, despite the fact that several serpins do possess the ability to 425

ch19

Jan. 10, 2007

8:57

SPI-B429

426


J. Potempa et al.

inhibit bacteria-derived proteases, there is no evidence that this strategy was employed to generate inhibitors specific for prokaryotic proteases. On the contrary, the literature is full of reports showing that serpins are target for efficient proteolytic inactivation by bacterial proteases. Does this mean that bacterial proteases have circumvented the protective properties provided by serpins against unwanted proteolytic activity? Our contention is that this is not necessarily the case. The available data are biased by the investigation of the interactions of predominantly human plasma serpins with bacterial pathogen-derived proteases. However, in this narrow spectrum of investigation, bacterial proteases do appear to win out over the serpins. Nonetheless, we do not know if any of the other 34 serpins identified in the human genome have evolved to control bacterial proteases. The same can be speculated regarding serpins from other mammalian species. Furthermore, newly emerging data suggests that serpins may function in invertebrate innate immunity as direct inhibitors of pathogen-derived proteases.11 In the following sections of this chapter, we describe the inactivation of serpins by pathogen-derived proteases, discuss potential consequences of such activity, and characterize inhibitory interactions of serpins with bacterial proteases.

2. Serpins as a Target for Proteolytic Inactivation through Cleavage of Peptide Bond within the Reactive Site Loop Serpins are responsible for inhibiting endogenous host proteases, and as such, regulate many cascade systems including fibrinolysis, coagulation and complement (Table 1). While the serpin/protease system is usually tightly regulated, any perturbance in this delicate balance, favoring either inhibitor or protease, can lead to disease states or serpinopathies. For example, pulmonary emphysema,12 Alzheimer’s disease,13,14 cystic fibrosis,15,16 adult respiratory distress syndrome,17 and sepsis18,19 have all been associated with unregulated proteolysis by host proteases. The nature of the imbalance may arise from a genetic mutation or deficiency, or it may be the results of inactivation of the inhibitor through a chemical reaction or proteolytic cleavage by non-target proteases. Perhaps one of the most studied diseases involving a serpin/protease imbalance is that of pulmonary

ch19

Jan. 10, 2007

8:57

SPI-B429


Serpin Interactions with Bacterial Peptidases Table 1

427

Serpins and their target proteinases.

Serpin (abbreviation)

Target proteinase

Physiological function of serpin

α1 -Protease inhibitor (α1 -PI)

Human neutrophil elastase (HNE)

Regulates elastase levels at sights of inflammation

α1 -Antichymotrypsin (α1 -Achy)

Cathepsin G, chymase

Regulates Cat G and chymase levels at sights of inflammation

α2 -Antiplasmin (α2 -AP)

Plasmin

Regulates fibrinolysis

Antithrombin III (AT III)

Thrombin

Regulates coagulation

Plasminogen activator inhibitor (PAI-1 and PAI-2)

Urokinase (u-PA) and tissue plasminogen activator (t-PA)

Regulates fibrinolysis

C1-inhibitor (C1-Inh)

C1 esterase

Regulates complement system and kallikrein-kinin cascade

emphysema, which is characterized by the destruction of elastin in the alveoli of the lungs as a result of an overabundance of human neutrophil elastase (HNE) relative to its functional inhibitor, α1 -protease inhibitor (α1 -PI). A genetic point mutation in a critical glutamic acid residue that stabilizes a salt bridge in α1 -PI causes a polymerization of this molecule and an inability to be secreted by hepatocytes. The ensuing deficiency of α1 -PI leads to a familial or inherited tendency toward the development of pulmonary emphysema (reviewed in Ref. 20). In a separate mechanism, it has been shown that the oxidation of the methionine residue within the “bait region” of α1 -PI to a methionine sulfoxide residue renders this inhibitor ineffective in regulating HNE.21 Therefore, chronic, heavy smoking can lead to deficiency states of α1 -PI, subsequently resulting in pulmonary emphysema in a similar manner as a patient who is genetically deficient. Finally, due to the extended conformation of the RSL common to all serpins, this domain is susceptible to proteolytic inactivation by non-host proteinases.22 It is well known that pathogens secrete various proteolytic enzymes. Once thought to be utilized only for the acquisition of nutrients for the bacteria, it is now commonly accepted that most bacterial proteases have

ch19

Jan. 10, 2007 8:57 SPI-B429

428


J. Potempa et al. Fig. 1. Cleavage of plasma serpins by pathogen-derived proteases. RSLs of four plasma serpins, α1 -PI, α1 -Achy, ATIII, and C1 Inh. Arrows indicate cleavage sites of specified pathogen-derived proteinases. Bold residues signify the “bait region” of the RSL for its endogenous proteinases. *Gingipain K cleavage site is unpublished observation of J. Potempa.

ch19

Jan. 10, 2007

8:57

SPI-B429


Serpin Interactions with Bacterial Peptidases

429

evolved to specifically modulate host immunity factors. Many bacterial proteases are not only resistant to inhibition by human serpins, but growing evidence indicates that several plasma serpins are cleaved and inactivated by bacterial proteases. Consequently, the serpin/host-protease ratio may be thrown out of balance, resulting in either dysregulation of coagulation or fibrinolytic systems, or generalized unregulated proteolysis from neutrophil derived HNE or cathepsin G (CatG). For instance, Pseudomonas aeruginosa enzymes inactivate α1 -PI, α1 -antichymotrypsin (α1 -Achy) and C1 Inhibitor (C1-Inh).23–26 Thermolysin and subtilysin both inactivate α1 -PI and α1 -Achy.27–29 Figure 1 gives most of the known cleavage sites within the RSL by bacterial proteases. However, this is only a partial list of proteases because cleavage sites are not routinely mapped out when reporting the activity of bacterial proteases on a serpin. For example, the metalloproteinase from Serratia marcescens inactivates α1 -PI, α1 -Achy, C1-Inh, α2 -antiplasmin (α2 -AP) and antithrombin III (ATIII); however, the cleavage sites are only mapped out for α1 -PI and α1 -Achy.25,30,31 It is interesting to note in Fig. 1 that several cleavage sites do conform to expected specificity of the protease; the V-8 protease from Staphylococcus aureus cleaves after Glu in the α1 -PI RSL32 and gingipain K from Porphyromonas gingivalis cleaves after Lys in the α1 -Achy RSL. In contrast, other proteases do not adhere to expected specificity. The only known substrate for periodontain, a protease from P. gingivalis, is the RSL of either α1 -PI or α1 -Achy. While this enzyme readily cleaves after aliphatic (Phe) and acidic (Glu) residues in the α1 -PI RSL, it does not cleave these residues in any other protein or synthetic substrate.33 Thus, it is suggested that the stressed conformation of the RSL may contribute to its susceptibility of cleavage.

3. Pathological Consequences of Serpin Inactivation by the C10 Family of Pathogen-Derived Proteases The CA clan of cysteine proteases is based on the archetypical protease, papain, and all members have the catalytic residues in the sequential order of Cys, His and Asn/Asp. The C10 family of this clan lacks the Asn/Asp of the catalytic triad and it is believed that a carbonyl group of a Trp may serve this role.34 The C10 family only contains two members that have

ch19

Jan. 10, 2007

8:57

SPI-B429

430


J. Potempa et al.

been characterized, i.e. streptopain, also known as SpeB from Streptococcus pyogenes and periodontain, from P. gingivalis.35 A third protease, PrtT, is a homologue of periodontain from P. gingivalis, but its activity has not been characterized. As presented below, a major physiological function of these proteases is thought to be the alteration of the serpin/protease balance. The anaerobe P. gingivalis is a causative organism of adult onset periodontal disease, which is characterized by edema and massive proteolytic degradation of connective tissues in the gingival pockets. The interactions of P. gingivalis with the host innate immune response have been paradoxical, with both pro- and anti-inflammatory responses reported. For example, the soluble P. gingivalis proteases gingipain-K (Kgp) and gingipain-R (Rgp) can stimulate the chemokine IL-836 as well as increase neutrophil chemotaxis by releasing C5a from C5 of the complement system.37 However, membrane bound versions of these same proteases, with a limited ability to diffuse beyond the plaque surface, completely degrade IL-836 and cleave the C5a receptor from infiltrating neutrophils,38 effectively ablating their localized chemotactic activity. These contradictions may be explained by an apparent compartmentalization at periodontal sites, whereby distal activation of chemotactic components and proximal paralysis of these same factors creates a “leukocyte wall” between the periodontal plaque and gingival epithelium.39 Activated neutrophils in the leukocyte wall undergo degranulation due to the inability to phagocytose foreign organisms, thereby expelling large quantities of HNE and CatG. While these proteases can cause abnormal connective tissue destruction, the presence of the serpins α1 -PI and α1 -Achy should minimize this process as they would complex with HNE and CatG, respectively. However, P. gingivalis secretes a potent serpinase, periodontain, that rapidly inactivates α1 -PI33 and α1 Achy (unpublished observation). At 1:1000 E:S ratios, complete conversion to cleaved, inactive α1 -PI is achieved in less than 60 min. Indeed, high levels of HNE are detected in gingival pockets of periodontal patients40,41 and the majority of α1 -PI present was shown to be inactivated.42 Thus, the leukocyte wall effectively becomes a digestive organ, providing nutrition to the bacteria at the expense of surrounding tissues. Streptococcus pyogenes is the primary cause of bacterial pharyngitis and impetigo, but invasive strains of this organism can cause streptococcal toxic shock syndrome and necrotizing fasciitis (NF), which is

ch19

Jan. 10, 2007

8:57

SPI-B429



431

characterized by the proteolytic liquefaction of the extracellular matrix of connective and soft tissues. Like P. gingivalis, S. pyogenes contains several pro- and anti-inflammatory mechanisms to modulate the immune system. Most S. pyogenes strains possess a C5a peptidase that blocks chemotaxis of leukocytes to the infection.43 Further more, in NF strains but not other S. pyogenes strains, a trypsin-like protease is secreted that degrades chemotactic IL-8.44 Indeed, histological staining of tissue sections taken from an NF patient reveals a leukocyte wall distal to the focus of infection, but no leukocytes are in contact with the bacteria. The dissemination of NF-causing S. pyogenes and the associated connective tissue destruction is believed to be due to the actions of host-derived proteases. Streptopain (SpeB), which is chromosomally encoded on all S. pyogenes isolates, contributes to the proteolysis through the cleavage of α1 -PI45 as well as α1 -Achy and α1 -AP (unpublished observation). Although only two members of the C10 protease family have been studied, both function to alter the serpin/host protease balance in favor of unregulated proteolysis beneficial to the microorganism. Additional members of the C10 family will have to be identified and studied to determine if this is a coincidence or a common theme in the family.

4. The Omptins Manipulate the Plasmin/α2 -antiplasmin System The omptins are a class of enterobacterial surface proteases that lack classical serine protease consensus sequences.46 Members include OmpT and OmpP of Escherichia coli, Pla of Yersinia pestis, PgtE of Salmonella typhimurium and SopA of Shigella flexneri.47–51 This class is characterized by antiparallel β-barrel transmembrane topology with five surface exposed loops.52,53 OmpT is the most studied enzyme in this class; although it is not associated with invasive infections, it is believed to be a housekeeping protein that functions to degrade denatured proteins.54 In contrast to OmpT, Pla and PgtE have been shown to aid the invasiveness of Y. pestis and S. typhimurium, respectively, by manipulating the serpin/protease ratio. Both enzymes not only convert plasminogen to active plasmin, but they also proteolytically inactivate α2 -AP, the endogenous inhibitor of plasmin, further exacerbating the imbalance of inhibitor to protease.53,55

ch19

Jan. 10, 2007

8:57

SPI-B429

432


J. Potempa et al.

Uncontrolled plasmin activity generated by PgtE is enhanced in bacteriainfected macrophages and thought to be a mechanism for the dissemination of Salmonella within the phagocytes through the degradation of extracellular matrix proteins such as laminin. Kukkonen et al. showed that residues in the exposed loops of L3 and L4 of Pla are responsible for the substrate specificity of Pla for plasminogen and α2 -AP.53 Significantly, a series of Pla-OmpT chimeras or substitution mutants to the L3 and L4 loops of OmpT resulted in an OmpT mutant that is able to activate plasminogen. Thus, this is an example of how an otherwise “housekeeping” protein may evolve into a virulence factor by subtle mutations at critical residues in an exposed loop, without altering the overall architecture of the protein.

5. Exploitation of C1 Inhibitor by StcE, a Metalloprotease Secreted by Escherichia coli O157:H7 Pathogenic microorganisms are masters at exploiting host defenses for their own benefits. Therefore, it is interesting that with the exception of C1-Inh abuse by Escherichia coli, there are no other examples of manipulating serpin activity by bacteria. E. coli O157:H7 (EHEC), a pathogen responsible for diarrhea, hemorrhagic colitis and the hemolytic uremic syndrome, expresses a unique metalloprotease with proteolytic activity limited to C1-Inh. This enzyme is referred to as secreted protease of C1-esterase inhibitor from EHEC (StcE) and it cleaves C1-Inh to produce a 60–65 kDa fragment.56 This cleavage, located in the heavily glycosylated N-terminal domain of C1-Inh, does not affect inhibitory activity of the serpin. However, the cleaved CI-Inh binds erythrocytes and immobilize C1-Inh on the surface. This increases the local concentration of inhibitor at the sites of potential lytic complex formation, protecting both E. coli O157:H7 and host cells, to which the bacterium adheres, from complement-mediated lysis.57

6. Inhibition of Bacterial Serine Peptidases by Serpins Serpins were described originally as inhibitors of serine proteases belonging to the chymotrypsin family (family S1). Recent advances in genome

ch19

Jan. 10, 2007

8:57

SPI-B429



433

sequencing and gene cloning have revealed that chymotrypsin-like proteases are of a very ancient evolutionary design and are widespread in three domains of life. Except for 15 bacterial species, which includes nine different Mycoplasma spp, Ureaplasma urealyticum, Burkholderia thailandensis, Francisella tularensis, Magnetospirillum magneticum, Mesoplasma florum, and Synechococcus sp. CC9902, genes coding proteases belonging to family S1 are present in all other genomes sequenced to date. These proteases play divergent roles in bacterial physiology from house-keeping functions and nutrient acquisition to virulence factors. Nevertheless, to our best knowledge, there is not a single report documenting inhibition of a bacterial chymotrypsin-like protease by a serpin inhibitor. The second largest group of serine proteases common in prokaryotes is constituted by subtilisin-like enzymes. Genes encoding these proteases are present in 145 out of 249 bacterial genomes sequenced to date. Although subtilisins are phylogenetically and structurally unrelated to chymotrypsinlike proteases, these two group of enzymes share very similar geometry among the catalytic Ser, His and Asp residues. Subtilisin-like proteases are also common in eukaryotes and in mammals where they play a crucial role in a variety of physiological processes; they are also involved in the pathology of diseases, including cancer, viral infection, and Alzheimer’s diseases. They function as proprotein convertases, converting a precursor to active protein by cleavage at basic motifs (usually Arg/Lys-Arg or Arg-Xaa-XaaArg, where Xaa can be any amino acid) and a list of activated substrates include neuropeptides, peptide hormones, growth and differentiation factors, receptors, enzymes, adhesion molecules, blood coagulation factors, plasma proteins, viral coat proteins, and bacterial toxins.58 Therefore, it is not surprising that serpins with the ability to stringently control proprotein convertases have evolved in metazoans.59–61 However, it is intriguing that several serpins, likewise, inhibit bacterial subtilisins (Table 2). In all cases, the amount of serpin required to inhibit the subtilisin is much higher than that required to inhibit the same amount of a target, endogenous protease. This large partition ratio (stoichiometry of inhibition), where 5–50 molecules of a serpin are required to inhibit one subtilisin molecule, is caused by more serpin molecules acting as substrates than inhibitors. In this way, several molecules of each serpin are consumed to inhibit one molecule of the protease. However, when inhibition occurs, the rate of

ch19

Jan. 10, 2007

8:57

SPI-B429


434

J. Potempa et al. Table 2

Kinetic parameters of subtilisins inhibition by human serpins.

Serpin

Subtilisin

SIa

kass [M/s]

Ki [pM]

Ref.

SERPINA1 (α1 -PI) SERPINA1 (α1 -PI) SERPINA2 (α1 -Achy) SERPINB9 (PI9) SERPINB8 (PI8)

Carlsberg Proteinase K Proteinase K Subtilisin A Subtilisin A

5.2 8.5 9.3 10 50

1.2 × 105 1.4 × 105 1.1 × 105 2.4 × 106 1.16 × 106

n.d. n.d. n.d. 3.6 8.4

27 27 27 62 63

aStoichiometry of inhibition or the partition ratio,64 the ratio of the inhibitor to the protease at the point of zero residual activity. n.d., not determined.

formation (kass ) and the stability of the complexes (Ki ) are similar to those formed between serpins and chymotrypsin family members or PI8 and furin, a proprotein convertase.59 The formation of stable inhibitory serpin–subtilisin complexes is especially noteworthy in light of the lack of inhibition of bacterial chymotrypsin-like proteases by serpins. Intuitively, one would expect that since mammalian chymotrypsin-like proteases are the major target for serpins, at least some of their prokaryotic counterparts would be inhibited by serpins. Thus far, no such interactions have been described and they either do not occur at all in nature, or we simply have not yet found a reactive pair of protease and serpin. Clearly, the interactions described to date between serpins and bacterial subtilisins have no physiological importance. However, taking into account that some subtilisin-like proteases, including C5a peptidase,65 PrtA protease66 and trepolisin (dentilisin),67 evolved as important virulence factors of major human pathogens, such as Streptococcus pyogenes, Streptococcus pneumoniae, and Treponema denticola, respectively, it would be interesting to investigate if any of the human serpins is able to control the activity of these proteases.

7. Inhibition of Bacterial Cysteine Peptidases by Serpins Several serpins exhibit cross-class specificity and inhibit cysteine proteases.7,9,68–79 Except viral serpins, all other human cross-class inhibitors characterized belong to an evolutionally related family of

ch19

Jan. 10, 2007

8:57

SPI-B429



435

ovserpins (Clade B serpins), which unlike circulating serpins, lack signal peptides and reside primarily within cells.80 Interestingly, the cross-class serpins can efficiently inhibit cysteine proteases belonging to two phylogenetically and structurally unrelated superfamilies of proteases (clans CA and CD of cysteine peptidases). The cowpox virus serpin CrmA is an effective inhibitor of serine protease granzyme B81 and caspases 1 and 8, clan CD proteases,82 apparently through their ability to readily accept an aspartic acid side chain of the P1 residue of the inhibitor in their primary substrate binding pockets (S1s). This may explain why Lys- and Arg-specific gingipains (Kgp and Rgp, respectively), cysteine proteases produced by P. gingivalis, a major human periodontopathogen,83 although structurally related to caspases84 are not inhibited by CrmA. However, the simple Asp → Lys P1 mutation rendered CrmA a good Kgp inhibitor (kass = 2.1 × 105 M−1 s−1 , Ki < 0.2 nM).85 This confirms the serpin family’s extraordinary adaptability for inhibiting diverse proteases but which, has no physiological meaning. Conversely, gingipain inhibition by (ATIII)86 may contribute to controlling the activity of this virulence factor in the development of periodontitis. It has been shown that in the absence of heparin, ATIII inhibits Rgp activity with kass = 5.65 × 104 M−1 s−1 , without formation of the SDS-stable complex. The heparin presence enhanced the kass and stabilized the complex, but in either case, inhibition was temporary and resulted in a reactive site cleaved (Arg393 –Ser394 ), an inactive inhibitor.87 To date, no inhibition of bacteria-derived papain-like cysteine protease (Clan CA) by a serpin has been reported. This is intriguing since papainlike cysteine proteases, while highly underrepresented in prokaryotes, are recognized as important virulence factors. These include streptopain (SpeB) of S. pyogenes, staphopains of S. aureus, and members of YopT peptidase family produced by Yersinia spp. and the plant pathogen Pseudomonas syringae.88 The other important group of cysteine proteases, structurally unrelated to papain, but strongly implicated in the pathogenicity of humans and plants caused by Y. pestis, Xantomonas campestris, respectively, are YopJ and XopD peptidases. These enzymes can be found inside eukaryotic cells of infected hosts, either being directly injected into the cytoplasm (YopT, YopJ, XopD) or may diffuse through the pores punched in the cell membrane by bacterial toxin (SpeB), or can be expressed intracellularly

ch19

Jan. 10, 2007

8:57

SPI-B429


436

J. Potempa et al.

during the cell invasion by the pathogen (SpeB and staphopains). Both in the extracellular and especially in the intracellular environments, these proteases should meet with ov-serpins, which are remarkable in their ability to inhibit cysteine proteases in comparison to other serpins. Therefore, it is our contention that in the near future, inhibitory reactions between bacterial cysteine proteases and serpins will be described.

8. Serpin Direct Functions in the Innate Immunity The role of serpins in the regulation of the innate immunity in invertebrates is known in detail.89–92 In this context, it is perplexing that virtually nothing is known with regard to direct inhibition of bacterial pathogen-derived proteases, although such a function is recognized.93 As mentioned in the introduction, serpin scaffolds provide an opportunity to generate a very broad spectrum of inhibitors with differing specificity. This is best documented in the case of a Spn4 gene of Drosophila melanogaster, where distinct RSL variants can inhibit serine proteases of subtilisin (S8), family S1, and cysteine proteases of papain family (C1).11 However, no real targets for these inhibitors have been identified so far. The lack of targets is due to the reliance of testing invertebrate serpins and only mammalian proteases. It can however be speculated that once the experimental set-up is changed and the bacterial proteases are tested, it would be just a matter of time before physiological target(s) for serpins are identified among pathogen-derived proteases.

9. Conclusions In general, serpin–bacterial protease encounters lead to inhibitor inactivation through limited proteolysis of the RSL. During infection, serpin inactivation may lead to local depletion of antiproteolytic activity and uncontrolled proteolytic degradation of host tissues or the deregulation of otherwise tightly controlled proteolytic cascades. This is apparently the downside for the mechanism of inhibition employed by serpins, which is based on a metastable structure and exposed RSL. On the other hand, the inhibition of several bacteria-derived proteases such as subtilisins and gingipains indicates that the serpin design allows for extraordinary adaptability at

ch19

Jan. 10, 2007

8:57

SPI-B429



437

inhibiting diverse proteases. This remarkable feature of serpins is employed widely by invertebrates to generate large numbers of inhibitors with varying specificity. Therefore, it can be predicted that serpins are also used as direct inhibitors of pathogen-derived proteases.

References 1. Cole EB, Miller D, Rometo D, Greenberg RM, Bromme D, Cataltepe S, Pak SC, Mills DR, Silverman GA and Luke CJ (2004) Identification and activity of a lower eukaryotic serine proteinase inhibitor (serpin) from Cyanea capillata: Analysis of a jellyfish serpin, jellypin. Biochemistry 43:11750–11759. 2. Jiang H, Wang Y and Kanost MR (1994) Mutually exclusive exon use and reactive center diversity in insect serpins. J Biol Chem 269:55–58. 3. Jiang H and Kanost MR (1997) Characterization and functional analysis of 12 naturally occurring reactive site variants of serpin-1 from Manduca sexta. J Biol Chem 272:1082–1087. 4. Jiang H, WangY, HuangY, Mulnix AB, Kadel J, Cole K and Kanost MR (1996) Organization of serpin gene-1 from Manduca sexta. Evolution of a family of alternate exons encoding the reactive site loop. J Biol Chem 271:28017–28023. 5. Kruger O, Ladewig J, Koster K and Ragg H (2002) Widespread occurrence of serpin genes with multiple reactive centre-containing exon cassettes in insects and nematodes. Gene 293:97–105. 6. Danielli A, Kafatos FC and Loukeris TG (2003) Cloning and characterization of four Anopheles gambiae serpin isoforms, differentially induced in the midgut by Plasmodium berghei invasion. J Biol Chem 278:4184–4193. 7. Schick C, Bromme D, Bartuski AJ, Uemura Y, Schechter NM and Silverman GA (1998) The reactive site loop of the serpin SCCA1 is essential for cysteine proteinase inhibition. Proc Natl Acad Sci USA 95:13465–13470. 8. Dahlen JR, Foster DC and Kisiel W (1998) The inhibitory specificity of human proteinase inhibitor 8 is expanded through the use of multiple reactive site residues. Biochem Biophys Res Commun 244:172–177. 9. Luke C, Schick C, Tsu C, Whisstock JC, Irving JA, Bromme D, Juliano L, Shi GP, Chapman HA and Silverman GA (2000) Simple modifications of the serpin reactive site loop convert SCCA2 into a cysteine proteinase inhibitor: A critical role for the P3 proline in facilitating RSL cleavage. Biochemistry 39:7081–7091. 10. Potempa J, Shieh BH and Travis J (1988) Alpha-2-antiplasmin: A serpin with two separate but overlapping reactive sites. Science 241:699–700.

ch19

Jan. 10, 2007

8:57

SPI-B429

438


J. Potempa et al.

11. Oley M, Lummer M, Smolenaars MM, Rodenburg KW and Ragg H, The SPN4 gene from Drosophila melanogaster is a multipurpose defense tool directed against proteases from three different peptidase families. J Biol Chem, in press. 12. Eriksson S (1964) Pulmonary emphysema and alpha-1-antitrypsin deficiency. Acta Med Scand 175:197–205. 13. Abraham CR, Selkoe DJ and Potter H (1998) Immunochemical identification of the serine protease inhibitor alpha-1-antichymotrypsin in the brain amyloid deposits of Alzheimer’s disease. Cell 52:487–501. 14. Smirnova IV, Ho GJ, Fenton JWII and Festoff BW (1994) Extravascular proteolysis and the nervous system: Serine protease/serpin balance. Sem Thromb Hemostasis 20:426–432. 15. Delacourt C, LeBourgeois M, D’Ortho MP, Doit C, Scheinmann P, Navarro J, Harf A, Hartmann DJ and Lafuma C (1995) Imbalance between 95 kDa type IV collagenase and tissue inhibitor of metalloproteinases in sputum of patients with cystic fibrosis. Am J Respir Crit Care Med 152:765–774. 16. Birrer P (1995) Proteases and antiproteases in cystic fibrosis: Pathogenetic considerations and therapeutic strategies. Respiration 62:25–28. 17. Bertozzi P, Astedt B, Zenzius L, Lynch K, LeMaire F, Zapol W and Chapman HA Jr (1990) Depressed bronchoalveolar urokinase activity in patients with adult respiratory distress syndrome. New Engl J Med 322:890–897. 18. Nuijens JH, Eerenberg-Belmer AJM, Huijbregts CCM, Schreuder WO, FeltBersma RJF, Abbink JJ, Thijs LG and Hack CE (1998) Proteolytic inactivation of plasma C1 inhibitor in sepsis. J Clin Invest 84:443–450. 19. Nuijens JH, Abbink JJ, Wachtfogel YT, Colman RW, Eerenberg JM, Dors D, Kamp AJM, Strack van Schijndel RJM, Thijs LG and Hack CE (1992) Plasma elastase alpha-1 antitrypsin and lactoferrin in sepsis: Evidence for neutrophils as mediators in fatal sepsis. J Lab Clin Med 119:159–168. 20. Salvesen G and Travis J (1989) Properties of naturally occurring elastase inhibitors, in Robert L and Hornebeck W (eds.) Elastin and Elastases, 2 CRC Press, Boca Raton, Inc. pp. 95–108. 21. Janoff A and Carp H (1979) Possible mechanisms of emphysema in smokers. Am Rev Respir Dis 116:65–69. 22. Nelson D, Potempa J and Travis J (1998) Inactivation of a-1 proteinase inhibitor as a broad screen for detecting proteolytic activities in unknown samples. Anal Biochem 260:230–236. 23. Kress LF (1986) Inactivation of human plasma serine proteinase inhibitors (serpins) by limited proteolysis of the reactive site loop with snake venom and bacterial metalloproteinases. J Cell Biochem 32:51–58.

ch19

Jan. 10, 2007

8:57

SPI-B429



439

24. Morihara K, Tsuzuki H, Harada M and Iwata T (1984) Purification of human plasma alpha 1-proteinase inhibitor and its inactivation by Pseudomonas aeruginosa elastase. J Biochem (Tokyo) 95:795–804. 25. Potempa J, Fedak D, Dubin A, Mast A and Travis J (1991) Proteolytic inactivation of alpha-1-anti-chymotrypsin. Sites of cleavage and generation of chemotactic activity. J Biol Chem 266:21482–21487. 26. Pemberton PA, Harrison RA, Lachmann PJ and Carrell RW (1989) The structural basis for neutrophil inactivation of C1 inhibitor. Biochem J 258:193–198. 27. Komiyama T, Gron H, Pemberton PA and Salvesen GS (1996) Interaction of subtilisins with serpins. Protein Sci 5:874–882. 28. Carrell RW, Evans DL and Stein PE (1991) Mobile reactive centre of serpins and the control of thrombosis. Nature 353:576–578. 29. Chang WS, Wardell MR, Lomas DA and Carrell RW (1996) Probing serpin reactive-loop conformations by proteolytic cleavage. Biochem J 314 (Pt 2):647–653. 30. Molla A, Akaike T and Maeda H (1989) Inactivation of various proteinase inhibitors and the complement system in human plasma by the 56-kilodalton proteinase from Serratia marcescens. Infect Immun 57:1868–1871. 31. Virca GD, Lyerly D, KregerA and Travis J (1982) Inactivation of human plasma alpha 1-proteinase inhibitor by a metalloproteinase from Serratia marcescens. Biochim Biophys Acta 704:267–271. 32. Potempa J, Watorek W and Travis J (1986) The inactivation of human plasma alpha 1-proteinase inhibitor by proteinases from Staphylococcus aureus. J Biol Chem 261:14330–14334. 33. Nelson D, Potempa J, Kordula T and Travis J (1999) Purification and characterization of a novel cysteine proteinase (periodontain) from Porphyromonas gingivalis. Evidence for a role in the inactivation of human alpha1-proteinase inhibitor. J Biol Chem 274:12245–12251. 34. Kagawa TF, Cooney JC, Baker HM, McSweeney S, Liu M, Gubba S, Musser JM and Baker EN (2000) Crystal structure of the zymogen form of the group A Streptococcus virulence factor SpeB: an integrin-binding cysteine protease. Proc Natl Acad Sci USA 97:2235–2240. 35. Barrett AJ, Rawlings ND and Woessner JF (eds.) (1998) Handbook of Proteolytic Enzymes, Academic Press: London. 36. Mikolajczyk-Pawlinska J, Travis J and Potempa J (1998) Modulation of interleukin-8 activity by gingipains from Porphyromonas gingivalis: Implications for pathogenicity of periodontal disease. FEBS Lett 440:282–286. 37. Wingrove JA, DiScipio RG, Chen Z, Potempa J, Travis J and Hugli TE (1992) Activation of complement components C3 and C5 by a cysteine proteinase

ch19

Jan. 10, 2007

8:57

SPI-B429

440

38.

39. 40.

41. 42.

43.

44.

45.

46.

47.

48. 49. 50.


J. Potempa et al.

(gingipain-1) from Porphyromonas (Bacteroides) gingivalis. J Biol Chem 267:18902–18907. Jagels MA, Travis J, Potempa J, Pike R and Hugli TE (1996) Proteolytic inactivation of the leukocyte C5a receptor by proteinases derived from Porphyromonas gingivalis. Infect Immun 64:1984–1991. Miyasaki KT (1991) The neutrophil: Mechanisms of controlling periodontal bacteria. J Periodontol 62:761–774. Cox SW and Eley BM (1992) Cathepsin B/L-, elastase-, tryptase-, trypsinand dipeptidyl peptidase IV-like activities in gingival crevidular fluid: A comparison of levels before and after basic periodontal treatment of chronic periodontitis patients. J Clin Periodontol 19:333–339. Uitto VJ, Nieminen A, Coil J, Hurttia H and Larjava H (1996) Oral fluid elastase as an indicator of periodontal health. J Clin Periodontol 23:30–37. Smith QT, WangY-D and Sim B (1994) Inhibition of crevicular fluid neutrophil elastase by a-1-antitrypsin in periodontal health and disease. Archs Oral Biol 39:301–306. Ji Y, McLandsborough L, Kondagunta A and Cleary PP (1996) C5a peptidase alters clearance and trafficking of group A streptococci by infected mice. Infect Immun 64:503–510. Hidalgo-Grass C, Dan-Goor M, Maly A, Eran Y, Kwinn LA, Nizet V, Ravins M, Jaffe J, Peyser A, Moses AE and Hanski E (2004) Effect of a bacterial pheromone peptide on host chemokine degradation in group A streptococcal necrotising soft-tissue infections. Lancet 363:696–703. Rapala-Kozik M, Potempa J, Nelson D, Kozik A and Travis J (1999) Comparative cleavage sites within the reactive-site loop of native and oxidized alpha1-proteinase inhibitor by selected bacterial proteinases. Biol Chem 380:1211–1216. Mangel WF, Toledo DL, Brown MT, Worzalla K, Lee M and Dunn JJ (1994) Omptin: An Escherichia coli outer membrane proteinase that activates plasminogen. Methods Enzymol 244:384–399. Grodberg J, Lundrigan MD, Toledo DL, Mangel WF and Dunn JJ (1988) Complete nucleotide sequence and deduced amino acid sequence of the ompT gene of Escherichia coli K-12. Nucleic Acids Res 16:1209. Kaufmann A, Stierhof YD and Henning U (1994) New outer membraneassociated protease of Escherichia coli K-12. J Bacteriol 176:359–367. Sodeinde OA and Goguen JD (1988) Genetic analysis of the 95-kilobase virulence plasmid of Yersinia pestis. Infect Immun 56:2743–2748. Guina T, Yi EC, Wang H, Hackett M and Miller SI (2000) A PhoP-regulated outer membrane protease of Salmonella enterica serovar typhimurium

ch19

Jan. 10, 2007

8:57

SPI-B429



51.

52.

53.

54.

55.

56.

57.

58. 59.

60.

61.

62.

441

promotes resistance to alpha-helical antimicrobial peptides. J Bacteriol 182:4077–4086. Egile C, d’Hauteville H, Parsot C and Sansonetti PJ (1997) SopA, the outer membrane protease responsible for polar localization of IcsA in Shigella flexneri. Mol Microbiol 23:1063–1073. Kramer RA, Zandwijken D, Egmond MR and Dekker N (2000) In vitro folding, purification and characterization of Escherichia coli outer membrane protease ompT. Eur J Biochem 267:885–893. Kukkonen M, Lahteenmaki K, Suomalainen M, Kalkkinen N, Emody L, Lang H and Korhonen TK (2001) Protein regions important for plasminogen activation and inactivation of alpha2-antiplasmin in the surface protease Pla of Yersinia pestis. Mol Microbiol 40:1097–1111. White CB, Chen Q, Kenyon GL and Babbitt PC (1995) A novel activity of OmpT proteolysis under extreme denaturing conditions. J Biol Chem 270:12990–12994. Lahteenmaki K, Kyllonen P, Partanen L and Korhonen TK (2005) Antiprotease inactivation by Salmonella enterica released from infected macrophages. Cell Microbiol 7:529–538. Lathem WW, Grys TE, Witowski SE, Torres AG, Kaper JB, Tarr PI and Welch RA (2002) StcE, a metalloprotease secreted by Escherichia coli O157:H7, specifically cleaves C1 esterase inhibitor. Mol Microbiol 45: 277–288. Lathem WW, Bergsbaken T and Welch RA (2004) Potentiation of C1 esterase inhibitor by StcE, a metalloprotease secreted by Escherichia coli O157:H7. J Exp Med 199:1077–1087. Taylor NA, Van De Ven WJ and Creemers JW (2003) Curbing activation: Proprotein convertases in homeostasis and pathology. Faseb J 17:1215–1227. Dahlen JR, Jean F, Thomas G, Foster DC and Kisiel W (1998) Inhibition of soluble recombinant furin by human proteinase inhibitor 8. J Biol Chem 273:1851–1854. Bentele C, Kruger O, Todtmann U, Oley M and Ragg H (2006) A proprotein convertase-inhibiting serpin with an ER targeting signal from Branchiostoma lanceolatum, a close relative of vertebrates. Biochem J 395:449–456. Osterwalder T, Kuhnen A, Leiserson WM, Kim YS and Keshishian H (2004) Drosophila serpin 4 functions as a neuroserpin-like inhibitor of subtilisin-like proprotein convertases. J Neurosci 24:5482–5491. Dahlen JR, Foster DC and Kisiel W (1997) Human proteinase inhibitor 9 (PI9) is a potent inhibitor of subtilisin A. Biochem Biophys Res Commun 238:329–333.

ch19

Jan. 10, 2007

8:57

SPI-B429

442


J. Potempa et al.

63. Dahlen JR, Foster DC and Kisiel W (1997) Expression, purification, and inhibitory properties of human proteinase inhibitor 8. Biochemistry 36:14874– 14882. 64. Patston PA, Gettins P, Beechem J and Schapira M (1991) Mechanism of serpin action: Evidence that C1 inhibitor functions as a suicide substrate. Biochemistry 30:8876–8882. 65. Park HS and Cleary PP (2005) Active and passive intranasal immunizations with streptococcal surface protein C5a peptidase prevent infection of murine nasal mucosa-associated lymphoid tissue, a functional homologue of human tonsils. Infect Immun 73:7878–7886. 66. Bethe G, Nau R, Wellmer A, Hakenbeck R, Reinert RR, Heinz HP and Zysk G (2001) The cell wall-associated serine protease PrtA: A highly conserved virulence factor of Streptococcus pneumoniae. FEMS Microbiol Lett 205:99–104. 67. Ellen RP, Ko KS, Lo CM, Grove DA and Ishihara K (2000) Insertional inactivation of the prtP gene of Treponema denticola confirms dentilisin’s disruption of epithelial junctions. J Mol Microbiol Biotechnol 2:581–586. 68. Takeda A, Yamamoto T, Nakamura Y, Takahashi T and Hibino T (1995) Squamous cell carcinoma antigen is a potent inhibitor of cysteine proteinase cathepsin L. FEBS Lett 359:78–80. 69. Annand RR, Dahlen JR, Sprecher CA, De Dreu P, Foster DC, Mankovich JA, Talanian RV, Kisiel W and Giegel DA (1999) Caspase-1 (interleukin1beta-converting enzyme) is inhibited by the human serpin analogue proteinase inhibitor 9. Biochem J 342(Pt 3):655–665. 70. Schick C, Pemberton PA, Shi GP, Kamachi Y, Cataltepe S, Bartuski AJ, Gornstein ER, Bromme D, Chapman HA and Silverman GA (1998) Crossclass inhibition of the cysteine proteinases cathepsins K, L, and S by the serpin squamous cell carcinoma antigen 1: A kinetic analysis. Biochemistry 37:5258–5266. 71. Jayakumar A, Kang Y, Frederick MJ, Pak SC, Henderson Y, Holton PR, Mitsudo K, Silverman GAAKEL-N, Bromme D and Clayman GL (2003) Inhibition of the cysteine proteinases cathepsins K and L by the serpin headpin (SERPINB13): A kinetic analysis. Arch Biochem Biophys 409:367–374. 72. Welss T, Sun J, Irving JA, Blum R, Smith AI, Whisstock JC, Pike RN, von MikeczA, Ruzicka T, Bird PI andAbts HF (2003) Hurpin is a selective inhibitor of lysosomal cathepsin L and protects keratinocytes from ultraviolet-induced apoptosis. Biochemistry 42:7381–7389. 73. Askew DJ, Askew YS, Kato Y, Luke CJ, Pak SC, Bromme D and Silverman GA (2004) The amplified mouse squamous cell carcinoma antigen gene

ch19

Jan. 10, 2007

8:57

SPI-B429



74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

443

locus contains a serpin (Serpinb3b) that inhibits both papain-like cysteine and trypsin-like serine proteinases. Genomics 84:166–175. Sakata Y, Arima K, Takai T, Sakurai W, Masumoto K, Yuyama N, Suminami Y, Kishi F, Yamashita T, Kato T, Ogawa H, Fujimoto K, Matsuo Y, Sugita Y and Izuhara K (2004) The squamous cell carcinoma antigen 2 inhibits the cysteine proteinase activity of a major mite allergen, Der p 1. J Biol Chem 279:5081–5087. Al-Khunaizi M, Luke CJ, Askew YS, Pak SC, Askew DJ, Cataltepe S, Miller D, Mills DR, Tsu C, Bromme D, Irving JA, Whisstock JC and Silverman GA (2002) The serpin SQN-5 is a dual mechanistic-class inhibitor of serine and cysteine proteinases. Biochemistry 41:3189–3199. Hook VY and Hwang SR (2002) Novel secretory vesicle serpins, endopin 1 and endopin 2: endogenous protease inhibitors with distinct target protease specificities. Biol Chem 383:1067–1074. Komiyama T, Ray CA, Pickup DJ, Howard AD, Thornberry NA, Peterson EP and Salvesen G (1994) Inhibition of interleukin-1 beta converting enzyme by the cowpox virus serpin CrmA. An example of cross-class inhibition. J Biol Chem 269:19331–19337. Messud-Petit F, Gelfi J, Delverdier M, Amardeilh MF, Py R, Sutter G and Bertagnoli S (1998) Serp2, an inhibitor of the interleukin-1beta-converting enzyme, is critical in the pathobiology of myxoma virus. J Virol 72:7830–7839. Krieg SA, Krieg AJ and Shapiro DJ (2001) A unique downstream estrogen responsive unit mediates estrogen induction of proteinase inhibitor-9, a cellular inhibitor of IL-1beta- converting enzyme (caspase 1). Mol Endocrinol 15:1971–1982. Silverman GA, Whisstock JC, Askew DJ, Pak SC, Luke CJ, Cataltepe S, Irving JA and Bird PI (2004) Human clade B serpins (ov-serpins) belong to a cohort of evolutionarily dispersed intracellular proteinase inhibitor clades that protect cells from promiscuous proteolysis. Cell Mol Life Sci 61:301–325. Quan LT, Caputo A, Bleackley RC, Pickup DJ and Salvesen GS (1995) Granzyme B is inhibited by the cowpox virus serpin cytokine response modifier A. J Biol Chem 270:10377–10379. Zhou Q, Snipas S, Orth K, Muzio M, Dixit VM and Salvesen GS (1997) Target protease specificity of the viral serpin CrmA. Analysis of five caspases. J Biol Chem 272:7797–7800. Potempa J, Sroka A, Imamura T and Travis J (2003) Gingipains, the major cysteine proteinases and virulence factors of Porphyromonas gingivalis: Structure, function and assembly of multidomain protein complexes. Curr Protein Pept Sci 4:397–407.

ch19

Jan. 10, 2007

8:57

SPI-B429

444


J. Potempa et al.

84. Eichinger A, Beisel HG, Jacob U, Huber R, Medrano FJ, Banbula A, Potempa J, Travis J and Bode W (1999) Crystal structure of gingipain R: An Arg-specific bacterial cysteine proteinase with a caspase-like fold. Embo J 18:5453–5462. 85. Snipas SJ, Stennicke HR, Riedl S, Potempa J, Travis J, Barrett AJ and Salvesen GS (2001) Inhibition of distant caspase homologues by natural caspase inhibitors. Biochem J 357:575–580. 86. Fishburn CS, Slaney JM, Carman RJ and Curtis MA (1991) Degradation of plasma proteins by the trypsin-like enzyme of Porphyromonas gingivalis and inhibition of protease activity by a serine protease inhibitor of human plasma. Oral Microbiol Immunol 6:209–215. 87. Curtis MA, Slaney JM, Carman RJ and Pemberton PA (1993) Interaction of a trypsin-like enzyme of Porphyromonas gingivalis W83 with antithrombin III. FEMS Microbiol Lett 108:169–174. 88. Potempa J and Pike RN (2005) Bacterial peptidases, in Herwald H (ed.) Bacterial Virulence Factors and their Molecular Mechanisms. 12, Karger, Basel, pp. 132–180. 89. Vodovar N, Acosta C, Lemaitre B and Boccard F (2004) Drosophila: A polyvalent model to decipher host-pathogen interactions. Trends Microbiol 12:235–242. 90. Reichhart JM (2005) Tip of another iceberg: Drosophila serpins. Trends Cell Biol 15:659–665. 91. Kanost MR, Jiang H and Yu XQ (2004) Innate immune responses of a lepidopteran insect, Manduca sexta. Immunol Rev 198:97–105. 92. Tong Y, Jiang H and Kanost MR (2005) Identification of plasma proteases inhibited by Manduca sexta serpin-4 and serpin-5 and their association with components of the prophenol oxidase activation pathway. J Biol Chem 280: 14932–14942. 93. Zang X and Maizels RM (2001) Serine proteinase inhibitors from nematodes and the arms race between host and pathogen. Trends Biochem Sci 26:191–197.

ch19

Jan. 10, 2007

8:57

SPI-B429


20 The Serpinopathies and Respiratory Disease David A. Lomas, Didier Belorgey, Elena Miranda, Meera Mallya, Peter Hägglöf, Lynda K. Sharp, Russell L. Phillips, Richard Page, Mark J. Davies and Damian C. Crowther

1. Introduction The serine proteinase inhibitors or serpins are found in most branches of life, including mammals, plants, viruses and prokaryotes where they inhibit a wide range of proteolytic cascades. Members of this family include α1 -antitrypsin, C1 inhibitor, antithrombin and tissue plasminogen activator, which play important roles in the control of proteinases involved in the inflammatory, complement, coagulation, and fibrinolytic pathways, respectively.1 The superfamily is defined by more than 30% sequence homology with the archetypal member α1 -antitrypsin and conservation of tertiary structure. This structure is composed of three β-sheets (A–C) and an exposed mobile reactive loop that presents a peptide sequence as a pseudosubstrate for the target proteinase2−8 (see Chap. 2). After docking, the enzyme cleaves the P1–P1 peptide bond of the serpin9 and the proteinase is inactivated by a dramatic conformational transition that swings it 70 Å from the upper to the lower pole of the protein in association with the insertion of the reactive loop as an extra strand (s4A) in β-sheet A.10−14 This remarkable conformational transition is central to the inhibitory activity of the serpins. However, as with most sophisticated mechanisms, the mobile domains are vulnerable to dysfunction. In the case of the serpins, mutations cause aberrant conformational transitions that result in the retention of the 445

ch20

Jan. 10, 2007

8:57

SPI-B429

446


D. A. Lomas et al.

serpin within the cell of synthesis. This gives rise to clinical conditions that result from either (i) intracellular protein overload and death of the cell, in which the serpin is synthesized, (toxic gain of function) such as Z α1 antitrypsin related cirrhosis and the dementia Familial Encephalopathy with Neuroserpin Inclusion Bodies (FENIB); or (ii) the loss of function such as deficiency of plasma antithrombin, C1-inhibitor, or α1 -antichymotrypsin.15 These cause diseases as diverse as thrombosis, angio-edema and emphysema, respectively. We have shown that there is a common mechanism underlying these conditions and so we have grouped them together as a new class of disease, the serpinopathies.15−17 We review here the molecular mechanisms underlying the serpinopathies and show how they provide new insights into the pathogenesis of respiratory diseases.

2. Gain of Function Serpinopathies: α1 -Antitrypsin Deficiency Related Cirrhosis and FENIB 2.1. PI*Z α1 -antitrypsin deficiency related cirrhosis α1 -Antitrypsin is an acute phase glycoprotein that is synthesized and secreted by the liver. It bathes all the tissues of the body with its primary role being to inhibit the enzyme neutrophil elastase.18,19 Most individuals carry two copies of the normal M allele and this results in plasma concentrations of 1.5–3.5 g/l. α1 -Antitrypsin deficiency was first described as a clinical entity in 1963 by Laurell and Eriksson who noted an absence of the α1 band on serum protein electrophoresis.20 We now recognize that the most important deficiency mutation is the Z allele (Glu342Lys). Approximately 4% of Northern Europeans are heterozygous for the Z allele (PI*MZ) with approximately 1 in 2000 being homozygotes (PI*Z). The Z allele results in the retention of synthesized α1 -antitrypsin within the endoplasmic reticulum of hepatocytes. The accumulation of abnormal protein starts in utero21 and is characterized by diastase-resistant, periodic acid-Schiff (PAS) positive inclusions of α1 -antitrypsin in the periportal cells.22,23 Seventy-three percent of Z α1 -antitrypsin homozygote infants have a raised serum alanine aminotransferase in the first year of life, but only in 15% of the people is there still abnormality by 12 years of age.24−27 Similarly, serum bilirubin is raised in 11% of PI*Z infants in the first 2–4 months, but falls to the normal

ch20

Jan. 10, 2007

8:57

SPI-B429


The Serpinopathies and Respiratory Disease

447

level by 6 months of age. One in 10 infants develops cholestatic jaundice and 6% develop clinical evidence of liver disease without jaundice. These symptoms usually resolve by the second year of life, but approximately 15% of patients with cholestatic jaundice progress to juvenile cirrhosis. The overall risk of death from liver disease in PI*Z children during childhood is 2 to 3%, with boys being at more risk than girls. All adults with the Z allele of α1 -antitrypsin have slowly progressive hepatic damage that is often sub-clinical and only evident as a minor degree of portal fibrosis. However, up to 50% of Z α1 -antitrypsin homozygotes present with clinically evident cirrhosis and occasionally with hepatocellular carcinoma.28 The retention of Z α1 -antitrypsin within hepatocytes explains the associated plasma deficiency. This lack of circulating plasma α1 -antitrypsin leaves the lungs exposed to enzymatic damage that is thought to underlie the adult onset emphysema. We have shown that the Z variant of α1 -antitrypsin is retained within hepatocytes as it causes a unique conformational transition and proteinprotein interaction. The mutation distorts the relationship between the reactive center loop and β-sheetA (Fig. 1). The consequent perturbation in structure allows the formation of an unstable intermediate that we have called M*.29−31 M* is characterized by inactivity and a raised melting temperature, and favors the formation of polymers in which the reactive center loop of one α1 -antitrypsin molecule sequentially inserts into β-sheet A of another.29,30,32−36 Spectroscopic analysis has demonstrated that oligomers of α1 -antitrypsin formed during an initial lag phase and condensed to form a heterogenous mixture of longer polymers.29,36 It is these polymers that accumulate within the endoplasmic reticulum of hepatocytes to form the PAS positive inclusions that are the hallmark of Z α1 -antitrypsin liver disease.32,37−39 The quality control mechanisms within the hepatocyte that handle polymers are now being elucidated.37,40−42 However, despite a more detailed understanding of the disposal pathway, it still remains unclear how the accumulation of Z α1 -antitrypsin within hepatocytes causes cell death and liver cirrhosis. The temperature and concentration dependence of polymerization,29,32 along with genetic factors,43 may account for the heterogeneity in liver disease among individuals who are homozygous for the Z mutation. The synthesis of α1 -antitrypsin rises during episodes of inflammation as part

ch20

Jan. 10, 2007

8:57

SPI-B429

448


D. A. Lomas et al.

Fig. 1. Pathways of serpin polymerization. The classic pathway of serpin polymerization is illustrated within the broken line. This is best described for Z α1 -antitrypsin and mutant neuroserpin. The polymerization of latent S49P neuroserpin is shown extending outside the broken line. The structure of a typical serpin such as α1 -antitrypsin (M) is centered on β-sheet A (green) and the mobile reactive centre loop (red).5 Mutations in the shutter domain (blue circle) result in the formation of an unstable intermediate (M*) that has an open β-sheet A.29−31 This patent β-sheet A can either accept the loop of another molecule (step 2) to form a dimer (D), which then extends into polymers (P), or its own loop to form a fully (step 3) or partially (step 4) loop-inserted latent conformation (Lf or Lp ).194−197 Ser49Pro neuroserpin spontaneously transforms into Lp under physiological conditions. The reactive loop is most likely to be inserted less into latent Ser49Pro neuroserpin than other latent serpins as the former has a lower melting temperature and less β-structure on CD analysis.72 Latent Ser49Pro neuroserpin is also able to form polymers under physiological conditions through either steps 5 to 2 or step 6. Stable oligomers (O) form on the pathway from 6 in the presence or absence of 0.5 M citrate showing that this pathway is distinct from that of steps 5 to 2. Reproduced from Ref. 72 with permission.

of the acute phase response. At these times, the formation of polymers is likely to overwhelm the degradative pathway, thereby exacerbating the formation of hepatic inclusions and the associated hepatocellular damage. Clearly, further prospective studies are required to assess whether pyrexial

ch20

Jan. 10, 2007

8:57

SPI-B429



449

episodes are more frequent and cause more intra-hepatic polymers in Z α1 antitrypsin homozygotes who develop liver disease than in those individuals who remain asymptomatic. Although many α1 -antitrypsin deficiency variants have been described, only two other mutants of α1 -antitrypsin have been similarly associated with profound plasma deficiency and hepatic inclusions: α1 -antitrypsin Siiyama (Ser53Phe)44,45 and Mmalton46 (deletion of phenylalanine at position 52, also known as Mnichinan47 and Mcagliari).48 The Siiyama mutation is the commonest cause of severe α1 -antitrypsin deficiency in Japan and the Mmalton variant is the commonest cause of severe α1 -antitrypsin deficiency in Sardinia. Both of these mutants are in the shutter domain underlying the bifurcation of strands 3 and 5 of β-sheet A (Fig. 1). The mutations disrupt a hydrogen bond network that is based on residue 334His and bridges strands 3 and 5 of the A sheet,49 causing it to partly allow the formation of folding intermediates50 and loop-sheet polymers in vivo.51,52 Polymerization also underlies the mild plasma deficiency of other variants that perturb the shutter domain: S (Glu264Val) and I (Arg39Cys) α1 antitrypsin.53,54 These point mutations cause less disruption to β-sheet A than does the Z variant. Thus, the rates of polymer formation are much slower than that of Z α1 -antitrypsin,29 and this results in less retention of protein within hepatocytes, milder plasma deficiency, and the lack of a clinical phenotype (Table 1). However, if a mild, slowly polymerizing I or S variant of α1 -antitrypsin is inherited with a rapidly polymerizing Z variant, then the two can interact to form heteropolymers within hepatocytes leading to inclusions and finally cirrhosis.54−56 Thus, the severity of retention of mutants of α1 -antitrypsin within hepatocytes can be explained by the rate of polymer formation. Those mutants that cause the most rapid polymerization cause the most retention of α1 -antitrypsin within the liver. This in turn correlates with the greatest risk of liver damage and cirrhosis, and the most severe plasma deficiency. 2.2. Familial encephalopathy with neuroserpin inclusion bodies Perhaps the most striking demonstration of polymer-associated disease results from mutations in neuroserpin. Neuroserpin is a member of the

ch20

Jan. 10, 2007

8:57

SPI-B429


450

D. A. Lomas et al. Table 1 Correlation between the rate of polymerization of mutants of α1 -antitrypsin and the severity of the accompanying plasma deficiency. Mutant name

Mutation

Rate of polymerization

Plasma deficiency

Z Siiyama Mmalton S I

Glu342Lys Ser53Phe 52Phe Glu264Val Arg39Cys

+++ +++ +++ + +

+++ +++ +++ + +

Based on data from Ref. 29. There is a striking genotype–phenotype correlation that is explicable by the rate of polymer formation. Reproduced from Ref. 198 with permission.

serpin superfamily that is predominantly expressed by neurons in the developing and adult brain. It is secreted from the axonal growth cones of the central and peripheral nervous system where it inhibits the enzyme tissue plasminogen activator (tPA).57−61 The expression pattern of neuroserpin and its in vitro inhibitory activity against tPA suggest that neuroserpin has a role in controlling synaptic connections, regulating emotional behavior and memory, reducing epileptic seizure activity, and limiting damage in cerebral infarction.62−66 We have recently described an autosomal dominant dementia, FENIB, that is characterized by inclusions of mutant neuroserpin as Collins’ bodies within cortical and subcortical neurons. The inclusions are PAS positive and diastase resistant, and bear a striking resemblance to those of Z α1 -antitrypsin that form within the liver (Fig. 2). The observation that FENIB was associated with a mutation Ser49Pro in the neuroserpin gene that was homologous to one in α1 -antitrypsin that causes cirrhosis (Ser53Phe),51 strongly indicated a common molecular mechanism. This was confirmed by the finding that the neuronal inclusion bodies of FENIB were formed by entangled polymers of neuroserpin with identical morphology to those isolated from hepatocytes from an individual with Z α1 -antitrypsin related cirrhosis.67 Four other families have now been identified with FENIB. These have allowed comparison of the severity of the mutation (as predicted by molecular modeling), the number of inclusions, and the age of onset of dementia (Table 2). Affected members in the original family with Ser49Pro neuroserpin (neuroserpin Syracuse) had diffuse small intraneuronal inclusions

ch20

Jan. 10, 2007

8:57

SPI-B429



451

Fig. 2. Z α1 -antitrypsin is retained within hepatocytes as intracellular inclusions. These inclusions are PAS positive and diastase resistant (a, arrowed) and are associated with neonatal hepatitis and hepatocellular carcinoma. (c) Electron micrograph of a hepatocyte from the liver of a patient with Z α1 -antitrypsin deficiency shows the accumulation of α1 -antitrypsin within the rough endoplasmic reticulum. These inclusions are composed of chains of α1 -antitrypsin polymers shown here from the plasma of a Siiyama α1 -antitrypsin homozygote (e) and from the liver of a Z α1 -antitrypsin homozygote (f ). Similar mutations in α1 -antitrypsin deficiency and neuroserpin encephalopathy result in similar intracellular inclusions of α1 -antitrypsin and neuroserpin. They are shown here in hepatocytes and neurons with PAS staining (a and b, respectively) and as endoplasmic reticulum aggregates of the abnormal proteins on electron microscopy (c and d, respectively). Electron microscopy confirms that the abnormal neuroserpin forms bead-like polymers and entangled polymeric aggregates identical to those shown here with Z α1 -antitrypsin (e and f, respectively). (Magnification left to right: ×200, ×20, 000, ×220, 000). Figure reproduced from Ref. 17 with permission.

of neuroserpin with an onset of dementia between the ages of 45 and 60 years.67−69 A second family, with a conformationally more severe mutation (neuroserpin Portland; Ser52Arg) had larger inclusions and an onset of dementia in early adulthood whilst a third family, with yet another mutation (His338Arg), had even more inclusions and the onset of dementia in adolescence. The most striking example was the family with the most

ch20

Jan. 10, 2007

8:57

SPI-B429


452

D. A. Lomas et al.

Table 2 Correlation between the rate of polymerization of mutants of neuroserpin, the number of inclusions and the severity of the associated dementia. Mutation

Rate of polymerization

Inclusions

Age of onset

Ser49Pro Ser52Arg His338Arg

+ ++ +++

+ ++ +++

45–63 20–40 15

Gly392Glu

++++

++++

13

Clinical features

Dementia, seizures Dementia, myoclonus Progressive myoclonus epilepsy Progressive myoclonus epilepsy, chorea

Based on data from Refs. 60, 67, 70, 71 and 73. There is a striking genotype–phenotype correlation that is explicable by the rate of polymer formation. Reproduced from Ref. 198 with permission.

“polymerogenic” mutation of neuroserpin, Gly392Glu. This replacement of a consistently conserved residue in the shutter region resulted in large inclusions with affected family members dying by age 20 years.70 The role of polymerization in disease is supported in our demonstration that recombinant Ser49Pro neuroserpin has a greatly accelerated rate of polymerization when compared with that in the wild-type protein60,71,72 and that Ser52Arg, which causes a more severe clinical phenotype, polymerizes even more rapidly.71 The cellular handling of neuroserpin has been assessed by transiently transfecting COS cells with wild-type neuroserpin and mutants of neuroserpin that cause FENIB. The most striking feature of the cell model is the retention of Syracuse (Ser49Pro) and Portland (Ser52Arg) neuroserpin as intracellular aggregates composed of polymers of mutant neuroserpin, similar to the loop-sheet polymers of mutant neuroserpin that can be isolated from the brains of individuals affected by FENIB.73 Once again, Portland (Ser52Arg) neuroserpin accumulates more rapidly than the Syracuse (Ser49Pro) mutant in keeping with the more severe clinical phenotype. Thus, FENIB shows a clear genotype-phenotype correlation, with the severity of disease correlating closely with the propensity of the mutated neuroserpin to form polymers (Table 2). The molecular pathway of polymerization is illustrated in Fig. 1. However, in serpins such as plasminogen activator inhibitor-1, antithrombin and α1 -antichymotrypsin population of M* can result in intramolecular loop

ch20

Jan. 10, 2007

8:57

SPI-B429



453

insertion and an inert latent configuration74−77 (Fig. 1). Other serpins such as α1 -antitrypsin and antithrombin can also be induced to form the latent conformer by heating in stabilizing concentrations of sodium citrate.78,79 This latent conformation cannot revert to the native species without incubation at high temperatures or refolding from denaturants.78,80 We have shown recently that the Ser49Pro mutation of neuroserpin also favors the formation of the latent conformer and that latent Ser49Pro neuroserpin can be detected in Collins’ bodies isolated from patients with FENIB.72 This transition, like that of polymerization, inactivates neuroserpin as an inhibitor of tPA.72 The inactivation of Ser49Pro neuroserpin will exacerbate the deficiency of neuroserpin at the synapse that arises from polymerization and the 100-fold decrease in inhibitory activity caused by the Ser49Pro mutation.60 These factors combine to increase the concentration of synaptic tPA, which may exacerbate the neuropsychiatric features of FENIB.66,69 The biochemical properties of latent Ser49Pro neuroserpin suggest that the reactive loop is less stably incorporated into β-sheet A than it is in other latent serpins (compare Lf and Lp in Fig. 1). This means that far from being inert, latent Ser49Pro neuroserpin spontaneously forms polymers under physiological conditions.72 Consequently, we have modified the pathways of latency and polymerization72 (Fig. 1). In addition to the standard pathway (shown within the dotted box), polymers may also form via a back reaction as shown in steps 5 and 2, or directly as shown in step 6. Only latent Ser49Pro has the ability to form oligomers in the presence of 0.5 M citrate, indicating that the second polymerization pathway from step 6 is distinct from that of step 2. It is currently not possible to define which pathway (either steps 5 and 2 or step 6) is most important in the polymerization of latent Ser49Pro neuroserpin in vivo. Nevertheless, the ability of latent Ser49Pro neuroserpin to polymerize will increase the burden of polymers within neurones, thus exacerbating the neuronal dysfunction and death that underlies FENIB. By analogy with the Z allele of α1 -antitrypsin, it is likely that the PAS positive inclusions of neuroserpin are mediating their effect by a toxic gain of function. However, both polymerization and latency inactivate neuroserpin as a proteinase inhibitor and any residual monomeric mutant protein is much less active as an inhibitor of tPA than the wild-type protein.60,71,72 Thus, in addition to the toxicity of the polymeric protein, there are likely to

ch20

Jan. 10, 2007

8:57

SPI-B429

454


D. A. Lomas et al.

be additional consequences from an excess of tPA at the synapse. This can be manifested as neuropsychiatric disease66 and epilepsy,65 and these features have been indeed reported in individuals with FENIB.69,70,81 Thus, the dementia caused by mutations of neuroserpin may result from a combination of loss of function and toxic gain of function. The inclusions of polymers that underlie FENIB and Z α1 -antitrypsin associated cirrhosis have been used as a paradigm to explain other conformational diseases such as Alzheimer’s disease, Huntington’s disease and amyloidosis.16,17,82 However, they differ from other conformational diseases in three fundamental aspects: (i) the mutants are retained as ordered, rather than disordered structures,32,33,67,73 (ii) they are retained within the endoplasmic reticulum,73,83 and (iii) they do not induce an unfolded protein response.84,85

3. Loss of Function Serpinopathies 3.1. Plasma deficiency of antithrombin, C1 inhibitor, α1 -antichymotrypsin, and heparin co-factor II The phenomenon of loop-sheet polymerization has now been reported in mutants of other members of the serpin superfamily in association with disease. Naturally occurring mutations have been described in the shutter (Fig. 1) and other domains of the plasma proteins C1-inhibitor (Phe52Ser, Pro54Leu, Ala349Thr, Val366Met, Phe370Ser, and Pro391Ser),86,87 antithrombin (Pro54Thr, Asn158Asp, and Phe229Leu)75,88 and α1 antichymotrypsin (Leu55Pro and Pro229Ala)30,89,90 (see later). In all cases, the residue numbers are based on the serpin template to allow comparison between members of the family.91 These mutations destabilize the serpin architecture to allow the formation of inactive loop-sheet polymers. The polymers probably form within the endoplasmic reticulum of the liver, but are cleared by the degradative pathway. They are not usually associated with the formation of inclusions and the liver disease seen in individuals who are homozygous for Z α1 -antitrypsin. One can speculate that this is because C1-inhibitor, antithrombin and α1 -antichymotrypsin are synthesized at approximately 10% of the rate of α1 -antitrypsin, and mutations are usually found in heterozygotes. However, one variant of

ch20

Jan. 10, 2007

8:57

SPI-B429



455

α1 -antichymotrypsin (Pro229Ala) has been shown to form granular inclusions within hepatocytes, analogous to those formed by Z α1 -antitrypsin.89 This mutation also allows the spontaneous formation of polymers in vitro (B. Gooptu and D. Lomas, unpublished observations). The individual with the Pro229Ala α1 -antichymotrypsin mutation was infected with the hepatitis C virus and it seems likely that the viral infection drove the inflammatory response which increased the production of α1 -antichymotrypsin polymers to form the inclusions. Clearly, it is impossible to know whether it was the hepatitis C or the inclusions of mutant α1 -antichymotrypsin that caused the associated hepatic damage, but the case illustrates that mutants of other serpins can form hepatic inclusions analogous to those of Z α1 -antitrypsin. The lack of circulating protein in individuals with C1-inhibitor, antithrombin and α1 -antichymotrypsin deficiency allows uncontrolled activity of proteolytic cascades, and hence angio-edema, thrombosis, and chronic obstructive pulmonary disease, respectively (see Refs. 15–17 for reviews). More recently, a mutation in heparin co-factor II (Glu428Lys) has been associated with plasma deficiency, but this has not yet been shown to cause disease.92 The mutation is of particular interest as it is the same as the Z allele that causes polymerization and deficiency of α1 -antitrypsin. We have shown that this same mutation also causes temperature dependent polymerization and inactivation of the Drosophila serpin, necrotic.93

4. Loss of Function, Gain of Function, and Respiratory Disease 4.1. Z α1 -antitrypsin deficiency and respiratory disease Emphysema was noted in some of the first individuals who were reported to have an absence of the α1 band on serum protein electrophoresis.20 It was confirmed by family studies94 and is now the only genetic factor that is widely accepted to predispose smokers to emphysema. The respiratory disease associated with α1 -antitrypsin deficiency usually presents with increasing dyspnoea with cor pulmonale and polycythemia occurring late in the course of the disease. Chest radiographs typically show bibasal emphysema with paucity and pruning of the basal pulmonary vessels.

ch20

Jan. 10, 2007

8:57

SPI-B429

456


D. A. Lomas et al.

Upper lobe vascularization is relatively preserved. Ventilation/perfusion radioisotope scans and angiography also show abnormalities with a lower zone distribution.94 Lung function tests are typical for emphysema with a reduced ratio of forced expiratory volume in 1 s to forced vital capacity (FEV1 /FVC), gas trapping (raised ratio of residual volume to total lung capacity), and low gas transfer factor. The onset of respiratory disease can be delayed to the sixth decade in never-smokers with PI*Z α1 -antitrypsin deficiency and these individuals often have a normal life span.96 Highresolution CT scans with 1 to 2 mm collimation are the most accurate method of assessing the distribution of panlobular emphysema and for monitoring progress of the pulmonary disease97,98 although this currently has little clinical value beyond clinical trials. The severity of lung function abnormalities and health status correlates well with the quantification of emphysema by CT scans.99 Plasma deficiency of α1 -antitrypsin is also associated with an increased prevalence of asthma100,101 and Wegener’s granulomatosis.102,103 Moreover, there have been reports of an increased prevalence of bronchiectasis in individuals with α1 -antitrypsin deficiency.104,105 This was recently investigated in a survey of 202 patients with bronchiectasis. In 121 of these patients, the cause of the bronchiectasis was unknown. The allelic frequency of S and Z α1 -antitrypsin was no different in the patients with bronchiectasis, compared with that in healthy blood donor controls living in the same geographic area.106 Thus, the authors concluded that α1 -antitrypsin does not predispose to the development of bronchiectasis. They did however find an over representation of PiZ alleles in patients with bronchiectasis and co-existing emphysema, and suggested that the bronchiectasis may be a consequence of emphysema in Z α1 -antitrypsin homozygotes rather than a primary effect. 4.2. Extra-pulmonary conditions associated with α1 -antitrypsin deficiency Panniculitis is a rare complication of α1 -antitrypsin deficiency that is characterized by an acute inflammatory infiltrate and fat necrosis. The first case of panniculitis in association with α1 -antitrypsin deficiency was described in 1972.107 Since then, there have been 28 case reports with a mean age of

ch20

Jan. 10, 2007

8:57

SPI-B429



457

approximately 40 years and an equal sex distribution (reviewed in Ref. 108). 64% of individuals have a Pi*Z phenotype and there is strong anecdotal evidence that replacing the deficient α1 -antitrypsin brings about a resolution of symptoms.108,109 The risk of other extra-pulmonary conditions has been assessed in 605 patients with α1 -antitrypsin deficiency enrolled in the Danish registry, because of respiratory symptoms or through family screening.110 A case-control analysis revealed that there was an increased risk of liver cirrhosis (relative risk 6.0), pancreatitis (relative risk 3.1) and gall stones (relative risk 1.9), and a reduced risk of cerebrovascular disorders (relative risk 0.5) in individuals with α1 -antitrypsin deficiency. This study found no increased risk of vasculitis, rheumatoid arthritis, arterial diseases or aneurysms. This is in contrast to case reports of vascular aneurysms in individuals with MZ and ZZ α1 -antitrypsin phenotypes.111,112 4.3. Polymerization of Z α1 -antitrypsin and emphysema Emphysema associated with plasma deficiency of α1 -antitrypsin is widely believed to be due to the reduction in plasma levels of α1 -antitrypsin to 10–15% of normal. This in turn markedly reduces the α1 -antitrypsin that is available to protect the lungs against proteolytic attack by the enzyme neutrophil elastase.113 The situation is exacerbated as the Z mutation reduces the association rate between α1 -antitrypsin and neutrophil elastase by approximately fivefold.114−117 Thus, the α1 -antitrypsin available within the lung is not as effective as the normal M protein. This combination of α1 -antitrypsin deficiency, reduction in the efficacy of the α1 antitrypsin molecule, and cigarette smoke can have a devastating effect on lung function,118,119 probably by allowing the unopposed action of proteolytic enzymes. The inhibitory activity of Z α1 -antitrypsin can be further reduced as the Z mutation favors the spontaneous formation of α1 -antitrypsin loop-sheet polymers within the lung.120 This conformational transition inactivates α1 -antitrypsin as a proteinase inhibitor, thereby further reducing the already depleted levels of α1 -antitrypsin that are available to protect the alveoli (see Ref. 16 for a review). Moreover, the conversion of α1 -antitrypsin from a monomer to a polymer converts it to a chemoattractant for human neutrophils.121,122 The magnitude of the effect

ch20

Jan. 10, 2007

8:57

SPI-B429

458


D. A. Lomas et al.

is similar to that of the chemoattractant C5a and present over a range of physiological concentrations (EC50 4.5 ±2 µg/ml). Polymers also induced neutrophil shape change and stimulated myeloperoxidase release and neutrophil adhesion.121 The chemotactic properties of polymers were confirmed by one group122 but refuted by another.123 More recently, we have used a monoclonal antibody to demonstrate polymers in emphysematous tissue associated with Z α1 -antitrypsin deficiency [Fig. 3(a)], but not in emphysema in individuals with normal levels of α1 -antitrypsin [Fig. 3(b)]. Neutrophils co-localized with polymers in the alveoli [Fig. 3(c)]. The pro-inflammatory properties of polymers were further confirmed by the demonstration that they caused a neutrophil influx when instilled into the lungs of mice.124 Therefore, the chemoattractant properties of polymers may explain the excess number of neutrophils in bronchoalveolar

Fig. 3. Polymers of α1 -antitrypsin and lung disease. Polymers can be detected in the emphysematous regions of the lungs from individuals with Z α1 -antitrypsin deficiency (a; brown staining) but not in regions of emphysema from M α1 -antitrypsin controls (b). The polymers co-localize with neutrophils in alveolar tissue (c; neutrophils in red and arrowed, polymers in brown). The chemoattractant properties of polymers are quite likely to be an important factor in the recruitment of excess neutrophils to the lungs of Z, rather than M α1 antitrypsin homozygotes with emphysema (d). Reproduced from Ref. 124 with permission.

ch20

Jan. 10, 2007

8:57

SPI-B429



459

lavage125 and in tissue sections of lung parenchyma [Fig. 3(d)] from individuals with Z α1 -antitrypsin deficiency. Moreover, polymers may contribute to the excess inflammation that is apparent even in individuals with Z α1 -antitrypsin deficiency, with very early lung disease126 and may drive the progressive inflammation that continues even after cessation of smoking. Any pro-inflammatory effect of polymers is likely to be exacerbated by inflammatory cytokines, cleaved or complexed α1 -antitrypsin,127 elastin degradation products128 and cigarette smoke, which themselves cause neutrophil recruitment. Thus, our understanding of the biological properties of α1 -antitrypsin provides novel pathways for the pathogenesis of emphysema in individuals who are homozygous for the Z mutation (Fig. 4). Emphysema associated with Z α1 -antitrypsin deficiency is likely to result from a combination of a loss of function of α1 -antitrypsin (deficiency of circulating proteinase inhibitor, reduced inhibitory activity, and intra-alveolar polymerization) and a toxic gain of function as a result of intra-alveolar polymer formation. For many years, the emphysema associated with Z α1 -antitrypsin deficiency has been a paradigm for emphysema seen in smokers who have normal levels of α1 -antitrypsin. However, this is clearly an over simplification as emphysema associated with Z α1 -antitrypsin deficiency has

Fig. 4. Proposed model for the pathogenesis of emphysema in patients with Z α1 antitrypsin deficiency. The plasma deficiency and reduced inhibitory activity of Z α1 antitrypsin may be exacerbated by the polymerization of α1 -antitrypsin within the lungs. These processes inactivate the inhibitor thereby further reducing the antiproteinase screen. α1 -Antitrypsin polymers may also act as a pro-inflammatory stimulus to attract and activate neutrophils, thereby increasing tissue damage. Reproduced from Ref. 15 with permission.

ch20

Jan. 10, 2007

8:57

SPI-B429

460


D. A. Lomas et al.

a different distribution (lower rather than upper lobe), different pathology (panlobular rather than centrilobular emphysema), the presence of pro-inflammatory lung polymers,120−122,124 and different patterns of gene expression.129 It seems likely that more differences will become apparent as we dissect the pathways of inflammation and tissue damage in individuals with α1 -antitrypsin deficiency. Understanding the chemotactic effect of α1 -antitrypsin polymers may partly explain some of the other inflammatory conditions that are associated with Z α1 -antitrypsin deficiency. Much of the tissue damage observed in bronchiectasis,104 vasculitis,102 panniculitis,130 and asthma101,131 is neutrophil derived and it is possible that α1 -antitrypsin polymers are one of the factors that drive inflammation and disease progression.

5. Conformational Transitions of α1 -Antichymotrypsin and Emphysema α1 -Antichymotrypsin is a 374 amino acid, 64 kDa, acute phase glycoprotein that is secreted from the liver132 and human bronchial epithelial cells.133 It circulates in the plasma at one tenth the concentration of α1 -antitrypsin and has inhibitory activity against cathepsin G,19,134,135 mast cell chymase,136 and chymotrypsin.19 The gene has been mapped to chromosome 14q32.1 and lies in close proximity to the α1 -antitrypsin gene.137 Alpha1-antichymotrypsin deficiency was first associated with airflow obstruction in five volunteers undergoing health screening.138,139 Further studies have reported decreased circulating levels of α1 -antichymotrypsin in association with the Pro229Ala,89,140 Leu55Pro90 and Met389Thr mutations141 and with a two base pair deletion at codon 391.142 To date, only heterozygotes with α1 -antichymotrypsin deficiency have been described and it is assumed that homozygote deficiency is incompatible with normal embryonic development. Pro229Ala and Leu55Pro α1 -antichymotrypsin mutations have been reported in 4%89,140 and 1.5%,90 respectively, of unrelated patients with chronic obstructive pulmonary disease (COPD), but in no healthy controls. However, the association of these two mutations with COPD was not replicated in a study of 168 patients with COPD and 61 control subjects.143 Moreover, the mutations are uncommon so that even if they do predispose to COPD, it will be in relatively few individuals.143,144

ch20

Jan. 10, 2007

8:57

SPI-B429



461

Wild-type α1 -antichymotrypsin in the lungs of patients with chronic obstructive pulmonary disease is intact but inactive.145 An understanding of the conformational transitions of the serpins has provided an explanation of the molecular basis for this inactivation. Biochemical evidence from bronchoalveolar lavage suggests that α1 -antichymotrypsin adopts a latent conformation.77 This loss of function conformational transition exacerbates the excess of proteolytic enzymes and so increases tissue damage. The role of the latent conformer of α1 -antichymotrypsin in the pathogenesis of COPD remains to be clarified. In particular, longitudinal studies are required to assess the interaction between the aberrant conformer, smoking, infection, and decline in the lung function. Finally, there is evidence of segregation of a signal peptide mutation (-Thr15Ala) that possibly increases the production of α1 -antichymotrypsin with an increased risk of late onset Alzheimer’s disease.146 This observation has not been confirmed by other groups.147−151 This polymorphism has been shown to increase the risk of chronic lung disease in Japanese smokers,152 but this was not reproduced in an Italian cohort.144

6. Treatment of α1 -Antitrypsin Deficiency The only treatment available for the cirrhosis associated with α1 -antitrypsin deficiency is supportive therapy. The liver disease progresses with varying speed and lacks specific features.153,154 Liver transplantation provides definitive treatment for patients with end-stage α1 -antitrypsin deficiency related cirrhosis and children who receive a transplant have an excellent clinical outcome.155 There is good evidence that many Z α1 -antitrypsin homozygotes would develop only mild lung disease if they abstain from smoking.96,156 Patients with α1 -antitrypsin deficiency related emphysema should receive advice on smoking cessation, and where appropriate, they should be offered nicotine replacement therapy or bupropion.157 Those patients with airflow obstruction should be assessed with lung function tests followed by trials of bronchodilators and inhaled corticosteroids.158 Many patients benefit from pulmonary rehabilitation,159 and where appropriate, assessment for long-term oxygen therapy.158 The most severely

ch20

Jan. 10, 2007

8:57

SPI-B429

462


D. A. Lomas et al.

affected individuals should be referred for consideration of lung transplantation. In appropriately selected patients with α1 -antitrypsin deficiency related emphysema, lung transplantation prolongs survival, improves functional capacity, and enhances the quality of life.160 However, rejection remains an obstacle to better medium-term results and currently heart lung transplantation has a five-year survival of approximately 60%. There is anecdotal evidence that patients with α1 -antitrypsin deficiency have more post-operative complications following lung transplantation. Assessment of lung lavage from 11 transplant recipients with α1 -antitrypsin deficiency revealed minimal or unmeasureable free elastase.161 However, free elastase was detected in three of seven patients during severe respiratory illnesses. The requirement for α1 -antitrypsin replacement therapy in patients with α1 -antitrypsin deficiency after lung transplantation requires further investigation. The role of lung volume reduction surgery (LVRS) in patients with α1 -antitrypsin is unclear. Cassina and colleagues reported 12 patients with α1 -antitrypsin deficiency who had undergone bilateral thoracotomy and lower lobe LVRS.162 They found a transient improvement in FEV1 and static lung volumes with deterioration to below the starting values by the 24th month of follow-up. The only exception was the six minute walk distance, which was still slightly improved, compared with the baseline after 24 months. In contrast, Gelb and colleagues reported six patients with α1 -antitrypsin deficiency in whom there was a significant improvement in lung function, following bilateral lower lobe video-assisted thoracoscopic LVRS.163 In four of the patients who were followed up for more than 22 months, there was a sustained improvement in breathlessness and exercise tolerance.163 Despite the conflicting reports, both authors agree that the benefit from LVRS is far less in patients with α1 -antitrypsin deficiency than in individuals with smoking related upper lobe emphysema with normal levels of α1 -antitrypsin.162,163 Thus, LVRS should not be offered routinely in patients with α1 -antitrypsin deficiency until further information is available.164 The genetic deficiency in the antielastase screen may be rectified biochemically by intravenous infusions of α1 -antitrypsin.113 There is registry data to suggest that this therapy may slow the rate of decline in the lung function of patients with an FEV1 of 35–49% predicted,165 but this has

ch20

Jan. 10, 2007

8:57

SPI-B429



463

yet to be proven in randomized controlled trials. The only controlled trial that has assessed α1 -antitrypsin replacement therapy showed that infusions of α1 -antitrypsin may retard the progression of emphysema as assessed by HRCT scans, but had no effect on the decline in FEV1 .166 A larger study is required to confirm these findings167 and α1 -antitrypsin replacement therapy (Prolastin®, Bayer) is currently not available in many parts of Europe. Other treatments at earlier stages of development include gene therapy and the administration of retinoic acid and chemical chaperones. Vectors carrying the α1 -antitrypsin gene have been targeted to liver,168 lung,169 and muscle170,171 in animals. There is good expression of α1 -antitrypsin, but further data is required to assess whether this can be achieved in man. In particular, it is important to determine the length of time of protein expression and whether the levels of α1 -antitrypsin in the epithelial lining fluid of the lung are sufficient to prevent ongoing proteolytic damage. Similarly, although the effects of retinoic acid on alveolar regeneration in the rat appear promising,172 they have yet to be demonstrated in patients with emphysema.

7. Novel Strategies to Prevent Conformational Transitions and Disease There is now strong evidence that polymers of α1 -antitrypsin, and indeed of mutants of all other serpins that are associated with disease, are formed by an aberrant linkage between the reactive center loop of one molecule and β-sheet A of another.2,32,34,173−176 This has allowed the development of new strategies to attenuate polymerization and thus treat the associated disease. Three strategies have been pursued to date: (i) chemical chaperones to stabilize the unstable mutant serpin, (ii) filling a surface cavity to block the conformational transition underlying polymer formation, and (iii) the use of reactive loop peptides that compete for binding to β-sheet A. 7.1. Chemical chaperones to stabilize serpins and block polymerization Chemical chaperones can stabilize intermediates on the folding pathway. Osmolytes such as betaine, trimethyamine oxide, and sarcosine

ch20

Jan. 10, 2007

8:57

SPI-B429

464


D. A. Lomas et al.

all stabilize α1 -antitrypsin against polymer formation.177 However, the chaperone trimethyamine oxide had no effect on the secretion of Z α1 antitrypsin in cell culture,178 as it favored the conversion of unfolded Z α1 -antitrypsin to polymers.179 In contrast, glycerol increased the secretion of Z α1 -antitrypsin from cell lines178 most likely as it binds to and stabilizes β-sheet A.49 4-Phenylbutyrate (4-PBA) also increased the secretion of Z α1 -antitrypsin from cell lines and transgenic mice.178 This agent has been used for several years to treat children with urea cycle disorders, and more recently, 4-PBA has been shown to increase the expression of mutant (F508) cystic fibrosis transmembrane regulator protein in vitro180 and in vivo.181 Unfortunately, 4-PBA had no effect on the secretion of α1 -antitrypsin in patients with α1 -antitrypsin deficiency.182 Finally, a recent study has shown that the chaperone α-crystallin can block the polymerization of α1 -antichymotrypsin, which is nucleation-dependent, but not that of α1 -antitrypsin, which is not dependent on nucleation.183 . Thus, strategies to reduce nucleation may be effective in preventing polymerization of some serpins. 7.2. Filling a surface hydrophobic pocket to block polymerization A second strategy comes from the identification of a hydrophobic pocket in α1 -antitrypsin that is bounded by strand 2A and helices D and E.5,184 The cavity is patent in the native protein, but is filled as β-sheet A accepts an exogenous reactive loop peptide during polymerization.5 The introduction of either Thr114Phe or Gly117Phe on strand 2 of β-sheet A within this cavity significantly raised the melting temperature of α1 -antitrypsin and retarded polymer formation. Conversely, Leu100Phe on helix D accelerated polymer formation, but this effect was abrogated by the addition of Thr114Phe. None of these mutations affected the inhibitory activity of α1 -antitrypsin. The importance of these observations was underscored by the finding that the Thr114Phe mutation reduced polymer formation and increased the secretion of Z α1 -antitrypsin from a Xenopus oocyte expression system. Moreover, cysteine mutants within the hydrophobic pocket were able to bind a range of fluorophores, illustrating the accessibility of the cavity to external agents. These data demonstrate the importance of this

ch20

Jan. 10, 2007

8:57

SPI-B429



465

cavity as a site for rational drug design to ameliorate polymerization and to treat the associated conformational disease.185 However, this approach may not be applicable to all serpins, as amphipathic organoligands can bind to this region of PAI-1 to induce polymer formation.186−188 7.3. Peptides with homology to the reactive center loop can compete for binding to β -sheet A and block polymerization We have shown previously that the polymerization of Z α1 -antitrypsin can be blocked by annealing of reactive loop peptides to β-sheet A.32,189 Such peptides were 11–13 residues in length and could bind to other members of the serpin superfamily.189,190 This was most clearly demonstrated by the finding that the reactive loop peptide of antithrombin inserted more readily to β-sheet A of α1 -antitrypsin and vice versa.191 These peptides, although

Fig. 5. The design of a selective inhibitor to block the polymerization of Z α1 -antitrypsin. The Z mutation (ringed in blue) allows partial insertion of the reactive centre loop (red) into β-sheet A (green) to form the intermediate M* (see Fig. 1). This opens the lower part of β-sheet A thereby favoring the incorporation of the reactive loop of another molecule and hence polymerization (upper pathway). Understanding the configuration of the reactive loop has allowed the design of a 6-mer peptide (yellow) that specifically binds to Z α1 -antitrypsin and so prevents polymer formation Ref. 31.

ch20

Jan. 10, 2007

8:57

SPI-B429


466

D. A. Lomas et al.

useful in establishing the mechanism of polymerization, are too long and too promiscuous to be suitable for rational drug design. More recently, the recognition that the Z mutation forces α1 -antitrypsin into a conformation similar to M* (see Fig. 1) has allowed the design of a 6-mer peptide that specifically anneals to the lower part of β-sheet A and blocks polymerization (Fig. 5).31,192,193 This peptide was specific to Z α1 -antitrypsin and did not anneal significantly to other serpins (such as antithrombin, α1 -antichymotrypsin, and PAI-1) with a similar tertiary structure. Indeed, smaller trimer peptides have been developed that will also anneal to a patent β-sheet A of antithrombin in vitro.49 The aim now is to convert these peptides into small molecule inhibitors that can be used to block aberrant polymerization in vivo.

Acknowledgments This work was supported by the Medical Research Council (UK), the Wellcome Trust (UK), and Papworth NHS Trust.

References 1. Silverman GA, Bird PI, Carrell RW, Church FC, Coughlin PB, Gettins P, Irving J, Lomas DA, Moyer RW, Pemberton P, Remold O’Donnell EGS, Travis J and Whisstock J (2001) The serpins are an expanding superfamily of structurally similar but functionally diverse proteins. Evolution, novel functions, mechanism of inhibition and a revised nomenclature. J Biol Chem 276:33293–33296. 2. Elliott PR, Lomas DA, Carrell RW and Abrahams J-P (1996) Inhibitory conformation of the reactive loop of α1 -antitrypsin. Nat Struct Biol 3:676–681. 3. Ryu S-E, Choi H-J, Kwon K-S, Lee KN and Yu M-H (1996) The native strains in the hydrophobic core and flexible reactive loop of a serine protease inhibitor: Crystal structure of an uncleaved α1 -antitrypsin at 2.7Å. Structure 4:1181–1192. 4. Elliott PR, Abrahams J-P and Lomas DA (1998) Wildtype α1 -antitrypsin is in the canonical inhibitory conformation. J Mol Biol 275:419–425. 5. Elliott PR, Pei XY, Dafforn TR and Lomas DA (2000) Topography of a 2.0Å structure of α1-antitrypsin reveals targets for rational drug design to prevent conformational disease. Protein Sci 9:1274–1281.

ch20

Jan. 10, 2007

8:57

SPI-B429



467

6. Kim S-J, Woo J-R, Seo EJ, Yu M-H and Ryu S-E (2001) A 2.1Å resolution structure of an uncleaved α1 -antitrypsin shows variability of the reactive centre and other loops. J Mol Biol 306:109–119. 7. Jin L, Abrahams J-P, Skinner R, Petitou M, Pike R and Carrell RW (1997) The anticoagulant activation of antithrombin by heparin. Proc Natl Acad Sci USA 94:14683–14688. 8. Li J, Wang Z, Canagarajah B, Jiang H, Kanost M and Goldsmith EJ (1999) The structure of active serpin 1K from Manduca sexta. Structure 7:103–109. 9. Wilczynska M, Fa M, Ohlsson P-I and Ny T (1995) The inhibition mechanism of serpins. Evidence that the mobile reactive centre loop is cleaved in the native protease-inhibitor complex. J Biol Chem 270:29652-29655. 10. Wilczynska M, Fa M, Karolin J, Ohlsson P-I, Johansson LB-A and Ny T (1997) Structural insights into serpin-protease complexes reveal the inhibitory mechanism of serpins. Nat Struc Biol 4:354-357. 11. Stratikos E and Gettins PGW (1997) Major proteinase movement upon stable serpin-proteinase complex formation. Proc Natl Acad Sci 4:453–458. 12. Stratikos E and Gettins PGW (1998) Mapping the serpin-proteinase complex using single cysteine variants of α1 -antitrypsin inhibitor Pittsburgh. J Biol Chem 273:15582–15589. 13. Stratikos E and Gettins PGW (1999) Formation of the covalent serpinproteinase complex involves translocation of the proteinase by more than 70Å and full insertion of the reactive centre loop into β-sheet A. Proc Natl Acad Sci USA 96:4808–4813. 14. Huntington JA, Read RJ and Carrell RW (2000) Structure of a serpin-protease complex shows inhibition by deformation. Nature 407:923–926. 15. Lomas DA and Mahadeva R (2002) alpha-1-antitrypsin polymerization and the serpinopathies: pathobiology and prospects for therapy. J Clin Invest 110:1585–1590. 16. Lomas DA and Carrell RW (2002) Serpinopathies and the conformational dementias. Nat Rev Genet 3:759–768. 17. Carrell RW and Lomas DA (2002) Alpha1 -antitrypsin deficiency: A model for conformational diseases. New Engl J Med 346:45–53. 18. Johnson D and Travis J (1978) Structural evidence for methionine at the reactive site of human α1 -proteinase inhibitor. J Biol Chem 253: 7142–7144. 19. Beatty K, Bieth J and Travis J (1980) Kinetics of association of serine proteinases with native and oxidized α1 -proteinase inhibitor and α1 antichymotrypsin. J Biol Chem 255:3931–3934. 20. Laurell C-B and Eriksson S (1963) The electrophoretic α1 -globulin pattern of serum in α1 -antitrypsin deficiency. Scand J Clin Lab Invest 15:132–140.

ch20

Jan. 10, 2007

8:57

SPI-B429

468


D. A. Lomas et al.

21. Malone M, Mieli-Vergani G, Mowat AP and Portmann B (1989) The fetal liver in PiZZ alpha-1-antitrypsin deficiency: a report of 5 cases. Pediatr Pathol 9:623–631. 22. Sharp HL, Bridges RA, Krivit W and Freier EF (1969) Cirrhosis associated with α1 -antitrypsin deficiency: A previously unrecognised inherited disorder. J Lab Clin Med 73:934–939. 23. Eriksson S and Larsson C (1975) Purification and partial characterization of PAS-positive inclusion bodies from the liver in alpha-1-antitrypsin deficiency. N Eng J Med 292:176–180. 24. Sveger T (1976) Liver disease in alpha1 -antitrypsin deficiency detected by screening of 200,000 infants. N Engl J Med 294:1316–1321. 25. Sveger T (1978) α1 -antitrypsin deficiency in early childhood. Pediatrics 62:22–25. 26. Sveger T (1988) The natural history of liver disease in α1 -antitrypsin deficient children. Acta Paed Scand 77:847–851. 27. Sveger T and Eriksson S (1995) The liver in adolescents with α1 -antitrypsin deficiency. Hepatology 22:514–517. 28. Eriksson S, Carlson J and Velez R (1986) Risk of cirrhosis and primary liver cancer in alpha1 -antitrypsin deficiency. N Engl J Med 314:736–739. 29. Dafforn TR, Mahadeva R, Elliott PR, Sivasothy P and Lomas DA (1999) A kinetic mechanism for the polymerization of α1 -antitrypsin. J Biol Chem 274:9548–9555. 30. Gooptu B, Hazes B, Chang W-SW, Dafforn TR, Carrell RW, Read R and Lomas DA (2000) Inactive conformation of the serpin α1 -antichymotrypsin indicates two stage insertion of the reactive loop: Implications for inhibitory function and conformational disease. Proc Natl Acad Sci (USA) 97:67–72. 31. Mahadeva R, Dafforn TR, Carrell RW and Lomas DA (2002) Six-mer peptide selectively anneals to a pathogenic serpin conformation and blocks polymerisation: implications for the prevention of Z α1 -antitrypsin related cirrhosis. J Biol Chem 277:6771–6774. 32. Lomas DA, Evans DL, Finch JT and Carrell RW (1992) The mechanism of Z α1 -antitrypsin accumulation in the liver. Nature 357:605–607. 33. Janciauskiene S, Dominaitiene R, Sternby NH, Piitulainen E and Eriksson S (2002) Detection of circulating and endothelial cell polymers of Z and wildtype alpha-1-antitrypsin by a monoclonal antibody. J Biol Chem 277:26540– 26546. 34. Sivasothy P, Dafforn TR, Gettins PGW and Lomas DA (2000) Pathogenic α1 -antitrypsin polymers are formed by reactive loop-β-sheet A linkage. J Biol Chem 275:33663–33668.

ch20

Jan. 10, 2007

8:57

SPI-B429



469

35. James EL and Bottomley SP (1998) The mechanism of α1 -antitrypsin polymerization probed by fluorescence spectroscopy. Arch Biochem Biophys 356:296–300. 36. Purkayastha P, Klemke JW, Lavender S, Oyola R, Cooperman BS and Gai F (2005) α1 -antitrypsin polymerisation: A fluorescence correlation spectroscopic study. Biochemistry 44:2642–2649. 37. Wu Y, Swulius MT, Moremen KW and Sifers RN (2003) Elucidation of the molecular logic by which misfolded α1-antitrypsin is preferentially selected for degradation. Proc Natl Acad Sci USA 100:8229–8234. 38. Janciauskiene S, Eriksson S, Callea F, Mallya M, Zhou A, Seyama KSH and Lomas DA (2004) Differential detection of PAS-positive inclusions formed by the Z, Siiyama and Mmalton variants of α1 -antitrypsin. Hepatology 40:1203–1210. 39. An JK, Blomenkamp K, Lindblad D and Teckman JH (2005) Quantitative isolation of alphalAT mutant Z protein polymers from human and mouse livers and the effect of heat. Hepatology 41:160–167. 40. Cabral CM, Liu Y and Sifers RN (2001) Dissecting the glycoprotein quality control in the secretory pathway. TIBS 26:619–623. 41. Teckman JH, Burrows J, Hidvegi T, Schmidt B, Hale PD and Perlmutter DH (2001) The proteasome participates in degradation of mutant α1 -antitrypsin Z in the endoplasmic reticulum of hepatoma-derived hepatocytes. J Biol Chem 276:44865–44872. 42. Novoradovskaya N, Lee J, Yu Z-X, Ferrans VJ and Brantly M (1998) Inhibition of intracellular degradation increases secretion of a mutant form of α1 -antitrypsin associated with profound deficiency. J Clin Invest 101: 2693–2701. 43. Wu Y, Whitman I, Molmenti E, Moore K, Hippenmeyer P and Perlmutter DH (1994) A lag in intracellular degradation of mutant α1 -antitrypsin correlates with liver disease phenotype in homozygous PiZZ α1 -antitrypsin deficiency. Proc Natl Acad Sci USA 91:9014–9018. 44. Seyama K, Nukiwa T, Takabe K, Takahashi H, Miyake K and Kira S (1991) Siiyama (serine 53 (TCC) to phenylalanine 53 (TTC)) A new α1-antitrypsindeficient variant with mutation on a predicted conserved residue of the serpin backbone. J Biol Chem 266:12627–12632. 45. Seyama K, Nukiwa T, Souma S, Shimizu K and Kira S (1995) α1 -Antitrypsindeficient variant Siiyama (Ser53 [TCC] to Phe53 [TTC]) is prevalent in Japan Status of α1 -antitrysin deficiency in Japan. Am Rev Respir Dis 152:2119– 2126.

ch20

Jan. 10, 2007

8:57

SPI-B429

470


D. A. Lomas et al.

46. Roberts EA, Cox DW, Medline A and Wanless IR (1984) Occurrence of alpha-1-antitrypsin deficiency in 155 patients with alcoholic liver disease. Am J Clin Pathol 82:424–427. 47. Matsunaga E, Shiokawa S, Nakamura H, Maruyama T, Tsuda K and FukumakiY (1990) Molecular analysis of the gene of the α1 -antitrypsin deficiency variant, Mnichinan. Am J Hum Genet 46:602–612. 48. Sergi C, Consalez GC, Fabbretti G, Brisigotti M, Faa G, Costa V, Romeo G and Callea F (1994) Immunohistochemical and genetic characterization of the M Cagliari α-1-antitrypsin molecule (M-like α-1-antitrypsin deficiency). Lab Invest 70:130–133. 49. Zhou A, Stein PE, Huntington JA and Carrell RW (2003) Serpin polymerisation is prevented by a hydrogen bond network that is centered on His-334 and stabilized by glycereol. J Biol Chem 278:15116–15122. 50. Kang HA, Lee KN and Yu M-H (1997) Folding and stability of the Z and Siiyama genetic variants of human α1 -antitrypsin. J Biol Chem 272: 510–516. 51. Lomas DA, Finch JT, Seyama K, Nukiwa T and Carrell RW (1993) α1 Antitrypsin Siiyama (Ser53 → Phe): further evidence for intracellular loopsheet polymerisation. J Biol Chem 268:15333–15335. 52. Lomas DA, Elliott PR, Sidhar SK, Foreman RC, Finch JT, Cox DWJCW and Carrell RW (1995) Alpha1 -antitrypsin Mmalton (52 Phe deleted) forms loop-sheet polymers in vivo: Evidence for the C sheet mechanism of polymerisation. J Biol Chem 270:16864–16870. 53. Elliott PR, Stein PE, Bilton D, Carrell RW and Lomas D (1996) A Structural explanation for the dysfunction of S α1 -antitrypsin. Nat Struct Biol 3: 910–911. 54. Mahadeva R, Chang W-SW, Dafforn T, Oakley DJ, Foreman RC, Calvin J, Wight D and Lomas DA (1999) Heteropolymerisation of S, I and Z α1 -antitrypsin and liver cirrhosis. J Clin Invest 103: 999–1006. 55. Cruz M, Molina JA, Pedrola D and Muñoz-López F (1975) Cirrhosis and heterozygous α1 -antitrypsin deficiency in a 4 year old girl. Helv Paediatr Acta 30:501–507. 56. Campra JL, Craig JR, Peters RL and Reynolds TB (1973) Cirrhosis associated with partial deficiency of alpha-1-antitrypsin in an adult. Ann Intern Med 78:233–238. 57. Osterwalder T, Contartese J, Stoeckli ET, Kuhn TB and Sonderegger P (1996) Neuroserpin, an axonally secreted serine protease inhibitor. EMBO J 15:2944–2953.

ch20

Jan. 10, 2007

8:57

SPI-B429



471

58. Osterwalder T, Cinelli P, Baici A, Pennella A, Krueger SR, Schrimpf SP, Meins M and Sonderegger P (1998) The axonally secreted serine proteinase inhibitor, neuroserpin, inhibits plasminogen activators and plasmin but not thrombin. J Biol Chem 237:2312–2321. 59. Hastings GA, Coleman TA, Haudenschild CC, Stefansson S, Smith EP, Barthlow R, Cherry S, Sandkvist M and Lawrence DA (1997) Neuroserpin, a brain-associated inhibitor of tissue plasminogen activator is localized primarily in neurones. J Biol Chem 272:33062–33067. 60. Belorgey D, Crowther DC, Mahadeva R and Lomas DA (2002) Mutant neuroserpin (Ser49Pro) that causes the familial dementia FENIB is a poor proteinase inhibitor and readily forms polymers in vitro. J Biol Chem 277: 17367–17373. 61. Barker-Carlson K, Lawrence DA and Schwartz BS (2002) Acyl-enzyme complexes between tissue type plasminogen activator and neuroserpin are short lived in vitro. J Biol Chem 277:46852–46857 62. Parmar PK, Coates LC, Pearson JF, Hill RM and Birch NP (2002) Neuroserpin regulates neurite outgrowth in nerve growth factor treated PC12 cells. J Neurochem 82:1406–1415. 63. Hill RM, Coates LC, Parmar PK, Mezey E, Pearson JF and Birch NP (2002) Expression and functional characterization of the serine protease inhibitor neuroserpin in endocrine cells. Ann NY Acad Sci 971: 406–415. 64. Yepes M, Sandkvist M, Wong MKK, Coleman TA, Smith E, Cohan SL and Lawrence DA (2000) Neuroserpin reduces cerebral infarct volume and protects neurones from ischaemia-induced apoptosis. Blood 96: 569–576. 65. Yepes M, Sandkvist M, Coleman TA, Moore E, Wu J-Y, Mitola D, Bugge TH and Lawrence DA (2002) Regulation of seizure spreading by neuroserpin and tissue-type plasminogen activator is plasminogen independent. J Clin Invest 109:1571–1578. 66. Madani R, Kozlov S, Akhmedov A, Cinelli P, kinter J, Lipp H-P, Sonderegger P and Wolfer DP (2003) Impaired explorative behavior and neophobia in genetically modified mice lacking or overexpressing the extracellular serine proteinase inhibitor neuroserpin. Mol Cell Neurosci 23:473–494. 67. Davis RL, Shrimpton AE, Holohan PD, Bradshaw C, Feiglin D, Sonderegger P, Kinter J, Becker LM, Lacbawan F, Krasnewich D, Muenke M, Lawrence DA,Yerby MS, Shaw C-M, Gooptu B, Elliott PR, Finch JT, Carrell RW and Lomas DA (1999) Familial dementia caused by polymerisation of mutant neuroserpin. Nature 401:376–379. 68. Davis RL, Holohan PD, Shrimpton AE, Tatum A, Daucher J, Collins GH, Todd R, Bradshaw C, Kent P, Feiglin D, Rosenbaum A, Yerby MS,

ch20

Jan. 10, 2007

8:57

SPI-B429

472

69.

70.

71.

72.

73.

74.

75.

76. 77.

78.

79.


D. A. Lomas et al.

Shaw C-M, Lacbawan F and Lawrence DA (1999) Familial encephalopathy with neuroserpin inclusion bodies (FENIB). Am J Pathol 155:1901–1913. Bradshaw CB, Davis RL, Shrimpton AE, Holohan PD, Rea CB, Fieglin D, Kent P and Collins GH (2001) Cognitive deficits associated with a recently reported familial neurodegenerative disease. Arch Neurol 58:1429–1434. Davis RL, Shrimpton AE, Carrell RW, Lomas DA, Gerhard L, Baumann B, Lawrence DA,Yepes M, Kim TS, Ghetti B, Piccardo P, Takao M, Lacbawan L, Muenke M, Sifers RN, Bradshaw CB, Kent PF, Collins GH, Larocca D and Holohan PD (2002) Association between conformational mutations in neuroserpin and onset and severity of dementia. Lancet 359:2242–2247. Belorgey D, Sharp LK, Crowther DC, Onda M, Johansson J and Lomas DA (2004) Neuroserpin Portland (Ser52Arg) is trapped as an inactive intermediate that rapidly forms polymers: implications for the epilepsy seen in the dementia FENIB. Eur J Biochem 271:3360–3367. Onda M, Belorgey D, Sharp LK and Lomas DA (2005) Latent S49P neuroserpin spontaneously forms polymers: identification of a novel pathway of polymerization and implications for the dementia FENIB. J Biol Chem 280:13735–13741. Miranda E, Römisch K and Lomas DA (2004) Mutants of neuroserpin that cause dementia accumulate as polymers within the endoplasmic reticulum. J Biol Chem 279:28283–28291. Lawrence DA, Strandberg L, Ericson J and Ny T (1990) Structure-function studies of the SERPIN plasminogen activator type 1. J Biol Chem 265: 20293–20301. Bruce D, Perry DJ, Borg J-Y, Carrell RW and Wardell MR (1994) Thromboembolic disease due to thermolabile conformational changes of antithrombin Rouen VI (187 Asn → Asp). J Clin Invest 94: 2265–2274. Zhou A, Huntington JA and Carrell RW (1999) Formation of the antithrombin heterodimer and the onset of thrombosis. Blood 94:3388–3396. Chang W-SW and Lomas DA (1998) Latent α1 -antichymotrypsin: A molecular explanation for the inactivation of alpha-1-antichymotrypsin in chronic bronchitis and emphysema. J Biol Chem 273:3695–3701. Lomas DA, Elliott PR, Chang W-SW, Wardell MR and Carrell RW (1995) Preparation and characterisation of latent α1 -antitrypsin. J Biol Chem 270:5282–5288. Wardell MR, Chang W-SW, Bruce D, Skinner R, Lesk A and Carrell RW (1997) Preparative induction and characterization of L-antithrombin: A structural homologue of latent plasminogen activator inhibitor-1. Biochemistry 36:13133–13142.

ch20

Jan. 10, 2007

8:57

SPI-B429



473

80. Hekman CM and Loskutoff DJ (1985) Endothelial cells produce a latent inhibitor of plasminogen activators that can be activated by denaturants. J Biol Chem 260:11581–11587. 81. Takao M, Murrell JR, Benson MD,Yazaki M, Piccardo P, Davis R, Holhan PD, Lawrence DA, Farlow MR and Ghetti B (2000) Neuroserpin S52R mutation causes progressive myoclonus epilepsy and a wide distribution of intraneuronal neuroserpin deposits in the CNS. Brain Pathol 4:597. 82. Carrell RW and Lomas DA (1997) Conformational diseases. Lancet 350: 134–138. 83. Sidhar SK, Lomas DA, Carrell RW and Foreman RC (1995) Mutations which impede loop/sheet polymerisation enhance the secretion of human α1 -antitrypsin deficiency variants. J Biol Chem 270:8393–8396. 84. Graham KS, Le A and Sifers RN (1990) Accumulation of the insoluble PiZ variant of human α1 -antitrypsin within the hepatic endoplasmic reticulum does not elevate the steady-state level of grp78/BiP. J Biol Chem 265:20463– 20468. 85. Lawless MW, Greene CM, Mulgrew A, Taggert CC, O’Neill SJ and McElvaney NG (2004) Activation of endoplasmic reticulum-specific stress responses associated with the conformational disease Z α1 -antitrypsin deficiency. J Immunol 172:5722–5726. 86. Aulak KS, Eldering E, Hack CE, Lubbers YPT, Harrison RA, Mast A, Cicardi M and Davis III AE (1993) A hinge region mutation in C1-inhibitor (Ala436 → Thr) results in nonsubstrate-like behavior and in polymerization of the molecule. J Biol Chem 268: 18088–18094. 87. Eldering E, Verpy E, Roem D, Meo T and Tosi M (1995) COOH-terminal substitutions in the serpin C1 inhibitor that cause loop overinsertion and subsequent multimerization. J Biol Chem 270:2579–2587. 88. Picard V, Dautzenberg M-D, Villoutreix BO, Orliaguet G, Alhenc-Gelas M and Aiach M (2003) Antithrombin Phe229Leu: A new homozygous variant leading to spontaneous antithrombin polymerisation in vivo associated with severe childhood thrombosis. Blood 102:919–925. 89. Faber J-P, Poller W, Olek K, Baumann U, Carlson J, Lindmark B and Eriksson S (1993) The molecular basis of α1 -antichymotrypsin deficiency in a heterozygote with liver and lung disease. J Hepatology 18: 313–321. 90. Poller W, Faber J-P, Weidinger S, Tief K, Scholz S, Fischer M, Olek K, Kirchgesser M and Heidtmann H-H (1993) A leucine-to-proline substitution causes a defective α1 -antichymotrypsin allele associated with familial obstructive lung disease Genomics 17:740–743.

ch20

Jan. 10, 2007

8:57

SPI-B429

474


D. A. Lomas et al.

91. Stein PE and Carrell RW (1995) What do dysfunctional serpins tell us about molecular mobility and disease? Nat Struct Biol 2:96–113. 92. Corral J, Aznar J, Gonzalez-Conejero R, Villa P, Minano A, Vaya A, Carrell RW, Huntington JA, and Vicente V (2004) Homozygous deficiency of heparin cofactor II: relevance of P17 glutamate residue in serpins, relationship with conformational diseases, and role in thrombosis. Circulation 110: 1303–1307. 93. Green C, Brown G, Dafforn TR, Morley T, Lomas DA and Gubb D (2003) Mutations in the Drosophila serpin Necrotic mirror disease-associated mutations of human serpins. Development 130:1473–1478. 94. Eriksson S (1965) Studies in α1 -antitrypsin deficiency. Acta Med Scand suppl.432:1–85. 95. Stein PD, Leu JD, Welsh MH and Guenter CA (1971) Pathophysiology of the pulmonary circulation in emphysema associated with alpha-1 antitrypsin deficiency. Circulation 43:227–239. 96. Seersholm N and Kok-Jensen A (1998) Clinical features and prognosis of life time non-smokers with severe α1 -antitrypsin deficiency. Thorax 53:265–268. 97. Dirksen A, Holstein-Rathlou N-H, Madsen F, Skovgaard LT, Ulrik CS, Heckscher T and Kok-Jensen A (1998) Long-range correlations of serial FEV1 measurements in emphysematous patients and normal subjects. J Appl Physiol 85:259–265. 98. DirksenA, Friis M, Olesen KP, Skovgaard LT and Sørensen K (1997) Progress of emphysema in severe α1 -antitrypsin deficiency as assessed by annual CT. Acta Radiologica 38:826–832. 99. Dowson LJ, Guest PJ, Hill SL, Holder RL and Stockley RA (2001) Highresolution computed tomography scanning in alpha1-antitrypsin deficiency: Relationship to lung function and health status. Eur Respir J 17:1097–2104. 100. Colp C, Pappas J, Moran D and Lieberman J (1993) Variants of α1 -antitrypsin in Puerto Rican children with asthma. Chest 103:812–815. 101. Eden E, Mitchell DBM, Khouli H, Nejat M, Grieco MH and Turino GM (1997) Atopy, asthma, and emphysema in patients with severe alpha-1antitrypsin deficiency. Am J Resp. Crit Care Med 156:68–74. 102. Griffith ME, Lovegrove JU, Gaskin G, Whitehouse DB and Pusey CD (1996) C-antineutrophil cytoplasmic antibody positivity in vasculitis patients is associated with the Z allele of alpha-1-antitrypsin, and P-antineutrophil cytoplasmic antibody positivity with the S allele. Nephrol Dial Transplant 11: 438–443. 103. Baslund B, Szpirt W, Eriksson S, Elzouki A-N, Wiik A, Wieslander J and Petersen J (1996) Complexes between proteinase 3 α1 -antitrypsin and proteinase 3 anti-neutrophil cytoplasmic autoantibodies: A comparison between

ch20

Jan. 10, 2007

8:57

SPI-B429



104.

105.

106.

107.

108.

109. 110. 111.

112. 113.

114.

115.

116.

475

alpha-1-antitrypsin PiZ allele carriers and non-carriers with Wegener’s granulomatosis. Eur J Clin Invest 26:786–792. King MA, Stone JA, Diaz PT, Mueller CF, Becker WJ and Gadek JE (1996) α1 -Antitrypsin deficiency: evaluation of bronchiectasis with CT. Radiology 199:137–141. Rodriguez-Cintron W, Guntupalli K and Fraire AE (1995) Bronchiectasis and homozygous (PiZZ) α1 -antitrypsin deficiency in a young man. Thorax 50:424–425. Cuvelier A, Muir J-F, Hellot M-F, Benhamou D, Martin J-P, Bénichou J and Sesboüé R (2000) Distribution of α1 -antitrypsin alleles in patients with bronchiectasis Chest 117: 415–419. Warter J, Storck D, Grosshans E, Metais P, Kuntz J-L and Klumpp T (1972) Syndrome de Weber-Christian associe a un deficit en alpha-1-antitrypsine; enquete familiale. Ann Med Interne (Paris) 123: 877–882. O’Riordan K, Blei A, Rao MS and Abecassis M (1997) α1 Antitrypsin deficiency-associated panniculitis Resolution with intravenous α1 -antitrypsin administration and liver transplantation. Transplantation 63: 480–482. Smith KC, Pittelkow MR and Su, WPD (1987) Panniculitis associated with severe α-1-antitrypsin deficiency. Arch Dermatol 123:1655–1661. Seersholm N and Kok-Jensen A (2001) Extrapulmonary manifestations of α1 -antitrypsin deficiency. Am J Resp Crit Care Med 163:A343. Schievink WI, Prakash UBS, Piepgras DG and Mokri B (1994) α1 -Antitrypsin deficiency in intracranial aneurysms and cervical artery dissection. Lancet 343:452–453. Cox DW (1994) α1 -antitrypsin: A guardian of vascular tissue. Mayo Clin Proc 69:1123–1124. Wewers MD, Casolaro MA, Sellers SE, Swayze SC, McPhaul KM, Wittes JT and Crystal RG (1987) Replacement therapy for alpha1 -Antitrypsin deficiency associated with emphysema. N Engl J Med 316:1055–1062. Ogushi F, Fells GA, Hubbard RC, Straus SD and Crystal RG (1987) Z-type α1 -antitrypsin is less competent than M1-type α1 -antitrypsin as an inhibitor of neutrophil elastase. J Clin Invest 80:1366–1374. Guzdek A, Potempa J, Dubin A and Travis J, (1990) Comparative properties of human α1 -proteinase inhibitor glycosylation variants. FEBS Lett 272: 125–127. Lomas DA, Evans DL, Stone SR, Chang W-SW and Carrell RW (1993) Effect of the Z mutation on the physical and inhibitory properties of α1 -antitrypsin. Biochemistry 32:500–508.

ch20

Jan. 10, 2007

8:57

SPI-B429

476


D. A. Lomas et al.

117. Llewellyn-Jones CG, Lomas DA, Carrell RW and Stockley RA (1994) The effect of the Z mutation on the ability of α1 -antitrypsin to prevent neutrophil mediated tissue damage. Biochim Biophys Acta 1227:155–160. 118. Larsson C (1978) Natural history and life expectancy in severe alpha1 antitrypsin deficiency, PiZ. Acta Med Scand 204:345–351. 119. Janus ED, Phillips NT and Carrell RW (1985) Smoking, lung function, and α1 -antitrypsin deficiency. Lancet i:152–154. 120. Elliott PR, Bilton D and Lomas DA (1998) Lung polymers in Z α1 -antitrypsin related emphysema. Am J Resp Cell Mol Biol 18: 670–674. 121. Parmar JS, Mahadeva R, Reed BJ, Farahi N, Cadwallader K, Bilton D, Chilvers ER and Lomas DA (2002) Polymers of α1 -antitrypsin are chemotactic for human neutrophils: A new paradigm for the pathogenesis of emphysema. Am J Resp. Cell Mol Biol 26:723–730. 122. Mulgrew AT, Taggart CC, Lawless MW, Greene CM, Brantly ML, O’Neill SJ and McElvaney NG (2004) Z α1 -antitrypsin polymerizes in the lung and acts as a neutrophil chemoattractant. Chest 125:1952–1957. 123. Janciauskiene S, Zelvyte I, Jansson L and Stevens T (2004) Divergent effects of α1 -antitrypsin on neutrophil activation, in vitro. Biochem Biophys Res Commun 315:288–296. 124. Mahadeva R, Atkinson C, Li J, Stewart S, Janciauskiene S, Kelley DG, Parmar J, Pitman R, Shapiro SD and Lomas DA (2005) Polymers of Z α1 antitrypsin co-localise with neutrophils in emphysematous alveoli and are chemotactic in vivo. Am J Pathol 166:377–386. 125. Morrison HM, Kramps JA, Burnett D and Stockley RA (1987) Lung lavage fluid from patients with α1 -proteinase inhibitor deficiency or chronic obstructive bronchitis: anti-elastase function and cell profile. Clin Sci 72:373–381. 126. Rouhani F, Paone G, Smith NK, Krein P, Barnes P and Brantly ML (2000) Lung neutrophil burden correlates with increased pro-inflammatory cytokines and decreased lung function in individuals with α1-antitrypsin deficiency. Chest 250S-251S. 127. Banda MJ, Rice AG, Griffin GL and Senior RM (1998) The inhibitory complex of human α1 -proteinase inhibitor and human leukocyte elastase is a neutrophil chemoattractant. J Exp Med 167: 1608–1615. 128. Senior RM, Griffin GL and Mecham RP (1980) Chemotactic activity of elastin derived peptides. J Clin Invest 66:859–862. 129. Golpon HA, Coldren CD, Zamora MR, Cosgrove GP, Moore MD, Tuder RM, Geraci MW and Voelkel NF (2004) Emphysema lung tissue gene expression profiling. Am J Respir Cell Mol Biol 31:595–600.

ch20

Jan. 10, 2007

8:57

SPI-B429



477

130. Bleumink E and Klokke AH (1985) Relationship between Weber-Christian panniculitis and the ZZ phenotype of alpha1 -antitrypsin. Arch Dermatol Res 277:328–329. 131. Sigsgaard T, Brandslund I, Omland O, Hjort C, Lund ED, Pedersen OF and Miller MR (2000) S and Z α1 -antitrypsin alleles are risk factors for bronchial hyperresponsiveness in young farmers an example of gene/environment interaction. Eur Resp. J 16:50–55. 132. Kurdowska A and Travis J (1990) Acute phase protein stimulation by α1 -antichymotrypsin-cathepsin G complexes. Evidence for the involvement of interleukin-6. J Biol Chem 265:21023–21026. 133. Cichy J, Potempa J, Chawla RK and Travis J (1995) Stimulatory effect of inflammatory cytokines on α1 -antichymotrypsin expression in human lungderived epithelial cells. J Clin Invest 95: 2729–2733. 134. Rubin H, Wang Z, Nickbarg EB, McLarney S, Naidoo N, Schoenberger OL, Johnson JL and Cooperman BS (1990) Cloning, expression purification, and biological activity of recombinant native and variant human α1 antichymotrypsins. J Biol Chem 265:1199–1207. 135. Lomas DA, Stone SR, Llewelyn-Jones C, Keogan M-T, Wang ZM, Rubin H, Carrell RW and Stockley RA (1995) The control of neutrophil chemotaxis by inhibitors of cathepsin G and chymotrypsin. J Biol Chem 270:23437–23443. 136. Schechter NM, Jordan LM, James AM, Cooperman BS, Wang ZM and Rubin H (1993) Reaction of human chymase with reactive site variants of α1 -antichymotrypsin. Modulation of inhibitor versus substrate properties. J Biol Chem 268:23626–23633. 137. Billingsley GD, Walter MA, Hammond GL and Cox DW (1993) Physical mapping of four serpin genes: α1 -antitrypsin α1 -antichymotrypsin, corticosteroid-binding globulin, and protein C inhibitor, within a 280 kb region on chromosome 14q311. Am J Hum Genet 52:343–353. 138. Eriksson S, Lindmark B and Lilja H (1986) Familial α1 -antichymotrypsin deficiency. Acta Med Scand 220:447–453. 139. Lindmark BE, Arborelius Jr M and Eriksson SG (1990) Pulmonary function in middle-aged women with heterozygous deficiency of the serine protease inhibitor alpha1 -antichymotrypsin. Am Rev Respir Dis 141:884–888. 140. Poller W, Faber J-P, Scholz S, Weidinger S, Bartholome K, Olek K and Eriksson S (1992) Mis-sense mutation of α1 -antichymotrypsin gene associated with chronic lung disease. Lancet 339:1538. 141. Tsuda M, Sei Y, Yamamura M, Yamamoto M and Shinohara Y (1992) Detection of a new mutant α1 -antichymotrypsin in patients with occlusivecerebrovascular disease. FEBS Lett 304:66–68.

ch20

Jan. 10, 2007

8:57

SPI-B429

478


D. A. Lomas et al.

142. Tsuda M, Sei Y, Ohkubo T, Yamamura M, Kamiguchi H, Akatsuka A, Tsuda T, Tachikawa H, Yamamoto M and Shinohara Y (1996) The defective secretion of a naturally occurring α-1-antichymotrypsin variant with a frameshift mutation. Eur J Biochem 235:821–827. 143. Sandford AJ, Chagani T, Weir TD and Paré PD (1998) α1 -Antichymotrypsin mutations in patients with chronic obstructive pulmonary disease. Dis Markers 13:257–260. 144. Benetazzo MG, Gile LS, Bombieri C, Malerba G, Massobrio M, Pignatti PF and Luisetti M (1999) α1 -Antitrypsin TAQ I polymorphism and α1 antichymotrypsin mutations in patients with obstructive pulmonary disease. Resp Med 93:648–654. 145. Berman G, Afford SC, Burnett D and Stockley, RA (1986) α-1Antichymotrypsin in lung secretions is not an effective proteinase inhibitor. J Biol Chem 261:14095–14099. 146. Kamboh MI, Sanghera DK, Ferrell RE and DeKosky ST (1995) APOE*4associated Alzheimer’s disease risk is modified by α1-antichymotrypsin polymorphism. Nature Genet 10:486–488. 147. Christie J, Lamb H, Singleton AB, Leake A, Melton LM, Perry RH, McKeith IG and Morris CM (1996) Determination of the alpha-1 antichymotrypsin polymorphism in Alzheimer’s disease. Alzheimer’s Res 2: 201–204. 148. Müller U, Bödeker R-H, Gerundt I, and Kurz A (1996) Lack of association between α1-antichymotrypsin polymorphism,Alzheimer’s disease, and allele ε4 of apolipoprotein E. Am Acad Neurol 47:1575–1577. 149. Haines JL, Pritchard ML, Saunders AM, Schildkraut JM, Growdon JH, Gaskell PC, Farrer LA, Auerbach SA, Gusella JF, Locke PA, Rosi BL, Yamaoka L, Small GW, Conneally PM, Roses AD and Pericak-Vance MA (1996) No genetic effect of α1 -antichymotrypsin in Alzheimher disease. Genomics 33:53–56. 150. Haines JL, Pritchard ML, Saunders AM, Schildkraut JM, Growdon JH, Gaskell PC, Farrer LA, Auerbach SA, Gusella JF, Locke PA, Rosi BL, Yamaoka L, Small GW, Conneally PM, Roses AD and PericakVance M (1996) No association between α1-antichymotrypsin and familial Alzheimer’s disease. Ann NY Acad Sci 802:35–41. 151. Murphy Jr GM, Sullivan EV, Gallagher-Thompson D, Thompson LW, van Duijn CM, Forno LS, Ellis WG, Jagust WJ, Yesavage J and Tinklenberg JR (1997) No association between the alpha 1-antichymotrypsin A allele and Alzheimer’s disease. Neurology 19:1313–1316.

ch20

Jan. 10, 2007

8:57

SPI-B429



479

152. Ishii T, Matsuse T, Teramoto S, Matsui H, Hosoi T, Fukuchi, Y and Ouchi Y (2000) Association between alpha-1-antichymotrypsin polymorphism and susceptibility to chronic obstructive pulmonary disease. Eur J Clin Invest 30: 543–548. 153. Nemeth A (1999) Liver transplantation in alpha-1-antitrypsin deficeincy.Eur J Pediatr 158 (Suppl 2):S85–S88. 154. Francavilla R, Castellaneta SP, Hadzic N, Chambers SM, Portmann B, Tung J, Cheeseman P, Rela M, Heaton ND and Mieli-Vergani G (2000) Prognosis of alpha-1-antitrypsin deficiency-related liver disease in the era of paediatric liver transplantation. J Hepatol 32: 986–992. 155. Prachalias AA, Kalife M, Francavilla R, Muiesan P, Dhawan, A, Baker A, Hadzic D, Mieli-Vergani G, Rela M and Heaton ND (2000) Liver transplantation for alpha-1-antitrypsin deficiency. Transpl Int 13:207–210. 156. Piitulainen E and Eriksson S (1999) Decline in FEV1 related to smoking status in individuals with severe alpha1-antitrypsin deficiency. Eur Respir J 13:247–251. 157. British Thoracic Society (1998) Smoking cessation guidelines and their cost effectiveness Thorax 53 (Suppl 5). 158. British Thoracic Society (1997) BTS guidelines for the management of chronic obstructive pulmonary disease, Thorax 52 (Suppl 5). 159. Lacasse Y, Wong E, Guyatt GH, King D, Cook DJ and Goldstein RS (1996) Meta-analysis of respiratory rehabilitation in chronic obstructive pulmonary disease. Lancet 348:1115–1119. 160. Trulock 3rd EP (1998) Lung transplantation for COPD. Chest 113 (Suppl 4):269S–276S. 161. King MB, Campbell EJ, Gray BH and Hertz MI (1994) The proteinaseantiproteinase balance in α-1-proteinase inhibitor-deficient lung transplant recipients. Am J Respir Crit Care Med 149:966–971. 162. Cassina PC, Teschler H, Konietzko N, Theegarten D and Stamatis G (1998) Two-year results after lung volume reduction surgery in α1 -antitrypsin deficiency versus smoker’s emphysema. Eur Respir J 12:1028–1032. 163. Gelb AF, McKenna RJ, Brenner M, Fischel R and Zamel N (1999) Lung function after bilateral lower lobe lung volume reduction surgery for α1 antitrypsin deficiency. Eur Respir J 14:928–933. 164. Berger RL, Celli BR, Meneghetti AL, Bagley PH, Wright CD, Ingenito EP, Gray A and Snider GL (2001) Limitations of randomized clinical trials for evaluating emerging operations: The case of lung volume reduction surgery. Ann Thorac Surg 72:649–657.

ch20

Jan. 10, 2007

8:57

SPI-B429

480


D. A. Lomas et al.

165. The Alpha-1-Antitrypsin Deficiency Registry Study Group (1998) Survival and FEV1 decline in individuals with severe deficiency of α1 -antitrypsin. Am J Resp Crit Care Med 158:49–59. 166. Dirksen A, Dijkman JH, Madsen F, Stoel B, Hutchison DCS, Ulrik CS, Skovgaard LT, Kok-Jensen A, Rudolphus A, Seersholm N, Vrooman HA, Reiber JHC, Hansen NC, Heckscher T, Viskum K and Stolk J (1999) A randomised clinical trial of α1 -antitrypsin augmentation therapy. Am J Respir Crit Care Med 160:1468–1472. 167. Schluchter MD, Stoller JK, Barker AF, Buist AS, Crystal RG, Donohue JF, Fallat RJ, Turino GM, Vreim CE and Wu MC (2000) Feasibility of a clinical trial of augmentation therapy for alpha-1-antitrypsin deficiency The alpha-1antitrypsin deficiency study group. Am J Respir Crit Care Med 161:796–801. 168. Song S, Embury J, Laipis PJ, Berns KI, Crawford JM and Flotte TR (2001) Stable therapeutic serum levels of human alpha-1 antitrypsin (AAT) after portal vein injection of recombinant adeno-associated virus (rAAV) vectors. Gene Ther 8:1299–1306. 169. Rosenfeld MA, Siegfried W, Yoshimura K, Yoneyama K, Fukayama M, Stier LE, Pääkkö PK, Gilardi P, Stratford-Perricaudet LD, Perricaudet M, Jallat S, Pavirani A, Lecocq JP and Crystal RG (1991) Adenovirus-mediated transfer of a recombinant α-antitrypsin gene to the lung epithelium in vivo. Science 252:431–434. 170. Song S, Morgan M, Ellis T, Poirier A, Chesnut K, Wang J, Brantly M, Muzyczka N, Byrne BJ,Atkinson M and Flotte TR (1998) Sustained secretion of human alpha-1-antitrypsin from murine muscle transduced with adenoassociated virus vectors. Proc Natl Acad Sci USA 95:14384–14388. 171. Bou-Gharios G, Wells DJ, Lu QL, Morgan JE and Partridge T (1999) Differential expression and secretion of alpha 1 anti-trypsin between direct DNA injection and implantation of transfected myoblast. Gene Ther 6: 1021–1029. 172. Massaro GD and Massaro D (1997) Retinoic acid treatment abrogates elastase-induced pulmonary emphysema in rats. Nature Med 3:675–677. 173. Schulze AJ, Baumann U, Knof S, Jaeger E, Huber R and Laurell C-B (1990) Structural transition of α1 -antitrypsin by a peptide sequentially similar to β-strand s4A. Eur J Biochem 194:51–56. 174. Mast AE, Enghild JJ and Salvesen G (1992) Conformation of the reactive site loop of α1 -proteinase inhibitor probed by limited proteolysis. Biochemistry 31:2720–2728. 175. Huntington JA, Pannu NS, Hazes B, Read R, Lomas DA and Carrell RWA (1999) 26Å structure of a serpin polymer and implications for conformational disease. J Mol Biol 293:449–455.

ch20

Jan. 10, 2007

8:57

SPI-B429



481

176. Dunstone MA, Dai W, Whisstock JC, Rossjohn J, Pike RN, Feil SC, Le Bonniec BF, Parker MW and Bottomley SP (2000) Cleaved antitrypsin polymers at atomic resolution. Protein Sci 9:417–420. 177. Chow MKM, Devlin GL and Bottomley SP (2001) Osmolytes as modulators of conformational changes in the serpins. Biol Chem 382: 1593–1599. 178. Burrows JAJ, Willis LK and Perlmutter DH (2000) Chemical chaperones mediate increased secretion of mutant α1 -antitrypsin (α1 -AT) Z: A potential pharmacologcial strategy for prevention of liver injury and emphysema. Proc Natl Acad Sci USA 97:1796–1801. 179. Devlin GL, Parfrey H, Tew DJ, Lomas DA and Bottomley SP (2001) Prevention of polymerization of M and Z α1 -antitrypsin (α1 -AT) with trimethylamine N-oxide Implications for the treatment of α1 -AT deficiency. Am J Resp Cell Mol Biol 24:727–732. 180. Rubenstein RC, Egan ME and Zeitlin PL (1997) In vitro pharmacologic restoration of CFTR-mediated chloride transport with sodium 4-phenylbutyrate in cystic fibrosis epithelial cells containing delta F508-CFTR. J Clin Invest 100:2457–2465. 181. Rubenstein RC and Zeitlin PL (1998) A pilot clinical trial of oral sodium 4-phenylbutyrate (Buphenyl) in deltaF508-homozygous cystic fibrosis patients: partial restoration of nasal epithelial CFTR function. Am J Respir Crit Care Med 157:484–490. 182. Teckman JH (2004) Lack of effect of oral 4-phenylbutyrate on serum alpha-1antitrypsin in patients with alpha-1-antitrypsin deficiency: A preliminary study. J Pediatr Gastroenterol Nutr 39:34–37. 183. Devlin GL, Carver JA and Bottomley SP (2003) The selective inhibition of serpin aggregation by the molecular chaperone, alpha-crystallin, indicates a nucleation-dependent specificity. J Biol Chem 278:48644–48650. 184. Lee C, Maeng J-S, Kocher J-P, Lee B and Yu M-H (2001) Cavities of α1 antitrypsin that play structural and functional roles. Prot Sci 10:1446–1453. 185. Parfrey H, Mahadeva R, Ravenhill N, Zhou A, Dafforn TR, Foreman RC and Lomas DA (2003) Targeting a surface cavity of α1 -antitrypsin to prevent conformational disease. J Biol Chem 278:33060–33066. 186. Egelund R, Einholm AP, Nielsen RW, Pedersen KE, Christensen A and Andreasen PA (1998) Probing and Modulating the Conformation of the Flexible Joints Region of Plasminogen Activator Inhibitor-1 With a Fluorescent Ligand, in Proceeding of the 2nd International Symposium on the Structure and Biology of Serpins, Cambridge, UK. 187. Egelund R, Einholm AP, Pedersen KE, Nielsen RW, Christensen A, Deinum J and Andreasen PA (2001) A regulatory hydrophobic area in the flexible joint

ch20

Jan. 10, 2007

8:57

SPI-B429

482

188.

189.

190.

191.

192.

193.

194.

195. 196. 197.

198.


D. A. Lomas et al.

region of plasminogen activator inhibitor-1, defined with fluorescent activityneutralizing ligands. Ligand-induced serpin polymerization. J Biol Chem 276:13077–13086. Pedersen K, Einholm AP, Chritensen A, Schack L, Wind T, Kenney JM and Andreasen PA (2003) Plasminogen activator inhibitor-1 polymers, induced by inactivating amphipathic organochemical ligands. Biochem J 372:747–755. Skinner R, Chang W-SW, Jin L, Pei X, Huntington JA, Abrahams J-P, Carrell RW and Lomas DA (1998) Implications for function and therapy of a 29Å structure of binary-complexed antithrombin. J Mol Biol 283:9–14. Fitton HL, Pike RN, Carrell RW and Chang W-SW (1997) Mechanisms of antithrombin polymerisation and heparin activation probed by insertion of synthetic reactive loop peptides. Biol Chem 378:1059–1063. Chang W-SW, Wardell MR, Lomas DA and Carrell RW (1996) Probing serpin reactive loop conformations by proteolytic cleavage. Biochem J 314: 647–653. Parfrey H, Dafforn TR, Belorgey D, Lomas DA and Mahadeva R (2004) Inhibiting polymerisation: New therapeutic strategies for Z α1 -antitrypsin related emphysema. Am J Respir Cell Mol Biol 31:133–139. Zhou A, Stein PE, Huntington JA, Sivasothy P, Lomas DA and Carrell RW (2004) How small peptides block and reverse serpin polymerization. J Mol Biol 342:931–941. Mottonen J, Strand A, Symersky J, Sweet RM, Danley DE, Geoghegan KF, Gerard RD and Goldsmith EJ (1992) Structural basis of latency in plasminogen activator inhibitor-1. Nature 355:270–273. Carrell RW, Stein PE, Fermi G and Wardell MR (1994) Biological implications of a 3Å structure of dimeric antithrombin. Structure 2:257–270. Im H, Woo M-S, Hwang KY and Yu M-H (2002) Interactions causing the kinetic trap in serpin protein folding. J Biol Chem 277:46347–46354. Lawrence DA, Olson ST, Palaniappan S and Ginsburg D (1994) Engineering plasminogen activator inhibitor 1 mutants with increased functional stability. Biochemistry 33:3643–3648. Lomas DA (2005) Molecular mousetraps and the serpinopathies. Biochem Soc Transac 33:321–330.

ch20

Jan. 10, 2007

8:57

SPI-B429


21 Liver Disease in α 1 -Antitrypsin Deficiency David H. Perlmutter

1. Introduction Alpha-1-antitrypsin (α1 -AT) deficiency is an autosomal co-dominant disorder associated with premature development of pulmonary emphysema, chronic liver disease and hepatocellular carcinoma. In the classical deficiency state, the so-called PIZ form, a point mutation results in a protein that is retained in the endoplasmic reticulum of liver cells rather than being secreted into the blood and body fluids. Loss-of-function permits uninhibited proteolytic destruction of the connective tissue matrix of the lung, ultimately leading to emphysema. Cigarette smoking exacerbates the development of emphysema because residual α1 -AT molecules are functionally inactivated by the increased load of active oxygen intermediates that are produced by smokers’pulmonary phagocytic cells. In contrast, liver disease results from a gain-of-toxic function mechanism whereby the retention of mutant α1 -ATZ in liver cells triggers a series of events that lead to liver injury and a predilection for hepatocellular carcinoma. Although this is the most common genetic cause of liver disease in children, only ∼10% of affected homozygotes develop clinically significant liver disease. This observation is the basis for the notion that genetic modifiers and environmental factors play a role in susceptibility to and/or protection from liver disease in the α1 -AT deficiency. Although there are several attractive new concepts for chemoprophylaxis and treatment of liver disease in this deficiency, the only effective treatment currently available is liver transplantation. 483

ch21

Jan. 10, 2007

8:57

SPI-B429


484

David H. Perlmutter

2. Epidemiology The incidence of the deficiency is ∼one in 2000 live births in most populations that have been carefully studied.1 Most of the studies of lung and liver disease in this population have been biased in ascertainment because they involve patients referred to specialty clinics. The only unbiased study comes from nationwide screening of all newborns in Sweden in the 1970s.2,3 Over 200,000 newborns were screened and 127 homozygotes for the α1 -ATZ allele were identified and most of them have now been followed for almost 30 years. The results show that only 14 of these individuals (11%) had prolonged obstructive jaundice in infancy and only 9 (7%) have developed clinically significant liver disease. Of the remaining α1 -AT-deficient population, 85% have had persistently normal transaminase levels as they have aged. Liver biopsies have not been done in this study and therefore it is not known whether some of these seemingly unaffected individuals have sub-clinical histological abnormalities and will develop clinical signs of liver disease as they reach the fourth and fifth decades of life subsequently. Although it is widely believed that lung disease affects a higher percentage of homozygotes,4 a valid determination of incidence will not be possible until the Swedish prospective cohort reaches the peak age range for emphysema, 40–60 years of age.

3. Pathogenesis of Liver Injury Liver injury in the classical form of α1 -AT deficiency appears to involve a gain-of-toxic function mechanism, whereby retention of the mutant α1 -ATZ molecule in the ER is hepatotoxic. The strongest evidence for the gainof-toxic function mechanism is the development of liver disease in mice transgenic for mutant human α1 -ATZ.5,6 As there are normal levels of endogenously derived anti-elastases in these mice, the liver injury cannot be attributed to a loss-of-function mechanism. To understand how the retention of mutant α1 -ATZ in the ER is toxic to liver cells, it is necessary to elucidate some of what is known as the biology of α1 -AT. As the archetype of the serine protease inhibitor (SERPIN) family, α1 -AT mainly functions as a blood-borne inhibitor of destructive neutrophil proteases, including elastase, cathepsin G, and proteinase 3.

ch21

Jan. 10, 2007

8:57

SPI-B429


Liver Disease in α1 -Antitrypsin Deficiency

485

It is predominantly derived from parenchymal liver cells, constituting the most abundant glycoprotein secreted by the liver. It is also considered a positive acute phase reactant, because its plasma concentration increases during the host response to inflammation/tissue injury. The classical deficiency mutant, α1 -ATZ, is characterized by a point mutation that results in the substitution of lysine for glutamate 342, and by itself, it accounts for defective secretion. The mutant α1 -ATZ molecule is retained in the ER of liver cells. Studies by Lomas and Carrell have shown that this substitution reduces the stability of α1 -ATZ as a monomer and increases the formation of polymers in vitro by a “loop-sheet” insertion mechanism.7 Moreover, their work showed the presence of polymers in liver cells by electron microscopic examination of a liver biopsy from a patient with the deficiency.8 Experiments with purified α1 -ATZ also showed that increases in temperature led to increased polymerization and, therein, that variations in the liver disease phenotype among deficient individuals could relate to variations in numbers or severity of intercurrent febrile illnesses.8 However, recent studies have shown that there is no increase in polymerization of α1 -ATZ in the liver of the PiZ mouse model exposed to increased body temperature,9 a result which is consistent with the in vivo supposition that increases in body temperature have many effects on biochemical processes and would most likely be complex. Several in vivo studies have suggested that polymerization is the cause of retention of α1 -ATZ in the ER of liver cells. The most powerful of these are studies showing partial correction of the secretory defect by insertion of a second mutation into the α1 -ATZ protein that suppresses loop-sheet polymerization.10−12 However, those studies do not exclude the possibility that there is an abnormality in folding that is distinct from the tendency to polymerize but is also partially corrected by the second, experimentally introduced mutation. More recent studies cast some doubt on the concept that polymerization is the cause of ER retention. Firstly, naturally occurring variants of α1 -AT, in which the carboxyl terminal tail is truncated, are retained in the ER of liver cells, even though they do not form polymers.13 Secondly, only ∼17% of the intracellular pool of α1 -ATZ at steady state is in the form of insoluble polymers in cell lines, which model the ER retention of this mutant.13,14 Moreover, because the remainder of α1 -ATZ in the ER in vivo is in heterogeneous soluble complexes with multiple ER

ch21

Jan. 10, 2007

8:57

SPI-B429

486


David H. Perlmutter

chaperones,14 the principles by which purified α1 -ATZ polymerizes in vitro are probably not applicable to what happens in live cells in vivo. Thus, existing data suggest that polymerization is not the cause by which α1 -ATZ is retained in the ER of liver cells, but rather the result of its retention. Nonetheless, the polymerogenic properties of the α1 -ATZ mutant are still likely to be critical determinants in the pathobiology of liver disease. An understanding of the pathways by which α1 -ATZ that is retained in the ER is degraded is also important in elucidating the gain-of toxic function mechanism and those pathways are obvious candidates for genetic variations that are associated with protection or susceptibility to liver disease in sub-groups of the deficient population. Studies by Wu et al. also provided a basis for this concept by showing that disposal of α1 -ATZ was delayed in genetically engineered cell lines from α1 -AT-deficient individuals who were homozygous for the Z allele and had clinically significant liver disease, “susceptible hosts,” when compared with homozygotes that had no evidence of liver disease, “protected hosts”.15 Subsequent investigations have shown that the degradation of α1 -ATZ is more complex than initially envisioned. Many studies have indicated that the proteosomal system is involved in ER degradation of α1 -ATZ in yeast and mammalian cells.16−22 This involvement appears to include both the classical ubiquitin-dependent proteosomal mechanism as well as a ubiquitin-independent proteosomal mechanism.19 However, it is not exactly clear how α1 -ATZ on the luminal side of the ER membrane is accessed by the proteosome from the cytoplasm. Although retrograde translocation from the ER to the cytoplasm has been demonstrated for some luminal substrates of the proteosome, there is very limited evidence for retrograde translocation of α1 -ATZ. Werner et al. detected α1 -ATZ in the cytosolic fraction of yeast when the proteosome was inhibited,17 but it was only a small fraction of the total α1 -ATZ in the ER and there has been no other evidence for retrotranslocation. This may mean that there are other mechanisms for transport from the ER to the cytoplasm, such as proteosome-mediated extraction through the ER membrane, as has been demonstrated for model ER degradation substrates.23 Moreover, there is evidence for non-proteosomal mechanisms for the degradation of α1 -ATZ.19 Cabral et al. have provided evidence for a nonproteosomal degradation pathway that is sensitive to tyrosine phosphatase

ch21

Jan. 10, 2007

8:57

SPI-B429



487

inhibitors.24 We have shown that autophagy contributes to the disposal of retained α1 -ATZ, using chemical inhibitors25 and cell lines genetically engineered for deficient autophagy.21 It is not yet known if there are still other mechanisms for ER degradation of α1 -ATZ or even if the different mechanisms that have already been identified contribute to the overall disposal pathway in specific ways. Interestingly, the studies of α1 -ATZ in autophagy-deficient cell lines21 as well as completely independent studies in yeast26,27 have suggested that autophagy may be particularly important for insoluble polymers/aggregates of α1 -ATZ. There is relatively limited information about how ER retention of mutant α1 -ATZ causes liver injury by a gain-of-toxic function mechanism. Recent studies have shown that liver from α1 -AT-deficient patients is characterized by significant mitochondrial injury, mitochondrial autophagy as well as activation of caspases-3 and -9.28,29 These mitochondrial changes and caspase activation were also seen in transfected cell line and transgenic model systems. Indeed, treatment of one transgenic mouse model with cyclosporine A, which inhibits the mitochondrial permeability transition pore (MPT), resulted in diminished hepatic mitochondrial injury, caspase activation, and improved survival.28 This may mean that ER retention of α1 -ATZ causes mitochondrial injury by a direct interaction between the distended ER and adjacent mitochondria. In fact, our most recent studies in model cell lines with inducible expression have shown that ER retention of α1 -ATZ activates BAP31,30 an ER protein that mediates direct interactions between ER and mitochondria.31 It is also possible that mitochondria are injured as innocent bystanders of an overexubant autophagic response that is activated by ER retention of α1 -ATZ. Autophagy is a ubiquitous, highly conserved cellular mechanism by which senescent and/or denatured constituents in the cytoplasm and intracellular organelles or whole organelles are sequestered from the rest of the cytoplasm within newly formed vacuoles that then fuse with lysosome for degradation. It is believed to be a mechanism for the turnover of cellular constituents during nutritional deprivation, stress states, morphogenesis, differentiation and aging (Fig. 1). Our previous studies have shown that ER retention of α1 -ATZ is a powerful stimulus for the autophagic response.27,28,31 A marked increase in autophagosomes has been observed in several different model cell lines genetically

ch21

Jan. 10, 2007

8:57

SPI-B429

488


David H. Perlmutter

Fig. 1. Autophagy as it occurs in the normal liver cell (upper panel) and might occur in the α1 -AT-deficient liver cell.

ch21

Jan. 10, 2007

8:57

SPI-B429



489

engineered to express α1 -ATZ, including human fibroblasts, murine hepatoma and rat hepatoma cell lines.27 Moreover, in a HeLa cell line engineered for inducible expression of α1 -ATZ, autophagosomes appear as a specific response to the expression of α1 -ATZ and its retention in the ER. There is a marked increase in autophagosomes in hepatocytes in transgenic mouse models of α1 -AT deficiency and a disease-specific increase in autophagosomes in liver biopsies from patients with the deficiency.25 Mutant α1 -ATZ molecules can be detected in autophagosomes by immune electron microscopy, often together with the ER chaperone calnexin. Intracellular degradation of α1 -ATZ is partially abrogated by chemical inhibitors of autophagy27 and when it is expressed in autophagy-deficient (atg5knockout) cell lines,21 indicating that autophagy also contributes to the quality control mechanism for the disposal of α1 -ATZ. Recently, we examined the autophagic response to ER retention of α1 -ATZ in vivo by testing the effect of fasting on the liver of the PiZ mouse model of α1 -AT deficiency.32 Starvation is a well-defined physiologic stimulus of autophagy, as well as a known environmental stressor of liver disease in children. The results showed that there is a marked increase in fat accumulation and in α1 -ATZ-containing, ER-derived globules in the liver of the PiZ mouse induced by fasting. These changes were particularly exaggerated at 3–6 months of age. Three-month-old PiZ mice had a significantly decreased tolerance for fasting when compared with that in non-transgenic C57/BL6 littermates. Although fasting induced a marked autophagic response in wild-type mice, the autophagic response was already activated in PiZ mice to levels that were more than 50% higher than those in the liver of fasted wild-type mice. In contrast to wild-type mice, there was no increase in autophagosomes in the liver of PiZ mice during fasting. These results indicate that autophagy is constitutively activated in the liver in α1 -AT deficiency and that the liver is unable to mount an increased autophagic response to physiologic stressors. Because variation in liver disease phenotype could also be attributed to variation in the cellular response pathways that are activated by accumulation of mutant α1 -ATZ in the ER, we have recently investigated how cells respond to this pathologic state using cell line and transgenic mouse model systems with inducible expression of the mutant gene. These models permit us to see the earliest responses and to separate them from compensatory

ch21

Jan. 10, 2007

8:57

SPI-B429

490


David H. Perlmutter

adaptations that permit cells or organisms to tolerate the accumulation of a mutant protein in the ER. Because the relative expression of the mutant gene can also be regulated in these systems, it is also possible to determine the effect of expressing the mutant gene product at specific concentrations, at specific stages of development, and for specific intervals on the cell, organ and organism. We were particularly interested in determining whether ER retention of α1 -ATZ in these systems elicited several very important signal transduction systems as exemplified by the unfolded protein response (UPR) and the ER overload pathway. The UPR is a signaling pathway activating a number of genes in response to the accumulation of unfolded proteins in the ER.33 In addition to new synthesis of ER chaperones such as BiP, and enzymes that facilitate disulfide bond formation, bolstering the protein-folding capacity of the ER, and lipids for synthesis of new ER membrane required to handle the increased protein load, an increase in synthesis of proteins that participate in degradative and other cellular translocation mechanisms occurs. There is also a decrease in initiation of translation in such a way that only specific mRNAs can be translated.34 To our surprise, accumulation of α1 -ATZ in the ER does not activate the UPR under any circumstances.30 We still do not know precisely why the UPR is not activated by ER retention of α1 -ATZ, but we believe it has to do with the specific intrinsic properties of this mutant because ER retention of other non-polymerogenic mutants α1 -AT Saar and α1 -AT SaarZ does activate the UPR.30 In contrast, the ER overload pathway was activated when α1 -ATZ accumulated in the ER in cell line models and in the liver of transgenic mouse models.30 This pathway activates the transcription factor NFκB in cells treated with brefeldin A, and in cells in which the adenovirus E3 accumulates in the ER.35 The fact that ER accumulation of α1 -ATZ activated NFκB but not the UPR provides strong corroborating evidence for the previously held contention that the ER overload pathway is distinct from the UPR.35 Activation of this pathway also has important implications for understanding target organ injury in α1 -AT-deficient patients. First, through NFκB activation, accumulation of α1 -ATZ in the ER of liver cells and respiratory epithelial cells36 could mediate the characteristic inflammation that occurs in the liver and lung of these patients, particularly neutrophil infiltration mediated by the NFκB target interleukin-8. Second, activation

ch21

Jan. 10, 2007

8:57

SPI-B429



491

of NFκB has been shown to play a key role in inflammation-associated carcinogenesis37,38 and therein could be involved in the mechanism of hepatocellular carcinoma in α1 -AT deficiency. We also examined two other signal transduction pathways that have been associated with ER stress. For one, we found that ER accumulation of α1 -ATZ led to cleavage/activation of the ER caspases,39,40 caspase-12 in mouse cells, and caspase-4 in human cells.30 These results indicate that accumulation of mutant α1 -ATZ in the ER results in the activation of the caspase cascade via both the ER and mitochondrial pathways. There is almost nothing known about the pathogenesis of hepatocellular carcinoma in α1 -AT deficiency. Results of recent studies by Rudnick et al.29 have suggested an interesting hypothesis. Using BrdU labeling in the PiZ transgenic mouse model of α1 -AT deficiency, these authors show that there is an increase in the proliferation of hepatocytes in the liver of these mice under resting conditions. Although the increase was 5 to 10-fold above the controls and was highly statistically significant, there were still a relatively low number of BrdU-positive hepatocytes at one time (2 to 3% detected over 72 h of continuous BrdU labeling). These data indicate that liver injury in the mouse model is relatively mild and nicely corresponds to the smoldering and slowly progressing liver disease that is seen in most α1 -ATdeficient patients. Interestingly, the increase in hepatocellular proliferation was entirely accounted for by male mice and correlated with increased numbers of hepatocytes that had PAS+, diastase-resistant globules as well as increased steady-state levels of human α1 -AT mRNA and protein. Systemic administration of testosterone to female PiZ mice led to an increase in the number of globule-containing hepatocytes, steady-state levels of α1 -AT mRNA and protein, and an increase in BrdU labeling that was comparable to that in male PiZ mice. These results were consistent with previous studies showing that androgens had a positive regulatory effect on the human α1 -AT gene — in this mouse that would represent a positive regulatory effect on the human transgene. However, more importantly, these results indicated that the increase in hepatocellular proliferation in the PiZ mouse liver was proportional to the number of globule-containing cells and/or the level of α1 -ATZ accumulation within these cells. These studies also demonstrated that almost all of the BrdU-positive cells in the PiZ mouse liver were globule-devoid, and conversely, very few

ch21

Jan. 10, 2007

8:57

SPI-B429

492


David H. Perlmutter

of the globule-containing cells were BrdU-positive. These results indicated that the globule-devoid hepatocytes had a selective proliferative advantage in the PiZ liver. To further characterize the selective proliferative advantage, the hepatocellular proliferative response to partial hepatectomy in the PiZ mice was determined. Although partial hepatectomy resulted in increased mortality among PiZ mice when compared with their C57/BL6 non-transgenic littermates, the PiZ mice that survived had a similar proliferative response to the C57 littermates. In particular, there was a comparable increase in BrdU labeling of globule-containing and globule-devoid hepatocytes. These data indicate that the block in proliferation of globule-containing cells is relative — i.e. when the stimulus is as powerful as the one that follows partial hepatectomy, those cells are able to replicate. Finally, these studies showed that there was increased caspase-9 and caspase-3 activation in the PiZ mouse liver. Together with the observation that the liver-to-body-weight ratio in PiZ mice was identical to that observed in wild-type mice, these data suggested that hepatocytes undergo programmed cell death at a rate equivalent to the rate of increased cellular proliferation. However, increased apoptosis was not detected either histologically or by TUNEL staining in the livers of these animals. These data indicate that the increase in cell death that must be occurring in the PiZ mouse liver is at a rate low enough to fall below the level of detection. Again, this is consistent with the slowly progressing nature of the disease. Most informative, however, is the lack of apoptosis histologically and by TUNEL staining of cells that obviously have globules. If we take this together with our data on how cells respond to accumulation of α1 -ATZ, we would suggest that the globule-containing cells are “sick but not dead” — they have activated ER- and mitochondrial-caspases, NFκB and autophagy with a relative block in proliferation, but they are not apoptotic. These data have led to a model in which the accumulation of α1 -ATZ in the ER activates ER and mitochondrial caspases, NFκB and autophagy but the caspase pathway is blocked at terminal steps so the globule-containing hepatocytes are “sick but not dead”. An “injury/regeneration” signal is generated by these cells in proportion to the amount of α1 -ATZ accumulation per cell or the number of cells with α1 -ATZ accumulation. Hepatocytes with lesser amounts of α1 -ATZ, globule-devoid hepatocytes, are therein

ch21

Jan. 10, 2007

8:57

SPI-B429



493

chronically stimulated “in trans” to divide. The cancer-prone state is then engendered by having cells that are unable to die at the appropriate time and cells that are chronically dividing in an inflamed milieu (Fig. 2). We would anticipate that some of the globule-containing hepatocytes eventually die, but because the block in their proliferation is relative, they can be replenished at least to the extent that there are always some of these cells present in the affected liver. This model is consistent with the observations of Geller et al. using the Z#2 transgenic mouse model of α1 -AT deficiency.5 These authors found that there were increasing areas of the liver that were negative for α1 -AT immunostaining as the mice aged, so that by the age of 12 months, more than 90% of the liver was α1 -AT-negative, corresponding to the globuledevoid hepatocytes. Moreover, by 6 months, adenomas began developing in the α1 -AT-negative areas and by 18 months, more than 80% of these mice had hepatocellular carcinoma arising from the α1 -AT-negative areas.41,42 It is not clear why adenomas and carcinomas arise rather specifically from the α1 -AT-negative regions. There are many possible explanations. Certainly, this is where the most rapid cell proliferation is occurring. The known anti-tumorigenic effect of autophagy, which is relatively specifically activated in the globule-containing hepatocytes of the PiZ mice and PIZZ patients,27,32 may also play a role.43−45 The mechanism by which the caspase pathway is blocked in globulecontaining hepatocytes is not known, but one possibility is the anti-apoptotic effect of heat shock proteins.46 We have shown that the accumulation of α1 ATZ leads to an increase in the expression of several heat shock proteins.47 Recent studies from the Grompe lab have indicated that a similar model may be involved in the mechanism of liver injury and hepatocarcinogenesis in the FAH mouse model of tyrosinemia.48 In these studies, the liver of the FAH mouse was found to contain hepatocytes that were damaged but not dead. In fact, these cells were found to be resistant to cell death, and therein prone to carcinogenesis. By the paradigm that we are proposing here, these hepatocytes would be considered equivalent to the globulecontaining hepatocytes in the PiZ mouse liver. They are damaged by an entirely different mechanism than are the globule-containing hepatocytes in the PiZ mouse liver and the damage is much more severe and rapidly progressing. However, they share the property of being TUNEL-negative

ch21

Jan. 10, 2007

8:57

SPI-B429

494


David H. Perlmutter

Fig. 2. Hypothetical model for hepatocarcinogenesis in α1 -AT deficiency. Globulecontaining hepatocytes (pale pink) tend to be periportal. They are “sick but not dead” and are generating a chronic regenerative stimulus, which is received in “trans” by globule-devoid hepatocytes (deep pink), which tend to be in the centrilobular regions leading to mitosis (bright red) and adenomatosis (dark red).

ch21

Jan. 10, 2007

8:57

SPI-B429



495

and upregulation of anti-apoptotic heat shock proteins has been implicated in the block in cell death.48 There is also evidence in the FAH mouse for chronic stimulation in “trans” of hepatocytes that are not damaged and have a selective proliferative advantage. According to the paradigm we are proposing here, these hepatocytes correspond to the globule-devoid hepatocytes in the PiZ mouse liver. The mechanism by which some undamaged hepatocytes arise in the FAH mouse as well as in tyrosinemic patients is known and appears to involve mutation reversion,49,50 whereas the origin/mechanism of globule devoid hepatocytes in the PiZ mouse and α1 -AT-deficient patient is still unknown. Evidence for a “trans” effect in the FAH mouse comes from studies with transplanted normal hepatocytes that are capable of undergoing multiple rounds of replication in the FAH liver, but only if disease is present. Once the mouse is given the drug NTBC, which prevents the accumulation of toxic intermediates in the damaged cells, transplanted hepatocytes will no longer replicate in the FAH mouse liver.51 A similar paradigm has been described for hepatocarcinogenesis in hepatitis B surface antigen (HBsAg) transgenic mice.52 In this model of hepatitis B-associated liver injury, like the PiZ model, the degree of hepatocellular proliferation correlated with the presence of injury and carcinomas, and with the cellular load of HBsAg, but the proliferating cells, adenomas and carcinomas had significantly lesser amounts of retained HBsAg. However, because this transgenic model only expresses HBsAg and not complete HB virus, it is not clear whether the paradigm applies to hepatocarcinogenesis in chronic hepatitis B virus infection. It is of some interest that HBsAg is known to be selectively retained in the ER of hepatocytes and this process is thought to be responsible for the so-called “ground glass hepatocyte”.53 Prolonged ER retention has also been observed for several proteins encoded by hepatitis C virus in infected hepatocytes,54 raising the possibility that a similar paradigm applies to liver injury and cancer predisposition in this increasingly important cause of liver disease.

4. Clinical Characteristics In many cases, this liver disease first becomes apparent at 4–8 weeks of age because of persistent jaundice. Conjugated bilirubin and transaminase

ch21

Jan. 10, 2007

8:57

SPI-B429

496


David H. Perlmutter

levels in the blood are mildly to moderately elevated. The liver may be enlarged, but rarely is there more in the way of symptoms, signs or laboratory abnormalities that suggest significant liver injury. It is very difficult to clinically differentiate these infants from infants affected by many other causes of neonatal liver disease, including infections, metabolic diseases and even the destructive hepatobiliary lesions of biliary atresia, and so α1 AT deficiency is usually thought of as one of the causes of a broad diagnostic category termed “neonatal hepatitis syndrome”. In some cases, there may be a cholestatic picture with icterus, pruritus and laboratory abnormalities such as hypercholesterolemia. Indeed, this subgroup of α1 -AT-deficient infants may have severe biliary epithelial cell damage and even paucity of the intrahepatic bile ducts detected in their liver biopsies.55 Rarely does α1 -AT deficiency manifest itself with severe progressive liver disease in the first year of life.56 The liver disease of α1 -AT deficiency may also be diagnosed later in childhood because of asymptomatic hepatomegaly, elevated transaminases detected incidentally or jaundice that develops during an intercurrent illness. Finally, this disease can first present in childhood, adolescence or adult life with complications of portal hypertension including splenomegaly, hypersplenism, gastrointestinal bleeding from varices, ascites, and/or hepatic encephalopathy. It should be considered in any adult patient with chronic liver disease, cryptogenic cirrhosis or hepatocellular carcinoma. An autopsy study done in Sweden suggested that as many as 25% of α1 -AT-deficient men who die between the ages of 40 and 60 have evidence of inflammation, necrosis, and/or carcinoma detected in their liver.57 The natural history of liver disease in α1 -AT deficiency is quite variable. Most infants that present with prolonged jaundice are asymptomatic by the time they reach 1 year of age. In the majority of these cases, there is no further evidence of liver disease for many years. As this diagnosis has only been known for 35 years or so, it is not yet known what proportion of these individuals develop liver disease and/or hepatocellular carcinoma. Even patients with severe liver disease caused by α1 -AT deficiency may have a stable or relatively slow-progressing course.58 It has not yet been possible to identify specific clinical and/or laboratory signs that can be used to predict a poor prognosis for liver involvement in α1 -AT deficiency. In the experience of most pediatric liver specialty clinics, the first definitive evidence for poor

ch21

Jan. 10, 2007

8:57

SPI-B429



497

prognosis comes in the form of a complication that affects the overall life functioning of the patient. It is still unclear whether heterozygotes for the classical form of α1 -AT deficiency are predisposed to liver disease. This is because all existing data on this issue have an inherent bias in ascertainment. Nevertheless, my experience with numerous PiMZ individuals with severe liver disease and no other plausible explanation leads me to believe that the predisposition does exist. There is still limited information about the incidence of liver disease in α1 -AT-deficient individuals with established emphysema. In one study involving 22 PIZZ patients with emphysema, there was elevated transaminase levels in 10 patients and cholestasis in 1 patient.59 Liver biopsies were not done in this study and so it might underestimate the extent and incidence of liver disease in adults with emphysema as their predominant clinical problem.

5. Diagnosis This deficiency should be considered in any of those with elevated transaminases, elevated conjugated bilirubin levels, asymptomatic hepatomegaly, signs or symptoms of portal hypertension, signs or symptoms of cholestasis, as well as bleeding/bruising with a prolonged prothrombin time. It should be considered in adults with chronic idiopathic hepatitis, cryptogenic cirrhosis and hepatocellular carcinoma. The diagnosis is established by means of serum α1 -AT phenotype (PI type) determination in isoelectric focusing electrophoresis or agarose electrophoresis at acid pH. Serum concentrations can be used for screening with follow-up PI typing of any values below normal (85–215 mg/dl). In my experience, it is wise to get both the serum concentration and PI typing when seriously considering this diagnosis. The distinctive histological feature of homozygous PIZZ α1 -AT deficiency, periodic acid-Schiff-positive, diastase-resistant globules in the ER of hepatocytes, substantiates the diagnosis. The presence of these inclusions should not be interpreted as diagnostic of α1 -AT deficiency. Similar structures are occasionally present in other liver diseases.60 The inclusions are eosinophilic, round to oval, and 1 to 40 µm in diameter. They are most prominent in periportal hepatocytes,

ch21

Jan. 10, 2007

8:57

SPI-B429


498

David H. Perlmutter

but may also be seen in kupffer cells and biliary epithelial cells.61 The liver biopsy may also be characterized by the variable degrees of hepatocellular necrosis, inflammatory cell infiltration, periportal fibrosis, and/or cirrhosis. There is often evidence of biliary epithelial cell destruction.

6. Treatment There is no specific therapy for α1 -AT deficiency-associated liver disease. Clinical care involves avoidance of cigarette smoking to prevent exacerbation of destructive lung disease/emphysema, supportive management of symptoms caused by liver dysfunction and prevention of complications of liver disease. Cigarette smoking markedly accelerates the lung disease associated with α1 -AT deficiency, reduces the quality of life, and significantly shortens the longevity of these patients.62 Progressive liver dysfunction and failure in children with α1 -AT deficiency has been managed successfully with liver transplantation with survival rates well over 92% for 5 years.63,64 Nevertheless, a number of homozygotes with severe liver disease, even cirrhosis or portal hypertension, may have a relatively low rate of disease progression and lead a relatively normal life for extended periods. With the availability of livingrelated donor transplantation techniques, it may be possible to observe these patients for some time before transplantation becomes necessary. Patients with α1 -AT deficiency and emphysema have been treated with purified plasma or recombinant α1 -AT administered intravenously or by means of aerosol as replacement therapy.65 This therapy is associated with improvement in serum and bronchoalveolar lavage fluid α1 -AT concentrations and neutrophil elastase inhibitory capacity in lavage fluid, without significant side effects. Although results of initial studies have suggested that there is a slower decline in forced expiratory volume in patients undergoing replacement therapy, this occurred only in a subgroup of patients and the study was not randomized.66 This therapy is designed for established and progressive emphysema. Protein replacement therapy is not being considered for patients with liver disease because there is no evidence that deficient serum levels of α1 -AT play a role in the development of liver injury. A number of patients with severe emphysema from α1 -AT deficiency are being treated with lung transplantation. Over a 13-year experience,

ch21

Jan. 10, 2007

8:57

SPI-B429



499

86 patients with α1 -AT deficiency underwent lung transplantation in St Louis with an ∼ 60% 5-year survival.67 A number of novel strategies for chemoprophylaxis and treatment of α1 -AT deficiency have been proposed recently. Because the treatment of the PiZ mouse with cyclosporine A (CsA) resulted in reduced hepatic mitochondrial damage, reduced hepatic caspase-3 activation and improved tolerance of starvation,28 CsA and other drugs that prevent the mitochondrial permeability transition appear to be candidates for the treatment of α1 -AT deficiency-associated liver disease. This strategy is particularly attractive because it involves a mechanism of action at a distal step in the pathway of liver damage that is still effective, even though the primary pathologic phenomenon, accumulation of α1 -ATZ in the ER continues. Several studies have shown that a class of compounds called chemical chaperones can reverse the cellular mislocalization or misfolding of mutant proteins.68–70 These compounds include non-specific chaperones such as glycerol, trimethylamine oxide, deuterated water and 4-phenylbutyric acid (PBA), as well as specific chaperones that have antagonistic or agonistic pharmacological bases. Burrows et al. found that glycerol and PBA mediate a marked increase in the secretion of α1 -ATZ in a model cell line.71 Moreover, oral administration of PBA was well tolerated by the PiZ mouse and consistently mediated an increase in blood levels of human α1 -AT, reaching 20–50% of the levels present in PiM mice and normal humans. However, PBA did not mediate an increase in the blood levels of α1 -AT in a recent pilot human trial.72 Studies of other candidate chemical chaperones will be important because the approach has the potential to prevent both liver damage by reducing the burden of α1 -ATZ that accumulates in the cell and to prevent lung damage by increasing the secretion of α1 -AT, and therein the amount of α1 -AT that reaches the lung to inhibit elastases. It also appears that several iminosugar compounds may be potentially useful for the chemoprophylaxis of liver and lung disease in α1 -AT deficiency. These compounds are designed to interfere with oligosaccharide side chain trimming of glycoproteins and are now being examined as potential therapeutic agents for viral hepatitis and other types of infection.73,74 Although we initially examined several of these compounds to determine the effect of inhibiting glucose or mannose trimming from the carbohydrate side chain of mutant α1 -ATZ on its fate in the ER, we found to our

ch21

Jan. 10, 2007

8:57

SPI-B429

500


David H. Perlmutter

surprise that one glucosidase inhibitor, castanospermine (CST) and two α-mannosidase I inhibitors, kifunensine (KIF) and deoxymannojirimicin (DMJ), actually mediate increased secretion of α1 -ATZ.75 Importantly, the α1 -ATZ that is secreted is partially functional as an elastase inhibitor. KIF and DMJ are less attractive candidates for chemoprophylactic trials because they delay degradation of α1 -ATZ, in addition to increasing its secretion and therefore having the potential to exacerbate susceptibility to liver disease. However, CST has no effect on the degradation of α1 -ATZ, and therefore may be targeted for development as a chemoprophylactic agent. Alternative strategies for at least partial correction of α1 -AT deficiency may result from further studies of the fate of α1 -ATZ in the ER. For instance, delivery of synthetic peptides to the ER to insert into the gap in the A sheet or into a particular hydrophobic pocket of the α1 -ATZ 76 and prevent polymerization of α1 -ATZ might result in the release into the extracellular fluid and prevent accumulation in the ER. Even if polymerization results from rather than causes the secretory defect, this strategy may be effective if peptide insertion leads to a change in conformation that is associated with translocation competence. Further understanding of the biochemical mechanism by which abnormally folded α1 -ATZ undergoes intracellular degradation might allow the development of pharmacological agents that can manipulate this degradative system, such as agents like interferonγ that enhance proteosomal activity, for prophylactic application to the subpopulation of deficient individuals predisposed to liver disease. Replacement of α1 -AT by means of somatic gene therapy has been discussed in the literature for a number of years.77 This strategy is potentially less expensive than replacement therapy with purified protein and may avert the need for weekly or even monthly administration. As a form of replacement therapy, however, this strategy will only be useful for lung disease in α1 -AT deficiency. There are still significant issues that need to be addressed before gene therapy becomes a realistic alternative.78 The most important pre-requisite will be demonstration that replacement therapy with purified plasma α1 -AT is truly associated with an ameliorative effect. Several novel types of gene therapy, such as repair of mRNA by trans-splicing ribozymes,79,80 chimeric RNA/DNA oligonucleotides,81–83 triplex-forming oligonucleotides,84 small fragment homologous replacement 85 or by RNA silencing 87,88 are theoretically attractive alternative

ch21

Jan. 10, 2007

8:57

SPI-B429



501

strategies for the prevention of liver disease associated with α1 -AT deficiency, because they would prevent the synthesis of mutant α1 -ATZ and ER retention. Other studies have shown that transplanted hepatocytes can repopulate the diseased liver in several mouse models,51,86 including a mouse model of the childhood metabolic liver disease hereditary tyrosinemia. Replication of transplanted hepatocytes occurs only when there is injury and/or regeneration in the liver. The results provide evidence that it may be possible to use hepatocyte transplantation techniques to treat hereditary tyrosinemia and perhaps other metabolic liver diseases, in which the underlying defect is cell-autonomous. For instance, α1 -AT deficiency involves a cell-autonomous defect and would be an excellent candidate for this strategy.

References 1. de Serres FJ, Blanco I and Fernandez-Bustillo E (2003) Genetic epidemiology of alpha-1 antitrypsin deficiency in NorthAmerica andAustralia/New Zealand: Australia, Canada, New Zealand and the United States of America. Clin Genet 64(5):382–397. 2. Sveger T (1976) Liver disease in α1-antitrypsin deficiency detected by screening of 200,000 infants. N Engl J Med 294:1316–1321. 3. Sveger T (1995) The natural history of liver disease in α-1-antitrypsin deficient children. Acta Paediatr Scand 77:847–851. 4. DeMeo DL and Silverman EK (2004) Alpha1-antitrypsin deficiency. 2: genetic aspects of alpha(1)-antitrypsin deficiency: Phenotypes and genetic modifiers of emphysema risk. Thorax 59(3):259–264. 5. Dycaico MJ, Grant SG, Felts K, Nichols WS, Geller SA, Hager JH and Pollard AJ, et al. (1988) Neonatal hepatitis induced by alpha 1-antitrypsin:A transgenic mouse model. Science 242:1409–1412. 6. Carlson JA, Rogers BB, Sifers RN, Finegold MJ, Clift SM, DeMayo FJ and Bullock DW, et al. (1989) Accumulation of PiZ antitrypsin causes liver damage in transgenic mice. J Clin Invest 83:1183–1190. 7. Lomas DA and Mahadeva R (2002) α1 -Antitrypsin polymerization and the serpinopathies: Pathobiology and prospects for therapy. J Clin Invest 110: 1585–1590.

ch21

Jan. 10, 2007

8:57

SPI-B429

502


David H. Perlmutter

8. Lomas DA, Evans DL, Finch JT and Carrell RW (1992) The mechanism of Z α1-antitrypsin accumulation in the liver. Nature 357:605–607. 9. An J-K, Blomenkamp K, Lindblad D and Teckman JH (2005) Quantitative isolation of α1 -AT mutant Z protein polymers from human and mouse livers and the effect of heat. Hepatology 41:160–167. 10. Kim J, Lee KN, Yi G-S and Yu M-H (1995) A thermostable mutation located at the hydrophobic core of α1-antitrypsin suppresses the folding defect of the Z-type variant. J Biol Chem 270:8597–8601. 11. Sidhar SK, Lomas DA, Carrell RW and Foreman RC (1995) Mutations which impede loop-sheet polymerization enhance the secretion of human α1-antitrypsin deficiency variants. J Biol Chem 270:8393–8396. 12. Kang HA, Lee KN and Yu M-H (1997) Folding and stability of the Z and Siiyama genetic variants of human α1-antitrypsin. J Biol Chem 272:510–516. 13. Lin L, Schmidt B, Teckman J and Perlmutter DH (2001) A naturally occurring non-polymerogenic mutant of α1-antitrypsin characterized by prolonged retention in the endoplasmic reticulum. J Biol Chem 276:33893–33898. 14. Schmidt BZ and Perlmutter DH (2005) Grp78, Grp 94 and Grp 170 interact with alpha-1-antitrypsin mutants that are retained in the endoplasmic reticulum. Am J Physiol 289:G444–G455. 15. Wu Y, Whitman I, Molmenti E. Moore K, Hippenmeyer P and Perlmutter D (1994) A lag in intracellular degradation of mutant α1-antitrypsin correlates with the liver disease phenotype in homozygous PiZZ α1-antitrypsin deficiency. Proc Natl Acad Sci USA 91:9014–9018. 16. McCracken AA and Brodsky JL (1996) Assembly of ER-associated protein degradation in vitro: dependence on cytosol, calnexin and ATP. J Cell Biol 132:291–298. 17. Werner ED, Brodsky JL and McCracken AA (1996) Proteasome-dependent endoplasmic reticulum- associated protein degradation: An unconventional route to a familiar fate. Proc Natl Acad Sci USA 93:13797–13801. 18. Qu D, Teckman HJ, Omura S and Perlmutter DH (1996) Degradation of mutant secretory protein, α1-antitrypsin Z, in the endoplasmic reticulum requires proteasome activity. J Biol Chem 271:22791–22795. 19. Teckman JH, Gilmore R and Perlmutter DH (2000) The role of ubiquitin in proteasomal degradation of mutant α1-antitrypsin Z in the endoplasmic reticulum. Am J Physiol 278:G39–G48. 20. Teckman JH, Burrows J, Hidvegi T, Schmidt B, Hale PD and Perlmutter DH (2001) The proteasome participants in degradation of mutant α1-antitrypsin Z in the endoplasmic reticulum of hepatoma-derived hepatocytes. J Biol Chem 48:44865–44872.

ch21

Jan. 10, 2007

8:57

SPI-B429



503

21. Kamimoto T, Shoji S, Hidvegi T, Mizushima N, Umebayashi K, Perlmutter DH and Yoshimori T (2006) Intracellular inclusions containing mutant α1antitrypsin Z are propagated in the absence of autophagic activity. J Biol Chem 281:4467–4476. 22. Cabral CM, Liu Y, Moremen KW and Sifers RN (2002) Organizational diversity among distinct glycoprotein endoplasmic reticulum-associated degradation programs. Mol Biol Cell 13:2639–2650. 23. Mayer T, Braun T and Jentsch S (1998) Role of the proteasome in membrane extraction of a short lived ER-transmembrane protein. EMBO J 17:3251–3257. 24. Cabral CM, Choudhury P, Liu Y and Sifers RN (2000) Processing by endoplasmic reticulum mannosidases partitions a secretion-impaired glycoprotein into distinct disposal pathways. J Biol Chem 275:25015–25022. 25. Teckman JH and Perlmutter DH (2000)Retention of the mutant secretory protein α1-antitrypsin Z in the endoplasmic reticulum induces autophagy. Am J Physiol 279:G961–G974. 26. Kruse KB, Bordsky JL and McCracken AA (2006) Characterization of an ERAD gene as VPS30/ATG6 reveals two alternative and functionally distinct protein quality control pathways: One for soluble Z variant of human alpha-1 proteinase inhibitor (A1PiZ) and another for aggregates of A1PiZ. Mol Biol Cell 17:203–212. 27. Kruse KB, Dear A, Kaltenbrun ER, Crum BE, George PM, Bernnan SO and McCrackenAA (2006) Mutant fibrinogen cleared from the endoplasmic reticulum via endoplasmic reticulum-associated protein degradation and autophagy: An explanation for liver disease. Am J Pathol 168:1299–1308. 28. Teckman JH, An JK, Blomenkamp K, Schmidt B and Perlmutter D (2004) Mitochondrial autophagy and injury in the liver in alpha 1-antitrypsin deficiency. Am J Physiol Gastrointest Liver Physiol 286:G851–G862. 29. Rudnick DA, Liao Y, An JK, Muglia LJ, Perlmutter DH and Teckman JH (2004) Analyses of hepatocellular proliferation in a mouse model of alpha-1antitrypsin deficiency. Hepatology 39:1048–1055. 30. Hidvegi T, Schmidt B, Hale P and Perlmutter DH (2005) Accumulation of mutant α1 -ATZ in the ER activates caspases-4 and -12, NFκB and BAP31 but not the unfolded protein response. J Biol Chem 280:39002–39015. 31. Wang B, Nguyen M, Breckenridge DG, Stojanovic M, Clemons PA, Kuppig S and Shore GC (2002) Uncleaved BAP31 in association with A4 protein at the endoplasmic reticulum is an inhibitor of fas-initiated release of cytochrome c from mitochondria. J Biol Chem 278(16):14461–14468. 32. Teckman JH, An J-K, Loethen S and Perlmutter DH (2003) Effect of fasting on liver in a mouse model of α1-antitrypsin deficiency: Constitutive activation of the autophagic response. Am J Physiol 283:G1117–G1124.

ch21

Jan. 10, 2007

8:57

SPI-B429

504


David H. Perlmutter

33. Ron D (2002) Translational control in the endoplasmic reticulum stress response. J Clin Invest 110:1382–1388. 34. Zhang K and Kaufman RJ (2004) Signaling the Unfolded protein response from the endoplasmic reticulum. J Biol Chem 279:25935–25938. 35. Pahl HL, Sester M, Burgert H-G and Baeuerle PA (1996) Activation of transcription factor NF-κB by the adenovirus E3/19K protein requires its ER retention. J Cell Biol 132:511–522. 36. Hu C and Perlmutter DH (2002) Cell-specific involvement of HNF-1 α in α1-antitrypsin gene expression in human respiratory epithelial cells. Am J Physiol 282:L757–L765. 37. Pikarsky E, Porat RM, Stein I, Abramovitch R, Amit S, Kasem S, GutkovichPyest E et al. (2004) NFκB functions as a tumor promoter in inflammationassociated cancer. Nature 431:461–465. 38. Greten FR, Eckmann L, Greten TF, Park JM, Li ZW, Egan LJ, Kagnoff M et al. (2004) IKKbeta links inflammation and tumorigenesis in a mouse model of colitis-associated cancer. Cell 118:285–296. 39. Nakagawa T, Zhu H, Morishima N, Li E, Xu J, Yankner BA and Yuan J (2000) Caspase-12 mediates endoplasmic-reticulum-specific apoptosis: A cytotoxicity by amyloid-β. Nature 403:98–103. 40. Hitomi J, Katayama T, Eguchi Y, Kudo T, Taniguchi M, Koyama Y, Manabe T, et al. (2004) Involvement of caspase-4 in endoplasmic reticulum stress-induced apoptosis and β-induced cell death. J Cell Biol 165:347–356. 41. Geller SA, Nichols WS, Kim S,Tolmachoff T, Lee S, Dycaico MJ, Felts K, et al. (1994) Hepatocarcinogenesis in the sequel to hepatitis in Z#2 α1-antitrypsin transgenic mice: histopathological and DNA ploidy studies. Hepatology 9:389–397. 42. Geller SA, Nichols WS, Dycaico MJ, Felts KA and Sorge JA (1990) Histopathology of alpha-1-antitrypsin liver disease in a transgenic mouse model. Hepatology 12:40–47. 43. Qu Y, Yu J, Bhagat G, Furuya N, Hibshoosh H, Troxel A, Rosen J, et al. (2003) Promotion of tumorigenesis by heterozygous disruption of beclin1 autophagy gene. J Clin Invest 112:1809–1820. 44. Liang XH, Jackson S, Seaman M, Brown K, Kempkes B, Hibshoosh H and Levine B ( 1999) Induction of autophagy and inhibition of tumorigenesis by beclin 1. Nature 402:672–676. 45. Yue Z, Jin S, Yang C, Levine AJ and Heintz N (2003) Beclin 1, an autophagy gene essential for early embryonic development, is a haploinsufficient tumor suppressor. Proc Natl Acad Sci USA 100:15007–15082.

ch21

Jan. 10, 2007

8:57

SPI-B429



505

46. Bruey JM, Ducasse C, Bonniaud P, Ravagnan L, Susin SA, Diaz-Latoud C, Gurbuxani S, et al. ( 2000) Hsp27 negatively regulates cell death by interacting with cytochrome c. Nat Cell Biol 2:645–652. 47. Perlmutter DH, Schlesinger MJ, Pierce JA, Punsal PI and Schwartz AL (1989) Synthesis of stress proteins is increased in people with homozygous PIZZ α1-antitrypsin deficiency and liver disease. J Clin Invest 84:1555–1561. 48. VogelA, van den Berg IET,Al-Dhalimy M, Groopman J, Ou C-N, Ryabinina O, Iordanov MS, Finegold M and Grompe M (2004) Chronic liver disease in murine hereditary tyrosinemia type I induces resistance to cell death. Hepatology 39:433–443. 49. Kvittingen EA, Rootwelt H, Berger R and Brandtzaeg P (1994) Self-induced correction of the genetic defect in tyrosinemia type I. J Clin Invest 94: 1657–1661. 50. Manning K, Al-Dhalimy M, Finegold M and Grompe M (1999) In vivo suppressor mutations correct a murine model of hereditary tyrosinemia type I. Proc Natl Acad Sci USA 96:11928–11933. 51. Overturf K, Al-Dhalimy M, Tanguay R, Brantly M, Ou CM, Finegold M and Grompe M (1996) Hepatocytes corrected by gene therapy are selected in vivo in a murine model of hereditary tyrosinemia type I. Nat Genet 12:266–273. 52. McLachlan A, Milich DR, Raney AK, Riggs MG, Hughes JL, Sorge J and Chisari FV (1987) Expression of hepatitis B virus surface and core antigens: Influences of pre-S and precore sequences. J Virol 61:683–692. 53. Wang H-C, Wu H-C, Chien CF, Fausto N, Lei H-Y and Su I-J (2003) Different types of ground glass hepatocytes in chronic hepatitis B virus infection contain specific pre-S mutants that may induce endoplasmic reticulum stress. Am J Pathol 163:2441–2449. 54. Dubuisson J (2000) Folding, assembly and subcellular localization of hepatitis C virus glycoproteins. Curr Top Microbiol Immunol 242:135–148. 55. Hadchouel M and Gautier M (1976) Histopathologic study of the liver in the early cholestatic phase of alpha-1-antitrypsin deficiency. J Pediatr 89: 211–215. 56. Grishan FR, Gray GF and Green HL (1983) α1-Antitryspin deficiency presenting with ascites and cirrhosis in the neonatal period. Gastroenterology 85:435–438. 57. Eriksson S, Carlson J and Velez R (1986) Risk of cirrhosis and primary liver cancer in α-1-antitrypsin deficiency. N Engl J Med 314:736–739. 58. Volpert D, Molleston JP and Perlmutter DH (2001) α1-Antitrypsin deficiencyassociated liver disease may progress slowly in some children. J Pediatr Gastroenterol Nutr 32:265–269.

ch21

Jan. 10, 2007

8:57

SPI-B429

506


David H. Perlmutter

59. von Schonfeld J, Breuer N, Zotz R, Liedmann H, Wencker H and Beste M (1996) Liver function in patients with pulmonary emphysema due to severe alpha-1-antitrypsin deficiency (PIZZ). Digestion 57:165–169. 60. Qizibash A and Yong-Pong O (1983) Alpha-1-antitrypsin liver disease: Differential diagnosis of PAS-positive diastase-resistant globules in liver cells. Am J Clin Pathol 79:697–702. 61. Yunis EJ, Agostini RM and Glew RH (1976) Fine structural observations the liver in α1-antitrypsin deficiency. Am J Pathol 82:265–286. 62. Wilson-Cox D (1989) Alpha-1-antitrypsin deficiency, in Scriver CB, Beaudet AL, Sly WS, Valle D, Vogelstein B and Childs E (eds.) The Metabolic Basis of Inherited Disease. McGraw-Hill; New York, pp 2409–2437. 63. Francavilla R, Castellaneta S, Hadzic N, Chambers SM, Portmann B, Tung J, Cheeseman P, Rela M, Heaton N and Mieli-Vergani G (2000) Prognosis of alpha-1-antitrypsin deficiency-related liver disease in the era of pediatric liver transplantation. J Heptol 32:986–992. 64. Kayler LK, Merion RM, Lee S, Sung RS, Punch JD, Rudich SM, Turcotte J, Campbell DA, Holmes R and Magee JC (2002) Long-term survival after liver transplantation in children with metabolic disorders. Pediatr Transplantation 6:295–300. 65. Wewers MD, Casolaro MA, Sellers SE, Swayze SC, McPaul KM, Witte JT and Crystal RG (1987) Replacement therapy for alpha 1-antitrypsin deficiency associated with emphysema. N Engl J Med 316:1055–1062. 66. The Alpha-1-Antitrypsin Deficiency Registry Study Group (1998) Survival and FEV1 decline in individuals with severe deficiency of α1-antitrypsin. Am J Respir Crit Care Med 158:49–59. 67. Cassivi SD, Meyers BF, Battafarano RJ, Guthrie TJ, Trulock EP, Lynch JP, Cooper JD and Patterson GA (2002) Thirteen-year experience in lung transplantation for emphysema. Ann Thorac Surg 74:1663–1670. 68. Sato S, Ward CL, Krouse ME, Wine JJ and Kopito RR (1996) Glycerol reverses the misfolding phenotype of the most common cystic fibrosis mutation. J Biol Chem 271:635–638. 69. Tamarappoo BK and Verkman AS (1998) Defective aquaporin-2 trafficking in nephrogenic diabetes insipidus and correction by chemical chaperones. J Clin Invest 101:2257–2267. 70. Brown CR, Hong-Brown LQ, Welch WJ (1997) Correcting temperaturesensitive protein folding defects. J Clin Invest 99:1432–1444. 71. Burrows JA, Willis LK and Perlmutter DH (2000) Chemical chaperones mediate increased secretion of mutant alpha 1-antitrypsin (α1-AT) Z: A potential

ch21

Jan. 10, 2007

8:57

SPI-B429



72.

73. 74.

75.

76.

77. 78. 79. 80. 81.

82. 83. 84. 85.

507

pharmacological strategy for prevention of liver injury and emphysema in α1-AT deficiency. Proc Natl Acad Sci USA 97:1796–1801. Teckman JH (2004) Lack of effect of oral 4-phenylbutyrate on serum alpha1-antitrypsin in patients with α-1-antitrypsin deficiency: A preliminary study. J Pediatr Gastroenterol Nutr 39:34–37. Jacob GS (1995) Glycosylation inhibitors in biology and medicine. Curr Opin Struct Biol 5:605–611. Zitzmann N, Mehta AS, Carrouee S, Butters TD, Platt FM, McCauley J, Blumberg BS, Dwek RA and Block TM (1996) Imino sugars inhibit the formation and secretion of bovine viral diarrhea virus, a pestivirus model of hepatitis C virus: Implications for the development of broad spectrum anti-hepatitis virus agents. Proc Natl Acad USA 96:11878–11882. Marcus NY and Perlmutter DH (2000) Glucosidase and mannosidase inhibits mediate increased secretion of mutant α1-antitrypsin Z. J Biol Chem 275:1987–1992. Parfrey H, Mahadeva R, Ravenhill NA, Zhou A, Dafforn TR, Foreman RC and Lomas DA (2003) Targeting a surface cavity of alpha-1-antitrypsin to prevent conformational disease. J Biol Chem 278:33060–33066. Crystal RG (1990) Alpha-1-antitrypsin deficiency, emphysema and liver disease: genetic basis and strategies for therapy. J Clin Invest 95:1343–1352. Anderson WF (2003) The current status of clinical gene therapy. Hum Gen Ther 13:1261–1262. Long MB, Jones JP, Sullenger BA and Byun J (2003) Ribozyme-mediated revision of RNA and DNA. J Clin Invest 112:312–318. Garcia-Blanco MA (2003) Messenger RNA reprogramming by spliceosomemediated RNA trans-splicing. J Clin Invest 112:474–480. Kren BT, Bandyopadhyay P and Steer CJ (1998) In vivo site-directed mutagenesis of the factor IX gene by chimeric RNA/DNA oligonucleotides. Natl Med 4:285–290. Metz R, Dicola M, Kurihara T, Bailey A, Frank B, Roecklyn B and Blaese M (2002) Mode of action of RNA/DNA oligonucleotides. Chest 121:915–925. Kmiec EB (2003) Targeted gene repair — in the arena. J Clin Invest 112: 632–626. Seidman MM and Glazer PM (2003) The potential for gene repair via triple helix formation. J Clin Invest 114:487–494. Gruenert DC, Bruscia E, Novelli G, Colosimo A, Dallapiccola B, Sangiuolo F and Goncz KK (2003) Sequence-specific modification of genomic DNA by small DNA fragments. J Clin Invest 112:637–641.

ch21

Jan. 10, 2007

8:57

SPI-B429

508


David H. Perlmutter

86. Davidson BL (2003) Hepatic diseases — hitting the target with inhibitor RNAs. N Engl J Med 349:2357–2359. 87. Rubinson DA, Dillon CP, Kwiatkowski AV, Sievers C, Yang L and Kopinja J (2003) A lentivirus-based system to functionally silence genes in primary mammalian cells, stem cells and transgenic mice by RNA interference. Nat Gen 33:401–406. 88. Rhim JA, Sandgen EP, Degen JL, Palmiter RD and Brinster RL (1994) Replacement of diseased mouse liver by hepatic cell transplantation. Science 263:1149–1152.

ch21

Jan. 10, 2007

14:27 SPI-B429


22 Regulation of Hemostasis by Heparin-Binding Serpins Frank C. Church, Robert N. Pike, Douglas M. Tollefsen, Ashley M. Buckle, Angelina V. Ciaccia and Steven T. Olson

1. Introduction Hemostasis is described as the balance between procoagulant and anticoagulant forces that act in concert to maintain the fluidity of blood under physiologic conditions.1–4 This system responds rapidly to vascular injury via the coagulation cascade, whose primary role is to prevent the flow of blood from the vasculature. When a vessel is injured, hemostasis is achieved by several mechanisms, including vessel spasm, formation of a platelet plug, blood coagulation, and eventual healing of the injured area. Platelets adhere to extracellular matrix proteins that are exposed from damaged blood vessel endothelial cells, and then become activated, and secrete factors that also attract and activate nearby platelets. The coagulation “cascade” is activated both by factors exposed on and secreted from the injured tissues and platelets. These events ultimately result in the formation of fibrin threads that enmesh platelets, blood cells, and plasma proteins to form the hemostatic plug or thrombus. If left unchecked, the clotting cascade would result in widespread thrombosis and vessel blockage; thus, “anticoagulant” pathways are also immediately activated that regulate the activities of procoagulant factors. Hemostasis is regulated by both procoagulant and anticoagulant components to insure that thrombus formation does not spread beyond 509

ch22

Jan. 10, 2007

14:27 SPI-B429


510

F. C. Church et al. Antithrombin Heparin Cofactor II Protein C Inhibitor Plasminogen Activator Inhibitor-1

Blood Coagulation (VIIa, IXa, Xa, XIa,Thrombin)

Protein C System

Hemostasis

(Thrombin/thrombomodulin, Activated Protein C, Protein S)

Protein C Inhibitor Plasminogen Activator Inhibitor-1 α1-Antitrypsin

Fibrinolysis (Tissue plasminogen activator, Urokinase/Urokinase Receptor, Plasmin)

Plasminogen Activator Inhibitor-1 Plasminogen Activator Inhibitor-2 α2-Antiplasmin Protein C Inhibitor

Fig. 1. Regulation of hemostasis by serpins. The overall host-defense mechanism of hemostasis is represented here by blood coagulation, fibrinolysis, and the protein C system. Each system has it own group of proteases that are activated from zymogens, and each system has its own group of serpins that are responsible for their down-regulation. Disruption of these proteases–serpin interactions may lead to a pathological state, which results in thrombosis. Shown in the boxes are the serpins that regulate each part of the hemostatic pathway, while the serine proteases and their accompanying components for blood coagulation, fibrinolysis, and the protein C system are given in parentheses.

the site of injury (Fig. 1). Several major anticoagulant mechanisms exist. Since the exposure of blood to tissue factor is the primary cause of the initiation of blood coagulation, regulation of factor VIIa/tissue factor activity is of utmost importance. The factor VIIa/tissue factor complex is regulated by tissue factor pathway inhibitor (TFPI), a multivalent Kunitztype protease inhibitor (TFPI will not be discussed further since it is a Kunitz-type inhibitor).5 Another mechanism involves the inhibition of the coagulation proteases by members of the serine protease inhibitor (serpin) superfamily.6–10 The protein C pathway, which itself is regulated by serpins, also plays a key role in hemostasis.11–15 Activated protein C down-regulates thrombin formation by proteolytically degrading coagulation factors Va and VIIIa. There is yet another mechanism that involves the fibrinolytic pathway of proteases, activators and serpins, which results

ch22

Jan. 10, 2007

14:27 SPI-B429


Regulation of Hemostasis by Heparin-binding Serpins

511

in the degradation of fibrin strands.16–20 In all of these mechanisms, the endothelium and its underlying basement membrane of the vessel wall play an active role. The annual incidence of venous thrombosis, one of the leading causes of mortality and morbidity, increases from 1 per 100,000 during childhood to 1 per 100 in old age.21–23 For almost two decades, inherited antithrombin deficiency was the only known genetic cause of thrombophilia. In the 1980s, deficiencies of protein C and protein S were linked to venous thromboembolism. Dahlback described resistance to activated protein C in 1993, caused by a single amino acid substitution in coagulation factor V protein (Arg506Gln), as a common thrombophilic disorder. In 1996, the prothrombin gene substitution (G20210A), which results in elevated levels of prothrombin was found to be another cause of thrombophilia. Evidence is accumulating that both genetic and non-genetic risk factors act in concert to precipitate thromboembolic events.

2. Regulation of Blood Coagulation by Antithrombin, Heparin Cofactor II, Protein C Inhibitor, and Plasminogen Activator Inhibitor-1 The majority of coagulation proteases are inhibited by two serpins, antithrombin (AT; systematic name SERPINC1) and heparin co-factor II (HCII; systematic name SERPIND1) (Fig. 1).9,10,24,25 The serpin– protease complexes are then cleared from the plasma by the non-specific low-density lipoprotein receptor related protein found on hepatocytes.26 Antithrombin inhibits thrombin, factors IXa, Xa, XIa, XIIa, kallikrein, and plasmin.27 However, a comparison of the inhibition rates indicates that only the reactions with thrombin, factor Xa, and factor IXa are physiologically important.28,29 Thrombin is the only coagulation protease that HCII inhibits.30 Target protease inhibition by antithrombin and HCII is greatly accelerated by glycosaminoglycans, such as heparin and heparan sulfate. The anticoagulant effect of commercial heparin is based on its ability to accelerate the inhibition of thrombin, factor Xa, and factor IXa by antithrombin. In vivo, heparan sulfate-containing proteoglycans that are localized to the endothelial cell membrane serve to accelerate this reaction. HCII inhibition of thrombin is also accelerated

ch22

Jan. 10, 2007

14:27 SPI-B429

512


F. C. Church et al.

by dermatan sulfate and dermatan sulfate-containing proteoglycans, which are localized on the plasma membranes of extravascular cells. Thus, it is postulated that HCII plays an important role in the regulation of thrombin activity in the extravasculature.31,32 Additionally, both protein C inhibitor (PCI) and plasminogen activator inhibitor-1 (PAI-1)/vitronectin can also inhibit thrombin, although the physiological relevance of these serpinthrombin inhibition reactions is not clear.

3. Regulation of the Protein C System by Protein C Inhibitor, Plasminogen Activator Inhibitor-1, and α1 -Antitrypsin While thrombin promotes blood coagulation by cleaving fibrinogen to fibrin and by participating in positive feedback loops that generate more thrombin, namely the activation of cofactors V and VIII and factor XI, it also participates in a negative feedback loop that involves the protein C system (Fig. 1).11–15 The protein C system is an important anticoagulant system that is activated in parallel with thrombin production. This pathway is activated when thrombin binds thrombomodulin (TM), an integral membrane protein on the endothelial cell surface. Thrombin bound to TM has altered protease specificity, such that it now cleaves the zymogen protein C and no longer recognizes fibrinogen and factor V. Binding of zymogen protein C is provided by the endothelial protein C receptor (EPCR), which presents protein C to thrombin-TM. Cleavage of protein C generates activated protein C (APC), which associates with its cofactor protein S and proteolytically inactives the procoagulant cofactors Va and VIIIa. The primary physiological regulator of APC activity is thought to be the serpin protein C inhibitor (PCI; also known as plasminogen-activator inhibitor-3; systematic name SERPINA5), whose activity is accelerated in the presence of glycosaminoglycans, such as heparin.33–42 PCI is also a potent inhibitor of the thrombin-thrombomodulin complex. Therefore, PCI inhibits the protein C system at the level of both zymogen activation and protease inhibition. Interestingly, it is important to note that antithrombin does not inhibit APC. Additionally, both PAI-1/vitronectin and α1 -antitrypsin have the ability to inhibit APC, which indicates that numerous serpins with inhibition rates from slow to fast are capable of regulating the protein C system.

ch22

Jan. 10, 2007

14:27 SPI-B429



513

4. Regulation of the Fibrinolytic System by Plasminogen Activator Inhibitor-1, α2 -Antiplasmin, Plasminogen Activator Inhibitor-2, and Protein C Inhibitor The fibrinolytic system serves to degrade fibrin polymers formed during the coagulation process (Fig. 1).16–20 The key protease in fibrinolysis is plasmin, which lyses fibrin clots by proteolyzing fibrin polymers after specific lysine residues. Plasmin is formed by limited proteolysis of plasminogen by two distinct physiological plasminogen activators: tissue-type plasminogen activator (t-PA) and urokinase-type plasminogen activator (u-PA). t-PA is the primary vascular physiological plasminogen activator since it is synthesized by the endothelium, binds specifically to fibrin, and induces fibrinolysis at lower concentrations than u-PA. u-PA is involved in pericellular plasminogen activation, since it is primarily localized to fibroblasts and other mesenchymal cells and requires binding to its cell-surface receptor, u-PAR, to express full activity. Inhibition of fibrinolysis occurs both at the level of plasmin and plasminogen activation. Plasmin is inhibited by the serpin α2 antiplasmin (systematic name SERPINF2). Another serpin, PAI-1 (PAI-1; systematic name SERPINE1), is the principal inhibitor of both plasminogen activators, since it inhibits both t-PA and u-PA efficiently and functions in solution and is bound to the surfaces.43,44 PAI-1 is also involved in regulation of tumor cell invasion, metastasis, and angiogenesis by contributing to a cytostatic environment and maintaining extracellular matrix (ECM) integrity by inhibition of uPA.45 PAI-1 is a unique serpin because of its very short half-life. It has a strong dependence on cytokines/growth factors for its expression, and PAI-1 is unstable in its native form and rapidly converts to a latent form.46 PAI-1 circulates in complex with the ECM protein vitronectin, which stabilizes the serpin in the active form. PAI-1 also has a heparin-binding site that resides in its D-helix region.47,48 PAI-2 is an intracellular serpin whose primary target is thought to be urokinase, although its exact role in regulating fibrinolysis is not fully understood.49,50 Both t-PA and u-PA are inhibited by PCI; however, the rates of inhibition are significantly slower than that found for PAI-1. PCI inhibition of these proteases may be more physiologically relevant in the kidney/urinary tract from where it was originally purified and named PAI-3.

ch22

Jan. 10, 2007

14:27 SPI-B429

514


F. C. Church et al.

5. Heparin-binding Serpins: Antithrombin, Heparin Cofactor II, and Protein C Inhibitor 5.1. Antithrombin properties and function Antithrombin is one of the several abundant serpin family protease inhibitors in blood, which circulates in plasma at a concentration of ∼ 2 µM. This serpin carries out a critical anticoagulant function by inhibiting blood clotting cascade proteases that have escaped from a site of vascular injury and is thereby thought to prevent the dissemination of blood clotting beyond an injury site.27 Such an anticoagulant function has been clear for many years from the well-established association of congenital or inherited deficiencies of antithrombin with an increased risk of developing thrombotic disease.51 Antithrombin is a glycoprotein consisting of 432 amino acids, 3 disulfide bonds, and 4 identical biantennary carbohydrate chains terminated by variable numbers of sialic acid residues in the predominant alpha form of the protein.52 A less prominent beta form lacks the carbohydrate chain at Asn 135.53 In addition to a ∼350 residue core serpin sequence homologous to other serpin sequences, antithrombin has a unique N-terminal extension.6 Antithrombin appears to have evolved sufficient amino acid differences from other serpins to be classified in a clade of its own.54 An important feature of antithrombin’s anticoagulant function is that it is regulated by certain species of mast cell heparin55 and endothelial cell-associated heparan sulfate.56 These sulfated, highly negatively charged glycosaminoglycans are thus key activators of antithrombin and greatly accelerate the reactions of the serpin with its target clotting proteases, principally thrombin, factor Xa and factor IXa, by 3–5 orders of magnitude to yield diffusion-limited, physiologically relevant rates, i.e. bimolecular rate constants of ∼ 107 M−1 s−1 .29,57–59 Such rate enhancements are the basis for the widespread clinical use of heparin as an anticoagulant drug for the treatment of venous thrombosis. 5.2. Mechanisms of heparin activation of antithrombin The mechanism by which heparin or heparan sulfate glycosaminoglycans activate antithrombin’s inhibitory function has received considerable attention over the past 30 years. A unique feature of the mechanism is

ch22

Jan. 10, 2007

14:27 SPI-B429



515

the involvement of a sequence-specific pentasaccharide site in heparin or heparan sulfate chains, which is required for antithrombin to bind with high-affinity and be fully activated by the polysaccharide (Fig. 2).57,60,61 High affinity heparin chains containing the pentasaccharide sequence bind antithrombin with a KD value of 10–20 nM under physiologic conditions whereas low-affinity polysaccharide species lacking this sequence bind about ∼1000-fold weaker.62 Anticoagulantly active heparan sulfate species associated with endothelial cells bind antithrombin with a similar high affinity.63 The binding of high-affinity heparin chains is a two-step process involving an initial low-affinity electrostatic interaction with the serpin, followed by a conformational change, which greatly increases heparin affinity through an induced-fit mechanism and enhances the reactivity of the inhibitor with its target clotting proteases.57,64,65 The heparin pentasaccharide is sufficient to bind antithrombin with high-affinity and induce full conformational activation of the serpin, although full-length heparin chains somewhat enhance the binding affinity due to their ability to make an additional ionic interaction outside the main pentasaccharide binding site.57 The activating conformational change induced in antithrombin by the heparin pentasaccharide contributes severalhundred-fold to the overall 104 –105 -fold enhancements in the rates of antithrombin inhibition of factor Xa and factor IXa.57,59 However, this conformational change contributes insignificantly (∼ 2-fold) to the severalthousand-fold enhancement in the rate of thrombin inhibition. Full-length heparin chains containing the pentasaccharide and having a minimal length of 16–18 saccharides are required to produce the maximal enhancement in the rate of thrombin inhibition.29,66,67 The rate enhancement in this case is due to the longer heparin chain binding thrombin alongside antithrombin in a ternary complex and thereby acting as a bridge or template to promote the encounter of the protease with the serpin (Fig. 2).57,68 Conformational activation of antithrombin thus serves principally to promote high affinity binding of the serpin to the heparin pentasaccharide site in this mechanism and minimally to affect the interaction of the serpin with thrombin in the ternary bridging complex. Thrombin binding to heparin in the ternary complex is mediated through an electropositive exosite on the protease that lies close to the active-site.69,70 Factors Xa and IXa contain a homologous heparin binding exosite, which is made accessible under

ch22

Jan. 10, 2007

14:27 SPI-B429


516

F. C. Church et al.

Antithrombin

Antithrombin - Heparin

Heparin Cofactor II

Cleaved Protein C Inhibitor

Antithrombin - Thrombin

Heparin Cofactor II - Thrombin

ch22

Jan. 10, 2007

14:27 SPI-B429



517

physiologic conditions by the presence of calcium ions. Calcium binding to the acidic Gla-domain of these proteases, which is absent in thrombin, thus prevents the acidic domain from interacting with and blocking the basic exosite.58,71,72 As a result, physiologic calcium levels further enhance the accelerating effect of full-length heparin chains containing the pentasaccharide on the reactions of antithrombin with factor Xa and factor IXa by 30–400-fold through the ternary complex bridging mechanism.29,58,59,72 5.3. Molecular details of the antithrombin–heparin interaction The heparin binding site of antithrombin was initially localized by studies of natural or chemically modified antithrombin variants with defects in heparin binding.27 The X-ray crystal structure of the antithrombin– pentasaccharide complex later confirmed this gross localization and further revealed the details of the interactions between antithrombin and the pentasaccharide (Fig. 2).73 The heparin binding site is formed by three distinct regions in the protein, which are brought together in the folded serpin: the N-terminus, the A helix and the D helix, including the N- and C-terminal extensions of the latter helix. Comparison of X-ray crystal structures of free and pentasaccharide-complexed antithrombin73,74 shows that binding of the pentasaccharide induces conformational changes that cause these regions to close up around the pentasaccharide and form a complementary binding site in which antithrombin basic residues are positioned for optimal hydrogen bonding and ionic interactions with negatively charged sulfate

Fig. 2. Structures of three heparin-binding serpins that regulate the proteases of hemostasis. (Top) The structures of antithrombin without (left panel) and with (right panel) the heparin pentasaccharide bound. (Middle) Heparin cofactor II (left panel) and cleaved protein C inhibitor (right panel) are shown as indicated. In all of the structures, the A β-sheet is shown in red, the B β-sheet in green, and the C β-sheet in yellow. The reactive center loop is shown in magenta and the D-helix in blue. The H-helix is shown in dark pink in the protein C inhibitor structure. Positively charged residues on the D- and H-helices and the heparin pentasaccharide are shown in ball and stick format. (Bottom) Encounter complex structures of antithrombin-thrombin (S195A)-heparin (left panel) and of heparin cofactor II–thrombin (S195A active site mutant) (right panel).

ch22

Jan. 10, 2007

14:27 SPI-B429

518


F. C. Church et al.

and carboxylate groups of the saccharide. The relative importance of these interactions has more recently been established by mutagenesis studies.75 Notably, three antithrombin residues, Lys 114, Lys 125, and Arg 129, contribute the most to the overall binding energy and show cooperativity in binding the pentasaccharide.76–78 Presumably, their interactions with the pentasaccharide drive the activating conformational changes, which fully engage the network of electrostatic and hydrogen bonding interactions in the final complex and lead to the high affinity interaction. Other residues, namely, Lys 11, Arg 13, and Arg 47, are less important in activating the inhibitor and make non-cooperative ionic interactions with the pentasaccharide once the inhibitor is activated, which serve to stabilize the serpin in the activated state.79,80 Two exposed contiguous phenylalanine residues in the D helix, Phe 121 and Phe 122, contribute indirectly to binding the pentasaccharide by making non-polar interactions with the stems of the critical basic residues in the activated state, which position the latter for optimal interactions with the pentasaccharide.81 Other residues in the periphery of the binding site, namely, Arg 24 and Trp 49, also indirectly contribute to pentasaccharide binding by participating in interaction networks, which are important for stabilizing the activated state of the serpin.79,82,83 Conformational activation of antithrombin is largely driven by the preferential binding of the pentasaccharide ligand to the activated state of the inhibitor. In this respect, the pentasaccharide is acting as a classical allosteric activator by shifting an equilibrium between low and high activity conformational states of the protein through preferential binding to the less favored high activity state.84 The binding energy of all residues involved in the interaction is thus fully realized only after conformational activation. Little binding energy is expended in making the initial low affinity interaction with the native inhibitor. Lys 125 makes the greatest contribution (∼10-fold) to the initial docking interaction with the pentasaccharide. Other basic residues contribute to a lesser degree, but collectively provide a positive electrostatic surface that attracts the negatively charged pentasaccharide through an electrostatic steering effect.76,84,85 Lys 114 is particularly critical for inhibitor activation as evidenced by a massive 105 -fold reduction in pentasaccharide affinity resulting from mutation at this position, which is more than 100-fold greater than the affinity losses produced by mutation of Lys 125 or Arg 129.78 As a consequence, the substantial

ch22

Jan. 10, 2007

14:27 SPI-B429



519

stabilization of the activated state over the native state upon pentasaccharide binding is lost when Lys 114 is mutated. A 3-O-sulfate on the central pentasaccharide residue, which is a marker of the pentasaccharide sequence, mediates the critical Lys 114 interaction.73 The Lys 114 interaction requires the loop at the N-terminal end of the D helix in which Lys 114 resides to form a new P helix, which may trigger the conformational changes that activate the inhibitor and produce the high-affinity pentasaccharide interaction. A disulfide bond between the D helix and the N-terminal regions links these conformational changes in the heparin-binding site. Studies with modified pentasaccharides with saccharide deletions on the non-reducing or reducing ends have shown that the non-reducing end trisaccharide unit is sufficient to form the initial low affinity interaction and induce the activating conformational change in antithrombin. The reducing end disaccharide makes no contribution to the initial pentasaccharide interaction and binds only after conformational activation to lock the inhibitor in the activated state.64 Lys 114, Lys 125, andArg 129 of antithrombin and the non-reducing end trisaccharide of the pentasaccharide thus appear to comprise the core elements of the allosteric activating machinery. 5.4. Reactive center loop structural changes induced by heparin activation The nature of the conformational changes in antithrombin that are induced by heparin binding and are responsible for the enhanced reactivity of the serpin with its target proteases has been intensively investigated for almost two decades. The first crystal structures of antithrombin were solved in 1994.86,87 and showed that the reactive center loop (RCL) of the serpin in the native state was partially inserted into the A β-sheet (Fig. 2). It must be noted at this juncture that almost all structures of antithrombin solved to date show the serpin in a dimeric form, where one native molecule is in intimate contact with another antithrombin molecule in the latent form. The RCL of the native molecule is inserted into the space in the C β-sheet of the latent molecule left vacant by strand s1C being pulled out of the C β-sheet as a result of the transition to latency. This of course led to some skepticism as to the reality of the conformation of the RCL of the native molecule. Nevertheless, in addition to the partial insertion into the

ch22

Jan. 10, 2007

14:27 SPI-B429

520


F. C. Church et al.

A β-sheet, the RCL of the native molecule was shown to have the critical P1 arginine residue pointing inwards to the body of the molecule. This orientation in the crystal structure was stabilized by a hydrogen bond to the main chain carbonyl group of Glu 255 on the body of the molecule, while the side-chain of Glu 255 made a salt bridge with the side-chain of Arg 399, located at a bend in the RCL. The inwards orientation of the P1Arg residue and the bonding involved was thought to be in accord with results from natural variants of antithrombin,88 where in particular the substitution of a His residue was found to not only essentially inactivate the serpin in terms of its inhibitory activity, but also to alter its heparin affinity.89 This supported the hypothesis that the conformational activation step induced by heparin binding culminated in the rearrangement of the RCL conformation of antithrombin such that the P1 residue changed its orientation from being inward facing to being externally orientated. The crystal structure of antithrombin bound to the heparin pentasaccharide demonstrated73 that the RCL was expelled from the A β-sheet, in agreement with biochemical evidence and predictions from modeling that this was the case.90–92 Frustratingly, the structure was of dimeric antithrombin once again and therefore the conformation of the rest of the RCL did not change from that seen in the structure for native antithrombin in the absence of heparin. Biochemical evidence demonstrated that the sidechain of the P1 arginine residue becomes accessible upon binding of heparin pentasaccharide and, therefore, most likely externally orientated.73,93 This experiment supported the theory that the full conformational change in antithrombin resulted in a complete rearrangement of the conformation of the RCL, which partially explained the increased rate of interaction with target proteases. Later studies questioned the importance of the proposed rearrangement of the P1 arginine orientation to the rate of association with proteases and suggested that this mechanism did not account for the increased rate of association with target proteases such as factor Xa.94–96 Nevertheless, the bulk of the studies confirm that there is a rearrangement of the positioning of the P1 residue.93,97 However, in solution there is an equilibrium between the forms with P1 facing inwards and those with P1 facing outwards, with the inward facing form favored in the absence of heparin and the outward facing form favored in the heparin bound form.94–96 A recent structure of monomeric antithrombin lacking the RCL crystal contacts of the dimeric

ch22

Jan. 10, 2007

14:27 SPI-B429



521

antithrombin structures confirms that the PI residue is inward facing due to an interaction with Glu237 in the native conformation.201 Moreover the recent crystal structures of antithrombin bound to thrombin in the presence of a full-length heparin mimetic consisting of serpin and protease binding domains connected by an uncharged linker show that the P1 arginine is in an external orientation interacting deeply with the S1 pocket of the protease (Fig. 2).98,99 The RCL of the serpin is apparently in a canonical state such as that observed in other serpin–protease Michaelis complex structures100–102 and provides the sole contact with thrombin, which is essentially held away from the serpin. Surprisingly, the orientations of the RCL complexed with thrombin are somewhat different in the two reported structures owing to the RCL hinge being partially inserted into sheet A in a native-like state in one of the structures99 and fully expelled in the other.98 This would imply that the loop-inserted and loop-expelled conformations are similarly capable of a productive thrombin interaction, in keeping with pentasaccharide activation of antithrombin not greatly influencing the reaction with thrombin. The suggestion that changes in the P1 Arg residue orientation upon conformational activation of antithrombin by the heparin pentasaccharide account for only a small part of the increased reactivity of the serpin with factors Xa and IXa in the activated state have prompted a number of mutagenesis studies aimed at identifying the residues in antithrombin and in the proteases, which are responsible for the increased rate of inhibition seen in the presence of heparin pentasaccharide. Studies on the proteases have suggested that Gln 61 of factor Xa makes contacts with the peptide backbone of the P1 or P2 residues of antithrombin,103 and that the Arg150 loop of this enzyme and factor IXa42,104 is also critical in the mediation of the rate enhancement seen with heparin–pentasaccharide bound antithrombin. Other studies have strongly suggested the existence of an additional site (exosite) on antithrombin,94,95 located on strand 3 of the C β-sheet,105,202 which is made accessible upon heparin binding and then mediates specific interactions with factors IXa and Xa. This would argue that the interaction between these proteases and heparin bound antithrombin is different from that for thrombin and may rather resemble the interaction between heparin cofactor II and thrombin, where the contact between the protease and the serpin is far more intimate than that seen for antithrombin-thrombin and

ch22

Jan. 10, 2007

14:27 SPI-B429

522


F. C. Church et al.

involves an exosite.102 This expectation has been recently confirmed by solution of the crystal structure of the encounter complex between heparinbound antithrombin and S195A factor Xa.203 The structure thus reveals an intimate exosite interaction of Arg150 of factor Xa with strand 3C and contiguous residues of antithrombin, in agreement with mutagenesis studies. Moreover, achieving this exosite interaction requires the expulsion of the reactive loop from sheet A which is induced by heparin activation, consistent with studies showing that an engineered disulfide which prevents reactive loop expulsion blocks the enhancement in antithrombin reactivity with factors Xa and IXa upon heparin pentasaccharide activation.106 5.5. Mechanism of transmission of conformational change from the heparin binding site to the RCL The mechanism by which conformational changes in the heparin binding site are transmitted to the RCL to cause expulsion of the RCL from sheet A, reorientation of the P1 residue, and exposure of the exosite has been suggested by modeling and mutagenesis studies. The global conformational changes induced in the serpin by heparin binding have been deconstructed into rotations of rigid body fragments, with marked changes in surface residue electrostatic and hydrogen bonding interactions between the fragments in the two states.107 However, mutagenesis studies have failed to demonstrate the critical importance of these surface residue interaction networks in mediating the conformational activation switch, in particular, the P13 Glu residue, which was thought to be key to switching between the native and activated states.108 The critical change in the heparin binding site, which is required to induce the RCL changes, appears to involve a two turn extension of the D helix into the loop at the C-terminal end of the helix.73,92 Deletions or mutation of Ala 131 to Pro in this region thus uncouple the conformational changes in the heparin binding site from those in the RCL and impair conformational activation.109,110 The structure of a presumed intermediate antithrombin-pentasaccharide complex in which the conformational changes in the heparin binding site have occurred without the D helix extension or reactive loop expulsion have suggested that an equilibrium exists between the RCL-inserted and expelled states in the complex and that the D helix extension and closure of sheet A can only

ch22

Jan. 10, 2007

14:27 SPI-B429



523

occur once the RCL has been expelled from the sheet.111 The mechanism of transmission of the conformational changes in the heparin binding site thus appears to initially require the RCL hinge to be expelled from sheet A through a mobile equilibrium, which then allows the D helix to extend. The helix extension causes the loop connecting the D helix to strand 2 of sheet A to shorten and forces the sheet to contract, thereby preventing the RCL from reentering the sheet and locking the inhibitor in the reactive loop expelled high activity state. Expulsion of the RCL reorients the P1 residue and makes exosites on the serpin body accessible to promote an intimate interaction with factors Xa and IXa. 5.6. Role of the RCL in determining antithrombin specificity for proteases The extent to which the reactive loop sequence of antithrombin contributes to the recognition of its target blood clotting proteases has been a question of considerable interest in the case of antithrombin since these target proteases exhibit major differences in their specificity for protein substrates.84 The P1 Arg residue is clearly of paramount importance in recognizing the trypsin-like clotting proteases, given the massive reductions in reactivity that accompany mutations of this residue.94,95 However, there appears to be surprisingly little specificity in the flanking sequence, except for the P1 residue interaction with thrombin, which is restricted to small P1 sidechains due to the occlusion of the S1 binding pocket by the 60 loop.112,113 Changing the P6–P3 residues to a consensus sequence for thrombin recognition based on the recognition sequences of its macromolecular substrates or to the RCL sequence of the thrombin-specific serpin, protease nexin-1, substantially enhances antithrombin reactivity with thrombin, with the greatest effect resulting from changing the unfavorable P2 Gly to Pro.59,94 However, these changes in reactivity are largely independent of the reactivity enhancements resulting from heparin activation. Factor Xa and factor IXa reactivity with the serpin is reduced, not surprisingly, by the RCL sequence changes designed to enhance thrombin reactivity, but the reductions are quite modest except for the P1 residue, and as with thrombin the changes in reactivity are largely independent of the reactivity enhancements produced by heparin activation.59,95 The RCL

ch22

Jan. 10, 2007

14:27 SPI-B429

524


F. C. Church et al.

P1 residue thus appears to be the principal determinant of the reactivity of antithrombin with its target proteases in the absence of heparin and this contribution is only modestly enhanced in the presence of the polysaccharide activator. Heparin activation thus increases antithrombin reactivity largely through the presentation of new protease interaction sites (exosites) either on the serpin body outside the RCL or on the bound heparin chain. Intriguingly, attempts to engineer thrombin specificity into the serpin, α1 -protease inhibitor (α1 PI), showed that changing the P1 Met to Arg in addition to other changes to a thrombin-preferred consensus sequence produced a highly reactive thrombin inhibitor.114 However, this thrombinspecific α1 PI variant retained significant reactivity toward the anticoagulant protease, activated protein C, making it ineffective as a potential antithrombotic agent. Only when the flanking sequence was changed to the sequence in antithrombin did the variant α1 PI show selectivity in rapidly inhibiting thrombin without significantly inhibiting activated protein C.115 It thus appears that the RCL sequence of antithrombin evolved not for optimal recognition of its target procoagulant proteases, but rather to prevent recognition by the anticoagulant protease, activated protein C. To overcome the poor recognition of its target procoagulant proteases mainly through the RCL P1 residue, the heparin activation mechanism evolved to enhance this recognition through the presentation of exosites on the serpin or on a bridging heparin, which could form additional interactions with the protease. These exosites minimally affect the reaction with activated protein C and thereby account for antithrombin’s selectivity for procoagulant proteases over the anticoagulant protease, despite the latter protease’s similar trypsin-like specificity for P1 Arg-containing substrates.29

5.7. Antiangiogenic activity of conformationally altered forms of antithrombin Purification of an antiangiogenic factor found to be present in bovine serum identified the factor to be cleaved antithrombin.116 The native protein was shown to be devoid of this activity, suggesting that the conformational changes induced by cleavage in the RCL by proteases, wherein the cleaved RCL is inserted as a sixth strand into the center of sheet A,117 generates a

ch22

Jan. 10, 2007

14:27 SPI-B429



525

new antiangiogenic epitope in the serpin. Subsequent studies showed that the latent form of antithrombin in which the RCL has inserted into sheet A without cleavage86 also expresses antiangiogenic activity,118 suggesting that similar conformational changes in antithrombin mediate the antiangiogenic activity. The antiangiogenic activity of both antithrombin forms has been demonstrated by the ability to inhibit basic fibroblast growth factor or vascular endothelial growth factor-stimulated endothelial cell proliferation or capillary growth in the chick chorioallantoic membrane, and, most likely due to this ability, has been shown to exhibit antitumor activity in mice.116,118 Interestingly, several other serpins such as pigment epitheliumderived factor (PEDF), PAI-1, and kallistatin have also been shown to exhibit antiangiogenic activity.18,119,120 However, the latter serpins exhibit this activity in their native forms, suggesting that a distinct mechanism is involved in the generation of antiangiogenic activity in conformationally altered antithrombin forms. The generation of both cleaved and latent forms under physiologic conditions121,122 suggests that the antiangiogenic activity of these forms may have physiologic relevance possibly in the regulation of tissue remodeling and wound healing. A puzzling observation has been the finding of a third antiangiogenic form of antithrombin, termed prelatent, which was isolated as a byproduct in the preparation of latent antithrombin from the native serpin by mild heating (50◦ C).123 Purification of the heat-treated native antithrombin by heparin-agarose chromatography revealed, in addition to the low heparin affinity latent antithrombin peak, a high heparin affinity peak thought to be residual native antithrombin. However, unlike the starting native antithrombin, the high heparin affinity antithrombin fraction was found to have a potent antiangiogenic activity. Further characterization of this “prelatent” antithrombin showed that it bound heparin with an affinity similar to that of the native protein and inhibited proteases in the absence and presence of heparin with rates similar to native antithrombin. The only difference discernable between native and prelatent forms of the serpin was that the latter appeared to show a distinct susceptibility to cleavage by certain non-target proteases. The unique cleavage site implicated the region between residues 325 and 375, involving a mobile loop at the distal end of sheet A, as having been altered in structure. Whether an antiangiogenic epitope resides in this region and is a common feature of the three antiangiogenic forms of the inhibitor remains to be demonstrated.

ch22

Jan. 10, 2007

14:27 SPI-B429

526


F. C. Church et al.

Mechanisms shown to mediate the antiangiogenic effect of latent antithrombin include an increased rate of apoptosis of endothelial cells and perturbed cell-matrix interactions resulting from the inhibition of focal adhesion kinase activation, focal adhesion contact formation, and actin reorganization.118 The mechanism of antithrombin antiangiogenic action has been further probed by the cDNA microarray approach to identify genes whose expression is altered by antiangiogenic forms of the serpin.124 A well-established proangiogenic heparan sulfate proteoglycan known as perlecan125 was found to be greatly down-regulated from 6to 10-fold in growth factor-stimulated primary HUVEC cultures treated with cleaved or latent forms of antithrombin, but not in cells treated with the native serpin. This downregulation was confirmed at the mRNA level by Northern blotting and semi-quantitative RT-PCR and at the protein level by immunoblotting. Down-regulation of the perlecan gene was also observed in the absence of growth factor stimulation of the cells, suggesting that antiangiogenic antithrombins interact with an endothelial cell receptor to mediate the down-regulation. That decreased perlecan levels were correlated with the antiproliferative effects of the antiangiogenic antithrombins on growth factor-stimulated endothelial cells was shown by the ability of TGFβ1, a known stimulator of perlecan expression, to increase perlecan expression and overcome the inhibitory effects of cleaved antithrombin on endothelial cell growth.124 Down-regulation of perlecan expression using antisense approaches similarly blocks bFGF-stimulated proliferation of fibroblast and tumor cell lines and tumor cell-induced angiogenesis, due to the fact that heparan sulfate chains on perlecan function as a coreceptor for bFGF binding to its receptors.126,127 Antiangiogenic forms of antithrombin may thus in part exert their antiangiogenic effects by downregulating the expression of perlecan coreceptors required to mediate the action of proangiogenic growth factors. However, the microarray analysis also showed that antiangiogenic forms of antithrombin additionally affect the expression of other proangiogenic and antiangiogenic genes, suggesting that antiangiogenic antithrombins inhibit angiogenesis through global effects on gene expression involving multiple genes analogous to the mode of action of endostatin.128,204 More recently, it has been shown that the heparin binding site of antithrombin also plays an important role in mediating the antiangiogenic

ch22

Jan. 10, 2007

14:27 SPI-B429



527

activity of cleaved and latent forms of the serpin.129 Cleaved and latent antithrombins have a greatly reduced affinity for heparin, but a synthetic pentasaccharide designed to bind native antithrombin with much higher affinity than the natural pentasaccharide130 was found to also bind cleaved and latent antithrombins with high affinity. This high affinity pentasaccharide was shown to block the inhibitory effects of cleaved or latent antithrombins in endothelial cell proliferation, migration, capillary-like tube formation, bFGF signaling and perlecan gene expression assays without affecting assays in the absence of antithrombin or in the presence of the native serpin. Moreover, mutation of Lys 114 when combined with modifications of Asn-linked carbohydrate chains, which reduce heparin affinity, was able to abolish the antiproliferative activity of the cleaved serpin. These findings together suggested that the heparin binding site of cleaved antithrombin is a critical mediator of its antiangiogenic activity, presumably because it makes a low affinity interaction with an endothelial cell-associated matrix or cell surface heparan sulfate.129 Most interesting, the Lys 114 mutation made in a higher heparin affinity glycoform showed the expression of antiangiogenic activity in both native and cleaved forms of the mutant, indicating that the conformational change was not necessary to express the antiangiogenic activity in this mutant. Since the Lys 114 mutation blocks conformational activation of the serpin and locks it in the native state in which the RCL is partially inserted into sheet A,78 this finding suggests that the critical determinant of antithrombin antiangioigenic activity is the partial or complete insertion of the RCL into sheet A and the conformational changes associated with this insertion. It follows that the inability of the native protein to express antiangiogenic activity results from the expulsion of the RCL, which accompanies heparin binding and activation. The heparininduced conformational change in native antithrombin may thus constitute an antiangiogenic switch that is responsible for activating the anticoagulant function and turning off the antiangiogenic function of the serpin.129 Whether an endothelial receptor in conjunction with a heparan sulfate coreceptor mediates the antiangiogenic effects of antithrombin and how antiangiogenic forms of the serpin block the proangiogenic effects of bFGF and VEGF on endothelial cells remains to be determined in future studies.

ch22

Jan. 10, 2007

14:27 SPI-B429

528


F. C. Church et al.

5.8. Heparin cofactor II, properties and function Heparin cofactor II mRNA has been detected only in human liver, and the normal concentration of HCII in blood plasma is ∼ 1µM.131 The mature protein is a single chain glycoprotein with an Mr ∼65,600.132 Heparin cofactor II has 480 amino acids with 3 Cys residues, but no disulfide bridges are apparently formed.133 The gene for HCII is located on chromosome 22 and has been mapped to 22q11.134 Heparin cofactor II is another vertebrate serpin, which is activated by the binding of glycosaminoglycans to specifically inhibit the blood clotting protease, thrombin. Whereas both heparin and dermatan sulfate glycosaminoglycans are thought to be physiologic activators, almost any polyanion, including polyphosphates, polysulfates, and polycarboxylates, will accelerate HCII inhibition of thrombin.25,135,136 This broad specificity can be explained by the glycosaminoglycan binding mechanism of HCII, whereby binding frees the amino terminus of HCII to act as a ligand for thrombin.

5.9. Reactive center loop and acidic domain of heparin cofactor II Heparin cofactor II is a unique thrombin inhibitor in that it contains Leu 444-Ser 445 as its reactive center bond,137,138 as opposed to other thrombin inhibitors such as AT or PCI, which contain an Arg-Ser reactive center bond. The P1 reactive center Leu residue, like that of α1 -antichymotrypsin, enables HCII to inhibit chymotrypsin and cathepsin G.30,139 The P1 L444R mutation of HCII shows a gain of antithrombotic activity in the absence of glycosaminoglycan and loses the ability to inhibit chymotrypsin. However, in the presence of glycosaminoglycan, the L444R HCII mutant has lower rates of antithrombin activity due to an increase in substrate character.140–142 Despite the non-ideal Leu reactive site sequence, HCII is still a potent thrombin inhibitor with inhibition rates increasing more than 10,000-fold in the presence of glycosaminoglycans. An N-terminal acidic region further contributes to the unique nature of HCII. The HCII acidic domain has a high degree of homology with the C-terminus of hirudin. The first of these HCII regions interact with anion-binding exosite-1 of thrombin (Fig. 2). The acidic domain is

ch22

Jan. 10, 2007

14:27 SPI-B429



529

contained in the first 75 amino acids of HCII.137,143,144 Within this region are two distinct clusters of acidic amino acids [acidic region-1 (AR-1) and acidic region-2 (AR-2)]: AR-1 is Glu 56-Asp 57-Asp 58-Asp 59-Tyr 60-Leu 61-Asp 62; and AR-2 is Glu 69-Asp 70-Asp 71-Asp 72-Tyr 73-Ile 74-Asp 75 (both Tyr residues are sulfated). Interestingly, deletion of the entire acidic domain did not drastically affect rates of thrombin inhibition by HCII in the absence of glycosaminoglycan.145 However, deletion of the acidic region resulted both in an increase in heparin–Sepharose affinity along with a significant loss of glycosaminoglycan accelerated thrombin inhibition. This suggests that the acidic domain of HCII mediates the glycosaminoglycan effect on thrombin inhibition. Recently, the acidic domain (including residues 1-75) of HCII was fused onto the amino-terminus of M358R α1protease inhibitor (α1-PI).205 This acidic domain- α1-PI chimera showed enhanced inhibition of α-thrombin, but not γ-thrombin missing exosite 1. These results suggest that by transferring the acidic region of HCII to α1PI, enhanced the ability of the chimera serpin to interact with exosite 1 of thrombin and this was revealed by enhanced inhibition rates.205 5.10. Glycosaminoglycan-binding site of heparin cofactor II The glycosaminoglycan binding site of HCII has been identified as the D helix region based on chemically modified, engineered and natural variants of HCII with glycosaminoglycan binding defects and on synthetic peptide studies. The D helix and N-terminal loop of HCII from Lys 173 to Phe 195 shows extensive homology to the Lys 114 to Lys 136 loop-D helix sequence of antithrombin, which binds heparin (Fig. 2). A synthetic peptide corresponding to HCII residues 165–195 blocked glycosaminoglycan-accelerated thrombin inhibition, and assumed a more α-helical structure in the presence of either heparin or dermatan sulfate, supporting this region as the glycosaminoglycan-binding site. A naturally occurring HCII mutant, HCIIOslo , with an Arg 189→His mutation was found to have a decreased affinity for dermatan sulfate but not for heparin, initially suggesting that the binding sites for heparin and dermatan sulfate were distinct.146 However, more extensive studies of natural and engineered HCII variants showed that the HCII binding sites for heparin

ch22

Jan. 10, 2007

14:27 SPI-B429

530


F. C. Church et al.

and dermatan sulfate are overlapping rather than distinct. Such studies have implicated Lys 173, Arg 184, Lys 185, and Arg 193 in heparin binding and Arg 184, Lys 185, Arg 189, Arg 192, and Arg 193 in dermatan sulfate binding.102,145,147–157 5.11. Mechanism of glycosaminoglycan activation of heparin cofactor II The predominant mechanism of glycosaminoglycan acceleration appears to be allosteric and involves conformational activation of the serpin. HCII possesses a unique amino terminal extension that contains two tandem repeats rich in acidic amino acids and containing two sulfated tyrosines (residues 56–75). Heparin and dermatan sulfate binding to HCII is thought to allosterically activate the serpin by displacing the acidic amino-terminus from an intramolecular interaction with the basic glycosaminoglycan binding site and freeing it for binding to thrombin exosite 1.145,150,152,153 In support of this mechanism, HCII mutants lacking the acidic domain and thrombin variants with a proteolytically altered or mutated exosite 1 show essentially no glycosaminoglycan-enhanced protease inhibition. Moreover, the N-terminally deleted HCII variants show enhanced glycosaminoglycan binding affinity in keeping with the glycosaminoglycan-binding site being more exposed in the absence of the N-terminal acidic domain interaction. An alternative allosteric mechanism has been more recently proposed, based on the recently described structures of native HCII and a Michaelistype complex of HCII with catalytically inactive S195A thrombin.102 While the structure of the complex verifies the interaction of the HCII N-terminal acidic domain with thrombin exosite 1 as proposed from the studies of HCII and thrombin variants, most of the N-terminus is not visible in the structure of the free serpin precluding any assessment of the validity of the proposed interaction of the N-terminus with the glycosaminoglycan binding site prior to activation. Surprisingly, the native HCII structure revealed that the reactive loop hinge is partially inserted in β-sheet A in a manner similar to that in native antithrombin and the short segment of the N-terminus that was visible suggested that it might be interacting with an alternative basic site on the serpin near the reactive loop. Glycosaminoglycan activation of HCII was thus proposed to resemble that of antithrombin wherein glycosaminoglycan binding to the D helix causes the expulsion of the buried

ch22

Jan. 10, 2007

14:27 SPI-B429



531

loop hinge from sheet A, which in turn disrupts the N-terminal tail interaction to promote binding to thrombin exosite 1 (Fig. 2). This allosteric process enhances the rate of HCII inhibition of thrombin several thousandfold to the diffusion limit by utilizing the N-terminal acidic domain interaction to position thrombin for optimal interaction with the HCII reactive center loop. Using equilibrium binding and kinetic studies, another alternative pathway was proposed for HCII inhibition of thrombin-heparin and thrombin-dermatan sulfate complexes.206 These studies revealed that the predominant species formed with glycosaminoglycan were with thrombin, which then allowed enhanced HCII inhibition to proceed.206 Thus, it was proposed that high molecular weight glycosaminoglycan acts as a template to form a ternary complex, preceding covalent thrombin-HCII complex stabilization. This preferred pathway of ternary complex formation would proceed through a binary thrombin-glycosaminoglycan intermediate. Binding of glycosaminoglycan in this thrombin-glycosaminoglycan complex to free HCII might further promote the allosteric interaction between acidic domain of HCII and thrombin exosite 1.206 5.12. Extravascular thrombin regulation by heparin cofactor II There is evidence that dermatan sulfate proteoglycans (DSPGs) on the surface of cultured fibroblasts and vascular smooth muscle cells and purified biglycan and decorin DSPGs bound to immobilized collagen type V accelerate the rate of thrombin inhibition by HCII.31,135,158–164 These findings suggest that HCII has a major role in thrombin regulation at extravascular tissue sites following vessel injury.162 DSPGs in the extracellular matrices and on certain cell surfaces may localize HCII to sites appropriate for inhibiting not only the coagulant activity of thrombin but also its mitogenic and chemotactic activities.160,163 5.13. Protein C inhibitor, properties and function Normal PCI concentration in human blood plasma is ∼90 nM,165 it is also in urine and other body fluids (e.g. tears, saliva, cerebral spinal fluid, and amniotic fluid), and it is in seminal fluid at ∼200 µg/ml, which is almost 40 times the amount in plasma.35,166–173 The PCI gene is on

ch22

Jan. 10, 2007

14:27 SPI-B429

532


F. C. Church et al.

chromosome 14 in a gene cluster at 14q32.1, which also includes other serpin genes, α1 -antitrypsin, α1 -antichymotrypsin, and corticosteroid-binding globulin.174,175 The mature protein is a single chain glycoprotein with Mr ∼ 57, 000.33,176 Protein C inhibitor has 387 amino acids, with a single Cys and no intramolecular disulfide bonds. With a reactive center bond of Arg 354-Ser 355, PCI has a broad protease inhibition profile, including trypsin, thrombin, activated protein C, acrosin, kallikrein, uPA, tPA, factor Xa, and factor XIa.33,34,38,39,166–168,171,176–186 Recently, a type II transmembrane serine protease encoded by the mouse differentially expressed in squamous-cell carcinoma (DESC) 1 gene was found to form a stable bimolecular complex with PCI.187 5.14. Glycosaminoglycan-binding site of protein C inhibitor A variety of glycosaminoglycans and polyanions accelerate activated protein C inhibition by PCI and there is no evidence for any sequence-specific binding of heparin to PCI. Binding of heparin to PCI has been studied by chromatography of the serpin on heparin-agarose or from the effect of heparin on the kinetics of PCI inhibition of proteases. The heparin-binding site in protein C inhibitor appears to be localized not to the D helix as in antithrombin and HCII, but to the H helix region with possible contributions from the N-terminal A+ helix region (Fig. 2).37,188–191 Both regions have sequences of basic residues consistent with a general heparin-binding motif. The recent crystal structure of cleaved PCI revealed that the amino terminal region is near the H helix.41 Although synthetic peptides corresponding to the A+ helix region were unable to bind heparin, an H helix peptide did bind and competed with PCI for heparin binding. Other studies in which the H helix was deleted in a recombinant fragment of PCI residues 1 to 294, expressed as a fusion protein in Escherichia coli, suggested a primary role for the H helix in heparin binding. Mutagenesis of all four basic residues in the H helix of PCI, Lys 266, Arg 269, Lys 270, and Lys 273, in a fulllength recombinant PCI has shown that all of these residues are important for heparin binding. Mutation of three basic residues in the D helix region of PCI has shown that these residues also contribute to binding heparin, but they are not necessary for heparin accelerated inhibition of either thrombin

ch22

Jan. 10, 2007

14:27 SPI-B429



533

or activated protein C. Thus, the heparin-binding site of PCI is centered on the H helix region with contribution most likely from other basic residues within this vicinity (including Arg 26, Arg 27, Arg 213, Arg 234, Arg 229, Lys 255, and Arg 362).41 To date, the physiologic proteoglycan responsible for acceleration of PCIs activity in vivo has not been clearly identified.

5.15. Mechanism of glycosaminoglycan acceleration of protein C inhibitor Heparin binds and activates the serpin, PCI, to inhibit the anticoagulant protease, activated protein C, in blood and the sperm protease, acrosin, in the seminal fluid. There is no evidence for an allosteric activation mechanism in PCI and instead the activation appears to involve only a ternary complex bridging or template effect of heparin.36,40,41,135,167,177,178,192 As with other proteases containing γ-carboxyglutamic acid (Gla) domains, heparin bridging of PCI and activated protein C is only modest unless calcium ions are present to bind the Gla-domain and prevent its interaction with the heparin-binding site of the protease.40,185,186 Thrombin is also inhibited by PCI and the inhibition is accelerated by heparin, but even the heparin-enhanced rate does not appear to be physiologically significant. A more physiologically relevant rate of thrombin inhibition by PCI results when thrombin is bound to the endothelial cell receptor, TM.38,39,42,182 This is consistent with PCI regulating the anticoagulant protein C pathway since thrombin binding to TM initiates this pathway by transforming the enzyme into an activator of protein C. Interestingly, the increase in PCI reactivity with thrombin upon binding to TM does not involve an associated chondroitin sulfate glycosaminoglycan chain of TM but rather is mediated by the EGF domains of the receptor. Whether PCI interactions with the receptor contribute to this rate enhancement in a manner similar to the receptor promotion of protein C activation by bound thrombin is presently not well understood. To study the mechanism of how TM accelerates thrombin inhibition by PCl and to probe the specificity of the PCl H-helix region for this complex, an antithrombin mutant was made where its H-helix region was replaced with that from PCl.42 From these studies, Yang et al. showed that TM accelerates the PCl-thrombin inhibition reaction

ch22

Jan. 10, 2007

14:27 SPI-B429


534

F. C. Church et al.

by not only modulating the extended binding pocket of the protease but also “substrate presentation” mechanism.42

6. Serpin Knockout and Transgenic Mice Much new knowledge has been obtained using gene ablation technology for many genes, including serpins. We present a short review of the mouse models and their phenotypes from removal of the antithrombin, HCII, and PCI genes, and related transgenic mouse models. 6.1. Antithrombin The antithrombin gene in mice was inactivated by replacement of exon 2 with the neomycin resistance gene (neo).193 AT+/− ×AT+/+ breeding pairs produced AT+/− offspring at approximately the expected mendelian frequency (50%), indicating that heterozygous antithrombin deficiency is compatible with normal fetal development. AT+/− mice were morphologically indistinguishable from wild-type mice and, despite having only 50% of normal antithrombin antigen and activity in their plasma, had no sign of spontaneous thrombosis up to 14 months of age.194 Female AT+/− mice were fertile and did not develop thrombosis during pregnancy. When mice were challenged by intraperitoneal injection of endotoxin, more fibrin deposition occurred in renal glomeruli, liver sinusoids, and small myocardial blood vessels in AT+/− mice than in wild-type animals. Restraint for 20 h in a narrow conical tube also caused greater renal fibrin deposition in the AT+/− mice. No fibrin deposition was observed in unchallenged AT+/− mice. No live AT−/− pups were born to heterozygous (AT+/− ×AT+/− ) breeding pairs.193 AT−/− embryos died in late gestation (day 15.5– 16.5) with signs of extensive subcutaneous and intracranial hemorrhage. Widespread tissue degeneration was observed with evidence of fibrin(ogen) deposition in the myocardium and liver but not in the brain, lungs, or kidneys. Fibrin(ogen) was undetectable in areas of hemorrhage, suggesting that plasma fibrinogen levels in the embryo were decreased due to consumption or hepatic insufficiency. Antithrombin is generally thought to be activated by endogenous heparan sulfate (HS) chains containing 3-O-sulfated glucosamine residues.

ch22

Jan. 10, 2007

14:27 SPI-B429



535

The rate-limiting enzyme for biosynthesis of these residues is glucosaminyl 3-O-sulfotransferase-1 (3-OST-1). Mice lacking the 3-OST-1 gene had a large reduction (75–98%) in the ability of extracts from most tissues to accelerate inhibition of factor Xa by antithrombin.195 The residual activity was possibly due to modification of HS by another 3-O-sulfotransferase (e.g. 3-OST-5). Unexpectedly, these mice had no increase in the rate of fibrin deposition in tissues under normal or hypoxemic conditions, nor did they have accelerated thrombosis in a carotid arterial injury model, despite the absence of high-affinity binding sites for 125 I-labeled AT in the vessel wall. Whereas 3-OST-1-deficient mice appeared normal on a mixed C57BL/6×129S4/SvJae genetic background, intrauterine growth retardation and early postnatal lethality were observed on a pure C57BL/6 background. Although the cause of neonatal mortality in these mice was unexplained, histologic examination did not reveal evidence of the focal thrombosis observed in AT−/− embryos. The lack of a prothrombotic phenotype in 3-OST-1 knockout mice challenges the concept that 3-O-sulfate modified HS plays a major role in hemostasis. Low levels of 3-O-sulfation produced by another enzyme, however, may be sufficient to activate AT physiologically. Alternatively, antithrombin may be activated by unmodified HS, which binds to antithrombin with a much lower affinity but may be present at much higher concentrations in blood vessels relative to 3O-sulfated HS. In contrast, to study the function of the heparin binding site of antithrombin in vivo, a transgenic mouse with the antithrombin heparin-binding site mutant R48C was generated.196 Mice homozygous for the antithrombin mutation developed life-threatening thrombosis, presumably due to the inability of antithrombin to interact with heparin/heparan sulfate.196 These results suggest an important in vivo function of the heparin binding site of antithrombin, which is different from that found in the 3-OST-1 knockout mice described earlier. 6.2. Heparin cofactor II The murine HCII gene was disrupted by replacement of exon 1 with the neo gene.197 HCII−/− pups were born at the expected mendelian frequency (25%) to heterozygous (HCII+/− ×HCII+/− ) breeding pairs, and they had normal growth and survival up to at least 1 year of age. HCII−/− male

ch22

Jan. 10, 2007

14:27 SPI-B429

536


F. C. Church et al.

and female mice were fertile and produced litters of normal size. Although HCII−/− mice showed no evidence of spontaneous thrombosis, formation of an occlusive thrombus in the carotid artery following photochemically induced endothelial injury was significantly faster in HCII−/− mice than in wild-type controls. The antithrombotic effect may result from activation of HCII by dermatan sulfate in the wall of the injured artery. To test this hypothesis, HCII−/− mice were injected prior to arterial injury with recombinant native HCII or with HCII mutants having decreased affinity for either heparin (K173Q) or dermatan sulfate (R189H). Both native HCII and the mutant with defective heparin binding restored the thrombotic occlusion time to normal, while the mutant with defective dermatan sulfate binding did not [He, Vicente and Tollefsen, unpublished observations]. These results provide evidence for a physiologic interaction between HCII and dermatan sulfate present in the arterial wall. Although dermatan sulfate-dependent thrombin inhibitory activity and HCII antigen were both undetectable in the plasma of HCII−/− mice, plasma AT activity was normal. Intravenous administration of dermatan sulfate prolonged the carotid occlusion time in wild-type mice but not in HCII−/− mice, indicating that the antithrombotic effect of dermatan sulfate is mediated primarily by HCII.198 In contrast, heparin prolonged the occlusion time equally in wild-type and HCII−/− mice, consistent with its ability to activate AT. 6.3. Protein C inhibitor Synthesis of PCI in humans occurs in many organs and tissues, predominantly the liver, kidneys, and reproductive organs. PCI expression is notably absent in mouse liver, and PCI antigen and activity are undetectable in mouse plasma. This important species difference has complicated investigation of the physiology of PCI in mice. Disruption of the PCI gene in mice was accomplished by replacement of exons 2–5 with the neo gene.173 Heterozygous breeding pairs (PCI+/− ×PCI+/− ) yielded homozygous PCI−/− pups at the expected Mendelian frequency (25%), which indicates that PCI deficiency does not impair embryogenesis. PCI−/− mice grew and survived normally and did not show evidence of abnormal hemostasis. Whereas PCI−/− female mice

ch22

Jan. 10, 2007

14:27 SPI-B429



537

were fertile and produced litters of normal size, PCI−/− males were infertile. Sperm from PCI−/− males were morphologically abnormal and were incapable of fertilizing wild-type oocytes in vitro. Histologic examination of the testis revealed destruction of Sertoli cells lining the seminiferous tubules, which is most likely to have resulted in the premature release and degeneration of spermatocytes. The precise cause of the Sertoli cell dysfunction is unclear. Excessive proteolysis in the absence of PCI may play a role, since increased amidolytic activity was demonstrated in extracts of testes from PCI−/− mice. To investigate the function of plasma PCI, two groups have generated transgenic mice expressing human PCI. The first group inserted the human PCI cDNA into a vector containing mouse albumin enhancer/promoter sequences that direct liver-specific expression.199 Homozygous transgenic mice expressed human PCI in their plasma at concentrations around 10 µg/ml, which was about twofold higher than the concentration of PCI in human plasma. These mice were normal in development and reproduction, had no apparent hemostatic abnormalities, and did not differ from wild-type mice in cytokine production or mortality after endotoxin administration. The second group constructed transgenic mice using the entire 15-kilobase human PCI gene and reported widespread expression of the transgene in liver, kidneys, brain, lungs, and reproductive organs.200 No abnormalities of development or reproduction were observed. In contrast to the previous study, these mice had higher levels of human PCI in their plasma (about 18 µg/ml), and they showed signs of enhanced disseminated intravascular coagulation and tissue damage in response to endotoxemia manifested by greater prolongation of the activated partial thromboplastin time, depression of antithrombin and fibrinogen levels, production of tumor necrosis factorα and interleukin-6, and release of hepatocellular enzymes in comparison with wild-type controls. The authors presented evidence that human PCI in transgenic mouse plasma can inhibit endogenously generated mouse activated protein C, which may explain the deleterious effects of PCI observed in the endotoxemia model and is consistent with the beneficial effects of activated protein C in humans with severe sepsis. It is unknown whether human PCI can also efficiently inhibit murine thrombin–TM, procoagulant proteases, or fibrinolytic proteases and to what extent inhibition of these enzymes might affect the phenotype of PCI transgenic mice.

ch22

Jan. 10, 2007

14:27 SPI-B429


538

F. C. Church et al.

Acknowledgments This work was supported by Research Grants HL-32656 (to FCC), HL55520 (to DMT), and HL-39888 and HL-64013 (to STO) from the National Institutes of Health, by BCTR0503475 from the Susan G. Komen Breast Cancer Foundation (to FCC), and by the National Health & Medical Research Council of Australia, the Australian Research Council, the National Heart Foundation of Australia (to RNP and AMB).

References 1. Mann KG (1999) Biochemistry and physiology of blood coagulation. Thromb Haemost 82:165–174. 2. Colman RW, Hirsh J, Marder VJ, Clowes AW and George JN (2001) Hemostasis and Thrombosis: Basic Principles and Clinical Practice, 4th ed. Lippencott, Williams and Wilkins: Philadelphia. 3. Dahlback B (2000) Blood coagulation. Lancet 355:1627–1632. 4. Hoffman MM and Monroe DM (2005) Rethinking the coagulation cascade. Curr Hematol Rep 4(5):391–396. 5. Broze GJ Jr (2003) The rediscovery and isolation of TFPI. J Thromb Haemost 1(8):1671–1675. 6. Huber R and Carrell RW (1989) Implications of the three-dimensional structure of alpha1-antitrypsin for structure and function of serpins. Biochemistry 28:8951–8966. 7. Gettins PGW, Patston PA and Olson ST (1996) Serpins: Structure, Function and Biology, 1st ed. Austin T, Landes, RG Publishers: 202 p. 8. Church FC, Cunningham DD, Ginsburg D, Hoffman M, Stone SR and Tollefsen DME (1997) Chemistry and Biology of Serpins. Advances in Experimental Medicine and Biology Plenum Press: New York, p. 358. 9. Silverman GA, Bird PI, Carrell RW, et al. (2001) The serpins are an expanding superfamily of structurally similar but functionally diverse proteins: evolution, mechanism of inhibition, novel functions, and a revised nomenclature. J Biol Chem 276(36):33293–33296. 10. Gettins PG (2002) Serpin structure, mechanism, and function. Chem Rev 102(12):4751–4804. 11. Esmon CT (2003) The protein C pathway. Chest 124(Suppl 3):26S–32S. 12. Esmon CT (2005) Coagulation inhibitors in inflammation. Biochem Soc Trans 33(Pt 2):401–405. 13. Esmon CT and Owen WG (2004) The discovery of thrombomodulin. J Thromb Haemost 2(2):209–213.

ch22

Jan. 10, 2007

14:27 SPI-B429



539

14. Dahlback B (2005) Blood coagulation and its regulation by anticoagulant pathways: genetic pathogenesis of bleeding and thrombotic diseases. J Intern Med 257(3):209–223. 15. Dahlback B and Villoutreix BO (2005) Regulation of blood coagulation by the protein C anticoagulant pathway: Novel insights into structure-function relationships and molecular recognition. Arterioscler Thromb Vasc Biol 25(7):1311–1320. 16. De Cicco M (2004) The prothrombotic state in cancer: Pathogenic mechanisms. Crit Rev Oncol Hematol 50(3):187–196. 17. Hertig A and Rondeau E (2004) Role of the coagulation/fibrinolysis system in fibrin-associated glomerular injury. J Am Soc Nephrol 15(4): 844–853. 18. Stefansson S, McMahon GA, Petitclerc E and Lawrence DA (2003) Plasminogen activator inhibitor-1 in tumor growth, angiogenesis and vascular remodeling. Curr Pharm Des 9(19):1545–1564. 19. Cesarman-Maus G and Hajjar KA (2005) Molecular mechanisms of fibrinolysis. Br J Haematol 129(3):307–321. 20. Stefansson S and Lawrence DA (2003) Old dogs and new tricks: Proteases, inhibitors, and cell migration. Sci STKE 2003(189):pe24. 21. Linkins LA and Weitz JI (2005) New anticoagulant therapy. Annu Rev Med 56:63–77. 22. Weitz JI and Bates SM (2005) New anticoagulants. J Thromb Haemost 3(8):1843–1853. 23. Weitz JI, Middeldorp S, Geerts W and Heit JA (2004) Thrombophilia and new anticoagulant drugs. Hematology (Am Soc Hematol Educ Program) 424–438. 24. Huntington JA (2003) Mechanisms of glycosaminoglycan activation of the serpins in hemostasis. J Thromb Haemost 1(7):1535–1549. 25. Tollefsen DM (1997) Heparin cofactor II. Adv Exp Med Biol 425 (in Church FC, Cunningham DD, Ginsburg D, Hoffman M, Stone SR and Tollefsen DM (eds.) Chemistry and Biology of Serpins): 35–44. 26. Kounnas MZ, Church FC, Argraves WS and Strickland DK (1996) Cellular internalization and degradation of antithrombin III-thrombin, heparin cofactor II-thrombin, and alpha1-antitrypsin-trypsin complexes is mediated by the low density lipoprotein receptor-related protein. J Biol Chem 271(11): 6523–6529. 27. Björk I and Olson ST (1997) Antithrombin: A bloody important serpin. Adv Exp Med Biol 425 (in Church, FC, Cunningham DD, Ginsburg D, Hoffman M, Stone SR and Tollefsen DM (eds.) Chemistry and Biology of Serpins): 17–33.

ch22

Jan. 10, 2007

14:27 SPI-B429

540


F. C. Church et al.

28. Buchanan MR, Boneu B, Ofosu F and Hirsh J (1985) The relative importance of thrombin inhibition and factor Xa inhibition to the antithrombotic effects of heparin. Blood 65(1):198–201. 29. Olson ST, Swanson R, Raub-Segall E, et al. (2004) Accelerating ability of synthetic oligosaccharides on antithrombin inhibition of proteinases of the clotting and fibrinolytic systems. Comparison with heparin and lowmolecular-weight heparin. Thromb Haemost 92(5):929–939. 30. Parker KA and Tollefsen DM (1985) The protease specificity of heparin cofactor II. Inhibition of thrombin generated during coagulation. J Biol Chem 260(6):3501–3505. 31. Shirk RA, Parthasarathy N, San Antonio JD, Church FC and Wagner WD (2000) Altered dermatan sulfate structure and reduced heparin cofactor IIstimulating activity of biglycan and decorin from human atherosclerotic plaque. J Biol Chem 275(24):18085–18092. 32. Tollefsen DM (2002) Heparin cofactor II deficiency. Arch Pathol Lab Med 126(11):1394–1400. 33. Suzuki K, Nishioka J and Hashimoto S (1983) Protein C inhibitor: Purification from human plasma and characterization. J Biol Chem 258:163–168. 34. Suzuki K (1985) Biochemistry and physiology of protein C inhibitor. 43–57. 35. Laurell M and Stenflo J (1989) Protein C inhibitor from human plasma: Characterization of native and cleaved inhibitor and demonstration of inhibitor complexes with plasma kallikrein. Thromb Haemostasis 62(3):885–891. 36. Pratt CW and Church FC (1992) Heparin binding to protein C inhibitor. J Biol Chem 267:8789–8794. 37. Cooper ST, Whinna HC, Jackson TP, Boyd JM and Church FC (1995) Intermolecular interactions between protein C inhibitor and coagulation proteases. Biochemistry 34:12991–12997. 38. Rezaie AR, Cooper ST, Church FC and Esmon CT (1995) Protein C inhibitor is a potent inhibitor of the thrombin-thrombomodulin complex. J Biol Chem 270:25336–25339. 39. Mosnier LO, Elisen MG, Bouma BN and Meijers JC (2001) Protein C inhibitor regulates the thrombin-thrombomodulin complex in the up- and down regulation of TAFI activation. Thromb Haemost 86(4):1057–1064. 40. Glasscock LN, Gerlitz B, Cooper ST, Grinnell BW and Church FC (2003) Basic residues in the 37-loop of activated protein C modulate inhibition by protein C inhibitor but not by alpha(1)-antitrypsin. Biochim Biophys Acta 1649(1):106–117. 41. Huntington JA, Kjellberg M and Stenflo J (2003) Crystal structure of protein C inhibitor provides insights into hormone binding and heparin activation. Structure (Camb) 11(2):205–215.

ch22

Jan. 10, 2007

14:27 SPI-B429



541

42. Yang L, Manithody C, Walston TD, Cooper ST and Rezaie AR (2003) Thrombomodulin enhances the reactivity of thrombin with protein C inhibitor by providing both a binding site for the serpin and allosterically modulating the activity of thrombin. J Biol Chem 278(39):37465–37470. 43. Andreasen PA, Kjoller L, Christensen L and Duffy MJ (1997) The urokinasetype plasminogen activator system in cancer metastasis: A review. Int J Cancer 72:1–22. 44. Andreason PA, Egelund R and Peterson HH (2000) The plasminogen activation system in tumor growth, invasion, and metastasis. Cellular Mol Life Sci 57:25–40. 45. DeClerck YA and Imren S (1994) Protease inhibitors: Role and potential therapeutic use in human cancer. Eur J Cancer 30A(14):2170–2180. 46. Hekman CM and Loskutoff DJ (1985) Endothelial cells produce a latent inhibitor of plasminogen activators that can be activated by denaturants. J Biol Chem 260(21):11581–11587. 47. Mimuro J, Sawdey M, Hattori M and Luskutoff DJ (1989) cDNA for bovine type 1 plasminogen activator inhibitor (PAI-1). Nucleic Acids Res 17(21):8872. 48. Wiman B, Almquist A, Sigurdardottir O and Lindahl T (1988) Plasminogen activator inhibitor 1 (PAI) is bound to vitronectin in plasma. FEBS Lett 242(1):125–128. 49. Dougherty KM, Pearson JM, Yang AY, Westrick RJ, Baker MS and Ginsburg D (1999) The plasminogen activator inhibitor-2 gene is not required for normal murine development or survival. Proc Natl Acad Sci USA 96: 686–691. 50. Antalis TM, Linn ML, Donnan K, et al. (1998) The serine protease inhibitor (serpin) plasminogen activation inhibitor type 2 protects against viral cytopathic effects by constitutive interferon alpha/beta priming. J Exp Med 187(11):1799–1811. 51. van Boven HH and Lane DA (1997) Antithrombin and its inherited deficiency states. Semin Hematol 34(3):188–204. 52. Bock SC, Wion KL, Vehar GA and Lawn RM (1982) Cloning and expression of the cDNA for human antithrombin III. Nucleic Acids Res 10(24): 8113–8125. 53. Peterson CB and Blackburn MN (1985) Isolation and characterization of an antithrombin III variant with reduced carbohydrate content and enhanced heparin binding. J Biol Chem 260(1):610–615. 54. Irving JA, Pike RN, Lesk AM and Whisstock JC (2000) Phylogeny of the serpin superfamily: implications of patterns of amino acid conservation for structure and function. Genome Res 10(12):1845–1864.

ch22

Jan. 10, 2007

14:27 SPI-B429

542


F. C. Church et al.

55. Petitou M, Casu B and Lindahl U (2003) 1976–1983, a critical period in the history of heparin: the discovery of the antithrombin binding site. Biochimie 85(1–2):83–89. 56. Marcum JA, Atha DH, Fritze LMS, Nawroth P, Stern D and Rosenberg RD (1986) Cloned bovine aortic endothelial cells synthesize anticoagulantly active heparan sulfate proteoglycan. J Biol Chem 261(16):7507–7517. 57. Olson ST, Bjork I, Sheffer R, Craig PA, Shore JD and Choay J (1992) Role of the antithrombin-binding pentasaccharide in heparin acceleration of antithrombin-proteinase reactions. Resolution of the antithrombin conformational change contribution to heparin rate enhancement. J Biol Chem 267:12528–12538. 58. Rezaie AR (1998) Calcium enhances heparin catalysis of the antithrombinfactor Xa reaction by a template mechanism: Evidence that calcium alleviates Gla domain antagonism of heparin binding to factor Xa. J Biol Chem 273(27):16824–16827. 59. Bedsted T, Swanson R, Chuang YJ, Bock PE, Bjork I and Olson ST (2003) Heparin and calcium ions dramatically enhance antithrombin reactivity with factor IXa by generating new interaction exosites. Biochemistry 42(27): 8143–8152. 60. Choay J, Petitou M, Lormeau JC, Sinay P, Casu B and Gatti G (1983) Structure-activity relationship in heparin: A synthetic pentasaccharide with high affinity for antithrombin III and eliciting high anti-factor Xa activity. Biochem Biophys Res Commun 116(2):492–499. 61. Thunberg L, Bäckström G and Lindahl U (1982) Further characterization of the antithrombin-binding sequence in heparin. Carbohydr Res 100: 393–410. 62. Streusand VJ, Bjork I, Gettins PG, Petitou M and Olson ST (1995) Mechanism of acceleration of antithrombin-proteinase reactions by low affinity heparin. Role of the antithrombin binding pentasaccharide in heparin rate enhancement. J Biol Chem 270(16):9043–9051. 63. de Agostini AI, Watkins SC, Slayter HS, Youssoufian H and Rosenberg RD (1990) Localization of anticoagulantly active heparin sulfate proteoglycans in vascular endothelium: antithrombin binding on cultured endothelial cells and perfused rat aorta. J Cell Biol 111(3):1293–1304. 64. Desai UR, Petitou M, Bjork I and Olson ST (1998) Mechanism of heparin activation of antithrombin. Role of individual residues of the pentasaccharide activating sequence in the recognition of native and activated states of antithrombin. J Biol Chem 273(13):7478–7487.

ch22

Jan. 10, 2007

14:27 SPI-B429



543

65. Olson ST, Srinivasan KR, Bjork I and Shore JD (1981) Binding of high affinity heparin to antithrombin III. Stopped flow kinetic studies of the binding interaction. J Biol Chem 256(21):11073–11079. 66. Petitou M, Herault JP, BernatA, et al. (1999) Synthesis of thrombin-inhibiting heparin mimetics without side effects. Nature 398(6726):417–422. 67. Danielsson A, Raub E, Lindahl U and Bjork I (1986) Role of ternary complexes, in which heparin binds both antithrombin and proteinase, in the acceleration of the reactions between antithrombin and thrombin or factor Xa. J Biol Chem 261(33):15467–15473. 68. Olson ST and Bjork I (1991) Predominant contribution of surface approximation to the mechanism of heparin acceleration of the antithrombinthrombin reaction. Elucidation from salt concentration effects. J Biol Chem 266(10):6353–6364. 69. Sheehan JP and Sadler JE (1994) Molecular mapping of the heparin-binding exosite of thrombin. Proc Natl Acad Sci USA 91:5518–5522. 70. Gan ZR, Li Y, Chen Z, Lewis SD and Shafer JA (1994) Identification of basic amino acid residues in thrombin essential for heparin-catalyzed inactivation by antithrombin III. J Biol Chem 269(2):1301–1305. 71. Rezaie AR and Olson ST (2000) Calcium enhances heparin catalysis of the antithrombin-factor Xa reaction by promoting the assembly of an intermediate heparin-antithrombin-factor Xa bridging complex. Demostration by rapid kinetics studies. Biochemistry 39:12083–12090. 72. Yang L, Manithody C and Rezaie AR (2002) Localization of the heparin binding exosite of factor IXa. J Biol Chem 277(52):50756–50760. 73. Jin L, Abrahams JP, Skinner R, Petitou M and Pike RN (1997) The anticoagulant activation of antithrombin by heparin. Proc Natl Acad Sci USA 94:14683–14688. 74. Skinner R, Abrahams JP, Whisstock JC, Lesk AM, Carrell RW and Wardell MR (1997) The 2.6A structure of antithrombin indicates a conformational change at the heparin binding site. J Mol Biol 266:601–609. 75. Olson ST, Bjork I and Bock SC (2000) Identification of critical molecular interactions mediating heparin activation of antithrombin: Implications for the design of improved heparin anticoagulants. Trends Cardiovasc Med 12(5):198–205. 76. Schedin-Weiss S, Desai UR, Bock SC, Gettins PG, Olson ST and Bjork I (2002) Importance of lysine 125 for heparin binding and activation of antithrombin. Biochemistry 41(15):4779–4788. 77. Desai U, Swanson R, Bock SC, Bjork I and Olson ST (2000) Role of arginine 129 in heparin binding and activation of antithrombin. J Biol Chem 275(25):18976–18984.

ch22

Jan. 10, 2007

14:27 SPI-B429

544


F. C. Church et al.

78. Arocas V, Bock SC, Raja S, Olson ST and Bjork I (2001) Lysine 114 of antithrombin is of crucial importance for the affinity and kinetics of heparin pentasaccharide binding. J Biol Chem 276(47):43809–43817. 79. Schedin-Weiss S, Desai UR, Bock SC, Olson ST and Bjork I (2004) Roles of N-terminal region residues Lys11, Arg13, and Arg24 of antithrombin in heparin recognition and in promotion and stabilization of the heparin-induced conformational change. Biochemistry 43(3):675–683. 80. Arocas V, Bock SC, Olson ST and Bjork I (1999) The role of Arg46 and Arg47 of antithrombin in heparin binding. Biochemistry 38(31):10196–10204. 81. Jairajpuri MA, Lu A, Desai U, Olson ST, Bjork I and Bock SC (2003) Antithrombin III phenylalanines 122 and 121 contribute to its high affinity for heparin and its conformational activation. J Biol Chem 278(18): 15941–15950. 82. Monien BH, Krishnasamy C, Olson ST and Desai UR (2005) Importance of tryptophan 49 of antithrombin in heparin binding and conformational activation. Biochemistry 44(35):11660–11668. 83. Fitton HL, Skinner R, Dafforn TR, Jin L and Pike RN (1998) The N-terminal segment of antithrombin acts as a steric gate for the binding of heparin. Protein Sci 7(3):782–788. 84. Olson ST and Chuang YJ (2002) Heparin activates antithrombin anticoagulant function by generating new interaction sites (exosites) for blood clotting proteinases. Trends Cardiovasc Med 12(8):331–338. 85. Schreiber G and Fersht AR (1996) Rapid, electrostatically assisted association of proteins. Nat Struct Biol 3(5):427–431. 86. Carrell RW, Stein PE, Fermi G and Wardell MR (1994) Biological implications of a 3A structure of dimeric antithrombin. Structure 2:257–270. 87. Schreuder HA, de Boer B, Dijkema R, et al. (1994) The intact and cleaved human antithrombin III complex as a model for serpin-proteinase interactions. Nat Struct Biol 1(1):48–54. 88. Stein PE and Carrell RW (1995) What do dysfunctional serpins tell us about molecular mobility and disease? Struct Biol 2(2):96–113. 89. Owen MC, Beresford CH and Carrell RW (1988) Antithrombin Glasgow, 393 Arg to His: a P1 reactive site variant with increased heparin afinity but no thrombin inhibitory activity. FEBS Lett 231:317–320. 90. Futamura A and Gettins PG (2000) Serine 380 (P14) → glutamate mutation activates antithrombin as an inhibitor of factor Xa. J Biol Chem 275(6): 4092–4098. 91. Huntington JA, Olson ST, Fan B and Gettins PG (1996) Mechanism of heparin activation of antithrombin. Evidence for reactive center loop

ch22

Jan. 10, 2007

14:27 SPI-B429



92.

93.

94.

95.

96.

97.

98.

99.

100.

101.

545

preinsertion with expulsion upon heparin binding. Biochemistry 35(26): 8495–8503. Grootenhuis PDJ and Van Boeckel CAA (1991) Constructing a molecular model of the interaction between antithrombin III and a potent heparin analogue. J Am Chem Soc 113:2743–2747. Pike RN, Potempa J, Skinner R, et al. (1997) Heparin-dependent modification of the reactive center arginine of antithrombin and consequent increase in heparin binding affinity. J Biol Chem 272(32):19652–19655. Chuang YJ, Swanson R, Raja SM, Bock SC and Olson ST (2001) The antithrombin P1 residue is important for target proteinase specificity but not for heparin activation of the serpin. Characterization of P1 antithrombin variants with altered proteinase specificity but normal heparin activation. Biochemistry 40(22):6670–6679. Chuang YJ, Swanson R, Raja SM and Olson ST (2001) Heparin enhances the specificity of antithrombin for thrombin and factor Xa independent of the reactive center loop sequence. Evidence for an exosite determinant of factor Xa specificity in heparin-activated antithrombin. J Biol Chem 276(18):14961–14971. Futamura A, Beechem JM and Gettins PG (2001) Conformational equilibrium of the reactive center loop of antithrombin examined by steady state and time-resolved fluorescence measurements: Consequences for the mechanism of factor Xa inhibition by antithrombin-heparin complexes. Biochemistry 40(22):6680–6687. Gettins PGW, Fan B, Crews BC, Turko IV, Olson ST and Streusand VJ (1993) Transmission of conformational change from the heparin binding site to the reactive center of antithrombin. Biochemistry 32:8385–8389. Dementiev A, Petitou M, Herbert JM and Gettins PG (2004) The ternary complex of antithrombin-anhydrothrombin-heparin reveals the basis of inhibitor specificity. Nat Struct Mol Biol 11(9):863–867. Li W, Johnson DJ, Esmon CT and Huntington JA (2004) Structure of the antithrombin-thrombin-heparin ternary complex reveals the antithrombotic mechanism of heparin. Nat Struct Mol Biol 11(9):857–862. Dementiev A, Simonovic M, Volz K and Gettins PG (2003) Canonical inhibitor-like interactions explain reactivity of alpha1-proteinase inhibitor Pittsburgh and antithrombin with proteinases. J Biol Chem 278(39): 37881–37887. Ye S, Cech AL, Belmares R, et al. (2001) The structure of a Michaelis serpinprotease complex. Nat Struct Biol 8(11):979–983.

ch22

Jan. 10, 2007

14:27 SPI-B429

546


F. C. Church et al.

102. Baglin TP, Carrell RW, Church FC, Esmon CT and Huntington JA (2002) Crystal structures of native and thrombin-complexed heparin cofactor II reveal a multistep allosteric mechanism. Proc Natl Acad Sci USA 99(17):11079–11084. 103. Quinsey NS, Whisstock JC, Le Bonniec B, Louvain V, Bottomley SP and Pike RN (2002) Molecular determinants of the mechanism underlying acceleration of the interaction between antithrombin and factor Xa by heparin pentasaccharide. J Biol Chem 277(18):15971–15978. 104. Manithody C, Yang L and Rezaie AR (2002) Role of basic residues of the autolysis loop in the catalytic function of factor Xa. Biochemistry 41(21):6780–6788. 105. Izaguirre G, Zhang W, Swanson R, Bedsted T and Olson ST (2003) Localization of an antithrombin exosite that promotes rapid inhibition of factors Xa and IXa dependent on heparin activation of the serpin. J Biol Chem 278(51):51433–51440. 106. Langdown J, Johnson DJ, Baglin TP and Huntington JA (2004) Allosteric activation of antithrombin critically depends upon hinge region extension. J Biol Chem 279(45):47288–47297. 107. Whisstock JC, Skinner R, Carrell RW and Lesk AM (2000) Conformational changes in serpins: the native and cleaved conformations of alpha-1 antitrypsin. J Molec Biol 296:685–699. 108. Johnson DJ and Huntington JA (2004) The influence of hinge region residue Glu-381 on antithrombin allostery and metastability. J Biol Chem 279(6):4913–4921. 109. Belzar KJ, Zhou A, Carrell RW, Gettins PG and Huntington JA (2002) Helix D elongation and allosteric activation of antithrombin. J Biol Chem 277(10):8551–8558. 110. Meagher JL, Olson ST and Gettins PGW (2000) Critical role of the linker region between helix D and strand 2A in heparin activation of antithrombin. J Biol Chem 275(4):2698–2704. 111. Johnson DJ and Huntington JA (2003) Crystal structure of antithrombin in a heparin-bound intermediate state. Biochemistry 42(29): 8712–8719. 112. Olson ST, Stephens AW, Hirs CH, Bock PE and Bjork I (1995) Kinetic characterization of the proteinase binding defect in a reactive site variant of the serpin, antithrombin. Role of the P1 residue in transition-state stabilization of antithrombin-proteinase complex formation. J Biol Chem 270(17): 9717–9724.

ch22

Jan. 10, 2007

14:27 SPI-B429



547

113. Rezaie AR and Olson ST (1997) Contribution of lysine 60f to S1 specificity of thrombin. Biochemistry 36(5):1026–1033. 114. Hopkins PCR, Crowther DC, Carrell RW and Stone SR (1995) Development of a novel recombinant serpin with potential antithrombotic properties. J Biol Chem 270(20):11866–11871. 115. Hopkins PC, Pike RN and Stone SR (2000) Evolution of serpin specificity: Cooperative interactions in the reactive-site loop sequence of antithrombin specifically restrict the inhibition of activated protein C. J Mol Evol 51(5):507–515. 116. O’Reilly MS, Pirie-Shepherd S, Lane WS and Folkman J (1999) Antiangiogenic activity of the cleaved conformation of the serpin antithrombin. Science 285:1926–1928. 117. Mourey L, Samama JP, Delarue M, Petitou M, Choay J and Moras D (1993) Crystal structure of cleaved bovine antithrombin III at 3.2 A resolution. J Mol Biol 232:223–241. 118. Larsson H, Sjoblom T, Dixelius J, et al. (2000) Antiangiogenic effects of latent antithrombin through perturbed cell-matrix interactions and apoptosis of endothelial cells. Cancer Res 60(23):6723–6729. 119. Bouck N (2002) PEDF: Anti-angiogenic guardian of ocular function. Trends Mol Med 8(7):330–334. 120. Miao RQ, Agata J, Chao L and Chao J (2002) Kallistatin is a new inhibitor of angiogenesis and tumor growth. Blood 100(9):3245–3252. 121. Marciniak E (1981) Thrombin-induced proteolysis of human antithrombin III: An outstanding contribution of heparin. Br J Haematol 48(2):325–336. 122. Zhou A, Huntington JA and Carrell RW (1999) Formation of the antithrombin heterodimer in vivo and the onset of thrombosis. Blood 94(10): 3388–3396. 123. Larsson H, Akerud P, Nordling K, Raub-Segall E, Claesson-Welsh L and Bjork I (2001) A novel anti-angiogenic form of antithrombin with retained proteinase binding ability and heparin affinity. J Biol Chem 276(15): 11996–12002. 124. Zhang W, Chuang YJ, Swanson R, et al. (2004) Antiangiogenic antithrombin down-regulates the expression of the proangiogenic heparan sulfate proteoglycan, perlecan, in endothelial cells. Blood 103(4):1185–1191. 125. Aviezer D, Hecht D, Safran M, Eisinger M, David G and Yayon A (1994) Perlecan, basal lamina proteoglycan, promotes basic fibroblast growth factorreceptor binding, mitogenesis, and angiogenesis. Cell 79(6):1005–1013. 126. Aviezer D, Iozzo RV, Noonan DM and Yayon A (1997) Suppression of autocrine and paracrine functions of basic fibroblast growth factor

ch22

Jan. 10, 2007

14:27 SPI-B429

548

127.

128. 129.

130.

131.

132.

133.

134.

135. 136.

137.

138.


F. C. Church et al.

by stable expression of perlecan antisense cDNA. Mol Cell Biol 17(4): 1938–1946. Sharma B, Handler M, Eichstetter I, Whitelock JM, Nugent MA and Iozzo RV (1998) Antisense targeting of perlecan blocks tumor growth and angiogenesis in vivo. J Clin Invest 102(8):1599–1608. Abdollahi A, Hahnfeldt P, Maercker C, et al. (2004) Endostatin’s antiangiogenic signaling network. Mol Cell 13(5):649–663. Zhang W, Swanson R, Izaguirre G, Xiong Y, Lau LF and Olson ST (2005) The heparin-binding site of antithrombin is crucial for antiangiogenic activity. Blood 106(5):1621–1628. Herbert JM, Herault JP, Bernat A, et al. (1998) Biochemical and pharmacological properties of SANORG 34006, a potent and long-acting synthetic pentasaccharide. Blood 91(11):4197–4205. Tollefsen DM and Pestka CA (1985) Heparin cofactor II activity in patients with disseminated intravascular coagulation and hepatic failure. Blood 66(4):769–774. Tollefsen DM, Majerus DW and Blank MK (1982) Heparin cofactor II: purification and properties of a heparin-dependent inhibitor of thrombin in human plasma. J Biol Chem 257(5):2162–2169. Church FC, Meade JB and Pratt CW (1987) Stucture-function relationships in heparin cofactor II: Spectral analysis of aromatic residues and absence of a role for sulfhydryl groups in thrombin inhibition. Arch Biochem Biophys 259(2):331–340. Herzog R, Lutz S, Blin N, Marasa JC, Blinder MA and Tollefsen DM (1991) Complete nucleotide sequence of the gene for human heparin cofactor II and mapping to chromosomal band 22q11. Biochemistry 30(5):1350–1357. Pratt CW, Whinna HC and Church FC (1992) A comparison of three heparinbinding serine proteinase inhibitors. J Biol Chem 267:8795–8801. Pratt CW, Whinna HC, Meade JM, Treanor RE and Church FC (1989) Physicochemical aspects of heparin cofactor II. Ann NY Acad Sci 556: 104–114. Blinder MA, Marasa JC, Reynolds CH, Deaven LL and Tollefsen DM (1988) Heparin cofactor II: cDNA sequence, chromosome localization, restriction fragment length polymorphism, and expression in Escherichia coli. Biochem 27:752–759. Griffith MJ, Noyes CM, Tyndall JA and Church FC (1985) Structural evidence for leucine at the reactive site of heparin cofactor II. Biochemistry 24: 6777–6782.

ch22

Jan. 10, 2007

14:27 SPI-B429



549

139. Church FC, Noyes CM and Griffith MJ (1985) Inhibition of chymotrypsin by heparin cofactor II. Proc Natl Acad Sci USA 82:6431–6434. 140. Derechin VM, Blinder MA and Tollefsen DM (1990) Substitution of arginine for Leu444 in the reactive site of heparin cofactor II enhances the rate of thrombin inhibition. J Biol Chem 265(10):5623–5628. 141. Han J-H, VanDeerlin VMD and Tollefsen DM (1997) Heparin facilitates dissociation of complexes between thrombin and a reactive site mutant (L444R) of heparin cofactor II. J Biol Chem 272(13):8243–8249. 142. CiacciaAV, WillemzeAJ and Church FC (1997) Heparin promotes proteolytic inactivation by thrombin of a reactive site mutant (L444R) of recombinant heparin cofactor II. J Biol Chem 272(2):888–893. 143. Inhorn RC and Tollefsen DM (1986) Isolation and characterization of a partial cDNA clone for heparin cofactor II. Biochem Biophys Res Commun 137(1):431–436. 144. Ragg H (1986) A new member of the plasma protease inhibitor gene family. Nuc Acids Res 14(2):1073–1087. 145. VanDeerlin VMD and Tollefsen DM (1991) The N-terminal acidic domain of heparin cofactor II mediates the inhibition of α-thrombin in the presence of glycosaminoglycans. J Biol Chem 266(30):20223–20231. 146. Blinder MA, Andersson TR, Abildgaard U and Tollefsen DM (1989) Heparin cofactor II Oslo. Mutation of Arg-189 to His decreases the affinity for dermatan sulfate. J Biol Chem 264(9):5128–5133. 147. Blinder MA and Tollefsen DM (1990) Site-directed mutagenesis of arginine 103 and lysine 185 in the proposed glycosaminoglycan-binding site of heparin cofactor II. J Biol Chem 265(1):286–291. 148. Fortenberry YM, Whinna HC, Gentry HR, Myles T, Leung LL and Church FC (2004) Molecular mapping of the thrombin-heparin cofactor II complex. J Biol Chem 279(41):43237–43244. 149. Jeter ML, Ly LV, Fortenberry YM, et al. (2004) RNA aptamer to thrombin binds anion-binding exosite-2 and alters protease inhibition by heparinbinding serpins. FEBS Lett 568(1–3):10–14. 150. Liaw PCY, Austin RC, Fredenburgh JC, Stafford AR and Weitz JI (1999) Comparison of heparin- and dermatan sulfate-mediated catalysis of thrombin inactivation by heparin cofactor II. J Biol Chem 274(39): 27597–27604. 151. Myles T, Church FC, Whinna HC, Monard D and Stone SR (1998) Role of thrombin anion-binding exosite-I in the formation of thrombin-serpin complexes. J Biol Chem 273(47):31203–31208.

ch22

Jan. 10, 2007

14:27 SPI-B429

550


F. C. Church et al.

152. Ragg H, Ulshofer T and Gerewitz J (1990) On the activation of human leuserpin-2, a thrombin inhibitor, by glycosaminoglycans. J Biol Chem 265(9):5211–5218. 153. Ragg H, Ulshofer T and Gerewitz J (1990) Glycosaminoglycan-mediated leuserpin-2/thrombin interaction: Structure-function relationships. J Biol Chem 265(36):22386–22391. 154. Rogers SJ, Pratt CW, Whinna HC and Church FC (1992) Role of thrombin exosites in inhibition by heparin cofactor II. J Biol Chem 267(6): 3613–3617. 155. Sheehan JP, Tollefsen DM and Sadler JE (1994) Heparin cofactor II is regulated allosterically and not primarily by template effects. Studies with mutant thrombins and glycosaminoglycans. J Biol Chem 269:32747–32751. 156. Sheehan JP, Wu Q, Tollefsen DM and Sadler JE (1993) Mutagenesis of thrombin selectively modulates inhibition by serpins heparin cofactor II and antithrombin III. Interaction with the anion-binding exosite determines heparin cofactor II specificity. J Biol Chem 268(5):3639–3645. 157. Whinna HC, Blinder MA, Szewczyk M, Tollefsen DM and Church FC (1991) Role of lysine 173 in heparin binding to heparin cofactor II. J Biol Chem 266:8129–8135. 158. Tollefsen DM, Pestka CA and Monafo WJ (1983) Activation of heparin cofactor II by dermatan sulfate. J Biol Chem 258(11):6713–6716. 159. Tollefsen DM, Peacock ME and Monafo WJ (1986) Molecular size of dermatan sulfate oligosaccharides required to bind and activate heparin cofactor II. J Biol Chem 261:8854–8858. 160. McGuire EA and Tollefsen DM (1987) Activation of heparin cofactor II by fibroblasts and vascular smooth muscle cells. J Biol Chem 262(1):169–175. 161. Tollefsen DM, Maimone MM, McGuire EA and Peacock ME (1989) Heparin cofactor II activation by dermatan sulfate. Ann NY Acad Sci 556:116–122. 162. Maimone MM and Tollefsen DM (1990) Structure of a dermatan sulfate hexasaccharide that binds to heparin cofactor II with high affinity. J Biol Chem 265(30):18263–18271. 163. Whinna HC, Choi HU, Rosenberg LC and Church FC (1993) Interaction of heparin cofactor II with biglycan and decorin. J Biol Chem 268:3920–3924. 164. Shirk RA, Church FC and Wagner WD (1996) Arterial smooth muscle cell heparan sulfate proteoglycans accelerate thrombin inhibition by heparin cofactor II. Arterioscler Thromb Vasc Biol 16(9):1138–1146. 165. Suzuki K, Deyashiki Y, Nishioka J and Toma K (1989) Protein C inhibitor: Structure and function. Thromb Haemostas 61(3):337–342.

ch22

Jan. 10, 2007

14:27 SPI-B429



551

166. Espana F, Gruber A, Heeb M, Hanson S, Harker L and Griffin J (1991) In vivo and in vitro complexes of activated protein C with two inhibitors in baboons. Blood 77:1754–1760. 167. Geiger M, Priglinger U, Griffin JH and Binder BR (1991) Urinary protein C inhibitor. Glycosaminoclycans synthesized by the epithelial kidney cell line TCL-598 enhance its interaction with urokinase. J Biol Chem 266(18): 11851–11857. 168. Ecke S, Geiger M, Resch I, et al. (1992) Inhibition of tissue kallikrein by protein C inhibitor. Evidence for identity of protein C inhibitor with the kallikrein binding protein. J Biol Chem 267:7048–7052. 169. Laurell M, Christensson A, Abrahamsson PA, Stenflo J and Lilja H (1992) Protein C inhibitor in human body fluids. Seminal plasma is rich in inhibitor antigen deriving from cells throughout the male reproductive system. J Clin Invest 89(4):1094–1101. 170. Radtke KP, Fernandez JA, Greengard JS, et al. (1994) Protein C inhibitor is expressed in tubular cells of human kidney. J Clin Invest 94(5):2117–2124. 171. Hermans JM, Jones R and Stone SR (1994) Rapid inhibition of the sperm protease acrosin by protein C inhibitor. Biochem 33:5440–5444. 172. Espana F, Sanchez-Cuenca J, Fernandez PJ, et al. (1999) Inhibition of human sperm-zona-free hamster oocyte binding and penetration by protein C inhibitor. Andrologia 31(4):217–223. 173. Uhrin P, Dewerchin M, Hilpert M, et al. (2000) Disruption of the protein C inhibitor gene results in impaired spermatogensis and male infertility. J Clin Invest 106(12):1531–1539. 174. Meijers JCM and Chung DW (1991) Organization of the gene coding for human protein C inhibitor (plasminogen activator inhibitor-3). Assignment of the gene to chromosome 14. J Biol Chem 266(23):15028–15034. 175. Billingsley GD, Walter MA, Hammond GL and Cox DW (1993) Physical mapping of four serpin genes: a1-antitrypsin, a1-antichymotrypsin, corticosteroid-binding globulin, and protein C inhibitor, within a 280-kb region on chromosome 14q32.1. Am J Hum Genet 52:343–353. 176. Suzuki K, Deyashiki Y, Nishioka J, et al. (1987) Characterization of a cDNA for human protein C inhibitor. A new member of the plasma serine protease inhibitor superfamily. J Biol Chem 262:611–616. 177. Heeb MJ, Espana F, Geiger M, Collen D, Stump DC and Griffin JH (1987) Immunological identity of heparin-dependent plasma and urinary protein C inhibitor and plasminogen activator inhibitor-3. J Biol Chem 262: 15813–15816.

ch22

Jan. 10, 2007

14:27 SPI-B429

552


F. C. Church et al.

178. Geiger M, Heeb MJ, Binder BR and Griffin JH (1988) Competition of activated protein C and urokinase for a heparin-dependent inhibitor. 2: 2263–2267. 179. Meijers JCM, Kanters DHAJ, Vlooswijk RAA, van Erp HE, Hessing M and Bouma BN (1988) Inactivation of human plasma kallikrein and factor XIa by protein C inhibitor. Biochemistry 27:4231–4237. 180. Heeb MJ, Espana FHGJ (1989) Inhibition and complexation of activated protein C by two major inhibitors in plasma. Blood 73(2):446–454. 181. Elisen MG, van Kooij RJ, Nolte MA, et al. (1998) Protein C inhibitor may modulate human sperm-oocyte interactions. Biol Reprod 58(3): 670–677. 182. Elisen MG, von dem Borne PA, Bouma BN and Meijers JC (1998) Protein C inhibitor acts as a procoagulant by inhibiting the thrombomodulin-induced activation of protein C in human plasma. Blood 91(5):1542–1547. 183. Elisen MGLM, Bouma BN, Church FC and Meijers JCM (1998) Inhibition of serine proteases by reactive site mutants of protein C inhibitor (plasminogen activator inhibitor-3). Fib Prot 12(5):283–291. 184. Shen L, Villoutreix BO and Dahlback B (1999) Involvement of Lys 62(217) and Lys 63(218) of human anticoagulant protein C in heparin stimulation of inhibition by the protein C inhibitor. Thromb Haemost 82:72–79. 185. Friedrich U, Blom AM, Dahlbäck B and Villoutreix BO (2001) Structural and energetic characteristics of the heparin-binding site in antithrombotic protein C. J Biol Chem 276:24122–24128. 186. Rezaie AR (2003) Exosite-dependent regulation of the protein C anticoagulant pathway. Trends Cardiovasc Med 13(1):8–15. 187. Hobson JP, Netzel-Arnett S, Szabo R, et al. (2004) Mouse DESC1 is located within a cluster of seven DESC1-like genes and encodes a type II transmembrane serine protease that forms serpin inhibitory complexes. J Biol Chem 279(45):46981–46994. 188. Kuhn LA, Griffin JH, Fisher CL, et al. (1990) Elucidating the structural chemistry of glycosaminoglycan recognition by protein C inhibitor. Proc Natl Acad Sci USA 87:8506–8510. 189. Shirk RA, Elisen MGLM, Meijers JCM and Church FC (1994) Role of the H helix in heparin binding to protein C inhibitor. J Biol Chem 269: 28690–28695. 190. Elisen MGLM, Maseland MHH, Church FC, Bouma BN and Meijers JCM (1996) Role of the A+ helix in heparin binding to protein C inhibitor. Thromb Haemost 75(5):760–766.

ch22

Jan. 10, 2007

14:27 SPI-B429



553

191. Neese LL, Wolfe CA and Church FC (1998) Contribution of basic residues of the D and H helices in heparin binding to protein C inhibitor. Arch Biochem Biophys 355:101–108. 192. Priglinger U, Geiger M, Bielek E, Vanyek E and Binder BR (1994) Binding of urinary protein C inhibitor to cultured human epithelial kidney tumor cells (TCL-598). The role of glycosaminoglycans present on the luminal cell surface. J Biol Chem 269(20):14705–14710. 193. Ishiguro K, Kojima T, Kadomatsu K, et al. (2000) Complete antithrombin deficiency in mice results in embryonic lethality. J Clin Invest 106(7): 873–878. 194. Yanada M, Kojima T, Ishiguro K, et al. (2002) Impact of antithrombin deficiency in thrombogenesis: Lipopolysaccharide and stress-induced thrombus formation in heterozygous antithrombin-deficient mice. Blood 99(7):2455–2458. 195. HajMohammadi S, Enjyoji K, Princivalle M, et al. (2003) Normal levels of anticoagulant heparan sulfate are not essential for normal hemostasis. J Clin Invest 111(7):989–999. 196. Dewerchin M, Herault JP, Wallays G, et al. (2003) Life-threatening thrombosis in mice with targeted Arg48-to-Cys mutation of the heparin-binding domain of antithrombin. Circ Res 93(11):1120–1126. 197. He L,Vicente CP, Westrick RJ, Eitzman DT and Tollefsen DM (2002) Heparin cofactor II inhibits arterial thrombosis after endothelial injury. J Clin Invest 109(2):213–219. 198. Vicente CP, He L, Pavao MS and Tollefsen DM (2004) Antithrombotic activity of dermatan sulfate in heparin cofactor II-deficient mice. Blood 104(13):3965–3970. 199. Wagenaar GT, van Vuuren AJ, Girma M, et al. (2000) Characterization of transgenic mice that secrete functional human protein C inhibitor into the circulation. Thromb Haemost 83(1):93–101. 200. Hayashi T, Nishioka J, Kamada H, et al. (2004) Characterization of a novel human protein C inhibitor (PCI) gene transgenic mouse useful for studying the role of PCI in physiological and pathological conditions. J Thromb Haemost 2(6):949–961. 201. Johnson DJD, Langdown J, Li W, Luis SA, Baglin TP and Huntington JA (2006) Crystal structure of monomeric native antithrombin reveals a novel reactive center loop conformation. J Biol Chem, published online at http://www.jbc.org/cgi/doi/10.1074/jbc.M607204200 202. Izaguirre G and Olson ST (2006) Residues Tyr253 and Glu255 in strand 3 of b-sheet C of antithrombin are key determinants of an exosite made accessible

ch22

Jan. 10, 2007

14:27 SPI-B429

554

203.

204.

205.

206.


F. C. Church et al.

by heparin activation to promote rapid inhibition of factors Xa and IXa. J Biol Chem 281:13424–13432. Johnson DJD, Li W, Adams TE and Huntington JA (2006) AntithrombinS195A factor Xa-heparin structure reveals the allosteric mechanism of antithrombin activation. EMBO J 25:2029–2037. Zhang W, Chuang Y-J, Jin T, Swanson R, Xiong Y, Leung L and Olson ST (2006) Antiangiogenic antithrombin induces global changes in the gene expression profile of endothelial cells. Cancer Res 66:5047–5055. Sutherland JS, Bhakta V, Filion ML and Sheffield WP (2006) The transferable tail: Fusion of the N-terminal acidic extension of heparin cofactor II to alpha(1)-proteinase inhibitor M358R specifically increases the rate of thrombin inhibition. Biochemistry 45:11444–11452. Verhamme IM, Bock PE and Jackson CM (2004) The preferred pathway of glycosaminoglycan-accelerated inactivation of thrombin by heparin cofactor II. J Biol Chem 279:9785–9795.

ch22

Jan. 10, 2007

8:57

SPI-B429


23 C1 Inhibitor: Structure, Function, Biologic Activity and Angioedema Philip A. Patston and Alvin E. Davis III

1. Introduction C1-inhibitor is a serpin that is the only inhibitor of the C1 component of the classical pathway of complement and is an important regulator of plasma kallikrein. Deficiency of C1-inhibitor results in angioedema. In this review, we summarize information on the structure, protease inhibitory properties, biosynthesis, anti-inflammatory properties and clinical significance of C1-inhibitor. A comprehensive overview of C1-inhibitor, particularly regarding angioedema and clinical aspects, can also be obtained by reference to earlier extensive articles.1–6

2. Structure of C1-inhibitor C1-inhibitor was cloned, sequenced, and its identity as a serpin confirmed, independently by a number of groups.7–12 Alignment with other serpin sequences revealed that it was likely to have the same structure of three β-sheets and nine α-helices.13 The alignments confirmed the identity of an arginine at the P1 position.14 The gene for C1-inhibitor has been localized to Chromosome 11.8,10,15 The gene has 17 Alu repeat sequences which are responsible for some of the defects seen in C1-inhibitor associated with type I hereditary angioedema.16–19 C1-inhibitor has been assigned to Clade G and is designated as SERPING1.20 555

ch23

Jan. 10, 2007

8:57

SPI-B429

556


Philip A. Patston and Alvin E. Davis III

The two most intriguing aspects of the structure of C1-inhibitor are that it is highly glycosylated and that it has an amino-terminal domain prior to the start of the serpin domain.8 This two-domain structure is evident from a variety of biophysical measurements.21–23 Homology with α1 -antitrypsin begins at approximately residue 120. There are six known N-linked carbohydrate attachment sites (Asn-3, -47, -59, -216, -231, -330) and seven O-linked carbohydrate attachment sites (Ser-42, Thr-26, -49, -61, -66, -70, -74). There are another potential seven O-linked sites (Thr77, -84, -85, -89, -93, -96, -97), but it is not clear if they are used. Thus, the N-linked sites are in the first half of the N-terminal region and in the serpin domain, whereas the majority of the O-linked sites are in the second half of the N-terminal domain where the O-linked consensus sequence, Glx-Pro-Thr-Thr, or minor variations, occurs seven times between residues 63 and 97.8 Perkins et al.23 confirmed that there are seven O-linked sites, consistent with the results of Bock et al.8 By the use of digoxigenin-labeled lectins, Schoenberger24 showed that sialic acids and complex and hybrid type N-linked sugars were present, but high mannose type carbohydrates were absent. Cai and Davis25 showed that the sialyl-Lewisx was present on the N-linked sugars, the implications of which are discussed in more details below. Analysis of the carbohydrate chains of C1-inhibitor isolated from individuals with hereditary angioedema indicated that although the O-glycan chains were unaltered, there was very little sialic acid present except for mannose-6-phosphate, perhaps to allow secretion via an alternate pathway.26 The amino-terminal domain is attached to the serpin domain via two disulfides (Cys101-Cys406 and Cys108-Cys183). Reduction of the cysteines results in the loss of inhibitory activity as a result of polymer formation.27 A C1-inhibitor mutant with a deletion of residues Asp62 to Thr116 formed inactive polymers, consistent with the importance of the disulfides in maintaining the metastable structure of the native serpin.28 Polymerization of normal C1-inhibitor also readily occurs after denaturation and on prolonged storage,29 and spontaneously with a hinge region mutant,30 carboxy-terminal mutants,31 and a deletion mutant.32 For many years, the function of the N-terminal domain remained a mystery with no obvious role being assigned to it, or to the carbohydrates,33,34 with the exception of a potential role of the sialic acid in mediating binding

ch23

Jan. 10, 2007

8:57

SPI-B429


C1 Inhibitor: Structure, Function, Biologic Activity and Angioedema

557

to neutrophils35 and regulating clearance from the circulation.36 However, recent new data suggest that it might have very important anti-inflammatory and immunity roles unrelated to the protease inhibitory activity of the serpin domain. These data are discussed in details below. Finally, it has been reported that a novel metalloproteinase from E. coli O157:H7 called StcE can enhance the activity of C1-inhibitor in complement erythrocyte lysis assays. StcE binds to the erythrocytes and to C1-inhibitor via the amino-terminal domain and thereby increases the local concentration of C1-inhibitor associated with the cell. In addition, StcE can cleave C1inhibitor in the amino-terminal domain. As expected, the StcE cleaved C1-inhibitor remains active as a kallikrein inhibitor.37–39 Recently, it has been shown that plasma kallikrein can bind to endothelial cells by an as yet uncharacterized mechanism, and that cell-bound kallikrein is resistant to inhibition by C1-inhibitor. However, StcE-cleaved C1-inhibitor can inhibit the cell-bound kallikrein, suggesting that a novel function of this domain is to prevent inhibition of cell-bound kallikrein.39 The physiological significance of these findings is remains unknown.

3. Inhibitory Specificity of C1-inhibitor C1-inhibitor was first identified in the 1960s as an inhibitor of the esterase activity of activated C1.40 Since then, it has been shown that C1-inhibitor is the only inhibitor of the activated C1s and C1r subunits of C1.41–45 Inhibition of C1r and C1s within C1 results in dissociation of C1q from the C1r and C1s complexes with C1-inhibitor.43,46,47 Although inhibition of C1 is a very important function of C1-inhibitor, the fact that the deficiency of C1inhibitor results in angioedema mediated by bradykinin shows that the inhibition of plasma kallikrein is a critical process.48 The inhibition of plasma kallikrein by C1-inhibitor has been studied in detail in plasma and with purified proteins.39,49–54 In terms of the inhibitory mechanism, C1-inhibitor inhibits its serine protease targets by the typical serpin suicide substrate mechanism with the formation of stable covalent C1-inhibitor-protease complexes and some reactive center loop cleaved C1-inhibitor.41,54 Other studies are consistent with the importance of the typical serpin conformational change, reactive center loop insertion into β-sheet A, and covalent

ch23

Jan. 10, 2007

8:57

SPI-B429

558



complex formation in the inhibition process.53,55–62 Confirmation of the importance of the reactions with C1s and kallikrein is given by studies that have shown that complexes of C1-inhibitor with these proteases formed in vivo.63–65 In studies with plasma and with purified proteins, inhibition of factor XIIa, β-factor XIIa, factor XIa, thrombin, plasmin, t-PA, glandular kallikrein, the protease subunit of NGF, plasma hyaluronan binding protein, hepatitis C virus NSC protease, chymotrypsin, C1r-like protease, MASP1, and MASP2 by C1-inhibitor have all been reported.66–90 Of these reactions, inhibition of factor XIIa and β-factor XIIa are most likely to be physiologically relevant.66–68 Inhibition of factor XIa by C1-inhibitor might be physiological, although α1 -antitrypsin and protease nexin I probably also play a role.65,70–72 The majority of plasmin inhibition in vivo is mediated by α2 -antiplasmin, but in DIC, complexes of plasmin and C1-inhibitor have been detected.74 Similarly, even though C1-inhibitor is a poor inhibitor of t-PA in vitro, complexes have been detected after exercise and electric shock.76–78,91 The remaining reactions that are likely to be important involve the inhibition of MASP-1 and MASP-2. Although the kinetics of these reactions have not been reported and detection of complexes from in vivo samples has not yet been shown, it is likely that C1-inhibitor is the main or the only inhibitor of these proteases which share considerable homology with C1r and C1s.86–90 Tables 1 and 2 give the second-order rate constants for some of these inhibition reactions. A number of serpins are activated by ligand binding,92 most notably antithrombin activation by heparin and heparan sulfate.93 Although heparin has no effect on the inhibition of plasma kallikrein or factor XIIa by C1inhibitor,51,94,95 glycosaminoglycans have been reported to increase the rate of inhibition of C1s by C1-inhibitor,45,55,96–101 along with a concomitant increase in reactive center loop cleaved C1-inhibitor.102,103 As there is no clearly defined heparin binding domain on C1-inhibitor, and as binding to endogenous glycosaminoglycans has not been reported, the significance of these results remains unclear. In addition, although in vitro, C1-inhibitor can bind to collagen and laminin103 and is a substrate for tissue transglutaminase,104 it is not known if these interactions have any regulatory or physiological role. Further information on the binding of C1-inhibitor to extracellular matrix components is given below. As with other

ch23

Jan. 10, 2007

8:57

SPI-B429



559

Table 1 The second order rate constants of inhibition of activated C1s and C1r by C1-inhibitor. Protease

Second order rate constant (M−1 s−1 )

Temperature (◦ C)

Reference

C1s C1s C1s C1s C1s C1s C1s C1r C1r C1r

1.2 × 104 1.4 × 105 4.7 × 105 6 × 104 11 × 104 4.8 × 104 6.2 × 104 2.8 × 103 2.5 × 103 4.4 × 104

37 25 37 30 32 37 37 37 25 37

45 55 55 97 97 130 243 45 55 55

The table shows values for the second order rate constants reported in the literature. The differences in the values for the second order rate constants can arise from the temperature at which the inhibition reaction was performed, and due to the different methods used, as discussed by Lennick et al.97 In addition, the most likely other source of variation is due to variable amounts of inactive polymers in the C1-inhibitor preparations.

serpin-protease complexes, the complexes of C1-inhibitor with C1s, plasma kallikrein, and β-factor XIIa are cleared from the circulation by LRP.105–108

4. Synthesis of C1-inhibitor The hepatocyte is the main site of C1-inhibitor biosynthesis,109 although expression has been shown in monocytes,110,111 endothelial cells,112 platelets,113 fibroblasts,114 and the brain.115,116 The gene for C1-inhibitor has an interferon-γ response element,117 and in cell culture studies, it is the major stimulator of synthesis in a variety of cell types.109,114,118–123 Phosphatase 2A and STAT1 are implicated in the signal transduction mechanisms leading to an increase in C1-inhibitor expression.117,124,125 It is often assumed that danazol and other attenuated androgens, which are used in the treatment of C1-inhibitor deficiency work by increasing C1-inhibitor synthesis, although there is no evidence to support this idea. In fact, danazol has no effect on C1-inhibitor synthesis in mice.126 Recombinant C1-inhibitor also has been produced by a variety of expression systems. Eldering et al.127 first expressed C1-inhibitor in

ch23

Jan. 10, 2007

8:57

SPI-B429


560

Philip A. Patston and Alvin E. Davis III Table 2 The second order rate constants of inhibition of plasma kallikrein and other proteases by C1-inhibitor. Protease

Plasma kallikrein Plasma kallikrein Plasma kallikrein Plasma kallikrein Factor XIIa Factor XIIa Factor XIIa β-Factor XIIa β-Factor XIIa Factor XIa Factor XIa Thrombin Plasmin Plasmin Plasmin t-PA

Second order rate constant (M−1 s−1 )

Temperature (◦ C)

Reference

1.7 × 104 5 × 104 4.5 × 104 2.5 × 104 3.7 × 103 8 × 103 4.5 × 103 3 × 103 9.8 × 103 2.3 × 103 1.8 × 103 19 3.6 × 104 1 × 104 5.5 × 103 5

23 37 37 37 37 37 37 23 37 37 37 37 23 23 37 37

49 51 50 95 68 95 243 66 95 69 95 73 49 245 130 91

Values are the second order rate constants reported in the literature. As indicated in Table 1, differences in values for given protease can arise due to the methods used, as well as the amount of inactive C1-inhibitor present. In addition, in the case of kallikrein, in particular, the inhibition reaction is very sensitive to temperature. At 37◦ C the stoichiometry of inhibition (SI) is about 2,54,243 but rapidly increases as the temperature of the inhibition reaction is decreased to an SI of over 6 at 4◦ C. 54 Thus the temperature of the reaction will affect the SI and the measured rate. To compensate for this it is necessary to multiply the measure second rate constant by the SI to give the true rate of inhibition.244

COS cells. Expression has also been reported using E. coli,128 Pichia pastoris,28 and the baculovirus system.129 Although studies using recombinant C1-inhibitor have produced useful structure-function data,34,61,130–133 the expression levels of active protein are low in all cases. These problems stem from the inherent difficulty in expressing serpins in the native metastable state, the requirement for two disulfide bonds to form correctly, the unusual amino-terminal domain, and the extensive glycosylation. To avoid the problems with C1-inhibitor produced in cells in vitro, the protein

ch23

Jan. 10, 2007

8:57

SPI-B429



561

has been expressed in the milk of trangsenic rabbits,134,135 with the goal of producing protein for clinical use. Transgenic mice expressing human C1-inhibitor have also been produced.136−138

5. The Biologic Importance of C1-inhibitor By virtue of its protease inhibitory activities, C1-inhibitor regulates activation of both the complement and contact systems. Therefore, these functions play a role in the modulation of inflammatory responses. The biologic importance of regulation of contact system activation is best demonstrated by the relatively dramatic symptoms of hereditary angioedema, which develops in people who are heterozygous for deficiency of C1-inhibitor. In the past several years, a number of other C1-inhibitor-mediated activities, which include interactions with extracellular matrix components, gramnegative endotoxin, and bacteria, and selectin adhesion molecules, have been described. In addition, an interaction with C3b that results in the inhibition of alternative complement pathway activation and does not depend on protease inhibition has been partially characterized. Although the biologic importance of these activities that are not the result of protease inhibition remains incompletely defined, these also may play important roles in the regulation of inflammation. Therapy with C1-inhibitor clearly is useful in reversing the symptoms of hereditary angioedema. In addition, a number of studies, including both animal models and preliminary studies in human disease, suggest that therapy with C1-inhibitor may be useful in a variety of inflammatory diseases.

6. Hereditary Angioedema C1-inhibitor is required for the maintenance of normal vascular permeability. Two phenotypic variants of hereditary angioedema are generally recognized. Type I is characterized by decreased antigenic and functional levels of a normal C1-inhibitor protein. Type II is characterized by the presence in plasma of normal or elevated antigenic levels of a dysfunctional mutant protein in addition to reduced levels of the normal protein. There is no difference in clinical presentation between the two types. It

ch23

Jan. 10, 2007

8:57

SPI-B429

562



also should be borne in mind that there is overlap in these categories. In particular, a number of individuals classified as type I by these criteria have mutations that result in a dysfunctional protein that is either catabolized or retained within the cell, or is secreted in very small amounts. Angioedema results from the dysfunction of the endothelial barrier at the level of the post-capillary venule with movement of fluid and protein from the intravascular to the extravascular space. During asymptomatic periods, C1-inhibitor levels in patients with HAE are approximately 30% of the normal level. The explanation for the decrease in C1-inhibitor levels below the expected 50% of the normal level is very likely multifactorial. First, metabolic turnover studies suggest that a C1-inhibitor concentration that is 50% of the normal level is insufficient to regulate activation of target proteases. As C1-inhibitor is a suicide inhibitor, this dysregulated activation results in the depression of its concentration.139,140 In addition, there is also evidence that at least in some patients, there may be supression of protein translation or secretion, or enhancement of catabolism of the normal C1-inhibitor protein by the mutant mRNA or protein.141,142 Development of an angioedema attack is accompanied by a further decrease in the C1inhibitor level together with evidence for both complement and contact system activation. This evidence includes decreased levels of C2 and C4, the presence of circulating complexes of C1-inhibitor with C1r and C1s, cleavage of high-molecular weight kininogen, and the presence of increased levels of bradykinin.143–148 As both the contact and the complement systems clearly are activated, it was not immediately clear which pathway, or whether both pathways, were responsible for the mediation of symptoms. Several early studies suggested that bradykinin was the mediator,149–151 while others suggested mediation via the complement system.152–156 More recent studies, one of which directly analyzed the production of mediators in C1-inhibitor deficient plasma, and another that detected bradykinin in patients’ plasma during attacks of angioedema, indicated mediation via contact system activation.143,157 The fortuitous discovery by Wisnieski et al.158 of a family among whom several members were heterozygous for a dysfunctional C1-inhibitor molecule was subsequently shown to contain a substitution of the P2 Ala-443 with a Val.132 This variant was a normal inhibitor of kallikrein, but deficient in C1s inhibition.132,159 The propositus presented with systemic lupus erythematosus, but no family member had

ch23

Jan. 10, 2007

8:57

SPI-B429



563

ever developed angioedema. These observations further strengthened the conclusion that bradykinin was the most likely mediator. Detailed descriptions of diagnosis and the treatment of C1-inhibitor deficiency, which represent the current standards of care, have been published recently.160,161

7. C1-Inhibitor Deficient Mice C1-inhibitor deficient mice were developed primarily to test the hypothesis that angioedema is mediated primarily via contact system activation with bradykinin generation. Although these mice, with rare exceptions, do not develop symptomatic angioedema, and are thus not a perfect model for angioedema, the analysis has provided a great deal of valuable information that is relevant to the pathogenesis of HAE and toward understanding the multiple biologic roles of C1-inhibitor.48 Both heterozygous (C1INH+/− ) and homozygous (C1INH−/− ) C1-inhibitor deficient mice have no increased fetal loss, are normal at birth, and subsequently grow and develop normally. As expected, C1INH−/− mice have no detectable C1-inhibitor mRNA or protein, while C1INH+/− mice have somewhat less than half normal levels, similar to humans with HAE. Both have evidence of complement activation with decreased C4 levels. Neither have episodes of detectable subcutaneous angioedema, although a very small number of deficient mice have developed gastrointestinal angioedema. Those episodes were not reproducible and factors that induce such episodes have not yet been found. Therefore, it cannot be unequivocally stated that these were a result of C1-inhibitor deficiency. However, intravenous injection of Evans blue dye, which complexes with albumin and is therefore a marker for the movement of fluid and protein across the endothelial barrier, revealed a rather striking increase in the vascular permeability in both C1INH+/− and C1INH−/− mice. Subcutaneous dye extravasation was exacerbated to a much greater extent than in wild-type mice at the site of application of mustard oil, an agent that induces a local inflammatory response. As expected, the increased vascular permeability was reversed following intravenous infusion of C1-inhibitor. To test the hypothesis that the increased vascular leak in these mice was caused by bradykinin release, C1-inhibitor deficient mice were treated with a variety of agents that either inhibit bradykinin release, its metabolism, or

ch23

Jan. 10, 2007

8:57

SPI-B429

564



its interaction with its receptor. Both DX88 (Dyax Corporation, Cambridge, MA, USA), a Kunitz domain variant that is a highly specific kallikrein inhibitor, and C1-inhibitor Ala-443 → Val, described above, completely reversed the increased vascular permeability. Furthermore, treatment with Icatibant (Hoe140) (Sigma Chemical Company, St. Louis, MO, USA), an effective bradykinin type 2 receptor (Bk2R) antagonist, also reversed the dye leakage from the vasculature. Bradykinin is the only known ligand for this receptor. On the other hand, treatment with the angiotensin converting enzyme inhibitor, captopril, dramatically enhanced vascular permeability in the deficient mice. Angiotensin converting enzyme, in addition to the function for which it is named, also provides a primary route for the degradation of bradykinin. Lastly, C1INH−/− mice were mated with Bk2R−/− mice. The resulting heterozygous double deficient mice were mated to obtain the C1INH−/− , Bk2R−/− mice. The vascular leakage in these mice was identical to that of wild-type mice. The increased vascular permeability in the C1-inhibitor deficient mice is therefore clearly mediated by bradykinin. These observations, combined with the data described in the preceding sections, provide additional support for the hypothesis that symptoms in hereditary angioedema are also mediated by bradykinin. These studies also support the concept that intravascular contact system activation takes place constantly, but that it is tightly regulated by C1-inhibitor.162 Regulation of the contact system by C1-inhibitor, in addition to its effects on endothelial barrier function, also regulates the other activites mediated by this system. These include effects on blood pressure regulation and both pro- and anti-thrombotic effects. Although attacks appear to be mediated by bradykinin, the pathogenesis of attack initiation remains unclear. The most common inducers of attacks include physical trauma, such as dental manipulations, and intercurrent febrile illness, although attacks also develop with no apparent predisposing factors.163,164 It seems quite likely that in an individual with HAE, with a pre-existing constantly low-plasma C1-inhibitor level, activation of any potential C1-inhibitor target protease (C1r, C1s, the mannan binding lectin associated proteases, factors XIa or XIIa, plasma kallikrein, tPA, plasmin) may consume the inhibitor to such an extent that unregulated activation of the complement and contact system is initiated. The decreased C1-inhibitor synthesis in a patient with HAE must be insufficient to maintain homeostasis

ch23

Jan. 10, 2007

8:57

SPI-B429



565

in the presence of such increased consumption. Based on current data, the role of complement system activation is therefore to initiate and/or perpetuate attacks rather than to generate a mediator of angioedema.

8. C1-inhibitor Functions that do not Depend on Protease Inhibition It is probably correct that the most crucial functions of C1-inhibitor are those that depend on its protease inhibitory activities that result in the regulation of complement and contact system activation, as described in the preceding sections. However, it is very likely that some or all of the activities that are not dependent on protease inhibition also play biologically relevant roles. The following sections will briefly review the interaction of C1-inhibitor with gram-negative endotoxin, selectin adhesion molecules, extracellular matrix components, and with C3b, none of which depends on protease inhibition. The interaction with gram-negative endotoxin and with selectins result in significant anti-inflammatory effects in animal models. Other effects of C1-inhibitor have also been described, the biologic relevance of which remains unclear. These include the inhibition of T-lymphocyte proliferation, apparently via C1s inactivation,165,166 the enhancement of neutrophil chemotaxis,167 and the suppression of collagen and ADP-induced platelet aggregation.168 The final section will discuss a number of disease models in which C1-inhibitor therapy has proven to be effective. It is possible that its effectiveness in some of these models may result from protease inhibition combined with other effects.

9. The Interaction of C1-Inhibitor with Gram-Negative Endotoxin Treatment with C1-inhibitor improves a number of different physiologic parameters and survival in several different animal models of endotoxin shock and sepsis.169–174 C3 and C4 deficient mice are very susceptible to endotoxin shock, which suggests that C3 may be required for endotoxin clearance.175 During endotoxin shock in these deficient mice, C1-inhibitor levels became depleted. Treatment with intravenous C1-inhibitor improved survival suggested that the contact system might play an important role in

ch23

Jan. 10, 2007

8:57

SPI-B429

566



pathogenesis. However, other data suggest that complement activation with C5a generation may be involved in the pathogenesis of septic shock.176–178 Several other studies suggest that the contact system may mediate some aspects of endotoxin shock,179–181 while some suggest that the contact system is of little importance.182 We hypothesized that the effectiveness of C1-inhibitor in endotoxin shock might result from some activity other than protease inhibition. In fact, both intact active C1-inhibitor and reactive center cleaved, inactive C1-inhibitor protected mice from endotoxin shock.183 Furthermore, C1-inhibitor prevented the binding of Salmonella typhimurium LPS to RAW 264.7 cells (a murine macrophage cell line) and to human blood cells. It directly interacted with LPS and blocked LPS-induced TNF-α mRNA expression. LPS is a component of the outer membrane of gram-negative bacteria. It activates monocytes to produce inflammatory mediators, particularly TNF-α.184 LPS binds to LPS-binding protein (LBP) and transfers LPS to CD14.185–188 The formation of LPSCD14 complexes initiates intracellular signaling by binding to Toll-like receptors (TLRs) expressed on mononuclear phagocytes and other cells.189 LPS interacts with a number of glycoproteins, including the three components of the cell membrane LPS receptor, CD14, MD-2 and TLR4.190–195 C1-inhibitor in which the amino terminal non-serpin domain was deleted did not bind to LPS.183 N-glycosidase treated C1-inhibitor also did not protect mice from lethal LPS-induced shock and did not inhibit the binding of LPS to RAW 264.7 cells or to human blood cells.196 N-linked glycosylation of C1-inhibitor is therefore required for its binding to LPS. This is also true of some other proteins that bind to LPS. Binding of LPS to its receptor requires N-linked glycosylation of both MD-2 and TLR4,190 which probably functions to maintain the conformation of these receptor components rather than as a binding site. The lipid-binding motifs of other LPS-interacting proteins197–201 contain clusters of positively charged amino acids. We therefore hypothesized that some or all of the four positively charged amino acids within the amino terminal domain of C1inhibitor (Arg-18, Lys-22, Lys-30 and Lys-55) might interact with the negatively charged phosphate groups of the diglucosamine backbone of lipid A and that one or more N-linked carbohydrates within this region (at Asn-3, Asn-47 and Asn-59) might be required to maintain the binding conformation. Using site-directed mutagenesis, we prepared single

ch23

Jan. 10, 2007

8:57

SPI-B429



567

amino acid substitutions at each site, in addition to combinations of substitutions.202 The results demonstrated that replacement of Asn-3 with Ala leads to nearly complete loss of the ability to interact with LPS, while little effect was observed with substitution at the other N-linked sites. On the other hand, the effect of replacement of the positively charged residues was additive. The near complete loss of the LPS interaction was observed only after replacement of all four amino acids. Based on these data, the following hypothesis was proposed: the binding site for LPS on C1 inibitor is discontinuous and contains Arg-18, Lys-22, Lys-30 and Lys-55. Removal of the carbohydrate at Asn3 is likely to result in a conformational rearrangement that obscures the positively charged binding site.

10. The Effect of C1-inhibitor on Leukocyte Migration As therapy with C1-inhibitor is useful in a number of different inflammatory diseases, because it has been shown to bind to endothelial cells under some conditions,203,204 and it is heavily glycosylated and has been shown to be fucosylated,205 we investigated the possibility that C1-inhibitor might express the sialyl LewisX tetrasaccharide and bind to selectins via this moiety. These studies showed that C1INH does express sialyl LewisX , and that it binds to both E- and P-selectins when expressed on the surface of CHO cells, and to the isolated immobilized selectins.25 It also binds to TNF-α-treated endothelial cells. In vitro, C1-inhibitor prevented the selectin-mediated adherence of the human U937 macrophage-like cell line to endothelial monolayers.25 Previously published data suggest that E-selectin expression is up-regulated during cold perfusion.206 It is therefore possible that the binding of C1-inhibitor to endothelial cells incubated in the cold is mediated via binding to E- or P-selectin. This may be a mechanism to concentrate C1-inhibitor on the endothelium at the sites of inflammation where it will be available for the modulation of both contact and complement system activation. It is also possible that the binding of C1-inhibitor may have a direct anti-inflammatory effect by competing with leukocytes for interaction with selectins. In support of this suggestion, recent data have extended these observations.207 C1-inhibitor interfered with the interaction of E-selectin with one of its ligands, carcinoembryonic antigen. Reactive center cleaved

ch23

Jan. 10, 2007

8:57

SPI-B429

568



C1-inhibitor inhibited binding to a great extent, but the N-deglycosylated protein completely lost this ability. C1-inhibitor also inhibited leukocyte rolling in vitro using a flow chamber model with both immobilized Eselectin and with P-selectin transfected cells. Inhibition was also demonstrated in vivo, using a mouse model of TNF-α-induced leukocyte rolling. These activities also were retained in reactive center cleaved C1-inhibitor, but were lost following N-deglycosylation. Lastly, C1-inhibitor suppressed the migration of leukocytes into the peritoneal cavity in sterile peritonitis induced with intraperitoneal injection of thioglycollate. This inhibition had the same characteristics as those described for the other assays. These data are consistent with the suggestion that inhibition results from the interaction of the sialyl-LewisX tetrasaccharide on C1-inhibitor with selectins. Therefore, in addition to its anti-inflammatory activities mediated via inhibition of complement and contact system proteases, these data suggest that plasma C1-inhibitor may play a direct role in modulating leukocyte adhesion during inflammation. Consistent with the observations described here, De Simoni et al.208 reported that the powerful neuroprotective effect of C1-inhibitor on brain ischemia-reperfusion injury is independent of C1qmediated activation of the classical complement pathway. All these data suggest that C1-inhibitor has broader based anti-inflammatory activities than what has been previously believed.

11. The Binding of C1 Inhibitor to Extracellular Matrix Components C1 inhibitor binds to a variety of constituents of the extracellular matrix, including type IV collagen, laminin, and entactin, as demonstrated by both blotting assay and ELISA.103 Digestion of C1-inhibitor with Crotalus atrox α-proteinase, which cleaves at Pro-36 had no effect on binding. Therefore, this binding, must differ in mechanism from the binding to gram-negative endotoxin which, as described above, requires the N-linked carbohydrate at *Asn-3 in addition, three of the four positively charged residues that participate in binding (Arg-18, Lys-22, Lys-30) are within this region. Although binding to these proteins did not alter rate constants for the reaction between C1-inhibitor and C1s, the interaction with collagen did result in an increase in the stoichiometry of inhibition. In addition to this binding, the mechanism

ch23

Jan. 10, 2007

8:57

SPI-B429



569

of which has not been determined, it has also been shown that tissue transglutaminase, but not factor XIII, can cross-link C1-inhibitor to immobilized fibrin.104 Using 3 H-putrescine, the cross-linked residue was shown to be Gln-453. The fibrin cross-linked C1-inhibitor retained inhibitory activity against its target proteinases. Although biologic studies remain to be done, it is quite likely that these binding activities provide a mechanism to immobilize and concentrate C1-inhibitor at extra-vascular sites of inflammation in order to maximize its ability to regulate local complement and contact system pathway activation.

12. C1-inhibitor Mediated Suppression of Alternative Complement Pathway Activation C1-inhibitor unexpectedly also has the ability to inhibit alternative pathway activation.209 It was shown to inhibit binding of factor B and C3 to alternative pathway activators and to block the ability of factor B to restore alternative pathway activity. Furthermore, removal of C1-inhibitor from serum resulted in enhanced alternative pathway activation. The data suggest that the mechanism almost certainly results from binding of C1-inhibitor to C3b, which inhibits factor B binding to C3b, analogous to the mechanisms of inhibition by both factor H and the C3b receptor, CR1. Furthermore, the binding and inhibition did not appear to require protease inhibitory activity, but it is not known which portion of the molecule mediates binding or if binding has any effect, either positive or negative, on protease inhibition. Unfortunately, this phenomenon has not been studied further, so it remains unclear whether this mechanism plays a significant role in vivo in comparison with the other mechanisms of alternative pathway regulation.

13. Therapeutic Uses of C1-inhibitor C1-inhibitor has a beneficial effect in a relatively wide variety of inflammatory disease models. These include most prominently ischemia-reperfusion injury, hyperacute transplant rejection, and gram-negative endotoxemia (discussed in previous section). In addition, a variety of data in both animal models and in limited human studies suggest that C1-inhibitor may be beneficial in pancreatitis,210–213 in traumatic and hemorrhagic shock,214,215 and

ch23

Jan. 10, 2007

8:57

SPI-B429


570


in the prevention of the vascular leak syndromes associated with thermal injury,216–218 IL-2 therapy,219–221 and cardiopulmonary bypass.222 In myocardial ischemia-reperfusion injury, C1-inhibitor, when given just before the time of reperfusion, resulted in decreased infarct size, myocardial neutrophil accumulation, plasma levels of creatine kinase, C3a and C5a, and decreased expression of P-selectin and ICAM-1 within the cardiac endothelium.223–226 Beneficial effects of C1-inhibitor have also been demonstrated in both skeletal muscle and middle cerebral artery ischemiareperfusion models.227–229 As described earlier, one study has suggested that the C1-inhibitor-mediated beneficial effect on brain injury is independent of classical complement pathway activation.208 Although the data are limited, some studies in humans have suggested a beneficial effect.230,231 In most instances of reperfusion injury, with the exception of at least some brain injury,208 the effect of C1-inhibitor is probably secondary to the inhibition of complement activation, which is clearly involved in the mediation of damage.232 Cardioprotection with a small molecule inhibitor of C1s supports this suggestion.233 The potential role of both contact system inhibition and of inhibition of selectin-mediated transmigration remain yet to be analyzed. Hyperacute transplant rejection in both allo- and xenotransplantation is complement mediated. Inhibition of natural antibody-mediated complement activation is required for the survival of xenotransplants. C1-inhibitor is beneficial in a variety of in vitro, ex vivo and in vivo models of xenotransplantation. In vitro, C1-inhibitor protects porcine or rodent endothelial or ovary cells from human complement-mediated damage.234–236 It also prolonged the survival of pig kidneys or lung perfused ex vivo with human blood,237,238 and reduced complement activation, and platelet and neutrophil activation. C1-inhibitor also protects pig kidneys from hyperacute rejection in cynomolgus monkeys.239,240

Conclusions Although there have now been a number of specific activities ascribed to C1-inhibitor, all these appear to relate to two primary biologic roles: the regulation of vascular permeability and the modulation of inflammation. The role of C1-inhibitor in the regulation of vascular permeability is most

ch23

Jan. 10, 2007

8:57

SPI-B429



571

obvious in patients with C1-inhibitor deficiency, in whom unregulated contact system activation leads to excessive generation of bradykinin with the induction of angioedema. The anti-inflammatory effects of C1-inhibitor relate to multiple specific activities, including inhibition of complement system activation via all three pathways, of contact system activation, of the activity of gram-negative endotoxin, and of leukocyte transmigration. Although not specifically discussed here, C1-inhibitor may also play roles in both blood pressure regulation and in coagulation/fibrinolysis.147,241,242 Bradykinin has a dramatic hypotensive effect; therefore, C1-inhibitor, by inhibition of kallikrein with the resulting suppression of bradykinin generation, plays an indirect role in blood pressure regulation. However, as kallikrein may also convert prorenin to renin, the overall effect of kallikrein regulation is complex. Lastly, bradykinin induces endothelial cell tPA release and inhibits thrombin-induced platelet aggregation, and kallikrein activates urokinase plasminogen activator. Therefore, inhibition of contact system activation results in an antifibrinolytic and procoagulant effect. In addition, C1-inhibitor may directly inhibit tPA under some circumstances. All these findings emphasize the fact that the biologic effects of C1-inhibitor are much broader than was originally appreciated.

References 1. Caliezi C, Wuillemin WA, Zeerleder S, Redondo M, Eisele B and Hack CE (2000) C1-Esterase inhibitor: an anti-inflammatory agent and its potential use in the treatment of diseases other than hereditary angioedema. Pharmacol Rev 52:91–112. 2. Davis III AE (1988) C1 inhibitor and hereditary angioneurotic edema. Ann Rev Immunol 6:595–628. 3. Kirschfink M (2002) C1-inhibitor and transplantation. Immunobiology 205:534–541. 4. Bergamaschini L and Cicardi M (2003) Recent advances in the use of C1 inhibitor as a therapeutic agent. Mol Immunol 40:155–158. 5. Cugno M, Nussberger J, Cicardi M and Agostoni A (2003) Bradykinin and the pathophysiology of angioedema. Int Immunopharmacol 3:311–317. 6. Davis III AE (2005) The pathophysiology of hereditary angioedema. Clin Immunol 114:3–9.

ch23

Jan. 10, 2007

8:57

SPI-B429

572



7. Tosi M, Duponchel C, Bourgarel P, Colomb M and Meo T (1986) Molecular cloning of human C1 inhibitor: sequence homologies with alpha-1antitrypsin and other members of the serpins superfamily. Gene 42:265–272. 8. Bock SC, Skriver K, Nielsen E, Thogersen HC, Wiman B, Donaldson VH, Eddy RL, Marrinan J, Radziejewska E, Huber R, Shows TB and Magnussen S (1986) Human C1 inhibitor: primary structure, cDNA cloning and chromosomal localization. Biochemistry 25:4292–4301. 9. Que BG and Petra PH (1986) Isolation and analysis of a cDNA coding for human C1 inhibitor. Biochem Biophys Res Commun 137:620–625. 10. Davis IIIAE, WhiteheadAS, Harrison RA, DauphinaisA, Bruns GAP, Cicardi M and Rosen FS (1986) Human inhibitor of the first component of complement, C1: characterization of cDNA clones and localization of the gene to chromosome 11. Proc Natl Acad Sci USA 83:3161–3165. 11. Rauth G, Schumacher G, Buckel P and Muller-Esterl W (1988) Molecular cloning of the cDNA coding for human C1 inhibitor. Protein Seq Data Anal 1:251–257. 12. Carter PE, Dunbar B and Fothergill JE (1988) Genomic and cDNA cloning of the human C1 inhibitor. Intron-exon junctions and comparison with other serpins. Eur J Biochem 173:163–169. 13. Huber R and Carrell RW (1989) Implications of the three-dimensional structure of alpha 1-antitrypsin for structure and function of serpins. Biochemistry 28:8951–8966. 14. Salvesen GS, Catanese JJ, Kress LF and Travis J (1985) Primary structure of the reactive site of human C1-inhibitor. J Biol Chem 260:2432–2436. 15. Theriault A, Whaley K, McPhaden A, Boyd E and Connor J (1989) Regional assignment of the human C1-inhibitor gene to 11q11-q13.1. Hum Genet 84:477–479. 16. Stoppa-Lyonnet D, Carter PE and Meo T (1990) Clusters of intragenic Alu repeats predispose the human C1 inhibitor locus to deleterious rearrangements. Proc Natl Acad Sci USA 87:1551–1555. 17. Stoppa-Lyonnet D, Duponchel C, Meo T, Laurent J, Carter PE, AralaChaves M, Cohen JHM, Dewald G, Goetz J, Hauptmann G, Lagrue G, Lesavre P, Lopez-Trascasa M, Misiano G, Moraine C, Sobel A, Spath PJ and Tosi M (1991) Recombinational biases in the rearranged C1-inhibitor genes of hereditary angioedema patients. Am J Hum Genet 49:1055–1062. 18. Ariga T, Carter PE and Davis III AE (1990) Recombinations between Alu repeat sequences that result in partial deletions within the C1 inhibitor gene. Genomics 8:607–613.

ch23

Jan. 10, 2007

8:57

SPI-B429



573

19. Carter P, Duponchel C, Tosi M and Fothergill J (1991) Complete nucleotide sequence of the gene for human C1 inhibitor with an unusually high density of Alu elements. Eur J Biochem 197:301–308. 20. Silverman GA, Bird PI, Carrell RW, Church FC, Coughlin, PB, Gettins PG, Irving JA, Lomas DA, Luke CJ, Moyer RW, Pemberton PA, RemoldO’Donnell E, Salvesen GS, Travis J and Whisstock JC (2001) The serpins are an expanding superfamily of structurally similar but functionally diverse proteins. Evolution, mechanism of inhibition, novel functions and a revised nomenclature. J Biol Chem 276:33293–33296. 21. Odermatt E, Berger H and Sano Y (1981) Size and shape of human C1-inhibitor. FEBS Lett 131:283–285. 22. Perkins S (1990) Hydrodynamic data show that C1-inhibitor of complement forms compact complexes with C1-r and C1-s. FEBS Lett 271:89–92. 23. Perkins SJ, Smith KF,Amatayakul S,Ashford D, Rademacher TW, Dwek RA, Lachmann PJ and Harrison RA (1990) Two-domain structure of the native and reactive centre cleaved forms of C1 inhibitor of human complement by neutron scattering. J Mol Biol 214:751–763. 24. Schoenberger OL (1992) Characterization of carbohydrate chains of C1-inhibitor and of desialylated C1-inhibitor. FEBS Lett 314:430–434. 25. Cai S and Davis III AE (2003) Complement regulatory protein C1inhibitor binds to selectins and interferes with endothelial-leukocyte adhesion. J Immunol 171:4786–4791. 26. Zhang F, Bries AD, Lang SC, Wang Q, Murhammer DW, Weiler JM and Linhardt RJ (2004) Metabolic alteration of the N-glycan structure of a protein from patients with a heterozygous protein deficiency. Biochem Biophys Acta 1739:43–49. 27. Simonovic I and Patston PA (2000) The native matastable fold of C1-inhibitor is stabilized by disulfide bonds. Biochim Biophys Acta 1481:97–102. 28. Bos IGA, Lubbers YTP, Roem D, Abrahams JP, Hack CE and Eldering E (2003) The functional integrity of the serpin domain of C1-inhibitor depends on the unique N-terminal domain, as revealed by a pathological mutant. J Biol Chem 29463–29470. 29. Patston PA, Hauert J, Michaud M and Schapira M (1995) Formation and properties of C1-inhibitor polymers. FEBS Lett 368:401–404. 30. Aulak KS, Eldering E, Hack CE, Lubbers YPT, Harrison RA, Mast A, Cicardi M and Davis III AE (1993) A hinge region mutation in C1-inhibitor (Ala436Thr) results in nonsubstrate-like behavior and in polymerization of the molecule. J Biol Chem 268:18088–18094.

ch23

Jan. 10, 2007

8:57

SPI-B429

574



31. Eldering E, Verpy E, Roem D, Meo T and Tosi M (1995) COOH-terminal substitutions in the serpin C1 inhibitor that cause loop overinsertion and subsequent multimerization. J Biol Chem 270:2579–2587. 32. Zahedi R, Aulak KS, Eldering E and Davis III AE (1996) Characterization of C1 inhibitor-Ta: A dysfunctional C1 inhibitor with deletion of lysine-251. J Biol Chem 271:24307–24312. 33. Prandini MH, Reboul A and Colomb MG (1986) Biosynthesis of complement C1 inhibitor by HepG2 cells. Reactivity of different glycosylated forms of the inhibitor with C1s. Biochem J 237:93–98. 34. Coutinho M, Aulak KS and Davis III AE (1994) Functional analysis of the serpin domain of C1 inhibitor. J Immunol 153:3648–3654. 35. Chang NS, Boackle RJ and Leu RW (1991) Characterization of C1-inhibitor binding to neutrophils. Immunology 73:95–101. 36. Minta JO (1981) The role of sialic acid in the functional activity and the hepatic clearance of C1-INH. J Immunol 126:245–249. 37. Lathem WW, Grys TE, Witowski SE, Torres AG, Kaper JB, Tarr PI and Welch RA (2002) StcE, a metalloprotease secreted by Escherichia coli O157:H7, specifically cleaves C1 esterase inhibitor. Mol Microbiol 45:277–288. 38. Lathem WW, Bergsbaken T and Welch RA (2004) Potentiation of C1 esterase inhibitor by StcE, a metalloprotease secreted by Escherichia coli O157:H7. J Exp Med 199:1077–1087. 39. Ravindran S, Grys TE, Welch RA, Schapira M and Patston P A (2004) Inhibition of plasma kallikrein by C1-inhibitor: Role of endothelial cells and the amino terminal domain of C1-inhibitor. Thromb Haemost 92:1277–1283. 40. Pensky J, Levy L and Lepow I (1961) Partial purification of a serum inhibitor of C’1 esterase. J Biol Chem 236:1674–1679. 41. Harpel PC and Cooper NR (1976) Studies on human plasma C1-inactivatorenzyme interactions. I. Mechanisms of interaction with C1s, plasmin and trypsin. J Clin Invest 55:593–604. 42. Ziccardi RJ and Cooper NR (1976) Activation of C1r by proteolytic cleavage. J Immunol 116:504–509. 43. Arlaud GJ, Reboul A, Sim RB and Colomb MG (1979) Interaction of C1-inhibitor with the C1r and C1s subcomponents in human C1. Biochim Biophys Acta 576:151–162. 44. Chesne S, Villiers CL, Arlaud GJ, Lacroix MB and Colomb MG (1982) Fluidphase interaction of C1 inhibitor (C1 Inh) and the subcomponents C1r and C1s of the first component of complement, C1. Biochem J 201:61–70.

ch23

Jan. 10, 2007

8:57

SPI-B429



575

45. Sim R, Arlaud G and Colomb M (1980) Kinetics of reaction of human C1-inhibitor with the human complement system proteases C1r and C1s. Biochim Biophys Acta 612:433–449. 46. Sim RB, Arlaud GJ and Colomb MG (1979) C1 inhibitor dependent dissociation of human complement component C1 bound to immune complexes. Biochem J 179:449–457. 47. Ziccardi RJ and Cooper NR (1979) Active disassembly of the first complement component C1 by C1-inhibitor. J Immunol 123:788–792. 48. Han ED, MacFarlane RC, Mulligan AN, Scafidi J and Davis III AE (2002) Increased vascular permeability in C1 inhibitor-deficient mice is mediated by the bradykinin type 2 receptor. J Clin Invest 109:1057–1063. 49. Schapira M, Scott CF and Colman RW (1981) Protection of human plasma kallikrein from inactivation by C1 inhibitor and other protease inhibitors. The role of high molecular weight kininogen. Biochemistry 20:2738–2743. 50. van der Graaf F, Koedam JA and Bouma BN (1983) Inactivation of kallikrein in human plasma. J Clin Invest 71:149–158. 51. Nilsson T (1983) On the interaction between human plasma kallikrein and C1-esterase inhibitor. Thromb Haemost 49:193–195. 52. Silverberg M, Longo J and Kaplan AP (1986) Study of the effect of high molecular weight kininogen upon the fluid-phase inactivation of kallikrein by C1 inhibitor. J Biol Chem 261:14965–14968. 53. de Agostini A, Patston PA, Marottoli V, Carrel S, Harpel PC and Schapira M (1988) A common neoepitope is created when the reactive center of C1inhibitoris cleaved by plasma kallikrein, activated factor XII fragment, C1 esterase, or neutrophil elastase. J Clin Invest 82:700–705. 54. Patston PA, Gettins P, Beechem J and Schapira M (1991) Mechanism of serpin action: evidence that C1 inhibitor functions as a suicide substrate. Biochemistry 30:8876–8882. 55. Nilsson T, Sjoholm I and Wiman B (1983) Structural and circular-dichroism studies on the interaction between human C1-esterase inhibitor and C1s. Biochem J 213:617–624. 56. de Agostini A, Schapira M, Wachtfogel YT, Colman RW and Carrel S (1985) Human plasma kallikrein and C1 inhibitor form a complex possessing an epitope that is not detectable on the parent molecules: Demonstration using a monoclonal antibody. Proc Natl Acad Sci USA 82:5190–5193. 57. Lennick M, Brew SA and Ingham KC (1985) Changes in protein conformation and stability accompany complex formation between human C1 inhibitor and C1-s. Biochemistry 24:2561–2568.

ch23

Jan. 10, 2007

8:57

SPI-B429

576



58. Nuijens JH, Huijbregts CC, van Mierlo GM and Hack CE (1987) Inactivation of C-1 inhibitor by proteases: Demonstration by a monoclonal antibody of a neodeterminant on inactivated, non-complexed C-1 inhibitor. Immunology 61:387–389. 59. Bruch M, Weiss V and Engel J (1988) Plasma serine proteinase inhibitors (serpins) exhibit major conformational changes and a large increase in conformational stability upon cleavage at their reactive sites. J Biol Chem 263: 16626–16630. 60. Skriver K, Wikkoff WR, Patston PA, Tausk F, Schapira M, Kaplan AP and Bock SC (1991) Substrate properties of C1 inhibitor Ma (alanine 434 glutamic acid). Genetic and structural evidence suggesting that the P12-region contains critical determinants of serine protease inhibitor/substrate status. J Biol Chem 266:9216–9221. 61. Davis III AE, Aulak K, Parad RB, Stecklein HP, Eldering E, Hack CE, Kramer J, Strunk RC, Bissler J and Rosen FS (1992) C1-inhibitor hinge region mutations produce dysfunction by different mechanisms. Nat Genet 1:354–358. 62. Perkins SJ, Smith KF, Nealis AS, Haris PI, Chapman D, Bauer CJ and Harrison RA (1992) Secondary structure changes stabilize the reactive-centre cleaved form of SERPINs. A study of 1H nuclear magnetic resonance and Fourier transform infrared spectroscopy. J Mol Biol 228:1235–1254. 63. Inman RD and Harpel PC (1983) C1 inactivator-C1s complexes in inflammatory joint disease. Clin Exp Immunol 53:521–528. 64. Harpel PC, Lewin MF and KaplanAP (1985) Distribution of plasma kallikrein between C1 inactivator and a2-macroglobulin in plasma utilizing a new assay for a2-macroglobulin-kallikrein complexes. J Biol Chem 260:4257–4263. 65. Wachtfogel YT, Harpel PC, Edmunds LH, Jr and Colman RW (1989) Formation of C1s-C1-inhibitor, kallikrein-C1-inhibitor and plasmin-alpha 2plasmin-inhibitor complexes during cardiopulmonary bypass. Blood 73: 468–471. 66. de Agostini A, Lijnen HR, Pixley RA, Colman RW and Schapira M (1984) Inactivation of factor-XII active fragment in normal plasma: predominant role of C1-inhibitor. J Clin Invest 93:1542–1549. 67. Kaplan AP, Gruber A and Harpel PC (1985) Assessment of Hageman factor activation in human plasma: Quantification of activated Hageman factor-C1 inactivator complexes by an enzyme-linked differential antibody immunosorbent assay. Blood 66:636–641. 68. Pixley RA, Schapira M and Colman RW (1985) The regulation of human factor XIIa by plasma proteinase inhibitors. J Biol Chem 260:1723–1729.

ch23

Jan. 10, 2007

8:57

SPI-B429



577

69. Meijers JC, Vlooswijk RA and Bouma BN (1988) Inhibition of blood coagulation factor XIa by C-1 inhibitor. Biochemistry 27:959–963. 70. Wuillemin WA, Minnema M, Meijers JC, Roem D, Eerenberg AJ, Nuijens JH, ten Cate H and Hack CE (1995) Inactivation of factor XIa in human plasma assessed by measuring factor XIa-protease inhibitor complexes: Major role for C1-inhibitor. Blood 85:1517–1526. 71. Wuillemin WA, Hack CE, Bleeker WK, Biemond BJ, Levi M and ten Cate H (1996) Inactivation of factor XIa in vivo: Studies in chimpanzees and in humans. Thromb Haemost 76:549–555. 72. Knauer DJ, Majumdar D, Fong PC and Knauer MF (2006) SERPIN regulation of factor XIa. The novel observation that protease nexin 1 in the presence of heparin is a more potent inhibitor of factor XIa than C1 inhibitor. J Biol Chem 275:37340–37346. 73. Cugno M, Bos I, Lubbers Y, Hack CE and Agostoni A (2001) In vitro interaction of C1-inhibitor with thrombin. Blood Coagul Fibrinolysis 12:253–260. 74. Levi M, Roem D, Kamp AM, de Boer JP, Hack CE and ten Cate JW (1993) Assessment of the relative contribution of different protease inhibitors to the inhibition of plasmin in vivo. Thromb Haemost 69:141–146. 75. Brown EW, Ravindran S and Patston PA (2002) The reaction between plasmin and C1-inhibitor results in plasmin inhibition by the serpin mechanism. Blood Coagul Fibrinolysis 13:711–714. 76. Booth NA, Walker E, Maughan R and Bennett B (1987) Plasminogen activator in normal subjects after exercise and venous occlusion: t-PA circulates as complexes with C1-inhibitor and PAI-1. Blood 69:1600–1604. 77. Bennett B, Croll A, Ferguson K and Booth NA (1990) Complexing of tissue plasminogen activator with PAI-1, alpha 2-macroglobulin and C1-inhibitor: Studies in patients with defibrination and a fibrinolytic state after electroshock or complicated labor. Blood 75:671–676. 78. Chandler WL, Levy WC and Stratton JR (1995) The circulatory regulation of TPA and UPA secretion, clearance and inhibition during exercise and during infusion of isoproterenol and phenylalanine. Circulation 92:2984–2994. 79. Sulikowski T and Patston PA (2001) The inhibition of TNK-t-PA by C1-inhibitor. Blood Coagul Fibrinolysis 12:75–77. 80. Heeb MJ and Espana F (1998) Alpha2-macroglobulin and C1-inactivator are plasma inhibitors of human glandular kallikrein. Blood Cells Mol Dis 24:412–419. 81. Faulmann EL, Young M and Boyle MD (1987) Inactivation of the proteolytic activity of mouse nerve growth factor by human C1-inhibitor. J Immunol 138:4336–4340.

ch23

Jan. 10, 2007

8:57

SPI-B429

578



82. Choi-Miura NH, Saito K, Takahashi K,Yoda M and Tomita M (2001) Regulation mechanism of the serine protease activity of plasma hyaluronan binding protein. Biol Pharm Bull 24:221–225. 83. Drouet C, Bouillet L, Csopaki F and Colomb MG (1999) Hepatitis C virus NS3 serine protease interacts with the serpin C1 inhibitor. FEBS Lett 458:415–418. 84. Aulak KS, Donaldson VH, Coutinho M and Davis III AE (1993) C1-inhibitor: Structure/function and biologic role. Behring Inst Mitt 204–213. 85. Ligoudistianou C, Xu Y, Garnier G, Circolo A and Volanakis JE (2004) A novel human complement-related protein, C1r-like protease (C1r-LP) specifically cleaves pro-C1s. Biochem J, Epub ahead of print, 2004. 86. Matsushita M, Thiel S, Jensenius JC and Terai Ift (2000) Proteolytic activities of two types of mannose-binding lectin associated serine protease. J Immunol 165:2637–2642. 87. Petersen SV, Thiel S, Jensen L, Vorup-Jensen T, Koch C and Jensenius JC (2000) Control of the classical and the MBL pathway of complement activation. Mol Immunol 37:803–811. 88. Rossi V, Cseh S, Bally I, Thielens NM, Jensenius JC and Arlaud GJ (2001) Substrate specificities of recombinant mannan-binding lectin-associated serine proteases-1 and -2. J Biol Chem 276:40880–40887. 89. Ambrus G, Gal P, Kojima M, Szilagyi K, Balczer J, Antal J, Graf L, Laich A, Moffatt BE, Schwaeble W, Sim RB and Zavadszky P (2003) Natural substrates and inhibitors of mannan-binding lectin-associated serine protease-1 and -2: A study on recombinant catalytic fragments. J Immunol 170:1374–1382. 90. Presanis JS, Hajela K, Ambrus G, Gal P and Sim RB (2004) Differential substrate and inhibitor profiles for human MASP-1 and MASP-2. Mol Immunol 40:921–929. 91. Huisman LG, van Griensven JM and Kluft C (1995) On the role of C1-inhibitor as inhibitor of tissue-type plasminogen activator in human plasma. Thromb Haemost 73:466–471. 92. Patston PA, Church FC and Olson ST (2004) Serpin-ligand interactions. Methods 32:93–109. 93. Olson ST, Bjork I and Bock SC (2002) Identification of critical molecular interactions mediating heparin activation of antithrombin: Implications for the design of improved heparin anticoagulants. Trends Cardiovasc Med 12:198–205. 94. Olson ST, Sheffer R and Francis AM (1993) High molecular weight kininogen potentiates the heparin-accelerated inhibition of plasma kallikrein by

ch23

Jan. 10, 2007

8:57

SPI-B429



95.

96. 97. 98.

99.

100.

101.

102.

103.

104. 105.

106.

579

antithrombin: Role for antithrombin in the regulation of kallikrein. Biochemistry 32:12136–12147. Wuillemin WA, Eldering E, Citarella F, de Ruig CP, ten Cate H and Hack CE (1996) Modulation of contact system proteases by glycosaminoglycans: selective enhancement of the inhibition of factor XIa. J Biol Chem 271:12913–12918. Caughman GB, Boackle RJ and Vesely J (1982) A postulated mechanism for heparin’s potentiation of C1 inhibitor function. Mol Immunol 19:287–295. Lennick M, Brew SA and Ingham KC (1986) Kinetics of interaction of C1 inhibitor with complement C1s. Biochemistry 25:3890–3898. Hortin GL and Trimpe BL (1991) C1 inhibitor: different mechanisms of reaction with complement component C1 and C1s. Immunol Invest 20: 75–82. Wuillemin WA, te Velthuis H, Lubbers YT, de Ruig CP, Eldering E and Hack CE (1997) Potentiation of C1 inhibitor by glycosaminoglycans: Dextran sulfate species are effective inhibitors of in vitro complement activation in plasma. J Immunol 159:1953–1960. Caldwell EE, Andreasen AM, Blietz MA, Serrahn JN, VanderNoot V, Park Y, Yu G, Linhardt RJ and Weiler JM (1999) Heparin binding and augmentation of C1 inhibitor activity. Arch Biochim Biophys 361:215–222. Bos IG, van Mierlo GJ, Bleeker WK, Rigter GM, te Velthuis H, Dickneite G and Hack CE (2001) The potentiation of human C1-inhibitor by dextran sulphate is transient in vivo: Studies in a rat model. Int Immunopharmacol 1:1583–1595. Weiss V and Engel J (1983) Heparin-stimulated modification of C1-inhibitor by subcomponent C1s of human complement. Hoppe Seylers Z Physiol Chem 364:295–301. Patston PA and Schapira M (1997) Regulation of C1-inhibitor function by binding to type IV collagen and heparin. Biochem Biophys Res Commun 230:597–601. Hauert J, Patston PA and Schapira M (2000) C1 inhibitor cross-linking by tissue transglutaminase. J Biol Chem 275:14558–14562. de Smet BJ, de Boer, J-P, Agterberg J, Rigter G, Bleeker WK and Hack CE (1993) Clearance of human native, proteinase-complexed and proteolytically inactivated C1-inhibitor in rats. Blood 81:56–61. Patston PA, Qi M, Schifferli JA and Schapira M (1995) The effect of cleavage by a Crotalus atrox alpha-proteinase fraction on the properties of C1-inhibitor. Toxicon 33:53–61.

ch23

Jan. 10, 2007

8:57

SPI-B429

580



107. Malek R, Aulak KS and Davis III AE (1996) The catabolism of intact, reactive centre-cleaved and proteinase complexed C1 inhibitor in the guinea pig. Clin Exp Immunol 105:191–197. 108. Storm D, Herz J, Trinder P and Loos M (1997) C1 inhibitor-C1s complexes are internalized and degraded by the low density lipoprotein receptor-related protein. J Biol Chem 272:31043–31050. 109. Zuraw B and Lotz M (1990) Regulation of the hepatic synthesis of C1 inhibitor by the hepatocyte stimulating factors interleukin 6 and interferon gamma. J Biol Chem 265:12664–12670. 110. Yeung LaiwahAC, Jones L, HamiltonAO and Whaley K (1985) Complementsubcomponent-C1-inhibitor synthesis by human monocytes. Biochem J 226:199–205. 111. Lappin D, Birnie G and Whaley K (1990) Modulation by interferons of the expression of monocyte complement genes. Biochem J 268:387–392. 112. Schmaier AH, Murray SC, Heda GD, Farber A, Kuo A, McCrae K and Cines DB (1989) Synthesis and expression of C1 inhibitor by human umbilical vein endothelial cells. J Biol Chem 264:18173–18179. 113. Schmaier AH, Amenta S, Xiong T, Heda GD and Gewirtz AM (1993) Expression of platelet C1 inhibitor. Blood 82:465–474. 114. KatzY and Strunk R (1989) Synthesis and regulation of C1 inhibitor in human skin fibroblasts. J Immunol 142:2041–2045. 115. Walker DG, Yasuhara O, Patston PA, McGeer EG and McGeer PL (1995) Complement C1 inhibitor is produced by brain tissue and is cleaved in Alzheimer disease. Brain Res 675:75–82. 116. Veerhuis R, Janssen I, Hack CE and Eikelenboom P (1996) Early complement components in Alzheimer’s disease brains. Acta Neuropathol 91:53–60. 117. Zahedi K, Prada AE, Prada JA and Davis III AE (1997) Characterization of the interferon-gamma responsive element in the 5’ flanking region of the C1 inhibitor gene. J Immunol 159:6091–6096. 118. Hamilton AO, Jones L, Morrison L and Whaley K (1987) Modulation of monocyte complement synthesis by interferons. Biochem J 242: 809–815. 119. Lappin D, Guc D, Hill A, McShane T and Whaley K (1992) Effect of interferon-gamma on complement gene expression in different cell types. Biochem J 281:437–442. 120. Patston PA, Medcalf RL, Kourteva Y and Schapira M (1993) C1-inhibitorserine proteinase complexes and the biosynthesis of C1-inhibitor by Hep G2 and U 937 cells. Blood 82:3371–3379.

ch23

Jan. 10, 2007

8:57

SPI-B429



581

121. Lotz M and Zuraw BL (1987) Interferon-gamma is a major regulator of C1-inhibitor synthesis by human blood monocytes. J Immunol 139: 3382–3387. 122. Heda G, Mardente S, Weiner L and Schmaier A (1990) Interferon gamma increases in vitro and in vivo expression of C1 inhibitor. Blood 75:2401–2407. 123. Zahedi K, Prada A and Davis III AE (1994) Transcriptional regulation of the C1 inhibitor gene by gamma interferon. J Biol Chem 269:9669–9674. 124. Heda GD, Kehoe KJ, Mahdi F and Schmaier AH (1996) Phosphatase 2A participates in interferon-gamma’s induced upregulation of C1 inhibitor mRNA expression. Blood 87:2831–2838. 125. Liu D, Scafidi J, Prada AE, Zahedi K and Davis III AE (2002) Nuclear phosphatases and the proteasome in suppression of STAT1 activity in hepatocytes. Biochem Biophys Res Commun 299:574–580. 126. Han Lee E, Pappalardo E, Scafidi J and Davis III AE (2003) Approaches toward reversal of increased vascular permeability in C1 inhibitor deficient mice. Immunol Letters 89:155–160. 127. Eldering E, Nuijens JH and Hack CE (1988) Expression of functional human C1 inhibitor in COS cells. J Biol Chem 263:11776–11779. 128. Lamark T, Ingebrigtsen M, Bjornstad C, Melkko T, Mollnes TE and Nielsen EW (2001) Expression of active human C1 inhibitor serpin domain in Escherichia coli. Protein Expr Purif 22:349–358. 129. Wolff MW, Zhang F, Roberg JJ, Caldwell EEO, Kaul PR, Serrahn JN, Murhammar DW, Linhardt RJ and Weiler JW (2001) Expression of C1 esterase inhibitor by the baculovirus expression vector system: Preparation, purification and characterization. Protein Expression and Purification 22:414–421. 130. Eldering E, Huijbregts CC, Lubbers YT, Longstaff C and Hack CE (1992) Characterization of recombinant C1 inhibitor P1 variants. J Biol Chem 267:7013–7020. 131. Eldering E, Haijbregts CC, Nuijens JH, Verhoeven AJ and Hack CE (1993) Recombinant C1 inhibitor P5/P3 variants display resistance to catalytic inactivation by stimulated neutrophils. J Clin Invest 92:1035–1043. 132. Zahedi R, Bissler JJ, Davis III AE, Andreadis C and Wisnieske JJ (1995) Unique C1 inhibitor dysfunction in a kindred without angioedema. II. Identification of an Ala443-Val substitution and functional analysis of the recombinant mutant protein. J Clin Invest 95:1299–1305. 133. Zahedi R, MacFarlane RC, Wisnieski JJ and Davis III AE (2001) C1 inhibitor: Analysis of the role of amino acid residues within the reactive center loop in target protease recognition. J Immunol 167:1500–1506.

ch23

Jan. 10, 2007

8:57

SPI-B429

582



134. Koles K, van Berkel PH C, Mannesse MLM, Zoetemelk R, Vliegenthart JF and Kamerling JP (2004) Influence of lactation parameters on the Nglycosylation of recombinant human C1 inhibitor isolated from the milk of transgenic rabbits. Glycobiology 14:979–986. 135. Koles K, van Berkel PHC, Pieper FR, Nuijens JH, Mannesse MLM, Vliegenthart FG and Kamerling JP (2004) N- and O-glycans of recombinant human C1 inhibitor expressed in the milk of transgenic rabbits. Glycobiology 14:51–64. 136. Grunenfelder J, Inderbitzin D, Zund G, Avital I, Burkhardt T, Candinas D and Turina M (2002) Increased susceptibility of the left compared to the right ventricle to remote ischemia/reperfusion injury in human C1-inhibitoroverexpressing transgene mice. J Invest Surg 15:281–286. 137. Inderbitzin D, Beldi G, Avital I, Vinci G and Candinas D (2004) Local and remote ischemia-reperfusion injury is mitigated in mice overexpressing human C1 inhibitor. Eur Surg Res 36:142–147. 138. Vinci G, Lynch NJ, Duponchel C, Lebastard TM, Milon G, Stover C, Schwaeble W and Tosi M (2002) In vivo biosynthesis of endogenous and of human C1 inhibitor in transgenic mice: Tissue distribution and colocalization of their expression. J Immunol 169:5948–5954. 139. Woo P, Lachmann PJ, Harrison RA, Amos N, Cooper C and Rosen FS (1985) Simultaneous turnover of normal and dysfunctional C1 inhibitor as a probe of in vivo activation of C1 and contact activatable proteases. Clin Exp Immunol 61:1–8. 140. Lachmann P and Rosen F (1984) The catabolism of C1-inhibitor and the pathogenesis of hereditary angioedema. Acta Pathol Microbiol Immunol Scand 284 (Suppl):35–39. 141. Ernst SC, Circolo A, Davis III AE, Gheesling-Mullis K, Fliesler M and Strunk RC (1996) Impaired production of both normal and mutant C1 inhibitor proteins in type I hereditary angioedema with a duplication in exon 8. J Immunol 157:405–410. 142. Kramer J, Rosen F, Colten H, Rajezy K and Strunk R (1993) Transinhibition of C1 inhibitor synthesis in type I hereditary angioneurotic edema. J Clin Invest 91:1258–1262. 143. Nussberger J, Cugno M, Amstutz C, Cicardi M, Pellacani A and Agostoni A (1998) Plasma bradykinin in angio-oedema. Lancet 351:1693–1697. 144. Laurell AB, Martensson U and Sjoholm AG (1976) C1 subcomponent complexes in normal and pathological sera studied by crossed immunoelectrophoresis. Acta Pathol Microbiol Scand 84:455–464.

ch23

Jan. 10, 2007

8:57

SPI-B429



583

145. Buhler R, Hovinga JK, Aebi-Huber I, Furlan M and Lammle B (1995) Improved detection of proteolytically cleaved high molecular weight kininogen by immunoblotting using an antiserum against its reduced 47 kDa light chain. Blood Coagul Fibrinolysis 6: 223–232. 146. Berrettini M, Lammle B, White T, Heeb MJ, Schwarz HP, Zuraw B, Curd J and Griffin JH (1986) Detection of in vitro and in vivo cleavage of high molecular weight kininogen in human plasma by immunoblotting with monoclonal antibodies. Blood 68:455–462. 147. Nielsen EW, Johansen HT, Hogasen K, Wuillemin W, Hack CE and Mollnes TE (1996) Activation of the complement, coagulation, fibrinolytic and kallikrein-kinin systems during attacks of hereditary angioedema. Scand J Immunol 44:185–192. 148. Cugno M, Cicardi M, Coppola R and Agostoni A (1996) Activation of factor XII and cleavage of high molecular weight kininogen during acute attacks in hereditary and acquired C1-inhibitor deficiencies. Immunopharmacology 33:361–364. 149. Curd JG, Yelvington M, Burridge N, Stimler NP, Gerard CLJ, Prograis J, Cochrane CG and Muller-Eberhard HJ (1982) Generation of bradykinin during incubation of hereditary angioedema plasma. Mol Immunol 19:1365 (Abstract). 150. Curd JGLJ, Prograis J and Cochrane CG (1980) Detection of active kallikrein in induced blister fluids of hereditary angioedema patients. J Exp Med 152:742–747. 151. Fields T, Ghebrewihet B and Kaplan A (1983) Kinin formation in hereditary angioedema plasma: evidence against kinin derivation from C2 and in support of spontaneous formation of bradykinin. J. Allergy Clin Immunol 72:54–60. 152. Donaldson VH (1973) Kinin formation in hereditary angioneurotic edema (HANE) plasma. Int Arch Allergy 45:206–209. 153. Donaldson VH, Merler E, Rosen FS, Kretschmer KW and Lepow IH (1970) A polypeptide kinin in hereditary angioneurotic edema plasma: Role of complement in its formation. J Lab Clin Med 76:986. 154. Donaldson VH, Ratnoff OD, Silva WDd and Rosen FS (1969) Permeabilityincreasing activity in hereditary angioneurotic edema plasma. J Clin Invest 48:642–653. 155. Strang CJ, Auerbach KS and Rosen FS (1986) C1s-induced vascular permeability in C2-deficient guinea pigs. J Immunol 137:631–635. 156. Strang C, Cholin S, Spragg J, Davis III AE, Schneeberger E, Donaldson V and Rosen F (1988) Angioedema induced by a peptide derived from complement component C2. J Exp Med 168:1685–1698.

ch23

Jan. 10, 2007

8:57

SPI-B429

584



157. Shoemaker LR, Schurman SJ, Donaldson VH and Davis IIIA E (1994) Hereditary angioneurotic edema: Characterization of plasma kinin and vascular permeability-enhancing activities. Clin Exp Immunol 95:22–28. 158. Wisnieski JJ, Knauss TC, Yike I, Dearborn DG, Narvy RL and Naff GB (1994) Unique C1 inhibitor dysfunction in a kindred without angioedema. I. A mutant C1 inhibitor that inhibits C1s but not C1r. J Immunol 152:3199–3209. 159. Zahedi R, Wisnieski J and Davis III AE (1997) Role of the P2 residue of complement 1 inhibitor (Ala443) in determination of target protease specificity. J Immunol 159:983–988. 160. Gompels MM, Lock RJ, Abinun M, Bethune CA, Davies G, Grattan C, Fay AC, Longhurst HJ, Morrison L, Price A, Price M and Watters D (2005) C1 inhibitor deficiency: Consensus document. Clin Exp Immunol 139:379–394. 161. Bowen T, Cicardi M, Farkas H, Bork K, Kreuz W, Zingale L, Varga L, Martinez-Saguer I, Aygoren-Pursun E, Binkley KE, Zuraw B, Davis III AE, Hebert J, Ritchie B, Burnham J, Castaldo AJ, Menendez A, Nagy I, Harmat G, Bucher C, Lacuesta G, Issekutz A, Warrington R, Yang W, Dean J, Kanani A, Stark D, McCusker C, Wagner E, Rivard GE, Leith E, Tsai E, MacSween M, Lyanga J, Serushago B, Leznoff A, Waserman S and de Serres J (2004) Canadian 2003 International Consensus Algorithm for the Diagnosis, Therapy and Management of Hereditary Angioedema. J Allergy Clin Immunol 114:629–637. 162. Schmaier AH (2002) The plasma kallikrein-kinin system counterbalances the renin-angiotensin system. J Clin Invest 109:1007–1009. 163. Agostoni A and Cicardi M (1992) Hereditary and acquired C1-inhibitor deficiency: biological and clinical characteristics in 235 patients. Medicine 71:206–215. 164. Frank M, Gelfand F and Atkinson J (1976) Hereditary angioedema: the clinical syndrome and its management. Ann Int Med 84:580–593. 165. Eriksson H and Sjogren HO (1995) Inhibition of activation of human T lymphocytes by the complement C1 esterase inhibitor. Immunology 86:304–310. 166. Nissen MH, Bregenholt S, Nording JA and Claesson MH (1998) C1-esterase inhibitor blocks T lymphocyte proliferation and cytotoxic T lymphocyte generation in vitro. Int Immunol 10:167–173. 167. Patrick RA, Hollers JC, Liu DY, Giese BH and Smith CW (1980) Effects of human complement component 1 inactivator on neutrophil chemotaxis and chemotactic deactivation. Infect Immun 28:700–707.

ch23

Jan. 10, 2007

8:57

SPI-B429



585

168. Coppola L, Guastafierro S, Verrazzo G, Coppola A, De Lucia D and Tirelli A (2002) C1 inhibitor infusion modifies platelet activity in hereditary angioedema patients. Arch Pathol Lab Med 126:842–845. 169. Guerrero R, Velasco F, Rodriguez M, Lopez A, Rojas R, Alvarez MA, Villalba R, Rubio V, Torres A and Castillo Dd (1993) Endotoxin-induced pulmonary dysfunction is prevented by C1-esterase inhibitor. J Clin Invest 91:2754–2760. 170. Scherer RU, Giebler RM, Schmidt U, Paar D and Kox WJ (1996) The influence of C1-esterase inhibitor substitution on coagulation and cardiorespiratory parameters in an endotoxin-induced rabbit model of hypercoagulability. Semin Thrombo Hemost 22:357–366. 171. Giebler R, Schmidt U, Koch S, Peters J and Scherer R (1999) Combined antithrombin III and C1-esterase inhibitor treatment decreases intravascular fibrin deposition and attenuates cardiorespiratory impairment in rabbits exposed to Escherichia coli endotoxin. Crit Care Med 27:597–604. 172. Schmidt W, Stenzel Z, Gebhard MM, Martin E and Schmidt H (1999) C1-esterase inhibitor and its effects on endotoxin-induced leukocyte adherence and plasma extravasation in postcapillary venules. Surgery 125: 280–287. 173. Schmidt W, Stenzel K, Walther A, Gebhard MM, Martin E and Schmidt H (1999) Influence of C1-esterase inhibitor on tissue oxygenation of jejunal mucosa during endotoxemia. Int J Surg Investig 1:277–283. 174. Jansen PM, Eisele B, de Jong IW, Chang A, Delvos U, Taylor JFB and Hack CE (1998) Effect of C1 inhibitor on inflammatory and physiologic response patterns in primates suffering from lethal septic shock. J Immunol 160: 475–484. 175. Fischer MB, Prodeus AP, Nicholson-Weller A, Ma M, Murrow J, Reid RR, Warren HB, Lage AL, Moore JFD, Rosen FS and Carroll MC (1997) Increased susceptibility to endotoxin shock in complement C3- and C4deficient mice is corrected by C1 inhibitor replacement. J Immunol 159: 976–982. 176. Barton PA and Warren JS (1993) Complement component C5 modulates the systemic tumor necrosis factor response in murine endotoxic shock. Infect Immun 61:1474–1481. 177. Czermak BJ, Sarma V, Pierson CL, Warner RL, Huber-Lang M, Bless NM, Schmal H, Priedl HP and Ward PA (1999) Protective effects of C5a blockage in sepsis. Nature Med 5:788–792.

ch23

Jan. 10, 2007

8:57

SPI-B429

586



178. Strachan AJ, Woodruff TM, Haaima G, Fairlie DP and Taylor SM (2000) A new small molecule C5a receptor antagonist inhibits the reverse-passive Arthus reaction and endotoxic shock in rats. J Immunol 164:6560–6565. 179. Colman RW (1999) Biologic activities of the contact factors in vivo. Potentiation of hypotension, inflammation and fibrinolysis and inhibition of cell adhesion, angiogenesis and thrombosis. Thromb Haemost 82:1568–1577. 180. Pixley RA, De La Cadena R, Page JD, Kaufman N, Wyshock EG, Chang A, Taylor Jr FB and Colman RW (1993) The contact system contributes to hypotension but not disseminated intravascular coagulation in lethal bacteremia. In vivo use of a monoclonal anti-factor XII antibody to block contact activation in baboons. J Clin Invest 91:61–68. 181. Jansen PM, Pixley RA, Brouwer M, de Jong IW, Chang AC, Hack CE, Taylor Jr FB and Colman RW (1996) Inhibition of factor XII in septic baboons attenuates the activation of complement and fibrinolytic systems and reduces the release of interleukin-6 and neutrophil elastase. Blood 87:2337–2344. 182. Feletou M, Jamonneau I, Germain M, Thurieau C, Fauchere JL, Villa P, Ghezzi P and Canet E (1996) Bradykinin B2 receptor involvement in rabbit and murine models of septic shock. J Cardiovasc Pharmacol 27:500–507. 183. Liu D, Cai S, Gu X, Scafidi J, Wu X and Davis III AE (2003) C1 inhibitor prevents endotoxin shock via a direct interaction with lipopolysaccharide. J Immunol 171:2594–2601. 184. Glauser MP, Zanetti G, Baumgartner JD and Cohen J (1991) Septic shock: Pathogenesis. Lancet 338:732–736. 185. Poltorak A, He X, Smirnova I, Liu MY, van Huffel C, Du X, Birdwell D, Alejos E, Silva M, Galanos C, Freudenberg M, Ricciardi-Castagnoli P, Layton B and Beutler B (1998) Defective LPS signaling in C3H/HeJ and C57Bl/10ScCr mice: Mutations in the Tlr4 gene. Science 282: 2085–2088. 186. Schumann RR and Latz E (2000) Lipopolysaccharide-binding protein. Chem Immunol 74:42–60. 187. Kitchens RL (2000) Role of CD14 in cellular recognition of bacterial lipopolysaccharides. Chem Immunol 74. 188. Fenton MJ and Golenbock DT (1998) LPS-binding proteins and receptors. J Leukocyte Biol 64:25–32. 189. Aderem A and Ulevitch RJ (2000) Toll-like receptors in the induction of the innate immune response. Nature 406:782–787. 190. da Silva Correia J and Ulevitch RJ (2002) MD-2 and TLR4 N-linked glycoproteins are important for a functional lipopolysaccharide receptor. J Biol Chem 277:1845–1854.

ch23

Jan. 10, 2007

8:57

SPI-B429



587

191. Stelter F, Bernheiden M, Menzel R, Jack RS, Witt S, Fan X, Pfister M and Schutt C (1997) Mutation of amino acids 39-44 of human CD14 abrogates binding of lipopolysaccharide and Escherichia coli. Eur J Biochem 243: 100–109. 192. Stelter F, Loppnow H, Menzel R, Grunwald U, Bernheiden M, Jack RS, Ulmer AJ and Schutt C (1999) Differential impact of substitution of amino acids 9-13 and 91-101 of human CD14 on soluble CD14-dependent activation of cells by lipopolysaccharide. J Immunol 163:6035–6044. 193. Viriyakosol S and Kirkland TN (1995) A region of human CD14 required for lipopolysaccharide binding. J Biol Chem 270:361–368. 194. Viriyakosol S, Mathison JC, Tobias PS and Kirkland TN (2000) Structurefunction analysis of CD14 as a soluble receptor for lipopolysaccharide. J Biol Chem 275:3144–3149. 195. Juan TS, Hailman E, Kelley MJ, Busse LA, Davy E, Empig CJ, Narhi LO, Wright SD and Lichenstein HS (1995) Identification of a lipopolysaccharide binding domain in CD14 between amino acids 57 and 64. J Biol Chem 270:5219–5224. 196. Liu D, Gu X, Scafidi J and Davis III AE (2004) N-linked glycosylation is required for C1 inhibitor-mediated protection from endotoxin shock in mice. Infect Immun 72:1946–1955. 197. Schumann RR, Lamping N and Hoess A (1997) Interchangeable endotoxinbinding domains in proteins with opposite lipopolysaccharide-dependent activities. J Immunol 159:5599–5605. 198. Ferguson AD, Welte W, Hofmann E, Lindner B, Holst O, Coulton JW and Diederichs K (2000) A conserved structural motif for lipopolysaccharide recognition by procaryotic and eucaryotic proteins. Structure 8:585–592. 199. Mancek M, Pristovsek P and Jerala R (2002) Identification of LPS-binding peptide fragment of MD-2, a Toll-receptor accessory protein. Biochem Biophys Res Commun 292:880–885. 200. Lamping N, Hoess A, Yu B, Park TC, Kirschning, C-J, Pfeil C, Reuter D, Wright SD, Herrman F and Schumann RR (1996) Effects of site directed mutagenesis of basic residues (Arg 94, Lys 95, Lys 99) of LPS-binding protein on binding and transfer of LPS and subsequent immune cell activation. J Immunol 157:4648–4656. 201. Visintin A, Latz E, Monks BG, Espevik T and Golenbock DT (2003) Lysines 128 and 132 enable lipopolysaccharide binding to MD-2, leading to toll-like receptor-4 aggregation and signal transduction. J Biol Chem 278: 48313–48320.

ch23

Jan. 10, 2007

8:57

SPI-B429

588



202. Liu D, Cramer CC, Scafidi J and Davis III AE (2005) N-linked glycosylation at Asn3 and the positively charged residues within the amino terminal domain of C1 inhibitor are required for its interaction with Salmonella typhimurium lipopolysaccharide and lipid A. Infect Immun 73:4478–4487. 203. Bergamaschini L, Gobbo G, Gatti S, Caccamo L, Prato P, Maggioni M, Braidotti P, Di Stefano R and Fassati LR (2001) Endothelial targeting with C1-inhibitor reduces complement activation in vitro and during ex vivo reperfusion of pig liver. Clin Exp Immunol 126:412–420. 204. Bergamaschini L, Gatti S, Caccamo L, Prato P, Latham L, Trezza P, Maggioni M, Gobbo G and Fassati LR (2001) C1 inhibitor potentiates the protective effect of organ preservation solution on endothelial cells during cold storage. Transplant Proc 33:939–941. 205. Strecker G, Ollier-Hartmann MP, van Halbeek H, Vliegenthart J F, Montreuil J and Hartmann L (1985) Primary structure of the glycan chains of normal C1 esterase inhibitor (C1-INH) after NMR analysis at 400 MHz. C R Acad Sci III 301:571–576. 206. Takada M, Nadeau KC, Shaw GD, Marquette KA and Tilney NL (1997) The cytokine-adhesion molecule cascade in ischemia/reperfusion injury of the rat kidney. Inhibition by a soluble P-selectin ligand. J Clin Invest 99: 2682–2690. 207. Cai S, Dole V, Bergmeier W, Scafidi J, Feng H, Wagner DD and Davis III AE (2005) A direct role for C1 inhibitor in regulation of leukocyte activation. J Immunol 174:6462–6466. 208. De Simoni MG, Rossi E, Storini C, Pizzimenti S, Echart C and Bergamaschini L (2004) The powerful neuroprotective action of c1-inhibitor on brain ischemia-reperfusion injury does not require C1q. Am J Pathol 164: 1857–1863. 209. Jiang H, Wagner E, Zhang H and Frank MM (2001) Complement 1 inhibitor is a regulator of the alternative complement pathway. J Exp Med 194: 1609–1616. 210. Testoni PA, Cicardi M, Bergamaschini L, Guzzoni S, Cugno M, Buizza M, Bagnolo F and Agostoni A (1995) Infusion of C1-inhibitor plasma concentrate prevents hyperamylasemia induced by endoscopic sphincterotomy. Gastrointest Endosc 42:301–305. 211. Yamaguchi H, Weidenbach H, Lurs H, Lerch MM, Dickneite G and Adler G (1997) Combined treatment with C1 esterase inhibitor and antithrombin III improves survival in severe acute experimental pancreatitis. Gut 40: 531–535.

ch23

Jan. 10, 2007

8:57

SPI-B429



589

212. Schneider DT, Nurnberger W, Stannigel H, Bonig H and Gobel U (1999) Adjuvant treatment of severe acute pancreatitis with C1 esterase inhibitor concentrate after haematopoietic stem cell transplantation. Gut 45:733–736. 213. Niederau C, Brinsa R, Niederau M, Luthen R, Strohmeyer G and Ferrell LD (1995) Effects of C1-esterase inhibitor in three models of acute pancreatitis. Int J Pancreatol 17:189–196. 214. Kochilas L, Campbell B, Scalia R and Lefer AM (1997) Beneficial effects of C1 esterase inhibitor in murine traumatic shock. Shock 8:165–169. 215. Horstick G, Kempf T, Lauterbach M, Bhakdi S, Kopacz L, Heimann A, Malzahn M, Horstick M, Meyer J and Kempski O (2001) C1-esteraseinhibitor treatment at early reperfusion of hemorrhagic shock reduces mesentery leukocyte adhesion and rolling. Microcirculation 8:427–433. 216. Radke A, Mottaghy K, Goldmann C, Khorram-Sefat R, Kovacs B, Janssen A, Klosterhalfen B, Hafemann B, Pallua N and Kirschfink M (2000) C1 inhibitor prevents capillary leakage after thermal trauma. Crit Care Med 28: 3224–3232. 217. Khorram-Sefat R, Goldmann C, Radke A, Lennartz A, Mottaghy K, Afify M, Kupper W and Klosterhalfen B (1998) The therapeutic effect of C1-inhibitor on gut-derived bacterial translocation after thermal injury. Shock 9:101–108. 218. Henze U, LennartzA, Hefemann B, Goldmann C, Kirkpatrick CJ and Klosterhalfen B (1997) The influence of the C1-inhibitor BERINERT and the proteinfree haemodialsateACTIHAEMYL20% on the evolution of the depth of scald burns in a porcine model. Burns 23:473–477. 219. Ogilvie AC, Baars JW, Eerenberg AJM, Hack CE, Pinedo HM, Thijs LG and Wagstaff J (1994) A pilot study to evaluate the effects of C1 esterase inhibitor on the toxicity of high-dose interleukin 2. Br J Cancer 69:596–598. 220. Hack CE, Ogilvie AC, Eisele B, Jansen PM, Wagstaff J and Thijs LG (1994) Initial studies on the administration of C1-esterase inhibitor to patients with septic shock or with a vascular leak syndrome induced by interleukin-2 therapy. Prog Clin Biol Res 388:335–357. 221. Hack CE, Ogilvie AC, Eisele B, Eerenberg AJ, Wagstaff J and Thijs LG (1993) C1-inhibitor substitution therapy in septic shock and in the vascular leak syndrome induced by high doses of interleukin-2. Intens Care Med 19 (Suppl 1):S19–S28. 222. Tassani P, Kunkel R, Richter JA, Oechsler H, Lorenz HP, Braun SL, Eising GP, Haas F, Paek SU, Bauernschmitt R, Jochum M and Lange R (2001) Effect of C1-esterase-inhibitor on capillary leak and inflammatory response syndrome during arterial switch operations in neonates. J Cardiothorac Vasc Anesth 15:469–473.

ch23

Jan. 10, 2007

8:57

SPI-B429

590



223. Buerke M, Murohara T and Lefer AM (1995) Cardioprotective effects of a C1 esterase inhibitor in myocardial ischemia and reperfusion. Circulation 91:393–402. 224. Buerke M, Prufer D, Dahm M, Oelert H, Meyer J and Darius H (1998) Blocking of classical complement pathway inhibits endothelial adhesion molecule expression and preserves ischemic myocardium from reperfusion injury. J Pharmacol Exp Ther 286:429–438. 225. Horstick B, Berg O, Heimann A, Gotze O, Loos M, Hafner G, Bierbach B, Petersen S, Bhakdi S, Darius H, Horstick M, Meyer J and Kempski O (2002) Application of C1-esterase inhibitor during reperfusion of ischemic myocardium: dose-related beneficial versus detrimental effects. Circulation 104:3125–3131. 226. Horstick G, Heimann A, Gotze O, Hafner G, Berg O, Bohmer P, Becker P, Darius H, Rupprecht HJ, Loos M, Bhakdi S, Meyer J and Kempski O (1997) Intracoronary application of C1 esterase inhibitor improves cardiac function and reduces necrosis in an experimental model of ischemia reperfusion. Circulation 95:701–708. 227. De Simoni MG, Storini C, Barba M, Catapano L, Arabia AM, Rossi E and Bergamaschini L (2003) Neuroprotection by complement (C1) inhibitor in mouse transient brain ischemia. J Cereb Blood Flow Metab 23:232–239. 228. Akita N, Nakase H, Kaido T, Kanemoto Y and Sakaki T (2003) Protective effect of C1 esterase inhibitor on reperfusion injury in the rat middle cerebral artery occlusion model. Neurosurgery 52:395–400. 229. Nielsen EW, Mollnes TE, Harlan JM and Winn RK (2002) C1-inhibitor reduces the ischaemia-reperfusion injury of skeletal muscles in mice after aortic cross-clamping. Scand J Immunol 56:588–592. 230. Bauernschmitt R, Bohrer H and Hagl S (1998) Rescue therapy with C1-esterase inhibitor concentrate after emergency coronary surgery for failed PTCA. Intens Care Med 24:635–638. 231. de Zwaan C, KleineAH, Diris JH, Glatz JF, Wellens HJ, Strengers PF, Tissings M, Hack CE, van Dieijen-Visser MP, and Hermens WT (2002) Continuous 48-h C1-inhibitor treatment, following reperfusion therapy, in patients with acute myocardial infarction. Eur Heart J 23:1670–1677. 232. Weisman HF, Bartow T, Leppo MKHC, Marsh J, Carson GR, Concino MF, Boyle MP, Roux KH, Weisfeldt ML and Fearon DT (1990) Soluble human complement receptor type 1: In vivo inhibitor of complement suppressing post-ischemic myocardial inflammation and necrosis. Science 249:146–151.

ch23

Jan. 10, 2007

8:57

SPI-B429



591

233. Buerke M, Schwertz H, Seitz W, Meyer J and Darius H (2001) Novel small molecule inhibitor of C1s exerts cardioprotective effects in ischemiareperfusion injury in rabbits. J Immunol 167:5375–5380. 234. Dalmasso AP and Platt JL (1993) Prevention of complement-mediated activation of xenogeneic endothelial cells in an in vitro model of xenograft hyperacute rejection by C1 inhibitor. Transplantation 56:1171–1176. 235. Matsunami K, Miyagawa S,Yamada M,Yoshitatsu M and Shirakura R (2000) A surface-bound form of human C1 esterase inhibitor improves xenograft rejection. Transplantation 69:749–755. 236. Fukuta D, Miyagawa S, Yamada M, Matsunami K, Kurihara T, Shirasu A, Hattori H and Shirakura R (2003) Effect of various forms of the C1 esterase inhibitor (C1-INH) and DAF on complement mediated xenogeneic cell lysis. Xenotransplantation 10:132–141. 237. Fiane AE, Videm V, Johansen HT, Mellbye OJ, Nielsen EW and Mollnes TE (1999) C1-inhibitor attenuates hyperacute rejection and inhibits complement, leukocyte and platelet activation in an ex vivo pig-to-human perfusion model. Immunopharmacology 42:231–243. 238. Schelzig H, Simon F, Krischer C, Vogel A and Abendroth D (2001) Ex vivo hemoperfusion (eHPS) of pig-lungs with whole human blood: Effects of complement inhibition with a soluble C1-esterase-inhibitor. Ann Transplant 6: 34–39. 239. Przemeck M, Lorenz R, Vangerow B, Klempnauer J, Winkler M and Piepenbrock S (2002) Pretreatment with C1-esterase inhibitor improves cardiovascular stability in small primates undergoing porcine kidney xenotransplantation. Transplant Proc 34:2383. 240. Hecker JM, Lorenz R, Appiah R, Vangerow B, Loss M, Kunz R, Schmidtko J, Mengel M, Klempnauer J, Piepenbrock S, Dickneite G, Neidhardt H, Ruckoldt H and Winkler M (2002) C1-inhibitor for prophylaxis of xenograft rejection after pig to cynomolgus monkey kidney transplantation. Transplantation 73:675–677. 241. Schmaier AH (2003) The kallikrein-kinin and the renin-angiotensin systems have a multlilayered interaction. Am J Physiol Regul Integr Comp Physiol 285:R1–R13. 242. Cugno M, Hack CE, Boer JPd, Eerenberg AJ, Agostoni A and Cicardi M (1993) Generation of plasmin during acute attacks of hereditary angioedema. J Lab Clin Med 121:38–43. 243. Bos IG, de Bruin EC, Karunta YA, Modderman PW, Eldering E and Hack CE (2003) Recombinant human C1-inhibitor produced in Pichia pastoris has

ch23

Jan. 10, 2007

8:57

SPI-B429

592



the same inhibitory capacity as plasma C1-inhibitor. Biochim Biophys Acta 1648:75–83. 244. Gettins PGW (2002) Serpin structure, mechanism and function. Chem Rev 102:4751–4803. 245. Patston PA, Roodi N, Schifferli JA, Bischoff R, Courtney M and Schapira M (1990) Reactivity of alpha 1-antitrypsin mutants against proteolytic enzymes of the kallikrein-kinin, complement and fibrinolytic systems. J Biol Chem 265:10786–10791.

ch23

Jan. 10, 2007

8:57

SPI-B429


24 Neuroserpin in Neurological Disease Manuel Yepes and Daniel A. Lawrence

1. Introduction Extracellular proteolysis mediated by serine proteinases has been implicated in a number of processes in the central nervous system (CNS) during development and in response to physiological and pathological events. One of these serine proteinases is tissue-type plasminogen activator (tPA). The primary substrate for tPA in vivo is the zymogen plasminogen, which tPA activates to the broad specificity proteinase plasmin. Outside the CNS, tPA is primarily a thrombolytic enzyme, since plasmin’s principal substrate is fibrin.1 In contrast, the roles for tPA and plasmin within the CNS are not well characterized, and their primary substrates are not known. Nevertheless, there is a growing body of evidence suggesting that in the CNS, tPA plays an important role in physiological and pathological events such as learning,2,3 memory,4–6 regulation of the permeability of the neurovascular unit,7 and neurodegeneration.8 Neuroserpin (serpin PI12) is an axonally secreted member of the serpin gene family,9,10 which is primarily expressed in the brain.10 Sequence analysis of neuroserpin’s cDNA initially suggested that neuroserpin was likely to be a functional serine proteinase inhibitor with specificity for trypsin-like enzymes.11 Subsequent studies showed that neuroserpin is a fully inhibitory serpin, that reacts preferentially with tPA (Table 1), suggesting that neuroserpin is a selective inhibitor of tPA in the CNS.

593

ch24

Jan. 10, 2007

8:57

SPI-B429


594

M. Yepes & D. A. Lawrence Table 1 Kinetic constants for the interaction of neuroserpin with various proteinases. Enzyme

Human two-chain tPA Human single-chain tPA Trypsin Human high molecular weight uPA Human low molecular weight uPA NGF-γ Plasmin Thrombin

k (s−1 )

K (µM)

ki (M−1 s−1 )

0.078 0.17 0.0096 0.0050 0.013 0.0086 0.000052 0.00013

0.12 2.1 0.16 0.20 1.4 1.3 0.15 0.64

6.2 × 105 8.0 × 104 5.9 × 104 2.5 × 104 9.2 × 103 6.5 × 103 3.6 × 102 2.1 × 102

2. Genetics of Neuroserpin The human neuroserpin gene is located in chromosome 3q2612,13 and consists of nine exons and eight introns. The open reading frame of the cDNAs of human,12 chicken,9 mouse,14 and rat15 neuroserpin each encode a protein of 410 amino acids. Furthermore, the positions and phases of the exonintron borders in the mouse are completely conserved between neuroserpin, protein nexin-1 (PN-1), and plasminogen inhibitor-1 (PAI-1).13

3. Biochemical Properties of Neuroserpin A comparison between neuroserpin inhibitory activity with that of PAI-1 and PN-1 suggests that neuroserpin may have a biological role different from these other serpins. Neuroserpin reacts ∼30-fold slower with tPA than does PAI-1; however, its rate of 6.2 × 105 M−1 s−1 is ∼20-fold faster than that of PN-1. Moreover, neuroserpin’s rate of inhibition of tPA is 25-fold faster than its rate of inhibition of uPA. This is in contrast with PAI-1, which inhibits tPA and uPA at essentially the same rate, and PN-1, which reacts ∼fivefold faster with uPA than it does with tPA.10 The efficient inhibition of tPA by neuroserpin has been confirmed in vivo in transgenic mice overexpressing neuroserpin in neurons where there is a significant decrease in basal tPA activity.16 Together, these observations support the hypothesis that neuroserpin is the primary inhibitor of tPA in the CNS. Finally, in vitro

ch24

Jan. 10, 2007

8:57

SPI-B429


Neuroserpin in Neurological Disease

595

studies have demonstrated that active neuroserpin and the tPA-neuroserpin complex are internalized into the endosomal compartment in a process mediated by the low-density lipoprotein receptor-related protein (LRP). However, the fact that neuroserpin does not directly bind to LRP on its own suggests the presence of an as yet unidentified co-factor that mediates its internalization.17

4. Distribution of Neuroserpin in the Central Nervous System 4.1. Neuroserpin in the developing CNS The highest levels of neuroserpin mRNA and protein expression do not correlate with PAI-1 or PN-1, but instead appear to be found in sites where either tPA message or tPA protein has been detected.14 In the developing mouse, embryo neuroserpin mRNA is detected on day 13 and reaches maximal levels shortly before birth in the neuronal precursors of most of the CNS regions immediately after becoming postmitotic and migrating from the ventricular zones.14 Likewise, neuroserpin protein has been detected in neurons that have settled in the cortical plate and extended axons that eventually will establish synaptic contacts.14

4.2. Neuroserpin in the adult CNS In the adult mouse, neuroserpin protein is present in three molecular weight forms: 47, 57 and 105 kDa, which are thought to correspond to the cleaved, native and proteinase-complex forms, respectively (Fig. 1).10 Furthermore, neuroserpin antigen is found predominantly in neurons in the neocortex, the hippocampus, the olfactory bulb, the basolateral amygdala, the Purkinje cells of the cerebellum, the subthalamic and medial mamillary hypothalamic nuclei, and in the myelinated axons of the commissura.10 In the adult human, neuroserpin mRNA is predominantly expressed in the CNS with low levels found in pancreas and testes (Fig. 2),9,10 and staining for neuroserpin protein has been detected in neurons in the neocortex, the midbrain, the dorsal and ventral horns of the spinal cord, the hippocampus, the ependymal cells lining the ventricles, the choroids plexus and the

ch24

Jan. 10, 2007

8:57

SPI-B429

596


M. Yepes & D. A. Lawrence

Fig. 1. Neuroserpin immunoblot analysis of normal murine brain extracts. Lanes 1 and 2 show a Coomassie stained 10% SDS-PAGE, and lanes 3–6 show immunoblots developed with chemiluminescence. The primary antibody in lanes 3 and 4 was rabbit antineuroserpin, and in lanes 5 and 6 was normal rabbit IgG. Lanes 1, 3, and 5 were loaded with 1 µl (18 µg) of brain extract, lane 2 was loaded with 500 ng of purified neuroserpin, and lanes 4 and 6 were loaded with 2 ng of purified neuroserpin. Reprinted from Journal of Biological Chemistry (1997; 272:33062–33067).

hypophysis.18 Neuroserpin has also been found in granules of cells in the anterior, intermediate and posterior lobes in the rat19 and human18 pituitary gland, as well as in the medullary cells of the adrenal gland in the rat,15,20,21 suggesting that this serpin may also play a role in the regulation of the secretory vesicle function.21

5. Neuroserpin in Normal CNS Function 5.1. Synaptic plasticity Neuronal plasticity is associated with critical physiologic and pathologic processes in the developing and mature CNS. One of the most important

ch24

Jan. 10, 2007

8:57

SPI-B429



597

Fig. 2. Northern blot analysis of RNA from various adult human tissues. The top panel shows the neuroserpin message, and the lower panel shows actin as a control of RNA loading. Each tissue is indicated above the lane. Sm Int (lane 6), small intestine; PBL (lane 8), peripheral blood lymphocytes. Reprinted from Journal of Biological Chemistry (1997; 272:33062–33067).

characteristics of this phenomenon is an activity-dependent remodeling of neuronal connectivity and synaptic transmission. It is known that plasminogen activators are axonally produced and secreted by neurons22–24 and that their occurrence during development and regeneration is important in supporting the advancement of the growth cone through the tissue during the period of axonal growth.25–27 Accordingly, it has been demonstrated that the gene for tPA is induced by neuronal activity2,3,6 and that tPA is released in association with morphological differentiation.22,23 A role for neuroserpin in synaptic plasticity has been suggested by the in vitro observation of an activity-dependent upregulation of neuroserpin in the hippocampus,28 by the description of changes in the number of neurites of Att-2015 and PC1229 cells in response to variations in neuroserpin expression, and by the fact that neuroserpin is released in response to neuronal depolarization.15,28 These observations have been supported in vivo by the finding that neuroserpin is primarily expressed in the developing CNS14 and by experiments demonstrating that neuroserpin plays a critical role in the development of the primary visual cortex.30

ch24

Jan. 10, 2007

8:57

SPI-B429


598


5.2. Behavior A role for tPA in behavior has been suggested by the observation that tPA−/− mice are impaired in their ability to respond to a negative stimulus5 and to stress-inducing situations.31 The importance of neuroserpin in behavior is supported by experiments demonstrating a decrease in locomotor activity in novel environments and in the response to anxiety-like situations in mice deficient in neuroserpin (Ns−/− ). However, it should be kept in mind that tPA activity is unchanged in Ns−/− mice,32 suggesting that the role of neuroserpin in behavior may be independent of its effects on tPA activity.

6. Neuroserpin in CNS Disease Neuroserpin has been implicated in the development of a number of pathologic events such as cerebrovascular disease,7,16,33–37 epilepsy,38,39 and dementia.40–43 6.1. Cerebrovascular disease Cerebral infarction (stroke) is a leading cause of disability and the second most common cause of death in the world after heart disease.44–46 It is caused by an occlusion of a cerebral blood vessel, which if not quickly resolved, leads to a reduction in blood flow and irreversible ischemic brain injury. Ischemia triggers a cascade of pathophysiological events, including the activation of components of the plasminogen system,33,47 with resultant development of cerebral edema7 and cell death.33,48,49 6.2. TPA and neuroserpin in cerebral ischemia Early after the onset of experimental cerebral ischemia, there is an increase in tPA activity around the blood vessels7 and in microglial cells16 in the area of ischemic penumbra, followed by a transient rise in neuroserpin expression in neurons located in the same area16,33 (Fig. 3). The subsequent decrease in neuroserpin expression is then followed by cell death and the expansion of the necrotic core.36,37

ch24

Jan. 10, 2007

8:57

SPI-B429



599

Fig. 3. Immunohistochemical staining of neuroserpin in brain 48 h after reperfusion. Panel A shows the area of penumbra ipsilateral to the stroke and, panel B shows a similar area of the cortex contralateral to the stroke (magnification 100×). This research was originally published in Blood. Yepes M, Sandkvist M, Wong MK, Coleman TA, Smith E, Cohan SL, Lawrence DA. Neuroserpin reduces cerebral infarct volume and protects neurons from ischemia-induced apoptosis. Blood 2000; 96:569–576. © The American Society of Hematology.

Clinically, tPA has been approved for the treatment of patients with acute ischemic stroke.50 However, in contrast to the clinical setting, results obtained from animal studies have shown that tPA may also have an adverse effect on cerebral ischemia. In these reports, tPA deficient animals when compared with wild-type animals were shown to have an approximately 50% decrease in infarct size and a 41% increase in neuronal survival following middle cerebral artery occlusion (MCAO). Moreover, administration of tPA at doses similar to those used for the treatment of acute stroke in humans resulted in a 33% increase in stroke volume in wild-type animals.47 In contrast, other studies have shown that tPA is beneficial in cerebral ischemia.51,52 This apparent disparity between different studies may be explained at least in part by the characteristics of the animal models used in each case. Accordingly, in models that involve the formation of a clot in the middle cerebral artery,52,53 or in which tPA enhances reperfusion,52,53 tPA is beneficial since it restores blood flow, and a lack of tPA exacerbates the damage. For example, one study of transient cerebral ischemia-reperfusion showed that tPA deficiency increases cerebrovascular fibrin deposition and that this is associated with increased brain injury.51 This study suggested that endogenous tPA may protect the brain during cerebral ischemia, presumably through its local thrombolytic action. In contrast, most of the

ch24

Jan. 10, 2007

8:57

SPI-B429

600



stroke models in which a blood clot does not form, or in the case of permanent MCAO, the need for the thrombolytic activity of tPA is obviated. In these cases, the lack of tPA actually protects against damage, implying that tPA can also be harmful in stroke.47,54 These results suggest a dual role for tPA during cerebral ischemia; although it plays an important role in thrombolysis after stroke, its presence in the extravascular (intracerebral) space following cerebral ischemia might be deleterious. In contrast to tPA, the genetic deficiency of urokinase-type plasminogen activator (uPA) has not been shown to have any effect on either the volume of the ischemic lesion54 or the disruption of the blood brain barrier following the onset of cerebral ischemia.7

6.3. Neuroprotective role of neuroserpin in cerebral ischemia A protective effect of neuroserpin in cerebral ischemia was demonstrated by the observation that either treatment with neuroserpin injected directly into the ischemic cortex immediately after the ischemic insult (Fig. 4)33 or the overexpression of the neuroserpin gene,16 results in a significant decrease in the volume of the ischemic area following MCAO. After the early occlusion of a cerebral blood vessel, there is the formation of two ischemic areas: the necrotic core which is a densely ischemic zone where neurons are irreversibly damaged, and the area of “ischemic penumbra” where cerebral blood flow is sufficiently decreased to abolish electrical potentials, yet sufficient to allow maintenance of membrane potentials and cellular ionic homeostasis.55–57 With time, this potentially salvageable area of penumbra tends to become infarcted, with expansion of the necrotic core. Among the most important events that occur in the area of the penumbra are the excitotoxic release of glutamate,58 microglial cell activation,16,59 apoptotic cell death,33 and the increase in tPA activity.7,16,33,36 The importance of neuroserpin in the preservation and protection of this potentially salvageable ischemic area was initially demonstrated by the observation that treatment with neuroserpin following MCAO, results in a considerable decrease in the tPA activity in the area of ischemic

ch24

Jan. 10, 2007

8:57

SPI-B429



601

Fig. 4. Quantitative analysis of infarct volume 72 h after reperfusion. Quantitation of the stroke volume was performed as described in “Materials and methods.” PBS indicates animals injected with PBS (n = 8); Ns, animals injected with neuroserpin (n = 8); Cl-Ns, animals injected with elastase-cleaved inactive neuroserpin (n = 2). P values (p < 0.01) relative to the PBS-treated animals are shown, and errors represent SEM. This research was originally published in Blood. Yepes M, Sandkvist M, Wong MK, Coleman TA, Smith E, Cohan SL, Lawrence DA. Neuroserpin reduces cerebral infarct volume and protects neurons from ischemia-induced apoptosis. Blood 2000; 96:569–576. © The American Society of Hematology.

penumbra.33 Later histopathological and neuroradiological studies showed significant cell survival in the ischemic area of the brain treated with neuroserpin.36 However, although neuroserpin treatment is associated with cell survival, treatment with neuroserpin does not improve the cerebral blood flow following MCAO.36 These observations suggest that the main role of neuroserpin following cerebral ischemia is neuroprotection through its ability to inhibit the rise in ischemia-induced intracerebral tPA activity. This hypothesis was supported by the observation that either treatment with neuroserpin or overexpression of the neuroserpin gene results not only in a considerable attenuation of tPA activity in the area of ischemic penumbra, but also in a significant decrease in both the number of cells with apoptotic features (Fig. 5)33 and in the extent of microglial activation.16

ch24

Jan. 10, 2007

8:57

SPI-B429

602



Fig. 5. Neuronal apoptosis within the ischemic penumbra. TUNEL staining ipsilateral of the infarct in PBS-treated (panels A and B) and neuroserpin-treated (panel C) animals. In panel A, NC indicates the necrotic core and P indicates the area of the penumbra. Panels B and C are high-magnification images of the penumbra in control (panel B) and neuroserpintreated animals (panel C). Examples of cells considered to be apoptotic for the purposes of quantification are indicated with the open arrows; cells considered as necrotic are indicated with the closed arrows. Magnification in panel A is 40× and in panels B and C 400×. (D) Quantitative analysis of apoptosis in the area of penumbra 72 h after reperfusion. For quantitation only cells with apoptotic bodies present (panel B) were counted. Control represents animals injected with PBS (n = 6). Neuroserpin indicates animals injected with neuroserpin (n = 6). P-values < 0.05 are shown, and errors represent SEM. (E) Quantitation of apoptosis was performed as in panel D at the times indicated, (◦) PBS-treated animals, () neuroserpin-treated animals. This research was originally published in Blood. Yepes M, Sandkvist M, Wong MK, Coleman TA, Smith E, Cohan SL, Lawrence DA. Neuroserpin reduces cerebral infarct volume and protects neurons from ischemia-induced apoptosis. Blood 2000; 96:569–576. © The American Society of Hematology.

6.4. Effects of neuroserpin on ischemia-induced increases in blood brain barrier permeability Increases in blood brain barrier permeability and cerebral edema are leading causes of death in patients with acute ischemic stroke.60 Laminin is one of the main non-cellular components of the neurovascular unit. Previous studies have demonstrated that excitotoxin-induced laminin degradation is

ch24

Jan. 10, 2007

8:57

SPI-B429



603

mediated by tPA61,62 and that cerebral ischemia induces changes in the architecture of the basement membrane with exposure of laminin epitopes, in a process inhibited by treatment with neuroserpin.33 These observations suggested a role for tPA in cerebral ischemia-induced increases in blood brain barrier permeability. Subsequent in vivo experiments demonstrated that as early as one hour after the onset of cerebral ischemia, there is a rise in endogenous tPA activity around the blood vessels in the area of ischemic penumbra (Fig. 6), followed a few hours later by an increase in the permeability of the neurovascular unit in the same area.7 Likewise, genetic deficiency of tPA results in considerable attenuation of ischemia-induced increases in blood brain barrier permeability.7 The importance of neuroserpin in this process was demonstrated by the finding that the injection of neuroserpin into the cerebrospinal

Fig. 6. Temporal and spatial relationship between tPA activity and vascular permeability following MCAO in WT (C57BL/6J) mice. In A–D, tPA activity 1 h after MCAO is shown in red by in situ zymography, and cell nuclei are in blue (DAPI). (A) By 1 h after MCAO, there is significant tPA activity in the vessel wall and in the perivascular tissue surrounding a vessel bordering the necrotic area. (B and C) The same vessel in adjacent sections (5 µm), but with anti-tPA antibodies included (B), or without the addition of Plg (C) in the overlay. (D) The background tPA activity associated with a vessel in a corresponding area in the contralateral hemisphere from the same section shown in (A). In A–D, the original magnification was ×100. (E–H) Evans blue extravasation is shown in red and cell nuclei in blue (DAPI) 6 h after MCAO. (E) A low-magnification view of the entire ischemic area. (F) Evans blue extravasation from a vessel located in the area adjacent to the ischemic area, similar to the one seen in (A). (G) Electronic magnification of the box in F. The arrow indicates an area of Evans blue leakage outside the internal elastic lamina of the vessel. (H) Evans blue is shown adhering to the vessel wall, but no extravasation is seen in a vessel from the same section seen in F and G but located in the corresponding region of the contralateral hemisphere. Ipsi, ipsilateral; Contra, contralateral. Reprinted from J. Clin. Invest. 112: 1533–1540 (2003).

ch24

Jan. 10, 2007

8:57

SPI-B429


604


fluid (CSF) following MCAO results in the attenuation of ischemia-induced increases in blood brain barrier permeability in wild-type mice.7 A role for intravascular tPA, given as a thrombolytic agent, in the development of cerebral edema has also been documented. Treatment with tPA following embolic occlusion of the middle cerebral artery results in a significant increase in the permeability of the blood brain barrier,36,63 which is attenuated when neuroserpin is injected into the CSF before thrombolysis.36 These data allow us to postulate the following working model: after the onset of cerebral ischemia, there is an increase in tPA activity mainly in perivascular glial cells16 in the area of ischemic penumbra. In reaction to this increase in tPA activity and in an attempt to regulate its deleterious effects, neuroserpin is released from neurons. However, when the ischemic insult is too prolonged or too intense, the activity of tPA overcomes the inhibitory effect of neuroserpin and then through a process that involves LRP in the neurovascular unit,7,64 the free tPA induces an increase in blood brain barrier permeability. This leads to the passage of plasma and potentially harmful substances from the intravascular space into the brain, with resultant cell death and expansion of the necrotic core. Treatment with neuroserpin results in a more efficient inhibition of tPA released in response to the ischemic insult, which results in the protection of the blood brain barrier from the effects of tPA, which in turn enhances neuronal survival. 6.5. Seizures A seizure is characterized by the development of neuronal hyperexcitability (the tendency of a neuron to discharge repetitively) and hypersynchrony (the property of a population of neurons to discharge together). One of the most remarkable characteristics of synaptic plasticity is the activity-dependent remodeling of synaptic efficiency and neuronal connectivity, e.g. seizures. This sets in motion a cascade of events, including establishment of new synaptic contacts65 and changes in gene expression.6

6.6. Role of tPA and neuroserpin in seizures Early studies using in situ hybridization showed that one hour after the induction of seizures, there is an upregulation of the tPA gene throughout

ch24

Jan. 10, 2007

8:57

SPI-B429



605

the brain.6 Subsequent experiments with an in situ zymographic technique demonstrated that as early as 10 min after the induction of an electrographic seizure there is an increase in tPA activity followed by a rise in the neuroserpin expression in the same areas38 (Fig. 7). This rise in tPA activity precedes the spreading of the abnormal electrical activity characteristic of a seizure throughout the brain. Moreover, treatment with neuroserpin results not only in a significant attenuation of seizure-induced increases in tPA activity, but also in the blockade of the generation of the abnormal electrical discharges.38 This effect of tPA and neuroserpin is seen not only on the genesis and propagation of aberrant electrical activity throughout the brain (seizures), but also on the clinical manifestations of this event (convulsions). Initial studies showed that tPA−/− mice are resistant to the induction of convulsions.8 Subsequent experiments demonstrated that tPA mediates not only the onset of a convulsion, but also the progression of its severity.38 Furthermore, treatment with neuroserpin results not only in the suppression of the abnormal electrical activity (seizures) but also in a significant slowing in the onset and worsening of its clinical manifestation (convulsions).38 Together, these results demonstrate that tPA and neuroserpin play an important role in the regulation of cell-cell interactions in the CNS.

6.7. Neuroprotective effect of neuroserpin on seizures-induced cell death One of the most important consequences of a prolonged seizure is cell death in the hippocampal CA3 layer.66 Initial studies showed that the injection of kainic acid (KA) directly into the hippocampus induces cell death in the CA1 and CA3 layers. The same studies demonstrated that KA-induced cell death is significantly attenuated by the genetic deficiency of tPA (tPA−/− mice).8 However, it is known that the injection of KA directly into the hippocampus induces not only cell death in the CA3 layer, but also long-lasting seizures.67 Therefore, to differentiate between KA- and seizure-induced cell death, and to study the role of tPA and neuroserpin on these events, seizures were induced in wild-type and tPA−/− mice by the injection of KA into the amygdala, a structure distal to the hippocampus, followed by

ch24

Jan. 10, 2007

8:57

SPI-B429

606



ch24

Jan. 10, 2007

8:57

SPI-B429



607

the injection of neuroserpin directly into the hippocampus. Not only was it observed that the generation of seizures induces cell death mainly in the CA3 layer, but also that either the genetic deficiency of tPA or the treatment with neuroserpin resulted in a significant decrease in seizuresinduced cell death.38 Taken together, these results allow us to propose the following working model for tPA and neuroserpin in the pathogenesis of a seizure: the induction of a seizure induces an increase in local tPA activity. This increase in tPA activity is important for the establishment of the structural changes at the synaptic level needed for the spreading to contiguous cellular structures of the abnormal electrical activity characteristic of a seizure. In response to this increase in tPA activity, there is a rise in neuroserpin expression intended to control the development of the hypersynchrony needed for a seizure to involve neighboring structures. However, when the stimulus is too persistent or too intense, as proposed above for cerebral ischemia, the activity of tPA overcomes the inhibitory action of neuroserpin followed by the spreading of the seizure throughout the brain. Once the abnormal electrical activity reaches the hippocampus, neuroserpin expression increases in an attempt to protect the cells from the neurotoxic effects of tPA. However, in the absence of exogenous neuroserpin, this inhibitory activity is overcome again with ensuing hippocampal cell death.

Fig. 7. Kinetic analysis of tPA activity and neuroserpin antigen in the limbic system during seizure. Seizures were induced by kianic acid injection into the amygdala and the animals were euthanized at 10, 30, and 60 min after injection. Cryostat sections were stained with DAPI to visualize nuclei and then analyzed either for in situ tPA activity by plasminogencasein zymography with internally quenched fluorescent casein (red) or for neuroserpin antigen by immunofluorescence (green). The top panel of each part shows tPA activity, and the bottom panel of each part shows neuroserpin antigen (Ns). (a–c) Ten minutes after kianic acid injection; (d–f) 30 min after injection; (g–i) 60 min after injection. ( j–l) Top panels: 60 min after kianic acid injection, with anti-tPA antibodies (tPA) in the overlay; bottom panels: normal rat brain with no kianic acid injection. a, d, g, and j show the basolateral amygdala ipsilateral (Ipsi) to the injection site; b, e, h, and k show the ipsilateral hippocampus; c, f, i, and l show the contralateral (Contra) hippocampus. Original magnification 100×. Reprinted from J. Clin. Invest. 109: 1571–1578 (2002).

ch24

Jan. 10, 2007

8:57

SPI-B429

608



7. Familial Encephalopathy with Neuroserpin Inclusion Bodies (FENIB) Neurodegenerative diseases are associated with a wide variety of molecular mechanisms that lead to neuronal cell death. One of these diseases, dementia, is characterized by the development of a progressive degenerative brain process that leads to distinct patterns of behavioral and cognitive decline. Mutations of the human neuroserpin gene have been linked to a dominantly inherited form of familial encephalopathy and epilepsy characterized by the presence of intraneuronal neuroserpin inclusion bodies.42,43

7.1. Neuropathology The most important neuropathological feature of this disease is the presence of eosinophilic, round (5–50 µm in diameter), periodic acid/Schift reagent (PAS)-positive, and diastase-resistant intraneuronal inclusion bodies (Collins bodies) that are composed primarily of polymerized neuroserpin.42 These inclusion bodies are found throughout the brain mainly in the cerebral cortex (layers III–V), cingulated gyrus, amygdala, substantia nigra (pars compacta), red nucleus and pons.41,42 The inclusion bodies of FENIB are intraneuronal and differ from those found in other causes of dementia, such as Parkinson’s and Huntington’s disease, in that in those cases, the mutant protein forms aggregates within the cytoplasm.68

7.2. Genetics and pathophysiology of FENIB The shutter region is an area of the serpin structure that controls the opening of the β-sheet A during the inhibition reaction. This opening is required so that a segment of the serpin reactive center loop can be inserted to become the central strand of this β-sheet. This dramatic conformational rearrangement provides the force necessary to distort and thus trap the serpin target protease.69 To date, all the mutations identified in neuroserpin that result in FENIB are clustered in this shutter region.41 Molecular analysis of the neuroserpin gene allowed the initial identification of single point mutations in each of the two unrelated families (ser49-to-pro and ser52-to-arg).43 Since then, three additional families have been described with single point

ch24

Jan. 10, 2007

8:57

SPI-B429



609

mutations, two of which were novel (gly392-to-glu and his338-to-arg).41 The critical importance of the shutter region in the development of FENIB was confirmed by these two novel mutations, since their location more than 300 amino acids away in the linear sequence of neuroserpin from the residues first identified, they were in fact located in very close proximity to the original mutation sites in the shutter region of the neuroserpin tertiary structure. During the normal serpin inhibition reaction, the shutter region is thought to prevent the opening of β-sheet A until the reactive center loop of the serpin has been cleaved by its target protease. Upon cleavage of the serpin, sheet A opens and the reactive center loop moves into position as the central strand. However, the effect of mutations in the shutter region of neuroserpin allows the opening of β-sheet A prior to interaction with a protease. This then permits the intact reactive center loop of one serpin molecule to be inserted into the β-sheet A of a second molecule, a process that leads to the formation of polymers, which accumulate within the cells.70 This process also results in the reduction of the inhibitory activity of neuroserpin, as has been demonstrated in the ser49-to-pro mutations.71 The mechanism of this disease is very similar to the serpinopathy seen in patients with mutations in the serpin, α1 -antitrypsin, where similar polymers are found in the hepatocytes (the site of α1 -antitrypsin synthesis). This leads to the development of severe liver cirrhosis in severely affected individuals due to the formation of α1 -antitrypsin aggregates directly into the hepatocytes. In less severe cases, the patients present with emphysema, which is caused by the loss of α1 -antitrypsin activity in the peripheral tissues leading to the excessive activity of neutrophil elastase, which produces emphysema when upregulated in the lung.72,73 7.3. Clinical manifestations It has been postulated that the intracellular precipitation of abnormal neuroserpin and the formation of Collins bodies underlies the clinical manifestations of FENIB with a close association between the degree of conformational instability induced by the mutation and severity of the disease as revealed by the age of onset. The closure of the strands 3 and 5 of β-sheet A of neuroserpin is needed to prevent the polymerization, and this closure is strongly dependent on the interaction between those strands with

ch24

Jan. 10, 2007

8:57

SPI-B429


610


the underlying amino acid side chains of the shutter region. Two highly conserved residues that are thought to be particularly important in this process are histidine 338 and glycine 392. The glycine at position 392 underlies sheet A and permits the close packing of overlying hydrophobic residues from sheet A that is essential for the stability of neuroserpin. Thus, the patient with replacement of glycine at position 392 with a charged glutamic acid residue manifests with the most severe phenotype.41,74 The histidine 338 residue is found in strand 5 of the sheet A and it forms a network of hydrogen bonds that stabilize the closed sheet by interlinking strand 5 with strand 3. This histidine also directly interacts with serines 49 and 52 which underlie sheet A. The loss of these bonds, due to a mutation of histidine to arginine, allows the sheet to open with the resultant formation of polymers, and patents with this mutation also show an early onset of symptoms. Accordingly, individuals with the gly392-to-glu mutation have 9.5 times more inclusion bodies within the cerebral cortex than patients with ser49-to-pro mutation. This results in an age of onset of symptoms in individuals with the ser49-to-pro, ser52-to-arg His 338-to-Arg and gly392-toglu of 48, 27, 15, and 13 years, respectively. The clinical manifestations also change according to the mutation point: the most important manifestation is dementia for patients with the ser49-to-pro mutation, whereas in patients with ser52-to-arg, his338-to-arg and gly392-to-glu mutations, dementia and severe progressive myoclonus epilepsy are the most important clinical characteristics.41 The most common cognitive manifestations of FENIB are associated with frontal lobe deficits.40,41 These patients initially present restriction in attention, concentration and oral fluency, most likely reflecting the accumulation of intraneuronal Collins bodies in the frontal lobes and its connections. In the course of the disease, there is global brain dysfunction with severe visuospatial disorientation.40 One important clinical difference of FENIB from the dementia of the Alzheimer type is that in FENIB, there is relative sparing of recall memory and a lack of significant word-finding difficulties, both of which are considerably involved in the early stages of Alzheimer’s disease.40,41

Conclusions FENIB is the second sepinopathy to be described in humans. While the mechanism of polymer formation is similar to that of α1 -antitrypsin Z

ch24

Jan. 10, 2007

8:57

SPI-B429



611

mutations, the clinical manifestations are quite distinct and reflect neuroserpin’s expression predominantly within the CNS.

References 1. Bugge TH, Xiao Q, Kombrinck KW, Flick MJ, Holmback K, Danton MJ, Colbert MC, Witte DP, Fujikawa K, Davie EW and Degen JL (1996) Fatal embryonic bleeding events in mice lacking tissue factor, the cell-associated initiator of blood coagulation. Proc Natl Acad Sci USA 93:6258–6263. 2. Seeds NW, Williams BL and Bickford PC (1995) Tissue plasminogen activator induction in Purkinje neurons after cerebellar motor learning. Science 270:1992–1994. 3. Seeds NW, Basham ME and Ferguson JE (2003) Absence of tissue plasminogen activator gene or activity impairs mouse cerebellar motor learning. J Neurosci 23:7368–7375. 4. Baranes D, Lederfein D, Huang YY, Chen M, Bailey CH and Kandel ER (1998) Tissue plasminogen activator contributes to the late phase of LTP and to synaptic growth in the hippocampal mossy fiber pathway. Neuron 21: 813–825. 5. Huang YY, Bach ME, Lipp HP, Zhuo M, Wolfer DP, Hawkins RD, Schoonjans L, Kandel ER, Godfraind JM, Mulligan R, Collen D and Carmeliet P (1996) Mice lacking the gene encoding tissue-type plasminogen activator show a selective interference with late-phase long-term potentiation in both Schaffer collateral and mossy fiber pathways. Proc Natl Acad Sci USA 93:8699–8704. 6. Qian Z, Gilbert ME, Colicos MA, Kandel ER and Kuhl D (1993) Tissueplasminogen activator is induced as an immediate-early gene during seizure, kindling and long-term potentiation. Nature 361:453–457. 7. Yepes M, Sandkvist M, Moore EG, Bugge TH, Strickland DK and Lawrence DA (2003) Tissue-type plasminogen activator induces opening of the bloodbrain barrier via the LDL receptor-related protein. J Clin Invest 112: 1533–1540. 8. Tsirka SE, Gualandris A, Amaral DG and Strickland S (1995) Excitotoxininduced neuronal degeneration and seizure are mediated by tissue plasminogen activator. Nature 377:340–344. 9. Osterwalder T, Contartese J, Stoeckli ET, Kuhn TB and Sonderegger P (1996) Neuroserpin, an axonally secreted serine protease inhibitor. EMBO J 15: 2944–2953.

ch24

Jan. 10, 2007

8:57

SPI-B429

612



10. Hastings GA, Coleman TA, Haudenschild CC, Stefansson S, Smith EP, Barthlow R, Cherry S, Sandkvist M and Lawrence DA (1997) Neuroserpin, a brain-associated inhibitor of tissue plasminogen activator is localized primarily in neurons. Implications for the regulation of motor learning and neuronal survival. J Biol Chem 272:33062–33067. 11. Osterwalder T, Cinelli P, Baici A, Pennella A, Krueger SR, Schrimpf SP, Meins M and Sonderegger P (1998) The axonally secreted serine proteinase inhibitor, neuroserpin, inhibits plasminogen activators and plasmin but not thrombin. J Biol Chem 273:2312–2321. 12. Schrimpf SP, Bleiker AJ, Brecevic L, Kozlov SV, Berger P, Osterwalder T, Krueger SR, Schinzel A and Sonderegger P (1997) Human neuroserpin (PI12): cDNA cloning and chromosomal localization to 3q26. Genomics 40:55–62. 13. Berger P, Kozlov SV, Krueger SR and Sonderegger P (1998) Structure of the mouse gene for the serine protease inhibitor neuroserpin (PI12). Gene 214: 25–33. 14. Krueger SR, Ghisu GP, Cinelli P, Gschwend TP, Osterwalder T, Wolfer DP and Sonderegger P (1997) Expression of neuroserpin, an inhibitor of tissue plasminogen activator, in the developing and adult nervous system of the mouse. J Neurosci 17:8984–8996. 15. Hill RM, Parmar PK, Coates LC, Mezey E, Pearson JF and Birch NP (2000) Neuroserpin is expressed in the pituitary and adrenal glands and induces the extension of neurite-like processes in Att-20 cells. Biochem J 345(Pt 3): 595–601. 16. Cinelli P, Madani R, Tsuzuki N, Vallet P, Arras M, Zhao CN, Osterwalder T, Rulicke T and Sonderegger P (2001) Neuroserpin, a neuroprotective factor in focal ischemic stroke. Mol Cell Neurosci 18:443–457. 17. Makarova A, Mikhailenko I, Bugge TH, List K, Lawrence DA and Strickland DK (2003) The LDL receptor-related protein modulates protease activity in the brain by mediating the cellular internalization of both neuroserpin and neuroserpin: tPA complexes. J Biol Chem. 18. Teesalu T, Kulla A, Simisker A, Siren V, Lawrence DA, Asser T and Vaheri A (2004) Tissue plasminogen activator and neuroserpin are widely expressed in the human central nervous system. Thromb Haemost 92:358–368. 19. Fell B, Smith AM, Hill RM, Parmar PK, Coates LC, Mezey E and Birch NP (2002) Characterisation of two serine protease inhibitors expressed in the pituitary gland. Arch Physiol Biochem 110:26–33. 20. Hook V and Metz-Boutigue MH (2002) Protein trafficking to chromaffin granules and proteolytic processing within regulated secretory vesicles of neuroendocrine chromaffin cells. Ann N Y Acad Sci 971:397–405.

ch24

Jan. 10, 2007

8:57

SPI-B429



613

21. Hill RM, Coates LC, Parmar PK, Mezey E, Pearson JF and Birch NP (2002) Expression and functional characterization of the serine protease inhibitor neuroserpin in endocrine cells. Ann N Y Acad Sci 971:406–415. 22. Krystosek A and Seeds NW (1981) Plasminogen activator release at the neuronal growth cone. Science 213:1532–1534. 23. Krystosek A and Seeds NW (1981) Plasminogen activator secretion by granule neurons in cultures of developing cerebellum. Proc Natl Acad Sci USA 78:7810–7814. 24. Krystosek A and Seeds NW (1986) Normal and malignant cells, including neurons, deposit plasminogen activator on the growth substrata. Exp Cell Res 166:31–46. 25. Dent MA, Sumi Y, Morris RJ and Seeley PJ (1993) Urokinase-type plasminogen activator expression by neurons and oligodendrocytes during process outgrowth in developing rat brain. Eur J Neurosci 5:633–647. 26. Sumi Y, Dent MA, Owen DE, Seeley PJ and Morris RJ (1992) The expression of tissue and urokinase-type plasminogen activators in neural development suggests different modes of proteolytic involvement in neuronal growth. Development 116:625–637. 27. Hantai D, Rao JS, Kahler C and Festoff BW (1989) Decrease in plasminogen activator correlates with synapse elimination during neonatal development of mouse skeletal muscle. Proc Natl Acad Sci USA 86:362–366. 28. Berger P, Kozlov SV, Cinelli P, Kruger SR, Vogt L and Sonderegger P (1999) Neuronal depolarization enhances the transcription of the neuronal serine protease inhibitor neuroserpin. Mol Cell Neurosci 14:455–467. 29. Parmar PK, Coates LC, Pearson JF, Hill RM and Birch NP (2002) Neuroserpin regulates neurite outgrowth in nerve growth factor-treated PC12 cells. J Neurochem 82:1406–1415. 30. Wannier-Morino P, Rager G, Sonderegger P and Grabs D (2003) Expression of neuroserpin in the visual cortex of the mouse during the developmental critical period. Eur J Neurosci 17:1853–1860. 31. Pawlak R, Magarinos AM, Melchor J, McEwen B and Strickland S (2003) Tissue plasminogen activator in the amygdala is critical for stress-induced anxiety-like behavior. Nat Neurosci 6:168–174. 32. Madani R, Kozlov S, Akhmedov A, Cinelli P, Kinter J, Lipp HP, Sonderegger P and Wolfer DP (2003) Impaired explorative behavior and neophobia in genetically modified mice lacking or overexpressing the extracellular serine protease inhibitor neuroserpin. Mol Cell Neurosci 23:473–494. 33. Yepes M, Sandkvist M, Wong MK, Coleman TA, Smith E, Cohan SL and Lawrence DA (2000) Neuroserpin reduces cerebral infarct volume and protects neurons from ischemia-induced apoptosis. Blood 96:569–576.

ch24

Jan. 10, 2007

8:57

SPI-B429

614



34. Yepes M and Lawrence DA (2004) Neuroserpin: a selective inhibitor of tissuetype plasminogen activator in the central nervous system. Thromb Haemost 91:457–464. 35. Yepes M and Lawrence DA (2004) Tissue-type plasminogen activator and neuroserpin: a well-balanced act in the nervous system? Trends Cardiovasc Med 14:173–180. 36. Zhang Z, Zhang L,Yepes M, Jiang Q, Li Q, Arniego P, Coleman TA, Lawrence DA and Chopp M (2002) Adjuvant treatment with neuroserpin increases the therapeutic window for tissue-type plasminogen activator administration in a rat model of embolic stroke. Circulation 106:740–745. 37. Liang W, Chuan-Zhen L, Qiang D, Jian Q, Hui-Min R and Bao-Guo X (2004) Reductions in mRNA of the neuroprotective agent, neuroserpin, after cerebral ischemia/reperfusion in diabetic rats. Brain Res 1015:175–180. 38. Yepes M, Sandkvist M, Coleman TA, Moore E, Wu JY, Mitola D, Bugge TH and Lawrence DA (2002) Regulation of seizure spreading by neuroserpin and tissue-type plasminogen activator is plasminogen-independent. J Clin Invest 109:1571–1578. 39. Belorgey D, Sharp LK, Crowther DC, Onda M, Johansson J and Lomas DA (2004) Neuroserpin Portland (Ser52Arg) is trapped as an inactive intermediate that rapidly forms polymers: Implications for the epilepsy seen in the dementia FENIB. Eur J Biochem 271:3360–3367. 40. Bradshaw CB, Davis RL, Shrimpton AE, Holohan PD, Rea CB, Fieglin D, Kent P and Collins GH (2001) Cognitive deficits associated with a recently reported familial neurodegenerative disease: Familial encephalopathy with neuroserpin inclusion bodies. Arch Neurol 58:1429–1434. 41. Davis RL, Shrimpton AE, Carrell RW, Lomas DA, Gerhard L, Baumann B, Lawrence DA, Yepes M, Kim TS, Ghetti B, Piccardo P, Takao M, Lacbawan F, Muenke M, Sifers RN, Bradshaw CB, Kent PF, Collins GH, Larocca D. and Holohan PD (2002) Association between conformational mutations in neuroserpin and onset and severity of dementia. Lancet 359:2242–2247. 42. Davis RL, Holohan PD, Shrimpton AE, Tatum AH, Daucher J, Collins GH, Todd R, Bradshaw C, Kent P, Feiglin D, Rosenbaum A, Yerby MS, Shaw CM, Lacbawan F and Lawrence DA (1999) Familial encephalopathy with neuroserpin inclusion bodies. Am J Pathol 155:1901–1913. 43. Davis RL, Shrimpton AE, Holohan PD, Bradshaw C, Feiglin D, Collins GH, Sonderegger P, Kinter J, Becker LM, Lacbawan F, Krasnewich D, Muenke M, Lawrence DA, Yerby MS, Shaw CM, Gooptu B, Elliott PR, Finch JT, Carrell RW and Lomas DA (1999) Familial dementia caused by polymerization of mutant neuroserpin. Nature 401:376–379.

ch24

Jan. 10, 2007

8:57

SPI-B429



615

44. The World Health Organization (1999) The World Health Report, pp. 1–121. Geneva: Switzerland. 45. Thorvaldsen P, Asplund K, Kuulasmaa K, Rajakangas AM and Schroll M (1995) Stroke incidence, case fatality, and mortality in the WHO MONICA project. World health organization monitoring trends and determinants in cardiovascular disease. Stroke 26:361–367. 46. WHO MONICA Project Principal Investigators (1988) The World Health Organization MONICA Project (monitoring trends and determinants in cardiovascular disease): a major international collaboration. J Clin Epidemiol 41:105–114. 47. WangYF, Tsirka SE, Strickland S, Stieg PE, Soriano SG and Lipton SA (1998) Tissue plasminogen activator (tPA) increases neuronal damage after focal cerebral ischemia in wild-type and tPA-deficient mice. Nat Med 4:228–231. 48. Barber PA, Demchuk AM, Hirt L and Buchan AM (2003) Biochemistry of ischemic stroke. Adv Neurol 92:151–164. 49. Dirnagl U, Iadecola C and Moskowitz MA (1999) Pathobiology of ischaemic stroke: an integrated view. Trends Neurosci 22:391–397. 50. (1995) The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group Tissue plasminogen activator for acute ischemic stroke. N Engl J Med 333:1581–1587. 51. Tabrizi P, Wang L, Seeds N, McComb JG, Yamada S, Griffin JH, Carmeliet P, Weiss MH and Zlokovic BV (1999) Tissue plasminogen activator (tPA) deficiency exacerbates cerebrovascular fibrin deposition and brain injury in a murine stroke model: Studies in tPA-deficient mice and wild-type mice on a matched genetic background. Arterioscler Thromb Vasc Biol 19:2801–2806. 52. Zivin JA, Fisher M, DeGirolami U, Hemenway CC and Stashak JA (1985) Tissue plasminogen activator reduces neurological damage after cerebral embolism. Science 230:1289–1292. 53. Kilic E, Kilic U, Bassetti CL and Hermann DM (2004) Intravenously administered recombinant tissue-plasminogen activator attenuates neuronal injury after mild focal cerebral ischemia in mice. Neuroreport 15:687–689. 54. Nagai N, De Mol M, Lijnen HR, Carmeliet P and Collen D (1999) Role of plasminogen system components in focal cerebral ischemic infarction: A gene targeting and gene transfer study in mice. Circulation 99:2440–2444. 55. Symon L (1980) The relationship between CBF, evoked potentials and the clinical features in cerebral ischaemia. Acta Neurol Scand Suppl 78:175–190. 56. Hakim AM (1987) The cerebral ischemic penumbra. Can J Neurol Sci 14: 557–559.

ch24

Jan. 10, 2007

8:57

SPI-B429

616



57. Hossmann KA (1994) Viability thresholds and the penumbra of focal ischemia. Ann Neurol 36:557–565. 58. Benveniste H, Drejer J, Schousboe A and Diemer NH (1984) Elevation of the extracellular concentrations of glutamate and aspartate in rat hippocampus during transient cerebral ischemia monitored by intracerebral microdialysis. J Neurochem 43:1369–1374. 59. Zhang Z, Chopp M and Powers C (1997) Temporal profile of microglial response following transient (2 h) middle cerebral artery occlusion. Brain Res 744:189–198. 60. Ayata C and Ropper AH (2002) Ischaemic brain oedema. J Clin Neurosci 9: 113–124. 61. Chen ZL and Strickland S (1997) Neuronal death in the hippocampus is promoted by plasmin-catalyzed degradation of laminin. Cell 91:917–925. 62. Indyk JA, Chen ZL, Tsirka SE and Strickland S (2003) Laminin chain expression suggests that laminin-10 is a major isoform in the mouse hippocampus and is degraded by the tissue plasminogen activator/plasmin protease cascade during excitotoxic injury. Neuroscience 116:359–371. 63. Aoki T, Sumii T, Mori T, Wang X and Lo EH (2002) Blood-brain barrier disruption and matrix metalloproteinase-9 expression during reperfusion injury: Mechanical versus embolic focal ischemia in spontaneously hypertensive rats. Stroke 33:2711–2717. 64. Wang X, Lee SR, Arai K, Lee SR, Tsuji K, Rebeck GW and Lo EH (2003) Lipoprotein receptor-mediated induction of matrix metalloproteinase by tissue plasminogen activator. Nat Med 10:1313–1317. 65. WuYP, Siao CJ, Lu W, Sung TC, Frohman MA, Milev P, Bugge TH, Degen JL, Levine JM, Margolis RU and Tsirka SE (2000) The tissue plasminogen activator (tPA)/plasmin extracellular proteolytic system regulates seizure-induced hippocampal mossy fiber outgrowth through a proteoglycan substrate. J Cell Biol %20 148:1295–1304. 66. Represa A, Niquet J, Pollard H and Ben-Ari Y (1995) Cell death, gliosis, and synaptic remodeling in the hippocampus of epileptic rats. J Neurobiol 26: 413–425. 67. Ben-Ari Y and Cossart R (2000) Kainate, a double agent that generates seizures: Two decades of progress. Trends Neurosci 23:580–587. 68. Kopito RR (2000) Aggresomes, inclusion bodies and protein aggregation. Trends Cell Biol 10:524–530. 69. Huntington JA, Read RJ and Carrell RW (2000) Structure of a serpin-protease complex shows inhibition by deformation. Nature 407:923–926.

ch24

Jan. 10, 2007

8:57

SPI-B429



617

70. Miranda E, Romisch K and Lomas DA (2004) Mutants of neuroserpin that cause dementia accumulate as polymers within the endoplasmic reticulum. J Biol Chem 279:28283–28291. 71. Belorgey D, Crowther DC, Mahadeva R and Lomas DA (2002) Mutant Neuroserpin (S49P) that causes familial encephalopathy with neuroserpin inclusion bodies is a poor proteinase inhibitor and readily forms polymers in vitro. J Biol Chem 277:17367–17373. 72. Carrell RW and Gooptu B (1998) Conformational changes and disease– serpins, prions and Alzheimer’s. Curr Opin Struct Biol 8:799–809. 73. Carrell RW and Lomas DA (1997) Conformational disease. Lancet 350: 134–138. 74. Ryu SE, Choi HJ, Kwon KS, Lee KN and Yu MH (1996) The native strains in the hydrophobic core and flexible reactive loop of a serine protease inhibitor: Crystal structure of an uncleaved alpha1-antitrypsin at 2.7 A. Structure 4: 1181–1192.

ch24

Jan. 10, 2007 8:57 SPI-B429



ch25

Jan. 10, 2007

8:57

SPI-B429


25 Serpins and Alzheimer’s Disease H. Tonie Wright and Sabina Janciauskiene

1. Introduction The first implication of a serpin in Alzheimer’s disease (AD) came from the immunochemical identification of α1 -antichymotrypsin (ACT) as a component of the cerebral amyloid plaques characteristic of the Alzheimer’s diseased brain.1 The most abundant protein in these Alzheimer’s plaques is a 40–42 amino acid peptide, Aβ, which continues to be exhaustively studied as a possible primary causative agent of AD. ACT is also an abundant component of Alzheimer’s plaques and has received considerable attention for possible involvement in the etiology of AD for two reasons. Firstly, ACT can interact with Aβ directly, and influences the in vitro polymerization of Aβ into the characteristic fibrils of amyloid plaques. Secondly, the lesions characteristic of Alzheimer’s diseased brains (amorphous or classical amyloid plaques) are frequently in proximity to activated astrocytes or microglia, which are markers for inflammation. As an acute phase reactant, the high abundance of ACT in amyloid plaques is most likely a consequence of this inflammatory state. This chapter focuses almost entirely on the possible roles of one serpin, ACT, in AD. Due to the multiple roles played by proteases, including serine proteases, in the processing of the Alzheimer’s Aβ peptide and

619

ch25

Jan. 10, 2007

8:57

SPI-B429

620


H. T. Wright and S. Janciauskiene

in inflammatory pathways, it seems quite likely that serpins other than ACT have a bearing on AD. However, there is little direct information implicating them.

2. Brain Distribution and Forms of ACT Astrocytes are primarily responsible for the expression of ACT in the brain,2,3 and ACT appears to be overexpressed in the Alzheimer’s brain, consistent with a state of inflammation and an acute phase response in AD. It was noted in an early study of ACT in amyloid plaques1 that anti-ACT antibody bound more frequently to brain amyloid plaque than to blood vessel amyloid deposits, relative to staining with Congo red, which specifically binds to amyloid protein with β-sheet structure. This suggested a possible specificity of ACT for the fibrillar form of Aβ, but it was subsequently shown that anti-ACT immunoreactivity also occurs in diffuse senile plaques of the Alzheimer’s diseased brain,4,5 indicating that non-fibrillar Aβ may also be associated with ACT. Furthermore, anti-ACT only binds to amyloid plaque that contains Aβ,2,6 and not to other amyloids with different protein constituents lacking Aβ, suggesting specificity in the association of Aβ and ACT in AD plaque. Another study found anti-ACT staining regions contiguous with fibrillar Aβ,7 and suggested that ACT may be inhibiting the normal degradation of the amyloid precursor protein (APP) from which Aβ is formed by proteolytic processing, thereby promoting fibril formation. Using monoclonal antibodies for the native, inactivated and protease-complexed forms ofACT, Rozemuller et al.8 found thatACT occurred in both amorphous and classical plaques, and that it was predominantly in the inactivated (cleaved) form or in complex with protease. In light of the subsequent studies of the interaction of Aβ with ACT, it would be useful to know whether any of the monoclonal antibodies used in this study had specificity or cross-reactivity for ACT/Aβ complex. Subsequently, measurements of active ACT in Alzheimer’s diseased brains showed a correlation with apoE genotype. ACT levels were higher in the higher risk apoE4 genotype, but appeared to have a lower specific activity.9 This was interpreted to mean that a larger fraction of ACT was inactive in the apoE4 carriers, though this could be due to ACT

ch25

Jan. 10, 2007

8:57

SPI-B429


Serpins and Alzheimer’s Disease

621

polymer formation, as well as to Aβ complex formation with ACT or to the inactivation due to adventitious protease cleavage of the serpin.

3. Effects of ACT on Aβ Fibrillization Aβ exists in multiple forms in vivo: soluble monomeric and oligomeric forms, disordered aggregates and ordered fibrils. Much attention over the past two decades has been on the fibrillar form of Aβ, and it is only recently that researchers have turned to the examination of the biological activities of the smaller soluble forms.10,11 It was initially observed that ACT could form an SDS-stable complex with peptide fragments of Aβ corresponding to the amino terminal 1–12 and 1–28 sequence, and that ACT lost its inhibitor activity toward chymotrypsin as a result of preincubation with these Aβ fragments.12,13 Subsequently, a panel of Aβ peptides of different lengths (1–40, 1–28, 11–28, and a double mutant 1–40) was used to assess the effects of ACT on fibrillar Aβ.14 It was found that ACT associated with preformed fibrils in solution, except those lacking the amino terminal 10 amino acids. Furthermore, association of ACT with the fibrils caused fibril dissolution over a period of minutes in the physiological buffer. These results suggested that ACT may function to destabilize fibrillar forms of Aβ and thereby enrich the pool of smaller forms. They also pointed to a specificity of ACT for the amino terminus of Aβ in associating with the fibrillar Aβ. Other studies assessed the effects of ACT on freshly dissolved Aβ rather than on preformed fibrils.15,16 In the first of these, the presence of ACT at a ratio of 1:200 to Aβ1–42 , promoted polymerization of Aβ1–42 into fibrils and it was suggested that this chaperone function could play a role in plaque deposition. In contrast, at a 1:10 ratio of ACT:Aβ1–40 , the latter study showed that ACT inhibited formation of fibrils and promoted disaggregation of preformed Aβ fibrils, and at a 1:15 ratio of ACT:Aβ1–40 , another study also found inhibition of fibril formation but with no effect on the cytotoxic properties of Aβ.17 It was further found that this interaction of ACT with Aβ resulted in the stabilization of ACT to denaturation, indicative of strand insertion in β-sheet A of the serpin. At lower concentrations of ACT and Aβ, but at comparable molar ratios, ACT was found to have little or no effect on the aggregation of Aβ.18 Further evidence for a direct

ch25

Jan. 10, 2007

8:57

SPI-B429

622



association of Aβ with ACT came from the observation that preincubation of ACT with Aβ2–9 abrogated the enhanced cytotoxicity toward primary cortical neurons of Aβ1–42 that was incubated with ACT.19 The molecular basis for the interaction between ACT and Aβ was first inferred by Lukacs and Christiansen20 who noted that residues 3–9 of Aβ could be incorporated as a substitute for β-strand 1 of β-sheet C in ACT. This proposal was consistent with the observation that short Aβ fragments corresponding to the amino terminal region affected the profibrillar activity of ACT. Other investigators noted that it was possible to build a stereochemically plausible model of ACT with the carboxyl terminus of Aβ inserted into β-sheet A of ACT.21 This model was consistent with the observation that ACT/Aβ complex showed increased stability to denaturation, characteristic of strand inserted forms of serpins. These two models were not mutually exclusive and a combined model in which Aβ inserted bimodally through both its amino and carboxyl sequences was proposed.22 This work also showed how the fibrillization of Aβ1–42 depends on the ratio of ACT:Aβ, lower fractions of ACT promoting fibrillization, while higher fractions of ACT had no effect. It was also found that on extended incubation of Aβ1–42 with ACT, a higher molecular weight form of the complex formed had the size of a 2:1 complex of ACT:Aβ1–42 . The possibility of such a complex is implicit in the bimodal model of Aβ binding to ACT, since pairs of ACT molecules could be linked through asymmetric binding of the two termini of Aβ to the two different β-sheets. The effects of ACT on Aβ are dependent on both the concentrations of the reactants and their molar ratios, and also on the form of Aβ being used. In some regimes, ACT promotes the dissolution of fibrils, while at other concentration ratios it appears to promote fibril formation, acting as a pathogenic chaperone. This latter function may occur through imposition of a β-strand conformation on Aβ as a consequence of the insertion into sheet A of ACT, thereby priming Aβ for polymerization into fibrils. It is also possible that ACT binds a specific conformer in the population of ACT molecules and that this depletion enhances the kinetics of fibril formation. Linking these in vitro observations to the in vivo setting is difficult, since the relative amounts of ACT and Aβ and their distribution in the in vivo microenvironment cannot be reliably assessed.

ch25

Jan. 10, 2007

8:57

SPI-B429



623

4. Transgenic Models for the in vivo Role of α1 -Antichymotrypsin in Alzheimer’s Disease The specificity of the interaction of ACT with Aβ in vitro prompted studies in transgenic mice to establish whether there is a link between ACT and the histopathology, and the cognitive deficits in this animal model that mimic those in humans. A knockout mouse model for ACT-related phenomena is problematic because there is no mouse analogue for ACT. Consequently, doubly transgenic mice expressing human ACT (hACT) and human amyloid precursor protein (hAPP) were created to study how these genes and their products interact to affect amyloid deposition and cognitive behavior in this animal model. Doubly transgenic mice expressing hACT in astrocytes, and a mutant APP (coding for a familial, three site AD variant that is amyloidogenic) in neurons, developed amyloid plaques in the hippocampus and neocortex to a greater extent than hACT or hAPP singly transgenic animals.23 The hAPP singly transgenic and the hACT/hAPP doubly transgenic mice, but not the singly transgenic hACT mice, developed astrocytosis, suggesting a link between hAPP expression and this inflammatory response. The hACT transgene alone had no effect on overall Aβ levels prior to plaque formation, nor did it correlate with the loss of SYN-IR presynaptic terminals, a hallmark of cognitive decline in AD. The acceleration of amyloid plaque accumulation in the hACT/hAPP double transgenic parallels the promotion by ACT of Aβ fibrillization in vitro in certain ratio ranges of the two components. The absence of a correlation between the presence of the hACT gene alone or between amyloid burden and the loss of presynaptic terminals characteristic of the AD brain, indicates that these putative AD risk factors are not linked to presynaptic terminal deficits, at least in this animal model. A similar mouse model transgenic for astrocyte expression of hACT and for hAPP with a single amyloidogenic mutation (one of the three used in the above work) gave somewhat different results from those described earlier.24 In the latter, there was increased Aβ expression in the doubly transgenic mouse, even at an early age, particularly in the hippocampus. The increased amyloid deposition paralleled increased expression of ACT, suggesting a possible synergy between these two, either in biosynthesis or in Aβ degradation. As in the other study, expression of astrocytic ACT

ch25

Jan. 10, 2007

8:57

SPI-B429

624



parallels astrocytosis and it is suggested that the expression of ACT is a consequence and not a cause of the latter. Comparison of these two studies is somewhat confounded by differences in the experimental models (e.g. different APP variant). A subsequent study of this mutant APP transgenic mouse model crossed with both apoE and ACT transgenics sought to measure the effects of the various combinations of expressed genes on cognition.25 This study showed that amyloid deposition tracks the expression of ACT in parallel with cognitive decline, consistent at least with a role for ACT in the events that lead to cognitive deficits.

5. Alzheimer’s Disease and the Genetics of ACT The identification of a small number of genes with links to AD naturally led to the suspicion that there may also be ACT mutants that confer higher susceptibility to AD. A point mutation in the signal sequence of the ACT gene was reported to confer an eight-fold increased risk of developing AD in carriers who also carry the apoE4 allele, an established risk factor for AD.26 This result has received support from some studies, but not from others and appears to be a weak risk factor. The occurrence of the mutation in the signal peptide of the translated ACT implies that the effects of ACT are dosage-dependent and not functional, since the signal peptide is removed in the mature ACT. Higher expression of the Thr → Ala mutant might be expected as a result of the slightly enhanced hydrophobicity of the mutant signal peptide in replacing Thr with Ala. A study of this T/A mutant ACT expression from transfected monkey kidney COS-7 and astrocytoma C6 cell lines showed an increased degree of glycosylation of the ACT from the mutant gene relative to the wild-type control, and this was attributed to a more efficient expression of the ACT through the secretory pathways of the endoplasmic reticulum where glycosylation occurs.27 The inflammatory cytokines interleukin-1 (IL-1) and -6 (IL-6), which regulate expression of ACT, have been reported at higher levels in the AD brain,28–30 and specifically in regions of amyloid deposition. Polymorphisms in the IL-1 and IL-6 genes were linked to increased risk of AD,31–35 and their effects may be propagated through the upregulation of ACT expression.

ch25

Jan. 10, 2007

8:57

SPI-B429



625

6. ACT as a Diagnostic Marker for Alzheimer’s Disease The possible involvement of ACT in the etiology of AD led to studies that determined whether its abundance in serum or cerebrospinal fluid could be used as a marker for AD. The long, asymptomatic progression time for AD dictates that any treatments for the disease will have to be prospective, and therefore are likely to depend on diagnosis prior to overt symptoms. The genetic risk factors for AD are weak and primarily predictive of the rarer, early onset forms of the disease, and diagnostic markers for AD are almost non-existent. The association of inflammation with AD and the implication of ACT in its pathogenesis at the molecular and cellular levels led to a large number of studies to determine whether ACT could be exploited for this purpose. One of the more thorough and well-controlled studies36 found statistically significant elevated plasma and cerebrospinal fluid levels of ACT in AD patients, and these levels appeared to correlate with the progression of AD dementia. Another population-based prospective cohort study in the Netherlands37 sought to determine whether high levels of inflammatory proteins in plasma samples are associated with an increased risk of dementia. ACT, among several other inflammatory proteins, was found to correlate with dementia, including AD. As the correlation is not strong and because ACT is an acute phase reactant associated with inflammation, which can have many causes, a high frequency of false positives would be anticipated from its use, and it is unlikely to be useful as a prospective marker specific for AD.

7. Serpins and Proteolytic Degradation of Aβ Studies of Aβ metabolism have been overwhelmingly focused on its biosynthesis and effects on cells in culture. Surprisingly, minimal work has been directed at the catabolism of the different forms of Aβ, including those in which Aβ is bound to other molecules such as ACT. The resistance of Aβ fibrils and Aβ amyloid plaque material to proteases has led to a tacit acceptance of chemical inertness in insoluble Aβ. This has further reinforced the accepted notion that amyloid plaque is the underlying cause of

ch25

Jan. 10, 2007

8:57

SPI-B429

626



AD pathology and reduced it to a “plumbing problem.” Aβ in some solute form(s) is susceptible to degradation by insulin degrading enzyme,38 gelatinase A,39 neprilysin,40 cathepsin D41 and a metalloproteinase,42 but it is unknown which, if any, of these are directly susceptible to serpin inhibition. Since proteolytic degradation of Aβ is a likely pathway in its turnover, a role for serpins in regulating that process is a strong possibility. One documented case for the involvement of ACT playing its role of serine proteinase inhibitor in the degradation of Aβ was found in a neuroblastoma cell model.43 A metalloendopeptidase (MP24.15) was identified that promotes degradation of soluble Aβ, but was itself not the catalyst for this reaction. Instead, it apparently activated a serine proteinase, which was present in the conditioned medium from cultured cells and appeared to be the catalyst for Aβ degradation. This activation could occur through a zymogen precursor or by the degradation of a peptide or protein inhibitor that specifically blocks the serine proteinase. The serine protease activity is inhibited by ACT, though it cannot be excluded that the effect of ACT is a consequence of the sequestration of the Aβ, which protects it from proteolysis. It may be significant that MP24.15 declines with age and at an accelerated rate in AD, and that its levels are most decreased in the brain regions where Aβ accumulation is the greatest. A possible implication of the plasminogen-plasmin pathway in Aβ degradation was drawn from observations that fibrillar Aβ and its analogues stimulate tissue plasminogen activator (tPA) to activate plasminogen to plasmin.44 This experiment was prompted by the observation of antibody cross-reactivity between anti-Aβ and anti-fibrinogen and the known activating effects of fibrin on tPA. Subsequently, it was shown that aggregated Aβ amyloid (but not soluble) induces expression of both tPA and the urokinase plasminogen activator,45,46 and that active plasmin degrades aggregated Aβ at a physiologically significant rate. These observations imply that elevated levels of plasminogen activator inhibitors (e.g. PAI-1), and possibly also of the plasmin inhibitor α2 -antiplasmin, will inhibit the degradation of Aβ. Significantly, it was shown in a mouse model in which Aβ is overexpressed, that PAI-1 is also overexpressed and the tPA-plasmin pathway is downregulated.47 Overexpression of PAI-1 and α2 -antiplasmin occur during inflammation and have the effect of suppressing plasmin-mediated degradation of Aβ.

ch25

Jan. 10, 2007

8:57

SPI-B429



627

8. Activity Gain and Loss from the ACT – Aβ Interaction Most research on the possible roles of ACT in AD has addressed the effects of ACT on Aβ, on the implicit assumption that Aβ is the active neurotoxic agent. Less attention has been paid to the loss of ACT function as a result of complex formation with Aβ or of possible new functions of the ACT/Aβ complex in the disease. ACT is an acute phase reactant that regulates proteases expressed in the primary immune response, thereby serving to control inflammation and tissue damage collateral to its immune function. In addition, one or more forms of ACT quench superoxide radicals48 by inhibiting a membrane bound NADH oxidase.49 ACT also binds polyanions, which may serve to regulate the inflammatory response by scavenging proteoglycans that are upregulated in inflammation and are components of amyloid plaque.50 It has been proposed that loss of ACT proteinase inhibitor activity could contribute to sustaining the state of chronic inflammation that is observed in most AD brains (ibid). It is not known whether the capacity of ACT to quench free radical activity is also affected by complex formation with Aβ, but loss of this activity would further diminish the anti-inflammatory functions ofACT. Oxidative stress has been widely investigated as a concomitant and possible proximal cause of the neuronal death observed in the AD brain.51,52 Loss of protection against oxidative damage, such as that provided by ACT, could be a factor in the complex of inflammatory reactions that is implicated in AD and could contribute to sustaining chronic inflammation. The breakdown of ACT functions that are important for suppressing inflammation could contribute to the establishment of a chronic inflammatory state that is responsible for neuronal damage over a long period of time. In addition to the loss of ACT function(s), the ACT/Aβ complex itself might have properties contributing to the chronic pro-inflammatory state that leads to neurodegeneration. The superoxide scavenging effects of ACT (see above) were in fact shown to be due to ACT inhibitor complex with chymotrypsin,53 and presumably other target proteases. Since Aβ ablates the inhibitor activity of ACT on forming complex with it, we can likewise infer that this superoxide scavenging activity of ACT, which requires complex formation with target proteinase, is compromised in the ACT/Aβ complex.

ch25

Jan. 10, 2007

8:57

SPI-B429

628

Fig. 1. in red.



Model of ACT/Aβ complex. ACT in blue, Aβ in yellow, reactive center loop (RCL)

The gain of function in the ACT/Aβ complex has also been observed in neuroblastoma and glioma cell lines. A solution containing ACT/Aβ1–42 complex was shown to perturb lipid metabolism and upregulate the transcription factors PPARγ and NFκB in a human neuroblastoma cell line.54 In addition to the proinflammatory effects of the upregulation of transcription factors, there is considerable evidence linking lipid metabolism to Aβ biogenesis.55–59 Effects of ACT/Aβ1–42 complex on lipid metabolism

ch25

Jan. 10, 2007

8:57

SPI-B429



629

could be another link in the chain of factors that propagates chronic inflammation. Another study of the effects of ACT/Aβ complex on glioma cells also found increased expression of the inflammatory cytokine TNF-α and increased uptake of LDL, elevated levels of mRNA for LDL receptor and HMG-CoA reductase, and of cellular lipid, consistent with a pro-inflammatory phenotype.60 It is likely that these effects are receptor mediated and they point to additional levels of feedback complexity in their interpretation, wherein a new molecular form such as ACT/Aβ1–42 could alter the expression of either of its constituents.

Conclusion A number of observations ranging from histochemical to molecular structure suggest that ACT may play a role in AD pathology: ACT is a major constituent of Alzheimer’s amyloid plaques; it is an acute phase reactant, which interacts directly with Aβ, resulting in altered properties and activities of both Aβ and ACT; a point mutation near the 5 terminus of the ACT gene correlates with increased risk of AD; elevated levels of ACT correlate with AD disease state; ACT can inhibit degradation of Aβ; and the ACT/Aβ complex induces proinflammatory responses in cell culture. One might plausibly argue that the links between AD and ACT are at least as strong as those for Aβ, whose function is not yet fully understood. It seems unlikely that a single molecule and its functions will provide a full explanation for the etiology of AD. Nevertheless, the molecular properties of ACT as a serpin, which determine its interactions with other molecules, and its cellular and systemic functions as an anti-inflammatory agent, offer intriguing threads to pursue in trying to understand this disease.

References 1. Abraham CR, Selkoe DJ and Potter H (1988) Immunochemical identification of the serine protease inhibitor α1 -antichymotrypsin in the brain amyloid deposits of Alzheimer’s disease. Cell 52:487–501. 2. Pasternack JM, Abraham CR, Van Dyke B, Potter H and Younkin SG (1989) Astrocytes in Alzheimer’s disease gray matter express Aβ1–40 mRNA. Am J Pathol 92:147–156.

ch25

Jan. 10, 2007

8:57

SPI-B429

630



3. Koo EH, Abraham CR, Potter H, Cork LC and Price DL (1991) Developmental expression of α1 -antichymotrypsin in brain may be related to astrogliosis. Neurobiol Aging 12:495–501. 4. Picken MM, Larrondo-Lillo M, Coria F, Gallo GR, Shelanski ML and Frangione B (1990) Distribution of the protease inhiitor α1 -antichymotrypsin in cerebral and systemic amyloid. J Neuropathol Exp Neurol 49:41–48. 5. Shoji M, Hirai S, Yamaguchi H, Harigaya Y, Ishiguro K and Matsubara E (1991) Alpha 1-Antichymotrypsin is present in diffuse senile plaques. Am J Path 138:247–257. 6. Abraham CR, Shirahama T and Potter H (1990) α1 -Antichymotrypsin is associated solely with amyloid deposits containing the β-protein. Amyloid and cell localization of α1 -antichymotrypsin. Neurobiol Aging 11:123–129. 7. Miyakawa T, Katsuragi S, Yamashita K, Araki K, Hashimura T, Kimura T and Ohuchi K (1992) The distribution of α1 -antichymotrypsin and amyloid production in the brain in Alzheimer’s disease. Virchows Archiv B Cell Pathol 61:331–335. 8. Rozemuller JM, Abbink JJ, Kamp AM, Stam FC, Hack CE and Eikelenboom P (1991) Distribution pattern and functional state of α1 -antichymotrypsin in plaques and vascular amyloid in Alzheimer’s disease. Acta Neuropathol 82:200–207. 9. Licastro F, Davis LJ, Pedrini S, Galasko D and Masliah E (1998) Prostaglandin E2 induced polymerization of human α1 -antichymotrypsin and suppressed its protease inhibitory activity: Implications for Alzheimer’s disease. Biochem Biophys Res Comm 249:182–186. 10. Wisniewski T, Ghiso J and Frangione B (1994) Alzheimer’s disease and solutble Aβ. Neurobiol of Aging 15:143–152. 11. Klein WL, Krafft GA and Finch CE (2001) Targeting small Aβ oligomers: The solution to an Alzheimer’s disease conundrum? Trends Neurosci 24:219–223. 12. Potter H, Abraham C and Dressler D (1991) The Alzheimer amyloid components α1 -antichymotrypsin and β-protein form a stable complex in vitro, in Iqbal K, McLachlan DRC, Winblad B Wisniewski HM (eds.), Alzheimer’s Disease: Basic Mechanisms, Diagnosis and Therapeutic Strategies, John Wiley, New York. 13. Potter H, Nelson RB, Das S, Siman R, Kayyali US and Dressler D (1992) The involvement of proteases, protease inhibitors, and an actue phase response in Alzheimer’s Disease. Ann NY Acad Sci 674:161–173. 14. Fraser PE, Nguyen JT, McLachlan DR, Abraham C and Kirschner DA (1993) α1 -Antichymotrypsin binding to Alzheimer Aβ peptides is sequence specific and induces fibril disaggregation in vitro. J Neurochem 61:298–305.

ch25

Jan. 10, 2007

8:57

SPI-B429



631

15. Ma J, Yee A, Brewer Jr H, Das S and Potter H (1994) Amyloid-associated proteins α1 -antichymotrypsin and apolipoprotein E promote assembly of Alzheimer β-protein into filaments. Nature 372:92–94. 16. Erikkson S, Janciauskiene S and Lannfelt L (1995) α1 -Antichymotrypsin regulates Alzheimer β-amyloid peptide fibril formation. Proc Nat Acad Sci USA 92:2313–2317. 17. Aksenova MV, Aksenov MY, Butterfield DA and Carney JM (1996) Alpha-1-antichymotrypsin interaction with A beta (1-40) inhibits fibril formation but does not affect the peptide toxicity. Neurosci Lett 211:45–48. 18. Webster S and Rogers, J (1996) Relative efficacies of amyloid β peptide (Aβ) binding proteins in Aβ aggregation. J Neurosci Res 46:58–66. 19. Ma J, Brewer Jr B and Potter H (1996) Alzheimer Aβ neurotoxicity: Promotion by antichymotrypsin, ApoE4; inhibition by Aβ-related peptides. Neurobiol Aging 17:773–780. 20. Lukacs C and Christianson DW (1996) Is the binding of β-amyloid protein to antichymotrypsin in Alzheimer plaques mediated by a β-strand insertion? Proteins Struct Funct Genet 25:420–424. 21. Janciauskiene S, Rubin H, Lukacs, CM and Wright HT (1998) Alzheimer’s peptide Aβ1–42 binds to two β-sheets of α1 -antichymotrypsin and transforms it from inhibitor to substrate. J Biol Chem 273:28360–28364. 22. Janciauskiene S, Erikkson S and Wright HT (1996) A specific structural interaction of Alzheimer’s peptide Aβ1–42 with α1 -antichymotrypsin. Nat Struct Biol 3:668–671. 23. Mucke L, Yu G-Q, McConlogue L, Rockenstein EM, Abraham CR and Masliah E (2000) Astroglial expression of human α1 -antichymotrypsin enhances Alzheimer-like pathology in amyloid protein precursor transgenic mice. Am J Path 157:2003–2010. 24. Nilsson LNG, Bales KR, DiCarlo G, Gordon MN, Morgan D, Paul SM and Potter H (2001) α1 -antichymotrypsin promotes β-sheet amyloid plaque deposition in a transgenic mouse model of Alzheimer’s disease. J Neurosci 21: 1444–1451. 25. Nilsson LNG, Arendash GW, Leighty RE, Costa DA, Low MA, Garcia MF, Cracciolo JR, Rojiani A, Wu X, Bales KR, Paul SM and Potter H (2004) Cognitive impairment in PDAPP mice depends on apoE and ACT-catalyzed amyloid formation. Neurobiol Aging 12:1153–1167. 26. Kamboh MI, Sanghera DK, Ferrell RE and DeKosky ST (1995) ApoE4associated Alzheimer’s disease risk is modified by α1 -antichymotrypsin polymorphism. Nat Genet 10:486–488.

ch25

Jan. 10, 2007

8:57

SPI-B429

632



27. Nilsson LNG, Das S and Potter H (2001) Effect of cytokines, dexamethazone and the A/T-signal peptide polymorphism on the expression of α1 antichymotrypsin in astrocytes: Significance for Alzheimer’s disease. Neurochem Int 39:361–370. 28. Griffin WST, Stanley LC, Ling C, White L, MacLeod V, Perrot LJ, White CL and Araoz C (1989) Brain interleukin-1 and S-100 immunoreactivity are elevated in Down syndrome and Alzheimer disease. Proc Nat Acad Sci USA 86:7611–7615. 29. Das S and Potter H (1995) Expression of the Alzheimer amyloid-promoting factor antichymotrypsin is induced in human astrocytes by IL-1. Neuron 14:447–456. 30. Sheng JG, Griffin WS, Royston MC and Mrak RE (1995) Distribution of interleukin-1-immunoreactive microglia in cerebral cortical layers: Implications for neuritic plaque formation in Alzheimer’s disease. Neuropathol Appl Neurobiol 21:290–301. 31. Du Y, Dodel RC, Eastwood BJ, Bales KR, Gao F, Lohmuller F, Muller U, Kurz A, Zimmer R, Evans RM, Hake A, Gasser T, Oertel WH, Griffin WST, Paul SM and Farlow MR (2000) Association of an interleukin 1alpha polymorphism with Alzheimer’s disease. Neurology 55:480–484. 32. Grimaldi LME, Casadei VM, Ferri C, Veglia F, Licastro F, Annoni G, Biunno I, De Bellis G, Sorbi S, Mariani C, Canal N, Griffin WST and Franceschi M (2000) Association of early-onset Alzheimer’s disease with an interleukin-1α gene polymorphism. Ann Neurol 47:361–365. 33. Nicoll JA, Mrak RE, Graham DI, Stewart J, Wilcock G, MacGowan S, Esiri MM, Murray LS, Dewar D, Love S, Moss T and Griffin WS (2000) Association of interleukin-1 gene polymorphisms with Alzheimer’s disease. Ann Neurol 47:365–368. 34. Papassotiropoulos A, Bagli M, Jessen F, Bayer TA, Maier W, Rao ML and Heun R (1999) A genetic variation of the inflammatory cytokine interleukin-6 delays the initial onset and reduces the risk for sporadic Alzheimer’s disease. Ann Neurol 45:666–668. 35. Licastro F, Pedrini S, Ferri C, Casadei V, Govoni M, Pession A, Sciacca FL, VEglia F, Anoni G, Bonafe M, Olivieri F, Franceschi C and Grimaldi LME (2000) Gene polymorphism affecting α1 -antichymotrypsin and interleukin-1 plasma levels increases Alzheimer’s disease risk. Ann Neurol 48:388–391. 36. DeKosky ST, Ikonomovic MD, Wang X, Farlow M, Wisniewski S, Lopez OL, Becker JT, Saxton J, Klunk WE, Sweet R, Kaufer DI and Kamboh MI (2003) Plasma and cerebrospinal fluid α1 -antichymotrypsin levels in Alzheimer’s disease: Correlation with cognitive impairment. Ann Neurol 53:81–90.

ch25

Jan. 10, 2007

8:57

SPI-B429



633

37. Engelhart MJ, Geerlings MI, Meijer J, Kiljaan A, Ruitenberg A, van Swieten JC, Stignen T, Hofman A, Witteman JC and Breteler MM (2004) Inflammatory proteins in plasma and the risk of dementia: The Rotterdam study. Arch Neurol 61:668–672. 38. Kurochkin IV and Goto S (1994) Alzheimer’s β-amyloid peptide specifically interacts with and is degraded by insulin-degrading enzyme. FEBS Lett 345:33–37. 39. Roher AE, Kasunic TC, Woods AS, Cotter RJ, Ball MJ and Fridman R (1994) Proteolysis of Aβ peptide from Alzheimer disease brain by gelatinase A. Biochem Biophys Res Comm 205:1755–1761. 40. Howell S, Nalbantoglu J and Crine P (1995) Neutral endopeptidase can hydrolyze β-amyloid(1-40) but shows no effect on β-amyloid precursor protein metabolism. Peptides 16:647–652. 41. Hamazaki H (1996) Cathepsin D is involved in the clearance of Alzehimer’s β-amyloid protein. FEBS Lett 396:139–142. 42. Mentlein R, Ludwig R and Martensen I (1998) Proteolytic degradation of Alzheimer’s disease amyloid β-peptide by a metalloproteinase from microglia cells. J Neurochem 70:721–726. 43. Yamin R, Malgeri EG, Sloane JA, McGraw WT and Abraham CR (1999) Metalloendopeptidase EC 3.4.24.15 is necessary for Alzheimer’s amyloid-β peptide degradation. J Biol Chem 274:18777–18784. 44. Kingston IB, Castro MJM and Anderson S (1995) In vitro stimulation of tissuetype plasminogen activator by Alzheimer amyloid β-peptide analogues. Nat Med 1:138–142. 45. Tucker M, Kihiko M,Caldwell JN, Wright S, Kawarabayashi T, Price D, Walker D, Scheff S, McGillis JP, Rydel RE and Estus S (2000) The plasmin system is induced by and degrades amyloid-beta aggregates. J Neurosci 20:3937–3946. 46. Wnendt S, Wetzels I and Gunzler WA (1997) Amyloid beta peptides stimulate tissue-type plasminogen activator but not recombinant prourokinase. Thromb Res 85:217–224. 47. Melchor JP, Pawlak R and Strickland S (2003) The tissue plasminogen activator-plasminogen proteolytic cascade accelerates amyloid-beta (Abeta) degradation and inhibits Abeta-induced neurodegeneration. J Neurosci 23:8867–8871. 48. Kilpatrick L, Johnson JL, Nickbarg EB, Wang Z, Clifford TF, Banach M, Cooperman BS, Douglas SD and Rubin H (1991) Inhibition of human neutrophil superoxide generation by α1 -antichymotrypsin. J Immunol 146:2388–2393.

ch25

Jan. 10, 2007

8:57

SPI-B429

634



49. Kilpatrick L, McCawley L, Nachiappan V, Greer W, Majumdar S, Korchak HM and Doublas SD (1992) α1 -Antichymotrypsin inhibits the NADPH oxidase-enzyme complex in phorbol ester-stimulated neutrophil membranes. J Immunol 149:3059–3065. 50. Janciauskiene S, Sun Y-X and Wright HT (2002) Interactions of Aβ with endogenous anti-inflammatory agents: A basis for chronic neurooinflammation in Alzheimer’s disease. Neurobiol Disease 10:187–200. 51. Markesbery WR (1997) Oxidative stress hypothesis in Alzheimer’s disease. Free Rad Biol Med 23:134–147. 52. Behl C (1999) Alzheimer’s disease and oxidative stress: Implications for novel therapeutic approaches. Prog Neurobiol 57:301–323. 53. Schuster MG, Enriquez PM, Curran P, Cooperman BS and Rubin H (1992) Regulation of neutrophil superoxide by antichymotrypsin-chymotrypsin complexes. J Biol Chem 267:5056–5059. 54. Sun YX, Wright HT and Janciauskiene S (2002) Alpha 1-antichymotrypsin/ Alzheimer’s peptide Abeta (1-42) complex perturbs lipid metabolism and activates transcription factors PPAR gamma and NFkappaB in human neuroblastoma (Kelly) cells. J Neurosci Res 67:511–522. 55. Puglielli L, Konopka G,Pack-Chung E, Ingano LA, Berezovska O, Hyman BT, Chang TY, Tanzi RE and Kovacs DM (2001) Acyl-coenzyme A: Cholesterol acyltransferase modulates the generation of the amyloid beta-peptide. Nat Cell Biol 3:905–912. 56. Frears E R, Stephens DJ, Walters CE, Davies H and Austen BM (1999) The role of cholesterol in the biosynthesis of beta-amyloid. Neuroreporth 10: 1699–1705. 57. Bodovitz S and Klein WL (1996) Cholesterol modulates alpha-secretase cleavage of amyloid precursor protein. J Biol Chem 271:4436–4440. 58. Galbete JL, Martin TR, Peressini E, Modena P, Bianchi R and Forloni G (2000) Cholesterol decreases secretion of the secreted form of amyloid precursor protein by interfering with glycosylation in the protein secretory pathway. Biochem J 348:307–313. 59. Simons M, Keller P, De Strooper B, Beyreuther K, Dotti CG and Simons K (1998) Cholesterol depletion inhibits the generation of beta-amyloid in hippocampal neurons. Proc Natl Acad Sci USA 95:6460–6464. 60. Sun YX, Wright HT and Janciauskiene S (2002) Glioma cell activation by Alzheimer’s peptide Aβ1–42 , α1 -antichymotrypsin, and their mixture. CMLS Cell Mol Life Sci 59:1734–1743.

ch25

Jan. 10, 2007

8:57

SPI-B429


Index α1 -antitrypsin, 412, 445, 446 α1 -antitrypsin deficiency, 483 α1 -antichymotrypsin, 74, 619 α1 -AT Saar, 490 α1 -ATZ, 490 Aβ, 619 acute phase reactant, 233, 619 acyl intermediate, 68 AFXa, 91 alpha-2-antiplasmin, 13 alternative exons, 230 alternatively-spliced RCL exons, 215 Alzheimer’s disease, 402 angioedema, 555 angiogenesis, 337, 346, 365 anhydrothrombin, 69 Anopheles gambiae, 231, 233 antiangiogenic, 524 antigen, 556 antithrombin, 13, 70, 212, 511, 534 antitrypsin-like, 11 apoE4, 620 apoptosis, 301, 371 apoptotic protease activating factor, 302 Arabidopsis thaliana, 282 astrocytes, 619 AT-Hook, 246 atherosclerosis, 411 autophagy, 487

Bowman–Birk, 67 bradykinin, 562, 563 branched pathway mechanism, 68 breach, 143 breast, 349 breast carcinoma, 369 C1, 555, 557 C1-inhibitor, 14, 555 C1-inhibitor biosynthesis, 559 C1-inhibitor deficient mice, 563 C14 family, 302 C3b, 569 Caenorhabditis elegans, 195, 232 cancer, 363 canonical conformations, 72 caspase, 89, 155, 302, 365 caspase inhibitors, 304 caspase-recruitment domain, 302 cathepsin, 202, 243, 312 CD loop, 401 CED3, 302 cell death, 301 cerebral edema, 598 cerebral ischemia, 598 cerebrovascular disease, 598 cervical cancer, 368 chaperone, 486, 622 chimeric serpins, 215 Chlamydomonas reinhardtii, 282 chromatin, 243, 250 chymotrypsin, 74, 558 cirrhosis, 446, 447 Clade B, 201, 243, 308, 401, 435 Clade G, 555 Clade L, 200 cleaved serpin, 41

β-factor XIIa, 558 bacterial proteases, 425, 426 beer, 282 blood brain barrier, 603 blood coagulation, 509 Bombyx mori, 215, 229 bootstrap neighbor-joining consensus, 150 635

index

Jan. 10, 2007

8:57

SPI-B429


636 coagulation factors, 510 collagen, 558 Collins bodies, 450, 608 complement, 555 complex dissociation, 87 conformational diseases, 119 conformations of serpin, 16, 35, 143, 267 corticosteroid binding globulin, 90 CrmA, 304 cross-class inhibitors, 434 crystal structures, 35 cysteine proteinases, 88, 365 cytotoxic T lymphocytes, 314 δ-conformation, 41 D-helix, 47 deacylation, 69 death effecter domain, 302 delta, 35 dementia, 608 diagnostic markers, 625 Drosophila melanogaster, 230, 233 Drosophila Serpins, 27A, 207, 236 dyslipidemia, 411 elastase, 319 emphysema, 460, 483 endometrium, 265 endotoxin shock, 565 energetic coupling, 85 entactin, 568 epithelial lining, 265 ERK, 398, 399 estradiol, 266 evolutionary relationships, 2 evolutionary tree, 211 extracellular matrix, 568 extrinsic pathway, 304 factor V, 511 factor XIa, 511, 558 factor XIIa, 511, 558 Familial Encephalopathy with Neuroserpin Inclusion Bodies (FENIB), 120, 446, 608 FRET, 88

Index functional compensation, 115 furin, 78, 214 G-helix, 47 gene structure, 6, 11 genetic redundancy, 208 gingipain, 430 glandular kallikrein, 558 glia-derived nexin-1, 13 glomeruli, 383 glutamine-rich reactive centers, 286 granzyme B, 365 head and neck cancer, 368 heat shock protein 47, 14 helix F, 86 helix I, 262 hemolymph, 230, 234 hemostasis, 509 heparan sulfate, 511, 558 heparin binding site, 517 heparin co-factor II, 13, 69, 511, 528 heparin-binding domain, 218 heparin-binding serpins, 509 hepatitis C virus NSC protease, 558 hepatocellular carcinoma, 483 hereditary angioedema, 561 heterochromatin, 244 hinge region motif, 21 histone, 244 horseshoe crabs, 237 hyperglycemia, 411 hypertension, 411 ICE, 302 inflammation, 619 inhibitor of apoptosis proteins, 304 innate immune response, 212, 235 insect, 229 insulin resistance, 411 interferon regulatory factor 6, 400 intracellular serpins, 195 intrinsic pathway, 302 kallikrein, 511 Kazal, 67

index

Jan. 10, 2007

8:57

SPI-B429


Index kidney, 387 Kunitz, 67, 510 L-DNase II, 321 laminin, 558, 568 latent, 35, 79 latent serpin, 40 LDL receptor, 393 leptin, 411 leukocyte adhesion, 568 liver disease, 483 loop insertion, 80 low-density lipoprotein receptor-related protein (LRP), 1, 394, 559, 595 LR Serpins, 287 lysosomal death pathway, 311 M-loop, 243, 251 malignancy, 337, 363 Manduca sexta, 215, 229 Manduca sexta 1B, 69 Manduca serpin-3, 236 MASP1, 558 MASP2, 558 maspin, 337, 365 maspin (SERPINB5), 365 matrix-associated regions, 247 maximum likelihood, 7 maximum parsimony, 7 Megsin, 383 MENT, 243 mesangium, 383 metabolic syndrome, 411 metastasis, 337, 367 Michaelis-like complex, 68 microglia, 619 myxoma virus, 164 myxomatosis, 169 native state, 35, 42 natural killer cells, 314 necrotic, 212 necrotic N-terminal extension, 217 neighbor-joining, 7 neonatal Z-ATT liver disease, 220 nephropathy, 383

637 neuroserpin, 14, 449, 593 neurovascular unit, 602 NK cells, 314, 365 nomenclature of the serpin superfamily, 10 non-inhibitory serpins, 207 nucleocytoplasmic, 246 nucleosomes, 244 obesity, 411 oligomerization, 252 Oryza sativa, 282 ov-serpins, 12 oxidative stress, 627 P35, 89, 304 PAI-1, 13, 365, 412 PAI-2, 364, 365 P. gingivalis, 430 pancreas, 387 pancreatic cancer, 349 pancreatic elastase, 321 papain, 155 pathogen, 429, 437 PEBP, 118 PEDF, 365 pentasaccharide, 517 phylogenetic analysis, 6, 20, 143 phylogenetic tree, 8 Pi-Z, 120 pigment epithelium derived factor, 13 Pittsburgh-ATT, 69, 208 Plant Kingdom, 279 plant serpins, 279 plasma hyaluronan binding protein, 558 plasma kallikrein, 555 plasmin, 511, 558 plasminogen, 598 polymerization, 35, 387, 447, 485, 556 pregnancy, 265 progesterone, 266 prokaryotes, 131 prokaryotic serpins, 22, 132 prophenoloxidase, 234 prophenoloxidase activating proteases, 235

index

Jan. 10, 2007

8:57

SPI-B429


638 protease cascade, 234 protease subunit of NGF„ 558 protein C, 511 protein C inhibitor, 531 protein C pathway, 510 protein S, 511 protein Z, 280 proteinase 3, 319 proteinase distortion, 84 proteolytic inactivation, 426 PSA, kallikrein2, 321 reactive center/site loop, 21, 45, 131, 207, 425 recombinant C1-inhibitor, 559 redundancy, 104 renal disease, 383 retinoblastoma protein, 401 reversible inhibition, 89 root of the serpin superfamily, 22 S → R transition, 221 S195A trypsin, 69 S. pyogenes, 431 scaffold-associated regions, 247 SCCA1, 248, 365 SCCA2, 365 scuPA, 90 serine proteases, 619 SERP1, 164, 177, 181 SERP2, 164, 172 SERP3, 164 serpin branched pathway mechanism, 68 serpin classification, 2 serpin complexes, 393 serpin fold, 16, 23, 35 serpin gene clusters, 209 serpin polymers, 41 serpin-1B, 231 serpin-1K, 231 serpin-derived allergens, 282 SERPINA1, 320 SERPINA3, 320, 371 SERPINB1, 320, 371 SERPINB2, 310, 364 SERPINB3, 310, 365

Index SERPINB4, 310, 365 SERPINB5 (maspin), 369 SERPINB6, 320 SERPINB9, 201, 307, 365 SERPINB10, 310, 365 SERPINB13, 312, 313, 365, 371 SERPINE1, 365 SERPINF1, 365, 371 SERPING1, 555 SERPING2, 365 serpinopathies, 208, 386, 426, 445, 446 shutter, 143 sialyl-Lewisx antigen, 556, 567 signaling receptors, 393 Siiyama variant α1 -antitrypsin, 221 smallpox, 164 SPI-1, 164, 170 SPI-2/crmA, 164, 172 SPI-3, 164, 177 Spn27A, 214 squamous-cell carcinoma, 369 squamous-cell carcinoma antigen, 363 SRP-2, 202 Streptopain, 435 stressed conformation (S), 215 stroke, 598 structural divergence, 4 structural features, 151 structural phylogeny, 14 structural similarity, 14 structures, 15 subtilisin, 155, 434 suicide substrate, 557 suicide substrate inhibitors, 67 synaptic plasticity, 596, 604 t-PA, 558 target-specificity, 217 temperature-dependent polymerization, 221 tetrahedral intermediate, 79 therapeutic uses of C1-inhibitor, 569 Thermobifida fusca, 45, 132 thermodynamics, 83 thermophilic organisms, 154 thermopin, 132

index

Jan. 10, 2007

8:57

SPI-B429


Index thrombin, 511, 558 thrombomodulin, 512 thrombosis, 221 thyroxine binding globulin, 90 tissue transglutaminase, 558, 569 tissue-type plasminogen activator (tPA), 593 TNF-α, 365 TNF-related apoptosis inducing ligand, 305 transgenic mice, 384, 561, 623 transglutamination, 401 type IV collagen, 568

639 unfolded protein response, 388 uPA, 399 uterine serpins, 261 vaccinia virus, 164 variant, 69 variola virus, 164 vaspin, 415, 418 VEGF, 365 x-ray structures, 69 Z variant of α1 -antitrypsin, 208, 220, 447

index

Molecular and Cellular Aspects of the Serpinopathies and Disorders in Serpin Activity

Molecular and Cellular Aspects of the Serpinopathies and Disorders in Serpin Activity

Molecular and Cellular Signaling

Molecular and Cellular Biophysics

Molecular and Cellular Signaling

Cellular and Molecular Neurobiology,

Molecular and Cellular Biophysics

Cellular and Molecular Neurobiology

Molecular and Cellular Signaling

Alzheimer's Disease: Cellular and Molecular Aspects of Amyloid beta

Salmonella Infections: Clinical, Immunological and Molecular Aspects (Advances in Molecular and Cellular Microbiology)

Salmonella Infections: Clinical, Immunological and Molecular Aspects (Advances in Molecular and Cellular Microbiology)

Lysosomal Disorders of the Brain: Recent Advances in Molecular and Cellular Pathogenesis and Treatment

Actin-based Motility: Cellular, Molecular and Physical Aspects

Brucella: Molecular and Cellular Biology

Molecular and Cellular Iron Transport

Molecular and Cellular MR Imaging

Cellulose. Molecular and Cellular Biology

Basic Neurochemistry, Seventh Edition: Molecular, Cellular and Medical Aspects

Yersinia: Molecular and Cellular Biology

Molecular and Cellular Iron Transport

Atherosclerosis: Molecular and Cellular Mechanisms

Integrated molecular and cellular biophysics

Cellular and Molecular Methods in Neuroscience Research

Advances in Molecular and Cellular Endocrinology

Molecular and cellular biology research in psychiatry

Techniques in Molecular and cellular Biology

Cellular and Molecular Methods in Neuroscience Research

Cellular and Molecular Methods in Neuroscience Research

Advances in Molecular and Cellular Endocrinology

Molecular and Cellular Aspects of the Serpinopathies and Disorders in Serpin Activity